GSA Special Publication No.21: Gondwana to Greenhouse: Australian Environmental Geoscience, 2001 by GSAustralia

AUSTRALIAN ENVIRONMENTAL GEOSCIENCE

H Darwin

Brisbane

Perth Sydney Adelaide Legend Location of studies by chapter i '

Melbourne VICTORIA

| Cretaceous & Tertiary Sedimentary Basins

P | | | | | $ Basement & Pre-Cretaceous Sedimentary Cover

•

TASMANIA 1000 K m

[ | Hobart

Numbers indicate the general locations of chapter topics on a simple geological map adapted from the AGSO National Digital Data Sets, 1998.

BRISBANE

PERTH SYDNEY ADELAIDE

CANBERRA MELBOURNE

m HOBART Population distribution in Australia. One dot = 1000 people based on statistical area boundaries - (Aust. Bureau of Statistics).

S P 2.1

GO ND WAN A TO

GREENHOUSE

2001 GEOLOGICAL SOCIETY OF AUSTRALIA INCORPORATED This volume is published in the Special Publications Series of the Geological Society of Australia

REFERENCE GENERAL REFERENCE: GOSTIN V. A. (Editor) 2001. Gondwana to Greenhouse: Australian Environmental Geoscience. Geological Society of Australia Special Publication 21

T W O FORMS OF REFERENCE TO SPECIFIC PAPERS ARE POSSIBLE: WILLIAMS M. A. J. 2001. Chapter 1—Quaternary climate changes in Australia and their environmental effects. Geological Society of Australia Special Publication 21, 3-11. WILLIAMS M. A. J. 2001. Chapter 1—Quaternary climate changes in Australia and their environmental effects. In: Gostin V. A. ed. Gondwana to Greenhouse: Australian Environmental Geoscience, pp. 3-11. Geological Society of Australia Special Publication 21.

First published in 2001 by the Geological Society of Australia Incorporated Typeset by Chapter 8 Pty Ltd, Sydney Cover Artwork produced by Artology Pty Ltd Printed and bound by Southwood Press, Sydney This book is copyright. Apart from any fair dealing for the purpose of private study, research, criticism or review, as permitted under the Copyright Act, no part may be reproduced by any process without the prior permission of the copyright owner. © 2001 Geological Society of Australia Incorporated ISSN 0072-1085 INQUIRIES AND ORDERS SHOULD BE DIRECTED TO: The Business Manager, Geological Society of Australia 706 Thakral House, 301 George Street N S W 2000 Australia

GONDWANA TO GREENHOUSE : AUSTRALIAN ENVIRONMENTAL GEOSCIENCE. Bibliography Includes index ISBN 1 876125 22 5

1. Geology - Australia. 2. Environmental geology - Australia. I. Gostin, V. A. (Victor A.). II. Geological Society of Australia. (Series: Special Publication (Geological Society of Australia); No. 21). 559.4

GONDWANATOGREENHOUSE AUSTRALIAN ENVIRONMENTAL GEOSCIENCE

Edited by V. A. GOSTIN

Geological Society of Australia Inc. Special Publication No. 21

THE GEOLOGICAL SOCIETY OF AUSTRALIA IS GRATEFUL FOR THE FINANCIAL SUPPORT OF THE FOLLOWING ORGANISATIONS AND INDIVIDUALS IN PUBLISHING THIS VOLUME: BUREAU OF RURAL SCIENCES MURRAY-DARLING BASIN COMMISSION AUSTRALIAN GEOLOGICAL SURVEY ORGANISATION AUSTRALIAN INSTITUTE OF GEOSCIENTISTS C S I R O ENVIRONMENTAL PROJECTS OFFICE CRCLEME JAMES COOK UNIVERSITY MARINE GEOPHYSICAL LABORATORY VIC SEMENIUK AND PETER HARRIS

AGSO

MURRAYDARLING B A S I N

COMMISSION AUSTRALIAN GEOLOGICAL SURVEY

CRCLEME

CONTENTS 1 wmMmmrnm

Foreword C. J. Chartres Introduction—Ancient Gondwana and an Australian environmental perspective V. A. Gostin

Theme 1—Ancient Australia and environmental changes Chapter 1 Quaternary climatic changes in Australia and their environmental effects M. A. J. Williams Chapter 2 Regolith: its history and environmental importance with particular reference to some engineering examples G. Taylor and G. H. McNally Chapter 3 Soil-related engineering problems: identification and remedial measures R. 1/1/ Fitzpatrick, P Slade and P. A. Hazelton Chapter 4 Avon River palaeodrainage system, Western Australia: geomorphological evolution and environmental issues related to geology M. J. Freeman Chapter 5 Ancient landforms of Kambalda and their significance to human activity J. D. A. Clarke

Theme 2—Geohazards in urban communities Chapter 6 Geohazards risk to urban communities in Australia K. Granger Chapter 7 Land instability: a case study from the lllawarra region, New South Wales B. G. Jones,. R. N. Chowdhury and M. H. Ghobadi

Theme 3—Mining and radioactivity Chapter 8 Acid drainage at minesites G. F. Taylor

Chapter 9 Hydrogeology and geochemistry of sediment banks contaminated with mine tailings in the King RiverJasmania D. C. McPhail, D. Green and 1/1/ C. Hooper

Chapter 10 Natural radioactivity, hazards, wastes and the environment in Australia R. Major

111

Theme 4—Water and sedimentary basins Chapter 11 Hydrology and environmental geology of the Great Artesian Basin, Australia M. A. Habermehl

127

Chapter 12 Water and mining in the Murray-Darling Basin: meeting the demands for a sustainable environment J. F. Lovering, P. Crabb and K. Goss

145

Chapter 13 Salinisation and environmental geoscience: a case study in the Murray Basin R. S. Brodie

] 53

Chapter 14 Environmental geoscience issues in the Gippsland Basin, Victoria L. Hodgson

15]

Chapter 15 Groundwater for Aboriginal communities in central Australia the Western Water Study (Wiluraratja Kapf), Northern Territory G. Jacobson and J. Wischusen

17 ]

Theme 5—Coastal and nearshore environments Chapter 16 Human impacts on geoheritage features of the Swan Coastal Plain and coastal zone, southwestern Australia V. Semeniuk and C. A. Semeniuk

181

Chapter 17 Late Quaternary sea-levels, climate change, and South Australian coastal geology N. Harvey, A. P. Belperio and R. P. Bourman Chapter 18 Value of estuarine sediments for understanding natural environmental chanae in y coastal regions S. L. Nichol

201

11 n

Chapter 19 Trace-metal pollution and sedimentation in coastal lagoons: an example from Lake lllawarra, New South Wales B. E. Chenhall, B. G. Jones and A. M. Depers

227

Chapter Stable 20 lead isotopes: application to sourcing of lead in the environment B. E. Chenhall, M. Chiaradia, B. L. Gulson, B. G. Jones and A. M. Depers

235

Chapter 21 Role of geology in environmental science: an example of pollution assessment of the estuarine and marine environment in central New South Wales G. F. Birch, S. E. Taylor and C. Matthai

243

Chapter 22 Contribution of geological process studies to environmental management: case studies from northeast New South Wales and southeast Queensland W. E. Boyd, R. T. Bush, M. W. Clark, J. V. Smith and L. A. Sullivan Chapter 23 Study of historical change as a contribution to environmental management issues: case studies from northeast New South Wales and southeast Queensland W. E. Boyd, M. M. Cotter, R. McGrath, S. Pathirana and A. Specht Chapter 24 Holocene Great Barrier Reef: sedimentary controls and implications for environmental management P. Larcombe

257

267

281

Chapter 25 dispersal along the inner shelf of the central Great Barrier Reef Sediment K. J. Woolfe and P. Larcombe

295

Chapter 26 Geomorphology and Holocene geology of coastal and estuarine plains of northern Australia J. Chappell

303

Theme 6—Marine geoscience Chapter 27 Environmental management of Torres Strait: a marine geologist's perspective P.T.Harris

317

Chapter 28 Australia's marine jurisdiction: baseline information for seabed management and monitoring climate change H. M. J. Stagg, N. F. Exon, P. J. Hill, D. T. Heggie and P. E. O'Brien Indexes Plates 1 to 16

329 339 341

Geological Society of Australia Special Publication 21, ix

FOREWORD Chief, Land and Water Sciences Division, Bureau of Rural Sciences., PO Box El I Kingston, ACT 2604, Australia.

Australia's combination of ancient geology, predominantly dry climates and vegetation has lead to the development of a range of environments, many of which are, if not unique, rarely encountered elsewhere on earth. This means that we are often unable to import solutions to a wide range of environmental, agricultural and resource assessment and management problems. Consequently, there is a major need for us to develop solutions to these problems that are tailored specifically to our environments. To do this we need to have a deep and broad understanding of geology, in terms of the distribution of rock and sediment types, the regolith and associated soils and the dynamic processes operating in the environment and how they interact with human activities. From a geological perspective the last 200 years of Australian history have focused on resource assessment and exploitation to the benefit of increasing our national wealth. The next 200 years will certainly be very different, even though Australia will still be a major supplier of coal, iron, non-ferrous metals and other mineral and hydrocarbon resources. National attention and that of many geoscientists, will increasingly turn to environmental-management issues. Whilst mining can have a major impact on land and water resources, the total area occupied by mines is <0.02% of the continental land surface. Furthermore, the pointsource nature of environmental problems associated with mining and the funds generally set aside by companies for rehabilitation mean that land management and degradation amelioration is easier to handle using a range of tested technologies. In contrast to mining, the problems of land degradation associated with agriculture and urban/industrial land uses, are increasingly becoming of concern both because of their areal extent and increasing severity. Recent estimates made by the Land and Water Resources Research and Development Corporation have suggested that the cost of five major resource degradation related issues (waterlogging and salinity; soil erosion, compaction and structural decline; acidification; loss of added fertilisers; and treatment of deteriorating water supplies) is in the order of $1.7 billion per year. The value of fixing up accumulated land degradation damage is probably immense and beyond our capacity to achieve, yet alone afford. The majority of this degradation results from the use of northern-hemisphere farming systems on Australia's old, fragile and droughtprone environments.

Geohazards are also becoming of increasing concern in Australia and the surrounding region. Whilst there is no definitive evidence to suggest that earthquakes or volcanic eruptions are becoming more frequent, these phenomena, as well as landslides and tsunamis, are impacting on more people when they occur because of rising population densities and consequent development on less-suitable terrain. Coupled with global warming and rising sea-levels and their attendant risks, geohazards present a major threat to the stability of 21st century infrastructure and communities. Consequently, there will be a growing demand for geohazard studies and management plans throughout the region. Currently, another critical issue to our legislators is that of managing our dwindling water supplies. Whilst there is a large literature on surface water, less is known and understood about groundwater. In southern Australia, virtually all surface waters have been allocated to users. This is causing an increasing demand for exploitation of groundwaters. Developing suitable management practices in order to enable sustainable groundwater use will also be a major focus for hydrogeologists over the next decade. It is, however, vital that we do so because our groundwaters are generally relatively free of anthropogenic contaminants, and a major contributor to river flow and numerous habitats of major significance to the maintenance of our biodiversity. The Bureau of Rural Sciences' role is to provide scientific advice to government. Recently, we have worked on issues as diverse as underpinning the Great Artesian Basin Sustainability Initiative, locating a site for Australia's lowlevel radioactive waste repository, and salinity and waterquality management. Every one of these and a myriad of other issues we deal with relate to the sustainability of our natural-resource base and environments. Geological knowledge is fundamental to managing sustainably. The chapters in this book present an excellent introduction to environmental geology and associated disciplines. The book helps us understand how palaeoenvironmental conditions and processes have led to the development of our current landforms, soils and environments and how current processes impact on a wide range of natural environments and land uses. Moreover, strong linkages are made between the biophysical environment and the human environment. Understanding these linkages is critical if we are to develop sustainable land-management paradigms and methodologies for the 21st century and beyond.

Geological Society of Australia Special Publication 21/xi-xvi

INTRODUCTION—Gondwana to Greenhouse: Australian environmental geoscience V. A. GOSTIN Department of Geology & Geophysics, University of Adelaide, SA 5005, Australia (victor.gostin@adelaide. edu. au).

INTRODUCTION Environmental concerns and sustainable development have become important issues for the whole planet. In order to better manage our interaction with the environment we need to understand the nature of the Earth's surface, its past and present climates, its soils, sediments, weathered rock and deeper geological structures, as well as the sea floor. While there exist many useful texts dealing with environmental geology, including natural hazards, they are mainly based on northern hemisphere areas with significantly different geology and they do not sufficiently address issues that are common to Australia. This book therefore aims to supplement existing texts with a compendium of key articles by Australian authors applying a geoscientific approach to Australian environments. The book is written for all science-trained people, from students to professionals, but it will also appeal to interested non-scientists wishing to inform themselves about environmental earth science, specifically relevant to Australia. The book title, Gondwana to Greenhouse, is intended to convey the concept of an ancient land that has been subject to millennia of environmental change, which currently incorporates human-generated climatic instability and sea-level rise. The purpose of this introduction is to identify major environmental geoscience issues that affect Australia, and hence the reasons for selecting the range of topics included in this book. Its broad scope indicates the wide variety of topics that are currently being addressed by Australian geoscientists. The reference map (inside front cover) gives the general location of studies covered by the specific chapters. Most chapters include comprehensive reference lists that will lead to both general and specialist publications, and the bibliography at the end of this introduction gives some useful references and web addresses on a number of environmental geoscience-related themes.

plines. Of major importance are studies in geochemistry, biogeochemistry, sedimentology, soil science and engineering geology, hydrogeology, geophysics, and remote sensing. The relationship of geology to other sciences and its way of engaging with the earth, is explored well in the tenth Mawson Lecture (Seddon 1996). Unlike most other sciences, geoscience has a deep time perspective as it engages with processes that have operated over very long time spans in the formation and modification of our planet. The time-scale over which various earth processes operate varies from seconds in the case of earthquakes to years for ENSO (El Nino Southern Oscillation). It extends to thousands and tens of thousands of years for climatic and sea-level oscillations, and to millions of years for massive crustal movements. Some processes are cyclic, some are random: together they have created a unique planetary history. As world human population has grown and as technological innovations keep crossing new frontiers of development, so human activities have become significant in seriously altering the natural environment. Indeed, in recent decades, humanity has become a very important player in global environmental issues. An example is greenhouse warming caused primarily by the burning of fossil fuels. The history of global warming has been reviewed by Christianson (1999). The Commission on Geological Sciences for Environmental Planning (COGEOENVIRONMENT) was established by the International Union of Geological Sciences (IUGS) in 1990 to address specifically the role of geoscience in environmental issues. Part of the mandate of COGEOENVIRONMENT is to increase awareness among the general public of the essential contribution of geological processes to sound planning and management of the environment. Publications of the commission include McCall et al (1996), and Berger and lams (1996).

Locating geoscience in environmental disciplines

Sources of primary environmental information: geological maps and data banks

Professional geoscientists recognise that geology is much more than the study of rocks and fossils, or the discovery of mineral and petroleum resources. Since the Earth is a very complex and large interactive system, earth science or geoscience necessarily involves consideration of the atmosphere, hydrosphere, cryosphere (ice), lithosphere, pedosphere (soils), as well as the biosphere. This overlap is critical to those practising environmental geoscience, requiring them to be conversant with many scientific disci-

The present environment is just a transitory phase produced by past processes and ever-changing environments. Unravelling the geological history of an area enables us to better understand its present geology, and hence local environmental implications and interactions. Many chapters in this book refer to geological maps and other databases. These show not only the age, nature and distribution of hard rocks and softer sediments, but also important relationships between rock formations, structures and cross-

xii

V. A. Gostin

sections of the crust. Geological Surveys at both State and Federal levels are important repositories of geological information relevant to the Australian environment. Seeing that many environmental issues include rock/soil/water factors, it is wise to begin any study by examining all relevant maps showing geological, hydrogeological and geomorphological features. Increasing amounts of geological information are now stored in electronic data banks, which may include useful geochemical, radiometric, engineering, as well as borehole information. A direct indication of flood-prone areas, for example, may be obtained from many geological maps by noting the distribution of 'Holocene alluvium'. Similarly, the extent of tidal inundation and the distribution of coastal sands are usually clearly defined. Geological maps also may indicate the location of granites and springs with high radioactivity, abandoned metal sulfide mines that may discharge toxic acid waters, and buried ancient river channels that carry useful (or salty) groundwater (Hodgson Chapter 14; Taylor & McNally Chapter 2). Remote sensing is playing an increasing role in research and mapping of problem soils, salinisation and pollution. As well as ground-penetrating radar, the techniques include satellite imaging, airborne hyperspectral remote sensing, airborne electromagnetics, and gamma-ray spectrometry. A multidisciplinary 'geological systems' approach using such techniques has recently proved successful in understanding the processes involved in land and water salinisation (Lawrie et al 2000; Jacobson & Wischusen Chapter 15).

Urban Australia and engineering geology Australia is a highly urbanised society, with 84% of its population residing in only 1% of its area (Australian Bureau of Statistics 2000). Clearly, the geology of urban environments is most important not only to engineering works, but to town planning, waste disposal, ground pollution and community health. And, since urban populations are fed and take their leisure by drawing on rural areas, the interaction between these is crucial to a proper appreciation of environmental issues. Space does not permit more than a few Australian case studies in this book, and the reader is referred to the selected bibliography. The most expensive of all geohazards are the problems of building houses and roads on problem soils. These include soils that are expansive, collapsing, creep-prone, dispersive, saline, and so on (Fitzpatrick et al Chapter 3; Taylor & McNally Chapter 2). The cumulative cost to communities worldwide from such situations far exceeds that of earthquakes and volcanism. Ground subsidence due to coal mining, groundwater extraction and fill settlement is becoming more significant. In Newcastle, for example, one can no longer avoid old mine workings in freeway cuttings or deep foundations (McNally 2000b). In Gippsland, Victoria, large areas are sinking due to lignite-mine dewatering (Hodgson Chapter 14).

Heritage The conservation of Australian geological and geomorphological heritage features has become increasingly important d other landscape modifications

have destroyed many important sites (Semeniuk & Semeniuk Chapter 16). In Australia, consideration of such geoheritage issues is the concern of the Geological Heritage Subcommittees of the Geological Society of Australia (Joyce 1999; Eberhard 1997; www.gsa.org.au).

GONDWANA TO GREENHOUSE: AUSTRALIA'S GEOLOGICAL CHARACTERISTICS Australia has many unique features and specific geological characteristics that are relevant to major or potential environmental impacts. Australia represents the ancient interior and more stable part of a bigger Indo-Australian crustal plate whose active edges extend to the surrounding volcanic zones of New Zealand, Fiji, Noumea, Papua New Guinea and Indonesia. While earthquakes are not numerous in Australia, those that do occur may be intense and remain unpredictable (McCue 1995; Gaull & Kelsey 1999). Tsunamis are also rare, and their record on the east coast has been described by Bryant and Young (1996). For a very long time, Australia formed part of the supercontinent of Gondwana (or Gondwanaland), and many of its landscape features go back to those times and to subsequent episodes of continental breakup. This fragmentation began about 154 Ma (Late lurassic) in Australia's northwest, then west, east and finally along the southern edge around 80 Ma (Late Cretaceous). A summary of the last billion years of earth history of Australia and its Gondwanan neighbours is given in Veevers (2000). An introduction and comprehensive guide to the Geology of Australia is available on 2 CD-ROMs (CD Solutions 1996). Unlike the continents of the northern hemisphere where recent mountain building and extensive surface modifications have taken place, Australia has had a relatively stable recent geological history. For the last 2 million years the continent has neither been scraped clean by ice nor fertilised by glacial flour. Furthermore, compared to some areas, Australia's soils have only very locally been affected by enriching volcanic activity. Apart from scattered rocky outcrops, the Australian landscape consists largely of a deep regolith—the cover of soils, sediments and deep weathering that overlies the hard bedrock. This regolith is largely a product of extensive water erosion, transport and deposition, modified somewhat by wind activity. It has also been altered by physical, chemical and biological processes over various time-scales. This has resulted in a mosaic of landforms, residual regolith and sediments that show a variety of profiles, often in close proximity, indicating different histories of formation. Locally, the regolith contains harder crusts enriched in iron and/or aluminium (ferricrete, bauxite), silica (silcrete), or limestone (calcrete). Environmentally, the nature of this regolith profoundly influences the character of the groundwater, vegetation, and by implication, human land use (McNally 1995; Taylor & McNally Chapter 2). Australia is the most low-lying and flattest of all continents, resulting in poor drainage, extensive flooding, and retention of salinity in many soils. This is also reflected in abundant coastal lowlands, especially in areas facing wide and shallow continental shelves. Many of these coastlines are not in equilibrium and are either eroding or accre -

Introduction This is because present sea-levels have existed for less than 7000 years, very recent in terms of coastal processes. The present climate is even younger, so that in the semiarid southwestern Murray Basin, the fresh groundwater derives from a wet climate merely 8000-4000 years ago (Leaney et al 1995). In contrast, some Great Artesian Basin waters accumulated in rains that fell over 1 million years ago (Habermerl Chapter 11). The low topography in central inland eastern Australia includes the catchments of the Gulf of Carpentaria in the north and the Murray-Darling Basin in the southeast, containing the lowest river gradients in the world. Furthermore, in central Australia the rivers flow inland to Lake Eyre, which is about 13 m below sealevel. In these lowlands are many lakes, marshes, and similar wetlands that are replenished by tropical cyclones. These are the foci of enormous biodiversity of great importance. The areas are very vulnerable to the impacts of development or other interference from human activities. Australia thus emerges as a relatively stable continent with its own profile of environmental issues and many of these are presented in this book. Six major themes can be identified, and notwithstanding the inevitable overlaps, the chapters are arranged according to these themes.

Theme 1, Ancient Australia and environmental changes Australia's Gondwanan inheritance has resulted in large areas with deep regolith of weathered rock, poor soils, ancient drainage and limited water supplies. Recognising this ancient profile is crucial to understanding the present soil degradation, salinisation and desertification in a country that maintains a strong agricultural base. To begin with, it is salutary to examine the geologically short-term environmental fluctuations that have followed one upon the other, in order to better appreciate the longterm evolution of the Australian landscape. The first chapter therefore examines the large variations of climate, sea-levels and changing land areas during the last 100 thousand years. Climatic conditions strongly affect the distribution of plant and animal species. Thus past and future changes in climate are crucial to the evolution and health of the biosphere. It is also important to realise that humans have modified the Australian environment for at least 60 thousand years. The second and third chapters deal with soils and the underlying deeply weathered rocks so common in Australia. This regolith is shown to vary greatly in age, structure and composition. The regolith controls plant ecology and faunal habitats, and it needs to be carefully mapped and understood not only for reliable foundations in the building industry, but for the management of extensive underground pipes and power lines. Chapters 4 and 5 deal with the environmental repercussions of the very ancient drainage systems and topography of southwestern Australia, including the Kambalda mining region.

xiii

inside front cover) and is subject to various geological hazards including landslides, the occasional earthquake and widespread soil heave. Cairns, in tropical Queensland (Chapter 6) is used as a case study to demonstrate the geohazard risk assessment methods and GIS support tools developed under the Cities Project of the Australian Geological Survey Organisation. Landslides and land instability affect only a minor part of Australia. Major urban areas affected include Cairns, and the areas immediately north, south and southwest of Sydney, represented by a case study of rockfalls, slumps, and mudflows in the Illawarra region (Chapter 7).

Theme 3, Mining and radioactivity While mining disturbs less than 0.02% of Australia's land area, its environmental effects are more widespread if one includes air-pollution from smelting and water pollution from abandoned mine workings and tailings. Land rehabilitation remains a lasting problem. Chapter 8 describes the environmental consequences resulting from various Australian mining activities, and addresses the required treatment of mine wastes. The next chapter discusses geochemical and hydrogeological characteristics of the mine wastes from Mt Lyell, Tasmania, redeposited along the banks of the King River. Natural or background geochemistry, including radioactivity, depends on the local geology and thus varies widely. Its nature and degree of toxicity needs to be better understood. Chapter 10 is devoted to natural radioactivity and uranium mining in Australia, given the widespread but largely uninformed public concern, and the need for accurate information. It also reports on the identification of geologically safe repositories for nuclear wastes.

Theme 4, Water and sedimentary basins Australia has two huge interior basins—the Murray-Darling and the older Great Artesian Basin, plus many smaller coastal basins, filled with Mesozoic to Holocene sediments and sedimentary rocks. These sediments and rocks carry abundant groundwaters that vary in salinity and therefore usefulness to this densely populated and intensely developed part of the continent. Chapters 11 to 13 deal with the Great Artesian and Murray-Darling Basins, and the ongoing or proposed activities aimed at reducing water wastage and the scourge of increasing salinity. The coastal Gippsland Basin in Victoria, containing huge opencut brown-coal mines is described in Chapter 14. Here groundwater is also important for domestic supplies, but the exploitation of water, coal and petroleum has resulted in subsidence, salinity and coastal changes. In regions outside such sedimentary basins, the availability and quality of groundwater is highly variable, and often poor. Chapter 15 addresses the cultural and geological issues involved in obtaining clean and dependable supplies of water for Aboriginal communities in central Australia.

Theme 2, Geohazards in urban communities

Theme 5, Coastal and nearshore environments

Australia's highly urbanised population is located largely along the eastern and southern coast of the mainland (see

Australia's very long coastal zone features wave-dominated beaches with rocky headlands, or else estuaries, lagoons,

xiv

V. A. Gostin

and in places, wide tidal flats. Fringing coral reefs and the largest barrier reef in the world, occur north of latitude 24°S, and are clearly environmentally important. The effects of changing sea-levels, coastal developments and pollution are therefore crucial, the more so since about 80% of Australia's population live along this coast. Eleven chapters ( 1 6 - 2 6 ) are devoted mainly to Australian coastal environmental issues beginning with the diverse geomorphology of the Swan Coastal Plain in Western Australia, and its geoheritage features. Local sealevel history reflects both changing global sea-levels and the elevation or subsidence of the earth's crust. Chapter 17 discusses the combined effect of these impacts on the South Australian coastal zone. Detailed analyses have resulted in improved estimates of current sea-level trends, greenhouse-accelerated sea-level rise predictions, and related Government legislation. Chapters 18 to 23 discuss the geological history of estuaries and various environmental issues, especially pollution, that affect the highly populated coastal zone of New South Wales and southeast Queensland. Chapter 18 examines the effects of sea-level rise, flooding and acidification. Chapters 19 and 2 0 study the pollution levels of a lagoon alongside an old smelting and metal processing complex near Wollongong, and the use of lead isotopes to identify the likely sources. Sydney's four major estuaries and the adjacent continental shelf are the focus of detailed pollution studies in Chapter 21. New insights are provided into the behaviour of heavy-metal contaminants that are regularly resuspended into the overlying water column. Nine separate case studies are presented in the complementary Chapters 22 and 23, covering northeast New South Wales and southeast Queensland. All show the importance of understanding the geological processes involved as well as the history of environmental changes that have affected specific areas. These studies cover water management, conflicting land use, remediation of mine and quarry sites, plantations and fire management, as well as environmental management relating to cultural heritage. The largest, most complex and most magnificent coral reef system in the world is the Great Barrier Reef of Australia. Chapters 24 and 25 describe the geological and biological aspects of its growth, and the potential threat to its health from increased sediment input and the occurrence of the destructive crown of thorns starfish. Also discussed are the effects of intense tropical cyclones and southeast trade winds on coastal sediment dispersal, as well as the impact of dredging. Proposals for improved methods of environmental management are also presented. The vast tropical coastal lowlands of northern Australia are dominated by macrotidal to mesotidal conditions, and regions with extensive wetlands of precious biodiversity. Chapter 2 6 describes the geomorphology, geology and present processes of several river and estuarine environments in this region with the aim of showing how knowledge of their geohistory is essential to their conservation and better management.

Theme 6, Marine geoscience ralia's marine jurisdiction extends to about twice the size - Marine geoscience addresses environmental

issues by providing essential baseline information that promotes improved seabed management. Chapter 27 examines the submarine environments across Torres Strait that could affect the laying of vulnerable gas pipelines, and that currently affect the dispersal of substantial Ok Tedi mine wastes, which flow down the Fly River of Papua New Guinea. The concluding Chapter 2 8 discusses the importance of marine geoscience to almost every sector of the marine economy and deals with methods of obtaining the necessary baseline information for seabed management. This includes systematic seabed sampling together with modern 3-D swath-mapping of the sea floor. The retrieval and study of marine sediment cores, going back millions of years, will also improve our knowledge of the climatic record and hence the modelling of climatic changes. Marine geoscience can thus improve our understanding and forecasting of Australia's highly variable climate which causes great problems for agriculture, water supply, droughts and floods.

CONCLUSION Arranged into six themes, the chapters of this book cover many parts of Australia, from the ancient and arid heartland, to the coastal zone, and out to the seaward limits of its jurisdiction (see map inside front cover). They are written by experts in their field, emphasising the centrality of geoscience in addressing many environmental problems. This book brings specialised modern geoscientific knowledge to the attention of a wider audience and hopefully will form part of the education of tomorrow's geoscientists. It presents an overview of the major geoscientific environmental issues facing Australia today, describes the many and varied ways of tackling the issues, and shows the vital role that geoscience plays in understanding the environment and its management.

ACKNOWLEDGEMENTS It has been an exciting adventure to conceive and compile this book concerning Australian achievements in environmental geoscience. I wish to thank sincerely all contributors for volunteering their chapters in this large compendium. Three years have elapsed since the arrival of the first chapters, and I offer my profound thanks to those early contributors for their patience and forbearance. It was my decision to seek the coverage and scope now encompassed in this volume. The final collected product is close to what was originally intended, and I hope that all contributors will feel the sense of exhilaration and completion that comes with the delivery of this volume. I wish to thank all reviewers of the chapters, and those who have helped with the numerous illustrations. The Honorary Editor of the Geological Society of Australia needs special praise for his painstakingly thorough work in producing the final corrected printed version; and the Executive Committee of the Society is thanked for the idea of accepting and publishing these chapters in a special publication. Landsat imagery of Perth from Space was provided by

Introduction Australian Centre for Remote Sensing (ACRES), AUSLIG, Canberra, and digitally enhanced and produced by Satellite Remote Sensing Services, Department of Land Administration, Western Australia. I wish to thank our sponsors, the Bureau of Rural Sciences, Murray-Darling Basin Commission, Australian Geological Survey Organisation, Australian Institute of Geoscientists, CSIRO Environmental Projects Office S.A., CRCLEME, James Cook University, Marine Geophysical Laboratory and Vic Semeniuk and Peter Harris.

REFERENCES AND SELECTED BIBLIOGRAPHY General environmental geoscience D . E . & FAIRBRIDGE R . W . 1 9 9 9 . Encyclopedia of Environmental Science. Kluwer Academic Publishers, Norwell. 1 9 9 8 . Environmental Geology. Principles and Practice. Blackwell Science, Oxford. BENNETT M. R. & DOYLE P. 1997. Environmental Geology. John Wiley & Sons, London. COGEOENVIRONMENT—The IUGS Commission on Geological Sciences for Environmental Planning (www.agso.gov.au/cogeo). KELLER E. A. 1992. Environmental Geology (6th edition). Macmillan, New York. MARSHALL C . P. & FAIRBRIDGE R . W . 1 9 9 9 . Encyclopedia of Geochemistry. Kluwer Academic Publishers, Norwell. MERRITTS D . , D E W E T A . & MENKING K . 1 9 9 8 . Environmental Geology. W.H. Freeman & Co., New York. MONTGOMERY C. W. 2000. Environmental Geology (updated 5th edition). McGraw-Hill, New York. MURCK B . W., SKINNER B . J . & PORTER S . C . 1 9 9 6 . Environmental Geology. John Wiley & Sons, New York. REYNOLDS J. M . 1 9 9 7 . An Introduction to Applied and Environmental Geophysics. Wiley and Sons, New York.

ALEXANDER BELL F. G .

Geology and environment of Australia AUSTRALIAN GEOLOGICAL SURVEY ORGANISATION. <www.agso.gov.au>

J. H. C. & DRAPER J. J. (Editors) 1997. North Queensland Geology. Australian Geological Survey Organisation Bulletin 240.

BAIN

B M R PALAEOGEOGRAPHIC GROUP 1 9 9 0 . Australia—evolution of a conti-

nent. Australian Government Publishing Service, Canberra. 1989. Geology and Mineral Resources of Tasmania. Geological Society of Australia Special Publication 15. C D SOLUTIONS 1 9 9 6 . Geology of Australia. An introduction to earth sciences and a comprehensive illustrated guide to the geology of Australia. ( 2 C D - R O M pack). C D Solutions, North Sydney, Australia. COCHRANE G. W., QUICK G. W. & SPENCER-JONES D. (Editors) 1991. Introducing Victorian Geology. Geological Society of Australia, Victorian Division, Melbourne. C S I R O LAND AND WATER, <www.clw.csiro.au/links/> DOUGLAS J. G. & FERGUSON J. A. (Editors) 1988. Geology of Victoria (2nd edition). Geological Society of Australia, Victorian Division, Melbourne. DREXEL J. F. & PREISS W. V. (Editors) 1995. The geology of South Australia. Vol. 2, The Phanerozoic. Geological Survey of South Australia, Bulletin 54. ENVIRONMENT Australia ONLINE, <www.environment.gov.au> GEOLOGICAL SURVEY OF WESTERN AUSTRALIA 1 9 9 0 . Geology and Mineral Resources of Western Australia. Geological Survey of Western Australia Memoir 3. JONES P. J . (Compiler) 1996. AGSO Phanerozoic Timescale 1995. Wallchart with Explanatory Notes. Oxford University Press, Melbourne. MURRAY-DARLING BASIN COMMISSION, <www.mdbc.gov.au/> PALFREYMAN W. D. 1984. Guide to the geology of Australia. Bureau of Mineral Resources Bulletin 181. SCHEIBNER E. & BASDEN H. (Editors) 1998. Geology of New South BURRETT C. F. & MARTIN E. L. (Editors)

Wales—Synthesis. Volume 2 Geological Evolution. Geological Survey of New South Wales, Memoir 1 3 ( 2 ) . and neighbours in Gondwanaland. GEMOC Press, Macquarie University, Sydney. YOUNG G. C. & LAURIE J. R. (Editors) 1996. An Australian Phanerozoic Timescale. Oxford University Press, Melbourne. VEEVERS J. J. (Editor) 2000. Billion-year earth history of Australia

Engineering geology, hydrogeology, mining BARNETT S . R . , HOWLES S . R . , MARTIN R . R . & GERGES N . Z .

2000.

Aquifer storage and recharge: innovation in water management. Australian Journal of Earth Sciences 4 7 , 1 3 - 1 9 . BENTLEY S. P. (Editor) 1995. Engineering Geology of Waste Disposal. Geological Society of London Engineering Geology Special Publication 11. COMMANDER D. P. 1994. Interaction of land use, water supply and geology on the Swan coastal plain. Geological Society of Australia (WA Division) Excursion Guidebook No.9. DOMENICO P. A . & SCHWARTZ F. 1 9 9 8 . Physical and Chemical Hydrogeology (2nd edition). John Wiley & Sons, Sydney. FITYUS S., HITCHCOCK P., ALLMAN M . A. & DELANEY M . (Editors) 1 9 9 8 . Geotechnical engineering and engineering geology in the Hunter Valley. Australian Geomechanics Society/Conference Books, Springwood. HUDAK P. F. 2000. Principles of Hydrogeology (2nd edition). Lewis Press, Boca Raton. LAWRIE K . C . , MUNDAY T. J . , DENT D . L., GIBSON D . L., BRODIE R . C . ,

WILFORD J . , REILLY N. S . , CHAN R . A . & BAKER P. 2000. A geologi-

cal systems approach to understanding the process involved in land and water salinisation. A G S O Research Newsletter 32, 13-15, 26-32. LEANEY F. W . , HERCZEG A . L. & LOVE A . J . 1 9 9 5 . Fresh groundwater in the S-W Murray Basin resulting from recharge during wet climatic periods. Australian Geological Survey Organisation Record 1995/61,

165-169.

W. 1998. Landscape Planning (3rd edition). John Wiley & Sons, Sydney. MCCALL G. J. H., D E MULDER E. F. J. & MARKER B. R. (Editors) 1996. Urban Geoscience. Balkema, Rotterdam. MCNALLY G. H. 1995. Engineering characteristics and uses of duricrusts in Australia. Australian Journal of Earth Sciences 42, 535-547. MCNALLY G. H. (Editor) 1997. Collected Case Studies No 3Engineering Geology Hydrogeology and Environmental Geology. Geological Society of Australia, Environmental, Engineering, Hydrogeology Specialist Group, Sydney MCNALLY G . H . (Editor) 2000a. Some issues in environmental geology in Australia. Australian Journal of Earth Sciences 47, 1-82. MCNALLY G. H . (Editor) 2000b. Geology and mining practice in relation to shallow subsidence in the Northern Coalfield, New South Wales. Australian Journal of Earth Sciences 47, 21-34. MCNALLY G. H. & FRANKLIN B . J . (Editors) 2000. Sandstone citySydney's dimension stone and other sandstone geomaterials. Geological Society of Australia, Environmental, Engineering, Hydrogeology Specialist Group, Sydney, Monograph 5. MCNALLY G. H. & JANKOWSKI J . (Editors) 1998. Environmental Geology of the Botany Basin. Geological Society of Australia, Environmental, Engineering, Hydrogeology Specialist Group Papers Series 4. MCNALLY G. H., CLARKE G. & W E B E R B. W . 2000. Porcellanite and the urban geology of Darwin, Northern Territory. Australian Journal of Earth Sciences 47, 35-44. MULLIGAN D. (Editor) 1 9 9 9 . Environmental Management in the Australian Minerals and Energy Industries—Principles and practice. University of New South Wales Press, Sydney. PECK W . A . , NEILSON J . L., OLDS R. J . & SEDDON K . D . (Editors) 1 9 9 2 . Engineering Geology of Melbourne. Balkema, Rotterdam. SHARMA P. 1997. Environmental and Engineering Geophysics. Cambridge University Press, Cambridge. SHEARD M.J. & BOWMAN G. M. 1996. Soils, stratigraphy and engineering geology of near surface materials of the Adelaide Plains. Mines & Energy South Australia Report Book 94/9 (also as a CD-ROM Data Package). SLOAN S. W. & ALLMAN M. A. (Editors) 1995. Engineering geology of the Newcastle-Gosford region. Australian Geomechanics Society/Conference Books, Springwood. MARSH

xvi

V. A. Gostin

Natural hazards BELL F. G . 1 9 9 9 . BRYANT E . A .

Cambridge.

BRYANT E. A.fieY O U N G

R. W . 1996. Bedrock-sculpturing by tsunami, South coast New South Wales, Australia. Journal of Geology 104, 565-582. GAULL B . A . fie KELSEY P. 1 9 9 9 . Historical felt intensities as a guide to earthquake hazard. Australian Journal of Earth Sciences 4 6 , 365-376. M C C U E K. 1995. Atlas of isoseismal maps of Australian earthquakes. Part 3. Australian Geological Survey Organisation Record 1 9 9 5 / 4 4 . NATURAL HAZARD RESEARCH CENTRE.

<www.es.mq.edu.au/NHRC/>

Coastal and marine environments B . fie KNECHT R . 1 9 9 8 . Integrated Coastal and Ocean Management. Concepts and Practices. Island Press, Washington. A . R. 1996. Coastal Zone Management Handbook. Lewis Publishers, Boca Raton. COMMONWEALTH OF AUSTRALIA 1995. Living on the Coast. The Commonwealth Coastal Policy, Canberra.

CICIN-SAIN CLARK

COOPERATIVE

RESEARCH

CENTRE

WATERWAY MANAGEMENT,

FOR

COASTAL

ZONE,

ESTUARY

<www.coastal.crc.org.au/>

ENVIRONMENT AUSTRALIA (marine website).

<www.environment.gov.au/marine/> 1 9 9 7 . Coastal and Estuarine Management. Routledge, London. HARVEY N. (Editor) 1996. Proceedings of the Australian Coastal Management Conference, Glenelg. University of Adelaide, Adelaide. KAY R. fie ALDER J. 1999. Coastal Planning and Management, EfieFN SPON/Routledge, London. V I L E S H. & SPENCER T. 1995. Coastal Problems: Geomorphology, Ecology and Society at the Coast. Edward Arnold, London. FRENCH P. W .

Geomorphology, greenhouse, hydrology, Quaternary geology, soils ANTARCTIC CO-OPERATIVE RESEARCH CENTRE.

<www.antcrc.utas.edu.au/antcrc/> D. 1997. Floodplain Processes. John Wiley and Sons,, London. ASWATHANARAYANA U. 1999. Soil Resources and the Environment. Scientific Publishers Inc., Enfield. B E R G E R A . R. fie IAMS W . I . (Editors). 1996. GEOINDICATORS— Assessing Rapid Environmental Changes in Earth Systems. Balkema, Rotterdam. BRIZGA S. fie FINLAYSON B . (Editors) 2000. River management: The Australasian experience. John Wiley and Sons, London. BUREAU OF RURAL SCIENCES, <www.affa.gov.au/> Volume CALOW P.fiePETTS G. E. (Editors) 1992 The Rivers Handbook. 1. Blackwell Scientific, Oxford. CALOW P.fiePETTS G. E. (Editors) 1994 The Rivers Handbook. Volume 2. Blackwell Scientific, Oxford. ANDERSON M. fie WALLING

1999. Greenhouse: The 200-year Warming. Walker and Co., New York.

CHRISTIANSON G . E .

Geological Hazards. Routledge, London. Natural Hazards. Cambridge University Press,

1991.

Story of

Global

COOPERATIVE RESEARCH CENTRE F O R LANDSCAPE EVOLUTION fie MINERAL EXPLORATION,

<http://leme.anu.edu.au> (Editors) 1 9 9 8 . Rehabilitation of and Implementation. John Wiley and Sons,

D E WAAL L., LARGE A .fieW A D E P. M .

Rivers: Principles London.

<www.ndsp.gov.au> <www.nlwra.gov.au> Hydrology and the River. Clarendon Press, Oxford. OLLIER C . D . 1 9 9 1 . Ancient Landforms. Belhaven Press, London. OLLIER C . D . fie PAIN C . F. 1 9 9 4 . Landscape evolution and tectonics in southeastern Australia. AGSO Journal of Australian Geology & Geophysics 15, 3 3 5 - 3 4 5 . PATON T . R . , HUMPHREYS G . S. fie MITCHELL P. B . 1 9 9 5 . Soils: A new global view. Allen fie Unwin, Sydney. PETTS G . E . fie AMOROS C. 1996. Fluvial Hydrosystems. Chapman and Hall, London. TAYLOR G. & PAIN C. (Editors) 1999. New Approaches to an Old Continent, Proceedings of Regolith '98, 3rd Australian Regolith Conference, Kalgoorlie, WA. Cooperative Research Centre for Landscape & Mineral Exploration, Perth. THOMS M. C.fieWALKER K . F. 1 9 9 2 . Channel changes related to lowlevel weirs on the River Murray, South Australia. In: Petts P. C. G. ed. Lowland Floodplain Rivers: Geomorphological Perspectives, pp. 2 3 5 - 2 4 9 . John Wiley & Sons, London. TWIDALE C . R. 1997a. The great age of some Australian landforms: examples of, and possible explanations for, landscape longevity. In: Widdowson M. ed. Palaeosurfaces: Recognition, Reconstruction and Palaeoenvironmental Interpretation, pp. 13-23. Geological Society of London Special Publication 120. TWIDALE C . R . 1997b. Persistent and ancient rivers—some Australian examples. Physical Geography 18, 291-317. TWIDALE C . R . 1 9 9 8 . Antiquity of landforms: an 'extremely unlikely' concept vindicated. Australian Journal of Earth Sciences 4 5 , 657-668. TWIDALE C . R . & CAMPBELL E . M . 1 9 9 3 . Australian Landforms: Structure, Process and Time. Gleneagles Publishing, Glen Osmond. NATIONAL DRYLAND SALINITY PROGRAM,

NATIONAL LAND AND WATER RESOURCES AUDIT, NEWSON M . 1 9 9 4 .

WILLIAMS M . , DUNKERLEY D . , D E DECKKER P., KERSHAW P. fie CHAPPELL

J. 1998. Quaternary Environments

(2nd edition). Arnold, London. The Great Ice Age:

W I L S O N R. C . L., D R U R Y S . A .fieCHAPMAN J. L. 2 0 0 0 .

Climate Change and Life. Routledge, London.

Heritage and others AUSTRALIAN Bureau OF STATISTICS 2000.

<www.abs.gov.au> R. 1997. Pattern & Process: Towards a Regional Approach to National Estate Assessment of Geodiversity. Technical Series No 2, Forest Task Force 2, Environment Australia, Canberra. JOYCE B. 1999. Different thinking. The Australian Geologist 112, 31-33. NATURAL HERITAGE TRUST, <www.nht.gov.au> SEDDON G. 1996. Thinking like a geologist: the culture of geology. Mawson Lecture 1996. Australian Journal of Earth Sciences 43, 487-495. W H I T E M. E. 1997. Listen—Our Land is Crying: Australia's Environment: Problems and Solutions. Kangaroo Press, Kenthurst. EBERHARD

THEME 1 ANCIENT AUSTRALIA AND ENVIRONMENTAL CHANGES

Geological Society of Australia Special Publication 21,3-11

CHAPTER 1—Quaternary climatic changes in Australia and their environmental effects M. A. J. WILLIAMS Department of Geographical and Environmental Studies, University of Adelaide, SA 5005, Australia. There is a strong possibility that fossil fuel burning and changing land use are contributing to changes in world climate. The future impact of these possible climatic changes on our river, lake and coastal ecosystems is hard to predict from current global atmospheric circulation models. The response of Australian landscapes to past climatic changes, particularly those of the well-dated late Quaternary, is a useful adjunct to numerical modelling. A critical appraisal of the evidence from several localities in central and southern Australia, including Lake Eyre and the Flinders Ranges, shows that landscape response to climate change is seldom straightforward. The interaction of different combinations of factors only indirectly linked to climate can lead to landscape responses that are complex and sometimes counterintuitive. Until we know a great deal more about feedback mechanisms, thresholds and the influence of biological processes, our attempts at predicting future landscape change will perforce remain very general. KEY WORDS: atmospheric carbon dioxide, climate change, environmental change, Lake Eyre, lakes, palaeogeography, photosynthesis, Quaternary, wind-blown dust.

INTRODUCTION Two aspects of human environmental impact are of increasing concern. One involves the accelerating impact of human activities upon our terrestrial and marine ecosystems. The other is the possibility that we may be contributing to changes in world climate through a combination of fossil fuel burning and changing land use, especially deforestation (Williams et al 1998). In 1990, carbon emissions from fossil fuel burning and land use change amounted to 6.0±0.5 Pg C and 1.6±1.0 Pg C, respectively (Houghton et al 1-992). Since many of the trace gases emitted are greenhouse gases which allow short-wave solar radiation to penetrate to the earth's surface but which trap a proportion of the outgoing long-wave terrestrial radiation, temperatures could increase in the lower atmosphere with as yet unknown effects on rainfall distribution. There has been an exponential increase in the atmospheric concentration of certain trace gases over the past two centuries, and especially since about 1950, when wholly synthetic gases such as the various chlorofluorocarbon gases began to be emitted (Figure 1.1, Table 1.1). This has prompted a resurgence of scientific research into the possible future impact of climatic change on global river, lake and coastal ecosystems, as well as on agricultural and other economic activities (Houghton et al 1990, 1996; Watson et al 1998). In seeking to establish the possible impact of climatic change on Australian ecosystems, several independent lines of enquiry have been pursued. One approach has involved detailed ecological studies of plant and animal associations in relation to present-day temperature, precipitation and evaporation gradients, including rigorous attempts to model distributions of particular species in relation to present-day climate. Another approach has focused

on the fossil record, most notably the pollen record of geologically recent vegetation changes. Complementary to this is the use of other natural archives of environmental change (Table 1.2), including the geomorphic and sedimentary evidence from rivers, lakes and dunes, to reconstruct the varying response of the Australian landscape to past changes in rainfall, runoff, plant cover and wind regime (Williams 1994, 2000). Unrelated to either of these approaches is the use of global atmospheric circulation models, some simple, others highly complex, which attempt to simulate the response of global climate to increases in atmospheric carbon dioxide concentration under existing boundary conditions. There are encouraging signs of increasing integration of these different approaches, and the active dialogue and exchange of data and ideas between practitioners in all three groups bodes well for our future understanding of the links between physical and biological processes in Australian landscapes and ecosystems. The aims of this chapter are to investigate how certain elements of the Australian landscape have adapted to Quaternary climatic changes, and to evaluate some of the difficulties involved in interpreting past climatic changes.

QUATERNARY ENVIRONMENTS IN AUSTRALIA The Australian landscape is akin to a palimpsest. Preserved through a combination of climatic aridity, tectonic stability and an almost total absence of widespread Pleistocene glacial erosion and deposition (Tasmania and the Snowy Mountains provide notable exceptions) are some of the oldest landscapes on earth (Williams 1991; Twidale 1998). Cheek by jowl with these ancient land surfaces, most of which have been modified by Late Cretaceous and Early

M. A. J. Williams

1850

1900

2000

YEAR

1900

1850

1900

2000

YEAR

> _Q

310

1850 CL CL

£a: iz

LU O

o o

1800

1850

1900

YEAR

1950

2000

1800

YEAR

1950

2000

Figure 1.1 The increasing concentration of carbon dioxide, methane, nitrous oxide and chlorofluorocarbon (CFC) gases over the last 250 years (after Houghton et al 1990). Table 1.1 Key greenhouse gases affected by human activities.

Atmospheric concentration Pre-industrial (1750-1800) Present day (1990) Current rate of increase/y (%) Atmospheric lifetime (y) Greenhouse forcing W m-2 (1765-1990)

Carbon dioxide

Methane

CFC 11

CFC 12

Nitrous oxide

ppmv 280 353 0.5 50-200 1.50

ppmv 0.8 1.72 0.9 10 0.56

pptv 0 280 4 65 0.06

pptv 0 484 4 130 0.14

ppbv 288 310 0.25 150 0.10

ppmv, parts per million by volume; ppbv, parts per billion (109) by volume; pptv, parts per trillion (10 12 ) by volume. Source: Houghton et al 1996. Cenozoic deep weathering and later erosion, are much younger landforms, many of which are no older than late Quaternary. Roughly two-fifths of the Australian arid zone is mantled in wind-blown sand dunes and sand plains, most of which were active at the height of the Last Glacial Maximum, some 25 000 to 15 000 years ago (25-15 ka). It takes an effort of the imagination to appreciate that relative to most of the past two million years of Quaternary time, the present geography of Australia is quite atypical of the past. The former coastlines may be reconstructed using the now well-dated evidence of late Quaternary sea-level fluctuations obtained from raised flights of coral terraces in the Huon Peninsula of New Guinea (Figure 1.2). The sealevel history is a measure of changes in global ice volume related to the waxing and waning of the mainly Northern Hemisphere ice caps (Williams et al 1998). The stable oxy-

gen isotopic composition of shallow and deep-sea foraminifers provides an independent measure of global ice volume change (Figure 1.2). Both the sea-level record and the marine isotope record reveal that over the last half million years roughly nine-tenths of each glacial-interglacial cycle involved a slow build-up to extreme glacial conditions, with progressive exposure of the continental shelf around Australia as ice caps expanded and global sea-levels fell to well over 100 m below present levels. Towards 18 ka, mainland Australia was up to 10°C colder, significantly drier and also much windier than today, with a land area nearly 25% greater and land bridge links to Tasmania and New Guinea (Figure 1.3a). As the ice caps melted and sea-levels rose, these land bridge connections were severed, and the climate became warmer and wetter (Figure 1.3b) (Williams et al 1998; Williams 2000).

Quaternary climatic change

Table 1.2 Characteristics of natural archives used in palaeoclimatic reconstruction. Archive Historical records Tree rings Lake sediments Ice cores Pollen Loess Ocean cores Corals Palaeosols Geomorphic features Sedimentary rocks

Best temporal resolution3

Temporal range (year)

Information derived

day/hour season/year 1-20 y year 100 y 100 y 1000 y year 100 y 100 y year

103 104 10 4 -10 6 105 105 106 107 104 105 107 107

T, H, B, V, M, L, S T, H, C a , B, V, M, S T, H, Cw, B, V, M T, H, Ca, B, V, M, S T, H, B H, B, M T, Cw, B, M CW>L T, H, C a , V T, H, V, L H, Ca, V, M, L

a Minimum sampling interval in most cases. T, temperature; H, humidity or precipitation; C, chemical composition of air (Ca), water (Cw) or soil (Cs); B, biomass and vegetation patterns; V, volcanic eruptions; M, geomagnetic field variations; L, sea-levels; S, solar activity. Source: Bradley 1990 table 1. Age (ka BP)

«o 5.0

0 -30

-60

2 00-120 -150 0

100

120

140

Age (ka BP)

160

180

200

220

240

260

Figure 1.2 Late Quaternary sea-level and associated oxygen isotope record derived from coral terraces, Huon Peninsula, Papua New Guinea, (a) Oxygen isotope record; (b) sea-level record (after Aharon & Chappell 1986).

Pollen evidence shows maximum expansion of tropical rainforest in northern Queensland between 8 - 6 and 3 ka (Figure 1.4) and geochemical, sedimentary and microfossil evidence shows highest lake levels in western Victoria between 7 and 5 ka (Figure 1.5). It is interesting to note that the Victorian lakes were becoming shallower and more saline after 5 ka, at precisely the time when the north Queensland rainforests were more extensive than today. One possible interpretation is that the winter westerlies in southeastern Australia had retreated further south several thousand years before the summer monsoon weakened in northeastern Australia—a speculative interpretation potentially testable using deep-sea cores from appropriate localities around Australia. The rainfall in the temperate southeast and southwest of the continent comes mainly from the westerly air masses that pass across the south of the continent most persistently in winter, when the Antarctic convergence is in its most northerly position. In northern Australia, short-term variations in the strength of the summer monsoon rains are modulated, among other factors, by El Nino Southern Oscillation events, and long-term changes reflect variations in the summer temperature regime over tropical oceans and land masses.

These long-term temperature changes are linked to variations in solar radiation, some of which are controlled by cyclical changes in the tilt of the earth's axis, the eccentricity of the earth's orbit around the sun, and the precession of the equinoxes. Williams et al (1998 chapter 5) provided a detailed discussion of the possible impact of these orbital perturbations upon the earth's climate. The short-term changes are of greater relevance to this discussion. Since the factors controlling the summer and winter air-mass systems are not identical, there is no necessary reason why they need to respond synchronously to any future global warming, since the thresholds and lag times will differ in both cases. At the present time, we do not know whether global warming will increase or decrease snow accumulation over Antarctica, so that it is hard to predict how the Antarctic convergence and hence the westerly air masses might respond in terms of their seasonal migrations north and south. Before we consider whether a knowledge of past environmental fluctuations is useful in understanding the present and, perhaps, future environments in Australia, it is instructive to consider how some other elements of the present Australian landscape have responded to late Quaternary climatic changes.

Late Quaternary palaeohydrology of Lake Eyre Perhaps the best-dated environmental history anywhere in Australia comes from the late Quaternary shorelines of Lake Eyre. Detailed stratigraphic work by Magee over the past decade (Magee 1998), supplemented by high-precision dating from a variety of independent chronometric techniques has provided a remarkably detailed record of lake level fluctuations over the last 150 thousand years (Figure 1.6). The lake-level histories of Lake Abhe in the Afar Rift of Ethiopia (Gasse 1975) and of Lake Chad on the south-central margin of the Sahara (Servant 1973; Servant & ServantVildary 1980), both show very high early Holocene lake levels, but Lake Eyre was not especially high at that time. This provides further grounds for wondering whether future changes in monsoon rainfall are likely to be synchronous in both hemispheres. In the case of Lake Eyre, the evidence

M. A. J. Williams

Uniform rainfall (> 1000mm) Summer rainfall (>400mm) Winter rainfall (>250mm) I •

Snow and ice Arid zone (<400mm rainfall in N <250mm in S.)

jggjjgj Active aeolian features Active dust plume

Cold ocean current Warm ocean current Cooler offshore

appears to suggest that a strong summer monsoon in the Northern Hemisphere does not necessarily connote an equally strong monsoon over northern Australia. Indeed, Lake Eyre attained its maximum levels towards 130-110 ka, with progressively lower levels thereafter, notably during 9 5 - 8 0 ka and 6 5 - 6 2 ka (Magee 1998). However, some words of caution are necessary here. To interpret climate from lake levels is not a simple matter. To then use the inferred climatic changes to 'explain' the lake fluctuations is to indulge in circular argument. Two aspects of this problem warrant discussion. First, fluctuations in the level of a given lake that may at any one time be fed in part from surface runoff, in part from groundwater influx, and in part from direct precipitation (with the relative proportions varying over time), are not of themselves direct indications of moisture source. Second, as Magee (1998) clearly notes, groundwater-controlled deflation during dry episodes in the Lake Eyre basin has long been a major process controlling lake-basin shape, volume and subsequent shoreline levels during ensuing lake transgressions. If the long-term export of lake-floor sediment through deflation exceeded the long-

Figure 1.3 Palaeogeography of Australia-New Guinea (a) during the Last Glacial Maximum (18 ka) and (b) during the early Holocene (9 ka) (after Williams 2000).

term replenishment of lake-floor sediment, successive lake levels would be lower, although the net input of water might not have diminished. A corollary to this might be that during the course of several glacial-interglacial cycles, with deflation especially active during the drier phases, the cumulative deepening of the lake floor might result in progressively lower lake shorelines. This would convey a possibly misleading impression that earlier interglacials were hydrologically more effective than later ones. This possibility certainly needs to be evaluated, even if only to be rejected. The Last Glacial Maximum was a time of total drying out of Lake Eyre and of massive deflation of the lake-floor sediments, causing a substantial lowering of the local baselevel. As noted above, this could explain the low elevation attained by the early Holocene lake. Alternatively, the early Holocene may not have been as wet as the previous wet phases at 130-110 ka, 9 5 - 8 0 ka and 6 5 - 6 2 ka. The Willandra lakes of western New South Wales (Figure 1.7) and Lake Keilambete and other maar lakes in western Victoria (Figure 1.5) were also dry during Last Glacial

Quaternary climatic change

Pollen site showing: © postglacial warming • wetter &/or warmer postglacial climate

| Uniform rainfall (>1000mm) Summer rainfall (>400mm) V//y\ •

Winter rainfall (>250mm) Snow and ice

Cold ocean current

| Arid zone

Warm ocean current

! Mostly vegetated aeolian features

Cooler offshore

Maximum and post-glacial times, prompting De Deckker (1986) to ask the highly pertinent question: 'What happened to the Australian aquatic biota 18 000 years ago?'

Impact of glacial aridity on the Australian aquatic biota A partial answer to this still unresolved conundrum, at least as far as the Australian tropics are concerned, lies in the presence of Pleistocene Lake Carpentaria. Lake Carpentaria was a huge freshwater lake located in the then exposed centre of what is now the Gulf of Carpentaria (Torgersen et al 1988). It was fed by runoff from the highlands of New Guinea during the the Last Glacial Maximum, when the sea was up to 135 m lower than today and the gulf was dry land. Other possible refuge areas include a lake in what is now Bass Strait, coastal wetlands on the then exposed continental shelf, and small lakes or wetlands in favoured localities such as the Flinders Ranges in South Australia, although even some of these may have become dry during the peak of glacial aridity (Cock et al 1999). A careful evaluation of how different components of the

Australia biota responded to post-glacial warming could be a useful guide to their future behaviour. The species or genera that adapted most rapidly to the change from cold dry glacial late Pleistocene climates to warm wet early Holocene climates (Figure 1.3) are likely to respond most effectively to any future global warming.

Impact of aeolian dust mantles on hillslope runoff and erosion A more subtle response to Quaternary climatic changes arises from the ambiguous response of certain desert hill slopes to climatic change. At intervals during the late Quaternary and earlier, extensive wind-blown dust mantles accumulated on desert hillslopes in every continent, including Africa, Asia and Australia. These aeolian dust plumes were both a consequence and a possible cause of climatic desiccation (Yung et al 1996). In Australia the best known dust mantles are located along the semiarid southern and eastern margins of the Australia arid zone (Walker & Costin 1971; Chartres et al 1988; Kiefert & McTainsh 1996). Many of these accumulations were quite thin and

M. A. J. Williams Lake Euramoo Quincan Bromfield Crater I Swamp

0 in

Lvnch's Crater

1 -

2 -

Maximum Rainforest Expansion

3 •

GQ S

4 5 -

6 7 Time of arrival of rainforest

8 "

Period of precipitation increases 10

11 1400 1600 1800 2000 2200 2400 2600

Present mean annual rainfall (mm) Figure 1.4 Vegetation changes in northeast Queensland deduced from pollen analysis of Holocene crater lake sediments (after De Deckker et al 1988).

easily incorporated into the soil mantle by eluviation and bioturbation. In certain localities, such as the Flinders Ranges in South Australia, a combination of rugged relief and a low cloud base provided an optimum combination of airflow divergence and humidity to facilitate dust entrapment, so that widespread and moderately thick mantles accumulated. The most recent phase of aeolian dust accretion spans the Last Glacial Maximum at 18 ka (McTainsh & Lynch 1996), and may have lasted from about 25 ka to 15 ka. Geomorphic and sedimentary evidence in Tunisia (Rognon & Coud£-Gaussen), Israel (Yair 1994), north-central India (Williams & Clarke 1984; 1995) and semiarid southeastern Australia (Williams et al 1991) suggests that the presence of these mantles may have altered hillslope runoff sufficiently to mask the more direct effects of late Quaternary climatic fluctuations upon the landscape. The aeolian mantles covered hitherto bare and relatively impermeable rock surfaces. Runoff was therefore reduced and infiltration increased. The result would be fewer flash floods and an increase in base flow. These inferred hydrological changes may well be the reason for the seemingly abrupt change from deposition of coarse sand and gravel by ephemeral or highly seasonal late Quaternary desert bed-load rivers to a depositional regime characterised by fine grained and widespread alluviation of the valley bottoms. In the Flinders Ranges of South Australia , the Matmata Hills of Tunisia and the Son and Belan valleys of India , these fine-grained valley fills were 15-25 m deep, and are now being entrenched by the present-day bed-load channels. The finegrained sediments in the resultant near-vertical sections remain as eloquent witnesses to this change in depositional regime.

P/E Estimates for 5 Victorian Maar lakes for the Holocene.

0 1

2 3 *c

Q. £

S >i

Keilambete

Gnotuk

Bullenmerri

West Basin

Medium and becoming low

Becoming Low - Medium

Low

Low?

Medium

Medium ?

East Basin

Low and changing

Low

Medium

Medium - Low

Low - Medium

Medium

High

Hlfjh

iSiiff:

High - Medium

High

Medium

Low

Medium - Low

Low

Lowest

Medium

Lowest

•

P|| Hip

High

50 kms Figure 1.5 Changes in Holocene precipitation/evaporation (P/E) ratios for five volcanic 'maar' lakes in western Victoria (after De Deckker et al 1988).

Quaternary climatic change

Impact on plants of changes in atmospheric carbon dioxide concentration

Figure 1.6 Fluctuations in the level of Lake Eyre during the last 150 thousand years (after Magee 1998).

There is one further aspect of the Last Glacial Maximum environment that is very belatedly starting to evoke interest. This concerns the possible impact on terrestrial ecosystems of fluctuations in the atmospheric concentration of carbon dioxide during glacial-interglacial cycles. Air bubbles trapped in successive ice layers of the Antarctic and Greenland ice sheets provide a remarkable record of past variations in the atmospheric content of methane (Chappellaz et al 1990) and carbon dioxide (Barnola et al. 1987) as well as of regional temperature fluctuations (Jouzel et al 1987). The record from the Antarctic Vostok ice cores now extends back to 400 ka, and confirms the existing published record from the last 160 ka which shows that glacial maxima were times of minimum levels of atmospheric methane and carbon dioxide, with important repercussions for the terrestrial biota (Figure 1.8). During times of low atmospheric carbon dioxide concentration, plants following the C4 photosynthetic pathway, notably grasses, would be at an advantage relative to most trees, which almost all follow the C3 pathway (Williams et al 1998 pp. 238-239; Farquuar 1997; StreetPerrott et al 1997). The Last Glacial Maximum was a time of prolonged drought, extreme cold, greater windiness and much reduced carbon dioxide levels, all of which would have had an adverse impact upon Australian trees. If we include the impact of fire, insect attack, less-efficient water use and drastically curtailed rates of photosynthesis, the adverse impact on trees could have been enormous. For instance, it is highly likely that the River Red Gum (Eucalyptus camalduensis) forests that are a feature of the riparian vegetation of the Murray-Darling basin today could have been virtually wiped out towards 18 ka. One predictable consequence of the demise of the deeprooted eucalyptus trees would have been a rise in the water-

Lacustral phases (wet) Pelletal clay dunes (dry)

i— 20 10 YEARS x 1000 (calibrated)

—j— 40

—i 50

Figure 1.7 Fluctuations in the levels of the Willandra lakes during the past 60 thousand years (after Bowler 1998).

M. A. J. Williams

CONCLUSIONS

o -

O z< 0

a: D -2 I<

a: LU CL

- -4 sLU

Some care is needed when interpreting the record of past environmental changes. The response of different ecosystems to past changes in climate was often complex and was not always in phase in different parts of the continent. There is therefore no compelling reason why the response to possible future climatic changes need be synchronous or similar in different parts of Australia. Some regions may become wetter as others suffer increasing drought. At this stage, neither our global atmospheric circulation models nor our knowledge of past environmental changes are sufficiently good to allow other than very general attempts at future prediction. One thing is sure: some of the responses of the Australian landscape may be counterintuitive and surprising. The precautionary principle is one we should follow, and our land management strategies will need to be flexible and even opportunistic if we are to reap the benefits and minimise the adverse impacts of future climatic change on the Australian environment.

Note added in proof

T—i—i—i—i 1 r~ 0 50 100 AGE (thousand years before present)

A recent theodolite survey of the fine-grain valley-fill deposits in the central Flinders Ranges indicates that they are wetland deposits (Williams et al. in press) and are not shallow lake deposits as first thought (Cock et al 1999) 150

Figure 1.8 Fluctuations in atmospheric carbon dioxide and methane obtained from analysis of air bubbles trapped in Antarctic ice at Vostok closely parallel local temperature fluctuations inferred from deuterium fluctuations over the past 160 thousand years (after Houghton et al. 1990, adapted from Jouzel et al. 1987, Barnola et al. 1987 and Chappellaz et al. 1990).

tables, with any salts present in the underlying rocks and sediments and in the groundwater being brought to the surface. Hence the paradox of high groundwater tables during times of aridity (Williams et al. 1991), reflecting the absence of the trees that act as natural groundwater pumps, maintaining the water-table well below the land surface. Any salt brought to the surface at this time would have added to the stresses noted above, leading to further loss of trees. The consequence of these cumulative pressures would have been an expansion of grassland at the expense of woodland and forest throughout Australia, a trend also observed on tropical mountain ecosystems in Africa during the Last Glacial Maximum (Street-Perrott et al 1997). The impacts of the destruction of woodland and forest during glacial maximum times in Australia were analogous to those of tree clearance in Australia during the past two centuries. It is only in the last fifteen or so years that we have begun to appreciate that many of the salinity problems in the agricultural regions of southern Australia stem from widespread tree clearance following European settlement. Since the slow rise in saline groundwater tables reflects events which occurred nearly a century ago, it is inevitable that any attempts to manage salinity through tree planting will need to take a long-term perspective.

REFERENCES P. & CHAPPELL J. 1986. Oxygen isotopes, sea level changes and the temperature history of a coral reef environment in New Guinea over the last 105 years. Palaeogeography, Palaeoclimatology, Palaeoecology 56, 337-379.

AHARON

BARNOLA J. M . , RAYNAUD D . , KOROTKEVICH Y. S . & LORIUS C .

1987.

Vostok ice core provides 160,000 year record of atmospheric C0 2 . Nature 3 2 9 , 4 0 8 - 4 1 4 . BOWLER J. M. 1998. Willandra Lakes revisited: environmental framework for human occupation. Archaeology in Oceania 33, 120-155. BRADLEY R. S. 1990. Global Changes of the Past. UCAR/Office for Interdisciplinary Earth Studies, Boulder, CO. CHAPPELLAZ J., BARNOLA J. M . , RAYNAUD D . , KOROTKEVICH Y. S . & LORIUS

C. 1990. Ice-core record of atmospheric methane over the past 160,000 years. Nature 345, 127-131. CHARTRES C . J., CHIVAS A. R. & WALKER P. H. 1 9 8 8 . The effects of aeo-

lian accessions on soil development on granitic rocks in southeastern Australia. II. Oxygen-isotope, mineralogical and geochemical evidence for aeolian deposition. Australian Journal of Soil Research 26, 17-31. COCK B. J., WILLIAMS M . A . J. & ADAMSON D . A . 1 9 9 9 . Pleistocene Lake Brachina: a preliminary stratigraphy and chronology of lacustrine sediments from the central Flinders Ranges, South Australia. Australian Journal of Earth Sciences 46, 6 1 - 6 9 . D E DECKKER P. 1986. What happened to the Australian aquatic biota 18 000 years ago? In: De Deckker P. & Williams W. D. eds. Limnology in Australia, pp. 487-496. CSIRO, Melbourne and Dr W. Junk Publishers, Dordrecht. D E DECKKER P., KERSHAW A . P. & WILLIAMS M . A . J. 1 9 8 8 . Past environmental analogues. In: Pearman G. I. ed. Greenhouse: Planning for Climate Change, pp. 4 7 3 - 4 8 8 . E.J. Brill, Leiden and CSIRO, Melbourne. FARQUUAR G. D. 1997. Carbon dioxide and vegetation. Science 2 7 8 , 1411-1412. GASSE F. 1 9 7 5 . Involution des lacs de I'Afar central (tthiopie et TFAI) du Plio-Pleistocene d I'Actuel. DSc thesis, University of Paris VI (unpubl.). H O U G H T O N J. T., CALLENDER B. A., & VARNEY S. K. (Editors) 1992. Climate Change 1992: The Supplementary Report to the IPCC

Quaternary climatic change Scientific Assessment Cambridge University Press, Cambridge. Climate Change: The IPCC Scientific Assessment. Cambridge University Press, Cambridge.

HOUGHTON J . T., JENKINS G . J . & EPHRAUMS J . J . (Editors) 1 9 9 0 .

HOUGHTON J . T . , MEIRO FILHO L. G . , CALLANDER B . A . HARRIS

N.,

WALKER P. H . & COSTIN A. B . 1 9 7 1 .

Atmospheric dust accession in south-eastern Australia. Australian Journal of Soil Research 9, 1-5.

WATSON R . T., ZINYOWERA M. C., Moss R . H . & DOKKEN D . J. (Editors)

1998. The Regional Impacts of Climate Change: An Assessment of Vulnerability. Cambridge University Press, Cambridge.

(Editors) 1 9 9 6 . Climate Change 1995: The Science of Climate Change. Cambridge University Press, Cambridge.

WILLIAMS M. A. J. 1991. Evolution of the landscape. In: Haynes C. D.,

JOUZEL J . , LORIUS C . , PETIT J - R . , GENTHON C . , BARKOV N . I . , KOTLYAKOV

WILLIAMS M. A. J. 1994. Some implications of past climatic changes in

KATTENBERG A .

& MASKELL K .

V. M . & PETROV V. M . 1 9 8 7 . Vostok ice core: a continuous isotope

temperature record over the last climatic cycle (160,000 years).

Nature

329,403-408.

KIEFERT L. & MCTAINSH G . H . 1 9 9 6 . Oxygen isotope abundance in the

quartz fraction of aeolian dust: implications for soil and ocean sediment formation in the Australasian region. Australian Journal of Soil Research 34, 4 6 7 - 4 7 3 . MAGEE J . M . 1 9 9 8 . Late Quaternary environments and palaeohydrology of Lake Eyre, arid central Australia. PhD thesis, Australian National University, Canberra (unpubl.). MCTAINSH G. H. & LYNCH A. W . 1 9 9 6 . Quantitative estimates of the effect of climate change on dust storm activity in Australia during the Last Glacial Maximum. Geomorphology 1 7 , 2 6 3 - 2 7 1 . ROGNON P. & COUD£-GAUSSEN G . 1 9 9 6 . Paleoclimates off northwest Africa (28°-35°) about 18, 000 yr B.P. based on continental eolian deposits. Quaternary Research 46, 118-126. SERVANT M. 1973 Sequences continentales et variations climatiques: Evolution du bassin du Tchad au Cenozoique superieur. DSc thesis, University of Paris VI (unpubl.). SERVANT M. & SERVANT-VILDARY S . 1 9 8 0 . L'environnement quaternaire du bassin du Tchad. In: Williams M. A. J. & Faure H. eds, The Sahara and the Nile, pp. 1 3 3 - 1 6 2 . Balkema, Rotterdam. STREET-PERROTT F. A., HUANG Y., PERROTT R. A. ETAL. 1997. Impact of lower atmospheric carbon dioxide on tropical mountain ecosystems. Science 278, 1422-1426. TORGERSEN T., LULY J . , D E DECKKER P., SEARLE D . E . , CHIVAS A . R . & ULLMAN W. J . 1 9 8 8 . Late Quaternary environments of the Carpentaria Basin, Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 6 7 , 2 4 5 - 2 6 1 . TWIDALE C. R. 1998. Antiquity of landforms: an 'extremely unlikely' concept vindicated. Australian Journal of Earth Sciences 45, 657-668.

Ridpath M. G. & Williams M. A. J. eds. Monsoonal Australia, pp. 5-17. Balkema, Rotterdam.

Australia. Transactions of the Royal Society of South Australia 118, 17-25.

WILLIAMS M. A. J. 2000. Quaternary Australia: extremes in the Last

Glacial-Interglacial Cycle. In: Veevers J. J., ed. Billion-year Earth History of Australia and Neighbours in Gondwanaland. pp 55-59. GEMOC Press, Sydney. WILLIAMS M. A. J. & CLARKE M. F. 1984. Late Quaternary environments in north-central India. Nature 308, 633-635. WILLIAMS M. A. J. & CLARKE M. F. 1995. Quaternary geology and prehistoric environments in the Son and Belan Valleys, north central India. Memoirs of the Geological Society of India 32, 282-308. WILLIAMS M . A . J . , D E DECKKER P., ADAMSON D . A . & TALBOT M . R . 1 9 9 1 .

Episodic fluviatile, lacustrine and aeolian sedimentation in a late Quaternary desert margin system, central western New South Wales. In: Williams M. A. J., De Deckker P. & Kershaw A. P. eds. The Cainozoic in Australia: A Re-appraisal of the evidence, pp. 258-287. Geological Society of Australia Special Publication 18.

WILLIAMS M . , DUNKERLEY D . , D E DECKKER P., KERSHAW P. & CHAPPELL

J. 1998. Quaternary Environments (2nd edition). Arnold, London.

WILLIAMS M . ,

PRESCOTT J . R . , CHAPPELL J . , ADAMSON D . , COCK B . ,

WALKER K. & GELL P. The enigma of a late Pleistocene wetland in

the Flinders Ranges, South Australia. Quaternary International (in press). YAIR A . 1 9 9 4 . The ambiguous impact of climate change at a desert fringe: northern Negev, Israel. In: Millington A. C. & Pye K. eds. Environmental Change in Drylands, pp. 1 9 9 - 2 2 7 . Wiley, Chichester. YUNG Y . L., LEE T., WANG C - H & SHIEH Y - T . 1 9 9 6 . Dust: a diagnostic of the hydrologic cycle during the Last Glacial Maximum. Science 271,

962-963.

Received 12 February 1999; accepted 3 May 2000

Geological Society of Australia Special Publication 21, 13-25

CHAPTER 2—Regolith: its history and environmental importance with particular reference to some engineering examples 7

Cooperative Research Centre for Landscape Evolution and Mineral Exploration, University of Canberra, ACT 2601 Australia. Sinclair Knight Merz, PO Box 164, St Leonards, NSW 1590, Australia. The regolith provides human's interface with planet Earth. It is that part of the solid Earth that interacts with the hydrosphere, atmosphere and biosphere; it is that part between fresh rock and fresh air. It contains weathered rock, sediments, volcanic ash, water, air and biota. Regolith drapes most of our landscapes and creates many of its landforms. It is from the regolith we extract most of what we rely on for life and its quality. Plants we use to sustain our life grow in it, and in so doing extract nutrients from it. Much of the water we drink comes from it or flows over and through it at some point in its journey to our taps or tanks. During this journey it extracts elements from the Earth that affect its purity. The regolith provides many of the materials necessary for life as we know it, for example aluminium, titanium for our paint base, glass sand, building materials and gold, to name but a few. The regolith acts as a buffer for C0 2 in our atmosphere, consuming very large amounts during the rock-weathering process. It is critical to our continued existence on this planet. This chapter illustrates many of these important functions of the regolith and goes further in providing some details of how it affects some aspects of engineering construction. It discusses aspects of land drainage and groundwater at Mt Gambier South Australia, both important issues to those who live in the district. In the Stuart Range, South Australia, people live in the regolith, draw their water from it and use it for construction materials, all of which present a special set of regolith-related problems. On Australia's Riverine Plain another set of problems present themselves including how to build roads in an area of expansive clay minerals. Regolith is the only interface we have with the planet, so by understanding how it formed and why it is important to us we can manage it sustainably for the continued benefit of humanity. KEY WORDS: construction materials, duricrust, Eastern Highlands, groundwater, problem soils, regolith.

INTRODUCTION The regolith is that part of the Earth's crust which results from the interaction of hydrological, chemical, physical, atmospheric and biological processes with the lithosphere. These interactions produce a great variety of materials that form the basis for most human activity on Earth. It is on the regolith that we generally live, work, grow our food and draw our water, and in which we are ultimately buried. The regolith provides many of the resources needed for human existence including building foundations, extracting materials for houses, roads and other necessary infrastructure. It provides much of the nutrients and water necessary for plant growth and food production. Many of the mineral commodities used in our everyday life come from the regolith including clay, sand, gravel, aluminium ore, titania (for paint), iron ore and other minerals. It plays a significant role in controlling vegetation ecology and faunal habitats. The regolith is our link between life, as we know it, and the Earth. Thus it is important for us to have some knowledge of the regolith and how it interacts with our being.

What is the regolith Regolith is defined by Bates and Jackson (1980) as 'A general term for the layer or mantle of fragmental or unconsol-

idated rock material, whether residual or transported and of highly varied character, that nearly everywhere forms the surface of the land and overlies or covers bedrock. It includes debris of all kinds, volcanic ash, glacial drift, alluvium, aeolian deposits, vegetal accumulations and soil. The term was originated by Merrill (1897)'. The etymology of the word is from the Greek rhegos = blanket and lithos = stone. Put more simply, regolith is everything between fresh rock and fresh air. The concept of regolith is generally restricted to terrestrial environments although obviously as sea-levels change, parts of what was terrestrial become drowned and visa versa so regolith may occur beneath the seas. The regolith comprises weathered rock, some fresh rock remnants, hard surface crusts (duricrusts), soils (in both the soil science and engineering senses), unconsolidated sediments, and chemical products of weathering, biological materials, water and its solutes. The processes of weathering and erosion are important in forming the regolith and determining its distribution within the landscape. Any rock materials exposed at the Earth's surface are subject to physical and chemical breakdown, forming particles of the original rock or new products (mainly clays and oxides of Fe and Al) respectively. These products are then distributed within the landscape by the processes of erosion and deposition.

G. Taylor and G. H. McNally

As a simple example of this, consider a hill slope and valley formed on granite (Figure 2.1). The granite shows an irregular weathering front between fresh rock and the granite saprolite with corestones. The latter represents the partial chemical decomposition of the granite to quartz and clay mineral, perhaps with some partially weathered feldspar also present, while preserving the original rock fabric. The corestones are as yet unweathered remnants of the granite. Above this, near the hilltop, is granite saprolite with none of the original minerals except quartz retained and there are no corestones (i.e. all the original rock has been chemically altered), but the original rock fabrics are preserved. Above this the original rock fabric is progressively lost and increasing pedogenetic organisation becomes apparent in the layer called soil. In the erosionally 'stable' area at the hill crest the soil and saprolite are thicker, but as the ground slopes downwards erosion becomes the dominant process and soils and saprolite are eroded leaving corestones exposed. Below this zone, as slopes decrease, deposition becomes the predominant process and the material eroded from upslope is deposited. This transported material also has soils developed on top of it, and these destroy the depositional fabrics and replace them with pedogenetic features. The sediment part of the regolith is comprised of deposits from colluvial, glacial, aeolian and alluvial processes operating on the lower slopes and in the valley bottom. These sedimentary materials generally consist of minerals that resist weathering or of new minerals formed during weathering. They may contain features indicative of their depositional mode (bedding, structures and textures), or they may be pedogenised to varying degrees. The rates of sedimentation and pedogenesis vary across the depositional surface enabling sedimentary features to be preserved in areas of more rapid sedimentation and soils to form where depositional rates are slower. The sedimentary part of the regolith is thus a mixture of recognisable sedimentary deposits and soils with varying degrees of development. The soil zone of the regolith is that part of the regolith occupied by soils with their constituent horizons. It is not the intent here to describe soils in detail, as many good texts exist on Australian soils (Isbell 1996; Paton et al.1996; Division of Soils CSIRO 1983).

SHORT HISTORY OF AUSTRALIA'S LANDSCAPES AND REGOLITH Australia's landscapes can be basically fall into three categories (Figure 2.2): (i) Precambrian shields or cratons; (ii) the Eastern Highlands; and (iii) sedimentary basins of both consolidated and unconsolidated deposits. It also follows that the age of the regolith and therefore its development and character is related to the time since the present day that the rocks or sediments were first exposed to subaerial conditions. The ages of exposure for Australia are shown in Figure 2.2.

Shields The shield areas are generally areas of relatively low relief landscapes developed on Precambrian rocks. Because of

Figure 2 . 1 A typical regolith profile developed on a granitic bedrock (see text for explanation).

the antiquity of the rocks, the relative tectonic stability and their duration of exposure, many of the landscapes are also very old. Stewart et al (1986) have identified possible extant Cambrian landscapes in the Davenport Ranges, Northern Territory, and it is clear that from the work of Bird and Chivas (1988) that kaolinite in many weathering profiles in the Yilgarn dates from the Permian or Mesozoic. This does not mean that all the landscape features of the shields are old, but there is an ancient skeleton to them. The largest shield area is the Yilgarn Plateau that is dominated by granitic and greenstone terrains with deep weathering and ferricreted and bauxitised surface crusts being relatively common. All these weathering features post-date the Permian, when glaciers swept the region. As well as cleaning up the surface the glaciers also carved out drainage channels still occupied by lakes (Oilier 1977), but little depositional evidence. Elsewhere on the Australian shields deposits formed by the Permian glaciers still form part of the regolith. Regolithic deposits of till and erratics at Rudall River in the northeastern Hamersley testify to the former extent of glacial activity and to the great age of much of Australia's regolith. Shields like the Kimberley and Arnhem Land Plateaux are dominated by quartzose sandstone outcrop. As a result of long exposure these sandstones have been karstified, forming spectacular landscapes like the Bungle Bungle Range of the East Kimberley and the Ruined Castle in eastern Arnhem Land. Groundwater movement through these rocks rather than metamorphism has silicified many of the Proterozoic quartzites of northern Australia. Silica leached out of the joints and framework has simply been deposited between the framework grains in the former sandstone. The Broken Hill Block forms the Barrier Ranges, a relatively high relief, fault-controlled shield landscape. Many of the upthrown areas are dominated by erosional processes and do not contain much regolith but are dominated by bedrock terrain. The downthrown areas are depositional regions where regolith is forming now and older regolith is preserved. In the upthrown areas remnants of earlier regolith that are not easily eroded, such as silcrete and ferricrete, are preserved as well. Unlike many other shield regions it has been recently demonstrated that the faulting in the Barrier Ranges is still relatively active over a wide area (Hill et al 1997). Thin deposits of younger regolith, probably derived during the Quaternary, cover large areas of the shield regions. These veneers consist of alluvial, saline and clay-rich

Importance of the regolith

D(M)-S

Legend T [; ; ; ; ;! Tertiary

C-D(U) KXH Carboniferous and Upper Devonian

K V/'/A Cretaceous

D(M)-S

Mid. Devonian and Silurian

J-R 1:: :| Jurassic and Triassic

O-e

Ordovician and Cambrian

P H i Permian

Precambrian

(L) Lower e.g. p

• •

D(M)"S

{

K(L)

0 D(M)-S

D(M)-S

(U) Upper

All of these areas have been continuously exposed to weathering and denudation since theiPe Permian

K(L) T J o

All Pre-Permian surfaces south and west of this line were affected by continental glaciation in

% the Permian and then re-exposea to weathering.

N.B. Quaternary cover and volcanics not shown Figure 2.2 Maximum age of exposure for various regions of Australia (from Beckmann 1983)

lacustrine, aeolian and colluvial sediments. Although derived from earlier regolith, they are distinct from the older deeply weathered regolith in that they were formed under very different climatic regimes (Williams 2001) and tend to be alkaline unlike the earlier highly acid older regolith. Regolith carbonates, particularly calcrete and dolocrete are widespread in lower topographic positions. Many of the salt lakes contain exotic clay minerals formed under alkaline conditions caused by the relatively arid Quaternary climates. Eastern Highlands The Eastern Highlands is an area of relatively high relief stretching from Tasmania to Cape York. The highlands are comprised predominantly of Palaeozoic sedimentary and volcanic rocks intruded by Palaeozoic granitic rocks.

Mesozoic basic to acid intrusive rocks and Late Cretaceous to Cenozoic volcanics cover substantial areas of the terrain. Tectonic activity has been widespread along the highlands, although the age of their original uplift remains obscure, but many hypotheses exist (Wellman 1987; Oilier 1982; Bishop 1988). Probably uplift was diachronous along the highlands. Taylor et al (1990) have clear evidence for >500 m relief on the eastern margin in southeastern New South Wales at about 60 Ma, but other definitive evidence is lacking. The fault control of many landscapes and regolith is well illustrated by Lake George (Abell 1985), Lake Bunyan (Taylor & Walker 1986a, b) and Cape York (Pain & Oilier 1992). Considerable evidence is accumulating to demonstrate the antiquity of many landscapes and regolith materials in the highland belt. Taylor et al (1985, 1990) and Bishop (1980) clearly show that many of the highland drainage net-

G. Taylor and G. H. McNally present. In some places (Taylor et al 1990) these basalt flows cover and preserve pre-basaltic deep weathering of Palaeozoic rocks. Many areas of the highlands are also covered by duricrusts including silcrete, ferricrete and bauxite related to basalt weathering, and in areas of lower rainfall, calcretes. Clearly some landscapes of the highland are every bit as old as those on the older rocks of the shield, but, because of their greater relative relief large areas are also covered by thin and younger regolith. The alternating drier and wetter climates of the Quaternary have lead to extensive stripping of regolith in areas of high to moderate relief and deposition on the lower hill slopes and in valley bottoms. Added to this is aeolian dust (parna) blown east during drier phases and incorporated into soils, lakes and on hill slopes of the highlands.

Sedimentary basins

Figure 2.3 Structural control of drainage and depocentres in the northeastern Lake Eyre catchment.

works in southern New South Wales have remained relatively fixed in position since the Early Cenozoic and probably from the Late Mesozoic. These rocks have been deeply weathered like those of the shield areas with dates on kaolin formation (Bird & Chivas 1988) ranging from Permian to Late Cenozoic. Palaeomagnetic dates (Ruxton & Taylor 1982) and more recently by Pillans et al (1999) record dates from the Early Palaeogene and Carboniferous respectively. Basalts preserve the terrain over which they flowed, and studies by Taylor et al (1985), Bishop (1988), Pain (1983) and many others demonstrate landscape dating from the earliest Palaeogene were similar to those in the highland at

Over the Mesozoic and Cenozoic basins that cover almost half the continent (Figure 2.2) landscapes are relatively flat, reflecting their depositional nature. Active tectonism in the basins has uplifted them since deposition ceased and continued faulting and folding also has a significant impact on contemporary landscapes. Some examples include: the Murray River has been diverted by the Cadell Fault (Hills 1975); silcretes in the Eromanga Basin are deformed by gentle folding (Senior et al 1978); and the Cooper Creek drainage system is essentially controlled by contemporary tectonism (Figure 2.3), as are other streams of southwest Queensland (Senior et al 1978). This gentle folding has allowed the development of extensive low plateaux, mesas and plains in the nearly flat-lying sediments of the Eromanga and Canning Basins. It has also caused strong deformation of Mesozoic sedimentary rocks along the margins of the Broken Hill and Wonnaminta Blocks in northwestern New South Wales (Hill et al 1997). Uplift during the Late Neogene has lifted the Eucla Basin above sea-level causing karst landscapes to develop.

Old valley

Breakaway Land surface Ferricrete Mottled zone vein Pallid zone

Fresh Bedrock

Figure 2 . 4 The Walther profile (from Walther 1915).

Importance of the regolith The Mesozoic sedimentary rocks of the Eromanga, Canning and Officer Basins are deeply weathered (100 m+) with typical Walther profiles (Figure 2.4) developed over wide areas. These profiles are typically capped by either ferricrete and/or silcrete, which in many places has been degraded to form extensive gibber plains. Gibber often appears to be a continuous sheet of pebbles and boulders, however it is often underlain by thick (5 m+) expansive desert loam over gibber and weathered rock. Gibber plains are often gilgaied. Gilgai develops as a result of the smectite content of the loams. On wetting and drying the smectite expands and contracts causing the formation of a microtopography with up to 2 m relief. It is common to find a duricrust developed above the saprolite zone. A duricrust is simply a hardened surface or near-surface layer where the regolith has been cemented, replaced and/or altered chemically to form sheets, nodules and pisoliths. The most common forms of duricrust are cemented Fe-oxides (ferricrete), Al-oxyhydroxides (bauxite), CaC0 3 (calcrete), or Si0 2 (silcrete). Less commonly C a S 0 4 . 2 H 2 0 (gypcrete), CaMg(C0 3 ) 2 , (dolocrete) or Mnoxides (manganocrete) form the bulk of the hardened layer. As their name implies, duricrusts are relatively resistant to weathering and erosion and often occur on what are presently topographic highs. It is not the purpose of this chapter to go into the formation of these crusts, but their origins and landscape significance are controversial (Oilier & Pain 1996). The flanks of the plateaux and mesas are draped by extensive colluvial and fan deposits composed of duricrusts and weathered sediments. Valley bottoms are extensively, but often thinly alluviated. Longitudinal sand dunes are common throughout most of the basins excepting the eastern and northern Eromanga Basin. Lunettes occur downwind of most lakes and palaeolakes across the basins, as well as on the southern shields and southern highland lakes. One of the major controls of the nature of the regolith is climate, as the amount of available water affects the amount of chemical weathering that can occur and the type of weathering products formed. Between the Carboniferous and the Late Neogene Australian climates were predominantly wet with significant glacial activity from the Late Carboniferous to the Early Permian and drier periods in the Early Triassic, Early Jurassic and mid-Cretaceous. This allowed almost continuous chemical weathering to occur through the Late Palaeozoic, Mesozoic and much of the Cenozoic, despite cool to cold temperatures during this time (Frakes et al 1992). Due to this extended period of weathering most of the regolith formed was highly leached, dominated by Fe oxides, Al-oxihydroxides, kaolin and quartz. With the abundant vegetation that existed through much of this time and the highly leached regolith, soils were very acid and nutrient poor as they are on such regolith still exposed today (e.g. the Yilgarn and southwestern Queensland). Increasing aridity from the Late Neogene and the oscillation of climates from warm wet to cool dry through the Quaternary (Williams 2001) diminished leaching and alkali and alkali-earth elements were retained in the regolith. This produced much more alkaline and nutrient-rich soils, but also lead to the retention of sodic salts in the regolith which

have periodically caused widespread salinisation during the Quaternary and are exacerbated at present due to human influences. Although Quaternary glacial climates had a major effect on southern and highland Australia, actual glacial activity was restricted to a small area around Mt Koscuiszko and to the highlands of Tasmania during the last glacial maximum about 18 000 years ago. They left some glacially derived regolith in their wake, but more important are the extensive periglacial deposits in the Eastern Highlands and Tasmania. Periglacial deposits are the result of freezing and melting of regolithic water and these consist of annual accumulations of rock debris accumulated on slopes as lobes and fans, and in some places (e.g. Snowy Mountains and Canberra) as boulder streams.

EXAMPLES OF THE ENVIRONMENTAL SIGNIFICANCE OF REGOLITH Australia's regolith is environmentally important: it is thick; it contains the major aquifers in the country; it is composed in places of saline, dispersive and expansive clays; and it controls the distribution of many biotic assemblages. The nature of the regolith and how humans use it has major effects on agriculture, water resources and biodiversity. Soil salinity is of major concern in southeastern and southwestern Australia. Much of the salt causing the problems is stored in the regolith and in its groundwater. As water-tables rise, salts are brought to the surface; this results in soil degradation and erosion, and takes good agricultural land out of production. Many soils contain exchangeable sodium so that when the soils are exposed to water they disperse easily, erode and harden on drying. Expansive clays cause agricultural problems resulting from gilgai formation and engineering problems relating to ground heaving and shrinkage. For example, the optic fibre cable between Adelaide and Melbourne where it crosses the lower Murray Basin has to be deeply buried and set in sand to avoid breakage. Similarly cracking clay soils developed on basalts of the Monaro in southeastern New South Wales control the distribution of Eucalyptus pauciflora in this region. The trees only survive where soils do not dry and crack, because here their taproots are not broken during early growth stages. This is the reason so much of Australia's black and grey soil plains tend to be treeless and steppe-like. Gilgaied soils in much of semiarid Australia support chenopod vegetation on the swells because of differing water availability in the depressions and swells. In the Broken Hill region pearl bluebush (Maureana sedifolia) only grows where pedogenetic calcrete occurs within 1.5 m of the surface. Elsewhere the shrubby vegetation is dominated by black bluebush and saltbush (M. pyramidata and Atriplex vesicaria). Palaeochannels in the regolith, but unrecognisable at the surface, are outlined at the surface by stands of sheoak (Casuarina sp.) that depend on groundwater stored in the channels for growth. A similar phenomenon occurs on the grey clay soil plains where native cypress (Callitris sp.) grows almost exclusively on sandy palaeochannels criss-crossing the plains. These channels are nutrient-poor compared to the grey soils and hold water more easily accessible to the trees. It is also interesting to

G. Taylor a n d G. H. McNally

Effect of Climate Models

Time (my)

(High Ca, Mg)

Figure 2.5 The dividing line between winter and summer rain that marks the boundary between various ecological associations on a continental scale; the boundary also marks significant differences in regolith (from Hill et al. 1999)

note that many of these palaeochannels are also the plague locust's (Gastrimargus musicus) prime hatching areas in this part of Australia, because the sandy soils are easier for the locust to lay its eggs, than in the surrounding harder grey clay soils. The antiquity of the regolith and the resultant high Fe oxide contents in many places causes problems with the use of phosphate fertilisers. The Fe oxides adsorb the phosphorous making it unavailable to plants. Many other regolith minerals interact with elements in or added to the environment. Clay minerals adsorb many heavy metals, pesticides and hold them in the environment until chemical changes cause their release. An example of this is that during the early stages of ultramafic rock weathering, smectite forms from pyroxenes and this may include up to 30% Ni compared to only 0.5% in the parent rock. The Ni concentration in the saprolite in such situations runs about 2%, a four-fold increase. Similarly at Weipa the smectite from the Rolling Downs Group contains 8 0 ppm Ni, but the kaolinitic weathering zone only 2 ppm, and the overlying bauxite 25 ppm. An interesting example of how the regolith affects fauna is the distribution of pink-tailed legless lizards around Canberra. These lizards live under small-sized rocks that occur only on mid-slopes of hills on acid volcanic bedrock (Hill 1995). They do not live under larger rocks higher up slopes or smaller ones lower in the landscape. This distribution is the result of the thermal properties of the mid-sized slabs of volcanic rock suiting the requirements of the lizards. The surface rock size in the landscape is related to colluvial processes and physical and chemical weathering, which break up the originally larger outcrops. Another example of the relationship between regolith and animals is the case of Buramus (mountain pygmy possum). It lives only in the Snowy Mountains in areas where the regolith is dominated by angular blocky boulder-field formed by periglacial activity and as neve deposits. The

Figure 2.6 Plot of C 0 2 vs time. RC0 2 is the mass of C 0 2 in the atmosphere at some time divided by that today. The parameters T and Ws represent the sensitivity of global mean temperature to atmospheric C 0 2 level and solar radiation respectively based on empirical equations fitted to the results of general circulation models (from Berner 1995).

close relationship between Buramus and these deposits has enabled a close approximation of the last glacial maximum extent of such conditions based on the distribution of its fossils (Ride et al 1989). There is a general relationship emerging across Australia between aspects of climate, regolith and biota. Figure 2.5 shows the distribution of pedogenetic calcrete and soil carbonates across Australia. The line shown on the map divides areas where pedogenetic calcrete is common from those where it is less so or absent. This line also divides areas of: (i) high groundwater salinity south of it from generally low salinity to the north; (ii) bluebush occurrences south of it; and (iii) summer rains north and winter rains south.

Regolith and greenhouse There is a significant relationship between silicate mineral weathering and atmospheric carbon dioxide levels. We know there are connections between C 0 2 levels and global climate so there is a feedback between weathering and global climate. If global climate is warming as a result of CO. increases in the atmosphere then there will be an increase in the rate of chemical weathering that consumes additional C 0 2 . A negative feedback loop is created. Similarly as global temperatures increase, sea-levels rise causing a reduction in the area of continental landmasses. This in turn means a decrease in weathering of silicates, providing a positive feedback loop. Vascular plants accelerate the rate of chemical weathering as they extract nutrients by secreting organic acid into the regolith to release elements they need from minerals. This compares to more primitive plants that show little evidence of increasing the weathering of silicate minerals. To test the above models it should be possible to plot time against C 0 2 in the atmosphere and see if values decreased as vascular plants evolved in the Devonian-Carboniferous

Importance of the regolith (Figure 2.6). This did indeed happen and the associated global cooling led to the Permo-Carboniferous glaciations. This gives strong support for the connection between climate, weathering, plants and atmospheric C 0 2 levels.

THREE SPECIFIC EXAMPLES OF THE ENGINEERING APPLICATIONS OF REGOLITH GEOLOGY Mt Gambier area, South Australia The South East region of South Australia, centred on Mt Gambier, is the most technically active part of Australia, having risen as much as 60 m in the past 800 000 years (Harvey et al Chapter 17). It also exhibits the most complete onshore Quaternary sequence and the most recent volcanism on the continent. In this geological context it is natural that the regolith has had a substantial impact on settlement and landuse. It has created two significant environmental problems: land drainage and protection of shallow groundwater resources. Some of the consequences of this interaction between geology and human activities in the southeast have been described by Smith (1983), Beal et al (1976) and Tyler et al (1983). Two geological peculiarities of the region are immediately obvious on a map. First, there are no surface streams between the Glenelg River, on the Victorian border, and the Murray River mouth, a distance of 300 km, despite this being the wettest part of the state. Nevertheless, there is a network of man-made channels, a single one of which (Drain M) has an annual discharge near Beachport of 1800 GL, comparable with that of the Murrumbidgee-Lachlan River system in New South Wales. The lack of rivers is due to the presence of highly permeable regolith, with most of the South East region blanketed by porous soils underlain at shallow depth by karstic limestone. The second peculiarity is the series of 'ranges', up to 13 in number, which parallel the present coastline and march inland for over a hundred kilometres. These are huge barrier dunes rising up to 30 m above the surrounding plains, which represent Plio-Pleistocene shorelines stranded by the rising landmass. They become older, lower, more quartzose and less defined as they progress inland; Late Tertiary remnants can be found inland as far as the vicinity of Wilcannia, New South Wales and Horsham, Victoria. LAND DRAINAGE The South East region was not at first attractive to settlers, since much was seasonal swampland or permanently inundated, though it was recognised that the soils were potentially productive if they could be drained (Turner & Carter 1989). Runoff from winter rain migrates northwest along the interdune corridors, which are not only uplifted but also tilted in this direction. The water appears to move by a combination of overland flow on the surface and interflow just below this. It accumulates in low ground at the foot of the dune 'ranges', where it slowly enters the groundwater system or evaporates. The solution was to short circuit this sluggish natural discharge by cutting channels to the sea through low points

in the dune ranges. This was done progressively, starting in the 1860s with swamps closest to the sea. After 130 years of sporadic excavation perhaps 2000 km of channels now dewater more than 5000 km2 of land. The earliest drains were hand-dug, often by workers on unemployment relief, but from 1912 draglines and face shovels ('steam navvies') supplemented manual labour (Turner & Carter 1989). In the 1950s these were supplemented by bulldozers and scrapers. The Woakwine Cutting, one of the drains excavated at this time, is 1 km long and up to 27 m deep and was dug by just two men with one bulldozer/scraper combination in less than three years (Nicol 1991)! The aeolianite dunes were easy to excavate, even by hand, being composed of cross-bedded mixtures of carbonate and quartz sand that is weakly cemented. Their internal structure is complex due to the effects of slow tectonic uplift, rapidly fluctuating Pleistocene sea-levels, solution and redeposition (Sprigg 1952). One practical result was the development of boxwork-like calcrete crusts that have reinforced the dunes, allowing excavation walls to be cut at about 60°, twice as steep as uncemented sand. The swamp soils, once reclaimed, were fertile and therefore suited to closer settlement—an important policy with most colonial governments, but especially so in South Australia, which had little farming land. Environmental problems such as loss of wetland habitats, acid sulfate soils, interference with groundwater recharge, ground shrinkage and subsidence in peat areas, and wind erosion of devegetated dunes were not recognised at the time. The main geotechnical problems encountered were those of shallow groundwater, hard calcrete bands in otherwise soft limestone and marl, and peat shrinkage on exposure. The latter could cause channels to shallow or reverse their flow, and in places the dry peat might even catch fire due to spontaneous combustion. Bogging of heavy construction plant on soft ground was a considerable nuisance, restricting most construction to the summer months. GROUNDWATER The Gambier Limestone aquifer is heavily drawn upon by domestic, industrial and pastoral consumers but, being unconfined, is especially at risk of contamination. It is a very permeable bryozoal coquinite of Oligo-Miocene age, which has a maximum thickness of about 300 m at the present coastline but thins inland (Smith 1983; Holmes & Waterhouse 1983). In addition to its high intergranular (primary) porosity this formation contains extensive solution cavities (secondary porosity). These karst features are almost everywhere buried under a thin cover of Late Cenozoic sediments and soils (Grimes 1994). Depth to the water-table varies from 0 to 35 m at present, depending on topographic location, but was about 100 m lower during Pleistocene glacial maxima (the latest around 18 ka BP). It was during these sea-level lowstands that the karst terrain—caves, open joints, collapse dolines (sinkholes), uvalas (connected dolines) and cenotes (doline lakes)—was formed. The effect of these has been to increase the transmissivity of the Gambier Limestone aquifer from about 300 m2/d (primary porosity) to as much

G. Taylor and G. H. McNally

Figure 2 . 7 Composite geological section (not to scale) through main regolith materials, Mt Gambier area of South Australia. Residual soils and tuffs not shown because of thinness and restricted areal distribution.

as 20 000 m2/d in places (Emmett & Telfer 1994). Water quality is very good: typically 300-600 mg/L total dissolved salts. GROUNDWATER AND THE REGOLITH The nature of the regolith covering the Gambier Limestone affects the shallow groundwater system in two ways. First, it controls the rate of surface water intake and hence the location of recharge areas. Second, it can make this aquifer vulnerable to pollution from surface sources such as street runoff and stockyard effluent. These regolith materials are complex and of variable permeability, but are generally 10-30 m thick and nearly ubiquitous; bare karst is exposed in only a few small areas, such as The Sisters southwest of Mt Gambier. The four main types of material covering the Gambier Limestone are: (i) dune aeolianites or calcarenites of the Bridgewater Formation (Pleistocene) which make up the 'ranges', the crests of which are up to 60 m ASL (though because of the rising Gambier Limestone floor, these barriers are only 15-30 m above the adjacent plains); (ii) the interdune flats are underlain by calcitic and dolomitic mudstone, sand sheets, flood deposits and lateritised marine sands; these Pliocene to Holocene deposits are of lagoonal, lacustrine and aeolian origin and are typically about 10 m thick (Schwebel 1983); (iii) a variety of residual soils ranging from podzols to terra rossas, plus peat and organic clays, are developed on the calcarenites and interdune sediments; and (iv) late Pleistocene to Holocene tuffs cover an area of over 100 km2 to average depths of 2-3 m, close to extinct volcanic vents. The relationship between these regolith materials and the unconfined aquifer is illustrated on Figure 2.7. All are permeable to some degree, but the interdune areas are less pervious than the dune sands. Within the interdune areas recharge is probably inversely proportional to regolith thickness. CONSTRUCTION MATERIALS A group of quarries located west of Mt Gambier produces Gambier Limestone blocks (ashlars) for building stone. This material is attractive because it is non-abrasive and soft, and can therefore be easily cut yet has adequate dry strength for

load-bearing walls of single-storey buildings. With porosity up to 60% this stone is unusually lightweight and an excellent insulator, but it suffers from high water absorption where exposed to the weather. It is the only significant source of block stone, as opposed to stone veneer cladding, in Australia. Current production is around 11 000 t/y. Calcrete from the Bridgewater Formation and tuff provide unsealed roadbase, though better quality aggregate has to be crushed and screened from basalt flows at Mt Schank and elsewhere. Mixtures of scoria and aeolianite have been used for pavement materials in adjacent areas of western Victoria, but such quarrying is now declining because of environmental objections to defacement of cinder cones. Fine aggregate is obtained from sand pits throughout the region, but only the quartz-rich varieties are suitable for high-strength concrete.

Stuart Range, northern South Australia The Stuart Range is a 60 m high escarpment, or 'breakaway', which trends northwest-southeast through the town of Coober Pedy in the far north of South Australia (Figure 2.8). It is the major topographic feature in this region, and acts as a watershed between the Lake Eyre hydrological basin to the east and an upland plateau to the west. The underlying Mesozoic sedimentary rocks are stratigraphically equivalent to those of the Great Artesian (Eromanga) Basin, and rest on Permian coal measures and granitic basement. West of the escarpment these sediments are up to 150 m thick, but to the east this reduces greatly, since the upper weathered (and opal-bearing) portion of the Bulldog Shale has been stripped. Coober Pedy in fact owes its location to the opal seams or 'levels' that were once exposed there in the breakaway face. The 110 m thick Bulldog Shale, of Early Cretaceous age, is remarkable because intense weathering has rendered it harder and more permeable than the parent rock. In its unweathered state this formation is a dark-grey, very stiff clayshale or weak mudstone that is pyritic and carbonaceous in part and contains boulders towards its base. The weathered Bulldog Shale is a white and mottled kaolinitic claystone, which resembles chalk. The extent of leaching is apparent in its low bulk density. The weathering profile is about 50 m deep and its base is roughly coincident with that of the escarpment.

Importance of the regolith

Opal mines silcrete and spoil heaps ^

Limited recharge Degraded landslides

Weathered Bulldog Shale

. Gilgai

Recharge along creeks

Unweathered Bulldog Shale

Mesozoic Aquifer

Unconfined SWL

GTO03-99

Basement granite

• Permian

Figure 2 . 8 Regolith geology of the Stuart Range escarpment near Coober Pedy, South Australia (not to scale). Throughout its 80-year history, Coober Pedy has been plagued by water supply problems, both of quantity and quality. In the late 1960s it became the first sizeable Australian community to be served by a desalinisation plant. At first a solar still was installed, but this was soon replaced by reverse osmosis equipment. The town's permanent population has increased twenty-fold since the 1950s, to about 5 0 0 0 , and the tourist numbers have grown even more rapidly. At the same time per capita water consumption has also greatly increased. This demand has been met by a combination of saline bore water for toilet flushing, low salinity (reverse osmosis) water for showers, and tank water for drinking. In addition, potable surface water is harvested from clay pans and ephemeral pools up to 100 km distant.

either untreated for washing or desalinated for drinking. The apparent paradox of an aquifer within the Bulldog Shale results from leaching and shrinkage during Early Tertiary weathering, which has created clusters of open and saturated joints in places. Although storage in this aquifer is small, its permeability is locally high. Leakage of surface waters through fractures and cavities is believed to be the main cause of lower salinity in parts of the underlying sand aquifer. A number of low-yielding, low-salinity wells in the region are significantly located adjacent to surface depressions in the weathered Bulldog Shale which become swamps in wet seasons.

Other environmental issues related to the regolith in the Coober Pedy area include: (i) excavatability and stability of mine tunnels and residential caverns ('dugouts'); (ii) availability and quality of roadbase and aggregate; (iii) landslides and potential foundation problems on swelling clay; and (iv) safety hazards caused by the presence of thousands of abandoned shafts, opencuts and mine-waste dumps.

Another surprising aspect of the weathered Bulldog Shale is its capacity to sustain unsupported excavations 10 m or more wide beneath only a few metres of cover. In other weathered sedimentary rocks artificial reinforcement provided by shotcrete, mesh and steel arches would be required. Possible explanations for this cavern-roof stability include the lack of bedding and of persistent jointing, negligible in situ stress, and a tendency for the ferruginous mottled zone to self-cement on exposure. Shaft sinking for opal mines is now generally carried out using truck-mounted Calweld-type bucket augers. These are drilled at 1 m diameter for exploration and reamed out to 2 m if completed as production shafts. This type of machine is primarily a soil auger, so it can only operate in weak and non-abrasive rock such as the weathered Bulldog Shale. The main impediments to what is otherwise an economical form of drilling are silicified bands, which have to be blasted through. Mine tunnels were formerly blasted laterally from the shaft bottom, but are now largely driven by small tunnel-boring machines that also excavate some dugouts.

GROUNDWATER SUPPLIES Groundwater has been obtained around Coober Pedy from the Mesozoic sand, the weathered Bulldog Shale and alluvial sand aquifers, though only the first is a significant producer (McNally 1977). The Mesozoic aquifer is a composite 6 0 - 8 0 m-thick, poorly cemented sand unit of the Upper Jurassic to Lower Cretaceous. The overlying Bulldog Shale is nominally a confining bed, though the sand aquifer is in most places either unsaturated or saturated but with only a few metres of head, owing to its elevated location at the edge of the Great Artesian Basin. This hydraulic geometry is illustrated on Figure 2.8. Its salinity is quite variable, in the range 1000 to 17 0 0 0 mg/L total dissolved salts (up to half that of seawater). The low salinity water, which occurs to the northwest of the town (Mason 1975) is most attractive for domestic supply,

UNDERGROUND EXCAVATION

AGGREGATE AND ROADBASE The Coober Pedy area is devoid of aggregate-quality rock, except for silcrete boulders and cobbles. These one-stone-

G. Taylor and G. H. McNally

thick lags could be 'broomed up' by bulldozer blading, but the damage to the ground surface would be out of all proportion to the amount of product obtained. Silcrete outcrops are more attractive, but these seldom cover more than 2-3 ha, and yield less than 100 000 t of roadbase; the proportion of screened and sized aggregate would be much less. Such deposits occur at sites of former groundwater discharge; many appear to be fossil mound springs (McNally & Wilson 1996). Furthermore, silcrete is only of marginal quality as an aggregate source. Its chips tend to be sharp-edged and brittle, causing tyre wear and often fracture under wheel impacts; they may strip (de-bond) from bitumen seals or polish (become less skid-resistant) under traffic. Opaline silica is potentially reactive in concrete, forming an expansive gel with excess alkalis in the cement paste. Crushing and screening silcrete materials generates large quantities of siliceous dust that may constitute a health hazard to workers. Quarry sites are shallow, hence extensive, and difficult to rehabilitate because of the low rainfall and selfhardening pallid zone beneath the duricrust. Roadbase for unsealed pavements is, nonetheless, available in large quantities from other partly cemented Tertiary and Quaternary surficial deposits. This material, especially nodular calcrete, is also suitable for sub-base in sealed roads, though basecourse for these higher standard pavements (like aggregate) has to be imported from outside the area. EXPANSIVE CLAYS The unweathered Bulldog Shale is an unlithified clay-shale composed largely of smectitic minerals. Self-mulching 'popcorn textured' soils and gilgai on treeless plains are characteristic surface features. The gilgai are defined by hummocky microtopography and mounds of silcrete gibbers on the plains, and by ridge-and-furrow topography on lower slopes. The latter, informally referred to as 'hilgais', are caused by a combination of shrink-swell ground heaving and downslope creep. Gypsum crystal clusters are abundant in the crumbly soil, sparkling in the sun like thousands of broken glass panes. Any buildings placed on these soils would be subject to large seasonal movements, salt damp and sulfate attack on concrete foundations. The main developments to date in the swelling clay areas northeast of Coober Pedy have been unsealed dry-weather roads that are relatively tolerant of subgrade heaving, although future sealed pavements will require more protection. Extensive landsliding has occurred further north in the Mt Barry-Arckaringa area, where unaltered Bulldog Shale is exposed in the lower slopes of the Stuart Range escarpment. These circular slips at first give the impression that multiple duricrusts are present, but are in fact repetitions of the same surface. The shale is dispersive as well as expansive, gullying rapidly where the surface veneer of gibbers is disturbed. ABANDONED MINING HAZARDS The Coober Pedy Precious Stones Field covers 4700 km 2 , though only about a fraction of this area, mostly within 10 km of the town, has been extensively mined. This has left

thousands of derelict mine shafts and tunnels, opencuts, waste dumps and minor excavations. Almost none of these have been backfilled, and few have collapsed naturally. The extent of dereliction is such that no site rehabilitation appears to be economically feasible at this stage. The main safety measures are warning signs erected along main roads, plus light fencing of especially dangerous areas close to tourist routes. Prior stream deposits on the Riverine Plain, New South Wales Prior stream sands are distinctive channel deposits which fill Late Pleistocene courses of the Murrumbidgee, Murray and Lachlan Rivers where these cross the eastern half of the Murray Basin in New South Wales, known as the Riverine Plain; they are also well developed in northern Victoria. Their existence was first noted during soil mapping in the 1940s, but their extent became clear only after airphoto coverage of the region became available in the 1950s (Langford-Smith 1960). This drainage system radiates from the New South Wales western slopes and from the central highlands of Victoria towards interior playas such as the Willandra Lakes (Butler 1950). It pre-dates the present-day Murray River outlet in South Australia and the tectonic activity which uplifted the Cadell Fault south of Deniliquin, and also two later styles of alluviation, the ancestral and modern streams. The differences between these are summarised by McNally and Sutherland (1997) and illustrated on Figure 2.9. Recent thermoluminescence dating (Page et al 1991) confirms that the prior stream deposits are generally more than 50 000 years old, despite their excellent state of preservation. What is the environmental significance of these prior streams? First, their sandbeds are the shallowest aquifers beneath the Riverine Plain. Though not significant groundwater producers themselves, because higher yielding and more widespread Late Tertiary aquifers occur only a little deeper, they are important factors in land salinisation and waterlogging. They act as conduits for excess irrigation water and for salt leached from surface soils, directing both towards low points in the topography. Second, and more important economically, the prior streams give rise to light-textured soils which are more favourable to irrigated agriculture than the underlying stiff and saline floodplain clays. During Late Pleistocene time the prior stream channels acted as conveyors for sand from the highlands, each moving perhaps 50 000 to 500 000 t/d at flood stage (Schumm 1968). Much of this was blown out of the streambeds during dry seasons, or redistributed from levees, creating a nearby veneer of sandy, low salinity soils. Wind-blown pelletal clay loess Cparna'), some of it calcareous, has improved the cultivation potential of the sand sheets bordering the prior streams. Finally, prior streams are the most important sources of road pavement material for a large portion of southwestern New South Wales that lacks quarryable rock. Well-graded (poorly sorted) sands from these channels provide a natural roadbase that can be compacted to high density. Though inferior to crushed rock roadbase under heavy traffic, it is quite adequate for shire roads and secondary

Importance of the regolith

Figure 2.9 Comparison of prior, ancestral and modern (misfit) fluvial landforms.

highways. This material is colloquially known as 'prior stream gravel' in the roadmaking sense, meaning a natural gravel-sand-fines mixture used for pavement courses. It would be better described as a sandy loam, since only a few percent of gravel-sized particles are present and these are in the fine range, 2 - 6 mm. CHANNEL FORM Prior streams are thought to have been most active between about 105 and 85 ka BP, early in the Last Glacial. They revived during the period 5 0 - 4 0 ka BP, when they were important feeders to the inland lake system (Page et al 1991). Their channels were relatively straight, wide and shallow. On airphotos the distance between levees is typically about 100 m, but this represents only the final waning phase of deposition; the underlying sandbed can be much broader. The channel tends to be steep-sided and flat-bottomed, with a depth of 3 - 6 m and a width-to-depth ratio of more than 50 (i.e. five times that of the present Murrumbidgee). Meanders, where visible, can have a wavelength of several kilometres. Individual channel seg-

ments can be traced for more than 20 km on airphotos, and may cross (or rather be eroded through by) younger ancestral and modern streams (Figure 2.9). Because of the low sinuosity of the prior stream channels, their gradient is relatively steep, about 1:3500. This compares with flatter than 1:7500 for the present-day Murrumbidgee (Schumm 1968). Hence they were more competent to move coarse sandy detritus by bedload traction than the present channels. Bankfull discharges estimated by Schumm are twice to thirty times those at present, while the rate of sand movement is claimed to have been 25 to 250 times greater. PRIOR STREAM SEDIMENTS Prior stream deposits are strikingly uniform: cross-bedded mixtures of silty sand, with few clay or gravel lenses. The deposits are usually linear, parallel-sided and elongate, with a stiff clay floor at 2 - 7 m depth. Though they may be very conspicuous on airphotos (but not all are), at ground level they are generally apparent only as a slight rise, perhaps 0.2 m, on an otherwise featureless plain. Particle-size distributions from borrow pits indicate that

G. Taylor and G. H. McNally

the prior stream 'gravels' are actually an even mixture of fine gravel (0-25%, but averaging 10%), coarse to fine sand and silt/clay fines (20-40%). These materials look like, and feel like, a yellow or off-white dirty sand with traces of 2 - 4 mm granules. Both sand and gravel particles are predominantly rounded quartz. The fines content is high for a pavement material, but it is relatively low in plasticity. This fraction was formerly thought to be mainly silt-sized ( 2 - 7 5 pm), but evidence of dispersive behaviour and from XRD tests suggests that much of this 'silt' is aggregated clay which may deflocculate. The provenance of these granular but gravel-poor sediments is thought to have been granites of the Australian Alps. Two possible mechanisms could explain their lack of gravel clasts. First, they may be reworked Pliocene channel sands and alluvial fans, equivalent to the deep leads of the Victorian highlands. Surface exposures occur in the Urana-Narrandera area and more widely in the subsurface, where they constitute the main aquifer system in the eastern part of the Riverine Plain (Woolley & Williams 1978). The second mechanism postulates periglacial heaving, solifluction and accelerated erosion of saturated, deeply weathered granitic regolith during spring thaws in the highlands. This presumes that Pleistocene periglacial action was much more widespread and of longer duration than glaciation, which was active for only a few millenia around 18 ka BP. Periglacial soil churning may have reached down to about 900 m asl, placing much of the Murrumbidgee catchment southwest of Canberra within its influence. Thin blankets of windblown clay (parna) occur widely in the eastern part of the Murray Basin and adjacent highlands, and it is likely that some of this has been incorporated in the prior stream alluvium. Long-term transfer of salt by westerly winds from Mallee playas to floodplain deposits of the Riverine Plain may also explain the sodicity of the swelling clays that surround the prior steam channels. ENGINEERING PROPERTIES The prior stream 'gravels' are actually well-graded sands with a large excess ( 2 0 - 4 0 % ) of fines. Like other clay-rich soils, their strength is very dependent on matrix suction (roughly, the degree of dryness) and dry compressive strength may be ten times that of soaked samples. Because of this moisture sensitivity, prior stream gravels are unsuitable in pavements that may become saturated. They therefore perform best in lower rainfall western areas, on embankments well above flood level, and beneath wide bitumen seals (which maximise the zone of stable moisture content beneath the blacktop). Conversely, shallow watertable locations, where capillary rise and vapour condensation beneath the seal can cause the basecourse to become progressively saturated, should be strongly avoided. The traffic life of a typical prior stream gravel pavement is less than that for crushed roadbase, though sufficient for all but the most heavily trafficked highways in this region. Unlike crushed roadbase, however, they are subject to shear failure under the wheels of braking heavy trucks at any pavement age because of their fine grainsize, rounded particles and abundant clay fines.

CONCLUSION Regolith forms the human interface with our planet Earth. It is in it we grow our food, from it we mine many resources, in it grow our native plants on which our native animals feed and it forms the foundations for many of our engineering enterprises. This chapter has demonstrated how we depend on the regolith in many ways, but in particular how some geotechnical problems of Australia are regolith related. For us it is critical that we understand the regolith, its materials and history so that we can manage it in a conservative way so it is available for future generations in as pristine state as possible.

REFERENCES ABELL R. S. 1985. Geology of the Lake George Basin, NSW. Bureau of

Mineral Resources Record 1995/4. L. & JACKSON J . A. 1 9 8 0 Glossary of Geology. American Geological Institute, Falls Church, Virginia.

BATES R .

BEAL J. C. BOUCAUT W . , SELBY J. ET AL. 1 9 7 6 . Hydrogeology a n d envi-

ronmental geology of south eastern SA. Excursion Guide 32AC, 25th International Geological Congress, Sydney. Development of soil landscapes. In: Soils an Australian Perspective, pp. 1 6 3 - 1 7 2 . Division of Soils, CSIRO, Melbourne. BERNER R. A. 1995. Chemical weathering and its effect on atmospheric C0 2 and climate. In: White A. F & Brantley S. L. eds. Chemical Weathering Rates in Silicate Minerals, pp. 565-583. Reviews in Mineralogy 31.

BECKMANN G. G. 1 9 8 3 .

BIRD M. I. & CHIVAS A. R. 1988. Stable-isotope evidence for low-tem-

perature kaolinitic weathering and post formational hydrogen-isotope exchange in the Permian. Chemical Geology (Isotope Geoscience Section) 77, 249-265. BISHOP P. 1980. Tertiary drainage on the southeastern highlands of New South Wales. Bureau of Mineral Resources Record 1980/67, 7-8. BISHOP P. 1988. The eastern highlands of Australia: the evolution of an interplate highland belt. Progress in Physical Geography 12, 159-182. BUTLER B. E. 1950. Theory of prior streams as a casual factor in the distribution of soils in the Riverine Plain of south-eastern Australia. Australian Journal of Agricultural Research 1 , 2 3 1 - 2 5 2 . DIVISION OF SOILS C S I R O . 1 9 8 3 Soils an Australian Perspective. CSIRO, Melbourne. EMMETT A. J.& TELFER A. L. 1994. Influence of karst hydrology on

water quality management in southeast South Australia. Environmental Geology 23, 149-155.

FRAKES L. A., FRANCIS J. F. & SYKTUS J. I. 1992. Climate Modes of the

Phanerozoic: the History of the Earth's Climate over the past 600 million years. Cambridge University Press, Cambridge. GRIMES K. G. 1994. The South-East karst province of South Australia. Environmental Geology 23, 134-148. HILL S. M. 1995. Biological applications of regolith mapping. In: McQueen K. G. & Craig M. A. eds. Developments and new approaches in regolith mapping, pp. 69-72. Centre for Australian Regolith Studies, University of Canberra, Canberra.

HILL S. M . , EGGLETON R. A. & TAYLOR G . 1 9 9 7 . A regional r e g o l i t h —

landform framework for mineral exploration models in the Broken Hill region. AusIMM 1997 Annual Conference, 12-15 March 1997, Ballarat. pp. 131-138.

HILL S. M . , MCQUEEN K. G . & FOSTER K. A. 1 9 9 9 . N e w S o u t h W a l e s

regolith-carbonate accumulation. In: Taylor G & Pain C. F. eds. Regolith '98 Proceedings, New Approaches to an old continent, pp. 191-208. Cooperative Research Centre for Landscape Evolution and Mineral Exploration, Perth. HILLS E. S. 1975. The Physiography of Victoria. Whitcombe & Tombes, Melbourne.

HOLMES J. W. & WATERHOUSE J. D. 1983. Hydrology. In: Tyler M . J.,

Twidale C. R., Ling J. K. & Holmes J. W. eds. Natural history of the South East, pp. 49-59. Royal Society of South Australia, Adelaide. R. F. 1996. The Australian Soil Classification. CSIRO, Collingwood.

ISBELL

Importance of the regolith T. 1 9 6 0 . The dead river systems of the Murrumbidgee. Geographical Review 50, 368-389. G. 1975. Groundwater near Giddi Giddinna Creek, northeast of Coober Pedy SA. Geological Survey of South Australia Quarterly Notes 56. MCNALLY G . H . 1 9 7 7 . Coober Pedy town water supply—groundwater prospects and completion report, Stuart Range Bores 7-10. South Australia Department of Mines Report RB 77/77 (unpubl.). MCNALLY G. H. & SUTHERLAND M. 1997. Engineering geology of prior stream deposits on the riverine plain, eastern Murray Basin NSW. In: McNally G. H. ed. Case studies in Engineering Geology, Hydrogeology and Environmental Geology, pp. 95-109. Geological Society of Australia, Environmental, Engineering, Hydrogeology Specialist Group Papers Series 3. MCNALLY G. H. & W I L S O N I. R. 1996. Silcretes of the Mirackina Palaeochannel, Arckaringa, South Australia. AGSO Journal of Australian Geology & Geophysics 16, 295-301. MERRILL G . P. 1 8 9 7 . A treatise on rocks, rock weathering and soils. McMillan, New York. NICOL S. 1991. South East: discovering South Australia. Royal Automobile Association, Adelaide. OLLIER C. D. 1977. Early Landform Evolution. In: Jeans D. N. ed. Australia: a Geography, pp. 85-98. University of Sydney Press, Sydney. OLLIER C. D. 1982. Geomorphology and tectonics of the Dorrigo Plateau, NSW. Journal of the Geological Society of Australia 29, 431-435. OLLIER C . D . & PAIN C . F. 1 9 9 6 . Regolith, soils and landforms. Wiley & Sons, Chichester. PAGE K . J., NANSON G . C . & PRICE D . M . 1 9 9 1 . Thermoluminescence chronology of Late Quaternary deposition on the Riverine Plain of southeastern Australia. Australian Geographer 2 2 , 1 4 - 2 3 . PAIN C. F. 1983. Geomorphology of the Barrington Tops area, New South Wales. Journal of the Geological Society of Australia 30, 187-194. PAIN C. F. & OLLIER C. D . 1 9 9 2 . Ferricrete in Cape York Peninsular, North Queensland. BMR Journal of Australian Geology & Geophysics 1 3 , 2 0 7 - 2 1 2 . PATON T . R . , HUMPHREYS G . S . & MITCHELL P. B . 1 9 9 6 . Soils a New Global View. University College London Press, London. PILLANS B . , TONUI E. & IDNURM M. 1 9 9 9 . Palaeomagnetic dating of regolith. In: Taylor G. & Pain C. F. eds. Regolith '98 Proceedings, New approaches to an old continent, pp. 237-244. Cooperative Research Centre for Landscape Evolution and Mineral Exploration, Perth. RIDE W. D . L, TAYLOR G. & DAVIS A. C. 1989. Zoological history of the Australian Alps—the mammal fossil-bearing deposits of the Monaro. In: Good R. ed. The significance of the Australian Alps, pp. 79-110. The Australian Alps National Liaison Committee, Canberra. RUXTON B. P. & TAYLOR G. 1982. The Cainozoic geology of the Middle Shoalhaven Plain. Journal of the Geological Society of Australia 29, 239-246. SCHUMM S . A. 1968. River adjustment to altered hydrologic regimen— LANGFORD-SMITH

MASON M .

Murrumbidgee river and paleochannels, Australia. US Geological Survey Professional Paper 859. Quaternary dune systems. In: Tyler M . J . et al. eds. Natural History of the South East Australia, pp. 1 5 - 2 4 . Royal Society of South Australia, Adelaide. SENIOR B. R., M O N D A & HARRISON P. L. 1978. Geology of the Eromanga Basin. Bureau of Mineral Resources Bulletin 167. SHEARD M . J . 1 9 9 5 . Quaternary volcanic activity and hazards. In: Drexel J. F. & Preiss W. V. eds. The Geology of South Australia, Vol. 2, pp. 264-268. Geological Survey of South Australia, Adelaide. SMITH P. C. 1983. A groundwater resource under stress at Mount Gambier SA. In: Knight M. J., Minty E. J. & Smith R. B. eds. Collected Case Studies in Engineering Geology, Hydrogeology and Environmental Geology, pp. 307-321. Geological Society of Australia Special Publication 11. SPRIGG R. C. 1952. The geology of the South East province, South Australia. Geological Survey of South Australia Bulletin 29. STEWART A. J . , BLAKE D . J . & OLLIER C. D . 1 9 8 6 . Cambrian river terraces and ridge tops in central Australia: oldest persisting landsurfaces? Science 233, 7 5 8 - 7 6 1 . TAYLOR G., TAYLOR G. R., BINK M. ETAL. 1985. Pre-basaltic topography of the northern Monaro and its implications. Australian Journal of Earth Sciences 32, 65-71. SCHWEBEL D . A. 1 9 8 3 .

TAYLOR G . , TRUSWELL E . M . , M C Q U E E N K . G . & BROWN M . C .

1990.

Early Tertiary palaeogeography, landscape evolution and palaeoclimatology of the southern Monaro, NSW, Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 78, 109-134. TAYLOR G. & WALKER P. H . 1986a. Tertiary history of Lake Bunyan, northern Monaro, NSW, Part I: geological setting and landscape history. Australian Journal of Earth Sciences 33, 219-229. TAYLOR G. & WALKER P. H . 1986b. Tertiary history of Lake Bunyan, northern Monaro, NSW, Part II: facies analysis and palaeoenvironmental interpretation. Australian Journal of Earth Sciences 33, 231-251. TURNER M. & CARTER D. 1989. Down the drain. South Eastern Drainage Board, Adelaide. TYLER M . J . , TWIDALE C . R., LING J . K . & HOLMES J . W . (Editors) 1 9 8 3 . Natural history of the South East. Royal Society of South Australia, Adelaide. WALTHER J . 1 9 1 5 . Laterit in Westaustralien. Zeitscrift der Deutschen Geologischen Gesellschaft 67B, 1 1 3 - 1 4 0 . WELLMAN P. 1987. Eastern highlands of Australia: their uplift and erosion. BMR Journal of Australian Geology & Geophysics 10,277-286. WILLIAMS M. A. J. 2001. Chapter 1—Quaternary climate changes in Australia and their environmental effects. Geological Society of Australia Special Publication 21, 3-11. WOOLLEY D. R. & WILLIAMS R. M. 1978. Tertiary stratigraphy and hydrogeology of the eastern part of the Murray Basin NSW. In: Storier R. R. & Kelly I. D. eds. The Hydrogeology of the Riverine Plain of Southeast Australia, pp. 45-65. Australian Soil Society, Riverina Branch, Sydney. Received 12 March 1998; accepted 27 April 2000

Geological Society of Australia Special Publication 21, 27-36

CHAPTER 3—Soil-related engineering problems: identification and remedial measures R. W, FITZPATRICK' P. SLADE' AND P. A. HAZELTON2 1 CSIRO Land and Water, 2 Environmental

Private Bag No. 2, Glen OsmondI SA 5064, Australia. Engineering, University of Technology, Sydney PO Box 123, Broadway NSW 2007, Australia.

The two case studies presented have demonstrated how movements, caused by soil shrink-swell properties, and corrosion, caused by saline soil solutions, damage optical fibre cable networks and water pipes. Such faults are very costly to the operator and if avoided can save millions of dollars. The case studies illustrate how pedological information has been used to predict and overcome such practical problems. KEY WORDS: corrosive soils, expansive soils, optical-fibre cables, pedology, saline soils, soil maps, water pipes.

INTRODUCTION The work of the soil scientist, especially in the field of pedology, has begun to move from a predominately agricultural context to one involving a broad range of stakeholders. This change has resulted in fundamental research and the practical applications of theoretical work (Bouma 1997). Pedology is an integrative and extrapolative component of soil science, which provides an organisational framework to explain and quantify spatial variability within landscapes. With the use of mechanistic models, it provides understanding of biogeochemical processes on both regional and global scales. Consequently, pedology provides an excellent framework for the extrapolation of the seen components of soils (hand specimens and soil layers) to the soil profile, and also to toposequences and catchments. Pedologists seek to apply this information to understand, predict and solve practical land-use problems. Citing several examples in the USA, Fanning and Fanning (1989) used the term 'pedotechnology' to describe the application of pedology in this way. However, in Australia, pedologists, have traditionally taken a narrower view of their work because they have, of necessity, placed much emphasis on soil surveys, classification systems and the methods used in describing soils for agricultural use (Northcote et al 1960-68; Isbell 1996; McDonald et al 1990; Fitzpatrick et al 1999). There is now a decline of interest in conventional soil mapping because most areas in Australia have been mapped at 1:250 000 scale or smaller and also because the perceived clientele for soil maps (e.g. farmers and land-use planners) have not directly used them to the extent formerly expected. Therefore more emphasis is now placed on special-purpose, detailed soil surveys such as investigations of soil degradation and pollution. Despite this change, the following four questions are most frequently asked by users of land resource information: (i) what soil properties are changing, vertically and laterally in landscape and with time; (ii) what are the most suitable approaches to characterise, monitor, predict and manage the changes in soils; (iii) what soil measurements

are required to make suitable predictions about soil and landscape conditions and about sustainable land use; and (iv) to what extent do soil processes and the management of soils, influence water quality? These four questions can be answered by utilising combinations of pedological data along with soil physical, chemical, hydrological and mineralogical data. The combined information assists the understanding of how soils vary in landscapes so that strategies can be developed for managing both spatial and temporal changes within them. Recently, pedologists in Australia have utilised existing information from soil maps, combining data such as soil physics and mineralogy, to work in non-agricultural contexts (Sheard & Bowman 1996). This chapter presents two brief case studies illustrating how this applied soil science information has been used to solve practical problems involving the movement and corrosion of optical fibre cables and water mains. Each of these approaches could have wider applications.

EXPANSIVE AND CORROSIVE SOILS The phenomena of swelling and corrosion by soils were recognised long before the development of soil science and geotechnical engineering. It is known that expansive soils cause more damage to structures, particularly light buildings and pavements, than any other natural hazard, including earthquakes and floods (Jones & Holz 1973). Damage caused by expansive soils to residences in the USA alone is expected to total US$ 997 million in the year 2000. The cost of this damage to other structures such as commercial/industrial buildings and transportation facilities could raise these estimated annual values by a factor of 2 to 3 times (Nelson & Millor 1992). In 1981, the Department of Housing and Urban Development in the USA estimated average annual losses caused by shrink-swell soils to be US$9 billion. In Australia, Isaacs (1950) used Prescott's Soil Map of Australia (1944) as a basis for foundation design for domes-

R. W. Fitzpatrick et al.

tic buildings. However, the first broad-scale use of soil classification for engineering purposes was carried by Aitchison et al (1954), who mapped the soils of Metropolitan Adelaide and classified them according to their shrink-swell properties. This information assisted designers of engineering structures, especially house footings. The classification system, which was used considered the soil profile to a depth of 0.8-1.0 m and was based on the pedological system developed by Stephens (1962). This work has subsequently been extended by Taylor et. al (1974) and Sheard and Bowman (1996). A similar classification was developed for Metropolitan Melbourne by Walsh et al (1978). The New South Wales Building Licensing Board (now the New South Wales Building Services Corporation) issued a classification of New South Wales Soils for Housing. The soils are ranked in five classes, depending on their reactivity, ranging from stable to extremely reactive for which the expected maximum surface movement is >60 mm (Hazelton & Murphy 1992). These classifications have been used widely in the respective cities to assist the designers of residential footings. Grant (1968, 1971, 1973, 1974) developed a different system based on landscape classification for engineering purposes, known as the P.U.C.E. (Pattern, Unit, Component, Evaluation) Program for Terrain Evaluation. Usually, concrete and steel structures can be designed by experienced engineers to perform successfully. This is achieved because the materials are carefully made and have either known, or accurately measurable, properties that remain substantially constant. However, optical-fibre cables (a glossary of terms is given in Appendix 3.1) and water pipes, for example, cannot be guaranteed to perform with the same degree confidence. Even after extensive site investigations, one cannot be certain that all of the soil properties relevant to the performance of such services have been identified. These latter difficulties arise because soil is extremely variable and can be very sensitive to moisture changes; this is particularly true for shrink-swell and corrosive (e.g. saline) soils.

CASE STUDY 1: METHODS FOR SOLVING SOIL-RELATED PROBLEMS CAUSING FAULTS IN THE AUSTRALIAN TELECOMMUNICATION NETWORK Optical-fibre cables are now used extensively in telecommunications throughout the world. Telstra Australia has developed an economical and rapid method for ploughing these cables directly into soils (Brass 1993). However, transmission faults (Appendix 3.1) can develop in some optical-fibre cables that are stretched by movements in soils, which shrink and swell seasonally. In addition, some soils were found to cause the stainless steel in cable marker tapes (Appendix 3.1) to corrode. If these faults that are potentially very costly, could be mitigated or avoided, the performance of telecommunication networks would be enhanced and millions of dollars saved. This challenge required a comprehensive investigation of the complex interactive effects of soil type, vegetation, climate, mineralogy and groundwater chemistry on optical-fibre cable networks in Australia.

This case study summarises the underlying approaches that have been used to develop suitable methods to reduce the impact of soil-related damage to the Australian opticfibre cable network (Fitzpatrick et al 1995a, b). An important aspect of the work was the close liaison with Telstra engineers in order to ensure that the approaches stemming from the research were suitably practical.

Soils and fault occurrences in optical-fibre cables Reconnaissance pedological investigations in the field revealed that soil-induced optical-fibre cable faults were confined to sections of routes through soils that have shrink-swell properties. These soils are often referred to as Cracking Clays (Northcote 1990-68), Vertisols (Soil Survey Staff 1999) or Vertosols (Isbell 1996). In these soils, optical-fibre cables can be stretched enough to cause the optical signals being transmitted though them to become distorted (Appendix 3.1). Subsequently, more detailed investigations were carried out on several aspects of the problem including the effects of: (1) The depth and swelling properties of various types of clay soils, indicated, for example by the presence of slickensides and gilgai (Soil Survey Division Staff 1993) (Figure 3.1, Appendix 3.1). Soil types can vary a number of times along a cable route (Figure 3.2). Soil types have been broadly mapped for the entire Australian continent and have been published as soils maps (Northcote et al 1960-68). (2) The particle sizes and mineralogy of the component clays in the soils. (3) The type and concentration of salts present, such as sodium chloride and lime (Figure 3.1). (4) Climatic conditions before, during and after cable laying (wetting and drying of soil). Faults develop during prolonged dry periods following wet ones, and are more likely in soils carrying woodland vegetation capable of drying the soil to greater than 1 m. Factors such as variability of rainfall, which leads to wetting and drying from one season to another, mean that the cause of faults will also vary from one place to another. (5) Vegetative conditions, particularly density of root growth and transpiration rates, before, during and after emplacement of cable. (6) Soil compaction during reinstatement of soil following cable laying. (7) Topographic conditions which influence water flow. (8) Groundwater properties, particularly salt concentration. When salts are leached from a highly saline shrinkswell soil by good quality water, such as rainwater, the soil may swell extensively. Such conditions can develop if deep cracks or furrows form above the cable due to poor reinstatement of soil following cable laying. (9) Rodent damage to cables in deeply cracked soils in the tropics. A simple but indicative, field test was specifically developed to easily show the swelling potential of soils in the field. This test was related to a range of laboratory-based physical and chemical tests on a broad range of samples taken from along existing cable routes. These laboratory tests included the examination of undisturbed samples with a membrane oedometer and by the shrink-swell method of

Soil-related engineering problems

CRACKS Polygonal Vertical

CaC0 3 Cable

Master Slickenside

Slickenside

Figure 3.1. Schematic section though a typical back swelling soil showing gilgai microrelief, cracks, zones with slickensides, where cable distortion occurs due to soil movement (shearing action) Sheard and Bowman (1996) to determine the instability index. Many geotechnical engineers regard this latter test as a most useful one for assessing vertical movement in shrink-swell soils. Because of the strong influence that climate and vegetation have on the shrink-swell properties of clay soils, the effects of climate, vegetation, mineralogy and groundwater chemistry on soil shrink-swell properties were closely studied (Fitzpatrick et al 1995a). These studies led to the development of a vegetation index, which rates the size and density of vegetation, because these factors modify the shrink-swell behaviour of soils. Similarly, a climatic risk map was developed, by using the Geographic Information System (GIS) combined with satellite remote sensing, to show seasonal variations in vegetation due to rainfall. It was found that a one-month dry period was sufficient to cause transmission faults in the standard optical-fibre cable buried in certain shrink-swell soils. A dry period was defined therefore as one in which the monthly rainfall total falls within the second decile (i.e. within the lowest 20% of rainfall totals from the meteorological station closest to the fault). Autumn is a high-risk season in winter-rainfall areas because then soils will be at their driest. Failed autumn rains, after a typically dry summer, compound the problem. In summer-rainfall areas where rainfall fails, as has happened recently in an El Nino period, mid summer can be a time of high risk. Although initially, chemical ameliorants were considered as possible agents to overcome soil swelling, the labo-

ratory studies that were undertaken showed that although some of these were successful, they would have to be applied at high concentrations, which would be uneconomic and environmentally hazardous. The highly impervious nature of swelling clay soils, together with the huge volume of soil to be treated for effective control of swelling, means that chemical ameliorants are totally impractical. Furthermore, the different mineralogy of the layer silicates in the various swelling clay soils, ranging from smectite to interstratified smectite-kaolinite to illite, means that no single treatment would be fully effective across the wide spectrum of the swelling clay soils identified in Australia.

Shrink-swell indices Following the more detailed work a simple and easy-to-use index to represent the shrink-swell potential of each major soil type known to occur in Australia was developed (Fitzpatrick et al 1995a, b). The index, represented by numbers from 1 to 10, integrated pedological, mineralogical, soil physical and soil chemical data from characteristic groups of Australian soils. A risk rating of 1 indicates hard rock shallower than cable depth. Soils posing little risk of swelling are rated 2 to 5 (sand to loam soils), while those soils posing high shrink-swell risk are rated from 8 to 10 because they are clay-rich with gilgai and slickensides (Figure 3.1). Soils with a moderate risk of swelling are rated 6 and 7 and usually have sandy topsoils that abruptly over-

R. W. Fitzpatrick et

al.

LEGEND

Shrink-swell index 2-5 6-7 8-10 Saline wet area

Risk to cable LOW MODERATE HIGH Risk of corrosion to marker tape HIGH

• A Towns

I : /

Proposed new optical fibre route (standard cable) avoiding high risk area Proposed new optical fibre route (high strength,rodent resistant, cable)

Existing optical fibre route (standard cable)

Application of water

Use non-corrosive cable marker

Figure 3.2. Schematic example, showing the relationship between soils and a suitable optical-fibre installation linking towns A, B, C and D. Selection of new routes, cable types and marker types is optimised according to local environmental factors. The need for temporary amelioration (watering) is indicated.

lie clay-rich subsoils. A shrink-swell index was allocated to each of the 3063 soil landscape (map) unit descriptions on the 1:2 million Atlas Soil Map of Australia (Northcote et al 1960-68). Following this, a catalogue and a computer database of indices was produced to assist the assignment of shrink-swell indices to broad groups of soils along proposed cable routes (Fitzpatrick et al 1995b). Soils and the corrosion of detectable marker tapes Persistently wet, saline soils were found to cause the stainless steel in cable marker tapes to corrode. Consequently, a corrosive soils risk (i.e saline and/or wet soils) was allocated to each of the 3063 soil landscape (map) unit descriptions on the 1:2 million Atlas Soil Map of Australia

(Northcote et al 1960-68). A catalogue and a computer database of wet and/or salinity risk was built up so that the corrosivity of the broad soil groups along proposed cable routes could be easily found. Soil assessment manual A flow chart, called a Soil Assessment Key, was devised to link the shrink-swell indices of soils with climatic, vegetation and rodent risk factors so that decisions about the best cable routes and the best optical-fibre cable types could be derived for any route across the continent. The catalogue of indices for the 3063 map units, the Soil Assessment Key and also the soil, climate and rodent risk maps have been assembled in a Soil Assessment Manual (Fitzpatrick et al

Soil-related engineering problems 1995b). This manual plays a central role in the planning of optical-fibre cable routes because it includes practical stepby-step instructions and is illustrated by colour photographs or diagrams of important soil features, such as slickensides and gilgai, indicative of swelling (Figure 3.1).

Office assessment phase In any specific case, the actual decision-making process to be followed, commences with an office assessment phase to broadly define the soil shrink-swell risk indices and corrosive soils. This procedure uses maps from the Atlas of Australian Soils and the specially prepared Catalogue of Risk Index Ratings for shrink-swell and corrosive soils for each map unit (including a computer database). Like all maps at scales of 1:2 000 000 they must be used with caution. For example, fine detail such as small geographic features and local variations in soil type cannot be shown. The climatic risk map and a vegetation cover index are also used during this process. For example, when the office assessment phase indicates that soils with shrink-swell risk from 2 to 5 are dominant in the map unit, then a standard optical cable is recommended (Figure 3.2). However, if soils with risk ratings from 8 to 10 are dominant in the map unit, and the climate risk is high, then a high-strength cable is required (Figure 3.2). Similarly if the office assessment phase indicates map units with the dominance of wet and/or saline soils then non-corrosive marker tapes must be used.

Field assessment phase If an office assessment has indicated the presence of map units consisting of soils with risk ratings of 6-7, a field investigation must be conducted to establish their shrinkswell properties and also whether they are deep and extensive enough to influence the proposed cable installation. The Soil Assessment Manual shows how to go about this process by using field checklist sheets. Similarly if the office assessment phase indicates map units with minor wet and/or saline soils, field measurements of electrical conductivity must be conducted in order to establish the concentration of salts (soil corrosivity). If the electrical conductivity (EC) is >1 dS/m, the soil is saline and strongly corrosive. Potentially corrosive soils necessitate the use of corrosion-resistant cable markers.

Choosing best routes, best cable type and cable markers All methods and procedures described in the Soil Assessment Manual are simple and inexpensive. Savings resulting from the use of soil information are potentially available to Telstra in two ways. First, by choosing the best cable routes and second, by choosing the correct type of optical-fibre cable. Overuse of the high-strength cable can be avoided and the standard cable used wherever it is deemed safe: a saving of approximately 25% will result. For example, the integration of both the office and field assessment phases enables (Figure 3.2): (1) Troublesome shrink-swell soils along any proposed optical fibre cable routes to be avoided and the most

appropriate cable types to be selected. In the top half of Figure 3.2, between towns A and B, the small area with a high shrink-swell risk (8-10) can easily be avoided so as to allow the cheaper standard cable to be ploughed into the low shrink-swell risk soil area (2-5). (2) The most suitable and cost-effective cable markers (non-corrosive) to be chosen. For example, because of the corrosive nature of the saline soils in the area between towns A and B a non-corrosive cable marker must be used; however, a cheaper standard stainless steel marker tape can be used for the majority of the route. (3) Troublesome shrink-swell soils to be identified where the simple application of water will sufficiently prevent shrinkage about previously installed affected cables. For example, between Towns C and D in Figure 3.2, wetting the cable route with copious amounts of water or onset of significant rains reduces the incidence of faults because it reduces the grip which clays have upon the cable sheath (i.e. it helps the soil to become more plastic). Water is difficult to distribute to the depth of the cable in swelling clay soils and is considered to be a short-term ameliorant. Unfortunately, increased water infiltration may also cause greater shrink/swell at cable depth and encourage the growth of tree roots, which will increase soil shrinkage during dry periods. Hence the permanent solution is to use a suitable cable or if practical select alternative routes. (4) The most suitable approach to the reinstatement of fragile soils following cable burial.

CASE STUDY 2: METHODS FOR SOLVING SOIL-RELATED PROBLEMS CAUSING FAILURE IN THE WATER MAINS IN THE ST GEORGE AREA, SYDNEY Water-main failures have a considerable effect on the public because during most failures the water supply is disrupted for more than one hour. Not only are the residents of the community inconvenienced during this period of time but also essential services such as hospitals and sewerage systems are effected. Large retail and industrial complexes are also affected and can be financially disadvantaged when failures occur. In 1993 the concept of customer satisfaction in delivery and disposal of water products was introduced in the St George area (Figure 3.3). An unplanned interruption to water supply of more than one hour, or for six hours in the case of planned maintenance, therefore, would be a breach of the proposed standards leading to customer rebates on water rates. The maintenance section responsible would be accountable for this cost. During, 1993-1994, there were 120 main failures per month reported in this area. An average number across districts would be 20-30 failures (St George Water Area Report 1994). While in other water-supply areas in the Sydney region the number of failures experienced were reported to be as low as 5 per month. Thus the number of failures in the St George Water Area was far in excess of the average expected disruptions so it was considered important to determine why these failures occurred. Therefore, risk, inconvenience and economic impact are the consequences as exemplified by the problems with water supply in the St George Water Area of Sydney.

R. W. Fitzpatrick et al.

SOURCE; Chapman and Murphy (1989) Saff Landscapes of the Sydney 100000 sheet Soft Conservation Service of NSW, Sydney

COOKS

BOTANY BAY

1991-92 Main Failures..... 1992-93 Main Failures 1993-94 Main Failures WfANAMATTA SHALES....................... I, i Soil Landscapes - Sirrong - Blacktown - Glenorie - Lucas Heights . . ... HAWKESBURY SANDSTONE ... L J Soil Landscapes - Gymsa - Hawkesbury - Lambert

... . ....

* •

WINDBLOWN SANDS.............. Soil Landscapes - Newport - Tuggerah REMAINING CLASSIFICATIONS, Soil Landscapes - Disturbed Terrain - Narrabeen - Warnewood - E it along - Mangrove Creek - Hornsby

Figure 3 . 3 : The relationship between soil landscape map units and the distribution of water-main failures in the St George Water Area.

Soil-related engineering problems This case study explains the methodology used to identify soils criteria underlying the most-at-risk main failure areas. Although it was a preliminary study, the findings have provided a set of guidelines for future pipe laying so as to minimise the risk of failure. The study area is located on the east coast of New South Wales, south of Sydney. It has an area of 95 km 2 and is bounded on the east by Botany Bay and to the south by the Georges River. It has a population of 222 000 living in 87 500 properties. The St George Water Area provides a service not only to the residential areas but also to a large major hospital (St George District Hospital), numerous private hospitals and nursing homes, schools, technical colleges, large shopping complexes, entertainment venues, and also to light industry such as panel beating and engineering works.

Reasons for water-main failures The reasons for water-main failure are extremely complex. The age of pipes and excessive corrosion are just two factors but many systems outside the St George area are also old and corroded but have had fewer failures. In 1994 the main-break report for Cost Period Seven for the St George Water Area emphasised the extremely high main failure frequency during the hot dry summer period 1993-1994. At the same time a geotechnical report was independently presented on the problem of cracks in the Administration Building at the Wiley Park Depot which is located in the same area. This indicated that the clays at the site were very sensitive to changes in moisture content. Some brief comments were made suggesting that the increase in main failures for the St George Water Area may be due to changes in expansive clays after prolonged dry weather. The soil surrounding the water-main networks therefore needed to be investigated. Soil variability and the related soil properties are specific parameters for any water-reticulation system (Hay 1984). There is thus a great possibility of a high degree of variability in the soil type and its properties for any given network. Soil properties such as shrink/swell potential, corrosivity, pH and EC can have a variety of effects on water mains and therefore upon their ability to survive other damaging effects such as traffic vibration and temperature extremes.

Affect of soil properties on water mains The movement of expansive soils (i.e. shrink-swell potential) can induce longitudinal bending stresses on the pipes, which particularly affect the smaller (100-150 mm) pipes. The prewar pipes laid in the St George Water Area were typically of cast iron with rigid joints; they were chiefly laid without external protection of a sand surround. The rigid joints are unable to accommodate the movement induced by the stresses resulting from a reactive soil (Palmer & Hazelton 1994). Although the influence of soil properties on the corrosion of underground services is not well-documented (Hay 1984), many water-supply authorities have nevertheless, attempted to classify soils by their corrosivity towards buried metal structures. Pipe manufacturers rely on at least some form of soil test prior to pipe laying. Ferguson (1992) determined that there were two basic forms of corrosion of buried water pipelines, extrinsic and intrinsic (Appendix

3.1). Starkey (1945) reported that anaerobic (intrinsic) corrosion could cause pipe failure within a few years and a steel pipe with a wall thickness of 5 mm could be penetrated in 8 to 10 years. pH is another contributing factor, especially where bacteria are most active in the range of 6.5-7.5 with lower pH levels indicating advanced bacterial activity. EC or soil salinity also contribute to this condition. Other variables include leaking pipes and extreme weather conditions both of which can have differing consequences for the mains depending on the soil type. If for example a more reactive (shrink/swell) soil underlies an area where pipes are leaking then more stress will be put on the pipes, especially during prolonged high seasonal temperatures, and the pipes may fail (Palmer & Hazelton 1994).

Methodology To ensure a useful analysis of the controllable causes, it was necessary to determine which variables were potentially specific to the high water-main failure frequency of the St George Water Area, and those which are more common to other water-supply areas of Sydney (i.e. those areas that experience far fewer main failures). The factors determined as being exclusive to the St George Area were soil type, local weather conditions and pressure fluctuations. In order to determine the correlation between soil properties and mains failure, historical mains-failure data was geocoded in conjunction with soil landscape information described by Chapman and Murphy (1989) (Figure 3.3). A soil landscape is a homogeneous unit incorporating both soil and topography that can be represented on a map. As a result of soil variability, the combination of landscape and soil provides more information than a soil map alone. The landscape topography, the geology and soil type and properties were correlated with the main failures for 1991-1992, 1992-1993 and 1993-1994 and also for the cost period 7 and 8 main failures for the St George Water Area (1993-1994 financial year) (Figure 3.3). The data used in the analysis focused only on cast-iron water mains between 100 and 150 mm in diameter (Palmer & Hazelton 1994). Using the encoded data three specific sites were chosen for excavation. During the hot dry summer period the St George Area Mainbreak Report-Cost Period Seven in 1994 emphasised the extreme failure frequency. Pressure fluctuation, is also an exclusive catalyst for main failure in the St George Water Area. Maintenance records indicated that a high proportion of main failures occurred in the Wiley Park and Penshurst mixed-feed reservoir zones. In these zones, water is fed either by gravity from reservoirs or directly from pumping stations via major trunk mains. A potential source of excessive pressure fluctuation is introduced with the distribution of water directly into the system through these pumping stations. The combination and interaction between these variables were considered exclusive to the St George Water Area and so to its high frequency of main failure. Using the historical information about the failure of mains and soil-landscape information, the following three sites were chosen for investigation: Victoria Road (Punchbowl Site); Broad Arrow Road (Narwee Site); and Earlwood Avenue (Earlwood Site). (1) Victoria Road (Punchbowl)—(Luddenham soil landscape). A large proportion of main failures had been

R. W. Fitzpatrick et al.

reported for this site over the financial years 1991-1994, which included a dry summer (1993-1994). The main was at the boundary of Wiley Park and Penshurst mixed-feed reservoir zones. In both zones there existed a high probability for pressure fluctuations. The soil had been weathered from Wianamatta Shale and therefore had the potential to be corrosive and expansive. The main, because it had been excavated at numerous locations for repair, was known to be corroded in some of the areas. (2) Broad Arrow Road (Narwee)—(Luddenham and Blacktown soil landscapes). This site was selected because it represented a location with extreme potential for mains failure. The site was notorious for catastrophic failures over the financial years 1991-1994 and represented a large proportion of the St George Water Area main failures. The mains, having been excavated previously at numerous points for repair, were known to be in extremely poor condition because of excessive corrosion. The mains were at the boundary of the Penshurst and Wiley Park mixed-feed reservoir zones. Therefore pressure fluctuations would also be expected at this site. (3) Earlwood Avenue (Earlwood)—(Hawkesbury and Gymea soil landscape). This site had experienced a comparatively low frequency of main failures between 1991-1994 compared with the other two sites. It is located at the boundary of the Allawah feed and Canterbury-Marrickville gravity-feed reservoir zone. Pressure fluctuations were expected to exist in this location as with Victoria Road and Broad Arrow Road sites. However, very few failures had been reported in this area and any excavations had shown the mains to be in good condition with little or no corrosion. At each site the soil was sampled for laboratory analysis: (i) in direct contact with the main; (ii) from the trench wall; and (iii) 300 mm above the main. VICTORIA ROAD SITE (PUNCHBOWL) The location of the first excavation was on the mid-slope position, so no waterlogging was expected because of good runon and runoff. However, the soil was wet so there was a strong possibility that the mains were leaking. Two sites, where trees were not present, were excavated on the lower slope. The concrete footpath had cracked in places but this was not associated with cracked control joints or expansion joints in the pipes so uneven volume changes in the soil were probably responsible. Six samples were taken from the sites. BROAD ARROW ROAD SITE (NARWEE) Three sites, two of which had experienced recent failures were excavated at this location between mid and lower slope. There were Melaleuca spp. (paper barks) along the footpath in the vicinity of the excavation. Three samples were taken from the sites. EARLWOOD AVENUE SITE (EARLWOOD) Three sites were excavated, one where a previous main failure had occurred, one downslope from that location at approximately mid-slope and one on the lower slope. Waterlogging problems were not expected at these sites.

Leptospernum conifertus (brush box) was planted on the footpath. Three samples were taken from each site and tested for corrosivity and expansivity (i.e. liquid limit, plastic limit, plasticity index, linear shrinkage, moisture content and expansivity ratio).

Results Pipes laid in texture-contrast soils (i.e. soils with marked differentiation in clay content increasing from topsoil to subsoil) formed from Wianamatta Shale (Blacktown and Luddenham Soil Landscape) developed failures at a rate three times more than in soils formed from Hawkesbury Sandstone (Figure 3.3) (Palmer & Hazelton 1994). Soils derived from the Hawkesbury Sandstone have a low clay content, and therefore do not greatly shrink or swell. Furthermore, based on the assessment of the data, the average main failure density for any two cost periods (breaks/km2) was determined to be twice the failure rate for soils from Hawkesbury Sandstone parent material (Hawkesbury and Gymea Soil Landscape). Overall, on average 83% of the mains failures occurred in the Blacktown and Luddenham Soil Landscapes where the underlying geology unit is the Wianamatta Shale, which weathers to a soil containing smectitic clays (Figure 3.3). These have a reactivity approximately 2.5 times that of illitic and kaolinitic clays found in the soils formed from the Hawkesbury Sandstone. In the Blacktown and Luddenham Soil Landscapes, where 72% of the main length exists, 83% of the main failures occurred in each year between 1991 and 1994. However, in the soils weathered from the Hawkesbury Sandstone and occurring in lower slopes where the kaolinitic clay content was highest, some shrink-swell also occurs under conditions of extreme wetting and drying.

SUMMARY AND CONCLUSIONS The two case studies presented have demonstrated how movements caused by soil shrink-swell properties, and corrosion caused by saline soil solutions, damage optical-fibre cables and water pipes. The case studies illustrate how pedological information can be used to overcome such practical problems. The first case study outlines an approach demonstrating how pedological data and land-resource-assessment information was applied to solve a significant Australianwide, telecommunications problem in a practical and costeffective manner. Telecommunication transmission faults can develop in some types of optical-fibre cables buried in particular soil types with shrink-swell properties. Such faults are very costly to the operator and if avoided can save millions of dollars. Field and laboratory investigations were conducted on a representative range of deleterious soils known to cause faults in optical-fibre cables. Close liaison between soil scientists and Telstra engineers ensured that these research investigations were practical. For example, a simple 1 to 10 soil risk rating of soil shrinkswell risk can be derived logically by using a series of questions and answers set out in a manual entitled 'Soil Assessment Manual: A Practical Guide for Recognition of Soils and Climatic Features with Potential to cause Faults

Soil-related engineering problems in Optical Fibre C a b l e s ' . T h e m a n u a l describes practical surrogate m e t h o d s to estimate soil shrink-swell indices by using either p u b l i s h e d soil m a p s by office a s s e s s m e n t , or b y undertaking relatively simple visual observations of soil properties in the field. This a p p r o a c h h a s assisted Telstra engineers to plan optical-fibre c a b l e routes a n d avoid the damaging effects of certain soils. T h e m a n u a l h a s b e e n incorporated in Telstra's planning operations procedure m a n u a l a n d specifically focuses on how to use soil features (with prior a s s i s t a n c e from the Atlas of Australian Soils m a p s ) , a s well a s climatic a n d soil c h e m i c a l information to: (i) avoid problem soils along optical fibre c a b l e routes; (ii) minimise the u s e of heavier duty but expensive c a b l e types without risking the reliability a n d performance of the optical link; (iii) correct problems affecting c a b l e s previously installed in troublesome soil types; a n d (iv) rectify problems of reinstatement of fragile soils following c a b l e burial. T h e s e c o n d c a s e study demonstrates how pedological data o b t a i n e d from soil landscape a s s e s s m e n t is used to identify areas with the greatest risk for water main failure. A technical paper entitled 'Guidelines for Site Selection to Minimise M a i n Failure in the St George W a t e r Area' was produced for Sydney W a t e r technical staff to: (i) predict, for urban areas, soil types that potentially could damage water mains b y soil m o v e m e n t s or corrosion; (ii) specify in areas of future urban development the type of pipes a n d the most suitable routes to avoid problem soils; (iii) recommend, for existing areas, that pipes b e relaid in s a n d or replaced with more resistant pipes where problem soils are present; (iv) undertake site-specific soil surveys where existing soil data is unsuitable or not available; a n d (v) extrapolate the methodology to other areas that were affected by frequent m a i n s failures.

Interpretation Manual, pp. 55-69. CSIRO, Melbourne. FITZPATRICK

W.,

WRIGHT

].,

SLADE

AL.

1995a.

Compendium of Technical Reports. R & D studies of fault occurrence in optic fibre cables buried in swelling clay soils and techniques for stabilising these soils: a collaborative research project with Telstra: July, 1995. Confidential Report to Telstra.

FITZPATRICK R . W . , W R I G H T M . ]., SLADE P. G . , HOLLINGSWORTH I . D . &

PETER P. 1995b. Soil Assessment Manual: a practical guide for recognition of soils and climatic features with potential to cause faults in optical fibre cables. Confidential Report to Telstra. GRANT K. 1968. A terrain evaluation system for engineering. CSIRO Division of Soil Mechanics Technical Paper 2. GRANT K. 1971. Terrain evaluation as a basis for engineering geology. CSIRO Division Applied Geomechanics Research Report 172. GRANT K. 1973. The PU.C.E. Programme for Terrain Evaluation—1. Principles. CSIRO Division of Applied Geomechanics Technical Paper 15. GRANT K. 1974. The P.U.C.E. Programme for Terrain Evaluation—II. Procedures for Terrain Classification. CSIRO Division of Applied Geomechanics Technical Paper 19 HAY L. 1984. The influence of soil properties on the performance of underground pipelines. MAgSc thesis, University of Sydney, Sydney (unpubl). HAZELTON P. A. & MURPHY B. W . (Editors) 1 9 9 2 . W h a t do all the

Numbers Mean?—A Guide for Interpretation of Soil Test Results. Department of Conservation and Land Management, Sydney. ISAACS D. W. 1950. Foundation soils in Australia. Commonwealth Experimental Building Station, Department of Works and Housing, Melbourne, Duplicated Document 31 (unpubl.). ISBELL R. F. 1996. The Australian soil classification system. CSIRO, Melbourne. JONES D. E. & HOLZ W. G. 1973. Expansive soils—the hidden disaster. Civil Engineering 43, 49-51. LEEDOM L. M. 1946. Graphic corrosion of cast iron. Journal of American Works Association 38, 1392-1396. MCDONALD R . C . , ISBELL R . F., SPEIGHT J . G . , WALKER J . & HOPKINS M .

S. 1990. Australian Soil and Land Survey Field Handbook (2nd Edition). Inkata Press, Melbourne. NELSON J. D. & MILLOR D. J. 1992. Expansive soils: problems and practice in foundation and pavement engineering. John Wiley & Sons, New York.

NORTHCOTE K . H . , BECKMANN G . G . , BETTENAY E . ETAL.

1960-68.

Atlas

of Australian Soils (Sheets 1-10 with explanatory booklets). CSIRO, Melbourne.

PALMER B. A. & HAZELTON P. A. 1994. Soil investigation for St George

ACKNOWLEDGMENTS W e are grateful to several engineers in Telstra for their assistance. Scientists a n d technical staff in C S I R O Land a n d W a t e r who contributed substantially to these studies include M a l c o l m Wright, Ian Hollingsworth, Paul Peter, Brian Richards, Phil Davies, D a m i a n M o w a t , Philipa Butterworth, Greg Rinder a n d David Wright. W e thank Stuart Murray for drafting Figure 3 . 3 .

REFERENCES AITCHISON G . D . , SPRIGG R . C . & COCHRANE G . W . 1 9 5 4 . T h e s o i l s a n d

geology of Adelaide and suburbs. Geological Survey of South Australia Bulletin 32. BOUMA J. 1997. The role of quantitative approaches in soil science when interacting with stakeholders. Geoderma 78, 1-12. BRASS K. 1993. The phone goes bush. Australian Geographic 29, 56-74.

CHAPMAN G. A. & MURPHY K. C. 1989. Soil Landscapes

of the

Sydney

1:100 000 Map Sheet. Soil Conservation Service of New South Wales, Sydney.

FANNING D . S. & FANNING M . C. B. 1 9 8 9 . Soil: Morphology,

Genesis,

and Classification. John Wiley and Sons, New York. FERGUSON P. 1992. Corrosion in buried water pipelines. Corrosion Australasia 17(4), 7-10. FITZPATRICK R . W . , M C K E N Z I E N . J . & MASCHMEDT D . 1 9 9 9 . S o i l m o r -

phological indicators and their importance to soil fertility. In: peverell K., Sparrow L. A. & Reuter D. J. eds. Soil Analysis: an

Water Area Main Failure Analysis. In: Soils in the City 2nd Conference, University of Melbourne, Parkville, December; 1994. PRESCOTT J. A. 1944. A Soil Map of Australia. CSIRO Bulletin 17.

SHEARD M. & BOWMAN G. B. 1996. Soils, stratigraphy and engineering

geology of near surface materials of the Adelaide Plains. Department of Mines and Energy South Australia Report Book 94/9.

SOIL SURVEY DIVISION STAFF 1 9 9 3 . Soil Survey Manual.

U S Department

of Agriculture Handbook 18. SOIL SURVEY STAFF 1999. Soil Taxonomy: A Basic System of Soil Classification for Making and Interpreting Soil Surveys (2nd edition). US Department of Agriculture Handbook 436. STARKEY R. L. 1945. Transformations of iron by bacteria in water. Journal of the American Water Works Association 37 pp. 963-967. STEPHENS C. G. 1962. A manual of Australian soils (3rd edition). C S I R O , Melbourne. ST GEORGE WATER AREA REPORT 1 9 9 4 . Report for the period on the high

frequency failure Nov 1993-Jan 1994 for the St George Water Area (unpubl.).

TAYLOR J . K . , THOMPSON B . P & SHEPHERD R . G . 1 9 7 4 . T h e s o i l s a n d

geology of the Adelaide area. Geological Survey of South Australia Bulletin 46.

VON WOLZOGEN KUHR C . A . H . & VAN DER VLUGT L . S .

1934.

The

graphitisation of cast iron as an electrobiological process in anaerobic soils. Water 18, 147-151.

WALSH P. F., HOLLAND J. E. & CAMPBELL K. D . 1 9 7 8 . Footing a n d s l a b

design in accordance with the new Victorian uniform building regulations. CSIRO, Division of Building Research Report, Melbourne.

Received 4 February 1999; accepted 9 June 2000

R. W. Fitzpatrick et al.

APPENDIX 3.1: GLOSSARY OF TERMS Distortion An undesired change in the shape of an electrical wave or signal Distortion results in the loss of clarity in reception or reproduction, or even the loss of information in a digital system. Extrinsic corrosion Extrinsic factors such as electrolysis, stray current and galvanic cell corrosion are more universal to water reticulation networks than localised soil effects on corrosion. Fibre optics (fibre optic communications) The transmission of information by the passage of light through flexible, glass fibres. Electrical impulses are converted into light, which is then transmitted through the optical fibre. The light is then reconverted into electrical impulses at its destination. Gilgai Gilgai (an Aboriginal word meaning small waterhole) are surface features consisting of a pattern of alternating mounds and depressions with a maximum difference in vertical interval of about 2 m (Figure 3.1). Water frequently ponds in depressions thereby helping to identify the presence of gilgai. Prominent shrinkage cracks occur in dry seasons. This microrelief is produced by swelling clays following prolonged expansion and contraction due to changes in moisture content; usually a succession of microbasins and micromounds in nearly level areas, or of microvalleys and microridges parallel to the direction of the slope. Intrinsic corrosion Intrinsic corrosion occurs when water-pipe mains are laid in anaerobic soil conditions. The bacterial activity in these soils occurs with little or no oxygen. Corrosion of cast iron under these conditions is characterised by a process known as graphitisation when iron ions migrate from the metal matrix leaving behind a high carbon content. This deposit becomes softer as the carbon content becomes higher (Von Wolzogen Kuhr & Van der Vlugt 1934 ) Outwardly the pipe may retain the appearance

of being in a good condition, however, its strength is dramatically reduced (Leedom 1946). Marker tape A stainless wire or other non-corrosive transponder buried above the optic fibre cable to help locate it. Optical fibre A glass thread that acts as a guide for light waves. Fibres used in telecommunications usually have a cladding of glass of a lower refractive index. In a communication system, several fibres are made up into a cable. Some 10 billion digital bits can be transmitted per second along an optical-fibre link in a commercial network, enough to carry tens of thousands of telephone calls. Hair-thin fibres consist of two concentric layers of high-purity silica glass—the core and the cladding—which are enclosed by a protective sheath. Light rays modulated into digital pulses with a laser or a light-emitting diode move along the core without penetrating the cladding. The light stays confined to the core because the cladding has a lower refractive index—a measure of its ability to bend light. Slickensides Natural shiny surfaces found on soil aggregates formed by the parallel orientation of clay particles during swelling and shrinking cycles (Figure 3.1). Refers to polished or grooved surfaces in rocks or soils resulting from part of the mass sliding or moving against adjacent material along a plane which defines the extent of the slickensides. In soils, they only occur in clay-rich materials with high swelling clay content. Soil salinity (EC) Soil salinity refers to the amount of dissolved salts in the soil solution. Soil salinity is commonly estimated by measuring the electrical conductivity (EC) of a soikwater extract or suspension because the EC is directly related to the electrolyte concentration of the extract or suspension. Soils with a very high salt content have a correspondingly low saturated resistivity. The lower the soil resistivity the greater the potential exists for free electron exchange between the water main and its soil environment. There is thus more likelihood of corrosion.

Geological Society of Australia Special Publication 21, 37-47

CHAPTER 4—Avon River palaeodrainage system, Western Australia: geomorphoiogical evolution and environmental issues related to geology ML 4. FREEMAN Land Access Branch, Department of Minerals and Energy; 100 Plain Street East Perth WA 6004, Australia (m. freeman@dme. wa.gov. au). The landforms of the Avon Palaeodrainage have been inherited from long-term weathering and erosion of the Archaean Yilgam Cratoa probably initiated during Permian times. Since those times, it is concluded that the Yilgarn has been an area with low surface gradients and relatively low relief, contributing some argillaceous sediment but little arenaceous sediments to surrounding basins. The Avon River previously drained an area of about 100 000 km2, from some time after the Permian until after Eocene times, in contrast to the present drainage system with a catchment of about 13 000 km2. During Pliocene times, it is inferred that uplift of the western margin of the Yilgarn Craton dammed the Avon River, creating a hugely expanded Yenyenning Lakes leading to the deposition up to 70 m of lake sediments. Later, a new channel was cut through the uplifted terrain, immediately east of the Darling Fault, some 50 km north of the previous valley position. Subsequently the rainfall decreased, the valleys became clogged with alluvium and, in the eastern 90% of the system, the river stopped flowing. The river channels are now replaced by lines of playas, and are the focus for evaporation of the exposed groundwater, concentrating and precipitating salts, principally gypsum. Gypsum, in the form of lunettes or arcuate dunes on the southeastern shore of the playas, is in demand for applying to farm soils to improve their texture and improve productivity. However, the mining is accompanied by clearing of the native vegetation which, on these lunettes, is commonly a significant proportion of the little amount of remnant vegetation in the region. Since settlement, the Wheatbelt Region has been extensively cleared for farming, leaving minimal areas covered with native vegetation. Many of the nature reserves in the Wheatbelt Region occur along the palaeodrainage channels where the saline soils were less attractive for farming, and now the mining of the gypsum from these nature reserves can be seen as a further threat to the conservation of what remains of the native vegetation. The replacement of the native vegetation throughout the Wheatbelt Region by perennial grasses with short growing periods has disrupted the former hydrological balance by reducing interception of rainfall and evapotranspiration losses, raising groundwater levels in the weathered bedrock and overlying alluvium. Where the level rises until groundwater saturates the near-surface soils, increased evaporation results and salt is deposited at and near the surface. This waterlogging of the soils renders once-productive farmland saline and non-productive, and is estimated to be increasing in area in the order of several hectares per day. The land clearance will lead to a new hydrological equilibrium becoming established, when the increased rainfall infiltration is balanced by the total loss from evapotranspiration, evaporation and active discharge of saline groundwater in low-lying parts of the terrain. The groundwater will migrate downslope towards the palaeodrainage valley floors where the rising groundwater levels cause, more prolonged flow of springs into the playas and, at worst, permanent flooding of playas and waterlogging surrounding land. These areas are the focus of many of the Wheatbelt Region nature reserves, leading to a grim prognosis for their long-term ability to protect the remnant native vegetation. There are, therefore, two significant environmental concerns in the Wheatbelt Region that owe their causes, at least in part, to the present geology and the geological history. These are: (i) gypsum mining from lunettes next to the playas, which is favoured as being the cheapest source of farm gypsum—therefore, the very activities that depleted the native vegetation in the Avon Palaeodrainage are now creating demand for gypsum mining in some remaining areas of native vegetation, thus potentially adding to the loss of native vegetation; and (ii) rising groundwater levels caused by the disruption of the previous hydrogeological equilibrium through removal of native vegetation, which poses an immediate and long-term threat to farm activities in lower areas and also threatens the value of conservation areas that are concentrated in the palaeodrainage valleys. KEY WORDS: Avon River, environmental management, geomorphology, groundwater, gypsum, palaeodrainage, Yilgarn Craton.

INTRODUCTION Each August the Avon River becomes Mecca for a flotilla of small water-craft, where the water roars through the Darling 'Plateau' east of Perth on

Australia's western seaboard. From Northam to the Swan Coastal Plain the crews race each other, vying for the honours of being the first to finish - or just to finish. The crews exert themselves on the water draining from an area of some 13 000 km2. Following good

M. J. Freeman winter rains, the crews can be pleased to handle their craft on flows of over 50 000 000 m3/day. However, it could be asked 'What would be the response of the crews if they had water actively draining from an area of some 100 000 km2, or some eight times the present catchment?'. 'A bit hypothetical', you say, 'the Wheatbelt Region is too dry and there are no river channels draining such an area\ Well, that may be the situation now, but it was not always the case. The Avon is now a mere tiny remnant of its earlier magnificent glory; a vestigial river near the mouth of a previously huge system.

The Wheatbelt region in the southwest of Western Australia is now a plateau, with low hills rising from a few tens to over a hundred metres above a broad, gently undulating plain or the nearest valley. Overall the terrain ranges over altitudes of 200-500 m above mean sea-level over distances of 400 km north-south and east-west. The western margin of the plateau is marked by the Darling Scarp, representing the Darling Fault, a major crustal feature of the Earth, extending north for about 1000 km from the south coast. This fault has a history of episodic movements, separated by long quiescent periods during the past 2000 million years. The fault probably experienced large-scale movement during Eocene times that may have extended into the Pliocene. Rainfall decreases from a maximum of about 500 mm/y along the western edge of the Wheatbelt region to a minimum of about 250 mm/y in the northeastern part. The rain dominantly falls in the cooler May-September period, with the remainder of the year receiving limited falls but having relatively high evaporation rates. However, erratic summer rainfall can be torrential, resulting from rain-depressions derived from distant tropical cyclones. These degraded cyclonic systems produce highly intermittent high-intensity wet periods that can produce flooding in inland surface water systems, particularly in the eastern parts of the Wheatbelt region. A number of large-scale networks of former major valleys dominate aspects of the geomorphology of the western half of Australia. Clearly evident when flying over the Wheatbelt and by examination of maps and satellite images, these were initially recognised from regional geological mapping and on regional contour maps of the land surface (van de Graaff et al 1977; Wilford & Truswell 1990). The valleys are referred to as palaeodrainage valleys or systems, and occur west from the Stuart Range in central South Australia and west from the Western Australia-Northern Territory border. Beard (1973) related the palaeodrainage systems to the vegetation, and used vegetation mapping to define the systems widely in Western Australia. These features are of such a scale that they are not readily apparent on the ground, merely being areas of dry salt lakes, uninteresting, even boringly flat with few of the features of an attractive landscape, and which can be overpowering and threatening when seen during searing summer heat. However, the interpretation of them provides a rich history of changes to the land surface and the climate of Western Australia over tens of millions of years. In the Wheatbelt region of Western Australia, the former integrated drainage of the Avon River system is now com-

monly marked by lines or areas of playas or dry salt lakes, and in places by some ill-defined watercourses that flow very intermittently into larger playas and then stop. These playas are rounded and shallow with circular, oval or more complex though still rounded outlines. Playa surfaces range from glaringly white, where gypsum, common salt or other salts have crystallised, to pale- to dark-brown and some are smooth while others have rough surfaces like ploughed paddocks where the surface has been uplifted and distorted by the crystallisation of salts just under the surface. A sparse growth of low (<20 cm), scattered plants that are salt-tolerant (halophytic), fringe some playas or in drier sites grow across the floor. Some of the playas have valuable mineral deposits. Access to these has led to debate because extraction requires the removal of any native vegetation growing on the gypsiferous lunettes. This chapter of the book dealing with environmental geological issues describes some background environmental geological aspects of the Avon Palaeodrainage. It is a region where change made to the pre-European vegetation has produced disequilibrium in the hydrological systems. A new balance will be reached in time, but over a much longer duration than normally conceived to be acceptable by humans. As well as taking a long time, the scale of the ultimate responses of the natural processes that will eventually take over, are difficult to predict. This will lead to the situation where those wanting access the Earth's resources confront those wanting nature conservation and protection of existing environmental issues, and with the mining needed for maintenance of farming productivity, there are extremely complex decision-making processes.

OUTLINE OF THE PALAEODRAINAGE SYSTEM This chapter specifically focuses on the Avon Palaeodrainage and its previous greater active extent, although to separate its description from adjacent, and probably genetically related, systems may be regarded as artificial by some. The extent of the Avon Palaeodrainage is shown on Figure 4.1. Former continuity of the valley floors and water flow is inferred, based on geological mapping, water-bore and geological drilling, ground-level information and geological reasoning. The system now consists of segments dominated by playas or groups of playas, some of which are interconnected, and with some tributaries with higher gradients providing intermittent run-off into the main channel that has a lower gradient. The Avon Palaeodrainage covers a large proportion of the Wheatbelt region of Western Australia, which is economically very important to Australia because of the large quantity of cereal crops it grows. Mulcahy and Bettenay (1972) coined the term Southwest Division, consisting of all drainages west of a north-trending line about 100 km west of Kalgoorlie and south of an east-trending line through Geraldton (Figure 4.1). Bettenay and Mulcahy (1972) subdivided this division into the Avon and a more northerly Monger Palaeodrainage and included all the many currently active and still-flowing coastal rivers. Within the Avon Palaeodrainage, they identified major tributaries as shown in Figure 4.1. The Ninghan Palaeodrainage was included by those authors as a tribu-

Avon River Palaeodrainage system

AVON RIVER PALAEODRAINAGE SYSTEM SOUTH WEST WESTERN AUSTRALIA

3ERALDTON

Yarra Yarra i Lake

\SWAN COASTALl/^NJORTHAM PERTH

LAIN INDIAN

OCEAN

COLLIE

/ALBANY

100 km

Active perennial watercourse Approximate boundary of palaeochannel

Land and Resources Access Branch Department ol Minerals and Energy

Oct 1999

PJaya lakes

d:\dgn\mjf\paleogeo_lat

Figure 4.1 Avon River palaeodrainage system, southwest Western Australia.

Approximate limit of Avon palaeochannel system Present catchment boundary of Avon River active system

M. J. Freeman

tary of the Avon Palaeodrainage. Examination of groundsurface levels and of regional geological maps does not necessarily support this model, and there is a strong possibility that the Ninghan Palaeodrainage may have connected to the more westerly Monger Palaeodrainage. The Avon Palaeodrainage immediately east of Brookton, and upstream from the presently active Avon River, is referred to on topographic maps as the Salt River. However, there are two other rivers with the same name on topographical maps within 50 km: one a tributary of the Mortlock River, East Branch and the second a tributary of the Mortlock River, North Branch. Clarification of this confusing terminology is clearly needed. Salama (1997) described the course east of Brookton as extending southwest from about Baandee for some 150 km to the Yenyening Lakes, adjacent to the present Avon River. This downstream part of the palaeodrainage system has a better defined path and flows more regularly than other parts further upstream, contributing to problems in the flowing Avon River because its name is quite apt—adding to the salt load in the present Avon River.

GEOMORPHOLOGICAL SETTING Palaeodrainage systems that were apparently active during the Cenozoic and are considered to have carried large amounts of run-off are common on the Australian landscape. Van de Graaff et al (1977) gave a synoptic view of the palaeodrainage systems of Western Australia. Pitt et al (1978) described them in South Australia, and Freeman (1986) described large alluvial fans in central Australia created by existing rivers that carried far larger quantities of water than is conceivable under present climatic conditions. Few detailed studies of the geology of the Avon Palaeodrainage have been undertaken, especially given the large area dominated by the feature and the agricultural value of the region, although because of increasing concerns regarding the effects of saline groundwater there is an increasing number of studies related to this matter. Mulcahy (1967) and Churchward (1969) described parts of and drew conclusions relating to those parts of the upstream palaeodrainage system that was inactive, but did not attempt to extend their interpretations to encompass the wider area of the total palaeodrainage. Salama et al (1992) discussed the evolution of the salt and sediment at Lake Deborah, concluding that accumulation post-dated the Eocene (<40 Ma), while Salama and Hawkes (1993) and Salama et al (1993) investigated three small subcatchments (up to 250 km2 each), concluding that an intricate and integrated network of relict channels extended over much of the Avon Palaeodrainage and contained thin sedimentary sequences. More recently, Salama (1997) described sediments in the Salt River portion in some detail, as elaborated upon later. In contrast to the Avon Palaeodrainage, the economic implications of the palaeochannels in the Goldfields as sources of groundwater for mining-related projects has resulted in more studies being completed there, including several by Clarke (2001). The present Avon River, as most people would recognise it, consists of an active channel, flowing west through a relatively youthful valley for some 90 km from Northam to the Darling Scarp (typically 150 m deep and 3 - 1 0 km across;

average gradient of 1.4 m/km). Downstream of this section is the Swan estuary extending 65 km from the Darling Scarp before entering the ocean at Fremantle (gradient <0.15 m/km). Upstream of the youthful section, the river flows along a broad, shallow valley for about 100 km in a northerly direction (<100 m deep and 20 km across; gradient 0.6 m/km). The implication of having these three sections (an upstream low gradient; a middle steeper gradient; and a lowermost nearly horizontal gradient) is that the present youthful valley is the result of rejuvenation of erosion caused by uplift of the western edge of the Yilgarn Craton through movement along the Darling Fault. However, there is a very large area to the north, east and southeast of Northam where parts of the watercourses only flow in the wettest seasons and where most of the drainage is dismembered and partially blocked. This is the palaeodrainage system of the Avon River. It would be highly speculative to indicate the likely flow rates of the ancestral Avon River, but the sheer dimensions of the former system suggests that the flow rate was very large. In general terms, it is likely that the river gradients were low and water movement sluggish. However, Salama (1997) noted that the basal palaeovalley sediment in his Yenyening Lakes investigation area is arenaceous implying that in the Pliocene this section was experiencing moderate flow rates. He calculated that the flow in this ancestral Salt River ranged between 80 m 3 /s and 17 000 m 3 /s and that the river gradient ranged 0.3 to 0.7 m/km within a valley similar to that of the present-day one downstream of Northam. If this was the situation, it presents difficulties with the interpretation of the Yilgarn having little relief since Eocene times and conflicts with the river gradient being similar to that of the present river upstream from Northam. Offshore from the river mouth is one of the largest submarine canyons in the western half of Australia, the Perth Canyon (von der Borch 1968). This can be traced for some 150 km from the continental shelf and onto the abyssal deep, and is up to 15 km wide and 3 km deep. It is suggested that for this to be eroded by turbidity currents would require the input of far larger quantities of detritus than the present river-flow carries. Of interest is the fact that the Perth Canyon does not extend onto the continental shelf, and the cause of this is considered further in this chapter. The actual genesis of the boundaries of the palaeodrainage system is not clear, and possibly was inherited from the time of the last significant tectonism to have affected part of the basement in the Neoproterozoic. Beard (1998) has investigated the central watershed that forms the eastern boundary of the system and concluded that it had stabilised by the mid-Cretaceous.

GEOLOGICAL SETTING Basement geology The basement underlying the Avon Palaeodrainage and the Darling Plateau is the Yilgarn Craton. This craton, underlying an area of some 660 000 km 2 , is one of two 'ancient shields' of the continental plate of Australia; the other being the Pilbara Craton to the north of the Yilgarn. These cratons are the oldest parts of the Australian continent and have

Avon River Palaeodrainage system been stable areas of rocks that have undergone no major changes for the past 2000 million years. They are dominated by plutons of granite up to 50 km across, felsic-gneissic terranes, and linear, generally north-northwest-trending greenstone belts of wqakly metamorphosed mafic and felsic volcanic and sedimentary rocks. The majority of the rocks were mostly formed between 3300 and 2600 Ma, along with the gold and nickel mineralisation for which the craton is renowned. Widespread tectonic activity was last recorded at about 2550 Ma, with an extensive mafic dyke swarm emplaced at 2400 Ma. A major tectonic overprinting at about 1100 Ma occurred along the southern margin. This, the latest tectonic events known to have affected any part of the Yilgarn Craton, was a continental plate-plate collision in which the Albany-Fraser Orogen rode northwesterly over the margin of the Yilgarn Craton. It is likely that this would have produced an extensive range of very high mountains from which erosion would have contributed large quantities of sediment to form, initially, a thick wedge of alluvium at the northern foot of the ranges on the Yilgarn. However, subsequent erosion has removed all traces of these sediments and even the core of the ranges, leaving only vestigial highs in, for example, the Stirling and Porongurup Ranges. There is evidence of only minimal epeirogenic or similar activity affecting the Yilgarn Craton since before Permian times. Most recently, the Yilgarn Craton has been ruptured by a number of relatively young, low, fault scarps, scattered throughout its area. The age of the oldest of these scarps is not known, although because of their morphology it is suggested that they are less than - 1 0 0 thousand years. The most important recent examples of these faults resulted in the town-destroying earthquake affecting Meckering in 1967 and the lesser Calingiri earthquake in 1974, both in the western part of the Yilgarn. Other than the young faults noted above, Cope (1975) hypothesised that the Ravensthorpe Ramp, located some 50-100 km north of the south coast may have developed relatively recently, and Commander (1999), reported indications that broad, gentle, northerly uplift has occurred. There is some evidence that could support a hypothesis that more localised, gentle updoming has occurred locally, as described later in this chapter. The granitic basement of the Yilgarn Craton is exposed as scattered bornhardts, or domes and rock outcrops, commonly culminating in striking landscape features. Twidale et al (1999) described some of these near Kellerberrin, as well as more widely, considering their genesis, identifying a multistage development and stating that the first development could have been as early as the Eocene. Many are economically important having been developed to collect rainfall by channelling run-off into tanks for reticulation to local communities in a region where there are no other potential natural potable water sources.

Phanerozoic geology Material eroded off the craton could be expected to be deposited in any sedimentary basins that existed around it at the time of the erosion. There was limited deposition of any sediment derived from the craton in the basins surrounding the Yilgarn Craton. The Perth Basin, to the west,

contains large thicknesses of sediment deposited during the past 200 million years, but new information actually implies that there was almost none derived from the Yilgarn. It is likely that a continental glaciation during Permian times (290-280 Ma: Brakel 1990) planed off the Yilgarn Craton surface before depositing sediments on it and in basins around the edge. Two adjacent basins (Perth Basin to the west, Officer Basin to the east) and one basin overlying the craton rocks (the Collie Basin in which sediments are preserved in a post-depositional graben) contain sediments deposited during Permian times after cessation of the glaciation. Wilson (1989) reported that the Collie Basin siliciclastic sediments represent braided stream channels and bars, with palaeocurrents dominantly towards the northwest, supporting a model that these sediments were widespread on the craton and not just restricted to the proximity of the basin. The Collie Basin sediments are probably remnants of former much thicker and extensive deposits. Le Blanc Smith (1993), using coal vitrinite reflectance data, interpreted that about 6500 m of cover was removed from the outcropping sediments, and deduced that the total thickness of 8 km of sediment deposited in the Collie Basin was similar to that in the southern part of the Perth Basin to the west. However, the Perth Basin thickness includes postPermian sediments. Kohn et al (1998), using fission track data, have interpreted that significant uplift of the Collie Basin sediments accompanied by unroofing the remaining rocks through removal of up to 2 km of sediments occured before Mesozoic times. Thus the thickness of Permian sediments in the Collie Basin on this basis was up to 3.5 km initially. It leads to the hypothesis that Permian sediments were deposited more widely across the Yilgarn Craton than just in the basin (Le Blanc Smith 1993), but then much eroded in a relatively short time. This is in conflict with Le Blanc Smith's (1993) contention, which implies that sedimentation continued until Cretaceous times, as occurred in the nearby Perth Basin. The source of these Permian sediments has not been determined. The Permian sedimentary rocks in the Collie Basin are overlain by a thin sequence of claystone to conglomerate of the Nakina Formation. Backhouse and Wilson (1989) dated these using palynomorphs as of Early Cretaceous age, and believed them to be equivalent to the Leederville Formation, a widespread unit in the Perth Basin. Similar equivalents occur near the edge of the craton at Donnybrook. Therefore, some deposition of sediments occurred over the western parts of the Yilgarn Craton following Permian times. Evidence of dating detrital zircons by Sircombe and Freeman (1999) suggests that there has been little erosion of the Yilgarn since at least Early Cretaceous time (130 Ma), or at least that contributed arenaceous sediments to the Perth Basin. Cawood and Nemchin (2000) have obtained similar results for a profile through the northern part of the Perth Basin. They concluded that the Perth Basin sediments were not derived from the Yilgarn Craton implying that the Yilgarn has been an area of low topography since just after Permian times or since ca 200 Ma. Based on the zircon ages, it is likely that the Perth Basin sediments were derived from Albany-Fraser Orogen and

M. J. Freeman

possibly Antarctica and India, both of which were adjacent to parts of the Perth Basin for considerable portions of its existence.

Geology of the palaeodrainage system Hocking (1990) reported that initial fluvial and lacustrine sediments in the palaeodrainage valleys east and south of the Avon were deposited during Eocene times and ceased during Oligocene times. Some of these sediments are organic-rich, locally becoming lignitic or grading into rocks similar to oil shales. Van de Graaff et al (1977) suggested this deposition was correlated with increasing aridity in the interior of Western Australia, arguing that prior to this the river flows were sufficiently strong to transport the detritus to the ocean out of the rivers, and not allowing any to accumulate. Following the Eocene fluvial deposition, a transgression resulted in marine incursions along the palaeodrainage channels, in some extending possibly hundreds of kilometres inland, with deposition of marine and estuarine sediments. There are no reports of marine incursions into the Avon Palaeodrainage, in contrast to other palaeovalleys to the south and east. However, there are suggestions that fluvial Eocene sedimentation was widespread over areas near but outside the Avon Palaeodrainage system, possibly as an extensive, thin sheet that has been largely eroded. For example, Hill and Merrifield (1993) described a diverse macroflora of Eocene-Oligocene age from the summit of the West Dale drainage. The widespread distribution of three probably equivalent units along the western portions of the Yilgarn Craton, the Kojonup Sandstone, the Kirup Conglomerate and the Harvey Formation, all possibly of Eocene age, may represent a basal, coarse-grained sediment of the extensive sheet related to widespread outflow of rivers from the Yilgarn Craton (Hocking & Cockbain 1990). Still further south, Eocene sediments were deposited in the Beaufort River valley as a palaeodrainage system (Waterhouse et al 1994). These deposits represent deposition in a low-relief, fluvial to lacustrine palaeoenvironment implying that the region may have had a low gradient for a considerable time with even less relief than now because a number of these outcrops occur on top of the plateau. A significant deposit possibly derived from erosion of the craton was identified by Playford et al (1976), who noted that the Kings Park Formation fills the continental shelf parts of the Perth Canyon where this had presumably been cut into the Cretaceous rocks of the Perth Basin. Salama (1994) enlarged on this and emphasised the interpretation that this formation constituted sediment deposited in the ancestral delta of the Avon River immediately offshore from Perth. This formation extends over some 14 500 km2 and is of Late Paleocene-Early Eocene age. He inferred that the volume of sediment was some 2.4 x 10 1 2 m 3 , equivalent to erosion of 25 m from the catchment area of nearly 100 000 km 2 , and this implies that there was erosion of -1.7 m/10 6 y for about 15 million years. This appears to be a large amount of erosion for such a large area with little relief and low gradients and the implications of Salama's interpretation require further research. Salama (1997) reported on a significant investigation into the palaeodrainage system along the Salt River.

Drilling of a number of cross-sections, coupled with shallow seismic surveying, electric and magnetic studies led to a clear picture of the nature of the sediments and shape of the buried valley. Up to 70 m of sediment occurs, with channel sands interbedded with overbank and freshwater fine-grained lake deposits. He hypothesised that the river was actively flowing in a valley up to about 140 m deep, similar in relief to the existing Avon River. He further hypothesised that the Kings Park Formation sediments were transported through an ancestral valley almost due west of the end of Salt River until the Darling Range uplift commenced prior to the deposition of the Salt River sediments. This uplift probably represents the latest large-scale movement along the Darling Fault, although P. E. Playford (pers. comm. 2000) has raised the possibility that some scarp-related features suggest that activity may not have ceased totally and that highly intermittent uplift may still be occurring. If this is correct, it would imply that Perth is not immune from very intermittent, potentially severe, local earthquakes. The river valley became blocked by this uplift in Pliocene time (or Miocene at the earliest), causing lacustrine sediment to accumulate in an extensive lake east of the present locality of Brookton. Eventually the water escaped through the range along the path of the present youthful valley, and some 80 km north of the former path. The presence of this large lake-filling of sediment shows that the Salt River was still receiving significant flows from upstream, contrary to the suggestions of other authors that the system became arid in Late Eocene times (37-33 Ma). The thickness of the palaeovalley sediments has been well-documented in palaeovalleys in the Eastern Goldfields in the search for industrial groundwater supplies (Commander et al 1991). In that area, the valleys contain up to 130 m of sediment. However, limited information is available from the Avon Palaeodrainage. Salama (1997) reported 70 m of sediment below the Yenyening Lakes. Water bores into the sediments usually do not penetrate through to the underlying rocks, although scattered holes do, but commonly the logging of the thickness of the sediments penetrated is unreliable. Examination of bores in the palaeodrainage system around the town of Hyden indicated that the clayey sediments typically have a maximum thickness of up to the order of 80-100 m. One bore sited only some 150 m north of granite outcrop near Wave Rock intersected 80 m of clay-rich material resting on weathered granite, demonstrating that the palaeovalleys locally had relatively steep sides, and all were not necessarily very broad with shallow-sloping sides.

HYDROGEOLOGY AND HYDROCHEMISTRY The palaeodrainages are filled river channels. They still carry water, but the surface flows are intermittent or nonexistent. In contrast, groundwater movement will still be occurring in the sediments of the palaeodrainage channels. In these palaeodrainage systems, the playas are recharged with freshwater after heavy rain. However, most of the time the wet playas are windows in the groundwater surface, exposing the water-table in the lowest part of the channel. Rainfall infiltrating into the soils permeates down to the

Avon River Palaeodrainage system local water-table, and then migrates downslope in the weathered bedrock and any overlying sediments, finishing in the palaeovalleys. Here the water discharges into the floor of the playas, evaporating and increasing the concentration of salts. Where this occurs, the valley material is full of water. Studies of groundwater in palaeodrainage channels by the Geological Survey of Western Australia in the Eastern Goldfield show that the salinity of the groundwater increases markedly downstream of significant playas, showing that playa evaporation produces a more saline plume which descends and continues downstream. Concentrations of salts rise until saturation is reached, causing crystallisation, mostly of gypsum but with halite and alunite. North and east of Lake Grace the rainfall decreases and the evaporation increases, resulting in the saturation of the water with respect to various salts. Gypsum is the most common salt deposited in the playas of the Avon system. Halite occurs in a limited number of playas. Commander (1999) commented upon the conspicuous absence of widespread halite crystallisation, noting that thin crusts formed during periods of high evaporation dissolve during the subsequent flooding, though without speculating on the reason for this deficiency of halite. The sodium chloride is meteoric, derived initially from sea sprays. Modern precipitation in rainfall is equivalent to 3 t NaCl/y/km2 throughout most of the Avon Palaeodrainage. Assuming this rate has continued since river water flows decreased, the total quantity of NaCl deposited could be well over 50 x 10 6 t/km2 and there is no evidence for this quantity. Order-of-magnitude calculations suggest there should be halite deposits throughout the floor of the palaeodrainage system and up to some 10-100 m thick. It appears that either groundwater transport of the halite is an effective transporter of halite into the Avon River from the upper reaches of the palaeodrainage system, or that extremely intermittent very large flooding events occur transporting the halite back to the ocean. Even with only a small proportion of this salt remaining, there is potential for large quantities to be held dissolved in storage in the subsurface waters. This heritage of salt, with more being added continually from rainfall, is generating a major environmental effect that has serious implications for future land uses in parts of the Wheatbelt region. Of great economic interest in the playas is the accumulation of alunite, K 3 ( S 0 4 ) 2 ( 0 H ) 6 , a potential source of the element potassium. Playas containing alunite occur from Lake Chandler, through Lake Baladjie to some playas near Mt Palmer southeast of Southern Cross. The alunite forms when dissolved potassium and sulfate ions react with clays on the playa floor. At Lake Chandler the State Alunite Industry set up a complex industrial plant to extract the potassium from alunite as an emergency World War II source of the essential element. Australia still has no indigenous, viable potassium deposits and imports all it uses. Research is still ongoing to try to extract the potassium using cost- and energy-efficient means, and extraction of the potassium may occur again in the future from this source. Bentonite clays occurring in some of the playa floors have a range of useful properties, paramount of which is the ability to absorb large quantities of fluids and organic

chemicals. They are used for industrial and domestic purposes, such as the absorbent in 'Kitty Litter' and are becoming more important commodities for an increasing range of applications with time. Lake Chandler was a source of bentonites in the 1960s and 1970s and future mining may occur. Some of the bentonites are particularly important and have very specialised markets. Further exploration of the playas may locate additional occurrences.

LAND DEVELOPMENT IN THE WHEATBELT Immediately after the establishment of Perth in 1829, it was recognised that the extremely sandy soils on modern and older coastal dunes from the coast almost to the Darling Scarp were not ideal for agriculture. Fortunately for the early settlers, the alluvial and colluvial soils at the base of the scarp proved better. However, there are restricted areas of those soils, and from the scarp eastwards for up to about 80 km, the soils were found to be extremely rocky, dominated by what later became known as a laterite profile under an extremely thin soil. The York area was first explored by Ensign Dale and his expeditionary party in 1830, only months after Perth was established, and was immediately recognised by him as the start of excellent farming land that extended eastwards. Clearing of the native vegetation was commenced soon after to allow planting of grains and grazing of sheep for meat and wool. This clearing continued throughout the Wheatbelt region for the subsequent 140 years. In that time, in the Wheatbelt region east of the Darling Fault, of a total of 13 500 000 ha in 44 shires, only about 900 000 ha (7%) of private land was not cleared. Of this total, in the Avon Palaeodrainage, covering 6 100 000 ha of land in 19 shires, only 360 000 ha (6%) on private land now has a coverage of remnant native vegetation. In some parts the clearing was extremely severe, such as in the Shire of Narembeen where only some 2200 ha of private land out of a total of 380 000 ha (0.6%) still retains a covering of native vegetation. There was a very strong development urge even as late as the 1960s, and Beard (1990 p. 8) described farm development at that time as a strongly supported ideology, and even botanical surveys prior to clearing during that decade were unacceptable because 'it was feared that the discovery of rare species would lead to demands for reserves'. Initially the best soils were cleared, referred to as 'York Gum' land because these trees dominate on the most productive faring soils. However, the clearing continued until in some places all that remained were some small areas reserved for flora protection, and a number of these were areas in which elevated salinity occurred—the palaeodrainage channels. These areas were locally dominated by samphire or halophytic vegetation near to playas, but included patches of other types of vegetation. The samphire vegetation consists typically of low succulent shrubs with some 30-70% coverage and dominated by Halosarcia and Sclerolaena (Beard 1990). Lyons (1999) reported that the flora of the playas has not been systematically studied, but he quoted estimates of 1000 taxa in 300 genera as possibly being present. Adjacent to the playas are lunettes, or

M. J. Freeman

lake-side dunes, and these had a flora distinctly different from the nearby low-lying areas. A proportion of the soils of the Wheatbelt region have high clay contents, referred to generally as the 'heavy clay soils'. Through a complex series of processes, including repeated ploughing, handling, turning, cropping and changing its chemistry through fertiliser addition, the clayey fraction becomes affected by sodium enrichment, producing hard-setting soils prone to 'scalding'. Scalded patches only support reduced growth rates, slow the infiltration of rainwater and generally become non-productive. Addition of soluble calcium salts has been found to be an effective treatment, at least in the short-term, by replacing the sodium in the clay with calcium, improving the texture of the soils and returning the farmland to high crop-production rates. A remedy to overcome the deterioration of soil textures is to apply gypsum to the sodic soils at a rate of about 2.5-5 t/ha. This improves the texture, allows better infiltration of rainwater and seed germination and markedly increases the productivity of the affected soils. The gypsum, however, is soluble in rainwater and its effectiveness is therefore limited in time. Clearing of the native vegetation and growing of cereal crops has resulted in the replacement of relatively deeply rooted plants that have the property of losing water by transpiration and evaporation for the whole year by those that only lose water for a relatively short (<6 month/y) portion of the time. The groundwater at the time of settlement was in a state of equilibrium between the input from rainfall and the losses through the evergreen plant demand and evaporation from the exposed groundwater surface in the playas. Removing the native vegetation has shifted the equilibrium status, reducing the water loss from plants. The response to this has been the raising of the groundwater level, and this will continue until groundwater exposures in lower parts of the land, and hence the rate of evaporation, expand to achieve a balance. Groundwater movement downslope towards the palaeodrainage valleys is slow, measured in metres to tens of metres per year, although with permeable formations it can rise to hundreds of metres. The consequence is that the time to reach the new state of groundwater equilibrium will take many decades. New equilibrium will occur when an increased output of water occurs through increased evaporation. Left to itself with the native vegetation gone, larger areas of exposed groundwater will develop, leading to an increase in the evaporation. A balance will be achieved when the increased infiltration of rainwater is equalled by the accelerated evaporation. Regrettably, because of the quantity of NaCl present in the groundwater, all the exposures will be of saline water, thereby destroying all plants that are intolerant to elevated salinity as they are exposed to this water. Another problem surfacing is the threat to buildings. A number of townsites as well as farms have been built in the lower parts of the landscape. The rising groundwater levels are now encroaching on some of these, with reports of cellars becoming increasing damp and lower parts of walls suffering salt-damp. Hence, the environmental impact is not restricted to farmland and native vegetation, but also to the cultural aspects of the Wheatbelt region. This rising water level, coupled with increasing salinisation, is regarded by some as one of the largest environmen-

tal threats facing agricultural pursuits in Western Australia (see Taylor & McNally 2001). However, it is not a new phenomenon, having been noted as long ago as 1897 (Mulcahy 1978), but the wider community now recognises the potential seriousness of the situation and it is to be hoped that work to undertake appropriate remedial actions will be accelerated.

GYPSUM, ALUNITE AND EVAPORITES As the rainfall over the catchment and consequent run-off decreased, the integrated, flowing Avon River Palaeodrainage dried probably commencing in Late Eocene or Oligocene times and continuing to the Pliocene. Alternatively, as the climate dried, there may have been periodic reversions to wetter conditions. As the flows decreased, the water lost its erosive power resulting in fans being formed where tributaries flowed into the main channel blocking the river flow and forming ponds. The same occurred where aeolian sands encroached onto the channel. Ultimately, the palaeodrainage came to consist of a broad, gentle valley with a sequence of isolated playas with limited interconnection between them in places, but with the dominant water movement being of groundwater. Where gradients are higher, more active flows along the system can follow unusually wet times, as occur when tropical cyclone remnants consisting of intense low atmospheric-pressure systems that, having lost the highest velocity winds, have turned into intense rain-bearing depressions. These track to the southeast from the Pilbara region, leaving behind flooded land. At these infrequent times, playas can become interconnected systems with drainage towards the west and southwest. In addition to blockages from sedimentation, gentle epeiric doming of the Yilgarn could generate sills or partial blockages of the system. For example, just downstream of Lake Brown, the palaeochannel becomes narrow and rock outcrops almost cross the channel, whereas upstream is a much wider and apparently sediment-clogged channel. Rain, when converted to groundwater by infiltration, dissolves soluble components from the weathered rock and overlying alluvium as it moves downslope and down the water gradient. Where it discharges onto playa floors and the water evaporates, the salts precipitate when concentrations reach saturation. Gypsum is one of the more common salts to precipitate. Many of the playas in the Wheatbelt region have gypsum lunettes, in addition to sand and clay-pellet lunettes, always on the southeastern sides of the playas. This side is favoured because the maximum drying of the playa surface occurs as winds blow progressively from the northeast, swinging to the north as a front approaches, finally blowing strongest as northwesterlies immediately in front of the cold, humid change. Therefore, the maximum drying of the playa surface, allowing the loosening of grains previously held down with surface tension from the moisture, occurs immediately before the strongest winds from the northwest. Consequently the largest lunettes occur on the southeast side of the playas in the WTieatbelt region. At Lake Brown, gypsum occurs in two forms: in the lake bed sediment and on the surface of the sediment. Gypsum

Avon River Palaeodrainage system is deposited from groundwater as individual crystals up to 2 mm across within the near-surface sediment. This sediment now consists of nearly pure gypsum with a limited grainsize, demonstrating a very stable process of crystallisation. Only rarely do the crystals continue to grow to larger sizes, but the mechanism preventing this occurring ubiquitously is not clear. In addition, gypsum is deposited seasonally as a near-continuous film on the surface, up to mm thick. As the surface dries, the gypsum film becomes disrupted and broken. Strong northwest winds then blow the gypsum landward where it accumulates to form the lunettes on the shore of the playa. The Lake Brown gypsum has been mined from the lake floor since the mid-1950s and is used as a prime source of gypsum for the manufacture of Plaster of Paris for Perth's ceiling and wall plastering. Mining has progressed for several kilometres across the playa floor, and originally mined areas are now generating more gypsum that is blowing to the lake shore and has started forming a new, low lunette, demonstrating that the lunette growth can be extremely rapid. Lake Brown is in the Lake Campion Nature Reserve and at about 50 km 2 is the largest of several tens of playas in the reserve, many of which have sand, clay or gypsum lunettes. Eight of the lunettes containing near-pure gypsum as loose crystals up to ~ 1 mm across are very attractive for farm gypsum supplies and many have therefore been targeted for mining. Botanical research conducted for the Department of Conservation and Land Management indicated that some 25% of the species conserved in the reserve are endemic to the gypsum lunettes, and therefore mining of all the lunettes seriously threatens a significant component of the flora in the nature reserve. Mattiske (1995) has researched a number of gypsum-bearing lunettes throughout the Wheatbelt. Freeman (1994) quantified the gypsum and determined there is ~6 Mt in all the lunettes, beside an unquantified but large resource on the floor of Lake Brown. This is a large resource in demand by the local farming community. Consideration of the environmental effects of mining the lunettes is receiving close scrutiny from the Environmental Protection Authority and is opposed by a number of conservation bodies. At Lake Chinocup, south of the township of Lake Grace, a clay-floored playa has a large gypsum lunette on the southeastern shore. A mining proposal for this gypsum quoted a total resource of 1 Mt of seed gypsum. The Environmental Protection Authority assessed the mining proposal and recommended against mining, and one of the grounds was that at least two species of rare flora were threatened by the mining. The Minister for the Environment subsequently allowed a limited trial mining to ascertain if the gypsum could be extracted in an environmentally acceptable manner and the native vegetation restored.

CONSERVATION AND THE THREAT TO REGIONAL BIODIVERSITY There are two long-term threats to the remnant native flora in the Avon Palaeodrainage. First, through the removal of the native vegetation on the upland slopes, in the long term, the reduced loss of water through decreased evapo-

transpiration will ultimately lead to the raising of groundwater levels. This will stress native vegetation in uncleared areas in low-lying parts of the landscape leading to mortality and loss of sensitive species. Second, the increased flux of groundwater will move down-gradient into the palaeodrainage channels, raising water-tables until a new equilibrium is achieved between the increased flux and loss through evaporation or outflow from the system. Outflow from the inland parts will not occur because of the clogging of the palaeovalley systems, at least not until water levels rise many metres to tens of metres above the valley floor to allow the water to flow over the blockages. Because of the high evaporation rates, the new equilibrium will probably be established through moderate raising of the groundwater levels, when larger areas of wet soil and sediment, and possibly some perennially inundated playas, result in increased evaporation that balances the increased groundwater flux. The raised water level will then threaten the remaining vegetation in the lowest parts of the palaeodrainage system. These are the parts of the region less attractive for farming, where nature reserves dominate, and now this is where the greatest long-term threat is from rising water-table levels. The rising saline groundwater levels also threaten the agriculturally valuable plants, resulting in loss of income by farmers with low-lying farmland. Coincidentally, there is increased demand for gypsum as a farmland additive to improve the soil textures and fabric. The gypsum lunettes also occur in the same low-lying areas of the palaeodrainage systems that are at greatest threat from rising saline groundwater levels. Mining of the gypsum to provide the additive required to improve the soil condition could further deplete the limited areas of native flora that remain after the clearing for farming.

CONCLUSION—OUTLOOK FOR THE FUTURE There are some 20 identified occurrences of gypsum adjacent to playas in the Wheatbelt region with about 12 in nature reserves. Because of the nature of the gypsum, only plant species that tolerate this mineral will grow, leading to a significant degree of endemism in them. The very fact that the gypsum has developed because of the geological history and environment in the palaeodrainages, has also now led to the concentration of reserves to protect the small remaining proportion of native vegetation over the same palaeodrainages. Gypsum applications are highly desirable to maintain farm productivity in areas of clayey soils. The majority of the gypsum occurs in the nature reserves and mining of the gypsum would deplete the nature conservation values of these reserves. The paradox then arises that if the native flora is protected from gypsum mining, the hydrogeological situation potentially poses a very serious long-term threat to the same areas along the lowest parts of the palaeodrainage channels. This is a very difficult and thought-provoking situation. Many unknown factors prevail, such as: (i) what is the long-term prognosis for gypsum applications to farm land; (ii) is it a worthwhile additive or will there be as-yetunrecognised adverse reactions from its use; (iii) if gypsum is not made available, will farm output from susceptible soils decrease to such an extent that funding for other sys-

M. J. Freeman

tems that now protect the environment decrease or cease, leading to additional environmental degradation; (iv) if gypsum is mined from the nature reserves, at what stage will the limit be reached where the depletion of the native flora species has gone far enough and how will such a limit be determined; and (v) how much more should the hydrogeology of the palaeodrainage c h a n n e l s be studied to assess the threat to the nature reserves from rising groundwater levels of saline water b e c a u s e if they rise significantly above the land surface the nature reserves will be seriously threatened? These questions deserve answers. It is clear that the most important aspect is to understand the groundwater regime a n d its host, the lithological media holding the water. B e c a u s e groundwater is unseen, members of the community commonly have perception difficulties in appreciating its nature, a n d understanding the meaning of 'rising salinity', except where it c o m e s to the surface a n d has already severely affected the land surface. Groundwater data collection is slow a n d expensive a n d until comprehensively understood is prone to misunderstanding and misinterpretation. Geoscientific input to the investigations is critically important a n d this factor is often not appreciated. It is most important to ensure there is an adequate input of geoscience to all considerations of cause, effect and remedy. The long-term solution to the loss of arable land to saline groundwater has to relate to increasing the rate of removal of water from the hydrogeological system until it b a l a n c e s the input of water. Given a suitably long period, the b a l a n c e will be achieved naturally, but in the meantime much farmland a n d remnant areas of native vegetation will be saturated or inundated by saline groundwater. There have to be sound outcomes to address these impacts, and the State has embarked on a program intended to produce acceptable results in the State's Salinity Action Plan

ACKNOWLEDGMENTS Appreciation is e x p r e s s e d to the reviewers Phillip Commander a n d John Waterhouse, who markedly helped to improve this paper, a n d to M a n j e e t Kumar for compiling the figure. T h e paper is published with the approval of the Director General of the Department of Minerals a n d Energy, Western Australia.

REFERENCES J. & W I L S O N A . C . 1 9 8 9 . New records of Permian and Cretaceous sediments from the southwestern part of the Yilgarn Block. Geological Survey of Western Australia Report 25, 1-5. B E A R D J. S . 1 9 7 3 . The elucidation of palaeodrainage patterns in Western Australia through vegetation mapping. Vegmap Publications, Applecross WA, Occasional Paper 1. B E A R D J. S. 1990. Plant life of Western Australia. Kangaroo Press, Sydney. B E A R D J . S. 1 9 9 8 . Position and development history of the central watershed of the Western Shield, Western Australia. Journal of the Royal Society of Western Australia 8 1 , 1 5 7 - 1 6 4 . BETTENAY E . & M U L C A H Y M . J . 1 9 7 2 . Soil and landscape studies in Western Australia. (2) Valley form and surface features of the southwest drainage division. Journal of the Geological Society of Australia 18, 3 5 9 - 3 6 9 . BRAKEL A. T . 1990. Permian. In: B M R Palaeogeographic Group, BACKHOUSE

Australia: Evolution of a Continent, pp. 48-54. Australian Government Publishing Service, Canberra. P. A. & N E M C H I N A. A. 2000. Provenance record of a rift basin: U/Pb ages of detrital zircons from the Perth Basin, Western Australia. Sedimentary Geology 134, 209-234. CHURCHWARD H . M . 1 9 6 9 . Erosional modification of a laterized landscape over sedimentary rocks. Its effect on soil distribution. Australian Journal of Soil Research 8, 1-19. CLARKE J. D. A. 2001. Chapter 5—Ancient landforms of Kambalda and their significance to human activity. Geological Society of Australia Special Publication 21, 49-59. COMMANDER D. P. 1999. Hydrogeology of salt lakes in Western Australia. In: Proceedings of the Salt Lake Ecology Seminar, Perth, July 1999. Chamber of Minerals and Energy, Perth (unpaginated). C O M M A N D E R D . P . , K E R N A . M . & S M I T H R. A . 1 9 9 1 . Hydrogeology of the Tertiary palaeochannels in the Kalgoorlie Region. Geological Survey of Western Australia Record 1991/10. C O P E R. N. 1975. Tertiary epeirogeny in the southern part of Western Australia. Geological Survey of Western Australia Annual Report 1974, 46-50. FREEMAN M . J . 1 9 8 6 . Huckitta N T . Northern Territory Geological Survey 1:250 000 Geological Series Explanatory Notes. FREEMAN M. J . 1 9 9 4 . Gypsum resource in the vicinity of Lake Campion Nature Reserve. Department of Minerals and Energy Western Australia, Land Access Report EV 100 (unpubl.). HILL R . S . & M E R R I F I E L D H . E. 1 9 9 3 . An Early Tertiary macroflora from West Dale, southwestern Australia. Alcheringa 1 7 , 2 8 5 - 3 2 6 . H O C K I N G R . M. 1990. Bremer Basin. In: Geology and Mineral Resources of Western Australia, pp. 561-563. Geological Survey of Western Australia Memoir 3. H O C K I N G R. M. & COCKBAIN A. E. 1990. Regolith In: Geology and Mineral Resources of Western Australia, pp. 592-602. Geological Survey of Western Australia Memoir 3. CAWOOD

KOHN B .

L . , O ' S U L L I V A N P. B . , G L E A D O W A . J . W . ,

KARNER G .

J. K. 1998. Late Palaeozoic cooling of southwest Australian terranes inferred from apatite fission track thermochronology. Geological Society of Australia Abstracts 41, 253. L E BLANC S M I T H G. 1 9 9 3 . Geology and Permian coal resources of the Collie Basin, Western Australia. Geological Survey of Western Australia Report 38. LYONS M. N. 1999. Botanical values of Wheatbelt salt lakes. In: Proceedings of the Salt Lake Ecology Seminar, Perth, July 1999. Chamber of Minerals and Energy, Perth (unpaginated). MATTISKE E. 1995. A review of botanical values in a range of gypsum dunes in the wheatbelt of Western Australia: final report for Australian Nature Conservation Agency Save the Bush Program 1993-4 Project SS6007. Mattiske Consulting report for the Department of Conservation and Land Management, Perth (unpubl.). M U L C A H Y M . J. 1967. Landscapes, laterites and soils in southwestern Australia. In: Jennings J. N. & Mabbutt J. A. eds. Landform studies from Australia and New Guinea, pp. 211-230. Australian National University Press, Canberra. M U L C A H Y M . J. 1978. Salinisation on the southwest of Western Australia. Search 9, 269-272. M U L C A H Y M . J & BETTENAY E . 1 9 7 2 . Soil and landscape studies in Western Australia. (1) The major drainage divisions. Journal of the Geological Society of Australia 18, 349-357. WEISSEL

P I T T G . M , B A N B O W M . C . , B A R N E S L . C . , H A R R I S W . K . & LINDSAY J . M .

1978. Report on drilling in the Tallarings Palaeodrainage System. South Australian Department of Mines and Energy Report 78/39 (unpubl.). PLAYFORD P. E., COCKBAIN A. E. & Low G. H. 1978. Geology of the Perth Basin. Geological Survey of Western Australia Bulletin 124. SALAMA R. B. 1994. The evolution of salt lakes in the relict drainage of the Yilgarn River of Western Australia. In: Renault R. & Last W. eds. Sedimentology and Geochemistry of Modern and Ancient Saline Lakes, pp. 189-199. Society of Economic Paleontologists and Mineralogists Special Publication 50. SALAMA R. B . 1997. Geomorphology, geology and palaeohydrology of the broad alluvial valleys of the Salt River system, Western Australia. Australian Journal of Earth Sciences 44, 751-765. SALAMA R., B A R B E R C., H O S K I N G J. & B R I E G E L D. 1992. Geochemical evolution of Lake Deborah east, prototype salt lake in the relict drainage in the Yilgarn River of Western Australia. Australian Journal of Earth Sciences 39, 577-590.

Avon River Palaeodrainage system SALAMA R . B . , FARRINGTON p . , BARTLE G . A . & WATSON G . D . 1 9 9 3 . T h e

role of geological structures and relict channels in the development of dryland salinity in the wheatbelt of Western Australia. Australian Journal of Earth Sciences 40, 45-56. SALAMA R . B . & HAWKES G . E . 1 9 9 3 . Surficial geology and stratigraphy of the Walatin Creek area: a West Australian model of Cainozoic sedimentary deposition in the Yilgarn Craton. CSIRO Water Resources Research Division Report 93/4. SIRCOMBE K. N. & FREEMAN M. J. 1999. Provenance of detrital zircons on the Western Australian coastline—implications for the geologic history of the Perth basin and denudation of the Yilgarn craton. Geology 27, 879-882. TAYLOR G . & MCNALLY G . H . 2001. Chapter 2—Regolith: its history and environmental importance with particular reference to some engineering examples. Geological Society of Australia Special Publication 21, 13-25. TWIDALE C. R . , BOURNE J. A. & ROMANI J. R . V. 1999. Bornhardt inselbergs in the Salt Valley south of Kellerberrin, Western Australia with notes on a tesselated pavement in granite and pinnacles in

laterite. Journal of the Royal Society of Western Australia 82, 33-49. VAN DE GRAAFF W . J . E . , CROWE R . W . A . , BUNTING J . A . & JACKSON M .

J. 1977. Relict early Cainozoic drainage in arid Western Australia. Zeitschrift fur Geomorphologie 21, 379-400. C. C. 1968. South Australian submarine canyons: their distribution and ages. Marine Geology 6, 267-279.

VON DER BORCH

WATERHOUSE J . D . , COMMANDER D . P., PRANGLEY C . & BACKHOUSE J .

1994. Newly recognised Eocene sedimentary rocks in the Beaufort River palaeochannel. Geological Survey of Western Australia Annual Review 1993-1994, 82-86. W I L F O R D G. E. & TRUSWELL E. M. 1990. Cainozoic. In: BMR Palaeogeographic Group, Australia: Evolution of a Continent, pp. 82-88. Australian Government Publishing Service, Canberra. W I L S O N A. C. 1989. Palaeocurrent patterns in the Collie Coal Measures—the implications for sedimentation and basin models. Geological Survey of Western Australia Report 25, 85-91. Received 20 September 1999; accepted 26 August 2000

Geological Society of Australia Special Publication 21, 49-57

CHAPTER 5—Ancient landforms of Kambalda and their significance to human activity J. D. A. CLARKE Cooperative Research Centre for Landscape Evolution and Mineral Exploration, Australian National University: Canberrai ACT 0200, Australia. The Kambalda region of western Australia is a good example of an ancient landscape in a presently arid environment. Arid environments have been present here for only the last 5 million years or less. Landscape-evolution models used for utilisation and management of the landscape that do not recognise both the present arid environment and its short history are flawed. The low relief of the area originated in the Permo-Carboniferous glaciation, traces of which are preserved to the east. The climate during the Late Permian remained cool and humid, and extensive weathering may have occurred in and perhaps beneath the sedimentary cover. The partially stripped deep weathering that controls much of the topography at Kambalda may have been initiated during this period. Rifting along the southern margin of Australia during the Jurassic resulted in drainage reorganisation and incision. Pre-existing weathering profiles were truncated. The eroded material formed the thick Jurassic to Eocene successions in the basins along the southern margin. Both terrestrial and marine biota in these sediments point to temperate climates and cool-water deposition. Marine transgressions would have invaded the palaeovalleys many times since this interval, but only those of the Eocene left any good evidence. Fluvial sands and gravels, lignites and shallow marine cool-water carbonates and spiculites point to fluvial estuarine and shallow-marine deposition in a high rainfall, temperate environment. Deep weathering was ongoing through this period. Neogene lacustrine redbeds were deposited in the depocentres of the old palaeovalleys. Aprons of ferruginised fluvial gravels and sands surround these. The disorganisation of drainage and more oxidised nature of the sedimentation suggest decreased rainfall, greater seasonality, and perhaps a savannah-like climate. Weathering was characterised by fluctuating water-tables and development of mottling. Continued Increase in aridity led to sulfate deposition in the lakes by the Early Pliocene to Mid-Pleistocene. Development of extensive dune fields further disturbed drainage, segmenting many larger lakes into chains and clusters. Blankets of wind-blown dust mantled the landscape with extensive development of alluvial deposits in areas of low relief. Human utilisation of the area consists of pastoral grazing, mining and local urbanisation. Models of regolith evolution are widely used in planning and interpreting geochemical exploration. Particular environmental geological concerns include increased mobility of surface materials and contamination of soils and surface ground water through spillage or leakage of highly saline water. KEY WORDS: environmental geology, geomorphology, Kambalda, marine sediments, mining, regolith.

INTRODUCTION Landscapes are commonly regarded as products of Late Cenozoic or even purely Quaternary processes. This is particularly the case with northern hemisphere geomorphologists, familiar with areas dominated by Quaternary glaciation. For example, Bloom (1978 p. 8) in a widely used introductory geomorphology text states: 'Practically the entire subaerial landscape has evolved in the late Cenozoic Era'. Australia, however, escaped continental glaciation during the Quaternary, and here evidence of much older landscapes are abundant (Oilier 1991). Even in the northern hemisphere, overemphasis on glaciation can blind observers to numerous relicts of older landscapes (Hall 1987). When working in ancient landscapes it is important to realise that many features of the landscape are relict, having been formed under environment regimes quite different to those presently active. It is equally important to recognise that not all elements in the landscape are equally ancient: the landscape consists of many superimposed or

palimpsest features of different ages, often formed under very different regimes. Recognising that landscape evolution in Australia has often occurred over a time-scale of hundreds of millions of years can be critical to developing appropriate resource development and land-use management. Landscape evolution has resulted in the formation of a regolith (sometimes defined as everything between fresh rock and fresh air) which may be several hundred metres thick. The regolith can consist of weathered bedrock, with an intact or partially stripped weathering profile, partly or wholly buried by several different sedimentary units, each with its own weathering history. The soil profile may be related to the weathering profile, to the covering sediments, or be independent of either. Understanding the regolith can only be achieved through understanding the history of the landscape. The regolith is important in many ways, it is an aid, a hindrance, and an asset to resource exploration, regolith processes can both obscure and enhance the surface expression of mineralisation, and degrade and upgrade ore bodies. The regolith also hosts many important resources,

J. D. A. Clarke grasses dominate sandy soils while stands of native cypress are present on dunes. Halophytic phreatophytes colonise the margins of salt lakes and the lower reaches of streams entering them. The beds of some smaller salt lakes are completely covered by such salt-tolerant vegetation.

LANDFORMS OF KAMBALDA Physiography

Figure 5 . 1 Location of Kambalda in Western Australia (from Clarke 1993 figure 1).

sand, gravels, clay, salts, groundwater, lignite, peat and uranium being but a few examples. The soil component of the regolith provides the basis of agriculture and the biosphere generally; it provides the foundation to most constructions and engineering works, and much on-land waste disposal directly effects the regolith. A good example of an ancient landscape occurs at Kambalda in Western Australia (Figure 5.1). Kambalda has been a site of gold prospecting since 1896 and of large-scale nickel mining since 1967 (Gresham 1991). Modern gold mining began in 1981. The landscape of Kambalda is typical of much of the Eastern Goldfields region of Western Australia. Earlier work on the geomorphology of Kambalda can be found in Clarke (1998) and much of the following section is adapted from that paper.

KAMBALDA ENVIRONMENT Climate The climate is semiarid with a mean annual rainfall of 242 mm with most rain falling on February and June. Temperatures for Kalgoorlie, some 50 km to the north, reach a mean monthly maximum of 33.6°C in January and a mean monthly minimum of 16.5°C in June. Annual pan evaporation at Kalgoorlie is 2701 mm; maximum evaporation occurs in January and minimum evaporation in August. There is thus a strong precipitation deficit in the region, with evaporation exceeding precipitation by a factor of more than 10. Dominant wind direction in Kalgoorlie is from the east, however strong winds have a northerly vector in summer and a westerly vector in winter.

Vegetation Kambalda lies within the Coolgardie Botanical District of the Southwestern Interzone (Beard 1979; Newbey 1984). The vegetation consists of eucalypt woodlands; a bluebush and saltbush understorey develops as the carbonate content of soils increases. Saltbush-bluebush steppe replaces woodland in areas of the most strongly calcareous soil. Spinifex

Most of the Kambalda landscape lies at elevations between 300 and 350 m above sea-level. A few Isolated peaks rise almost to 400 m, which is the elevation of much of the surrounding granite terrain. Major salt lakes occupy the lowest parts of the landscape. The largest salt lake at Kambalda is Lake Lefroy, whose bed occurs at an elevation of 286 m. Lake Lefroy occupies the Lefroy palaeovalley. This major valley system is 15-40 km wide, 200 m deep and is one valley among many that drained into the Eucla Basin beneath the Nullarbor Plain (van de Graaff et al 1977). The Lefroy palaeovalley is approximately half-filled by sediment, most of which is of Tertiary age (Clarke 1993). The thickest sediments are Eocene silts, sands, lignites and marine spongolites. Younger Tertiary red-brown sands and silts overlie these. Modern lake sediments form a comparatively thin covering.

Underlying lithology The Kambalda bedrocks form part of the Kalgoorlie terrane of the Late Archaean Norseman-Wiluna Greenstone Belt in the Yilgarn Craton (Swager et al 1992). The Archaean rocks consist of ultramafic, mafic and felsic volcanics, together with volcaniclastic and epiclastic sediments, formed between 2.7 and 2.6 Ga. Slightly younger intermediate and felsic rocks have intruded these. The entire succession has undergone greenschist- to amphibolite-grade metamorphism. Extensive areas of granite surround the greenstone belt. The surrounding granite areas tend to be slightly higher than greenstone areas, and have subdued relief. A degree of lithological control is evident in relief. Ridges tend to be of mafic and ultramafic rocks whereas low-lying plains are of sedimentary or felsic volcanic rocks. Small granitic intrusions within the greenstone belt may locally form inselbergs.

Weathering and duricrusts Much of the terrain is deeply weathered, forming a saprolite, often more than 50 m deep (Figure 5.2). Lithology is important in controlling original weathering depth. Mafic rocks and granites are the most shallowly weathered and sedimentary and volcanic rocks are the most deeply weathered. Ultramafic rocks have intermediate depths of weathering. The major palaeovalleys truncate weathering, indicating that the saprolite formed prior to drainage incision. Weathering profiles are also superimposed on many of the palaeovalley sediments showing that deep weathering continued during their infill. Duricrusts consist largely of ferricretes. Most of these consist of iron-cemented surficial sediments rather than in

Kambalda landforms and human activity

AEOLIAN DEPOSfTS BURIED RALAEODRAINAGE

SALT LAKE I I

ISAPROUTE (OFTEN WITH IC0LLUV1AL VENEER)

Figure 5.2 Regolith terrain map of the Kambalda region (from Clarke 1994b figure 2). situ weathered material. Ferricretes commonly occur at two elevations in the landscape, about 350 and 300 m. Local topographic inversion has occurred. Minor duricrusts include small valley calcretes, dolocretes and magcretes developed over ultramafic rocks. Silcretes are locally present on siliceous Tertiary sediments, and gypcretes are present on gypsum dunes on the lakes. Several stacked gypcrete horizons are present in many such dunes.

Erosional landforms The largest scale erosional landform is the Lefroy Palaeovalley and its major tributaries. These features are now relict. Smaller scale fluvial erosion is ongoing along gullies cut into higher areas. Much of Lake Lefroy is underlain by a surface cut into bedrock at an elevation of about 280 m above sea-level.

J. D. A. Clarke

Figure 5.3 Schematic cross-sections showing evolution of the Kambalda regolith. Vertical scale = hundreds of metres and horizontal scale = tens of kilometres, (a) Pre-Jurassic topography and saprolitic mantle (G, granite; M, mafic; U, ultramafic; S, sediment). (b) Jurassic-Eocene valley incision and deepening of weathering front, (c) Eocene valley infill and deepening of weathering front, (d) Miocene fluviolacustrine sedimentation, (e) Pliocene-Holocene incision and deposition of playa lake and dune systems. From Clarke 1994b figure 3.

Thus lake sediments are commonly only 5 - 1 0 m thick. A meandering palaeovalley is incised 80 m or more into this buried rock surface. The northern and western shores of Lake Lefroy are rocky, with outcrops of bedrock, varying in size from islands to small knobs, emerging above the lake bed offshore. The exposed rocks are eroded by a combination of salt weathering and aeolian deflation. Deflation by strong winds (northerly in summer, westerly in winter) removes the weathered material, transporting it across the lakes to be deposited on the opposite shore. Some material may also be removed and transported by water when the lakes are filled. Wind-driven currents would also tend to deposit sediments on the eastern and southern shores of the lakes. These shores are characterised by aeolian sediment deposition, usually in parallel dune complexes. This asymmetry is found in even the smallest lakes and pans. The lake beds themselves correspond to the regional watertable and are often shallowly flooded in winter. Some aeolian deflation also occurs on land. This is especially prevalent where vegetation cover has been

Figure 5.4 Maps showing evolution of the Kambalda landscape, (a) Inferred pre-Jurassic valley and topography, (b) Triassic-Eocene drainage diversion and incision, (c) Eocene marine and non-marine deposition, (d) Miocene lake system and younger ferricrete systems, (e) Pliocene-Holocene arid landscape. From Clarke 1994b figure 4. effected by saline groundwater discharge, rabbit warrens, fire or human disturbance. Such deflation results in etching of the soil surface and exposure of the roots of shrubs.

Depositional landforms Extensive alluvial deposition has occurred in low-relief alluvial fans where streams discharge from elevated areas onto the plains surrounding Lake Lefroy. Streams entering along the northern and western shores of Lake Lefroy do so via fan deltas. Those along the eastern and southern shores are either lost within the marginal dunefields or have cut deep estuaries through the dunes to the lake. Wind-driven currents to the eastern and southern shores probably transport some of the material deposited into Lake Lefroy by streams along the northern and western shorelines. Dunefields have accreted along the eastern and southern shores of Lake Lefroy. Dunefields have also built up and disrupted fluvial and lacustrine connections between Lake Lefroy and adjacent lakes, Lake Randall to the east and Lake Zot to the south. The dunes consist primarily of reddened quartz sand, although heavy minerals, ferruginous grains, clay, and gypsum are significant smaller components. The youngest dunes parallel the modern lake shore and may contain small elongate playas in their

Kambalda landforms and human activity

results in salt and gypsum efflorescence and fluffy ground in both creeks and along lake shores. Soils in more elevated areas are skeletal and relict through active erosion. Buried mottled gley horizons point to periods of water saturation in the past.

EVOLUTION OF THE KAMBALDA LANDSCAPE Three diagrams illustrate the history of the Kambalda landscape. The first of these shows the development of the regolith profile in schematic cross-section (Figure 5.3). The second (Figure 5.4), is a plan view of the developing landforms. The third shows the development of the landscape from a more regional perspective (Figure 5.5).

Permian glaciation: Year Zero?

Figure 5.5 Palaeogeographic evolution of the southeastern Yilgarn Craton. (a) Palaeovalley incision (pre-Jurassic). (b) Middle Eocene Tortachilla transgression, (c) Late Eocene Tuketja transgression. (d) Semi-permanent lake sedimentation (Miocene), (e) Pliocene-Holocene salt lakes. From Clarke 1994a figure 7. swales. The oldest dunes are often degraded to sandplains and have pisolitic dolocretes developed on them. Small lakes have lunettes on their southern and eastern shores. Lacustrine sediments in Lake Lefroy consist of slightly calcareous, gypsiferous, silty to sandy muds up to 9 m thick (Clarke 1994c). The muds tend to be red-brown in colour along the margins of Lake Lefroy and carbonaceous in the centre. The gypsum consists of clear prismatic and swallowtail crystals. In the relatively recent past the gypsum of the lake bed has been deflated to form gypsum dunes. The dunes occur both on the bed of Lake Lefroy and along its western and southern shores. The gypsum dunes are now in the process of erosion by lake waters and are bounded by cliffs up to 10 m high. The gypsum dunes are older than the youngest quartz-clay dunes, which partly bury them.

Soils Soils are both carbonate-rich and strongly reddened. Carbonate occurs both as disseminated powder and discrete nodules. The nodules are pisolitic and typically dolomitic or mixed calcite-dolomite. When developed over or adjacent to ultramafic rocks they are composed of magnesite. Ongoing and relict alluvial, colluvial and aeolian deposition along the margins of Lake Lefroy result in complex soil stratigraphy in these areas. Vadose discharge

Continental ice sheets eroded the Yilgarn Craton during the Late Carboniferous to Early Permian (BMR Palaeogeographic Group 1990). They eroded an average of 350 m of rock from the craton (van de Graaff 1981). The present low relief of the craton may have originated through this erosion. Older valleys may have survived locally (Finkl & Fairbridge (1979). Fluvioglacial deposition occurred along the ice cap margins, especially during periods of ice decay. By the end of glaciation these processes had formed a thick cover of glacigene sediment over much of the Yilgarn Craton. This is now preserved as isolated relicts of the Lower Permian Paterson Formation to the east of Kambalda (Griffin & Hickman 1988). Recycled Late Permian palynomorphs, both in Tertiary (C. Foster pers. comm. 1993) and Cretaceous sediments (Alley & Clarke 1992) of southern and western Australia suggest that Permian cover was much more extensive than it is now.

Permian to Jurassic: the hidden years Little can be inferred about the nature of the landscape from the Late Permian to Early Jurassic because of the absence from the Yilgarn Craton and adjacent basins of sediments of this age. We infer that the landscape was of low relief, consisting of plains of glacigene sediment. The wide headwaters of the main large palaeovalleys suggest that they were cut by rivers that had their source in what is now Antarctica, then joined to southern Australia (Oilier 1988). The climate was probably cool (not tropical) and weathering occurred in and beneath the sedimentary mantle. The deep weathering profile, now partially stripped to form the present topography of the Kambalda landscape, may date from this long interval.

Jurassic through Eocene: emergence of the landscape The inception of rifting along the southern margin of Australia during the Jurassic resulted in the extensive reorganisation of drainage. Palaeovalley headwaters were truncated and their flows diverted or even reversed. Rivers began to incise the landscape, stripping first the Permian sedimentary cover and then the deeply weathered land-

J. D. A. Clarke

scape without apparent regard for underlying lithological control. The valleys transported the eroded material into the developing rift separating Australia and Antarctica where they formed thick successions of Jurassic to Eocene clastic sediment (Clarke & Alley 1993). Sediments deposited in the valleys during sea-level highstands were eroded out during subsequent regressions. No Jurassic sediments are known in the palaeovalleys and Cretaceous sediments only from their lower reaches (Jones 1990). Eocene sediment is preserved in the palaeovalleys only because subsequent aridity minimised erosion.

Rise and fall of Eocene seas Similar sedimentary infill occurs in the lower reaches of all the palaeovalleys along the margin of the Eucla Basin. The Lefroy palaeovalley can be taken as typical. Two marine transgressions occurred during the Middle to Late Eocene, the first (Tortachilla transgression) at 40-38 Ma and the second (Tuketja transgression) at 38-37 Ma (McGowran et al 1992). The transgressions resulted in an irregular coastline of estuaries and embayments along the margins of the Eucla and Bremer Basins. Deposition on the coastal plain occurred in a mosaic of lagoons, peat bogs, and river channels. Terrestrial vegetation consisted of temperate rainforests with angiosperms such as Nothofagus and conifers, including podocarps and araucarians. Ferns were present as an understorey (Clarke 1994a, Carpenter & Pole 1995). The rainforests resembled those of both highland New Guinea and Tasmania. Most offshore sediments deposited during the first transgression were eroded from the Lefroy palaeovalley during the period of low sea-level prior to the second transgression. Deposition during this period in the lower parts of the palaeovalley consisted of estuarine sandbars. The second transgression deposited a very unusual sediment, composed almost exclusively of siliceous sponge spicules. This sedimentary event occurs in the shallow-water sediments of the Late Eocene of southern Australia from Walpole in the southwest of Western Australia to Adelaide. It is not known why siliceous sponges should proliferate to such an extent during such a narrow time interval. The nearshore sediments during both transgressions passed offshore into the open-water limestones of the Eucla Basin (Jones 1990).

Oligocene to Pliocene: the lake era The post-Eocene transgressions in the Eucla Basin did not come up the Lefroy palaeovalley as far as Kambalda, However, the high base-level that formed allowed finegrained deposits to accumulate in lakes. The sediments were mainly red silts and sands, but included some carbonates and organic-rich facies. The lakes were surrounded by low-relief alluvial fans and fan deltas, which deposited coarse-grained, iron oxide-cemented clastics. The volume of eroded and deposited material was minuscule compared to the vast quantities stripped from the region from Jurassic to Eocene times and erosional modification of the landscape was minor. Deep-weathering processes continued, resulting in mottled weathering profiles superimposed on Eocene sediments along the flanks of the palaeovalley. Reduced rates of erosion and disorganisation of the

palaeodrainage into a series of lakes implies a change to a drier, possibly savannah-like climate.

Pliocene and Pleistocene: onset of aridity The upper part of the palaeovalley sediments shows an abrupt beginning to evaporitic sedimentation. This reflects establishment of climates arid enough to allow deposition of sulfates. The reduced nature of some of these sediments has preserved pollen grains, dating the onset of gypsum deposition to the earliest Pliocene. Palaeomagnetic dating by Zheng et al (1998) suggests the aridity is more recent, being established in the mid-Pleistocene. Whichever date is ultimately proved to be correct, the sedimentary record of Lake Lefroy indicates that arid conditions have persistent to the present, through a complex and possibly cyclic history (Clarke 1994c). More arid conditions resulted in establishment of extensive dunefields. These disrupted the course of the palaeovalleys even further, with the dunefields segmenting many larger lakes into chains or clusters of smaller ones. Windblown silt blanketed areas of low-relief topography, including many areas underlain by residual bedrock. Although the current landscape is dominated by aridity, it is important to realise that these processes have only operated for the last 5 million years. Long though this period may seem in human terms, it represents only 2% of the total history of the landscape. The present is not (always) the key to the past.

Human impacts on the Kambalda landscape It is likely that Aboriginal burning of the region may have had a major impact on vegetation patterns, as suggested for other areas of Australia. The duration and extent of Aboriginal modification of the Kambalda region will remain unknown until the Pleistocene record of the area is better understood. The Kambalda region has experienced multiple land use since the 1890s. In addition to gold and nickel mining already mentioned, there has been in the past an extensive timber-cutting industry in support of mining. Sandalwood harvesting has also been important. Pastoral activity has been confined to sheep, which have been responsible for localised overgrazing of the near-ground vegetation and subsequent erosion. Anthropogenic effects on the Kambalda landscape are of two types, direct and indirect. Direct effects include the result of human constructions and excavations, most obviously mines, tailing dams and mullock dumps, but also water-catchment dams, causeways, buildings and roads. In addition to changing the shape of the landscape, many of these features also effect drainage and water movement. This in turn effects fluvial sediment transport and deposition. The discharge of hypersaline mine water into conveniently situated playa lakes has also had the local effect of changing, and in some cases building up, the playa lake surfaces. Playas formerly characterised by clay or gypsum deposition acquire thick crusts of halite through the evaporation of the hypersaline water. Indirect effects mainly result from disturbance of vegetation cover that then changes erosion and deposition of

Kambalda landforms and human activity sediments, particularly aeoliari sediments. Vegetation disturbance has resulted through fire, overgrazing, construction and excavation work, and the activities of rabbits. Effects to date are generally localised to the immediate vicinity of the disturbance.

APPLICATIONS AND IMPLICATIONS Exploration Much of what is sampled in the course of mineral exploration in the Kambalda region is regolith. Even what passes for fresh rock is generally partly weathered. Thus understanding landscape shaping and regolith-forming processes is of vital importance to mineral exploration. Regolith processes have both negative and positive effects on the surface expression of mineralisation. Negative effects include the attenuation of geophysical signatures of mineralisation through destruction of minerals responsible for an electric or magnetic anomaly, and attenuation of geochemical signatures by leaching of elements of interest to an exploration geochemist. In addition, the surface expression of mineralisation can be simply buried, sometimes by up to 100 m of palaeovalley deposits. Positive effects of regolith processes are that the size of the geochemical halo surrounding mineralisation may be increased, even though attenuated. The size can be increased through mechanical dispersion (erosion and dispersal of fragments of mineralisation) or chemical dispersion (migration of indicator elements such as gold, nickel, arsenic and copper in groundwater and then precipitation over a considerable area surrounding the deposit). The increased size of the halo greatly increases the chances of the mineralisation being discovered. Important studies of exploring in areas of thick regolith in Western Australia, including Kambalda, can be found in Mazzucchelli (1972), Anand et al (1998) and Butt et al (1992).

Mining The great age and complexity of the regolith at Kambalda and elsewhere on the Yilgarn Craton has important implications to mining, particularly in openpits. Three are probably most significant. The first consists of engineering aspects. Weathered rock and poorly cemented sediments have quite low strengths, necessitating much lower angles for the pit walls. This increases the stripping ratio—the amount of waste rock in comparison to ore that needs to be removed. The thickness and nature of the regolith can thus impact markedly on mining costs and thus both the feasibility on the project and the economic return on investment. The second impact is the effect of deep weathering on the ore itself. Supergene enrichment of gold and nickel ores can be important in some deposits, while in others the ore elements have been largely leached out of the regolith. Both processes can operate in the same deposit, resulting in a highly complex distribution of ore blocks, the understanding and prediction of which can be facilitated by understanding the regolith processes involved. Finally, the regolith itself may host significant deposits

of economic interest. Across the Yilgarn Craton as a whole, these include lignite (Elms et al 1982), uranium (Mann & Deutscher 1978; Fullwood & Barwick 1990), and gold (Smyth & Button 1989). All these deposits occur within the palaeovalley cover. Technical, political and environmental issues have precluded their exploitation at present. Mineralisation within the weathered sequence itself has proved easier to exploit. Most notable of these is the Boddington gold deposit (Hickman & Keats 1990) and the Darling Range bauxite deposits (Smurthwaite 1990). The Boddington gold deposit is the result of enrichment in the saprolite of otherwise subeconomic gold deposits. Bauxite is a duricrust from which almost all constituents except aluminium oxides are leached.

Groundwater The shortage of surface water resources over much of the Yilgarn and the limited capacity of the water pipeline from the Darling Ranges has forced extensive exploration for groundwater throughout the Yilgarn Craton (Commander 1993; Kern & Commander 1993; Morgan 1993). Even when the salinity is too high for human or animal use, it still has considerable value as process water in metallurgical plants. Disposal of such highly saline water must be carefully managed because of its destructive potential of vegetation and soils. Dundon (1988) gives a useful historical review of groundwater exploration and utilisation in the Goldfields region. In particular he noted the importance of groundwater contained in palaeovalleys (89% of the total) and noted the divergence between present drainage and subsurface groundwater flow, which may actually cross surface watersheds. Such groundwater capture may be due to subtle epeirogenic movements associated with the Jarrahwood Axis, which has been responsible for diversion of surface drainage through the Mesozoic and the Cenozoic (Clarke 1994a). Dundon (1988) also discussed, with examples, other aquifer types, including fractured bedrock aquifers and aquifers developed in weathered bedrock. Commander et al (1991) highlighted the major environmental groundwater issues in the Eastern Goldfields region. These included contamination of near-surface groundwater by saline water, contamination of soils from spilled or leaked saline water, and soil or groundwater contamination by cyanide leakage from gold-tailings dams. Prevention of such spills to date has been through usage of lined tailings and storage dams and recycling saline water as much as possible to minimise usage. A good understanding of the small confined and unconfined near-surface aquifers in the regolith of the region is likely to be of great assistance in managing and containing spills when they occur. These are poorly understood at this time, however.

Environmental geology Geological inputs into environmental studies at Kambalda were initially minimal. Some of the earliest studies of the environment focused on geographic factors (Brodie-Hall & Oliver 1976), climate and vegetation in urban planning (Verschuer 1976) and sociology (Reading 1976; Austin 1976). The impacts of pastoral activity on soils were not documented in these studies. Cahill (1991) discussed

J. D. A. Clarke

revegetation from a number of different mining and processing operations in the Goldfields, although soil-nutrient chemistry was the only regolith factor considered. Woolard (1991) reviewed past and present environmental management of exploration programs at Kambalda, paying particular attention to the impact of drillsite preparation and access, saline water from drilling, and drillsite and track rehabilitation on the soil. Particular issues included the ability of soils to support plant growth, contamination from fuel and hydraulic fluids, and soil erosion leading to sedimentation in dams. An understanding of groundwater and soil chemistry and modern surface processes of erosion, sedimentation and degradation of anthropogenic materials was demonstrated as being important in the planning, implementation, and rehabilitation phases of exploration. The potential significance in different geological substrates (saprolite, fresh rock, Tertiary sediments, dune sands, alluvial and lacustrine sediments, etc.) to management issues has not been publicly documented. Environmental geology at Kambalda is thus in its infancy.

CONCLUSION The Kambalda landscape, like most others in Australia, is a palimpsest landscape, one where the effects of many different environments have been superimposed. Understanding the Kambalda region requires understanding the history of at least 2 5 0 million years. The Permian glaciation, the cooltemperate forests of the Cretaceous to Early Tertiary, the Eocene transgressions, and the Late Tertiary savannas, were all dramatically different from the current semiarid environment. Semiarid environments have been important for about 2% of the total history of the Kambalda landscape. At Kambalda, and indeed for much of arid Australia, the present is not the key to the past, but a plumb line by which we can measure how different the past was to the present. Environmental change on a massive scale has shaped the landscape we know today. Effective utilisation and management of landscapes such as Kambalda require understanding of their history, although use of regolith data in such investigations has been limited, apart from geochemical exploration.

ACKNOWLEDGMENTS I wish to thank my former colleagues in W M C for discussions that helped shape the ideas in this paper and for the writers and commentators of previous writings on this subject. Others to whom I am indebted include Brian McGowran and Neville Alley who provided reality checks on many ideas. Finally, I am grateful to Vic Gostin for inviting me to contribute to this volume.

REFERENCES ALLEY N. F. & CLARKE J. D. A. 1992. Stratigraphy and palynology of

Mesozoic sediments from the Great Australian Bight area, southern Australia. BMR Journal of Australian Geology & Geophysics 13, 113-130.

ANAND R . R . , SMITH R . E . , PHANG C . , WILDMAN J . E . , ROBERTSON I . D .

M. & MUNDAY T. J. 1998. Geochemical exploration in complex lateritic environments of the Yilgarn Craton, Western Australia. CRC LEME open-file report for P240A. AUSTIN J. K. 1976. Kambalda case studies—4. Education and teenagers. In: Australian UNESCO seminar Kambalda (Western Australia), 18-23, August 1973, pp. 215-226. Australian Government Printing Service, Canberra. BEARD J. S. 1979. Phytogeographic regions. In: Gentilli J. ed. Western Landscapes, pp. 107-121. University of Western Australia Press, Perth. BLOOM A. L. 1978. Geomorphology. Prentice-Hall Inc, Englewood Cliffs. BMR PALAEOGEOGRAPHIC GROUP. 1990. Australia: Evolution of a Continent. Australian Government Publishing Service, Canberra. BRODIE-HALL L. C. & OLIVER J . B . 1976. Kambalda case studies—1. Man and the environment: new towns in isolated settings. In: Australian UNESCO seminar Kambalda (Western Australia), 18-23, August 1973, pp. 187-197. Australian Government Printing Service, Canberra. BUTT C . R . M . , SMITH R . E . , ANAND R . R . , LAWRENCE L. M . & DOWNES

P. N. 1992. Semiarid and arid terrains. In: Butt C. R. M. & Zeegers H. eds. Regolith exploration geochemistry in tropical and subtropical terrains, pp. 295-391. Elsevier, Amsterdam. CAHILL M. 1991. Rehabilitation methods and trials in the goldfields. Revegetation—our future. In: Papers from a workshop on practical environmental management in the goldfields, Western Australia, Section 9 (no pagination). Western Australian Chamber of Mines in conjunction with the Goldfields Rehabilitation Group. CARPENTER R . J . & POLE M. S . 1 9 9 5 . Mid-Late Eocene plant fossils of the Lefroy and Cowan palaeodrainages, Western Australia. Australian Systematic Botany 8 , 1 1 0 7 - 1 1 5 4 . CLARKE J. D. A. 1993. Stratigraphy of the Lefroy and Cowan palaeodrainages, Western Australia. Journal of the Royal Society of Western Australia 76, 15-22. CLARKE J . D. A. 1994a. Evolution of the Lefroy and Cowan palaeodrainages, Western Australia. Australian Journal of Earth Sciences 41, 55-68. CLARKE J. D. A. 1994b. Geomorphology of the Kambalda region of Western Australia. Australian Journal of Earth Sciences 41, 229-239. CLARKE J . D. A. 1994c. Lake Lefroy, a palaeodrainage playa in Western Australia. Australian Journal of Earth Sciences 41, 417-427. CLARKE J. D. A. 1998. Ancient landforms of Kambalda and Norseman. Geological Society of Australia Special Publication 20, 40-49. CLARKE J. D. A. & ALLEY N. F. 1993. Petrologic data on the evolution of the Great Australian Bight. In: Findlay R. H., Unrug R., Banks M. R. & Veevers J. J. eds. Gondwana Eight, Assembly, Evolution and Dispersal, pp. 585-596. A. A. Balkema, Rotterdam. COMMANDER D. P. 1993. Groundwater in the goldfields. In: Spottiswood D. J. & Misra V. N. eds. Kalgoorlie as a Regional Industrial Centre, pp. 11-13. Proceedings of the National Conference Kalgoorlie Branch Australasian Institute of Mining and metallurgy, Kalgoorlie Group Institution of Engineers Australia, WA School of Mines, and the Goldfields-Esperance Development Authority. Published by WA School of Mines and Curtin University. COMMANDER D. P., KERN A. M. & SMITH R. A. 1991. Hydrogeology of the Tertiary palaeochannels of the Kalgoorlie region. Geological Survey of Western Australia Record 1991/10. DUNDON P. J . 1 9 8 8 . Exploration for groundwater in the goldfields. In: R & D for the Minerals Industry, pp. 7 7 - 8 5 . Western Australian School of Mines Conference 88. ELMS R. G., MATTHEWS R. D. & CHAPMAN D. G. 1982. Notes on the geology of the lignite resources—Esperance, Western Australia. Australian Coal Geology 4, 101-110. FINKL C. W. & FAIRBRIDGE R. W. 1979. Palaeogeographic evolution of a rifted cratonic margin: S.W. Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 26, 221-252. FULLWOOD K . E . & BARWICK R. E . 1 9 9 0 . Mulga Rock Tertiary uranium deposit. In: Abstracts 2nd International Conference on Prospecting in Arid Terrain, Perth, Western Australia, April 1988, pp. 67-67. Australasian Institute of Mining and Metallurgy. GRESHAM J. J. 1991. Kambalda. History of a mining town. Western Mining Corporation Ltd, Melbourne. GRIFFIN T. J . & HICKMAN A. H . 1 9 8 8 . Widgiemooltha 1: 250 000 geological sheet SH51-14, Western Australia. Geological Survey of Western Australia, Perth.

Kambalda landforms and human activity A. V. 1987. Weathering and relief development at Buchan, Scotland. In: Gairdner V. ed. International Geomorphology 1986, pp. 991-1005. John Wiley & Sons Ltd, Chichester. HICKMAN A. H . & KEATS W. 1990. Gold. In: Geology and Mineral Resources of Western Australia, pp. 645-669. Geological Survey of Western Australia Memoir 3. JONES B. G. 1990. Cretaceous and Tertiary sedimentation on the western margin of the Eucla Basin. Australian Journal of Earth Sciences 37, 317-329. KERN A. M. & COMMANDER P. 1993. Cainozoic stratigraphy in the Roe Palaeodrainage of the Kalgoorlie region, Western Australia. Geological Survey of Western Australia Report 34, 85-95. MANN A. W. & DEUTSCHER R . L . 1 9 7 8 . Genesis principles for the precipitation of carnotite in calcrete palaeodrainages in Western Australia. Economic Geology 73, 1 7 2 4 - 1 7 3 7 . MAZZUCCHELLI R . H . 1 9 7 2 . Secondary geochemical dispersion patterns associated with the nickel sulphide deposits at Kambalda, Western Australia. Journal of Geochemical Exploration 1, 103-116. M C G O W R A N B . , Moss G . & BEECROFT A. 1992. Late Eocene and Early Oligocene in southern Australia: local neritic signals of global oceanic changes. In: Prothero D. R. & Berggren W. A. eds. EoceneOligocene Climatic and Biotic Evolution, pp. 178-201. Princeton University Press, Princeton. MORGAN K . H . 1 9 9 3 . Development, sedimentation, and economic potential of palaeoriver systems of the Yilgarn Craton of Western Australia. Sedimentary Geology 85, 6 3 7 - 6 5 6 . N E W B E Y K . R . (Editor) 1 9 8 4 . The Biological Survey of the eastern Goldfields of Western Australia. Records of the Western Australian Museum, Supplement 18. OLLIER C . D . 1 9 8 8 . The Regolith in Australia. Earth-Science Reviews

HALL

25,

355-361.

OLLIER C. D.

1991. Ancient landforms. Belhaven Press, London and New York. READING P. J. R . 1976. Kambalda case studies—3. Living in Kambalda. In: Australian UNESCO seminar Kambalda (Western

Australia), 18-23, August 1973, pp. 201-214. Australian Government Printing Service, Canberra. and Mineral Resources of Western Australia, pp. 615-624. Geological Survey of Western Australia Memoir 3. SMYTH P. R. & BUTTON A. 1989. Gold exploration in Tertiary palaeodrainage systems of Western Australia. In: Bhappu R. B. & Harton R. J. eds. World Gold '89. Gold forum on technology and practices, pp. 262-269. Proceedings of the 1st Joint International meeting between the Society of Mining Engineers and the Australasian Institute of Mining and Metallurgy, Reno, Nevada. SWAGER C., W I T T W . K., GRIFFIN T. J. ET AL. 1992. Geology of the Archaean Kalgoorlie terrane—an explanatory note. Geological Survey of Western Australia Record 1990/12. VAN DE GRAAFF W. J. E. 1981. Palaeogeographic evolution of a rifted cratonic margin: S.W. Australia—a discussion. Palaeogeography, Palaeoclimatology, Palaeoecology 34, 163-172. SMURTHWAITE A. J. 1990. Alumina. In: Geology

VAN DE GRAAFF W . J . E . , CROWE R . W . A . , BUNTING J . A . & JACKSON M .

J. 1977. Relict Early Cenozoic drainage in arid Western Australia. Zeitschrift fur Geomorphologie NF 21, 379-400. The preservation of the environment in a developing area. In: Australian UNESCO seminar Kambalda (Western Australia), 18-23, August 1973, pp. 1 9 8 - 2 0 0 . Australian Government Printing Service, Canberra. WOOLARD C. 1991. Environmental management of exploration impacts, KNO case histories. In: Jones I. O. & Misra V. eds. Geology, Mining, and Metallurgical Practices in the Eastern Goldfields, pp. 127-141. Proceedings of the Regional Conference, Kalgoorlie Branch, Australasian Institute of Mining and Metallurgy, pp. 127-141. ZHENG H., WYRWOLL K. H., LI Z . & POWELL C. MCA. 1998. Onset of aridity in southern Western Australia—a preliminary palaeomagnetic appraisal. Global and Planetary Change 18, 175-187. VERSCHUER J . 1 9 7 6 .

Received 20 July 1997; accepted 5 May 2000

THEME 2 GEOHAZARDS IN URBAN COMMUNITIES

Geological Society of Australia Special Publication 21, 61 -66

CHAPTER 6—Geohazards risk to urban communities in Australia K. GRANGER*

Australian Geological Survey Organisation, GPO Box 378, Canberra ACT 2601 Australia. The sciences that address natural hazards have traditionally focused on the study of specific phenomena. Doing something about those hazards and reducing the risks to the community that they pose has tended to be left to the engineering profession or, in the extreme, to disaster managers. The past decade, however, has seen a paradigm shift with increasing attention being paid to vrisk management' across many disciplines. Developing a comprehensive understanding of the risks faced by an urban community is the first challenge to be met if appropriate risk reduction strategies and response options are to be implemented and their effectiveness monitored. The Queensland city of Cairns has provided an excellent case study to demonstrate the scenario-based risk-assessment methods and GIS decision support tools developed under the Australian Geological Survey Organisation's Cities Project. KEY WORDS: Cairns, earthquakes, floods, geohazards, landslides, public risk, risk analysis (hazards), risk management.

INTRODUCTION According to estimates produced by Emergency Management Australia, the impact of significant natural disasters in Australia, excluding drought, costs the community, on long-term average, $1.25 billion annually. That equates to: (i) the total annual value of mineral royalties earned by all governments in Australia; or (ii) the total amount spent in Australia in 1996 on oil and gas exploration; or (iii) about half the value of the annual Australian wool clip. The 1989 Newcastle earthquake, one of Australia's worst natural disasters, killed 13 people and severely injured almost 200 others. In addition, it generated an insured loss of almost $1 billion and a total economic cost of perhaps $4 billion. As our communities grow and our buildings and infrastructure age, the risks posed by all natural hazards grows accordingly. It is not surprising, therefore, that there is a considerable interest in reducing the human, economic and environmental cost of natural disasters. In 1996 the Australian Geological Survey Organisation (AGSO) recognised the important role it could play in this process by establishing the National Geohazards Vulnerability of Urban Communities Project (better known as the Cities Project). This public-benefit science effort has brought together AGSO's expertise across a range of disciplines with the express purpose of undertaking research that will lead to safer, more sustainable and consequently more prosperous communities. This effort formed a significant part of the Commonwealth Government's contribution to the 1990s United Nations International Decade for Natural Disaster Reduction and the development of national risk-mitigation strategies.

GEOHAZARDS Under the Cities Project, we regard geohazards broadly as Earth processes with the potential to cause loss or harm to the

community or the environment. The work of the Cities Project is focused on the risks posed by the more acute and potentially fatal geohazards, especially earthquakes, landslides and inundation hazards (e.g. flood, storm tide and tsunami). The risks posed by the more chronic geohazards, such as dryland salinity, reactive clays, groundwater contamination and coastal erosion, clearly cause significantly greater economic harm, however, they do not pose the same level of threat to life. The risks they pose, however, may be considered at a later stage of the project.

RISK The Cities Project marks a significant departure for AGSO. As the national geoscience research agency, its focus for more than 50 years in the field of geohazards has been exclusively on hazard phenomena, especially earthquakes and regional volcanoes. The study of risk is much more complex. In developing its approach to risk reduction activities, the Cities Project has adopted the principles of risk management established in the Australian and New Zealand Standard AS/NZS 4360-1995 and its subsequent revision in 1999 (Standards Australia 1999). This generic guide provides the philosophical framework within which the Cities Project studies are developed. That process is outlined in Figure 6.1. The risk management standard (Standards Australia 1999 p. 3) defines 'risk' as: 'the chance of something happening that will have an impact upon objectives. It is measured in terms of consequences and likelihood.' This definition is really too general for our purposes, consequently we have chosen to follow the conceptual basis developed under the Office of the United Nations Disaster Relief Coordinator in 1979 and cited by Fournier d'Albe (1986) as follows: * Present address: 68 Ann-Mariee Drivee, Caboolture, Qld 4510, Australia (riskscience@bigpond.com.au).

K. Granger

(e.g. by evacuating people before the flood arrives or the volcano erupts); and/or reducing vulnerability (e.g. by improving building standards). This outcome is the principal objective of risk management.

RISK IDENTIFICATION

Figure 6.1 Risk management overview (after Standards Australia 1999 figure 3.1)

'Natural hazard means the probability of occurrence, within a specified period of time in a given area, of a potentially damaging natural phenomenon. Vulnerability means the degree of loss to a given element at risk or set of such elements resulting from the occurrence of a natural phenomenon of a given magnitude. Elements at risk means the population, buildings and civil engineering works, economic activities, public services, utilities and infrastructure, etc., at risk in a given area. Specific risk means the expected degree of loss due to a particular natural phenomenon: it is a function of both natural hazard and vulnerability. Risk (i.e. 'total risk') means the expected number of lives lost, persons injured, damage to property and disruption of economic activity due to a particular natural phenomenon, and consequently the product of specific risk and elements at risk'. Total risk can be expressed simply in the following pseudo-mathematical form: Risk(Total) = Hazard x Elements at Risk x Vulnerability This approach is not only elegant, it is also very practical. Given the complexity of urban communities and the degree to which the various elements at risk are interdependent, the 'total risk' approach is considered mandatory. It also lends itself well to the consideration of multiple hazards, many of which may be interrelated. For example, cyclones may produce storm tide, flood and landslide as well as severe wind hazards—traditionally, each of these phenomena has tended to be studied separately. The 'total risk' approach also lends itself to quantitative, qualitative and composite analytical approaches. Risk reduction (or 'mitigation' to use the term employed in the standard), then is seen as any activity that moderates the severity of the impact, either by eliminating, or reducing the magnitude of, the hazard (generally not possible); reducing the elements at risk that are potentially exposed

The first stage of the risk management process, once the study area (the context) has been defined, is to develop a good understanding of the history of, and potential for, hazard impacts, their consequences and the response of the affected community to that experience. A significant proportion of this information is derived from the networks of terrestrial and satellite-borne instruments that constantly monitor the environment for, amongst other things, the development of severe weather events, floods or developing volcanic activity; or to detect and locate the occurrence of bushfires and earthquakes. AGSO maintains national databases of earthquakes, landslides and tsunamis, whilst the Bureau of Meteorology maintains comprehensive databases of floods and severe weather events such as cyclones. Whilst this information on the hazard phenomena is invaluable, it is equally important to balance it with an understanding of the human experience of the event and the response or reaction of the community to that experience. Perhaps the best developed collection of such experiential information is that developed by the Newcastle Region Library relating to the 1989 earthquake in that city. This vital component in understanding community risk, however, is still very poorly done in most Australian communities.

RISK ANALYSIS AS/NZS 4360:1999 (Standards Australia 1999 p. 3) defines 'risk analysis' as: 'a systematic use of available information to determine how often specified events may occur and the magnitude of their likely consequences'. We have identified three distinct aspects of this process.

Phenomenon process knowledge The focus of hazard-science research is on the mechanisms that cause, create, generate or drive the hazard phenomena, e.g. what causes earthquakes and what influences the transmission of their energy through various strata. This is underpinned by information relating to the background climatic, environmental, terrain, ecological and geological aspects of the site that are relevant to hazard studies, e.g. the depth and nature of the sediments and their microtremor response. Whilst there is little that can be done to eliminate or reduce the severity or frequency of these phenomena, a good understanding of what drives them enhances our ability to forecast or predict their behaviour. It is also fundamental to establishing an understanding of event probabilities. In the Cities Project's Cairns case study, for example, a considerable effort went into capturing detailed microtremor and geotechnical (borehole) data on which to base the analysis of the earthquake hazard and detailed geomorphological and elevation data to underpin the analysis of land-

Geohazards risk in Australia Table 6.1 Relative contribution of building characteristics to vulnerability.

Building age Floor height or vertical regularity Wall material Roof material Roof pitch Large unprotected windows Unlined eaves Number of storeys Plan regularity Topography

Flood

Wind

Hail

Fire

Earthquake

** *

*****

* * * * *

* * * *

*****

* * *

*****

* * * *

*****

* * * *

* * *

* * * *

***

*****

* * * *

*****

***

* * * *

* * * * *

*****

* * *

*****

* * * *

** *

The number of stars reflect the significance of each attribute's contribution to building vulnerability. slide and inundation hazards such as flood and storm tide. A detailed technical report on the Cairns landslide study, for example, is provided in Michael-Leiba et al (1999). These data were required to supplement and improve the resolution of the traditional medium- and small-scale mapping of themes such as geology, soils and terrain.

Elements at risk and their vulnerability This is a relatively new area of study and is focused on developing an understanding of the vulnerability of a wide range of the elements that are at risk within the community i.e. the people, buildings and infrastructures. It involves disciplines as diverse as geography, demography, psychology, economics, engineering and anthropology. In the Cities Project case studies undertaken to date, a significant effort has been made to develop very detailed data on the principal elements at risk in the built environment, whilst comprehensive statistics of good resolution are available from the quinquennial national censuses to provide at least basic measures of human vulnerability. The broad groups of elements at risk for which data have been collected include the following. SETTING Basic regional data have been accumulated for themes including the physical environment (climate, vegetation, geology, soils, land use, topography, elevation, etc.), access (external links by major road, rail, air, marine and telecommunications infrastructures), population numbers and distribution, and administrative arrangements (local government, suburb and other boundaries). SHELTER The buildings that provide shelter to the community at home, at work and at play, vary considerably in their vulnerability to different hazards. A range of information relating to their construction is required. These building characteristics contribute to the relative degree of vulnerability associated with exposure to a range of hazards (Table 6.1). Access to shelter is also significant, so information on mobility within the community is needed. Details of the capacity and vulnerability of the road and other transport networks, for example, are required.

SUSTENANCE Modern urban communities are highly reliant on their utility and service infrastructures such as water supply, sewerage, power supply and telecommunications. These so-called lifelines are significantly dependent on each other and on other logistic resources such as fuel supply. The community is also dependent on the supply of food, clothing, medicine and other personal items. Information has been accumulated on all of these, as well as on the enterprises that wholesale, distribute and service these sectors (such as transport, material handling equipment and storage). SECURITY The security of the community can be measured in terms of its health and wealth and by the forms of protection that are provided. Physically, these may be assessed by the availability of hospitals, nursing homes, industries, commercial premises, agricultural land use, ambulance stations, fire stations, police stations and works such as flood-retention basins and levees. Also important are sociodemographic and economic issues related to the elderly, the very young, the disabled, household income, unemployment, home ownership and the resources available at the fire and police stations. Emergency plans are also a key component of community security. SOCIETY Here we find some of the more intangible measures such as language, ethnicity, religion, community and welfare groups, education, awareness, meeting places, cultural activities and so on. Some of these may be measured in terms of the facilities that they use, such as schools, churches and sporting clubs, however, the more meaningful measures relate specifically to the individuals, families and households that make up the community. SUMMARY Whilst these data provide a detailed quantitative description of specific aspects of the community's risk environment, they do not, of themselves, provide an adequate measure of overall community vulnerability. Nor do they individually reflect the relative levels of vulnerability across the city. To overcome these shortcomings we developed a

K. Granger

Table 6 . 2 Measures used to develop the vulnerability profile for Cairns. Total population Number of houses Number of flats Number of hotels, motels, etc Mean house occupancy 3 Mean flat occupancy 3 Total length of road Number of carsa Households with no car 3 Length of lifelines (power, water, etc) Number of logistic facilities Number of power & water supply sites Number of telecommunications sites a b

Number of public safety & medical sites Number of commercial & business sites Area under sugar cane Socio-economic disadvantage index b Economic resources index b Proportion of people under 5 years 3 Proportion of people over 65 years 3 Proportion of households renting 3 Number of community & public facilities Proportion of visitorsa Education & occupation index b Proportion of new residents (<5 years) 3 Proportion with no religious adherence 3

derived from the summaries of the 1996 census provided in CData96 (Australian Bureau of Statistics 1998a) derived from the Socio-Economic Indexes for Areas (Australian Bureau of Statistics 1998b)

vulnerability profile of Cairns (Granger et al 1999). In this, we rated each of the city's 41 suburbs according to their contribution to overall community risk based on the number and nature of the elements at risk they contain. Twentysix measures were used to develop this vulnerability profile for Cairns (Table 6.2). In subsequent work in Mackay (Middelmann & Granger 2000), some variables, such as the area of sugar cane, were dropped and other variables added. These included the number of road intersections (to give a better measure of network density), population density and, following the work of Morrow (1999), social measures, such as the number of large families and the number of single-parent households, have been added. Development of this technique and the measures included is still evolving.

Synthesis and modelling Clearly, the range and variety of information needed to fuel a comprehensive risk analysis is enormous. Granger (1998), for example, provided a detailed description of the datasets assembled for the Cairns study, their sources and the information infrastructure established to manage their use. Whilst there are many sources now available from which such information can be captured or derived, much of it with the essential spatial and temporal attributes needed, there remain important gaps. Our knowledge of hazard phenomena and the processes that drive them, for example, are far from perfect. It is necessary, therefore, to develop appropriate models to fill the knowledge gaps. The behaviour of some hazards, such as floods, have an established body of modelling research behind them, whilst others, such as cyclones and earthquakes are, as yet, less well served. The synthesis of data and the essential mapping of the spatial relationships between the hazard phenomena and the elements at risk requires the use of tools such as geographic information systems (GIS). In the work undertaken in the Cairns case study, at least 90% of the information used has some form of spatial content. Similarly, the relationships that are most significant in risk analysis are largely spatial. Whilst GIS have been used for almost two decades as tools to address specific aspects of the risk-management

problem, especially in hazard mapping and the spatial modelling of phenomena such as bushfires or storm-tide inundation, there are few examples of integrated risk-management applications. There are obvious advantages in developing a fusion between a philosophy of risk management and the power of GIS as a decision support tool, hence Risk-GIS, as it has been christened in the Cities Project.

RISK EVALUATION AS/NZS 4360:1999 (Standards Australia 1999 p. 3) defines 'risk evaluation' as: 'the process used to determine risk management priorities by evaluating and comparing the level of risk against predetermined standards, target risk levels or other criteria'. We see three key components to this.

Scenario analysis Scenario analysis is an emerging technique that contributes to 'future memory', an understanding of 'what will happen when...'. The output embraces forecasts or estimates of community risk including economic loss and potential casualties, or assessments of the impact of secondary or consequential hazards, such as the release of hazardous materials following an earthquake. It also provides essential input to both the development of risk-treatment strategies and to framing long-term impact forecasts or estimates. To cover the widest range of potential disaster events, these scenarios typically extend from the relatively small and more frequently occurring events to those in the so-called 'maximum probable' or 'maximum credible' range. We try to develop scenarios for events with an average recurrence interval of 50, 100, 500, 1 000 and 10 000 years. It is, however, not yet practical to apply all of these to hazards such as earthquakes because of the range of uncertainty that exists in their average recurrence interval estimates. Uncertainty aside, this approach enables us to not only assess the impact of a single event of specific magnitude, such as those established for planning or building code purposes (e.g. a 'Q100' flood), it also provides the basis for assessing cumulative risk. This approach provides a more

Geohazards risk in Australia

safe, sustainable & prosperous communities

Figure 6.2 Cities Project understanding of the risk management process.

realistic assessment of both the economic and social cost of hazard impact. Acceptability

In the approach to risk assessment set out in Standards Australia (1999), it is the practice to compare the level of risk found during the evaluation process with previously established risk criteria, so that it can be judged whether the risk is 'acceptable' or not. At first glance this may seem to be something of a chicken-and-egg process—if you do not know what the level of risk posed by earthquake is in Cairns, for example, how can you realistically determine what level of risk is acceptable? Levels of acceptability are, however, built in to such things as urban-planning design constraints and the Australian Building Code, where criteria are based on design levels. For example, under the earthquake loading code, AS 1170.4-1993 Earthquake Loads (Standards Australia 1993), the 'design level of earthquake shaking' is one in which there is an estimated 10% probability of ground motions being exceeded in a 50 y period, i.e. the acceptability criterion is set at a 10% chance of exceedence over the lifetime of a typical building. Similarly, many urbanplanning schemes establish constraints on development within the area likely to be affected by a flood or storm tide of a given average recurrence interval (typically 100 y). Risk prioritisation

Regardless of the quantification of risk achieved, the process of risk prioritisation remains largely subjective because of the acceptability factor. Prioritisation, however, is the first step in the allocation of resources to risk mitigation, especially if considered in a multihazard context. We are beginning to address the complex issue of comparing the risks posed by hazards with greatly different impact potential. In Cairns, for example, there is a strong spatial correlation between the areas that are most at risk from

major inundation hazards (river flooding, storm tide and tsunami) and those in which sediments are most likely to maximise earthquake impact. Conversely these are the areas that are at least risk from landslide impact and, to some degree, from severe wind impact. The ultimate responsibility for determining what levels of risk are 'acceptable' rests with the community and the respective local authority. RISK-MITIGATION STRATEGIES

Whilst the role of AGSO and the Cities Project is concerned primarily with risk identification, analysis and evaluation, these processes provide some insight into the risk-mitigation process. Monitoring and surveillance

One of the principal sources of historical hazard-event information and hazard-phenomenon knowledge is the extensive network of monitoring stations and remote-sensing resources that have been established across Australia. For example, AGSO has access to more than 150 seismographs across the country, whilst the Bureau of Meteorology maintains some 45 weather radar sites, 246 automatic weather stations and 3029 stream-gauging stations. The Bureau of Meteorology also takes data from the Japanese Geostationary Meteorological Satellite 26 times a day in addition to data taken from the polar orbiting US NOAA satellites. Warnings and forecasts

An effective warning and forecasting system, combined with a high level of community awareness and risk appreciation, is clearly one of the most potent mechanisms by which to achieve risk reduction. These are typically taken to mean short-term warnings, such as those issued by the

K. Granger

Bureau of Meteorology for the hazards that can literally be seen coming, such as cyclones, floods and severe storms. They may, however, also embrace the longer term estimates of the 'hazardousness' of areas such as those contained in the earthquake acceleration coefficient maps that accompany AS1170.4-1993. They can both be significantly enhanced through the scenario-analysis process.

Mitigation strategies and response options Risk assessments are made so that strategies may be developed that ultimately will lead to the elimination, reduction, transfer or acceptance of the risks, and to ensure that the community is prepared to cope with a hazard impact. Increasingly, the emphasis is shifting from structural responses, such as building levees and other 'flood mitigation' works, to non-structural responses. These include: well-maintained and appropriate information about risk linked to comprehensive monitoring and warning systems; widespread and ongoing community awareness programs based on risk history, scenario analysis and an effective risk communication capability; and emergency management plans, resources and training based on risk assessments.

Summary The key components of the Cities Project's understanding of the risk-management process are illustrated in Figure 6.2. The bottom line is that if we get all of this right, the outcome will be safer, more sustainable and more prosperous communities.

PROGRAM The Cities Project research and technique development program is evolving through a series of multihazard case studies of increasing complexity and scale. There are four distinct series of case studies involved. (1) Earthquake microzonation projects of Homebush Bay (New South Wales), Launceston (Tasmania) and Adelaide City (South Australia). These studies actually predate the establishment of the Cities Project but can be seen as the first steps in the evolution of the Project's approach to risk studies. (2) Comprehensive multihazard risk studies of the Queensland centres of Cairns, Mackay, Gladstone and the southeast Queensland urban complex (Caboolture to the Gold Coast). These are the core research and developmental case studies.

(3) The Newcastle 99 Project, in which the experience of the 1989 earthquake is being used to calibrate and validate the risk analysis and risk-evaluation techniques, models and tools against the reality of Australia's only major urban centre earthquake disaster. (4) A series of subject-specific studies including a landslide risk assessment of the Wollongong region and studies of the earthquake risk to lifeline utilities in the Australian Capital Territory and areas of Sydney such as the Botany Basin. Output from this program includes the publication of scientific and general studies, presentations to a wide range or conferences and workshops and comprehensive databases and mapping of the hazard phenomena and the elements at risk in each community. Priority is also given to the production of community awareness material, such as the Landslide Awareness and Shore Safety brochures produced in conjunction with Emergency Management Australia and Surf Lifesaving Australia respectively. These, and details of other Cities Project activities, are available in the Geohazards directory of the AGSO Home Page at www. agso .gov. au.

REFERENCES BUREAU OF STATISTICS 1998a. CData96, Final release. Australian Bureau of Statistics, Canberra. 1998b. Socio-Economic Indexes for Areas. Australian Bureau of Statistics, Canberra.

AUSTRALIAN

AUSTRALIAN BUREAU OF STATISTICS

FOURNIER D'ALBE E. M. 1986. Introduction: reducing vulnerability to

nature's violent forces: cooperation between scientist and citizen. In: Maybury R. H. ed. Violent Forces of Nature, pp. 1-6, Lomond Publications, Maryland. GRANGER K . 1 9 9 8 . ASDI (Australian Spatial Data Infrastructure) from the ground up: a public safety perspective. Australia New Zealand Land Information Council and Australian Geological Survey Organisation, Canberra. GRANGER K . , JONES T . , LEIBA M . & SCOTT G . 1 9 9 9 . Community risk in Cairns: a multi-hazard risk assessment. Australian Geological Survey Organisation, Canberra. MIDDLEMANN M . & GRANGER K. 2 0 0 0 . Community risk in Mackay: a

multi-hazard risk assessment. Australian Geological Survey Organisation, Canberra (unpubl.).

MICHAEL-LEIBA M . , BAYNES F. & SCOTT G . 1 9 9 9 . Quantitative landslide

risk assessment of Cairns. Australian Geological Survey Organisation Record 1999/36. MORROW B. H. 1999. Identifying and mapping community vulnerability. Disasters No. 23, Vol. 1, pp. 1-18. Overseas Development Institute, Oxford. STANDARDS AUSTRALIA 1 9 9 3 . Minimum design loads on structures Part 4: Earthquake loads AS 1170.4-1993. Standards Australia, Homebush. STANDARDS AUSTRALIA 1999. Australia New Zealand Standard AS/NZS 4360:1999 Risk management. Standards Australia, Homebush, and Standards New Zealand, Wellington. Received 27 April 1999; accepted 5 October 2000

Geological Society of Australia Special Publication 21, 67-74

CHAPTER 7—Land instability: a case study from the lllawarra region, New South Wales B. G. JONES1, R. N. CHOWDHURY2 AND M. H. GHOBADP ? School

of Geosciences, University of Wollongong, NSW 2522, Australia. of Civil Mining and Environmental Engineeringi University of Wollongong, NSW 2522, Australia. of Geology, Shahid Chamran University, Ahvaz, Iran

2 Department

3 Department

Land instability and landslips form part of the natural erosional and geomorphological processes occurring along the steep escarpment in the lllawarra region. Such mass movements result from the interplay between slope, rock strength, mineralogy, groundwater seepage and earthquakes. They are commonly accelerated by human activities which involve disturbance of equilibrium and environmental changes. This is especially important in the lllawarra region where steep slopes along the escarpment, a volcanic-rich rock sequence, high periodic rainfall events, extensive clearing of natural vegetation and slope modification following European settlement have all contributed to an increase in land instability in the region. Land-management policies must include methods for monitoring movements and assessment of landslide hazard, and the attendant risks to life, property and the environment in general. Detailed geomorphological and geotechnical investigations and maps are essential to allow suitable planning and design that would minimise the amount of damage caused by land movement. KEY WORDS: engineering geology, landslides, natural hazards, remediation, slumping.

INTRODUCTION Land instability can occur in any area where a mass of soil or rock is present in an unstable configuration on a slope. A general overview of slope processes and their effects on slope profiles and landscape evolution is provided by Ritter et al. (1995). Many landslides are initiated when the gravitational forces exceed the inertial forces resisting potential failure and movement. In general, the forces resisting potential movement relate to the shear strength along any potential surface of sliding. Shear strength in any frictional particulate material increases with increase both in the effective stress and the angle of shearing resistance. However, the effective normal stress on a plane decreases as the pore-water pressure increases. Often rainfall increases the pore-water pressures in sloping soil and rock masses and, therefore, the shear strength along any potential surface of sliding decreases. Thus land instability often occurs during or after heavy or prolonged rainfall. This is why the recent Thredbo landslide disaster in the Snowy Mountains region of Australia, which occurred at 11 20 pm on 30 July 1997 and claimed the lives of 18 people, has not yet been clearly understood. The landslide occurred apparently without warning and was not the result of rainfall immediately preceding the time of sliding. Land instability often causes adverse economic, environmental and social impacts, especially in urban sloping areas. Assessment of landslide hazards and risks is, therefore, very important for planning and decision making concerning further development, as well as landslide management including preventive and remedial measures. Accurate mapping of geology and landslide areas can be

facilitated by the use of a Geographic Information System (GIS). The use of GIS, together with a suitable landslide database, for the development of an effective hazard assessment strategy has been discussed recently by Flentje and Chowdhury (1996) and Chowdhury and Flentje (1996). Effective strategies for hazard and risk assessment require a thorough understanding of the causes and mechanisms of slope failure. Knowledge of regional and local geology and geomorphology is essential and quantitative assessments require geotechnical testing and stability analyses to be carried out.

MASS MOVEMENT Understanding the form and processes governing mass movement has increased considerably over the last century, and the variety and complexity of moving soils and rocks has become more obvious. The term mass movement is used here as a general term (after Varnes 1978) to include rockfalls, topples and debris flows, which involve little or no true sliding, and spreads, slides and flows where movement occurs along distinct slide planes. Gravity is the principal driving force; the movement is directed downwards and outwards on a sloping surface, and the displaced material may include soil, bedrock and/or constructed fills. Soil creep, a slow movement which occurs without a well-defined failure surface, is not considered here although many landslides creep before readily observed movement occurs. In urban areas prolonged creep movements can cause significant damage to some structures.

B.G.Jones etal.

Factors causing mass movements It is important to recognise the factors, processes and conditions that fundamentally cause a slope to become unstable, and also the events and factors that trigger the eventual failure and movement. Rainfall and earthquakes have long been known to contribute to slope instability and they may trigger slope failure. Environmental influences on slope stability include both external changes which increase the shearing stress, and internal changes which decrease the shearing resistance: these causes and processes were recognised and clearly explained by Terzaghi (1950). In many regions of the world rainfall is one of the main factors controlling the frequency of landslides. The magnitude of its influence depends on the frequency and rate of precipitation, topography, changes in temperature (especially freeze-thaw), and the permeability and other properties of the rocks and soils. Geomorphological information about slopes is essential for understanding the complex phenomena and many interacting factors which control slope behaviour. Landslides are more common on steep slopes and it is usually possible to determine lower limits (thresholds) of inclination below which landslides are unlikely to occur in specific areas. However, threshold slopes depend on local geological, geotechnical and hydrogeological conditions. Thus while a threshold slope angle of 11° has been interpreted for slopes on the Wianamatta Shale at Razorback, south of Sydney (Dunkerley 1976), at West Pennant Hills some landslides have occurred on the same unit with slope inclinations of 6° (Fell 1985). A good geological understanding of an area often provides the basis for assessing slope stability. Therefore, considerable research effort has been directed towards quantifying the influence of geological factors such as stratigraphy, sedimentology, petrography, weathering, structure, groundwater and earthquakes on slope stability. Depositional environments control the distribution of sedimentary strata and the physical and chemical properties of these facies affect slope stability. Particular attention should be directed towards recognition of thick sequences of weak rocks (claystone and siltstone) that are very important in slope development and have a notoriously high landslide potential. However, minor geological details can also be critically important. For example, thin layers of clay in an otherwise dense or strong soil can lead to instability and even complete failure. Thin layers of water-bearing sand also can have consequences adverse to stability. Likewise, coal seams commonly act as aquifers and their underclays often contribute to failure surface development in colluvial slopes. The percentages of principal rock-forming minerals and cements in a sequence also provide a key index to its mechanical performance. The total clay content, and especially the percentage of swelling clay minerals, are useful indicators of the potentially plastic behaviour of shale (Franklin & Dusseault 1989). The engineering behaviour and residual strength of colluvium derived from argillaceous rocks also depends on both its original and weathered petrography. In many landslide events chemical alteration, such as hydration and ion exchange in clays, are thought to have contributed to the development of instability and progression to landsliding (Zaruba & Mencel 1982).

Progressive action and progressive failure are important aspects of landslide mechanisms. Often engineering stability analyses are performed on the premise that failure along a slip surface is 'simultaneous'. This approach often works reasonably well within the limited context of design of slopes or remedial measures. However, in many cases, a full understanding of slope performance is not gained on this basis. Processes leading to the development of continuous slip surfaces in otherwise homogeneous earth masses are also progressive. Both spatial and temporal aspects of progressive action must be considered and methods should be developed to model these processes (Chowdhury 1976, 1980). Because of the many uncertainties concerning slopes and landslide processes, probabilistic approaches have also been developed to supplement deterministic approaches (Chowdhury 1992; Chowdhury & Zhang 1993). Two main groups of geologic factors distinguish slope stability problems in soil from those in rock. In general, rock masses, because of structural discontinuities, are best considered as possessing anisotropic strength, permeability and deformability characteristics to a much greater and more significant degree than soils. The presence, continuity, spacing, orientation and nature of bedding planes, joints, faults and shear zones in the weathered rock beneath the soil will have a major effect on the mass shear strength, the shape of the potential slip surfaces and on the hydrological conditions. In many cases, therefore, the geological discontinuities may control landsliding (Bell 1993). A knowledge of the location and orientation of zones having low shear strengths, due to previous displacements, chemical alteration and/or presence of clay gouge, is very important. Regional and local groundwater conditions are often critical to slope stability. Slope failures are commonly associated with high groundwater levels or high pore-water pressures following prolonged or intense rainfall. Groundwater levels fluctuate much more in jointed rock slopes than in many cohesive soil slopes due to the more open joint systems in rock masses. Rapid seepage from a permeable sequence can stabilise a region whereas if the rock mass is very tight, or if the soil mass is impermeable, excessive pore-water pressure can exist long after a rainfall or rainstorm event. Water pressure reduces the normal effective stress in a substrate resulting in shear strength decrease; consequently draining a rock or soil mass will reduce the water pressure and increase the shear strength. Many sloping rock masses have open joints due to stress relief which serves to retard the development of high water pressures, especially near the slope surface. Almost every earthquake in mountainous country has caused at least some landslides and/or rockfalls as a result of horizontal and vertical accelerations in the soil or rock masses. Earthquakes can also change the magnitude and distribution of the pore-water pressure in relatively loose cohesionless materials, thus reducing soil shear strength. Rapid increase of pore-water pressure in some coarsegrained soils occurs by repeated shear stress fluctuations. In loose sandy soils this cyclic shear loading may lead to liquefaction, i.e. total loss of shear strength. Keefer (1984) suggested that rockfalls, rockslides, earthfalls and earthslides can be triggered even by very weak seismic activity,

Land instability, lllawara region NSW

Figure 7.2 Geological cross-section of the escarpment in the Clifton area showing the relationship of debris deposits to the underlying geological units. The base of the cliff consists of a wave-swept rock platform. See Figure 7.3 for location.

Figure 7.1 Major regions in Australia affected by landslides (shaded) (after Blong & Eyles 1989). The Illawarra study area just south of Sydney is indicated by the arrow. while deep-seated slumps and earthflows are generally initiated by stronger and perhaps longer ground shaking. As Zaruba and Mencel (1982 p. 144) noted, 'preventive measures in a landslide or an area susceptible to sliding must be based on a detailed, integrated geological and geotechnical investigation. It is necessary to study the geological structure of the area, the petrographical and physical properties of rocks, and the local hydrogeological conditions. As the form of a slope is the end product of past geological processes, the morphological history of the slope must also be understood'. These statements certainly apply to the Illawarra area where marginally stable colluvial masses exist on many slopes.

ILLAWARRA REGION In the urban areas of New South Wales, landslides are common in the Illawarra, Campbelltown-Picton, Gosford-Wyong and Newcastle-Lake Macquarie regions (south, southwest and north of Sydney), and in the Lismore region (Figure 7.1). The Illawarra coastal escarpment has long been recognised as a region of significant landslide activity (Bowman 1972; Chowdhury 1976; Young 1978; Chestnut 1981; Chowdhury & Young 1987; Pitsis 1992; Ghobadi 1994, 1995; Flentje & Chowdhury 1996). The 300 m-high escarpment consists of flat-lying Permo-Triassic volcaniclastic coal measures plus fluvial sequences capped by a stable quartz-cemented quartz sandstone (Figure 7.2). The lower Narrabeen Group and upper Illawarra Coal Measures consist of interbedded strong lithic sandstone and weak shale, claystone and coal seams. This succession in the escarpment has been acted upon by erosion, stress relief, weathering, creep and sliding processes to produce masses of marginally stable colluvial soil slopes as well as zones of potentially unstable colluvium and rock masses on many of the steep hillsides. The majority of slope instability prob-

lems of economic significance in northern Illawarra involve translational or rotational slides, or extremely slow to rapid flows of soil, colluvium or constructional fill. The equilibrium of colluvial masses is frequently upset by prolonged periods of precipitation and by human activities, e.g. removal of toe support, loading the slope, and changing the surface and subsurface drainage. Such slope instability has had disastrous effects on urban areas as well as on road and rail routes. As long ago as 1890, Shellshear reported on the treatment of slips on the Illawarra railway at Stanwell Park. This was at the time when the railway line occupied the position of the present coastal road (Lawrence Hargrave Drive) which has a long history of slippage, rockfalls and mudslides. This is because the slopes on which the road is constructed consist of colluvium underlain by weak strata of the Wombarra Shale and are particularly vulnerable during periods of heavy rain.

Main causes of landslides in the Illawarra area Observation and study of areas of instability in the Illawarra region have shown that most slides are associated with specific sequences of the Illawarra Coal Measures and the Narrabeen Group. Colluvium cover on the slopes ranges from 0 to 20+ m. Depending on their topographical position, these deposits may be derived from the Hawkesbury Sandstone, the Narrabeen Group, the Illawarra Coal Measures or a combination of these older strata. Landslide occurrences in the Illawarra area are related to the following factors. PROPERTIES OF THE COLLUVIUM The colluvium varies in composition from a coarse sandstone debris with some clay matrix to a finer and more clayrich material that may still contain large sandstone blocks. With any heavy rain, pore-water pressures rise, decreasing the shear strength of these surficial deposits. As the shear strength reduces to the level of the applied shear stress, the clay matrix is mobilised and may lead to slow, moderate or rapid mass movement. Where natural drainage occurs, the colluvium has become consolidated and, with compaction and deposition of cementing materials from solution, it is relatively impervious to percolating waters. Such colluvium

B.G.Jones etol.

can be stable, especially on gentle topographic slopes, when pore-water pressures are low. However, many failures have been documented on these gentle slopes which involve excess pore pressure during periods of above average antecedent rainfall. PROPERTIES OF THE ILLAWARRA COAL MEASURES AND NARRABEEN GROUP The Illawarra Escarpment consists of a repetitious sequence of lithic sandstone interbedded with shale and coal. In the Narrabeen Group, thick sequences of weak rocks (Stanwell Park Claystone and Wombarra Shale) are more easily eroded than the sandstone strata and cause relatively rapid rates of recession to occur. Undermining along the contact between claystone and sandstone reduces support for the overlying vertically jointed sandstone and eventually leads to slabs falling off along the vertical joint faces. Thus rockfalls and toppling are common on the steep slopes in the area, especially in the upper parts of the escarpment. Jointed sandstone beds and coal seams in the Illawarra Coal Measures commonly act as aquifers, with claystone beds acting as aquitards. Slope instability is, in some cases, related to the presence of these aquifers which are a source of high pore-water pressures. PETROLOGY OF ILLAWARRA COAL MEASURES AND NARRABEEN GROUP Both the sandstone and shale beds in the upper Illawarra Coal Measures and Narrabeen Group were derived largely from the New England Fold Belt to the north and consist predominantly of volcanic rock fragments and finer volcanic ash. During post-depositional alteration, the original volcanic glass devitrified to produce smectite clays. These clays not only cause swelling and shrinkage near the surface as a response to wetting and drying, but also reduce the permeability near the surface of the rock mass. This latter factor has the potential to increase the aqueous pore pressures and hence increase the likelihood of surficial mass movement of both the rock mass and the overlying or adjacent colluvial deposits. Loss of carbonate cement during weathering accelerates the processes of shear strength decrease. LOCAL STRUCTURAL GEOLOGY The relationship of landslides to structural geology in the Illawarra area is marked. The high horizontal-stress environment known to exist in this area (Hilleard 1993) has a strong influence on slope failure since stress release at the escarpment face causes a near-surface opening of prominent joint sets, with an orientation between 005° and 025°, allowing easy ingress of water and clay. The regional geological dip of about 5° to the north-northwest causes joint blocks initially to tilt landward away from the escarpment. Stress-relief opening of joints parallel to the escarpment often negates the effects of this tilt. The intersection of joints and bedding planes produces additional surfaces of weakness which, together with lithological changes (interbedded sandstone, shale, claystone and coal), aids the development of slope instability in the Illawarra area.

Figure 7.3 Location map showing the relationship between the Illawarra escarpment and the coast. Potentially unstable conditions occur along most of the escarpment and the locations of other figures are noted.

The massive Scarborough Sandstone, with a typical joint spacing of up to 4 m, produces large joint blocks. In contrast, the Bulgo Sandstone, which is typically highly fractured with joint spacings of 1-50 cm, produces smaller sized joint blocks. As a result, more rockfalls occur from the Bulgo Sandstone than from the Scarborough Sandstone in the study area, but falls of Scarborough Sandstone are usually much larger and more destructive. Local structural features such as fractures and faults influence the underground water circulation and appear to be directly related to landslides. Even after prolonged dry spells the fault zones remain wet and hence a relatively small rainfall can thoroughly saturate the colluvium in the fault zone, where it is already in a marginally stable position. GEOTECHNICAL PROPERTIES OF ROCK AND COLLUVIUM A significant decrease in strength and slake durability was found to occur with a change in mineralogy and an increase in weathering from fresh to weathered rocks. Presence of carbonate cement (siderite or calcite) in the fresh shale causes high durability. Moderately and highly weathered claystone and shale in Narrabeen Group rocks have low to very low durability depending on the quantity of swelling clays versus remaining cement. In the Illawarra region durability of claystone beds in the sandstone units generally decreases upwards through the Narrabeen Group. Thus claystone interbedded in the Bulgo Sandstone and the highly weathered underlying Stanwell Park Claystone both

Land instability, lilawara region NSW have very low durability. This has a significant effect on slope stability in the Bulgo Sandstone, especially where the Stanwell Park Claystone acts as the bedrock for the colluvial mantle between Clifton and Stanwell Park (Figures 7.2, 7.3). Sandstone units in the Narrabeen Group, which contain expansive clay minerals and volcanic rock fragments, show significant strength loss on wetting. High proportions of expansive clay minerals were detected in volcanic rock fragments which suggest that clay softening in the presence of water is important in controlling moisture-related reduction of strength in sandstone in the Illawarra area. Some of the weathered sandstone units become completely disaggregated on 4 cycles of the slake durability test (wetting and drying). The geotechnical properties of the surficial deposits most related to their stability are the cohesion (c) and the angle of internal friction (([>), and a guide to these values is provided by the clay content and the plasticity index of the soil. These parameters, and the observed inclination of natural slopes, lead to the conclusion that many colluvial slopes in the Illawarra area often become unstable in the long-term at inclinations above approximately 10°. It is now well-known that the angle of internal friction reduces after large relative displacement along any surface within a soil, from a peak value, <(> , to a residual value, fr. The residual cohesion, c|>r, along slip surfaces in clays approaches zero. The long-term stable inclination of natural slopes may thus provide a good guide to the residual friction angle of the slope materials in a particular area. MARINE EROSION The vertically jointed coastal cliffs are sites for toppling and rockfalls because the toe of the slope is eroded by the sea. Fretting and weathering of low-strength shale and weathered volcanic sandstone exposed at sea-level also produces some falls and topples. INFLUENCE OF WATER Increase in the pore-water pressure in a colluvial deposit affects its stability adversely. The very low permeability of colluvial material, combined with a lack of adequate drainage, can cause a rapid increase in the pore-water pressure after heavy rainfall. This leads to a decrease in the shearing resistance of the surficial material, which often results in the onset of slope instability. Fracture permeability is the most important feature controlling groundwater movement, with most of the fractures occurring in areas of stress relief. Horizontal water flow within the rock mass is concentrated along the base of sandstone units, i.e. between claystone and sandstone beds. This increases the rate of weathering of the underlying claystone and causes fretting and weathering of the sandstone as well as larger toppling failures. Perched watertables are quite common in the Illawarra area because the many claystone sequences in the Narrabeen Group are relatively impermeable. These increase the potential hydraulic head leading to failure. It has been suggested (Young 1978) that the most unstable periods along the escarpment are those when the rainfall exceeds 400 mm per month. Comprehensive work concerning the duration

Figure 7.4 Rockfall and slide debris across Lawrence Hargrave Drive just north of Clifton. The slide has removed a netting barrier emplaced to stop smaller falls of rock debris. See Figure 7.3 for location.

and intensity of rainfall which causes sliding is now being completed at the University of Wollongong (Flentje 1998). In particular, the period and magnitude of antecedent rainfall are being correlated to accelerated slope movements at depth at monitored sites.

CASE STUDIES Rockfalls Along the road between Clifton and Coalcliff small rockfalls occur after most significant rainfall events whereas larger rockfalls and associated debris flows occur less frequently, e.g. the events which blocked Lawrence Hargrave Drive in both 1987 and 1988. In 1987 the main rockfall was associated with undercutting and toppling of about 500 t of rock from the vertically jointed Scarborough Sandstone located about 6 - 8 m above the roadway. The joints from which the slabs were detached belong to one of the regional joint sets oriented at 015° (Memarian 1993). Relaxation of the horizontal stress field, as a result of escarpment erosion, led to joints in the sandstone opening near the cliff face and becoming partly filled with expandable clay from the overlying soil. Widening of the roadway by cutting back into the underlying Wombarra Shale removed the natural toe slope and exposed the shale to active weathering. This erosion and weathering led to undercutting of the Scarborough Sandstone which, together with expansion of clay and increased pore-water pressure in the joints following a heavy rainfall event, led to the major rockfall and toppling event (Hutton et al 1989). Further rockfalls from both the Scarborough and overlying Bulgo Sandstone units, together with debris flows from the intervening shale units, also followed the heavy rains in April 1988 and again blocked the road (Figure 7.4). Subsequent remedial measures in this area have included rockbolting in the jointed Scarborough Sandstone, extending of concrete buttresses along the undercut portions of the sandstone, drilling horizontal drainage holes into the sandstone and underlying shale to reduce water pressure, and placing shotcrete over the exposed Wombarra Shale. The latter is not particularly

B. G. Jones et al.

Figure 7.5 Landslip activity in colluvial material on Mt Ousley Road during 1977 which instigated the emplacement of remedial measures (drained buttress-fills) to stabilise the slope. See Figure 7.3 for location.

effective as the shale continues to weather and swell as a result of groundwater movement leading to a breakup of the shotcrete surface. Slumps In the Illawarra region slumps range in size from very small to large and they are generally active following long periods of wet conditions and high rainfall. Numerous slumps along Mt Ousley Road in 1977 (Figure 7.5) resulted in the emplacement of drained buttress-fills to stabilise the slopes. After the rains in April 1988 a number of major slumps between Clifton and Coalcliff caused the road to be covered with debris and, in one place, led to the occurrence of collapse. The road was closed for a period of six months while extensive repairs were carried out. The most spectacular slump at this time was the collapse of Lawrence Hargrave Drive about 500 m north of Clifton (Figure 7.6). This collapse was caused by a rotational slump involving both colluvial material and a thick covering of weathered Wombarra Shale. The slide plane lay just above the top of the Coal Cliff Sandstone at the edge of the cliff and resulted in the accumulation of a debris apron, including road materials, at the base of the cliff. Instability in this area had probably been increased by the undercutting collapse of the edge of the cliff about three years before the landslip occurred. The head of the slump consisted of a number of smaller scarp faces and extensive en echelon tension crack arrays were developed around the margin of the slump. Elsewhere along the road arrays of tension cracks, and even small faults and folds, gave indications of instability and local downslope movement in the road materials (Hutton etal 1989). A second major slump occurred at Moronga Park, Clifton (Figure 7.7). This slump was particularly interesting because it moved relatively slowly over a period of 18 months as a retrogressive landslide. After the initial heavy rains in April, the head scarp of the slump showed a slip plane with up to 1 m of vertical movement through a thick colluvial deposit (Hutton et al 1989). At this time the toe of the slump could not be clearly defined on the cliff face where it converted into an earth flow, but it lay in the upper part of the Illawarra Coal Measures. The slump is now about 100 m long, 50 m

Figure 7.6 The collapse of Lawrence Hargrave Drive in 1988 was caused by a landslip in the upper weathered portion of the Wombarra Shale. A large volume of natural debris and road material formed a colluvial apron at the base of the adjacent cliff. See Figure 7.3 for location.

wide and has an almost vertical head scarp about 12 m high (Figure 7.8). The initial toe of the slump has subsequently been pushed over the coastal cliff to form a steep debris apron (Figure 7.7). Subsequent movement on the outer part of this slump has also involved removal of the upper approximately 3 m thick weathered layer of the Illawarra Coal Measures, including a thin coal seam (probable aquifer) and associated very weathered underclay. Tension cracks up to 1 m deep and 15 cm wide are a characteristic feature of the outer part of this flow and have developed at right angles to the flow direction in response to the removal of the toe of the slump at the cliff edge. At the intersection of the slump and the coastal cliff, shear failure has occurred on a plane dipping 30° seawards which is not directly jointcontrolled in this shaly sequence. Headward movement of the scarp face was recorded during the progressive slump failure and further potential movement is indicated by tension cracks around the scarp face. The opening of such cracks provides access for additional water to infiltrate the slump mass. During headward scarp retreat the northern part of the slump retreated at a greater rate resulting in translational fault movement within the colluvial material between the two halves of the slump mass (Hutton et al 1989). Two additional small slumps developed to the south of the main slump while a more active slump to the north converted into an earthflow and mudflow which came down the steep valley eroded along the Clifton Fault. Mudflows Mudflows occur in very wet debris where rapid development of excess pore-water pressure overcomes the frictional resistance of the mass and very quickly converts into fluidised movement. Most mudflows in the Illawarra region have been caused by destabilisation of colluvial slopes, commonly by removal of the toe of the slope during construction work for roads, railways or building sites or by blocking the natural drainage leading to ponding of runoff. When such destabilised material becomes saturated during or following heavy rainfall events it can move as a series of surges even on low slopes. In such flows the fluid appears to act as a lubricant between the moving soil particles and

Land instability, iliawara region NSW

Figure 7.7 The Moronga Park slump increased in size for many months by progressive slump failure following the major rainfall period in April 1988. This resulted in a large head scarp and an equivalent debris apron at the base of the cliff. Smaller slumps are evident both north and south of the main slip. See Figure 7.3 for location.

the flows have a relatively high density. These flows can move large debris which is commonly carried in the upper part of the flow due to dispersive pressures (Johnson 1970). The velocity and mass of mudflow material can be very destructive to any vegetation or structure in its path and these effects were made dramatically obvious during the recent Thredbo landslide disaster in the Snowy Mountains. Mudflows are most common in areas where surficial deposits directly overlie weathered shale. In the heavy rainfall periods in 1987 and 1988 a number of mudflows occurred on the two main roads coming down the escarpment into the Illawarra area, one causing partial collapse of the Mt Ousley Road. The most well-known case is the Coledale disaster of 1988 (Mostyn & Adler 1991; Davies & Christie 1996) where a fatal complex debris slump-mudflow at Rawson Street, Coledale, occurred on the night of 30 April. Before this event, water became impounded behind a railway embankment which resulted in scouring of the crest and face of the embankment and its subsequent collapse in a debris slump. During the slump, the material partially liquefied resulting in a mudflow that destroyed a house killing two of the occupants. This illustrates the great care required in slope and embankment safety management when natural drainage is restricted or when slopes are likely to be destabilised by removal of colluvial material.

REMEDIAL WORKS Artificially induced slides, many of which include underlying or adjacent colluvium, almost invariably result from poor site selection, poor drainage, poor design and construction practices or a combination of these factors. Many slope-instability evaluations in the Illawarra area are interdisciplinary geotechnical endeavours requiring concepts from engineering geology, soil mechanics and rock mechanics. Reliable evaluations of slope instability must begin with an understanding of regional and site geology, and the geomorphological processes which formed the site and which continue to influence the slope evolution. Once this level of

Figure 7.8 Part of the head scarp of the Moronga Park slump showing the very poorly sorted nature of the colluvial veneer in this area. See Figure 7.3 for location.

geologic understanding is reached, slope behaviour can often be assessed on the basis of appropriate geotechnical models using suitable site investigation, strength testing and stability analyses. The key factor in many cases is the recognition of old landslide deposits. Identified colluvial masses can often be stabilised with drained buttress fills or other suitable retaining structures. Stabilisation of a partially unstable colluvial mass by excavation alone generally requires removal of virtually the entire mass. This is seldom practical for large and extensive natural sloping areas. Improvement of subsurface and surface drainage to maintain reduced pore-water pressures and hence increase the soil shear strength is an important component of stabilisation measures for many colluvium landslides. However, drainage by itself may not always be sufficient for stabilisation. With rock slopes, a realistic design will frequently involve a combination of measures including: (i) control of passive factors, such as weathering; (ii) the use of structural restraints, such as rock anchors; and (iii) drainage. These control measures should be applied as soon as possible after exposure of the particular face in order to minimise joint dilation and subsequent groundwater flow. Subsurface groundwater flow in both jointed rock sequences and colluvium can only be avoided by tapping the groundwater far back from the exposed face with drainage adits or boreholes. This will usually only be warranted in potentially large scale failures or hazardous areas of significant extent. Coastal erosion, which causes oversteepening of the cliffs, is a critical factor in the northern Illawarra and large breakwater walls would need to be constructed to combat this process. An alternative to remedial work along the existing roadway and railway between Stanwell Park and Coalcliff would be the excavation of a tunnel to carry traffic past the dangerous section or to construct a bridge across the shore platforms. However, such proposals may not be feasible on the basis of cost and other considerations.

CONCLUSIONS Land instability and landsliding are part of the natural erosional and geomorphological processes occurring along the steep escarpment in the Illawarra region. They are acceler-

B. G. Jones et al.

ated due to human activities which involve disturbance of equilibrium and environmental changes. Good slope management can prevent some failures, reduce the frequency or scale of movement in some areas and also reduce impacts of landsliding in many situations. Management policies must include methods for monitoring movements and assessment of landslide hazard, and the attendant risks to life, property and the environment in general. Maps of geology and land instability, or hazard maps, need to be used as the basis for landuse planning and construction design as well as for risk assessment and management. Database support additional to these maps is needed for specific and detailed information about each site of land instability. Thus detailed geomorphological and geotechnical investigations are an essential first step to suitable planning and design that would minimise the amount of damage caused by land movement.

ACKNOWLEDGEMENTS The authors acknowledge the support of the GEME Research Centre, Department of Civil, Mining and Environmental Engineering and the School of Geosciences, all at the University of Wollongong. MHG acknowledges post-graduate scholarship funding provided by the Islamic Republic of Iran. The authors appreciate comments made by Phil Flentje on a draft of this paper.

REFERENCES F. G. 1993. Engineering Publications, London.

BELL

Geology.

Blackwell

Scientific

BLONG R. J. & EYLES G. O. 1989. Landslides: extent and economic sig-

nificant in Australia, New Zealand and Papua New Guinea. In: Brabb E. E. & Harrod B. L. eds. Landslides: Extent and Economic Significance, pp. 343-353. Balkema, Rotterdam. BOWMAN H. N. 1972. Natural slope stability in the City of Greater Wollongong. Geological Survey of New South Wales Report 14, 159-222.

CHESTNUT W.

1981. Wollongong-Port

Hacking

1:100,000

Sheet.

Engineering geological consideration and environmental geological hazards. Geological Survey New South Wales Report GS81/202 (unpubl.). CHOWDHURY R. N. 1976. Mechanism of natural slope failures in the Greater Wollongong area of New South Wales. Search 7, 3 9 6 - 3 9 7 . CHOWDHURY R. N. 1980. Landslides as natural hazards—mechanisms and uncertainties. Geotechnical Engineering 11, 135-180. CHOWDHURY R. N. 1992. Simulation of risk of progressive failure. Canadian Geotechnical Journal 29, 94-102. CHOWDHURY R. N. & FLENTJE P. N. 1996. Geological and land insta-

bility mapping using a GIS package as a building block for the development of a risk assessment procedure. In: Proceedings of the 7th International Symposium on Landslides, pp. 1 7 7 - 1 8 2 . ISL1996, Trondheim, Norway. CHOWDHURY R . N . & YOUNG A . R . M . 1 9 8 7 . Field guide to slope instability in the Wollongong-Picton-Nattai Region of New South Wales. 5th International Conference and Field Workshop on Landslides., University of Wollongong, Wollongong. CHOWDHURY R. N. & ZHANG S. 1993. Modelling the risk of progressive

slope failure—a new approach. Reliability Engineering and System Safety 40, 17-30. DAVIES W . N. & CHRISTIE H. D. 1 9 9 6 . T h e Coledale mudslide, New

South Wales, Australia—a lesson for geotechnical engineers. In: Proceedings of the 7th International Symposium on Landslides, pp. 701-706. ISL-1996, Trondheim, Norway. DUNKERLEY D. L. 1976. A study of long-term slope stability in the Sydney Basin, Australia. Engineering Geology 10, 1-12. FELL R. 1985. Slope stability in the Wianamatta Group. In: Pells P. J. N. ed. Engineering Geology of the Sydney Region, pp. 163-175. Balkema, Rotterdam. FLENTJE P. N. 1998. Computer based landslide hazard and risk assessment (northern Illawarra Region of New South Wales, Australia). PhD thesis, University of Wollongong, Wollongong (unpubl.). FLENTJE P. N. & CHOWDHURY R. N. 1 9 9 6 . Preparation a n d validation

of digital maps of geology and slope instability. In: Chacon J., Irigaray C. & Fernandez T. eds. Proceedings of the 8th International Conference and Field Workshop on Landslides, pp. 343-351. Balkema, Rotterdam. FRANKLIN J . A . & DUSSEAULT M. B . 1 9 8 9 . Rock Engineering. McGrawHill, New York. GHOBADI M. H. 1994. Geology and slope stability in the northern Illawarra, New South Wales, Australia. In: 7th Congress of the International Association of Engineering Geology, pp. 1307-1314. Balkema, Rotterdam. GHOBADI M. H. 1995. Geology and slope stability in the northern Illawarra, New South Wales, Australia. PhD thesis, University of Wollongong, Wollongong (unpubl.). HILLEARD P. R. 1993. Bedrock movement and the failure of Stanwell Park Railway viaduct. In: McNally G. H., Knight M. & Smith R. eds. Collected Case Studies in Engineering Geology, Hydrology and Environmental Geology, pp. 1-27. Butterfly Books, Sydney. HUTTON A . C . , FERGUSON C . L. & JONES B . G . 1 9 8 9 . L a n d s l i p i n t h e

northern Illawarra Coalfield. In: Proceedings of the 23rd Newcastle Symposium, Advances in the Study of the Sydney Basin, pp. 37-44. University of Newcastle, Newcastle. JOHNSON A. M. 1970. Physical Processes in Geology. Freeman, San Francisco. KEEFER D. K . 1984. Landslides caused by earthquakes. Geological Society of America Bulletin 95, 406-421. MEMARIAN H. 1993. Fracture history of Coal Cliff Sandstone in Coalcliff area, New South Wales. In: Proceedings of the 27th Newcastle Symposium, Advances in the Study of the Sydney Basin, pp. 113-121. University of Newcastle, Newcastle. MOSTYN G. R. & ADLER M. A. 1991. Design of the remedial works for

the Coledale Landslide. In: Proceedings of the 6th International Symposium on Landslides, pp. 7 9 1 - 7 9 6 . I S L - 1 9 9 2 , Balkema, Rotterdam. PITSIS S. E. 1992. Slope Instability along the Illawarra Escarpment. MEngSc, University of New South Wales, Sydney, (unpubl.). RITTER D. F., KOCHEL R . C. & MILLER J. R . 1995. Process Geomorphology. W. C. Brown Publishers, Boston. SHELLSHEAR W. 1890. On the treatment of slips on the Illawarra railway at Stanwell Park. Journal and Proceedings of the Royal Society of New South Wales 24, 58-62. TERZAGHI K. 1950. Mechanism of landslides. Geological Society of America, Berkley Volume, 83-123. VARNES D. J. 1978. Slope movement types and processes. In: Schuster R. L. & Krizek R. J. eds. Landslides, Analysis and Control, pp. 11-35. National Academy of Sciences, Washington, Special Report 176. YOUNG A. R. M. 1978. Influence of debris mantle and local climatic variations on slope stability near Wollongong, Australia. Catena 5, 95-107. ZARUBA Q. & MENCEL V. 1982. Landslides and Their Control (2nd edition). Elsevier, Amsterdam. Received 2 September 1997; accepted 1 May 2000

THEME 3 MINING AND RADIOACTIVITY

Geological Society of Australia Special Publication 21, 77-94

CHAPTER 8—Acid drainage at minesites CSIRO Environmental Projects Office, PMB 2, Glen Osmond SA 5064, Australia. Acid drainage (formerly acid mine drainage (AMD) or acid rock drainage (ARD)) from sulfidic minewastes and mines and the subsequent off-site contamination of water and soil is considered the major environmental issue of the mining industry. Although present management techniques have significantly reduced its generation, acid drainage from abandoned or old mines has in many instances had catastrophic impacts on the environment. Remediation and rehabilitation of such sites is extremely difficult and very expensive. The extent of the problem has led to the formation of national organisations, funded by federal governments and the mining industry, to develop new and improved techniques for prediction, management (prevention), treatment and control, monitoring and restoration. The most commonly used management technique is to cover the sulfidic waste to prevent oxygen and/or water infiltration. The long-term sustainability of these techniques are not yet proven under Australian climatic conditions and in some cases are believed only to delay the onset of acid drainage. The most effective management technique is the appropriate use of geoscientific data generated during exploration and pre-feasibility drilling. This should include geology, mineralogy, geochemistry, structure, geotechnical properties, hydrology etc. in planning the mining and milling of ore and disposal of minewastes. Notice should also be taken of the natural oxidation processes of gossan formation and secondary dispersion which are analogous to acid drainage and contaminant transport. This paper discusses the processes of acid mine drainage generation, natural attenuation, environmental implications, treatment, control, prediction and characterisation, and monitoring and analysis. Finally, future directions for management of sulfidic minewastes are discussed. KEY WORDS: acid drainage, Captains Flat, contamination, environmental management, Mt Lyell, Mt Morgan, oxidation, Pine Creek, rehabilitation, Rum Jungle, sulfides, Woodlawn.

ACID GENERATION Acid drainage results from the oxidation of sulfur in sulfides and sulfosalt minerals in the presence of oxygen to form H + ions (and low pH) and the transport of the oxidation products including sulfate in solution. There are over 2000 sulfur-bearing minerals (including sulfates) the more common of which are listed in Table 8.1. Although the oxidation of sulfide minerals is most commonly associated with mines, minewastes and mineral processing, it occurs in situ with the formation of gossans (Taylor & Thornber 1992), contamination of streams (e.g. Rio Tinto in Spain) and secondary dispersion of pathfinder and other trace elements (Thornber 1992). These natural processes are exploited in exploration geochemistry and have led to the discovery of mineral deposits such as Broken Hill (New South Wales), Mt Isa (Queensland), Mt Lyell (Tasmania) and Kambalda and Telfer (Western Australia), and a host of smaller deposits that were exploited during the 19th and 20th centuries. The processes of gossan formation often led to supergene enrichment of the mineralisation particularly copper, gold, silver and to a lesser extent nickel and lead (Taylor et al 1984). Differences between gossan formation and secondary dispersion and generation of acid drainage are due to the difficulty of oxygen and water infiltrating a mineral deposit, whereas in a mine or minewastes, oxygen and water access is much easier, specific surface areas much greater, and reaction rates faster due to fracturing, crushing and grinding.

Of the sulfide minerals, pyrite is the most common source of acid drainage because it is by far the most common of the sulfides and is a waste product of mining and mineral processing. Other iron sulfides such as pyrrhotite and marcasite contribute significantly to acid drainage. Pyrite oxidation is complex as it involves chemical, biological and electrochemical reactions. The chemical reactions are summarised by: FeS 2 + 7 / 2 0 2 + H 2 0 Fe 2 + + V 4 0 2 + H + Fe 3 + + 3 H 2 0

Fe 2 + + 2 S 0 4 2 " + 2H+ Fe 3 + + V 2 H 2 0

Fe(OH)3 + 3H +

FeS 2 + 14Fe 3+ + 8 H 2 0

15Fe 2+ + 2 S 0 4 2 " + 16H+

with oxygen and Fe 3 + being the oxidants, whilst water is a necessary reactant and a transport mechanism. When the pH is low (<4.5), pyrite is more rapidly oxidised by Fe 3 + , and provided there is a continuous source of Fe 3 + in solution the oxidation of pyrite occurs in the absence of oxygen. The processes are catalysed by the presence of bacteria such as Thiobacillus ferroxidans and Thiobacillus thiooxidans both of which are found in acid drainage. Other bacteria and algae have been isolated and characterised. Biotic oxidation of sulfides may increase the rate of reaction by 10 5 or 10 6 . Oxidation of pyrite by Thiobacilli is described by Suzuki et al (1994) and is summarised by Parker (1999).

G . F. Taylor

Table 8.1 Common sulfur minerals. Mineral name Sulfides Argentite Arsenopyrite Bornite Bravoite Chalcocite Chalcopyrite Covellite Galena Greigite Mackinawite Marcasite Molybdenite Pentlandite Pyrite Pyrrhotite Sphalerite Troilite Violarite Sulfosalts Enargite Proustite Pyrargyrite Tennantite Tetrahedrite Sulfates Alum Alunite Anglesite Anhydrite Barite Brochantite Copiapite Epsomite Gypsum Hexahydrite larosite Melanterite

Composition

Sulfur content (%) 12.9 19.7 24.5 -36 20.2 35 33.6 13.4 43.4 34.1 53.4 40.1 -36 53.4 -36 33 36.5 42.5

Ag2S FeAsS Cu 5 FeS 4 (Ni,Fe)S2 Cu2S CuFeS2 CuS PbS Fe 3 S 4 FeS FeS 2 MOS2

(Fe,Ni) 9 S 8 FeS 2 FEI-XS

ZnS FeS Ni 2 FeS 4

Ag3AsS3 Ag 3 SbS 3 (Cu,Fe,Zn,Ag)12As4S13 (Cu,Fe,Zn,Ag) 12 Sb 4 S 13

32.6 19.4 17.8 -28 -25

(K,Na,NH 4 )Al(S0 4 ) 2 .12H 2 0 KA1 3 (S0 4 ) 2 (0H) 6 PbS0 4 CaS0 4 BaS0 4 Cu 4 (S0 4 )(0H) 6 CUS0 4 .5H 2 0 MgS0 4 .7H 2 0 CaS0 4 .2H 2 0 MgS0 4 .6H 2 0 KFe 3 (S0 4 ) 2 (0H) 6 FeS0 4 .7H 2 0

-39.9 15.4 10.6 23.5 13.7 7.1 12.8 13 18.6 14 12.8 11.5

CU 3 ASS 4

The electrochemical oxidation of pyrite is less well understood, with some of the processes being described by Sato and Mooney (1960) who were the first to describe them as redox (oxidation-reduction) processes. A more detailed description is provided by Thornber and Taylor (1992). The reduction reaction that occurs at the cathode consumes electrons and oxidation reaction at the anode provides electrons as follows: cathodic reaction:

0 2 + 2 H 2 0 + 4e~

anodic reaction:

MS + 4 H 2 0

40H"

M 2 + + S 0 4 2 ~ + 8H + - 8e"

At the cathode the solution will have a high pH (alkaline) while at the anode the solution will have a low pH (acid). There are many factors which control the rate and extent of weathering. Mention has already been made of the competency of the ore and country rock (controlling the movement of oxygen and water) and the specific surface area of

the mineral grains. Other factors include the conductivity of the electrolyte solution paths between anode and cathode, coating of surfaces with secondary oxidation products, resistivity of the ore assemblage, depth of water-table (dissolved oxygen content of water is much lower than in air) and nature of sulfides. Even for pyrite, there is considerable debate concerning its oxidation rate and even more so for the other sulfide minerals (Bennett et al 2000). This is presently the subject of considerable research in Australia. The composition of the primary sulfide minerals is an important factor in determining the pH of any water in contact. For a metal sulfide M r S where r is the metal/sulfur ratio: 2M r S + ( 3 + 2 r ) H 2 0 + ( 3 + r ) 0 2 -> 2 r M 2 + + 2 S 0 4 2 " + 4 H + + 4rOH~ where r < l , such as pyrite (FeS 2 ), marcasite (FeS 2 ), monoclinic pyrrhotite ( F e 7 S 8 ) and violarite (FeN i ) 3 S 4 acid is produced; for monosulfides where r = 1, such as galena (PbS), sphalerite (ZnS) or chalcophrite (CuFeS 2 ), the overall reaction is neutral and the overall pH will not change if there is no hydrolysis of the metals; for sulfides with r > 1, such as chalcocite (Cu 2 S) and pentlandite [(NiFe) 9 S g ], OH" is in excess. Pyrrhotite is found in many metamorphic sulfide deposits and has a variable chemical composition F e ^ S where x can range from 0.125 ( F e 7 S 8 ) to 0.00 (FeS). When oxygen is the primary oxidant the overall reaction is: F e ^ S + (2-x/2)02 + XH20

(l-x)Fe2+ + S 0 4 2 " + 2xH+

with the F e 2 + being oxidised to form more acidity. Oxidation of pyrrhotite by F e 3 + at acid pH occurs through the following reactions: F e ^ S + (8-2x)Fe 3 + + 4 H 2 0

(9-3x)Fe 2 + + S 0 4 2 " + 8 H +

or F e ^ S + (2-2x)Fe 3 + -» (3~3x)Fe 2 + + S Field observations indicate that pyrrhotite oxidation will form elemental sulfur, together with pyrite and marcasite as intermediate products: 2 F e 1 - x S + (V 2 - x ) 0 2 + (2-4x)H + -> FeS 2 + ( l - 2 x ) F e 2 + + (1-2X)H 2 0 The intermediate pyrite or marcasite subsequently oxidises. Pyrrhotite is much more reactive than pyrite and oxidation may result in spontaneous combustion either underground or in surface minewastes. Abiotic oxidation reactions of common ore sulfides are given below: galena: PbS + 8 F e 3 + + 4 H 2 0

8H+ + S 0 4 2 " + Pb2+ + 8Fe2+

and anglesite P b S 0 4 may precipitate

Acid drainage at mine sites sphalerite: ZnS + 8Fe 3 + + 4 H 2 0

8H + + S 0 4 2 ~ + Zn 2 + + 8Fe 2 +

arsenopyrite: FeAsS + 13Fe 3+ + 8 H 2 0 14Fe 2+ + S 0 4 2 " + 13H+ + H 3 AS0 4 with the dissolved Fe 2 + reacting with the arsenate to form scorodite (FeAs0 4 .2H 2 0) chalcopyrite: CuFeS 2 + 16Fe 3 + + 8 H 2 0 Cu 2 + + 17Fe 3 + + 2 S 0 4 2 " + 16H + These reactions release metal cations into solution and consequent off-site contamination of water and soil They are also the basic reactions which lead to secondary dispersion haloes utilised in exploration geochemistry.

SOURCES OF ACID DRAINAGE AT MINESITES Any sulfide-bearing material exposed to the atmosphere is a potential source of acid drainage. Whereas in situ sulfide orebodies oxidise very slowly because infiltration of oxygen and water is very slow, the sulfides exposed by mining, milling and processing are more rapidly oxidised because of greater surface area and permeability. The acidity and contaminant loads resulting from mining activities are therefore much greater than those associated with passive in situ oxidation and weathering. Waste rock, low-grade ore, overburden and ROM (run of mill) stockpiles are potential sources of acid drainage, and because of the sheer tonnages involved generate more acid drainage than other sources. Because of mining techniques, the dumps are usually heterogeneous in rock type and size distribution. During deposition, both oxygen and water are incorporated and because of the up to 30% void space in unconsolidated rock piles, subsequent ingress of oxygen and water is extremely difficult to stop. Such infiltration is vertical from the atmosphere, by advection where the dumps are high and exposed, and laterally where the dumps are in valleys. Other characteristics of rock dumps which may have implications for acid drainage generation are the relatively low surface area of exposed minerals in comparison with those in tailings dams, the erodibility of weathered surface material to expose fresh sulfide, and the void space which may decrease the stability of surface covers through compaction. The heterogeneity complicates monitoring and modelling of rock dumps and makes representative sampling for prediction and characterisation difficult. In strip-mining activities such as in the coal mines of the Bowen Basin, Queensland, the near-surface regolith which supported flora and fauna is covered by highly dispersive, sodic and pyritic materials from lower in the regolith profile during the stripping operations. Not only does this increase the acid drainage potential but also makes ecosystem reconstruction difficult. Relocation and reshaping of rock dumps are both difficult and expensive (Roe et al 1996). Although the volume of tailings and other mine spoils is significantly less than that of waste rock, they nevertheless represent a large potential source of acid drainage and contamination. Previous disposal techniques such as dis-

charge into rivers and streams (Mt Lyell) or leaving exposed and not contained (Rum Jungle, Mt Morgan, Captains Flat in Australia; Summitville, Leadville, Central City in Colorado, USA) have led to considerable environmental problems and the expenditure of vast sums in rehabilitation (Pendleton et al 1995). It is now common practice to constrain tailings within natural or constructed impoundments and to cover them (see below). Problems associated with tailings are common to those generated by other mine wastes and presently utilised management techniques are similar. The major difference is the small grainsize distribution, fine grainsize and high surface area and relative homogeneity of the tailings. Whereas waste rock and low-grade ore are stockpiled at their intrinsic pH, tailings are often deposited at neutral to alkaline pH. This has led some companies to ignore the possibility of later acid drainage generation (despite predictive tests indicating otherwise). A source of acid drainage often ignored is abandoned mines. At Mt Lyell the major source of acidity and Cu load (2469 kg/day) in Haulage Creek is the fragmented rocks of the Prince Lyell workings (Miedecke et al 1997). Oxidation of the pyrite in the abandoned Iron Mountain Mine, California and movement of water through it results in a pH of -0.45 to -3.4 and high levels of Cu, Zn, Al, As, Sb and Cd (Alpers et al 1994) whilst the Reynolds Adit drains South Mountain at Summitville, Colorado with discharges up to 7000 L/min containing 300 mg/L Cu, 47 mg/L Zn at a pH <3 (Pendleton et al 1995; Plumlee et al 1995). There is little available information on the impact of underground mines on the pH, acidity and contaminant load of groundwaters resulting from movement of mine waters into aquifers. Flooding of the abandoned Eagle Pb-Zn mine in Colorado has increased the pH from 3 to 6 and decreased dissolved ions by an order of magnitude (Neukirchner & Hinrichs 1997). However stagnant water in openpit mines (pit lakes) can have a low pH of 2-3 and high contaminant load (Klapper & Schultze 1997). Again, there is little information available on resultant groundwater plumes.

FIELD DETECTION OF ACID DRAINAGE Any seepage or water flowing from a mine, minewaste, mill or ponded water in an opencut (void) may be acid. If this is the case there is a high risk that it also contains elevated levels of sulfate ions and contaminant metal ions. Detection of acid drainage is of vital importance and can be achieved in several ways: (i) the water has a distinctive red to brown colouration (Figure 8.1 on Plate 1); (ii) a brown-orange-yellow precipitate coats the material which is in contact with the water (Figure 8.2 on Plate 1); (iii) the presence of filamentous green, brown or black algae (Figure 8.3 on Plate 1); (iv) using a pH paper or pH meter to determine the actual pH of the water—many natural waters may have a pH~5.5 so that a pH below 4 is considered to be of potential risk; and (v) where the seepage has evaporated, a coating of white or greenish efflorescent secondary minerals may form (Figure 8.4 on Plate 1), which are commonly sulfates of Fe, Al, Mg, Na or K. Determination of pH is one of the most definitive techniques but care must be taken in the interpretation of nearneutral or alkaline pH as the acid drainage may have been

G. F. Taylor

Table 8.2 Geochemical associations of elements in various types of sulfide deposits.

Ag As Au Ba Bi Co Cr Cu Hg Ir Mn Mo Ni Pb Pd Pt Re Sb Se Sn Te TI U Zn

Volcanogenic Cu-Zn-Pb

Sedimentary black-shalehosted Cu

Porphyry Cu

Au-bearing Fe sulfides

X X X X X

X X

X X X

Ultramafic Ni-Cu -

Sedimentary exhalative Pb-Zn-Ag

Carbonatehosted Zn-Pb

X X

X X X -

X X

X X X

X X -

X X X X X

X X

X X X

neutralised by natural processes. In this case, there is usually a gelatinous brown-orange precipitate of iron oxyhydroxides present. The more insidious evidence of acid drainage is the obvious death of flora and fauna and destruction of both aquatic and terrestrial ecosystems (Figure 8.5 on Plate 1). The extent of the impact may differ from a few metres to many kilometres such as in the Finniss River downstream from the Rum Jungle Mine (Bennett & Lawton 1995) and in the Queen and King Rivers and Macquarie Harbour below the Mt Lyell workings in Tasmania (Koehnken 1997b). In some instances the impacts are not obvious and selected organisms are the only evidence of ecosystem destruction. Such impacts may be caused by low pH, high sulfate content or presence of contaminant elements.

ENVIRONMENTAL IMPLICATIONS OF ACID GENERATION Previous equations have shown that the metals in the sulfide minerals are liberated as cations during the oxidation process. However, these equations do not indicate the wide range of metals that can be solubilised. For example, there are wellknown metal associations with each of the mineral deposit types and ore minerals (Table 8.2), all of which can be released during oxidation and mineral breakdown. Although these associations are readily recognised, the concentration of each element varies between mineral deposits and within deposits (i.e. zoning is common in many mineral deposits). It is not only the chalcophile elements that are released

during oxidation. The gangue minerals and country rocks are more easily weathered in acid environments releasing lithophile elements such as aluminium, manganese, and for some feldspars and micas, barium, lead and thallium (which have substituted for potassium). The question is, why is the release of these elements of greater significance for acid drainage than for normal weathering processes? The latter usually occur at circum-neutral pH values causing many of these cations to precipitate as secondary minerals or be absorbed on surfaces of layer silicates, iron and manganese oxides. In the lower pH environments of acid drainage many of these cations remain in solution (Figure 8.6 on Plate 1) and are transported off-site. Because of their greater bioavailability in water-soluble form they are more ecotoxic and may well enter the food chain. Acid generation and contaminant release are not necessarily related. Indeed, some of the potential contaminants (e.g. Cu, Zn, Cd, Pb) may be released into solution at alkaline pH values (Figure 8.6 on Plate 1). Conversely at very low pH, secondary minerals such as alunite and jarosite may be precipitated, incorporating many of the contaminant elements on their surfaces or in the mineral structures (Scott 1987). Many of the secondary minerals formed by oxidation of sulfidic minewastes are water soluble sulfates and further contribute to off-site contamination. However, in Australia and other countries where evaporation far exceeds precipitation, these secondary minerals are precipitated at or near the surface as 'efflorescences'. Rather than the near-surface environment being leached of contaminant elements, they are concentrated, posing problems to the establishment of sustainable ecosystems.

Acid drainage at mine sites Table 8.3 North Adit Spring, Captains Flat, chemistry summary (post-rehabilitation data). Study

Statistic

Zn mg L'1

Zn load kg d'1

Flow L s' 1

Brooks (1980) Jan 1977 to Oct 1978 pH & conc (n=7); load & flow (n=6)

Max Min Median

3.4 2.7 3.2

315 170 230

55.5 3.2 34.3

3.1 0.2 1.9

Brown & Train (zn Brooks 1980) Dec 1976 to Aug 1978 (n=60)

Max Min Median

3.7 2.4 2.7

450 195 240.5

151 3 19

5.4 0.1 1.0

Max Min Median

3.2 2.8 3.0

180 153 160

February 1992 August 1992

3.4 3.1

120 105

Jun to Nov 1982 (n=4)

NSW DMRa a

New South Wales Department of Mineral Resources (unpublished), by kind permission of Kerry Brooks.

In addition to the release of lithophile elements, the accelerated weathering of gangue minerals may also lead to the formation of new secondary minerals. These secondary minerals can alter geotechnical, hydrological and erodibility properties of the minewastes or in extreme cases the strength of dam walls. An example is the overburden from some of the opencut coal mines in Queensland. Weathering of these dump materials may give rise to highly dispersive, sodic and saline soils which are difficult to revegetate, increase the salinity levels in void and groundwaters, and result in increased erosion. Unexpected generation of acid drainage from a minesite can result in several adverse effects (Lee 1999): (i) exceedance of discharge water quality standards (ANZECC/NHMRC 1992) and possible adverse effects on downstream ecosystems potentially leading to fines and litigation; (ii) non-budgeted cost of establishing a water-treatment plant to neutralise acidic water to recycling in the mill; (iii) cost of treating water prior to off-site discharge, a process which may have to be done for many years; (iv) additional rehabilitation costs which will be incurred to slow down the rate of acid generation; (v) possibility of never legally being released from a performance bond; (vi) discouraging potential investors from joint venturing or buying the operation; (vii) discouraging financial operators from lending money; and (viii) restrictions placed on the company to further access other potential mining areas.

CASE HISTORIES IN AUSTRALIA Captains Flat, New South Wales The Captains Flat Cu, Au, Pb, Ag, Zn and pyrite deposit, about 50 km southeast of Canberra was mined from 1882 to 1962. Processing waste water and mine drainage water polluted the Molonglo River upstream of Canberra (Brooks 1998). In the 1940s, 2 x 10 6 L d _ 1 of mill waste water and mine water containing approximately 300 mg L _ 1 Zn was continuously discharged directly into the river. This represents 600 kg d" 1 Zn load in addition to unknown quantities of other metals, acid, cyanides, sulfides, thiocyanates,

cresols and phenols. Over the complete period of mining, contaminated material was eroded from the tailings dumps and contaminated leachate discharged into the river. Significant problems arose on several occasions when the tailings dams collapsed and erosion and leaching of the minesite added to contamination of the river bed, floodplain sediments and pastures, resulting in landholders sueing the mining company for destruction of pastures. Because of this, and its proximity to the Australian capital, Canberra, substantial rehabilitation of the site was undertaken in 1976 (Figure 8.7 on Plate 1). This work effectively stabilised the tailings dams, eliminating risk of failure, and greatly reduced erosion and failure. However, despite these measures, some pollution of the Molonglo River continues to occur because of acid drainage generation (Figure 8.8 on Plate 1) and dispersion of soluble elements (Table 8.3). This case study serves to show that a combination of poor mine planning and local rehabilitation measures make it nearly impossible to prevent water and oxygen entering the mine and waste dumps. Dames and Moore (1993) have indicated that unless further remedial action is undertaken, the present situation will remain indefinitely. Additional remediation would be very costly and could not hope to prevent all future contamination.

Woodlawn, New South Wales The Woodlawn deposit occurs some 70 km northeast of Canberra in a similar geological environment to Captains Flat. Mining commenced in 1978 to produce Cu, Pb and Zn concentrates (with Au, Ag having been leached from the stockpiled gossan early in the operation) and ceased in 1998. The mine occurs in an environmentally sensitive area adjacent to Lake George and is visible from the main highway between Sydney and Canberra. Consequently, monitoring and remediation are important. Although not suffering from the catastrophic failure experienced at Captains Flat, several periods of high-intensity rainfall have caused problems of excess water in the tailings and evaporation dams and waste-rock dumps (Tarlington 1995). Carefully designed water-management

G. F. Taylor

Table 8.4 Typical Woodlawn waste-rock dump leachate analyses. Date 03 Jan 1992 29 Apr 1993 27 Feb 1995

Conductivity (mS m' 1 )

Fe (mg L' 1 )

S04

2.9 3.1 3.0

2670 1860 2060

230 125 88

7650 6500 4980

350 350 310

45 30 24

39 000 32 800 27 900

Table 8.5 Woodlawn south dam seepage and piezometer analyses. pH 3.2 6.00 4.91 4.56 4.56

Eh (mV)

South Dam SS-1 SS-2 SS-3 SS-3 SS-3R SS-5A SS-5B SS-5C SS-7

4.87 4.89 4.88 4.91

SE-1 SE-2-Rut SE-2-Pond SE-3 SE-4 SE-5

4.18 119 6.17 6 8.57 -130 7.33 - 6 0 11 6.09 6.53 -13

XI X2 Y1 Y2 Z1 Z2

7.62 7.47 6.94 7.18 6.97 7.27

16 80

Cond Cu Pb Zn Cd A1 Mn Ca Mg S04 (mS m"1) (pgmL'1) (pg mL"1) (pgmL"1) (pg mL"1) (pg mL'1) (pgmL"1) (pg ml/ 1 ) (pgmL"1) (pgmL"1) 40 0.87 5.39 110 120 100 15 15 17 20

2.1 0.74 2.7 3.79 3.8 3.8 6.15 5.97 6.6 3.4

1100 600 780 2660 2270 2100 1025 1075 1075 925

5.8 1.55 3.16 6.92 6.40 6.00 2.40 2.44 2.64 2.58

140 6.25 120 1300 1200 990 150 135 140 120

68 66 65 260 240 250 120 130 130 110

510 615 620 555 600 580 585 535 550 535

640 720 755 2510 2600 2450 1060 100 1075 910

7300 5420 6180 21090 20000 18000 8490 8070 8640 7480

4750 1707 4290 4140 4090 4060

2.73 <0.01 <0.01 0.04 1.06 1.96

2.95 0.06 0.03 0.08 1.18 0.06

700 1.94 0.75 195 390 855

3.02 0.002 <0.001 0.88 1.21 2.99

11 0.24 0.30 0.24 0.73 1.94

55 0.61 0.30 35 60 85

580 25 35 350 410 525

615 45 260 550 725 800

5750 135 280 3410 4870 5750

3230

ND ND ND 0.02 0.01 0.01

ND ND 0.05 0.03 0.02 0.03

0.33 0.15 1.02 1.09 0.77 0.51

0.004 0.003 0.002 0.007 0.004 0.005

0.34 0.68 ND 0.06 0.06 0.07

0.61 0.00 0.02 0.02 0.72 0.03

3.18 0.80 17 30 25 76

159 51 344 319 221 145

307 68 318 200 691 479

7600 4730 4380

8400

-69 -53 -23 -36 -24 -40

2060 4540 4770

2070 1501

South Dam is a sample of surface south dam water; SS is seepage from beneath the south wall, south dam; SE is seepage from beneath the east wall, south dam; XI to Z2 are piezometers. Table 8.6 Nature and composition of leachates from Mt Lyell abandoned Au mine.

Waste rock Mine water Abandoned adit

Cu mg L"1

S04

Flow m 3 d"1

2.5 2.9 2.8

1160 520 138

140 40 174

130 170 28

13 7 45

8200 3800 1100

4700 4300 400

techniques and constant monitoring ensure that off-site discharges meet acceptable criteria. The now disused openpit produced 60-70 Mt of waste rock which covers an area of 100 ha and is constructed in five levels to a total height of 80 m. This has been rehabilitated by compaction of the surface, covering with weathered rock and soil and revegetation with grasses, shrubs and trees (Figure 8.9 on Plate 2). Despite a high sulfide content of 6-7%, the vegetation is generally healthy with the roots seldom penetrating the compacted layer. However, there have been some small sections where acid leachate has resulted in the death of

some vegetation. Leachate flow has reduced from 3 L s _ 1 to 0.4 L s _ 1 over a four year period but still has a relatively high contaminant load (Table 8.4). The acid drainage generating potential is high because of the high pyrite content of the tailings. Seepage associated with two sheared contacts between massive sandstone and a banded siltstone below the southern wall of the southern tailings dam is highly contaminated (Table 8.5) (Niven 1990). The tailings in two dams were reprocessed, but this will not reduce the heavy-metal content (except for Zn) or the acid drainage potential.

Acid drainage at mine sites

Table 8 . 7 Water quality in sumps and seeps on the minesite and in tributaries of the Dee River.

Ml Nelsons

Al mg L" 1 Ca Cd Co Cu Fe Mn Na Ni Zn S 0 4 mg L"1 PH EC mS m" 1

610 478 0.14 1.80 20 3.56 136 236 0.47 41 13500 3.66 12460

Sumps (M1-M8) M2 M3 M4 M5 M6 No 2 Mill Opencut Frog Hollow Creek Pipe Sump 3020 509 0.34 5.8 121 554 389 306 1.60 104 43600 2.24 11200

330 550 0.09 1.23 20 104 46 792 0.49 11.4 8240 2.84 10420

803 437 0.36 2.50 106 1460 131 213 0.65 45 16100 3.29 13600

461 466 0.48 2.74 39 1920 153 325 0.71 59 15000 3.72 13400

Mt Lyell, Tasmania The Mt Lyell mineral field on the west coast of Tasmania was initially worked for Au following its discovery in 1883. Copper mineralisation was subsequently found in 1893 and the company developed a process for pyrite smelting in 1896 (McQuade et al 1995). These smelting operations combined with timber cutting, frequent bushfires and high annual rainfall resulted in extensive loss of vegetation and considerable erosion. In addition, waste-rock dumps (50 Mt containing 10% pyrite) and abandoned mine workings are significant sources of acid drainage. Tailings have been consistently discharged into the local river system and in excess of 90 Mt have been discharged into the Queen/King River systems and Macquarie Harbour to date (Wood 1991). Some natural revegetation of the estimated 1500 ha of completely denuded and 2 5 0 0 ha of substantially denuded countryside has occurred, and research into revegetation of the remaining area is continuing. Although welcomed by regulatory authorities, there have been some objections to the revegetation program by members of the local community who want the bare hills to remain as a 'tourist attraction'. Acid drainage and dissolved metals derived from wasterock dumps, mine dewatering, underground drainage and pyritic outcrops create the greatest environmental problems, particularly for aquatic ecosystems. The nature of leachates from three sources is shown in Table 8.6. A constant problem is rain falling into the old West Lyell opencut which then percolates through 3 5 0 m of fractured and caving pyritic rock before it is pumped out of the mine. Discharge from the abandoned adit shows that high levels of contaminants are still present after 5 0 years when these underground workings were abandoned. Acid drainage from tailings is reduced because of the practice of subaqueous disposal (a method recommended in Canada—see later). However, some tailings have been deposited on the banks of the King River and over a considerable period have been intermittent sources of acid drainage over a distance of 15 km down river (Wood 1991). The contamina-

628 475 0.49 2.72 102 1730 152 260 0.69 60 15800 3.33 13800

M8 T2 Shepherds Dairy Ck

640 450 0.05 1.37 26 51 90 277 0.28 15.4 11700 2.94 -

0.17 482 <0.01 0.56 0.11 0.13 33 731 0.13 1.42 7402 7.03 10350

Tributaries (T2-T7) T3 T4 T6 Dairy Ck Mundic Ck Shepherds

0.093 428 <0.01 0.54 0.46 46 48 388 0.07 2.30 9375 6.20 10580

1090 493 0.74 3.04 196 1420 124 172 0.72 55 17940 2.76 14900

340 446 0.05 1.82 16 2.25 152 266 0.28 11.3 10770 3.62 -

T7 Fletchers Ck <0.005 36 <0.01 <0.01 <0.005 0.043 0.10 50 <0.01 <0.01 12 7.15 590

tion at Mt Lyell has recently attracted state and federal funding and was the subject of a number of research projects (Koehnken 1997a). It is estimated (Koehnken 1997b) that acid drainage will continue for another 6 0 0 years with the present copper load being 2 0 0 0 kg/day.

Mt Morgan, Queensland Gold, Cu and Ag were mined at Mt Morgan, 35 km southsouthwest of Rockhampton, between 1882 and 1991. As a result of this long period of mining there is now a very large opencut filled with low-quality water (Figure 8.10 on Plate 3), numerous waste rock dumps, tailings dams and smelter material (Figure 8.11 on Plate 3). There have also been discharges of tailings, waste rock, mineral-processing solutions as well as acid drainage from the tailings and waste rock into the nearby Dee River resulting in considerable contamination and a consequent massive drop in plant and microinvertebrate species. Poor quality water from the opencut is also entering the groundwater zone. A recent study of the mine and drainage systems including 21 sites up to 5 0 km downstream on the Dee River (Figure 8.12) shows high levels of contamination (Table 8.7) and precipitates of iron oxyhydroxides, aluminium hydroxides, gypsum, jarosite and other hydrated sulfate minerals. Concentrations of metals in the sediments are also high (Table 8.8). As the Dee River is used as a water supply and passes through agricultural and grazing country the impact of contaminants is widespread. The Queensland Department of Mines and Energy is presently funding a detailed study to determine: (i) the locations and relative contributions of various sources of pollution in the Dee River; (ii) the potential for remobilisation into the water column of contaminants currently held in the river sediment; and (iii) the natural processes leading to removal of contaminants from the water column. Rehabilitation of the mine and dumps are considered to involve a considerable capital investment and continuing outlays for maintenance.

G. F. Taylor

To Rockhampton Horse Paddock Dump

Upper

West Mount Morgan Dump

% Nelson's \ o Sump

1Mb

Gully Dump

WASTE ROCK Graveyard Dump

OPEN COT SMELTER Mirndic West

WASTE ROCK

Mm&jrW* a East

u^tto

Mundic

.Outlet

OS >9

Frog Hallow Stop

N E B

^Shepherds Sumps

E H Distance downstream (kilometres)

DtO—

1000m

Figure 8 . 1 2 Map of Mt Morgan, Queensland, minesite showing water-sampling sites.

Rum Jungle, Northern Territory The Rum Jungle uranium mine 80 km south of Darwin in the Northern Territory was operated from 1954 and continued until shutdown in 1971. Opencut mining of White's orebody (Figure 8.13) required diversion of the Finniss

River with overburden being deposited in a dump on the south bank of the diversion channel (Harries & Ritchie 1988). Dyson's, another uranium orebody was mined in 1957-58. In 1963 the Intermediate copper orebody was mined with the low-grade sulfide and oxide ores being put

Acid drainage at mine sites

BORROW AREA1

To Ranger Figure 8.13 Rum Jungle, Northern Territory, minesite plan prior to rehabilitation.

in a pile where an attempt was made to extract the Cu by heap leaching. Until 1961, tailings were pumped to a disposal area, with supernatant water flowing into the Finniss River. After 1961, tailings were pumped into Dyson's and White's opencuts. Initially, highly acidic (pH 1.5) waste liquor from the treatment plant was also discharged to the tailings area, but after 1961 it was pumped straight into a holding dam in the bed of the Finniss River and was released downstream in the wet season. From 1967, part of the waste liquor was pumped into White's opencut. When the mine closed in 1971 there remained three waste rock dumps, three opencut pits, a tailings dam and a copper heap leach pad (Figure 8.13). All of these contributed significantly to pollution of the Finniss River with acid drainage and heavy metals (Cu, Mn, Zn), resulting in greatly reduced aquatic flora and fauna. An estimated $16 million was spent in early 1980s to rehabilitate the mine by putting all the tailings in one of the opencuts and revegetating the

area; the heap of copper ore was removed and the area revegetated; and the two pyritic waste dumps were reshaped and capped with a three-layer cover and sown with native species. Although there has been a marked improvement in the Finniss River (Table 8.9) (Bennett & Lawton 1995) the two pyritic waste rock dumps continue to generate acid and release heavy metals. The tailings are no longer a source of contamination, but low pH (-2.5) high Cu waters from the two remaining opencuts continues to enter groundwaters. The Rum Jungle Creek South uranium mine several kilometres south of the main Rum Jungle mine was operated from 1961 to 1963 resulting in an openpit and a slightly radioactive overburden heap of 2 x 10 6 m 3 of chloritic schists and slates. In 1990 for a cost of $1.9 million the area was rehabilitated by burying spilled radioactive material from around the site and replacement with imported non-radioactive material and top soil, reshaping and stabilising the waste rock dump, and revegetating all disturbed areas. A number of

G. F. Taylor

Table 8.8 Concentrations of metals in sediments from the Dee River (mg kg -1 dry weight unless otherwise indicated). T2

Dairy Ck fines

Ag A1 (%) As Ba Ca (%) Cd Co Cr Cu Fe (%) K Mg (%) Mn Na (%) Ni Pb Ti V Zn a b

<5 5.06 —

26 0.86 <5 30 5 448 3.96 208 1.58 321 0.36 <10 <50 46 42 67

Dairy Ck coarse b

<5 1.81 42.7 162 0.255

Dee at Dee at Dairy Ck Mundic Ck Mundic Ck fine fine coarse

33 0.10

20 0.11

<5 0.74 32.8 49 0.454

1306 20.62 782 0.53 205 0.21

7 14 945 9.08 1114 0.439 191 0.14

10 18 1187 11.58 1857 0.798 307 0.29

197 159 76

81 153 137

—

2.40

6.38

17 784 15.21 806 0.52 147 0.19

74 263 79 112

148 273 112 .205

Dll

1st ford fines 3

D14

D13

Walmul coarse

fines

coarseb

D15

Dee u/s Fletchers coarse

<5 4.32

0.80

<5 1.37

<5 0.97

<5 0.77

29 0.09

52 0.08 <5 < 12 784 8.33 781 0.34 178 0.05 <10 <50 70 95 138

30 0.126 <5 6 9 296 4.75 613 0.328 242 0.04 <10 <50 380 79 62

26 0.079 <5 10 9 370 6.85 614 0.283 228 0.03 <10 <50 435 87 74

28 0.14 10.4 19 17 903 17.58 688 1.72 1480 0.89 <10 <50 99 59 144

10 489 3.45 600 0.24 116 0.10 -

203 70 47

D16

Dee d/s Fletchers coarse

D16

Dee d/s Fletchers white ppt

0.59

<5 15.83

13 0.162

9 0.60 <5 38 <5 11945 0.22 446 0.36 1215 0.21 <10 <50 11 11 955

8 263 3.91 519 0.258 185 0.05 -

398 67 60

<63 ]im. between 63 pm and 1 mm.

Table 8.9 Historical load data in the Finniss River at a site 5.6 km downstream from Rum Jungle together with a record of flow and rainfall. Year 1969/70 1970/71 1971/72 1972/73 1973/74 1982/83 1983/84 1984/85 1985/86 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93

Flow (xl0 6 m 3 )

Rainfall (mm)

Copper total (t)

Zinc total (t)

7 33 31 22 69 9.5 48 11.7 11.4 13.2 6.3 35 3.1 40.5 7.1 29.9

896 1611 1542 1545 2000 1121 1704 1136 1185 1222 1064 1600 900 1590 1002 1421

44 77 77 67 106 23 28 9.1 3.7 5.6 3.2 5.4 1.8 14.9 3.8 11.9

24 24 22 30 5 9 4.1 2.7 2.7 2 4.4 1.6 7.4 2.7 3.9

Manganese total (t)

Sulfate (t)

46 110 84 77 87 6 21 7.2 8.2 8.6 5.4 19.2 3.9 30.5 9.1 24.7

3300 12000 6600 5500 13000 1520 3600 1600 4400 2870 1230 3940 760 4000 1260 2696

Source: Bennett and Lawton (1995).

facilities were constructed and the site is now used as a public recreational area including swimming in the old openpit.

Pine Creek, Northern Territory Gold was discovered at Pine Creek, 2 2 0 km south of Darwin in the 1870s. Recent mining operations began in 1985 and ceased in late 1995. The mine is subject to monsoonal rainfall during November to March with an average annual rainfall of 1147 mm.

Much of the initial mining was of oxidised ore from the Enterprise Pit (Figure 8.14), but more recently consisted of both oxidised and unweathered rock. Tailings are therefore a mixture of both oxidised and fresh material, the latter with an acid drainage producing potential. Both south and north waste dumps are producing acid drainage (Table 8.10) which is discharged into the process-water dam. The data in Table 8.11 demonstrate that since capping of the dumps began in 1991 in conjunction with a catchment management plan (Fawcett 1995) there has been a signifi-

Acid drainage at mine sites

NORTH CANDY'S PIT SOUTH CANDY'S PIT

SOUTH WASTE ROCK DUMP Figure 8 . 1 4 Site plan of the Pine Creek, Northern Territory, goldfields minesite. Table 8 . 1 0 Water analysis results from monitoring site CK7, east of north waste dump, Pine Creek.

pH EC (mS nr 1 ) Mg (mg L"1) S04 (mg L"1) Mn (mg L"1) A1 (mg L"1) Asa (pg L"1) CdMpgL" 1 ) Cua (pg L"1) Fe a (pg L"1) Pba (pg L"1) Zna (pg L"1) a

21.1.91

Date sampled 13.2.92

20.12.94

2.9 10 890 1 976 14 920 262 649 882 1 960 48 800 53 000 264 879 000

3.9 850 60 410 5.4 9.5 8 74 1 200 250 41 15 000

5.1 400 21 56 1.3 0.26 15 7 32 30 10 1 400

Measured in filtrate

cant reduction in the amount of acid drainage reaching the process-water dam. Decommissioning of the mine has left a number of openpits, waste-rock dumps and a tailings dam in the immediate vicinity of the town of Pine Creek. Water from Pine Creek has been diverted into the Enterprise Pit which provides a large lake for recreational uses, reduces risk to public safety by reducing the exposed height of pit walls and controls local groundwater levels. Waste rock dumps

Table 8.11 Water analysis results from the process water dam, Pine Creek.

pH EC (mS nr 1 ) Mg (mg L"1) S04 (mg L"1) Mn (mg L_1) A1 (mg L"1) Asa (pg L"1) Cda (pg L-1) Cua (pg L"1) Fe a (pg L"1) Pba (pg L' 1 ) Zna (pg L"1) a

23.4.91

Date sampled 18.3.92

3.2.95

3.6 2 400 240 1 800 27 70 12 300 4 700 4 800 280 102 000

4.2 1 700 110 1 000 12 19 10 10 3 500 3 600 10 30 000

4.5 1 200 51 530 5.7 3.7 20 19 680 280 17 6 900

Measured in filtrate

and the tailings have been covered and revegetated and are being monitored for a variety of physical, chemical, hydrological and biological characteristics.

MANAGEMENT OF ACID DRAINAGE Management of acid drainage falls into three distinct categories: (i) abandoned minesites; (ii) existingfestablished mines; and (iii) deposits in the pre-feasibility and planning stage.

G. F. Taylor

Table 8 . 1 2

Cost of components of rehabilitation work at Rum

Jungle. Component

Expenditure 1986 $ (Aus)

Copper heap leach pile/tailings dam/ Dyson's open cut

3 685 925

Waste-rock dumps

2 826 997

Opencut water treatment

6 234 221

Other areas

904 720

Management

4 666 040

Monitoring Total

322 346 18 6 4 0 2 4 9

The cost of rehabilitation of abandoned mines will vary according to the nature and extent of contamination. In USA it has been estimated the total cost of cleaning up the Summitville, Colorado gold mine to be US$100-120 million. Similarly, the Iron Mountain, California Cu mine which has mine drainage pH of <2 will cost somewhere in the order of US$150 million for the most effective and politically negotiable combination of treatments. Both of these mines are in high rainfall, mountainous areas and adjacent to major water supplies. In Wyoming US$253 million has been spent on abandoned mine projects since 1977. This includes about US$400 000 per hectare for subsidence; US$22 000 per hectare for reclaiming abandoned surface coal mines; an average of US$4000 to close each of the 254 small underground non-coal mines; and US$32 000 per hectare to reclaim surface uranium mines (Richmond 1995). Costs in Australia are comparable. Estimated cost for complete rehabilitation of the Brukunga Pyrite Mine is $50 million with present annual cost of treating acid drainage with lime, and covering and revegetating the tailings being in excess of $500 000. Detailed costing for the rehabilitation work at Rum Jungle uranium mine (Bennett & Lawton 1995) is given in Table 8.12. The cost of rehabilitating waste rock dumps, including reshaping before covering, was $68 500 per hectare. Despite this expenditure, rehabilitation has not been completely effective, although the Finniss River is much improved. At Pajingo, Queensland, capping of the waste-rock dump with three 0.5 m-thick compacted layers and a loose rock mulch layer cost approximately $50 000 per hectare (Tredinnick & Cornwell 1995). Revegetation, consisting of minor earthworks, seeding with natives and pasture cost an additional $1500 per hectare. At Mt Leyshon, Queensland the cost of building a runoff control scheme was $1 million with annual extensions costing around $40 000 (Orr 1995). To construct the benign porphyry seal (1 m thickness) on the waste-rock dumps costs $4100 per hectare and this does not include the cost of selectively handling the porphyry or shaping of the dump both of which are integral parts of the mining operations. The total cost of dump rehabilitation for flat surfaces is about $12 000 per hectare. Despite the considerable sum spent on rehabilitation ($2.3 million in 1976), pollution problems still exist at Captains Flat. Dames and Moore (1993) estimated that the only feasible way of overcoming these problems is by treatment of the point sources with chemical and/or biological processes. The estimated cost of a chemical treatment plant is estimated at $641 000 with annual operating costs of $81

000. Combined chemical and biological treatment would cost an estimated $679 500 with annual operating costs of $30 000. Broad hectare rehabilitation of the diffuse sources was estimated to cost over $1.5 million. No doubt these estimates would have to be substantially increased to undertake this work today. The costs of constructing wetlands to remove sediment and reduce pH and total dissolved solids from seepage can vary enormously and depend particularly on the topography of existing landforms, whether natural or synthetic liners are required, and the composition of the effluent. Typical price ranges quoted in the US are US$10-100 m~2. The various techniques for control and treatment of acid drainage are discussed below. This is then followed by methods for prediction of acid drainage, monitoring of remediation/restoration effectiveness and future directions.

CONTROL AND TREATMENT Covers Prevention is better than cure, because once oxidation of sulfidic minewastes has started it is extremely difficult to stop. The essential ingredients in the formation of acid drainage are sulfides, oxygen and water with bacteria to catalyse the reaction. At abandoned and existing mines, it is not possible to remove the sulfides so the majority of efforts are aimed at eliminating oxygen and/or water. The most commonly utilised technique is the construction of a cover over the sulfidic minewaste—'wet' covers to eliminate oxygen and 'dry' covers to stop water infiltration. Recent studies in Canada under the industry- and government-funded MEND program (Feasby et al 1997) have shown the use of water covers and underwater disposal to be the best prevention technology for unoxidised sulfidic minewastes. This technology is particularly well suited to Canadian geography and climate and may well be utilised in Scandinavia and Russia. The various types of 'wet' covers are: (i) waste deposition in lakes, the ocean or openpits; (ii) water ponded on the surface of sulfidic wastes; (iii) constructed wetland on the surface of sulfidic waste; (iv) watersaturated soil cover (perched water-table); and (v) elevated water-table. A typical example of water ponded on the surface of sulfidic tailings (Figure 8.15 on Plate 3) is at one of the old uranium mines at Elliot Lake, Canada. Although already partially oxidised, this water cover has substantially reduced acid drainage. The only area in Australia where 'wet' covers are practical is Tasmania because precipitation exceeds evaporation for most of the year. At the Renison tin mine, research has indicated that a constructed wetland is the most effective means for reducing acid drainage (Jones et al 1997) particularly if the tailings are first covered with a benign layer of non-sulfidic material. In those countries where evaporation exceeds precipitation making the maintenance of 'wet' covers extremely difficult, widespread use is made of 'dry' covers. These are usually composed of benign waste-rock, stock-piled topsoils or imported clays/soils. In the past and, unfortunately, in some existing cases, the purpose of such covers has been based on aesthetics—'out of sight, out of mind'—and pro-

Acid drainage at mine sites

viding a substrate for revegetation. Such techniques have failed over time and acid seepage reemerges (e.g. Captains Flat, New South Wales; Rum Jungle, Northern Territory) or never ceased (Brukunga, South Australia). Much of the present research and application is aimed at defining the optimum conditions for zero or low waterflux covers and oxygen-limiting covers based on the hydrological properties of the cover material (e.g. the design of covers for the Kidston, Queensland minewaste dumps: Bews et al 1997). These covers are designed to store water during wet seasons and lose it through evapotranspiration during the dry. An additional design feature is to optimise water management; runoff in humid climates and internally draining in semiarid to arid environments to enhance ecosystem reconstruction as at Oaky Creek and Newlands coal mines in Queensland (McNamara et al 1998). One of the potential consequences of evapotranspiration is the precipitation of secondary salts at or near the surface of the 'dry' cover (Elliott et al 1997). Such salts (usually soluble sulfates) result in the death of any vegetation planted to increase evapotranspiration and reduce erosion. To eliminate this effect, capillary breaks are incorporated within the 'dry' cover. 'Dry' covers may also be used as oxygen-limiting covers in humid climates by maintaining maximum saturation. Again, considerable research is being undertaken to determine the most appropriate cover material. This material is seldom found at the minesite and must be imported from 'borrow' pits. Under the Canadian MEND program several forms of organic wastes have been trialled as oxygen-reducing covers. Material such as sewage sludge (Elliott et al 1997), papermaking residue (Cabral et al 1997), lumber waste (Tasse et al 1997), municipal solid-waste compost (Pierce 1992; Pierce et al 1994) and peat (Elliott et al 1997) have all been shown to consume oxygen by biological action. However, there is some doubt as to the long-term efficacy of this technique as the organic matter is consumed. Rapid production of leaf litter on such surfaces may retain their integrity and permit development of sustainable ecosystems. Other techniques to reduce oxygen and water infiltration into sulfidic minewastes include covering with desulfurised tailings (Hanton-Fong et al 1997; Bussiere et al 1997), neutralisation sludge (Coleman et al 1997), blending and layering (Mehling et al 1997) and in-pit disposal (Orara et al 1997; Lewis-Russ et al 1997). Each technique has exhibited initial success in reducing acid drainage generation and contaminant release. There is considerable concern relating to the long-term stability of neutralisation sludge as it is composed primarily of gypsum which is partially water soluble thus releasing any coprecipitated contaminant elements. Neutralisation sludge is widely used to cover tailings and waste rock at the old Brukunga mine in South Australia.

Australia, tailings dam). Abandoned or old minesites such as Britannia (Canada), Summitville, Leadville, Central City (USA), Mt Lyell, Captains Flat, Mt Morgan, Rum Jungle, Brukunga (Australia) all require control or treatment of the acid drainage to prevent widespread off-site contamination. Neutralisation is still the most commonly utilised and effective technique to treat acid drainage. This involves the mixing of finely ground calcite, dolomite, magnesite or lime in the acid drainage to react as for example in the reaction 2H+ + S 0 4 2 + + CaC0 3

Ca 2+ + S 0 4 2 " + H 2 0 + C0 2

with a precipitate of gypsum and contaminant elements forming. Care should be exercised if dolomite or magnesite are used as magnesium sulfate is extremely soluble and acts as a strong purgative. If siderite (FeC0 3 ) is used, it will neutralise the acid drainage but oxidation and hydrolysis of the released Fe 2+ (ferrolysis) will create more acid (see under Acid Generation). Other factors which must be taken into consideration are: what resources (including finance) will be available for the long-term (up to 300 years) neutralisation; and how will the sludge be disposed of. Passive neutralisation is commonly used in coal mines in the USA but may also be effective in the short-term at metalliferous minesites. Again, the long-term viability is not assured as either the neutralising capacity may be exceeded or the secondary minerals formed by neutralisation may passivate the neutraliser. Anoxic limestone drains tried at a mine in Nova Scotia (Feasby et al 1997) were disappointing. Another potential technique is to mix sufficient limestone/lime with the sulfidic waste prior to deposition. This is potentially expensive and may add substantially to the volume to be disposed of. Wetlands Constructed or natural wetland systems have been utilised throughout the mining community as a possible means of increasing the pH, decreasing the contaminant load and sedimenting out any suspended load. Mines in Australia utilising wetlands include Ranger, Woodcutters, Cosmo Howley (Northern Territory); Hellyer, Tasmania and Mt Morgan (Queensland). Feasby et al (1997) noted that results in Canada have been disappointing but Sobolewski (1997) reported that natural wetlands improved water quality, which at many mines was to full environmental compliance. At Ranger, a constructed wetland is effective in removing U and Mn from near-neutral pH mine waters but is less effective in reducing Mg and S 0 4 2 " levels (Jones et al 1997). The difference between constructed and natural wetlands is the maturity of the latter with the impact of organic matter and micro-organisms playing a greater role.

Neutralisation

Bioremediation

There is general consensus that in other than humid climates where 'wet' covers are effective, prevention of acid drainage is not possible. The previously described techniques can only reduce the rate of formation. Many sulfidebearing wastes may be subject to advective oxygen infiltration (Garvie et al 1997) requiring specially designed covers, or lateral water flow (e.g. Brukunga, South

Bioremediation is an essential ingredient of wetland systems, particularly the activity of sulfate-reducing bacteria, resulting in the precipitation of stable sulfides below the water-table. The use of bactericides has also been tried in the control of acid drainage and has met with limited success (Sobek et al 1990). This technique has not received widespread application. The use of active bioremediation

G. F. Taylor

using bioreactors may be cost-effective for small seepages but at larger sites can only be used as part of an integrated system to recover ore minerals (Struthers et al 1997). A technique which has been tested in both the laboratory and the field is the development of porous reactive walls (Waybrant et al 1997; McRae et al 1999; McGregor et al 1999). The walls are composed of different organic mixtures (leaf mulch, sawdust, sewage sludge and wood chips together with calcite) and are effective in increasing the pH and alkalinity, and decreasing S 0 4 and many cations from seepage plumes. However, as this is a relatively new technology, its long-term viability has not been established. The technique may not have widespread applicability in Australia because of the relative lack of suitable organic matter (sparse vegetation cover and poor composting characteristics). Waring and Taylor (2000) have recently described an alternative to flooding underground mines to prevent acid drainage. The technique involves controlling the atmosphere in the mine using organic matter to passively displace air with carbon dioxide and methane produced by anaerobic bacteria. Where the contaminant load is sufficiently high SX/EW (solvent extraction/electro-winning) techniques may be used to recover contaminant elements (Miedecke et al 1997).

PREDICTION For operating mines and those in the feasibility stage, it is important to know whether or not a rock mass is likely to produce acid drainage or not, and if it is, how much and at what rate. A further consideration is whether the acid drainage will be just acid+sulfate+iron, or will it also contain other heavy metals. The first question is answered by undertaking predictive tests and the second by detailed sampling, mineralogical and geochemical analysis (see later). Prediction of acid drainage from sulfidic minewastes is dependent on static and kinetic chemical tests and the development and application of models. Presently utilised predictive techniques are described by Lawrence and Day (1997) and Comarmond (1997). Each of these techniques is based on the acid generated by the sulfide present and how much can be neutralised by carbonates and other minerals present in the gangue. This process is commonly called acid-base accounting. Where can acid-base accounting go wrong? Rocks and ore deposits are notoriously heterogeneous so that unless the sampling is thorough, the acid-base accounting may not be representative. Not all sulfur is present as pyrite, but as other sulfides, sulfates and organic sulfur which may lead to an erroneous estimation of acid production. Similarly, not all carbon is present as calcite and may be in the form of other carbonate minerals, organic carbon and/or graphite. Some silicates can contribute to acid neutralisation such as plagioclase feldspar: CaAl2 Si 2 O g + 2H + + 6 H 2 0 6 0 H" CaAl2 S i 2 0 8 + 2H + + H 2 0

Ca 2 + + 2A13+ + 2 H 2 S i 0 4 +

Ca 2 + + A l 2 S i 2 0 5 ( 0 H ) 4

Acid-base accounting can be improved by constant geological, mineralogical and geochemical control of sampling to take into account heterogeneity. Understanding the processes that will take place in the rocks after mining and processing will also assist. Another technique is the application of mathematical models to predict the oxidation rate of pyrite in waste-rock dumps and tailings dams and pollutant generation. A number of models based upon the rate of transport of oxygen to the reaction site are described by Ritchie (1994) and Wunderly et al (1996). Such techniques have been used to predict the mitigating effect of various covers on acid generating tailings at the Renison tin mine in Tasmania (Jones et al 1997).

RESTORATION The restoration of minesites producing acid drainage is expensive and extremely difficult. It is far cheaper and less difficult to plan and implement strategies at the start-up phase to ensure that sulfidic minewastes are properly managed to have minimal impact on the environment. What criteria are needed for minesite restoration? They should include: (i) ultimate land use—community and regulatory authorities' expectations; (ii) physical and chemical stability; (iii) self-sustainability—after say 5 years does not require further human intervention; (iv) cost-effectiveness; and (v) preservation of mining heritage. Recently the Minerals Council of Australia released its discussion paper (Minerals Council of Australia 1997) on mine closure that included the following six groups of issues—(i) stakeholder involvement; (ii) planning; (iii) financial provision; (iv) management; (v) standards; and (vi) relinquishment— which cover the five criteria above. The ultimate land use may well be the source of conflict between the various stakeholders with for example the mining company wishing to do very little as it may wish to reprocess or remine, the regulatory authorities requiring return to a native ecosystem, and the community suggesting retention as mining heritage (tourism), recreation or agriculture. Each will require different restoration techniques. In some instances, legislation leaves no doubt as to the expectations for restoration, e.g. the Ranger minesite adjacent to Kakadu National Park or the bauxite/gold mines in the jarrah forests of Western Australia. On the other hand there appears to be confusion as to whether the overburden from openpit coal mining in Queensland should be restored to native ecosystems or agricultural/pastoral usage. At Mt Lyell, the denuded hills caused by tree felling, acid rain and erosion are tourist attractions and the local community does not want them restored. However, tailings and acid drainage in the local drainage are of considerable concern particularly to owners of aquiculture in for example Macquarie Harbour, Tasmania. Reshaping and covering acid minewastes must not only meet the chemical requirements previously described above under Control and Treatment but also physical stability. Erosion and weathering are natural processes that are not easily stopped. However, these processes are normally very slow. The constructed landforms at minesites (voids, mines, waste-rock dumps, tailings-storage facilities) may weather and erode slightly faster because of greater exposed surface

Acid drainage at mine sites

area. The object is to prevent catastrophic events such as mass movement of waste rock or failure of tailings dam walls as at Marcopper, Philippines in 1996, Los Frailes, Spain in 1998 and Baia Mare, Romania in 2000. Other impacts are from dust and turbidity of waterways. Weathering may result in secondary minerals which are erodible, of lesser strength or change the hydrological properties. It is no longer acceptable to engineer a structure to hide acid-generating material and provide a substrate for agriculturally based revegetation. Although initially aesthetically acceptable, it soon breaks down with acid burning of vegetation and subsequent erosion (Figure 8.16 on Plate 3). What is now accepted as best practice is the construction of an engineered system based upon knowledge of what is to be achieved (Wilson 1997). One of the major outcomes of an engineered system is the development of a self-sustaining ecosystem (Kearns & Barnett 1998). The expense of even the most basic restoration is prohibitive and unless planned for during operational mine life can result in public moneys being used. Captains Flat, New South Wales and Rum Jungle, Northern Territory are two Australian examples where government finance was necessary to remediate acid drainage. Recent planning to remediate a tailings storage facility in the Northern Territory looked at a thick polypropylene cover to exclude oxygen and water. The projected cost was in excess of $13 million compared with $ 1-2 million for a conventional store and release cover. Finally, there are many mines in Australia that have a high heritage value. Reference has already been made to Mt Lyell in Tasmania. The Mt Morgan minesite in Queensland, which provided finance for the initiation of British Petroleum and the Walter and Eliza Hall Trust, has many historical structures, waste dumps and clay pits, which may be impediments to amelioration of acid drainage and restoration. Whereas at Glen Osmond, Moonta, Wallaroo and Burra in South Australia, the Cornish mining heritage is well-preserved and attracts many tourists.

MONITORING AND ANALYSES Rehabilitation is expensive and complicated, and its beneficial impacts require monitoring on a regular basis. This is particularly true in situations where acid drainage may not become obvious for several years due to remnant alkalinity in contained processing waters or where limited buffering capacity exists. A typical example is the former Gordonstone coal mine in Queensland where the washery wastes have a high pyrite content. However, the process water which is pumped from the mine has a pH of 8.5 and evidence of acid drainage did not appear for several years after deposition of the wastes. It is extremely important to decide what is to be monitored and for what purpose. In situ measurements of geotechnical, hydrological, physical and chemical characteristics of waste rock dumps, ROM dumps, tailingsstorage facilities, openpits and mines are not only important in acid drainage management but also in the overall environmental management of a minesite. Although not directly impacting on acid drainage, many of the geotechnical/physical characteristics will determine such properties as water

and oxygen content and flux. Of particular relevance are the geotechnical characteristics of dam walls and dry covers designed for long-term management of acid generation. These may include permeability, pore-water pressure, shear strength, density, compaction and settlement, all of which are commonly determined at the time of installation but are seldom monitored over longer periods. Perhaps such monitoring may have prevented or given adequate warning of some of the recent catastrophic tailings dams failures. The immediate state of sulfidic minewastes is determined by measurements of temperature, porewater and seepage pH, EC, Eh and chemistry, and the concentrations of oxygen, carbon dioxide, methane and gaseous sulfur compounds in pore gases. Changes over a period probably indicate that sulfide oxidation has commenced and that contaminant load is increasing. In particular, gaseous oxygen concentration in the minewastes is a direct indicator of the potential for sulfide oxidation. In situ, on-line measuring of oxygen ingress through a synthetic cover over pyriterich tailings using specially developed oxygen probes was reported by Patterson and Davis (1999). Off-site monitoring of the chemical and biological characteristics of soils, surface and groundwaters are a measure of contamination and the effectiveness of restoration. However, such monitoring may be of little consequence if detailed baseline studies have not been undertaken. Exploration data such as soil, stream sediment, and groundwater geochemistry used to locate the mineralisation may be used as baseline studies. It is often difficult to determine the nature and extent of groundwater contamination. Geophysical techniques such as electromagnetic and electrical methods may be used to detect and measure the extent of contaminant plumes (Buselli et al 1991). Recent developments in remote-sensing techniques such as imaging spectrometers have been utilised to map acidic minewastes and drainage (Warren & Hick 1996; Swayze et al 2000). Hyperspectral mapping can be similarly used or in monitoring impacts on vegetation. Rehabilitation of minesites is more commonly aimed at the development of sustainable ecosystems. But there has been no universally acceptable indicators of successful rehabilitation, particularly to allow companies to walk away from disused minesites. A system based on landscape function analysis has now been widely accepted in Australia as the most appropriate method of monitoring ecosystem rehabilitation success (Tongway et al 1997). Hyperspectral remote sensing can be used to rapidly monitor health of large-scale revegetation such as at bauxite, mineral sands and iron ore mines (Warren & Hick 1996). The various analytical techniques to determine the chemistry of minewastes and waters are detailed by Garvie and Taylor (2000) together with a number of monitoring techniques. But as is the case in any geoscientific examination, the chemical data are of little use unless the sampling and sample preparation are appropriate. These aspects are also covered by Garvie and Taylor (2000).

FUTURE DIRECTIONS The most significant factor in preventing or reducing acid drainage is the application to mine planning of geological

G. F. Taylor

data gained prior to mining. Data that are vital include mineralogy, geochemistry, structure, hydrology and geotechnical properties, many of which are available from exploration and pre-feasibility drilling and, in some instances, from exploratory shafts or declines. Predictive tests can also be formed on drill samples to ascertain the potential for acid drainage. Several new mines in Australia have taken this approach at the outset of mine development. At the Cadia Mine, New South Wales geological data were utilised to develop a block model (Scott et al 1997) which can be utilised to schedule mining of acid-producing waste and its placement to minimise acid drainage. Techniques to prevent acid drainage or ameliorate its impact on the environment described above are accepted and applied throughout the world. There are other techniques which are being tested including: (i) the formation of hardpans/cements by oxidation of sulfides at or near the surface to form an impermeable cover on tailings (Figure 8.17 on Plate 3) (Agnew & Taylor 2000a, b) by a process similar to the natural formation of ferricretes, silcretes and calcretes; (ii) removal of contaminant elements from tailings in a recycling process (Struthers et al 1997); (iii) codisposal of tailings with waste rock or coarse rejects as has already been utilised in the coal industry (Williams 1996) and is now being trialled at Kidston, Queensland; (iv) accelerated oxidation of sulfidic minewastes so that all the acid drainage has been generated during the life-time of the mine and the elements removed from solution by SX/EW techniques; (v) formation of stable secondary minerals by oxidation processes—many minerals formed at low pH are stable in the Australian environment and incorporate many heavy metal contaminants (Scott 1987; White 1996); (vi) whereas coarse-grained tailings have been used as underground minefill, newly developed 'paste technology' allows a mixture of coarse and fine tailings to be similarly used—other mines are utilising dewatered tailings underground; (vii) new water-treatment techniques such as kaolin amorphous derivative (Mackinnon et al 1997) and the green precipitate process (Taylor 1997) to remove contaminant elements are being trialled; and (viii) passivation of sulfidic wastes by using a variety of coating material has been trialled for more than decade—coating materials include phosphate rock (Fytas et al 1998) (which reacts with Fe 2 + in presence of acid to form a non-reactive iron phosphate layer), and silicates (Fytas et al 1999). In addition, Struthers et al (1997) examined a variety of uses for sulfidic tailings, in an attempt to revert waste to something of economic benefit.

into sulfuric acid which can be sold to other users. However, at some smelters such as Norilsk, Russia, sulfur dioxide from the nickel smelter has devastated forests within a diameter of tens of kilometres. Another source of acid drainage in coastal Australia are the acid sulfate soils. These have formed by action of sulfate-reducing bacteria on marine sulfates in anaerobic environments such as mangrove swamps. If these areas are disturbed for development or agriculture the fine-grained iron sulfides rapidly oxidise to form acid drainage (Dent 1986). Another implication of the oxidation of pyrite and marcasite together with carbonaceous matter in coal wastes may lead to spontaneous combustion. There has been considerable research to determine the proportions of these ingredients (and moisture) leading to spontaneous combustion and the nature of the gaseous products which contribute to greenhouse emissions and which may be carcinogenic. It may also occur in base-metal mine tailings particularly if the highly reactive pyrrhotite and marcasite are present. The processes have been described for the copper orebodies at Mt Isa (Lukaszewski 1968, 1969).

CONCLUSIONS Acid drainage formed principally by the oxidation of sulfides in mines and minewastes is a potentially massive liability to the mining industry. In cool to cold humid environments, the most effective technique of preventing acid drainage is to cover the sulfides with water, thus preventing oxygen entering the waste. This technique is not possible in many countries including Australia where evaporation exceeds precipitation. Covers designed to store and release water are the most commonly used technique to rehabilitate minewastes. They must also be designed for long-term physical and chemical stability as well as being the foundation of self-sustaining ecosystems. Continued research on economic use of minewastes, innovative technologies to prevent or ameliorate acid drainage and a commitment by the industry, regulators and the community to best practice and due diligence will ensure the environment is not compromised. Geoscientists have an important role in this and other environmental issues related to the mining industry.

REFERENCES AGNEW M. & TAYLOR G. 2000a. The classification, development and

OTHER SOURCES OF ACID DRAINAGE Acid drainage is not limited to sulfidic minewastes. Although most Australian coals have a low sulfur content, the combustion of 'dirty' coals in many countries leads to formation of S 0 2 which combines with moisture in the atmosphere. This may fall over a wide area as 'acid rain' killing native vegetation and lowering the pH of surface water and soils. A similar reaction occurs around smelters (nickel, copper, lead, zinc) which can emit hundreds of thousands of tonnes of S 0 2 into the atmosphere each year. Many smelters now have scrubbers which convert the S 0 2

cycling of hardpans and cemented layers in tailings storage facilities in Australia. In: Grundon N. J. & Bell L. C. eds. Proceedings of the 4th Australian Acid Mine Drainage Workshop, Townsville, March 2000, pp. 157-169. AGNEW M. & TAYLOR G. 2000b.Laterally extensive surface hardpans in tailings storage facilities as possible inhibitors of acid rock drainage. Proceedings of the 5th ICARD Conference, Denver, Colorado, May 2000, pp. 1337-1346. ALPERS C. N . , NORDSTROM D. K. & THOMPSON J. M. 1994. Seasonal variations of Zn/Cu ratios in acid mine water from Iron Mountain, California. In: Alpers C. N. & Blowes D. W. eds. Environmental Geochemistry of Sulfide Oxidation, pp. 324-344. American Chemical Society Symposium Series, Washington, DC. ANZECC/NHMRC 1992. Australian and New Zealand guidelines for the assessment of contaminated sites. Australian and New Zealand

Acid drainage at mine sites Environment and Conservation Council and National Health and Medical Research Council. Comparison of oxidation rates of sulfidic mine wastes measured in the laboratory and field. Australian Centre for Mining Environmental Research, Brisbane. BENNETT J. W. & LAWTON M. D. 1995. Assessment of the Rum Jungle strategy for acid mine drainage control. In: Grundon N. J. & Bell L. C. eds. 2nd Australian Acid Mine Drainage Workshop, pp. 179-190. Australian Centre for Mining Environmental Research, Brisbane.

BENNETT J . W . , COMARMOND M . J . & JEFFREY J . J . 2 0 0 0 .

B E W S B . E . , O ' K A N E M . A . , W I L S O N G . W . , WILLIAMS D . & CURREY N .

1997. The design of a low flux cover system, including lysimeters, for acid generating waste rock in semi-arid environments. Proceedings of the 4 th I CARD Conference, Vancouver, BC, June 1997, pp. 747-762. BROOKS K. A. 1980. Mine waste pollution of the Molonglo River: an investigation into the effect of remedial measures carried out in 1976 at Captains Flat. Chemical Laboratories New South Wales Department of Mineral Resources. BROOKS K. A. 1998. Captains Flat rehabilitation—20 years on. In: McLean R. W. & Bell L. C. eds. Proceedings of the 3rd Australian Acid Mine Drainage Workshop, Darwin, pp. 103-113. Australian Centre for Mining Environmental Research, Brisbane. BUSELLI G . , BARBER C . , DAVIS G . B . & WILLIAMSON D . R . 1 9 9 1 . T h e u s e

of electro-geophysical methods for groundwater pollution and soil salinity problems. Exploration Geophysics 22, 59-64.

BUSSIERE

B.,

NICHOLSON R .

V . , AUBERTIN M .

SERVANT S .

1997.

Effectiveness of covers built with desulfurised tailings: column tests investigations. Proceedings of the 4 th ICARD Conference, Vancouver, BC, June 1997, pp. 7 6 3 - 7 7 8 .

CABRAL A . , LEFEBVRE G . , BURNOTTE F., PROULX M - F . , AUDET C . , LABBE

1997. Use of deinking residues as cover material in the prevention of AMD generation at an abandoned mine site. Proceedings of the 4 th ICARD Conference, Vancouver, BC, June 1997, pp. 1109-1123. COLEMAN M . M . D . , WHALEN T . J . & LANDVA A . 1 9 9 7 . Investigation of the placement of lime neutralisation sludge on acid generating waste rock. Proceedings of the 4 th ICARD Conference, Vancouver, BC, June 1997, pp. 1 1 6 3 - 1 1 7 8 . COMARMOND J. 1997. Chemical methods for predicting the acid mine drainage/acid rock drainage potential of mine wastes in Australia. In: Harries J. ed. Acid mine drainage in Australia - its extent and future liability, pp. 71-84. Supervising Science Report 125. DAMES AND M O O R E 1993. Final report, Captains Flat mine site. Assessment of options for further remediation for New South Wales Environment Protection Authority. DENT D . 1986. Acid sulfate soils: a baseline for research and development. International Institute for Land Reclamation and Improvement, Holland, Publication 39. M . & MICHAUD C .

Conference, Vancouver, BC, June 1997, pp. 3 3 3 - 3 5 0 . 1997. Evaluation of low sulfur tailings in the prevention of acid mine drainage. Proceedings of the 4 th ICARD Conference, Vancouver, BC, June 1997, pp. 835-852. HARRIES J. R. & RITCHIE A. I. M. 1988. Rehabilitation measures at the Rum Jungle mine site. In: Salomons W. & Forstner U. eds. Environmental Management of Solid Wastes, pp. 131-151. Springer-Verlag, Berlin. JONES D. R., ELLERBROEK D. A. & LASZCZYK H. 1997. Evaluating closeout options for acid generating tailings. Proceedings of the 4th ICARD Conference, Vancouver, BC, June 1997, pp. 1303-1319. KEARNS A. & BARNETT G . 1 9 9 8 . Use of ecosystem function analysis in the mining industry. In: Asher C. J. & Bell L. C. eds. Proceedings of Workshop on Indicators of Ecosystem Rehabilitation Success, pp. 3 1 - 4 6 . Australian Centre for Mining Environmental Research, Brisbane. KLAPPER H . & SCHULTZE M . 1 9 9 7 . Sulfur acidic mining lakes in Germany - ways to controlling geogenic acidification. Proceedings of the 4th ICARD Conference, Vancouver, BC, June 1997, pp. HANTON-FONG C. J., BLOWES D. W. & STUPARYK R. A .

1727-1744.

L. 1997a. The Mount Lyell remediation research and demonstration program (MLRRDP) and other Tasmanian acid drainage issues. In: McLean R. W. & Bell L. C. eds. 3rd Australian Acid Mine Drainage Workshop, pp. 171-181. Australian Centre for Mining Environmental Research Brisbane. KOEHNKEN L. 1997b. Mount Lyell remediation. Final report. Supervising Scientist Report 126. LAWRENCE R . & DAY S. 1997. Chemical prediction techniques for A R D . Short Course, 4th I C A R D Conference, Vancouver, BC, June 1997. LEE M. 1999. Management of sulfidic mine wastes. Handbook. Australian Centre for Mining Environmental Research, Brisbane. LEWIS-RUSS A., BRONSON J., LUPO J. F. & LONG B . G. 1997. Control of acid rock drainage by pit backfilling. Proceedings of the 4 th ICARD Conference, Vancouver, BC, June 1997, pp. 1377-1388. LUKASZEWSKI G. M. 1968. Reaction of ANFO explosives with mineral sulphides. Australasian Institute of Mining and Metallurgy Proceedings 228, 61-70. LUKASZEWSKI G. M. 1969. Natural oxidation and the reaction of ANFO explosives in mineral sulphides at Mount Isa Mines Limited, Australia. Commonwealth Mining and Metallurgy Congress (preprint). Offices of the Congress and Institution of Mining and Metallurgy, London. MACKINNON I . D . R. EXELBY H . SING B. .& PAGE D . 1 9 9 7 . Amorphous derivatives of aluminosilicate clays: new materials for acid mine drainage remediation. In: McLean R. W. & Bell L. C. eds. 3rd Australian Workshop on Acid Mine Drainage, Darwin, NT, pp. 1 9 5 - 2 0 1 . Australian Centre for Mining Environmental Research, Brisbane. KOEHNKEN

M C G R E G O R R . , BLOWES D . , LUDWIG R . , PRINGLE E . , CHOI M . , BOOTH R . ,

1997.

1199-1207.

DUCHENE M . & POMEROY M . 1999. Remediation of heavy metalcontaminated groundwater using a porous reactive wall within an urban setting. In: Proceedings Mining and the Environment II, pp. 645-653. Sudbury, Ontario, September 1999. MCNAMARA R., LEFEBVRE N . & JOYCE J . 1998. Assessment of mine site rehabilitation performance at the Oaky Creek coal mine, Bowen Basin, Central Queensland. In: Asher C. J. & Bell L. C. eds. Indicators of Ecosystem Rehabilitation Success, pp. 125-137. Australian Centre for Mining Environmental Research, Brisbane. M C Q U A D E C. V., JOHNSTON J . F. & INNES S. M. 1995. Review of historical literature and data on the sources and quality of effluent from the Mount Lyell lease site. Supervising Scientist Report 104. M C R A E C. W. T., BLOWES D . W. & PTACEK C. J. 1999. In-situ removal of arsenic from groundwater using permeable reactive barriers: a laboratory study. Proceedings Mining and the Environment II. Sudbury, Ontario. September 1999. pp. 601-609. MEHLING P. E , DAY S . J. & SEXSMITH K . S . 1 9 9 7 . Blending and layering waste rock to delay, mitigate or prevent acid drainage: a case study review. Proceedings of the 4 th ICARD Conference, Vancouver, BC, June 1997, pp. 9 5 1 - 9 7 0 .

GARVIE A. M. & TAYLOR G .

F. 2000. Techniques to quantify processes associated with polluted effluent from sulfidic mine wastes. Australian Centre for Mining Environmental Research, Brisbane. GARVIE A . M., BENNETT J . W . & RITCHIE A . I . M. 1 9 9 7 . Quantifying the spatial dependence of the sulfide oxidation rate in a waste rock dump at Mt Lyell, Tasmania. Proceedings of the 4th ICARD

MIEDECKE J . , MILLER S . , GOWEN M . , RITCHIE I . , JOHNSTON J . & M C B R I D E ,

ELLIOTT L . C . M . , DAVISON J . G . , LIANGXUA LIU & STOGRAN S . W .

Evaporative losses, salt migration and mineralogical changes in dry covers over oxidised tailings .Proceedings of the 4 th ICARD Conference, Vancouver, BC, June 1997, pp. 1 2 1 1 - 1 2 3 0 . FAWCETT M. N. R. 1995. Management of AMD at the Pine Creek Mine. In: Grundon N. J. & Bell L. C. eds. 2nd Australian Acid Mine Drainage Workshop, Charters, Towers, pp. 83-90. Australian Centre for Mining Environmental Research, Brisbane. FEASBY D . G . , TREMBLAY G . A . & WEATHERALL C. J . 1 9 9 7 . A decade of technology improvement to the challenge of acid drainage—a Canadian perspective. Proceedings of the 4th ICARD Conference, Vancouver, BC, June 1997, pp. i-iv. FYTAS K., EVANGELOU B . & BORDELEAU D. 1998. Pyrite coating technology using phosphate. In: Proceedings of IV International Conference on Clean Technologies for the Mining Industry, Santiago, Chile. May 1998. pp. 79-86. FYTAS K . , BOSQUET P. & EVANGELOU B . 1 9 9 9 . Application of silicate coatings on pyrite to prevent acid mine drainage. In: Proceedings Mining and the Environment II, Sudbury, Ontario September 1999. pp.

P. 1997. Remediation options (including copper recovery by SX/EW) to reduce acid drainage from historical mining operations at Mt Lyell, Western Tasmania, Australia. Proceedings of the 4th ICARD Conference, Vancouver, BC, June 1997, pp. 1451-1468. MINERALS COUNCIL OF AUSTRALIA 1997 Towards a Strategy for the

G. F. Taylor

Australian Minerals Canberra.

Industry.

Minerals Council of Australia,

NEUKIRCHNER R. J. & HINRICHS D .R. 1997. Effects of ore body inun-

dation—a case study. Proceedings of the 4 th ICARD Conference, Vancouver; BC, June 1997, pp. 1469-1484. NIVEN R. 1990 The hydrogeology and geochemistry of seepage beneath the south tailings dam, Woodlawn mine, NSW. In: AMIC Environmental Workshop, Wollongong, Vol. 1, pp. 172-206. AMIC, Canberra ORAVA D . A . , TREMBLAY G . A . , TIBBLE P. A . & NICHOLSON R . V .

1997.

Prevention of acid rock drainage through the application of in-pit disposal and elevated water table concepts. Proceedings of the 4th ICARD Conference, Vancouver; BC, June 1997, pp. 971-988. ORR M. S. 1995. Development of a waste dump management strategy at Mt Leyshon gold mine. In: Grundon N. J.& Bell L. C. eds. 2nd Australian Acid Mine Drainage Workshop, pp. 91-100. Australian Centre for Mining Environmental Research, Brisbane. PARKER G. 1999. A critical review of acid generation resulting from sulfide oxidation: processes, treatment and control. In: Parker G. & Robertson A. eds. Acid Drainage, pp. 1-182. Australian Minerals and Energy Environment Foundation, Melbourne. Occasional Paper 11. PATTERSON B. M. & DAVIS G. B. 1999. Remote in situ monitoring of

gaseous and dissolved oxygen in sulfidic minewaste repositories. In: 2nd AMEEF Innovation Conference, Perth, pp. 57-60. Australian Minerals and Energy Environment Foundation, Melbourne.

PENDLETON J . A . , POSEY H . H . & LONG M .

1995.

Characterising

Summitville and its impacts: setting the scene. In: Posey H. H., Pendleton J. A. & van Jyl D. eds. Proceedings, Summitville Forum '95, pp. 1-12. Colorado Geological Survey Special Publication 38. PIERCE W. G. 1992. Reclamation of sulphide tailings using municipal solid waste compost: literature review and recommendations. CANMET, Ottawa. Project NEDEM 2.25.1(a). PIERCE W . G . , BELZILE N . , WISEMAN M . E . & WINTERHALDER K .

1994.

Composted organic wastes as anaerobic reducing covers for long term abandonment of acid-generating tailings. Proceedings of the 3rd ICARD Pittsburgh, Pennsylvania, Vol. 2, pp. 148-157. US Bureau of Mines Special Publication SP06A-94.

PLUMLEE G . S . , SMITH K . S . , MOSIER E . L . , FICKLIN W . H . , MONTOUR M . ,

BRIGGS P. & MEIER A. 1995. Geochemical processes controlling

acid-drainage generation and cyanide degradation at Summitville. In: Posey H. H., Pendleton J. A. & van Jyl D. eds. Proceedings, Summitville Forum '95, pp. 23-34. Colorado Geological Survey Special Publication 38. RITCHIE A. I. M. 1994. Sulfide oxidation mechanisms: controls and rates of oxygen transport. In: Jambor J. L. & Blowes D. W. eds. Short Course Handbook on Environmental Geochemistry of Sulfide Mine-wastes, pp. 201-245. MAC, Canada. RICHMOND T. C. 1995. Abandoned mine land reclamation costs in Wyoming. Sudbury '95 Conference on Mining and the Environment. Sudbury, Ontario, pp. 1221-1230. CANMET, Ottawa. ROE

P. A . ,

MULLIGAN

BELL

1996.

Environmental

Management of Coal Mines in the Bowen Basin, Central Queensland. In: Mulligan D. R. ed. Environmental Management in the Australian Minerals and Energy Industries, pp. 290-315. University of NSW Press, Sydney.

SATO M . & MOONEY H. M . 1 9 6 0 . T h e electrochemical m e c h a n i s m of

sulphide self potentials. Geophysics 25, 226-249. SCOTT K. M. 1987. Solid solutions in, and classification of, gossan derived members of the alunite-jarosite family, northwest Queensland, Australia. American Mineralogist 72, 178-187. SCOTT P. A . , EASTWOOD G . , JOHNSTONE G . & CARVILLE D .

1997.

Early

exploration and pre-feasibility drilling data for the prediction of acid mine drainage for waste rock. In: McLean R. W. & Bell L. C. eds. 3rd Australian Workshop on Acid Mine Drainage Darwin, NT, pp. 45-58. Australian Centre for Mining Environmental Research, Brisbane.

SOBEK A. A., BENEDETTI D . A. & RASTOGI V. 1 9 9 0 . S u c c e s s f u l r e c l a m a -

tion using controlled release bactericides: two case studies. In: Proceedings of the 1990 Mining Reclamation Conference, Charleston, West Virginia, pp. 33-41. SOBOLEWSKI A. 1997. The capacity of natural wetlands to ameliorate water quality: a review of case studies. Proceedings of the 4th ICARD Conference, Vancouver; BC, June 1997, pp. 1549-1564. STRUTHERS S. BRUMLEY J. & TAYLOR G. F. 1 9 9 7 . An integrated system for

recycling base metal mine tailings. In: Annual Meeting American Society for Surface Mining & Reclamation, Austin, Texas, pp. 579-592.

SUZUKI I., CHAN C. W . & TAKENCHI T. L. 1 9 9 4 . O x i d a t i o n of inorganic

sulfur compounds by Thiobacilli. In: Alpers C. N. & Blowes D. W. eds. Environmental geochemistry of sulfide oxidation, pp. 60-67. American Chemical Society Symposium Series 5 5 0 .

SWAYZE G . A . , SMITH K . S . , CLARK R . N . , SUTLEY S . J . , PEARSON R . M . , VANCE J . S . , HAGEMAN P. L . , BRIGGS P. H . , M E I E R A . L . , SINGLETON

M. J. & ROTH S. 2000. Using imaging spectroscopy to map acidic mine waste. Environmental Science and Technology 34, 47-54. TARLINGTON S. E. 1995. Management of AMD at Woodlawn Mines. In: Grundon N. J. & Bell L. C. eds. 2nd Australian Acid Mine Drainage Workshop, Charters Towers, pp. 75-82. Australian Centre for Mining Environmental Research, Brisbane. TASSE N . , GERMAIN D . , DUFOUR C . & TREMBLAY R . 1 9 9 7 . O r g a n i c - w a s t e

cover over the East Sullivan Mine Tailings: beyond the oxygen barrier. Proceedings of the 4 th ICARD Conference, Vancouver, BC, June 1997, pp. 1627-1642.

TAYLOR G. F. & THORNBER M . R. 1 9 9 2 . Gossan and ironstone surveys.

In: Butt C. R. M. & Zeigers H. eds. Regolith Exploration Geochemistry in Tropical and Subtropical Terrains, pp. 139-202. Elsevier, Amsterdam.

TAYLOR G . F., WILMSHURST J . R . , TOGASHI Y . & ANDREW A . S .

1984.

Geochemical and mineralogical haloes about the Elura Zn-Pb-Ag orebody, western New South Wales. Journal of Geochemical

Exploration

22, 265-290.

TAYLOR R. M. 1997. AMD remediation by induced hydrolysis reactions—the green precipitate process. In: McLean R. W. & Bell L. C. eds. 3rd Australian Acid Mine Drainage Workshop, Darwin NT, pp. 203-207. THORNBER M. J. 1992. The chemical mobility and transport of elements in the weathering environment. In: Butt C. R. M. & Zeigers H. eds. Regolith Exploration Geochemistry in Tropical and Subtropical Terrains, pp. 79-96. Elsevier, Amsterdam. THORNBER M . R. & TAYLOR G. F. 1 9 9 2 . T h e m e c h a n i s m s of sulphide

oxidation and gossan formation. In: Butt C. R. M. & Zeigers H. eds. Regolith Exploration Geochemistry in Tropical and Subtropical Terrains, pp. 119-138. Elsevier, Amsterdam.

TONGWAY D . , KEARNS A . , HINDLEY N . & BARNETT G . 1 9 9 7 . I n d i c a t o r s o f

ecosystem rehabilitation success and selection of demonstration sites. ACMRR Task 4.0 Final Report . Australian Centre for Minesite Rehabilitation Research, Brisbane.

TREDINNICK I. A. & CORNWELL P. J. 1 9 9 5 . A manager's operational per-

spective of AMD management at the Pajingo gold mine. In: Grundon N. J. & Bell L. C. eds. 2nd Australian Acid Mine Drainage Workshop, Charters Towers, pp. 43-51.

WARING C. & TAYLOR J. 1 9 9 9 . A new passive t e c h n i q u e for the pre-

vention of acid drainage from underground mines: GaRDs. In: 24th Annual MCA Environmental workshop, pp. 527-530. Minerals Council of Australia, Canberra.

WARREN P. & HICK P. 1 9 9 6 . Tracking our performance: airborne

remote sensing at Comalco's bauxite mine, Weipa, North Queensland. In: 21st Annual MCA Environmental workshop, pp. 67-78. Minerals Council of Australia, Canberra.

WAYBRANT K . R . , BLOWES D . W . , PTACEK C . J . & BENNER S . G . 1 9 9 7 . A

geochemical and mineralogical investigation of porous reactive walls for the prevention of acid mine drainage. Proceedings of the 4th ICARD Conference, Vancouver, BC, June 1997, pp. 1643-1658. WHITE J. 1996. Tactical research in support of the development of innovative apatite-type ceramics for the stabilisation of toxic metal wastes. In: Ian Wark Research Institute Annual Report p. 19. Ian Wark Research Institute, Adelaide. WILLIAMS D. J. 1996. Broadening the options—innovative tailings disposal: concepts and practices. In: 21st MCA Environmental Workshop, pp. 2 6 4 - 2 7 8 . Minerals Council of Australia, Canberra. WILSON W. 1997. Dry covers for mine tailings and waste rock. Short Course, 4th ICARD Conference, Vancouver, BC, June 1997. WOOD I. 1991. The Mount Lyell Mining and Railway Company Limited—an environmental case study. AMIC Environmental Workshop, Perth, Vol. 1, pp. 203-220. Australian Mineral Industry Council, Canberra. WUNDERLY M . D . , BLOWES D . W . , FRIND E . O . & PTACEK C . J .

1996.

Sulfide mineral oxidation and subsequent reactive transport of oxidation products in mine tailings impoundments: a numerical model. Water Resources Research 3 2 , 3173-3187.

Received 22 June 2000; accepted

16 November

2000

Geological Society of Australia Special Publication 21, 95-109

CHAPTER 9—Hydrogeology and geochemistry of sediment banks contaminated with mine tailings in the King River, Tasmania ffnrwirMiBimiw^

D. C. McPHAIL, D. GREEN AND W. C. HOOPER

Wk immmm, mm %. mm, m

Victorian Institute of Earth and Planetary Sciences, Department of Earth Sciences, Monash University, Vic. 38001 Australia. The King River in western Tasmania received approximately 100 Mt of sulfide-bearing tailings and slag from the Mt Lyell copper mine at Queenstown by the early 1990s. These tailings and slag now contaminate riverbanks, bottom sediment and delta sediment in the King River and Macquarie Harbour. The tailings in the sediment banks and delta, in particular pyrite, have oxidised and reacted to generate acid and potentially toxic concentrations of many elements. The impact of the contaminated sediment in one riverbank, Bank D, on water quality in the King River is estimated using a combination of groundwater and geochemical methods. Groundwater flow, velocity and discharge are estimated in sediment banks using piezometric measurements and the geochemistry of groundwater and sediment is measured using a combination of analytical methods. Mass loadings of acid and elements for both summer and winter 1997 are calculated using groundwater discharges and measured groundwater geochemistry. In most cases, groundwater flow is perpendicular to and towards the river, although in the upstream end of the bank and close to the river, flow can be away from the river. Estimated residence times of groundwater vary from <1 y to >20 y. Groundwater discharge from Bank D to the river is estimated to have been 8 m3/day in summer and 19 m3/day in winter. At most this represents only 0.04 vol% of the flow in the King River at the time. The groundwater in the bank was acidic (2.5 < pH < 6.6) and acidity increased towards the river. In winter, groundwater was up to 2pH units more acidic than in summer. Major elements in groundwater were dominated by iron and sulfate that result from the breakdown of pyrite, although high levels of aluminium (>100 mg/L) were measured in some samples. Many heavy metals were detected in groundwater samples where concentrations were up to an order of magnitude higher in winter than summer. The sediment consisted mainly of quartz and lithic fragments, with pyrite and other sulfide minerals making up to 1 vol%. In deeper parts of the bank, framboidal sulfide grains were observed, presumably the result of bacterial activity, and trapped up to more than 1000 ppm Co, Cu, Zn, Pb and Mn. Numerical modelling indicates that some dissolved elements (e.g. Fe, Si and possibly Ca, SO/ - , Mn and Al) are controlled by mineral solubility, whereas trace elements such as heavy metals may be controlled by adsorption. The overall mass loading of acid and metals from the groundwater and sediment in all riverbanks plus the delta of the King River, probably represents less than 1 wt% and 5 wt% of the total loading that results from the Mt Lyell minesite at Queenstown. KEY WORDS: acid drainage, contamination, geochemistry, groundwater, King River, mine tailings, sediment, sulfides, Tasmania.

INTRODUCTION River systems can be severely contaminated by mining activity, e.g. direct disposal of mine tailings, waste rock and mine waters into the river systems, waste from smelting or other processing methods and drainage from waste rock piles, tailings impoundments and slag heaps on mine sites. The Queen and King Rivers in western Tasmania have been severely contaminated by waste material released in the past from the Mt Lyell Copper Mine in Queenstown (Koehnken 1997). An estimated 97 Mt of mine tailings and 4.5 Mt of slag were discharged into the Queen and King Rivers between 1916 and 1994 (Locher 1997b). That material has been transported down the Queen River and now resides in bottom sediments and sediment banks of the King River and a large delta and bottom sediments in the Macquarie Harbour. Although the recent and current own-

ers of the mine impound the tailings, the legacy of the historical mining activity results in continuing contamination from the mine site. Acidic, metal-rich water is pumped out of the mine, partially treated and released into Haulage Creek, a tributary of the Queen River. Piles of waste rock and slag at the minesite are sources of acid and metals that are leached by rain and groundwater into the Queen River. As a result the Queen River is highly contaminated. For example, between December 1994 and April 1995, the Queen River downstream of the mine site had a pH of approximately 2.5 to 3.5 and median concentrations of 10 900 mg/L Cu, 12 500 mg/L Fe, 12 400 mg/L Mn and 26 200 mg/L Al, and this is after 3 to 7 times dilution of the element concentrations pumped directly from the mine (Koehnken 1997). A median value of approximately 2 t of dissolved copper per day enters the river system from the minesite, and up to 9 t of copper per day can enter the river system during storm events

D. C. McPhail

u\ MT LYEli [MINING COMPANY)

LAKE

John Butted PowerStation

iURBU!

MAINLAND AUSTRALIA •Melbourne

Queenstown

mceston

TASMANIA Figure 9.1 Location map of the catchment of the King River, Tasmania. For reference, Queenstown is shown in both inset and main map. Catchment is shown in white. Major roadways are shown in dark grey lines and labelled with letters and numbers, e.g., A10. White lines are the boundaries of World Heritage Areas. Modified from Taylor et al. (1996).

(Koehnken 1997). The Queen River and lower reaches of the King River have been devoid of aquatic life for decades and the receiving body of water, the Macquarie Harbour, is contaminated with mine tailings, acid and dissolved metals from the outflow of the King River. The environmental impact of past and present mining activity and strategies for remediating the environmental

problems were the subject of a major study sponsored, funded and directed by the Department of Environment and Land Management, Tasmania and the Office of the Supervising Scientist, Commonwealth of Australia. That study was the Mt Lyell Remediation, Research and Demonstration Program (MLRRDP) and consisted of 16 projects that focused on a wide range of topics. For exam-

Mine tailings, King River, Tas.

Figure 9.2 Location map for sediment banks on the King River, Tasmania. Banks are shown as stippled areas and labelled A-U. Dashed lines represent roads or tracks. King River catchment is shown in white. Modified from Taylor et al (1996).

pie, the projects consisted of reviews of existing information, scientific studies of water and sediment quality in the river and harbour systems, fluvial processes, biological surveys and revegetation and other remediation strategies. The results of the individual projects were published in reports of the Supervising Scientist (Australia) as well as a summary report by Koehnken (1997). The transport and environmental impact of mine tailings and slag that were released into the river system from the Mt Lyell mining operation have been the subject of several recent studies. Hie sediment transport of the tailings was studied as part of the MLRRDP and a PhD study by Helen Locher (Locher 1997a, b). The geochemistry, hydrogeology and environmental impact of mine-tailings in sediment banks and the delta of the King River have been studied as part of the MLRRDP (Taylor et al 1996) and university research projects (Green 1997; Hannan 1996; Hooper 1997). The main focus of this chapter is to highlight the results of the geochemical and hydrogeological studies, and show how mine tailings contained in sediment banks impact on the water quality of the King River system. In particular, the results are summarised for one sediment bank to show how detailed studies can help in understanding the spatial and seasonal geochemistry and hydrogeology of sediment banks contaminated with sulfide-rich mine tailings.

SETTING OF THE KING RIVER CATCHMENT The King River catchment has an area of 809 km2 and is situated in western Tasmania (Figure 9.1). The Mt Lyell mine is in the Queen River subcatchment (area of 73 km2), the largest subcatchment in the King River catchment. Mine tailings, slag and mine water were pumped into Haulage Creek, a small tributary of the Queen River. Although tailings are now pumped into an impoundment and mine water is treated by the present mine operators, acidic and metalrich mine water is still released into Haulage Creek. The Queen River flows through the town of Queenstown and drains into the King River, which then drains into the north-

ern end of the Macquarie Harbour, a stratified body of water with a surface area of 276 km2 (Figure 9.1). The climate of the area is temperate with average temperatures of approximately 8-22°C in summer and 2-13°C in winter (Locher 1997b). The annual rainfall ranges up to 2600 mm at the mine site to 1800 mm near the mouth of the King River (Taylor et al 1996). Precipitation results from westerly winds that bring moisture picked up from the Indian Ocean and rise when they encounter the mountains of the West Coast Range of Tasmania. Much of the King River catchment is densely forested (blackwood, myrtle, King Billy Pine, and along the rivers Huon Pine), although buttongrass plains are also present in upper reaches of the catchment (Locher 1997b). The mine site was deforested during historical mining operation, a result of timber cutting for fuel, acid rain from pyritic smelting, loss of topsoil during heavy rain and frequent bushfires (Locher 1997b). Soils consist mainly of yellow podzols common to areas underlain by siliceous rocks. The Mt Lyell copper deposit is hosted in the Cambrian Mt Read Volcanics (Corbett 1992 ). The geology, structure, stratigraphy, mineralogy and geochemistry of the deposit have been well studied (Walshe & Solomon 1981; Taylor et al 1996). The deposit consists mainly of massive pyrite + chalcopyrite and chalcopyrite + bornite ore that contains up to 50 vol% pyrite, 15 vol% sphalerite, 2 vol% galena and many other ore minerals (e.g. chalcocite, covellite, digenite, enargite, molybdenite, tetrahedrite-tennantite and gold). Gangue and alteration minerals consist of quartz, barite, sericite and chlorite. The geomorphology, hydrology and sediment transport and storage of the lower King River has been studied in detail by Locher (1997a, b). The average flow rate in the Queen River (below the Lynchford camp and ~ 4 km upstream of the King River) varied between 3.99 and 6.98 m3/s between 1988 and 1994 (Locher 1997b) and the maximum recorded rate was 107.6 m3/s. In the King River (below the Queen River confluence) the average flow rate varied between 44.4 and 61.7 m3/s between 1992 and 1994 (Locher 1997b), with a maximum flow rate of 316 m3/s in 1994. The water flow in the lower King River is now con-

D. C. McPhail

300

% n

200

D 100

+FM

>E3 •fif

Distance (m) Figure 9.3 Bank D of the King River, Tasmania (see Figure 9.2 for location). Piezometer locations are marked with symbols and labelled with letter and number. Piezometer nests are labelled with more than one number (e.g. C2/3). Solid lines in the western part of the bank are ephemeral streams. WREB, West River East Bank; EREB, East River East Bank; ERWB, East River West Bank. Lightly shaded areas north of the road and south of the King River are valley sides.

trolled by the John Butters Power Station, which is situated above the confluence of the King and Queen Rivers (Figure 9.1) and was commissioned in 1992. There are fewer and less extreme flood events now, and that has affected the sediment transport in the King River (Locher 1997b). Mine tailings and slag are present in bottom sediments and sediment banks of the King River, although approximately 96 vol% of the tailings and 19 vol% of the slag have been flushed through the King River to its delta and the Macquarie Harbour (Locher 1997b). Along the King River there are 21 sediment banks (Figure 9.2) that contained an estimated 3.4 Mt of tailings during 1993 to 1997 (Locher 1997b). The bottom sediments in the lower 8 km of the King River contained up to 10 Mt of tailings, resulting in the river bed being raised by up to 9 m (Locher 1997b). Upstream the sediment banks in the King River are mounded levee banks up to 6 m above the average river level. They range in length from 250 to 2820 m and in width from 10 to 140 m, and the estimated depth of tailings is between 1.25 and 4.5 m (Locher 1997b). The morphology of the banks changes downstream, where they become flatter and lower (approximately 1-3 m above average river level). The hydrogeology and geochemistry of four banks (D, H, N and R: Figure 9.2), two of each bank type, and the King River delta were studied in 1996 (Taylor et al 1996 ). Subsequently, one bank (Bank D: Figures 9.2, 9.3) was studied in detail to learn more about the geochemical and hydrogeological variability in the banks and seasonal variability in groundwater flow and geochemistry (Green 1997; Hooper 1997; S. Baker, D. Green, W. C. Hooper & D. C. McPhail unpubl. data). Bank D is a low, flat-topped bank -2.5 km upstream of the mouth of theriver.It is -800 m long and varies in width from <15 m at the ends to -100 m in the middle of the bank. It rises -1.5-2 m above average river level and the depth of mine tailings is on average - 3 m (Green 1997).

METHODS A combination of hydrogeological and geochemical methods was used to study the impact the sediment banks on the water quality of the King River (Green 1997; Hooper 1997; Taylor et al 1996 ). In addition, geophysical methods have also been used to study the water distribution and contamination in Bank D (S. Baker, W. C. Hooper & D. C. McPhail unpubl. data). Hydrogeological methods allow the estimation of groundwater flow and discharge within sediment banks and geochemical methods allow the geochemistry of groundwater and sediment to be measured. The combination allows the prediction the mass loadings of acid and elements that are entering or leaving the river system from sediment banks. Detailed geochemical studies result in a better understanding of atmosphere-water-sediment interaction that is especially important in aciddrainage environments. As part of the studies highlighted in this chapter, fieldwork was conducted on Bank D three separate times: July-August 1995 (winter: Taylor et al 1996 ) and February 1997 (summer) and June 1997 (winter: Green 1997; Hooper 1997; S. Baker, D. Green, W. C. Hooper & D. C. McPhail unpubl. data).

Hydrogeology A total of 33 piezometers were installed in nine lines approximately perpendicular to the river (Figure 9.3). The piezometers consisted of 50 mm PVC pipe that had a slotted interval cut into the bottom 20 cm of each pipe. The slotted interval was covered by a sleeve of 100 pm nylon mesh and secured with tape at the top of the sleeve and a nylon plug at the bottom of the pipe. Holes were handaugered into the sediment bank and the piezometers were

Mine tailings, King River, Tas.

inserted once the holes penetrated the water-table by at least several tens of centimetres. Because the sediments are unconsolidated they collapsed around the bottom of each piezometer, preventing any packing and sealing of the screened interval of the piezometers. The holes near the top of the piezometers were packed using sediment from the augered hole. In some areas of the bank, nests of piezometers were installed to measure the water levels and quality as a function of depth. Piezometer depths varied between approximately 1 and 4 m (Table 9.1). During installation the lengths of the piezometers were recorded and after installation, the relative elevations of were surveyed to approximately 1 cm accuracy. Table 9 . 1 Depths of piezometers installed in Bank D of the King River, Tasmania (Hooper 1997). Piezometer Al B1 CI C2 C3 D1 D2 D3 El E2 E3 E4 E5 F1 F2 F3 F4 F5 F6 F7 G1 G2 G3 G4 HI H2 H3 11 WREB1 WREB2 WREB3 EREB ERWB

Ground elevation (cm) 111.0 144.1 128.6 148.9 145.1 86.4 126.5 128.0 114.2 132.2 144.6 127.7 121.7 105.2 131.2 107.4 113.9 105.0 114.8 133.1 67.5 134.3 164.3 170.9 82.2 128.9 96.1 99.6 150.6 149.7 106.3 140.0 127.0

Depth (cm) 213.7 204.8 168.8 234.4 279.7 242.9 161.3 92.0 259.7 121.2 145.9 272.7 341.0 206.7 124.0 87.4 325.7 225.9 355.0 285.0 228.1 212.1 122.4 366.1 212.2 118.4 236.4 125.9 238.0 117.3 220.1 255.3 266.3

Locations of piezometers are shown in Figure 9.3. Ground elevation is relative to the average river level in February and June 1997.

Depth to the water level was measured in each piezometer using an electric water-level tape. Using the surveyed elevations of the piezometer tops, the hydraulic head was calculated from the raw measurements for both summer and winter. Hydraulic conductivities were measured in the field and the laboratory. Field measurements were made

using rising-head tests, where the water was bailed out using disposable nylon bailers and the subsequent rise in the water level in the piezometer was measured periodically. The results were used to calculate hydraulic conductivities by the Hvorslev method (Hvorslev 1951; Fetter 1994). Hydraulic conductivities were also measured in the laboratory using a constant-head permeameter (Fetter 1994); however, they were approximately an order of magnitude higher than those measured in the field. The field measurements are preferred because it is likely that the samples used in the permeameter were disturbed during sampling and transport (e.g. vibrations could have broken up grains or cemented clumps: Hooper 1997). Grainsize and porosity of sediment were measured for core samples taken near eight of the nine series of piezometers installed in Bank D (Figure 9.3), the exception being the upstream end of the bank (Series I). Grainsize was measured using a combination of sieving and a laser dispersion technique for grainsize less than 80 pm. Porosity was measured by saturating segments of the core samples with water and then drying the samples to constant weight. The porosity was calculated from the known initial volume and mass of sediment plus water, the dry weight, and assuming the density of water was 1.0 g/cm3.

Geochemistry Sediments from the bottom of all the augered holes were sampled in order to study sediment and groundwater samples that were from the same areas and depths in the bank. The mineralogy of the samples was determined using conventional petrographic, electron microprobe and X-ray diffraction techniques. Partial geochemistry of the sediments was measured by digesting samples in hot nitric acid and measuring Fe, Mn, Al, Cu, Zn, Co and Pb. Selected samples were also subjected to sequential extraction (Tessier et al 1979) to measure the speciation of trace metals in the sediments. The sequential extraction steps were (Green 1997): (i) exchangeable or leachable (1.0M MgCl2); (ii) acid soluble [1M NaOAc at pH = 5 (where OAc = acetate)]; (iii) reducible (0.04 M NH2OH.HCl + 25 vol% HOAc); (iv) oxidisable (0.02M HN0 3 + 30 vol% H 2 0 2 at pH = 2 and 85°C, followed by 30 vol% H 2 0 2 at 85°C and then 3.2 M NH4OAC + 20 vol% HN0 3 ); and (v) immobile (concentrated HNOs at 85°C). Element concentrations in filtered and acidified extractions were analysed by Atomic Absorption Spectroscopy (AAS: major and minor elements) or High Resolution Inductively Coupled Plasma Mass Spectrometry (HR-ICP-MS: trace elements). Groundwater was sampled in summer and winter in all piezometers after purging at least three times. As quickly as possible, water samples were measured in the field for temperature, pH, Eh, Electrical Conductivity (EC) and then filtered through 0.45 pm cellulose acetate filters. Several preparation steps were subsequently taken. Parts of the samples were acidified with 1M nitric acid for element analysis in the laboratory. Chemicals were added to separate parts of the sample to measure ferrous iron (orthophenanthroline), sulfide (Zn-acetate) or sulfate (barium chloride). In addition, chloride concentrations were measured using an ion-specific electrode. All samples were stored in acid-washed HDPE bottles and transported to the

100

D. C. McPhail

300-

Winter 1997

8C 2 0 0 CO

100-

Distance (m) Figure 9.4 Hydraulic head contours in summer and winter for Bank D of the King River, Tasmania. Contours are in 10 cm intervals relative to the average river level during times of hydraulic head measurement. The locations of piezometers used to calculate contours are marked with symbol and labelled with letter and number. Lightly shaded areas to the north of the road and south of the river are valley sides.

laboratory in closed (i.e. dark) and ice-packed containers. Element concentrations in the filtered and acidified samples were measured using the same techniques used for sediment analysis plus UV-Vis spectrophotometry for dissolved silica.

RESULTS AND DISCUSSION Hydrogeology The groundwater flow varied between summer and winter. In general the groundwater levels (i.e. hydraulic heads) were higher in Bank D compared with the river level at the

times of measurement, indicating that groundwater was flowing to the river from the bank. The same was true for three other sediment banks (H, N and R: Figure 9.2) that were studied by Taylor et al (1996). In Bank D, the hydraulic heads were - 1 0 - 5 0 cm higher in winter (June 1997), probably because of increased rainfall. The river level at Bank D, surveyed during the field trips in February and June 1997, varied by up to 20 cm, mainly following rainfall events but also because of the power needs of the John Butters Power Station upstream. Contours of hydraulic head were calculated using the values from shallow piezometers, boundary conditions at the river's edge (average river level) and north of the bank (ponded water beside Lowana Road: Figure 9.4) and inter-

Mine tailings, King River, Tas. polation methods (kriging option in the computer program SURFER). The two ephemeral creeks shown in Figure 9.3 are omitted for simplicity. Contours are shown for summer and winter (Figure 9.4). Most of the contours are subparallel to the river's edge and hydraulic head decreases towards the river. During summer the measured water levels in some piezometers near the river's edge were lower than the river level. Those piezometers are in the upstream end of Bank D (lines F, G and H). This indicates that during some times of year (lower rainfall and/or increased river level) that Bank D groundwater could be recharged by river water. Not only does this have an effect on the groundwater flow in the bank, but it could also affect water and sediment compositions in the bank. For example, river water probably contains higher dissolved oxygen levels and different element concentrations than the groundwater. Increasing the oxygen levels in the sediment bank may increase acid generation in the bank and element concentrations in groundwater could be concentrated or diluted, depending on the relative concentrations of the river water. The changing geochemical conditions could also result in the dissolution or precipitation of minerals, e.g. oxygenated waters could oxidise and dissolve sulfide minerals and result in the deposition of iron oxyhydroxide minerals. Groundwater discharge and velocity in the saturated zone in the bank can be estimated using Darcy's law, once the hydraulic conductivity, hydraulic gradients and porosity are known. Darcy's law is used to calculate specific discharge (Fetter 1994): v=

= -K— dl

where v is the specific discharge (m/s), Q J s the total discharge (m 3 /s), A is the cross-sectional area of groundwater flow (m 2 ), K is the hydraulic conductivity (m/s) and dh/dl is the hydraulic gradient (m/m; dimensionless). The average linear velocity is calculated by dividing the specific discharge (v) by the porosity expressed as a fraction. The values of hydraulic conductivity calculated from rising-head field tests range from 8.0 x 10~8 to 9.6 x 1 0 - 6 m/s (Hooper 1997), with an average of 2.6 x 1(T 6 m/s (19 measurements). These values are consistent with those of silty to fine sands (Fetter 1994), as are observed in the sediments. There are no clear spatial trends in the hydraulic conductivity in Bank D, although values generally increased from the upstream end of the bank to the downstream end. Hydraulic gradients vary throughout Bank D and are estimated for areas between the measured hydraulic head values at the bases of the piezometers. Approximately horizontal gradients are estimated between the shallowest piezometers and vertical gradients are estimated from piezometer nests. In general, approximately horizontal hydraulic gradients increase as the river edge is approached, as is seen by the closer spacing of the hydraulic head contours nearer the river (Figure 9.4). Hydraulic gradients perpendicular and closest to the river vary between approximately 0.007 and 0.022 towards the river in the summer, although in some parts of the bank (F, G and H series: Figure 9.4) the hydraulic gradients point away from the river, reflecting the recharge mentioned above. In winter, the approximately horizontal gradients

101

steepen to between approximately 0.015 and 0.060 and at the time of measurement (June 1997) all gradients pointed towards the river. Vertical gradients were measured in several piezometer nests: C2 and C3; E2, E4 and E6; F l , F5 and F7; F2 and F6; F3 and F4; G3 and G4; HI and H3 (Figure 9.3). Away from the river, vertical gradients were downward in both summer and winter, and greater in winter. Near the river (Fl, F5 and F7; HI and H3), the vertical gradients were near zero in both winter and summer in the F-line nest and changed from downwards in winter and slightly upwards in summer in the H-line nest. This reflects the decreased hydraulic heads in summer. Measured porosity values vary between 40 and 54% and in general increase towards the upstream end of Bank D. The average grainsize of sediments ranged between 0.05 and 0.28 mm and in general decrease towards the upstream end of Bank D. In most cases, the samples were poorly sorted according to the calculated standard deviations of grainsize (a = 1.00 to 2.00 (f>) and classification scheme of Folk (1980). Groundwater discharge to the river is calculated using the hydraulic conductivities, hydraulic gradients and Darcy's law. Summer and winter discharges are calculated by subdividing Bank D into sections corresponding to the piezometer lines (boundaries between sections are the midlines). The hydraulic conductivity for each piezometer series is an average of the measured values in that series and the summer and winter hydraulic gradients are calculated between the river's edge and the second row of piezometers along the bank. Flow is assumed to be horizontal only and the depth of the cross-sectional area of flow is assumed to be 4 m. The depth estimate is based on measured river channel depths at points above and below Bank D in 1988 and 1994 (Locher 1997b). The results are listed in Table 9.2. The total groundwater discharge into the river from Bank D during the times of our measurements is estimated to be 8 m3/day during summer (February 1997) and 19 m 3 /day during winter (June 1997), reflecting the higher hydraulic gradients in the bank during winter. Taylor et al (1996) estimated the total groundwater discharge from Bank D, based on three piezometer measurements made in the winter of 1996, to be 93 m3/day. This is approximately five times higher than the estimate of 19 m 3 /day for winter 1997 in this study. Although there could have been a higher discharge in 1996, much of the difference is more likely attributable to the lower hydraulic conductivity found in this study. They assumed a value of K = 1 xlO - 5 m/s, based on one rising-head test, whereas the average value found in this study was approximately three times lower K = 2.9 x 10" 6 m/s. In order to evaluate the effect groundwater discharge has on the river water, it is necessary to know the river flow. Accurate river flow rates are unknown for February and June 1997. The mean annual flow of the King River (below the confluence of the Queen River) between 1924 and 1984 and after the commissioning of the John Butters Power Station in 1993 and 1994 was approximately 5 x 10 6 m3/day (Locher 1997b). The groundwater discharge from Bank D is therefore only a small fraction of the mean river flow (e.g. 0.004%). Even during minimum river flow (e.g. 5 x 10 4 m3/day in 1994: Locher 1997b) it is still only 0.04 vol% at the maximum estimated groundwater discharge rate of 18.77 m3/day.

102

D. C. McPhail

Table 9.2 Calculated groundwater discharge from Bank D into the King River, Tasmania. Hydraulic gradient (m/m) Winter0 Summerc

Piezometer series

Section widtha'b (m)

Hydraulic conductivity13 (m/s)

A B C D E F G H I Total

109 40 55 85 90 105 110 95 145

8.52 x 10" 7 9.56 x 10" 6 8.55 x 10" 6 2.42 x 10" 6 4.93 x 10" 7 2.29 x 10" 6 7.94 x 10" 7 6.07 x 10" 7 3.06 x 10" 7

0.0134 0.0137 0.0124 0.0074 0.0125 0.0207 0.0087 0.0184 0.0221

0.0602 0.0381 0.0300 0.0154 0.0237 0.0481 0.0185 0.0208 0.0326

Discharge (m3/day) Summer Winter 0.43 1.81 2.02 0.53 0.19 1.72 0.26 0.37 0.34 7.67

1.93 5.04 4.88 1.09 0.36 4.00 0.56 0.41 0.50 18.77

Depth of cross-sectional area of groundwater flow is assumed to be 4 m. Section widths, hydraulic conductivities and hydraulic conductivities are from Hooper (1997). c Summer refers to February 1997 and Winter refers to June 1997 a

The groundwater velocity and residence time in Bank D are estimated by dividing the specific discharge (v) by the measured porosity. Assuming a typical porosity in Bank D of 0.45, groundwater velocity varies between approximately 0.4 m/y (Series E in summer) and 25 m/y (Series B in winter). The subsequent residence times for groundwater in the bank range from less than a year for the fastest velocities and narrowest part of the bank to - 2 6 y at the slowest velocities and widest parts of the bank. These estimates are based on hydraulic gradients calculated between piezometers and may represent only the approximate horizontal component of an overall hydraulic gradient. If the flow paths of groundwater were known, more accurate estimates of flow velocity and discharge could be calculated. Ideally the full three-dimensional flow system in Bank D should be known, so that the flow paths would be accurately known. There are not enough data from the 33 piezometers to describe three-dimensional flow, but the flow system could be calculated mathematically (e.g. by finite-difference or finite-element methods). At present it is not possible to calculate the flow system accurately because of the heterogeneous nature of the sediment layers (e.g. organic-rich, highly permeable layers observed in trenches, layers and sections of different grainsize, porosity and hydraulic conductivity). In addition we do not know with certainty the locations and types of boundary conditions below the bank and between the bank and the valley side. Accurate calculations of total discharge are difficult because the cross-sectional area of discharge is unknown, the hydraulic gradients and hydraulic conductivity are variable throughout the bank and it is likely the groundwater flow is transient (because of changes in river level and flood and/or rainfall events). In summary, the groundwater flow is variable throughout the bank. The estimated discharge of groundwater from Bank D is less than 0.05% of the river flow and up to two or three times higher in winter than in summer. By combining the geochemistry of the groundwater it is possible to calculate the mass loading of acid and elements to the King River from Bank D. The next section reviews the geochemistry of the groundwater and sediment, followed by calculations of the mass loadings and environmental impact of Bank D on the river water.

Groundwater geochemistry The geochemistry of the groundwater in Bank D was measured to study its variability within the bank and between summer and winter seasons. The mineralogy and geochemistry of the sediment were identified and measured to understand more about the interactions between groundwater and sediment in Bank D. The measured pH, Eh and major element concentrations in groundwater samples are presented in Tables 9.3 and 9.4. The temperature of groundwater was approximately 10°C during both field seasons. The pH was between 2.46 and 6.61 during the summer season and 2.44 and 6.02 in the winter season. In general, the groundwater was more acidic during winter, where the pH of samples taken from the same piezometers was up to 2 units lower than in summer, although the changes were variable and some waters were slightly less acidic in winter. The acidity of the groundwater increased towards the river. The average pH of summer samples was 4.5 in piezometers nearest the river, 4.8 in the second row of piezometers and 5.5 in the third row of piezometers farthest from the river. The corresponding winter values were 4.0, 4.6 and 5.3. The pH also changed with depth, although there are no clear trends. In some cases, pH increased with increasing depth and in others pH decreased with increasing depth. In one case, piezometers nearest the river in the E-series (Figure 9.2), the pH decreased and then increased with increasing depth. The variability reflects the transient nature of groundwater flow, heterogeneity of the bank sediments and possibly different bacterial species and populations. The oxidation potential of the groundwater samples, measured as Eh, was also variable within the bank, with values ranging from approximately 55 mV to 6 8 0 mV. There are no clear trends in Eh; however, in general more reducing conditions (lower Eh) were encountered with increasing depth (>2 m). In some cases, conditions were reducing enough to stabilise hydrogen sulfide (H 2 S), which was detected by smell and gravimetric analyses of waters in some piezometers. The presence of hydrogen sulfide was probably because of bacterial activity (sulfate-reducing bacteria) and resulted in the precipitation of iron sulfide

Mine tailings, King Rivet Tas.

103

Table 9.3 pH, Eh and major element concentrations in groundwater sampled in summer (February 1997) in Bank D of the King River, Tasmania (Green 1997). Piezometer

Eh (mV)

Cl-

so42-

A1 B1 CI C2 C3 D1 D2 D3 El E2 E3 E4 F1 F2 F3 F4 F5 G1 G2 G3 HI H2 H3 11

6.14 6.61 4.2 5.68 5.41 4.13 5.81 4.88 4.45 4.71 6.22 4.47 2.46 4.19 5.00 5.63 2.57 4.28 3.62 5.94 4.88 5.25 4.04 5.3

103 41 284 222 171 312 194 197 77 171 268 332 561 113 391 190 638 62 285 72 110 149 219 199

na 8.5 7.4 10.1 12.0 8.4 11.7 9.4 13.3 12.5 13.1 14.5 7.7 13.2 10.2 17.2 8.7 7.6 13.3 13.4 9.6 12.9 9.5 14.3

1.5 1.6 1.5 2.2 2.5 2.6 2.6 1.1 3.5 1.6 2.3 4.2 1.4 2.1 1.0 1.3 1.2 1.7 1.2 1.7 2.3 1.0 1.4 0.8

17.5 31.8 30.8 52.6 51.7 166.6 191.9 56.2 113.6 85.6 110.7 263.9 31.0 84.5 10.9 12.7 30.0 67.9 47.8 12.6 54.3 10.0 67.4 8.3

17.5 28.2 21.4 53.2 55.3 138.5 155.9 44.1 80.0 55.4 109.8 191.7 22.4 82.3 5.9 4.0 20.7 46.0 47.2 7.5 45.7 5.1 44.5 3.2

88.1 384.8 207.7 189.5 83.9 2141 1681 473.0 517.2 328.1 506.0 532.8 65.9 1002 43.6 5.1 88.6 186.1 660.9 52.3 408.8 53.0 300.9 36.4

3.6 8.1 4.6 14.0 5.8 52.9 85.1 12.5 23.7 23.0 57.7 87.0 4.1 32.0 1.1 0.4 3.7 12.7 7.4 2.8 15.8 1.0 13.2 0.5

<0.1 0.9 4.2 <0.1 <0.1 91.8 <0.1 2.2 8.5 4.2 <0.1 <0.1 19.1 13.3 0.4 <0.1 19.1 5.3 66.8 <0.1 4.0 <0.1 8.5 <0.1

21.0 64.5 81.5 29.1 12.9 60.8 19.2 104.3 71.2 76.7 67.8 62.8 79.9 48.6 25.1 8.6 88.3 66.2 61.4 20.8 66.0 37.9 71.8 31.5

49.6 39.0 24.1 63.8 46.1 31.9 46.1 34.7 42.5 39.0 113.4 39.0 6.0 39.0 42.5 39.0 5.7 na 35.5 88.6 na 46.1 19.9 53.2

235.9 920.2 571.4 689.8 495.4 5460 2312 1112 1628 1006 1831 2617 628.6 2403.9 122.3 29.8 640.8 746.2 1824 40.4 1083 128.2 929.2 103.5

na, not available. Element concentrations are in mg/L. Temperature of samples was approximately 10°C. Table 9.4 pH, Eh and major element concentrations in groundwater sampled in winter (June 1997) in Bank D of the King River, Tasmania (Green 1997). Piezometer

Eh (mV)

CI-

S042"

A1 B1 CI C2 C3 D1 D2 D3 El E2 E3 E4 F1 F2 F3 F4 F5 G1 G2 G3 HI H2 H3 11

5.42 4.23 3.38 3.97 5.21 4.26 5.43 3.62 4.17 3.63 6.02 3.37 2.47 3.63 4.43 5.95 2.44 4.94 4.33 6.32 4.65 5.40 3.92 4.19

161 315 558 402 214 142 244 507 120 494 273 589 670 127 207 144 681 57 147 152 157 137 147 192

11.8 7.4 8.5 12.5 11.7 8.1 9.6 8.6 11.7 10.8 11.0 13.2 10.8 13.1 13.4 16.0 10.0 6.9 9.9 14.0 9.4 12.9 9.2 13.2

1.7 1.4 1.7 1.9 2.1 2.4 2.4 1.1 3.6 0.6 2.0 4.7 1.7 1.8 0.9 1.4 1.5 1.7 0.8 1.3 2.6 1.0 1.6 1.0

18.6 43.8 64.1 56.7 45.1 152.3 171.1 86.5 101.3 71.2 90.0 262.2 55.5 98.9 10.5 7.9 52.0 59.7 24.6 28.2 97.6 9.4 70.4 6.8

17.4 33.5 40.1 42.6 40.3 128.1 142.4 52.6 67.7 65.4 95.5 181.6 46.3 95.1 7.1 2.8 48.6 35.5 28.3 13.4 79.3 6.4 47.0 3.8

66.2 402.6 292.6 126.8 67.5 1985 1623 788.2 435.3 671.2 542.8 422.8 168.8 1215 61.8 6.0 167.3 180.2 384.7 64.1 748.0 58.6 265.7 35.7

4.2 8.7 8.8 12.3 4.9 44.4 73.8 11.9 21.6 27.8 51.7 88.4 10.3 35.7 1.4 0.3 10.5 10.8 4.5 5.3 36.1 1.3 13.8 2.5

<0.1 14.2 42.4 <0.1 <0.1 140.3 <0.1 76.7 3.1 43.8 <0.1 <0.1 64.9 100.4 <0.1 <0.1 73.5 1.8 39.7 <0.1 <0.1 <0.1 25.4 <0.1

9.2 27.0 40.3 9.5 7.4 49.7 15.3 41.9 45.6 33.1 9.1 12.2 46.9 44.4 8.3 5.1 48.6 40.0 41.2 14.3 35.5 18.8 43.6 10.0

32.2 4.6 3.4 7.8 7.4 6.4 10.7 3.7 9.5 7.4 22.8 5.5 5.6 33.3 20.9 57.5 1.2 32.1 27.7 69.1 14.8 65.7 54.1 26.7

228.5 1012 1052 543.6 377.0 4919 3845 2085 1334 1812 1557 2422 1261 3183 176.9 9.9 1294 637.0 1016 92.2 1852 132.5 1043 78.2

Element concentrations are in mg/L. Temperature of samples was approximately 10°C.

104

D, C. McPhail

minerals (see below). Eh values in samples taken in winter were generally higher but this may have been due to lower pH values. Major element concentrations were dominated by iron and sulfate in all groundwater samples, a result of the oxidation and dissolution of pyrite, e.g. FeS 2 (s) + 7/20 2 (g) + H2 0 = F e 2 + + HS 0 4 " + H + Iron concentrations varied from several mg/L to > 2 0 0 0 mg/L. No systematic trends were observed; in some parts of the bank iron concentrations were higher in winter and in other parts they were lower. Sulfate concentrations varied from several tens of mg/L to > 5 0 0 0 mg/L and as with iron there was no clear spatial or seasonal trends. The concentrations of other major elements varied in differing amounts. Dissolved sodium and potassium were approximately constant in all samples and between seasons. Calcium, magnesium and manganese varied by approximately an order of magnitude and did not change systematically between seasons. Aluminium varied from below detection limit (0.1 mg/L) to greater than 100 mg/L, and had consistently higher concentrations during winter. Dissolved silicon varied less than an order of magnitude and had consistently lower concentrations in winter. The variation in the concentrations of dissolved elements is a function of many possible geochemical variables (e.g. temperature, pH, Eh, ligand concentrations) and processes (e.g. mineral solubility, adsorption and fluid mixing). Preliminary interpretations are presented below after a summary of the observed mineralogy and sediment geochemistry. Many trace elements were detected in groundwater samples. Concentrations were variable and the elements reflect the nature of the mine tailings and the original ore material. Only a brief summary is given here—more complete details are available in Green (1997). In approximate order of decreasing concentration, the detected elements were Cu (0.001-15 mg/L), Co (0.009-2.3 mg/L), Zn (0.02-1 mg/L), Ba ( 1 7 - 7 7 pg/L), Ni ( 3 2 - 2 5 1 pg/L), Pb ( 0 . 2 - 1 0 8 pg/L), Cr (0.07-15 pg/L), As (1-9.1 p^/L), U (0.1-8 pg/L), Cd (0.4-2.3 pg/L), Sn (0.01-0.15 pg/L) and in a few samples traces of Sb and Ti. During the summer season, only samples from piezometers in lines C, D, E and F were analysed for trace elements, whereas for the winter season data are available for all piezometers. In most cases where both summer and winter results are available, the dissolved trace-element concentrations were higher in winter than summer, sometimes by nearly an order of magnitude (Green 1997).

Sediment mineralogy and geochemistry The sediment of Bank D is variable, where discontinuous layers are identified based on colour changes, amount of organic matter, hard-pan and mineralogy. The stratigraphy of Bank D is complicated and the thickness of individual sediment layers varies from 1 0 - 2 0 cm down to <1 cm. Mine tailings dominate the surface sediment as well as sediment samples taken during piezometer installation. In one piezometer, F7 (285 cm deep), natural sediment was observed. That sample contained coarse (>2 cm diameter), rounded sediment grains and no primary pyrite. It did,

however, contain framboidal iron sulfide grains that probably resulted from biogeochemical reactions in the sediment bank (see below). In the top 3 0 - 1 0 0 cm the sediment is coloured red (i.e. oxidised), below which the sediment is grey (i.e. relatively unoxidised). The depth of oxidised sediment correlates approximately to the depth of the vadose zone, although the level of the water-table varies by tens of centimetres seasonally. Thin layers of relatively unoxidised tailings are present between oxidised layers in the shallower parts of the bank and, conversely, oxidised layers between relatively unoxidised layers in deeper parts of the bank. Some of the latter oxidised layers create a hard-pan, where quartz grains, lithic fragments, and perhaps significantly, slag grains (up to 3 0 vol% of observed hard-pan layers), are cemented together with iron oxyhydroxide precipitate. In addition, there are organic-rich layers that can be highly permeable and in some deeper parts of the bank contain framboidal sulfide grains. The mineralogy of the sediment is dominated by quartz, although in oxidised zones at or near the surface of the bank many of the grains are coated by orange-red iron oxyhydroxide precipitate (goethite?) that resulted from pyrite oxidation and dissolution. The coating is either amorphous or present in amounts too small to identify crystalline material unequivocally using conventional powder X-ray diffraction techniques. The following ranges of modal abundances were estimated based on petrographic analysis (Green 1997 ). Quartz makes up 5 0 - 9 0 vol% of the sediment and fine-grained clay (?) minerals make up 5 - 3 0 vol%. The remaining non-opaque minerals consist of sericite + quartz lithic fragments ( 5 - 3 0 vol%), muscovite ( 1 - 7 vol%), chlorite ( 0 . 5 - 5 vol%) and traces of other felsic and volcanic minerals. Organic matter (identified in hand specimen) content is variable and can represent as much as 10 vol%. Sulfide minerals typically make up approximately 1 vol% of the sediment. Pyrite predominates and there are lesser amounts of chalcopyrite, bornite, sphalerite and galena. Magnetite and hematite were identified in many samples, but in amounts much less than 1 vol%. In deeper parts of the bank, framboidal iron sulfide can be up to 1 vol% of the sediment. In more oxidised parts of the bank, typically at or near the surface, primary sulfide minerals (e.g. pyrite, chalcopyrite) were more prevalent ( 1 - 1 0 vol%) than in other parts of the bank. The presence of framboidal iron sulfide grains may be significant in affecting acid generation and acid and metal transport, as the precipitation of iron sulfide could result in changes in acidity. For example, assuming hydrogen sulfide (detected in samples from deeper parts of the bank) and the initial precipitation of an iron monosulfide (observed in Bank D sediment: Wilkin & Barnes 1996 ), the following reaction may generate increased acidity: F e 2 + + H 2 S(aq) = 'FeS' (s) + 2H+ The framboidal sulfide grains may also affect metal transport. Approximately 9 0 individual grains were handpicked from multiple sediment samples, dissolved and analysed for metal content (Green 1997). The measured average metal concentrations were 1520 ppm cobalt, 760 ppm copper, 147 ppm zinc, 140 ppm lead and 120 ppm manganese. This may be important in trapping, or at least

105

Mine tailings, King River, Tas.

Table 9.5 Ranges of selected element concentrations and percentages of those elements in different sediment fractions in sediment samples from Bank D of the King River (Green 1997).

Fe Mn Al Cu Zn Co Pb

'Bulk' (ppm)

Exchangeable

Acid soluble

Reducible

Oxidisable

Immobile

430-29,000 1-247 306-11,000 25-437 7-118 3-49 8-74

1-6 38-62 2-5 2-49 16-70 73-100 0-100

0-4 1-9 0-3 5-25 0-7 0 0-12

11-31 0-8 4-16 0-16 0-16 0-7 0-20

2-37 0-15 6-20 6-76 0-9 0-13 0

41-79 19-60 54-83 5-36 27-72 0-14 0-83

(%)

The ranges are for seven samples (piezometers CI, Dl, D3, E l , E2, F1 and G2: Figure 9.3). Details of the analytical techniques are given in text.

attenuating, metals as they travel with groundwater through the sediment bank. The geochemical composition of the sediment was measured in several ways. Samples and elements were selected based on locations in the bank (e.g. proximity to the river and depth) and the results of the measured groundwater compositions (i.e. samples where elements were in excess of recommended guidelines for water quality). Seven sediment samples (CI, Dl, D3, El, E2, F1 and G2) were digested in two ways: (i) hot (85°C) nitric acid for three hours to determine 'bulk' concentrations of selected elements (Fe, Mn, Al, Cu, Zn, Co and Pb); and (ii) sequential extraction (Tessier et al 1979 ) for the same elements to determine speciation of those elements in the sediment. The results are listed in Table 9.5. There are wide ranges in 'bulk' concentrations and percentage fractions of the elements in different parts of the sediment and reflect the difficulty in obtaining accurate results and the heterogeneous nature of the sediment. Notably, large fractions of Mn, Zn, Co and Pb and in some cases Cu are in the 'teachable' (i.e. extracted with MgCl2) part of the sediment, indicating that these elements might be most easily leached from the sediment. The source of manganese and some of the iron is most likely the oxyhydroxide coatings on mineral grains, although carbonate minerals may also be an important source. The more dissolvable Al probably resides in finegrained clay minerals (e.g. gibbsite, kaolinite and alunite). Cu, Zn, Co and Pb are most likely present as adsorbed species on, or coprecipitates with, the iron (and/or manganese) oxyhydroxide coatings on mineral grains, although Zn and Pb are also present in primary sulfide minerals. Slag particles may contain high concentrations of chalcophile elements (e.g. Cu, Pb, Zn, Co), as metal-containing sulfide droplets were observed in slag grains.

Geochemical controls on groundwater geochemistry Groundwater composition in Bank D is likely to be affected predominantly by water-sediment interaction because of the fine-grained sediment (i.e. high surface area) and redox reactions with sulfide minerals. In addition, acid generation probably requires interaction with the atmosphere and the framboidal sulfide grains are likely to be a result of bacterial activity. In order to understand more clearly what effects water-sediment interaction has on groundwater composi-

tion, aqueous speciation and mineral saturation indices were calculated using MINTEQA2 (Allison et al 1991). The saturation index (SI) is the logarithm of the ion activity product divided by the solubility product for the relevant mineral (Drever 1997): CT

SI =

I.A.P \og-— sp

where the I.A.P. is a product of the calculated activities of ions in a mineral dissolution reaction and the Ksp is the equilibrium constant of the same reaction. The activities of the ions are calculated using measured water compositions (temperature, pH, Eh, element concentrations) and the computer program. The predicted saturation states of many minerals vary within only an order of magnitude; however, the saturation states of sulfide minerals are more variable than for other minerals and show trends with depth. Table 9.6 is a brief summary of the modelling results, where the listed minerals were either observed in the sediment, likely to be present in the fine-grained clay fraction or common in acid-drainage environments. Only typical saturation indices are listed. The minerals that are predicted to be supersaturated are mainly iron- and/or aluminium-bearing oxides, sulfates and silicates (Table 9.6), reflecting the high concentrations of Fe, Al and sulfate in the groundwater of Bank D (Tables 9.3, 9.4). The concentrations of some dissolved elements, especially major elements, may be controlled by mineral solubility. Iron oxyhydroxide minerals (e.g. goethite and lepidocrocite) may limit or control dissolved iron concentrations. The predicted saturation indices for crystalline goethite and lepidocrocite are approximately two log units supersaturated (Table 9.6); however, the coatings on mineral grains are probably amorphous (preliminary X-ray diffraction measurements: Green 1997) and have a higher solubility than their crystalline counterparts. Their saturation indices would therefore be closer to zero. Dissolved silica is probably limited and/or controlled by Si0 2 minerals (e.g. quartz, amorphous silica) in a similar way as their saturation indices are close to zero (Table 9.6). Dissolved aluminium concentrations are close to being in equilibrium with some Al-bearing minerals (e.g. gibbsite and muscovite have saturation indices close to zero) but other Al-bearing minerals are supersaturated (e.g. alunite, pyrophyllite, jarosite and nontronite minerals), sometimes by greater than

106

D. C. McPhail

Table 9 . 6 Typical calculated saturation indices of some minerals in groundwater samples of Bank D of the King River (Green 1997). Mineral Supersaturated minerals Nontronite (Na, K, Mg, Ca) Hematite Magnetite Jarosite (Na, K) Kaolinite Pyrophyllite Alunite Montmorillonite Lepidocrocite Goethite Muscovite Diaspore Quartz Saturated minerals Barite Cristobalite Chalcedony Halloysite Gypsum S i 0 2 (amorphous) Leonhardite Gibbsite Sulfide minerals Pyrite (vadose zone) Pyrite (saturated zone) Greigite Chalcopyrite (saturated zone) Covellite (saturated zone) Galena Sphalerite CoS MnS

Saturation Index

Notes

9-16 9 7 4-7 3 5.5 5 3 2.1 1.9 1.2 1.2 1.1

Possible clay minerals Observed Observed Likely in oxidised zones Possible clay mineral Possible clay mineral Likely in oxidised zones Possible clay mineral Possible iron oxyhydroxide mineral Probable iron oxyhydroxide mineral (observed?) Observed (muscovite, sericite) Possible clay mineral Observed

0.8 0.6 0.4 0.2 -0.1 -0.4 -0.5 -1

Possible sulfate mineral Potential fine-grained mineral Potential fine-grained mineral Potential fine-grained mineral Possible/likely Probably part of fine-grained material Potential fine-grained mineral Potential aluminium clay mineral

-10 3-20 -0.5 to - 2 11 8 -0.8 to-1.2 -3 -5 -12

15 orders of magnitude (Table 9.6). Controls on aluminium concentrations are complicated and it appears that the kinetics of precipitation reactions inhibits the precipitation of many of the possible aluminium minerals. Dissolved calcium and sulfate concentrations might be controlled by gypsum and/or barite solubility, although those minerals have not been observed in the samples from Bank D (barite has been observed in delta sediment: Hannan 1996 ). Dissolved sulfate concentrations could be affected by other minerals (e.g. jarosite, alunite, Fe-sulfates, Fe-sulfides), but although some of these minerals are found in acid-drainage environments, most have not been identified unequivocally in our studies. Trace elements (i.e. the heavy metals) may be affected or controlled predominantly by adsorption (Drever 1997) or coprecipitation with iron oxyhydroxides and/or sulfides. Copper may be an exception where the saturation indices of chalcopyrite and chalcocite indicate those minerals are up to 11 orders of magnitude supersaturated. There are many complicating factors that limit our present understanding of the sediment-water interaction in Bank D. Those factors include the heterogeneity in the sediment and groundwater mineralogy and compositions, uncertainties in the thermodynamic properties of minerals and aqueous species, the kinetics of dissolution and pre-

Observed Observed (primary, secondary) Possible in framboidal sulfide Observed Possible copper sulfide Observed Observed Possible in framboidal sulfide Possible in framboidal sulfide

cipitation reactions, the effects of adsorption and the effects of elements and species that have not yet been considered in our studies (e.g. carbon), the effects of gas (e.g. C0 2 , CH4, H 2 S) solubility in groundwater, and the effects of biological activity. In addition, the pathways that groundwater takes through the bank will affect how the groundwater and sediment compositions change. Chemical reactions (and water flow) in the unsaturated zone are obviously important as the degree of oxidation in the tailings is highest above the water-table. Environmental impact of sediment banks along the King River The environmental impact of sediment banks can be estimated by calculating the mass loading of elements to the river. The estimated discharge of groundwater (m3/day) is multiplied by the concentration of dissolved elements in the groundwater. The effect of groundwater acidity on the river water is difficult to estimate accurately because measured pH values probably do not represent the total acidity of an acid-mine water. Following Taylor et al (1996 ) the acid loading is calculated by using the measured pH and the measured iron (i.e. Fell) concentrations. The reason for

Mine tailings, King River, Tas.

including iron is because of the acid generation that could result from the oxidation of iron and precipitation of ferric iron minerals, e.g. Fe 2 + + 5/2 H 2 0(1) + 1/2 0 2 ( g ) = Fe(OH) 3 (s) + 2 H + Mass loadings were calculated for each piezometer series and summed to estimate the total mass loading from Bank D. Acid loadings are calculated as sulfuric acid equivalent, and all loadings are calculated based on measured compositions of groundwater sampled in piezometers closest to the river. More complete water geochemistry was measured in the samples taken during the winter season (June 1997), so the following results are calculated for that time. The mass loading of the most important contaminants are: 13 kg/day H 2 S 0 4 - equivalent; 7 k^day Fe; 0.7 kg/day Al; 68 g/day Cu; 14 g/day Zn; and 15 g/day Co. Mass loadings in the summer season (February 1997) would have been several times lower because the groundwater discharge was 2.4 times less and the pH and element concentrations approximately 20-50% lower in summer. The total estimated discharge to the King River from Bank D is small relative to the river flow (0.04%: calculated above), so the groundwater, and the dissolved acid and elements in it, will be diluted by approximately 2500 times on entering the river. Taylor et al. (1996 ) estimated the overall impact of the sediment banks and delta sediments on the water quality in the King River and Macquarie Harbour. They used a similar method as that outlined in this chapter, basing their calculations on results from three to five piezometers in each of four sediment banks (D, H, N and R: Figure 9.2) and 19 piezometers installed in the north and south lobes of the King River delta. They estimated a total discharge of groundwater into the river and harbour from 18 of the sediment banks (A-R: Figure 9.2) and the north and south lobes of the delta, assuming that the fluxes of groundwater to the river are the same in unstudied banks as they were in the four studied sediment banks. Their estimated total groundwater discharge was 2.3 x 10 3 m3/day, which represents 0.05 vol% of the mean annual flow of the King River. They calculated mass loadings from groundwater, using their estimated discharge and measured groundwater chemistry and assuming that there is an equal contribution to mass loading from surface runoff. The sum of the two sources represents - 1 - 5 wt% of the loading in the King River sourced upstream at the Mt Lyell mine site at Queenstown (Taylor et al. 1996). The contribution from mine tailings and slag particles in bottom sediments of the King River is unknown. They suggested that although the contribution from ground and surface water is small relative to the loads already in the river, acute events (e.g. flood, rainfall) may have an adverse impact on the river water quality by flushing higher concentrations of acid and elements out of the bank. In addition, their estimates are uncertain because of the variability of groundwater discharge and composition in sediment banks and in different seasons. The results of the subsequent studies summarised in this chapter show that the total groundwater discharge from a sediment bank can vary by several times between summer and winter and the groundwater composition can vary by orders of magnitude within the bank and by several times between summer and winter.

107

The mine tailings in the river system are likely to contaminate the river and harbour water for a long time. Contamination may continue for approximately 2000 years, based on an approximate pyrite concentration of 2 wt%, a copper concentration of 0.085 wt%, an estimated 2.73 Mt of tailings in the sediment banks and delta and an acid-producing potential from pyrite of 33 kg H 2 S0 4 -equivalent/t and current groundwater discharge rates (Taylor et al 1996). This estimate is uncertain because the transient effects of climate, groundwater flow and discharge, river flow and the geochemistry of the contaminated sediment. In addition, some banks show some signs of revegetation and there has been limited success in revegetation trials, both of which may reduce the generation of acid and mass loadings of acid and toxic elements to the river system.

SUMMARY AND CONCLUSIONS The hydrogeology and geochemistry of sediment banks on the King River, Tasmania have been studied in order to estimate the impact of mine tailings on the water quality in the King River and Macquarie Harbour. Hydrogeological studies allow the estimation of groundwater discharge into the river. By combining with measured groundwater geochemistry, mass loadings of acid and elements to the river system have been estimated. Geochemical studies of the groundwater and sediment in the bank also help in understanding water-sediment interaction, including acid generation and element leaching, transport and deposition. One particular sediment bank (Bank D) approximately 2.5 km upstream of the King River mouth has been studied in detail to gain an understanding of spatial and seasonal variability in groundwater flow and geochemistry and the interaction between sediment, atmosphere and water and the controls on groundwater composition in the bank. Several other banks and the delta were studied in similar way but in less detail. The significant outcomes of these studies are as follows. (1) Groundwater flow in Bank D is mainly perpendicular towards the river; however, some of the upstream end of the bank, particularly in summer, has flow into the bank from the river. Water levels, hydraulic gradients and groundwater flow are greater in winter. (2) Estimated residence times for groundwater in Bank D vary from less than a year to approximately 14 y, although flow in some parts of the bank be higher and residence times lower because of observed highly permeable and discontinuous sediment layers. (3) Total discharge of groundwater into the King River from Bank D is estimated to have been approximately 8 3 m /day during summer (February 1997) and 19 m 3 /day during winter (June 1997). The highest discharge probably represents at most 0.04 vol% of the mean annual river flow in the King River. (4) The geochemistry of the groundwater in Bank D was highly variable. Groundwater was acidic (2.4 < pH < 6.6) and increased in acidity closer to the river. The pH was up to two units more acidic during the winter. Groundwater and sediment were oxidised near the surface (approximately the top 10-50 cm) and in some parts of the bank reducing conditions (i.e. presence of hydrogen sulfide and biogenic sulfide grains) were encountered at depths greater

108

D. C. McPhail

than approximately 2 m. Element concentrations were dominated by ferrous iron and sulfate, mainly due to the oxidation and dissolution of pyrite present in the mine tailings. Most major element concentrations varied by up to an order of magnitude and there was no systematic change between summer and winter. Aluminium concentrations were higher in winter than summer and silicon concentrations were lower in winter. Trace-element concentrations varied by several orders of magnitude and were generally higher in winter than in summer. (5) Sediment consists of fine-grained mine tailings, natural sediment and organic matter. The mineralogy of the sediment is dominated by quartz, with lesser amounts of lithic fragments, muscovite, chlorite, iron oxides and sulfide minerals (pyrite, chalcopyrite, bornite, sphalerite and galena). Framboidal sulfide (biogenic) grains were present in deeper parts of the bank. (6) Some dissolved element concentrations (e.g. Si, Fe, Ca, S 0 4 2 - and possibly Al and Cu) appear to be controlled by mineral solubility. Other elements may be controlled by adsorption and/or coprecipitation with iron oxyhydroxide coatings and framboidal iron sulfide minerals. (7) The impact of the groundwater discharge on presentday water quality in the King River and Macquarie Harbour is small—most of the acid and metal load in the river is sourced from the Mt Lyell mine site. The estimated groundwater discharge from all sediment banks and delta sediments is only 0.05 vol% of the river flow. The groundwater is contaminated with acid and potentially toxic elements, but the estimated mass loadings are on the order of 1-5 wt% of the load in the river. (8) There appears to be a natural attenuation process for metal transport in Bank D, where the precipitation of iron sulfide minerals, presumably resulting from bacterial sulfate reduction, traps heavy metals in deeper parts of the bank. It is difficult to estimate accurately the magnitude of environmental impact of contaminated sediment banks on river water quality. The heterogeneous nature of riverbanks and the transient flow makes it difficult to estimate flow velocity and discharge accurately. The geochemistry of sediment is also heterogeneous, especially because of the variation in oxidation potential (oxidised at surface, reducing at depth). Although the sediment geochemistry may not vary as much in time, the groundwater geochemistry varies in both space and time, where depending on the element their concentrations can be higher or lower in different seasons. The mass loading of acid and contaminants from a sediment bank into a river could be estimated by measuring water chemistry above and below the bank, but differences would probably be undetectable in cases like the King River, where the river flow is orders of magnitude higher than discharge rates from the bank. Finally, the environmental impact of sediment banks on river quality may not result entirely from groundwater discharge. Overland flow or aeolian transport may also carry contaminants into the river. Future research into problems should include sampling more often to understand the transient nature of groundwater flow and composition. The water flow and geochemistry in the unsaturated zone should be studied. Reactive transport modelling of sediment-water interaction will provide insight into how groundwater compositions change

along the flow path through contaminated sediment. Measuring and studying organic carbon compounds (solid and aqueous) and bacteria in sediment banks is also necessary for a complete understanding of the geochemical evolution of the sediment and groundwater.

ACKNOWLEDGMENTS I acknowledge and greatly appreciate the work of my students and colleagues during our studies of the King River. I was fortunate to work with Jeff Taylor and Nigel Murphy of Earth Systems Pty Ltd, Melbourne, and Tamie Weaver of the University of Melbourne in our initial study of the King River during 1995. That study inspired several Honours and Masters research projects at Monash University. I thank Mick Hannan, Deborah Green, Cam Hooper and Simon Baker for their work, much of which resulted in the data summarised in this chapter. Simon Baker calculated and provided Figure 9.4 and read a preliminary draft of the manuscript. Marion Anderson and Joel Brugger are also thanked for reading versions of the manuscript and making good suggestions for improvements. I also thank the editor of this book, Vic Gostin, for giving me the opportunity to contribute and for his interest and help in producing this chapter. Funding for the initial study (Taylor et al 1996 ) was provided by the Department of Environment and Land Management, Tasmania and the Office of the Supervising Scientist, Australia, and funding for the Monash University research projects was provided by the Australian Research Council Small Grants scheme in 1996 and 1997.

REFERENCES ALLISON

D.,

BROWN

NOVO-GRADAC

1991.

MINTEQA2/PRODEFA2, A Geochemical Assessment Model for Environmental Systems: Version 3.0 User's Manual. United States Environmental Protection Agency, Athens, Georgia. CORBETT K. D. 1992. Stratigraphic-volcanic setting of massive sulphide deposits in the Cambrian Mount Read Volcanics, Tasmania. Economic Geology 87, 564-586. DREVER J . I . 1 9 9 7 . The Geochemistry of Natural Waters: Surface and Groundwater Environments (3rd edition). Prentice-Hall, Upper Saddle River, New Jersey. FETTER C. W. 1994. Applied Hydrogeology (3rd edition). Prentice-Hall, Englewood Cliffs, New Jersey. FOLK R. L. 1980. Petrology of Sedimentary Rocks. Hemphill Publishing, Austin, Texas. GREEN D . 1 9 9 7 . The geochemistry of mine tailings and groundwater in a sediment bank of the King River, Tasmania. BSc (Hons) thesis, Monash University, Melbourne (unpubl.). HANNAN M. 1996. Acid generation from oxidised mine tailing sediments in the King River, Tasmania. BSc (Hons) thesis, Monash University, Melbourne (unpubl.). HOOPER W. C. 1997. The hydrogeology of a sediment bank on the King River, Tasmania. BSc (Hons) thesis, Monash University, Melbourne (unpubl.). HVORSLEV M . J . 1 9 5 1 . Time lag and soil permeability in ground water observations. US Army Corps of Engineers Waterway Experimentation Station Bulletin 36. KOEHNKEN L. 1 9 9 7 . Final Report. Mount Lyell Remediation, Research and Demonstration Program. Supervising Scientist Report 126, Supervising Scientist, Canberra. LOCHER H . 1997a. Sediment storage and transport in the King River, Tasmania. PhD thesis, Monash University, Melbourne (unpubl.). LOCHER H. 1997b. Sediment transport in the King River, Tasmania. Mount Lyell Remediation, Research and Demonstration Program.

Mine tailings, King River, Tas. Supervising Scientist Report 120. TAYLOR J . R . , WEAVER T . R . , MCPHAIL D . C . & MURPHY N . C .

1996.

Characterisation and impact assessment of mine tailings in the King River system and delta, Western Tasmania, Mount Lyell Remediation, Research and Demonstration Program. Supervising Scientist Report 105. TESSIER A., CAMPBELL P. G. C . & BISSON M. 1979. Sequential extraction procedure for the speciation of particulate trace metals. Analytical Chemistry 51, 844-880.

109

WALSHE J. L. & SOLOMON M. 1981. An investigation into the environ-

ment of formation of the volcanic-hosted Mt Lyell copper deposits using geology, mineralogy, stable isotopes, and a six-component chlorite solid solution model. Economic Geology 76, 246-284. WILKIN R. T. & BARNES H. L. 1996. Pyrite formation by reactions of iron monosulfides with dissolved inorganic and organic sulfur species. Geochimica et Cosmochimica Acta 60, 4167-4179. Received 26 June 2000, accepted 8 August 2000

Geological Society of Australia Special Publication 21, 111 -124

CHAPTER 10—Natural radioactivity, hazards, wastes and the environment in Australia R. MAJOR* 14 Hounslow Avenue, Mile End, SA 5031, Australia. Ionising radiation emanates from many natural sources, such as the Sun and outer space, air, rocks, soil, food, water and our bodies. Radiation also comes from artificial sources such as X-ray machines, TV sets, luminous watch dials, fertilisers and pharmaceutical products. The biggest exposures for the average person come from air (radon gas), rocks, soil, building materials and medical X-rays. Our individual exposure (for non-radiation workers) from these natural and artificial sources is 2-3 mSv/y. For the average person less than 1% of our exposure to ionising radiation comes from the coal and nuclear-fuel industries and from nuclear-weapons test fallout. Some people live in areas of very high natural background radiation (ten times the average) with no apparent ill effects. Geological knowledge is vital in the planning of nuclear-waste disposal facilities. KEY WORDS: cosmic rays, in situ leaching, nuclear radiation, Olympic Dam, potassium-40, radioactive wastes, radium, radon gas, uranium mines.

INTRODUCTION Radioactive elements are present in various amounts in all materials in and around us, and this chapter describes the sources and strengths of natural radioactivity. Radioactivity is discussed in relation to cosmic rays, the earth's crust, soils, coals, fertilisers, building materials, water and foods. Uranium mining in Australia is considered with special reference to mine-waste disposal at Olympic Dam, and the new application of the in situ leaching method to uranium mining. This chapter ends with a summary of Australian plans for a radioactive-waste repository site. Nuclear reactions occur during the splitting (fission) of the nucleus of a heavy atom (e.g. uranium) or by the joining together (fusion) of nuclei of light elements (e.g. hydrogen). Nuclear fission releases enormous energy because the nucleus is disrupted and some very strong nuclear binding energy is released. Consequently the fission of a small amount of uranium releases as much energy as the burning of a very large amount of coal. For example, a large 1000 M W electric nuclear-power reactor will fission about 1 t of uranium-235 in one year, releasing about the same amount of energy as that from a coal-fired power station burning about 2 700 000 t of black coal (Aswathanarayana 1986 p. 137). Technical aspects of nuclear-power reactors are described in Nero (1979) and comparisons of wastes and pollution from nuclear and coal power are found in Wilson and Jones (1974) and Eicholz (1976). Radiation is a general term to describe the dissemination of energy from a source where, in the absence of absorption, the energy falls off as the inverse square of the distance from the source. This term is applied to sound waves as well as to electromagnetic waves such as radio waves, microwaves, infrared (heat) waves, visible light, Xrays, gamma rays, etc. It is also applied to emitted particles (e.g. alpha and beta particles, protons, neutrons) from a radioactive or other source (Walker 1991). Details of the

physics of nuclear reactions may be obtained from standard texts, web sites (see references), and a summary is provided by Nero (1979). Almost all radiation may be harmful under certain conditions depending on the energy transmitted and whether that energy is absorbed by living or non-living things. When various types of radiation pass into a material they interact with the atoms and molecules of that material. Some of the radiation (called ionising radiation) has enough energy to remove electrons from those atoms and molecules. All radiation with a wavelength shorter than ultraviolet (e.g. gamma radiation) and alpha and beta particles are ionising. Radiation with wavelengths of ultraviolet or longer are nonionising. However, non-ionising radiation, including ultraviolet, infrared and microwaves, can cause biological damage (e.g. sunburn) when they deposit enough energy (Dale 1979 p. 20). The amount of damage depends on the nature of the radiation, strength of the source, its distance, the effect of any shielding and the length of exposure time. Ionising radiation can cause biological damage by two principal mechanisms: (i) direct damage by ionising atoms or molecules in the body resulting in their disruption: for example, the disruption of a DNA molecule in the nucleus of a cell prevents the information originally contained in the genes from being transmitted accurately to the next generation; the likelihood of such mutation is extremely low for doses <250 milliGrays (mG) (see Appendix 10.1) (Cember 1983 p. 179); (ii) indirect damage by the disruption of water molecules (most of the body is water) resulting in the production of highly reactive free radicals and molecules (e.g. hydrogen peroxide) that are chemically toxic to the body (Cember 1983 p. 180). Biological damage depends on the amount of energy deposited in the tissue, and the method and rate of deposi* former Uranium Projects Officer, South Australian Department of Mines and Energy.

112

R.Major

tion. The effect of deposition rate is that a large dose received in a short time causes more damage than would the same dose spread over a longer period. This is analogous to the effects of taking a hundred aspirin tablets in an hour, compared with taking one tablet a day for one hundred days (Dale 1979 p. 20; Canadian Nuclear Association 1975 p. 9).

NATURAL RADIOACTIVITY Radiation of natural origin pervades the whole environment. Every second, more than 100 cosmic rays penetrate our bodies, several radioactive atoms of the air we inhale disintegrate within our lungs, and about 4 0 0 0 potassium4 0 atoms together with two or three uranium atoms from our food and drink disintegrate inside us exposing us to ionising radiation. In addition, more than 5 0 0 0 0 gamma rays enter our bodies from the soil and building materials around us (Gaines 2000; Lapp 1979 p. 31). The following section describes the role of radioactive heat in the Earth's heat flow, the nature of cosmic radiation, and the major radiogenic elements at the Earth's surface.

Radioactivity and the internal heat of the Earth The Earth has two general sources of heat: that from the Sun and the internal heat from the Earth. The heat flowing from the Earth is very small, and is about 5 0 0 0 times less than the heat received from the Sun. The surface temperature and climate are almost totally controlled by solar radiation (Clark & Cook 1992 p. 369; Decker & Decker 1989 p. 163). There are four main sources of internal heat within the Earth: (i) original heat from the time of formation; (ii) heat from the gravitational energy of redistributing heavy elements (e.g. iron) toward the centre of the Earth (iii) heat from tidal friction slowing the Earth's rotation; and (iv) heat from the breakdown of radioactive elements (Lowrie 1997 p. 178). The heat produced by tidal friction arises from the slowing of the Earth's rotation and much of this energy is dissipated in the swirling tides of the Earth's oceans. The accretion of the Earth at the time of formation of the solar system generated immense amounts of heat from collision, gravitational collapse and infall of meteorites. This heat together with heat from radioactive decay, would have caused melting. The resulting gravitational segregation of iron to form the core would have greatly amplified the total heat generated, perhaps melting the entire Earth. However gravitational melting would have been a one-event process and its heat long gone unless sustained by radioactive decay. The radioactive decay of unstable atoms would have started warming fast at first because of many short-lived, highly radioactive elements, and then more slowly from the longer lived but less radioactive elements such as uranium238, thorium-232, and potassium-40. The average present surface heat flow from the continents is about 6 0 mW/m 2 of which approximately 24 mW/m 2 is generated within the crust by the decay of these radioactive isotopes (England 1992 p. 288).

Concentration of radioactivity in the Earth's crust The Earth's continental crust has an overall 'granitic' composition and has been derived by partial melting of the underlying mantle with an overall 'basaltic' composition. The continental crust, typically 3 5 - 4 0 km thick and covering about 40% of the surface of the Earth, accounts for only about 0.4% of the Earth's volume. However, it contains about 30% of the total Earth's (terrestrial) budget of the heat-producing radioactive elements (uranium, thorium and potassium-40), whose size or electric charge excludes them from the common minerals in the Earth's mantle, and causes them to be concentrated in the crust (Taylor 1990). The chemical differentiation of the Earth into the metallic core, siliceous mantle and siliceous crust is a result of thermal evolution involving melting, metamorphism and density stratification (Gill 1996 p. 252). Differentiation of the crust from the mantle has resulted from ongoing igneous, sedimentary and metamorphic activity throughout the Earth's history. W h e n a solid rock partially melts and that melt develops in equilibrium with the solid rock, then elements are fractionated (separated) in two ways: (i) minerals with lower melting temperatures (containing iron, aluminium, sodium, silicon) enter the melt preferentially, leaving the residual solids enriched in refractory minerals (mainly containing magnesium and calcium, that have a higher melting temperature); and (ii) during melting, crystals tend to exsolve certain trace elements that are more difficult to accommodate in their structure: these elements are termed incompatible elements (such as uranium, thorium, potassium and rubidium) because the ions of these elements are too large and/or have too high an electric charge to fit easily into the host crystal; these incompatible elements are more easily accommodated in the open disordered structure of a melt than in the lattice of the host crystal (Gill 1996 pp. 201, 252). Extraction of magma from the mantle progressively transports these incompatible elements from the mantle upwards into the crust, and thus crustal rocks become enriched in these elements. Repeated melting and metamorphism in the continental crust have led to its internal differentiation into a lower, more refractory crust relatively depleted in incompatible elements, and an upper crustal layer - 1 0 km-thick, enriched in silicon, sodium, calcium and aluminium together with almost all the crustal inventory of incompatible radioactive elements (uranium, thorium and potassium). Most of the crustal contribution to measured surface heat flow in continental areas originates in this upper 10 km layer (Gill 1996 p. 252).

Cosmic radiation Cosmic rays originate predominantly from the Milky Way galaxy and from the Sun, with very rare, but very energetic rays from other galaxies (Clay & Dawson 1997). Primary cosmic rays are mainly protons, but also alpha particles, electrons and the nuclei of atoms some of which impinge on the atmosphere at almost the speed of light, and from all directions of space. The total flux of these primary cosmic rays above the atmosphere is about 1 cosmic ray/cm 2 /s. Such nuclei with relativistic speeds are called primary cosmic rays to distinguish them from the cascade of sec-

Natural radioactivity in Australia

113

Table 10.1 Ranges and averages of the concentrations of 4 0 K, 2 3 2 Th and 2 3 8 U in typical rocks, soils and building materials (adapted from Eisenbud & Gesell 1997 tables 6-5, 6-6, 6-16, 6-17). Potassium-40 total K% Bq/kg Average granite Average basalt Ultrabasic igneous Quartz sandstone Arkose Shale Bituminous shale (Tennessee) Coal (all USA) Phosphate rock (North Africa) Phosphate rock (Florida) Carbonate rocks Cement Clay brick Wood Wallboard Average soil Beach sands All rocks (range) Continental crust (av.)

4 0.8 na <1 2-3 2.7 na na na na 0.3 0.8 2.3 11 0.3 1.5 <1 0.3-4.5 2.8

1000 300 —

<300 600-900 800 — —

Thorium-232 ppm Bq/kg 2-17 3-4 na <2 2? 12 na 1-79 (4.5 mean)

8-70 10-15 —

<8 <8 50 — —

—

70 237 666 3300 89 400 <300 70-1500 850

2 5.1 10.8

8 21 44

—

Uranium-238 ppm Bq/kg 3-5 0.5-1.0 0.03 <1 1-2? 3.7 50-80 0.2-43 (1.7 mean) 20-30 120 2 3.4 8.2

40-67 7-10 0.37 <10 10-25? 40 610-980 —

240-370 1500 25 46 111

—

3 9 6 1.6-20 10.7

12 37 25 7-80 44

1 1.8 3 0.5-5 2.8

14 22 40 7-60 36

na, not available.

ondary particles and rays generated by their impact with nuclei of atoms in the atmosphere. The secondary particles and rays (mesons, electrons, gamma rays, neutrons, etc.) shower down through the atmosphere, colliding and changing as they go, and are found with decreasing intensity all the way to the ground, and below ground (Rossi 1966 figure 11-4). At sea-level the mesons account for about 80% of the cosmic radiation flux and electrons for about 20% (Friedlander 1994; Chown 1998; Eisenbud & Gesell 1997 p. 185). Apart from direct exposure to cosmic radiation there is also indirect exposure from biologically active radionuclides such as tritium ( 3 H) and carbon-14 ( 14 C) which are produced by interaction of cosmic rays with atoms in the atmosphere. Other radionuclides from cosmic-ray interactions are isotopes of beryllium, sodium, phosphorus, sulfur and chlorine (Eisenbud & Gesell 1997 table 6.1a). Cosmic rays cause a low radiation exposure to people, with over 100 cosmic ray particles passing through our bodies every second. At sea-level this exposure is about 0.3 millisieverts per year (mSv/y) but the exposure increases with altitude because the rays have increasing energy at higher levels in the atmosphere. Residents of Denver, Colorado, at an altitude of 1600 m, for example, receive nearly twice the annual dose that is received at sea-level. The passengers and crew of high-flying aircraft also are subject to additional dose from cosmic rays depending on the routes flown. The dose rate also depends on solar activity and on the latitude, because the Earth's magnetic field shields some cosmic radiation from the Earth except at the poles (Eisenbud & Gesell 1997 p. 186). Aircrew receive annual doses of between 0.3 and 9.0 mSv (average 1.5 mSv/y), which is about five times that received from all natural background radiation by a person who lives at sealevel (Hore-Lacy 1999 pp. 57, 78; Eisenbud & Gessell 1997 p. 187). In Australia, it is estimated that international and

domestic cabin crews receive 4 mSv/y and 2 mSv/y respectively (Hockings et al 1999). Therefore the total annual exposure to some aircrew exceeds the average annual exposure to Australian uranium miners (see Table 10.6).

Ionising radiation from rocks, soil, food and our bodies Ionising radiation emanates from rocks, soil and any building materials derived from them, because of their content of radioactive potassium-40, together with uranium and thorium and their radioactive decay products (Table 10.1). Furthermore, the bodies of all people on Earth contain some radioactivity from ingestion and inhalation. The alpha-emitting radionuclides in a typical New York diet are listed in Table 10.2, and the estimate of radioactivity in the adult body is shown in Table 10.3. This indicates that the average adult is mildly radioactive with over 500 000 radioactive disintegrations per minute, mainly beta emissions from potassium-40 (Lapp 1979) (For dose rate comparisons see Tables 10.4 to 10.6). POTASSIUM-40 Of the three naturally occurring potassium isotopes only potassium-40 is radioactive, decaying to argon-40. Potassium-40 is about 0.01% of natural potassium, and imparts a radioactivity of about 30 000 Bq/kg potassium. Potassium found in seawater gives it a radioactivity of about 11 000 Bq/m3 (Eisenbud & Gessell 1997 p. 171). The potassium content of rocks varies widely from about 4% in granites to about 0.3% in limestones (Table 10.1). The potassium content of soils of arable lands is also enhanced by the use of fertilisers. A person who weighs 70 kg contains about 140 g of potassium, mainly in muscles,

114

R.Major

Table 10.2 Alpha-emitting radionuclides in a New York City diet (from Fisenne et al 1987). Food intake (kgfy) Fresh vegetables Canned vegetables Root vegetables Potatoes Dry beans Fresh fruit Canned fruit Fruit juice Bakery products Flour Whole-grain products Macaroni Rice Meat Poultry Eggs Fresh fish Shellfish Dairy products Rounded Total

48 22 10 38 3 59 11 28 44 34 11 3 3 79 20 15 8 1 200 637

with an activity of about 4000 Bq due to potassium-40. Because of its relative abundance and its energetic beta emission, potassium-40 is easily the predominant radioactive component in normal foods and in human tissues (Eisenbud & Gessell 1997 p. 171). As Table 10.3 indicates, the average adult person is mildly radioactive with 9300 radioactive disintegrations every second, mainly beta emissions from potassium-40. URANIUM Uranium is found in all rocks and soils. Typical concentrations are listed in Table 10.1 which show that acid igneous rocks such as granites contain concentrations in the order of 4 ppm, which is about five times greater than in basic rocks, and 100 times greater than in ultrabasic igneous rocks. However, some phosphate rocks in the USA contain as much as 120 ppm and are a potential source of uranium. Consequently phosphate fertilisers may have a high uranium content, as could some by-products of fertiliser manufacture (e.g. gypsum) used in buildings as plaster board. Because uranium occurs in soils and phosphate fertilisers, it is also present in food and therefore in human tissues (Tables 2, 3) with an activity of about 3 Bq. Lung, kidney and bone receive the highest annual doses from uranium, estimated to be 11, 9 and 6 pSv respectively (Eisenbud & Gesell 1997 pp. 135, 172).

U (total)

Radium-226 is a decay product of uranium-238. Radium226 and its decay products (e.g. radon-222 and its decay products) are responsible for a large part of the dose received by people from naturally occurring radioactive isotopes. Bone contains about 80% of the total radium in the

230Xh

0.9 0.03 0.01 0.007 0.09 0.007 0.004 0.02 0.4 0.17 0.07 0.007 0.004 0.2 0.01 0.015 0.011 0.03 0.07 2.0

232Th

0.8 0.3 0.06 0.004 0.08 0.007 0 0.01 0.12 0.09 0.03 0.003 0 0.16 0 0 0.01 0.03 0.05 1.5

226 R a 2.6 0.5 0.15 0.25 0.17 2.7 0.07 0.77 3.5 3 1.1 0.2 0.02 0.16 0.4 1.3 0.24 0.06 1.1 18

Table 10.3 Estimate of radioactivity in the adult human body (modified from Lapp 1979 p. 31). Radioisotope

Radioactivity (Bq)

Dose equivalent (mSv/y)

Tritium (3H) from nature pTritium (3H) man-made pCarbon ( 14 C) pPotassium ( 40 K) pRubidium ( 87 Rb) pStrontium ( 90 Sr) pCesium ( 137 Ce) pLead ( 210 Pb) pRadium ( 226 Ra) a Radium ( 228 Ra) pNatural Uranium a Total

30 1000 2960 4070 1075 48 104 22 3 2 3 9317

0.00001 0.0006 0.01 0.17 0.006 0.004 0.003 0.09 0.02 0.03 0.02 0.35

With about 9300 radioactive disintegrations per second the adult human has over half a million disintegrations per minute, making us all mildly radioactive

Table 10.4 Annual estimated average effective dose equivalent received by a person in USA from natural radiation (from Eisenbud & Gesell 1997 table 6-23). Source

RADIUM-226

Radionuclide intake (Bq/y)

2.2 0.2 0.2 0.07 0.18 0.2 0.04 0.03 2.3 0.3 0.5 0.02 0.02 0.3 0.03 0.04 0.2 4 0.4 12

Dose (mSv)

Inhaled (radon and decay products) Other internally deposited radionuclides ( 40 K, 2 1 0 Po) Terrestrial radiation Cosmic radiation Cosmogenic radioactivity (14C) Total (rounded)

2.0 0.39 0.28 0.27 0.01 3.0

Natural radioactivity in Australia human body, with an activity of about 0.85 Bq. The alpha particle emission from radium, radon and other emissions from the decay products to bone is about 170 pSv/y. Radium is chemically similar to calcium and is absorbed in the soil by plants and passed up the food chain to people. There is considerable variability in the radium content of foods (Table 10.2), because the radium content of soil and its parent rock is variable as is the metabolism of different plants. For example, brazil nuts are much more radioactive than other foods, in the order of 1000 times greater than the radium concentration in the foods that make up the average diet in USA. The radium content of surface waters is low (4-20 mBq/L) because dissolved radium adsorbs quickly to solids and does not migrate far from its place of release to groundwater (Eisenbud & Gesell 1997 p. 142). THORIUM-232 The thorium content of igneous rocks is - 1 2 ppm and of sandstones ~6 ppm (Table 10.1). The thorium content of igneous rocks is therefore about four times the uranium content, but because of the lower specific activity of thorium, the radioactivity due to thorium and uranium is about the same. Thorium-232 is present in biological materials in insignificant amounts due to its low solubility. It is found in the lungs, indicating that the principal source of human exposure is by inhalation of suspended soil particles. Table 10.2 indicates that the annual ingestion of thorium through food is about 3.5 Bq. RUBIDIUM-87 Rubidium-87 forms 28% of natural rubidium. It emits beta radiation resulting in an average whole body dose of ~3]iSv/y (Eisenbud & Gesell 1997 p. 172). RADON AND ITS DECAY PRODUCTS Radon-222, a member of the uranium radioactive decay series, is an inert, colourless and odourless gas having a half-life of 3.8 days. A major but variable contribution to the natural radiation dose rate arises from the inhalation of the short-lived radioactive decay products of radon-222, two of which are emitters of alpha particle radiation. These two short-lived decay products are polonium-218 (half-life 3 minutes) and polonium-214 (half-life 1.6 x 10" 4 s). If they decay in the lungs they may damage lung tissue with the emitted alpha particles possibly initiating lung cancer. Some scientific literature suggests that the risk for lung cancer from radon and smoking are multiplicative (Toussaint 1998 p. 15). Elevated levels of radon-222 in air can occur in uranium mines, and in non-uranium mines from groundwater. The enhancement of lung cancer incidence rates observed among uranium miners in the past is attributed to smoking and long-term exposure to high levels of radon decay products. However, uranium exists naturally in trace quantities in rocks and soil (Table 10.1) so that radon is always present in the atmosphere due to emanation from the rocks, soil, groundwater, natural gas, etc. Elevated levels of radon have been observed in dwellings in a number of countries

115

including the US, Sweden and the United Kingdom. These elevated levels often exceed international guidelines and are usually associated with a number of factors including the uranium content of the soil, soil porosity, soil moisture, atmospheric pressure, building material, mode of construction and ventilation. Depending upon the porosity of the soil, radon diffusion into dwellings can vary greatly. In addition, radon may also come from the domestic water supply, building materials, and the natural gas supply (Langroo et al 1990). Radon-222 also causes indirect exposure via one of its decay products, lead-210 (half-life 138 days), which deposits from the atmosphere onto vegetation. This results in elevated exposures to cigarette smokers (from lead-210 and its decay products in tobacco leaves), and to residents of northern countries who subsist on reindeer that consume lichens. Lichens absorb lead-210 and concentrate it because they are slow growing. Radon-220, which is a decay product of thorium-232, has a half-life of only 54 s, compared to 3.8 days for radon222. Therefore radon-220 has less time to diffuse out of its source into air, and because of its lower concentrations, it causes much less risk than does radon-222 (Eisenbud & Gessell 1997 p. 152-167).

High levels of natural radiation Whereas most background levels of radiation typically range from 1.5 to 3.5 mSv/y, in some places they are much higher, due to sands, rocks, and 'health spas' containing (variously) high levels of thorium, uranium, radium and radon. A high level of background radiation from monazitebearing sands, affects many people in Kerala and Madras States in India. There, some 140 000 people receive an annual dose rate of over 15 mSv from gamma rays (plus a similar amount from radon). High natural levels occur from sands and rocks in Brazil, Iran, China, Egypt, Kenya and Sweden, and from health spas in South America, Europe and Japan. Some places have natural background radiation with dose rates of more than 50 mSv/y. To date, no adverse health effects have been discerned from doses arising from these high natural levels (www.uic.com.au/ral.htm; Eisenbud & Gessell 1997 p. 191; UNSCEAR 1993 p. 40).

Natural radioactivity in coal Coal contains uranium, thorium and potassium radionuclides. The quantity of radionuclides discharged to the atmosphere per tonne of coal consumed depends on the concentration of radionuclides in the coal, the method of combustion, and the efficiency of fly-ash recovery. The mean value for uranium and thorium concentrations in mined coal is 1.7 ppm for uranium, and 4.5 ppm for natural thorium, which is similar to the average concentrations found in soils and rocks (Table 10.1). Some coals, however, have much higher concentrations (Eisenbud & Gesell 1997 pp. 175-180). About 90% of the mass of coal is consumed during combustion and the radioactive nuclides tend to concentrate in the nonvolatile fraction or 'ash\ The ash then undergoes partitioning, according to whether it separates within the furnace and stack, or passes with the hot gases to the 'fly

116

R.Major

ash'. Modern furnaces now burn pulverised coal and release 60-85% of the coal ash content into fly ash. In the past, it was common practice to allow the fly ash to escape to the environment, but this is no longer permitted in many countries, where electrostatic precipitators and scrubbers are used to reduce the amount of fly ash released. Because concentration of some radionuclides occurs in the stack, the amounts of lead-210 and polonium-210 in the fly ash are 5-10 times greater than in the original coal. In contrast, the emissions of radionuclides from a modern 1000 MW electric nuclear-power plant are mainly inert gases such as krypton and xenon, which do not combine biologically in the human body. The normal levels of uranium and thorium in the environs are sufficiently high that changes due to the emissions from coal-fired power stations are barely detectable. However, it has been found that the radium-226 content of snow downwind of coal-fired power plants is higher than its concentration in rainwater. The concentrations of naturally occurring radionuclides in airborne particles collected in the vicinity are elevated above background, and increase the dose to the lung (Eisenbud & Gesell 1997 p. 178). The first report on radioactive emissions from coal-burning power plants in 1964 concluded that the dose from the fossil-fuel emissions was greater than that from nuclear reactors. However, stringent requirements of the US Federal Clean Air Act have since resulted in substantial reductions in the lung dose from fly-ash emissions. A modern coal-fired power station causes an average exposure of about 0.0014 mSv/y to people downwind (Eisenbud & Gesell 1997 p. 178). Radioactive emissions from nuclear-power reactors The emissions of radionuclides from a nuclear-power reactor are mainly inert gases such as krypton and xenon, which do not combine biologically in the human body. Trace emissions of radioactive iodine however, do combine with human tissue, but the combined exposures of these emissions is minute. For example, in 1988 in the USA there were 110 nuclear power reactors in operation, and about 150 million people lived within 80 km of these plants. Their average whole-body exposure was only 0.000 005 mSv in that year (Eisenbud & Gesell 1997 p. 243) compared to an exposure of 0.0014 mSv/y from a modern coal-fired power station. In both cases these exposures are not significant to health.

RADIATION IN AUSTRALIAN HOMES Australians spend most of their time in buildings, particularly in their homes. Consequently the radiation levels from gamma radiation and from radon-222 decay products inside their homes are an important component of their exposure to radiation. For instance, international data now suggest that about 50% of the average dose to the population is from radon gas (Toussaint 1998). Radon concentrations and gamma-radiation exposures depend upon geological location of the home, and tend to be higher in solid brick and brick veneer dwellings than in

those of lighter construction. This is because bricks are made of clay containing traces of uranium, thorium and radioactive potassium-40. Radon concentrations tend to be enhanced also in homes which are poorly ventilated or which rely on air conditioning for ventilation (Langroo et al 1990 p. 14). In 1989 a nation-wide survey of Australian homes was conducted by the Australian Radiation Laboratory (in Melbourne) to determine the average annual dose to the Australian population from exposure to radon and gamma radiation. Dose meters were placed for 12 months in 3413 randomly distributed homes (representing about 1 in 1400 occupied dwellings). The measured annual average radon concentration in Australian homes was about 12 Bq/m3, and the average effective dose equivalents to a person in Australia were estimated to be 0.6 mSv/y for radon, and 0.9 mSv/y for gamma radiation (Langroo et al 1990). These are low values compared to the world average radon concentrations of about 40 Bq/m3. France and Germany have radon levels close to this, while Britain is about 22 Bq/m3. However, many houses built on granites or uranium-bearing shales in UK, Sweden, Austria and Hong Kong record many times this value (Selinger 1998 p. 476). Several countries have set guidelines for remedial action to reduce radon levels in homes that have high radon concentrations. One of the most stringent recommendations to date is that of the National Radiological Protection Board in the United Kingdom. It suggests a threshold level of 200 Bq/m3 for simple remedial action, such as the use of a fan to increase the air pressure within the home, and hence reduce radon emanation from below ground. In the Australian study only two of the 3413 homes surveyed were found to have radon concentrations in excess of 200 Bq/m3. If this proportion of high-concentration homes is properly representative of Australia as a whole, only 2000-3000 homes would be expected to exceed this value. In general, however, the values for radon concentration in Australian homes are lower than those found elsewhere and show that there is little need for concern (Toussaint 1998). Radon and gamma radiation levels are unevenly distributed in the environment and cluster in certain geological locations. While the mean of the exposure may be low in any region, there is the possibility of localised areas of relatively high radiation levels from the underlying rocks and soils. The Perth region provides an example of a wide range of radon and gamma radiation levels. From 1990-94 radiation-measuring instruments in 1912 homes showed a wide range of radiation levels from the Perth coastal plain, eastwards onto the Darling Scarp and to the south of Perth (Toussaint 1998). The mean indoor radon concentration was about 17 Bq/m3 with a range from 0.2 to 560 Bq/m3. Only 10 homes were found to have more than 100 Bq/m3, and these were generally on the Darling Scarp, where granites are common (mean 28 Bq/m3, max. 253 Bq/m3), and to the south of Perth. The higher gamma radiation levels were also located generally on the Darling Scarp. Clearly, knowledge of the local geology is relevant to better understanding and management of this issue. In Western Australia, the Radiological Council has set action levels of 100 Bq/m3 for new houses and 200 Bq/m3 for existing homes. While the data indicate that the radon

Natural radioactivity in Australia levels are generally low, there were three homes that exceeded the action levels. According to the International Commission on Radiological Protection, which sets recommended levels for the guidance of local authorities, a concentration of 60 Bq/m3 for radon-222 in homes (assuming 80% occupancy for a whole year) corresponds to an effective dose from the short-lived decay products of about 1.0 mSv/y (Toussaint 1998).

117

Other, including coal, nuclear power and fallout

Consumer items Medical Internal

Radon Gas

Summary of radiation exposure People receive background radiation from many sources both natural and artificial. Radiation exposure for most people comes from natural sources such as rocks, soils, food, water and building materials; it also comes from cosmic rays and radon gas in the air. Artificial sources are air travel and medical procedures, with very minor exposure from nuclear-power plants, coal-fired power plants, and the testing of nuclear weapons. These sources are summarised in Figure 10.1. Table 10.4 details the annual dose equivalent exposure from all natural sources in the USA. Table 10.5 summarises the exposure from all sources. This indicates that the total annual background exposure to a person from natural and artificial sources is about 2-3 mSv. Table 10.6 compares this typical background exposure of an average Australian to that of a variety of people.

URANIUM MINING IN AUSTRALIA The first reported (but unconfirmed) occurrence of uranium in Australia was in 1890 at the Mt Rhine mine, southeast of the Barossa Valley, South Australia (O'Neil 1982 p. 152). The first confirmed record of uranium was in 1894 from the Carcoar district of New South Wales (Card 1896). In 1896 Krause reported the mineral liebigite (hydrated calcium uranium carbonate) from Mt Ogilvie in the northern Flinders Ranges of South Australia. In 1906 uranium was discovered and mining commenced at Radium Hill in eastern South Australia (Dickinson et al 1954 pp. 7-9) and in 1910 at Mt Painter in the northern Flinders Ranges, South Australia (Coats & Blissett 1971 p. 152). Uranium was mined for its radium content from these two localities at various times until the 1930s. Radium was very valuable and was the only source of very penetrating gamma radiation for medical purposes—mainly to treat cancer in the early part of the 20th century (Landa 1987). Serious exploration for uranium in Australia for possible use in nuclear weapons commenced in 1944. The known deposits at Mt Painter and Radium Hill were re-examined, and uranium was discovered at Rum Jungle (Northern Territory) in 1949, in the South Alligator River Valley (Northern Territory) in 1953, and at Mary Kathleen (Queensland) in 1956 (Battey et al 1987). As a result of this exploration and discovery, 13 uranium mines were operating between 1954 and 1971. They produced -78001 of uranium. Exploration was revived in the late 1960s with the prospect of increased demand for uranium for the generation of electricity. Numerous deposits were found (Figure 10.2) and four mines were opened. In 1976 Mary Kathleen was recommissioned and closed in 1982. Nabarlek (Northern Territory), was mined out in 1979. Ranger

Terrestrial

AVERAGE RADIATION DOSE (USA) Figure 10.1 Average radiation dose estimates for the US public (after Eisenbud & Gessell 1997 p. 528).

Table 10.5 Sources of exposure to ionising radiation (from Hore-Lacy 1999 table 13). Typical

Natural Terrestrial + house: radon Terrestrial + house: gamma Cosmic (at sea level) +0.02 for every 100 m elevation Food, drink and body tissue Total Artificial From nuclear weapon tests Medical (X-ray, CT, etc. av.) From nuclear energy From coal burning From household appliances Total Behavioural Skiing holiday Air travel in jet airliner

Range

(mSv/y)

0.7

0.3-100

0.6

0.3

0.3 1.9 0.003 0.370 0.0003

0.1-1

0-0.5 0.1-1

up to 25

0.0001 0.0004 0.375 0.008 mSv per week 0.0015-0.005 per hour

up to5

The International Commission for Radiological Protection recommends, in addition to background, the following exposure limits: for general public, 1 mSv/y; for radiation workers, 20 mSv/y averaged over 5 consecutive years.

(Northern Territory) began producing in 1981 and Olympic Dam (South Australia) began producing in 1988. Updated references to Australia's uranium deposits are available from the Uranium Information Centre, Melbourne (www.uic.co.au including a map at www.uic.com.au/ozuran.htm). At present only Ranger and Olympic Dam are producing. In the calendar year 1999 their total production was about 7055 t of uranium oxide concentrate ('yellow cake') and total sales to nuclear-power companies in nine countries was about $345 million (Uranium Information Centre

118

R.Major

Table 10.6 Some comparative radiation doses (modified from Hore-Lacy 1999 pp. 58,78; Eisenbud &Gesell 1997 p. 191; UNSCEAR 1993 p. 40; www.uic.com.au) 1 mSv/y 2 mSv/y 3 mSv/y 1.5 mSv/y 2.9 mSv/y 5.0 mSv/y 10 mSv/y 20 mSv/y 5-50 mSv/y 50 mSv/y 350 mSv in lifetime 1000 mSv 5000 mSv 10 000 mSv

Current limit for public above that received from background and medical exposure. Typical background radiation to Australian public. Typical background radiation to North American public. Average incremental dose for aircrew. Average occupational dose to US nuclear industry employees. Average occupational dose to Australian uranium miners. Maximum actual dose to Australian uranium miners. Current limit for nuclear industry employees (or a max. of 100 mSv averaged over 5 years). Background radiation to people living in natural highly radioactive areas. Current maximum limit in single year for nuclear industry employees and uranium miners. Criterion for relocating people after Chernobyl accident. As short-term dose: probably causes (temporary) radiation sickness. As short-term dose: would kill about half those receiving it within a month. As short-term dose: fatal within days or weeks.

Newsletter, March-April 2000).

Mine-waste disposal at Olympic Dam The Olympic Dam deposit at Roxby Downs, in central South Australia, is a huge orebody mined underground. It is composed mainly of hematite (iron oxide) and some rareearth minerals, but only copper, uranium, gold and silver are extracted and sold. Discovered by Western Mining Corporation in 1975, it commenced production in 1988. The orebody is at least 7 km long and 4 km wide, and in 1999 had about 180 km of underground access tunnels. It originally contained about 1620 Mt of mineralised rock, about 566 Mt of which are ore grade. This ore (in 1988) contained about 11 Mt of copper, about 360 000 t of uranium oxide, about 2 700 t of silver and about 270 t of gold (Newton et al 1988). It is the sixth largest copper orebody and by far the single largest uranium orebody in the world. Annually, from 1999 more than 9 Mt of ore is mined to produce each year about 200 000 t copper, 4600 t uranium oxide, 2.4 t gold and 26 t silver, at a market value of ~A$650 million. At that production rate there are approximately 8.3 Mt of waste (tailings) to be disposed of. Tailings are the finely ground residue of the ore which remains after most of the copper, uranium, gold and silver have been extracted. The tailings are acidic with a pH of 1-2. The tailings are separated into two grainsizes: fine (mainly silt-sized and called 'slimes') and coarse (sand). Each year about 1.7 Mt of the sand-sized tailings are mixed with cement and pumped underground to fill the mined-out stopes from which the ore has been removed. In addition, about 6.6 Mt of slimes are pumped to the open-air tailings dams, which have a total area of about 190 ha. Excess water is pumped to lined evaporation ponds and the slimes are allowed to dry and consolidate in the dry climate. The radioactivity of the ore is low because the amount of uranium and its decay products is low, as there is only about 600 g/t of uranium in the ore. The tailings contain residual quantities of uranium and other radioactive isotopes in its decay chain. The overall radioactivity in the tailings is approximately 80% of the radioactivity of the original ore. Measurements in air near the tailings dams have shown

that natural ventilation is sufficient to disperse and dilute radon and radon decay products to very low levels within quite short distances. Similarly dust monitoring has shown that the tailings dams are not a major dust emission source. This is because the even grainsize and smooth flat moist surface of the tailings limit the processes that could lead to dust lift-off. In general, traces of radionuclides attributable to operations can be found at distances up to 5 km from the mine, plant and tailings dams. The resultant radiation doses to the public are a small proportion of that allowed by the international radiation protection standards. A final rehabilitation plan for the tailings storage facility is subject to ongoing trials. However, preliminary calculations indicate that a 1 m-thick soil cover over the surface, overlain by rock armour, would be sufficient to achieve an acceptable reduction in the long-term release of radon. (Olympic Dam Marketing 1987; WMC Olympic Dam Corporation 1997; Yeeles 1999; T. Dwyer pers. comm. 2000).

In situ leaching at Honeymoon and Beverley, South Australia In situ leaching is a mining technique whereby holes are drilled into the orebody, and a leaching solution is pumped into the orebody to dissolve the commodity (uranium or copper). The solution is then pumped to the surface, the uranium (or copper) removed, and the solution is recharged with chemicals and pumped underground in a continuous cycle. Because there are no underground or opencut workings, there is very little waste or disturbance on the surface. There are no surface dumps of ore or waste rock, no extensive buildings, smelter, refinery, tailing dams, roads, etc. However, only a few ores are amenable to in situ leaching, requiring the following conditions: (i) the chemistry of the ore must be amenable to being dissolved at normal (low) underground temperatures; (ii) the host rock must be porous and permeable (e.g. sandstone) to allow the leaching solution to flow through the ore zone; (iii) the ore zone must be saturated with groundwater (i.e. it is an aquifer); and (iv) the ore zone must be confined below and above by impermeable rock (usually clay) that prevents the escape of

Natural radioactivity in Australia

Jabiluka Darwin

119

.Nabarlek

Ranger

Koongarra. Westmoreland "Ben Lomond

Valhalla Manyingee

•

Kintyre

Mary Kathleen

NORTHERN TERRITORY

QUEENSLAND WESTERN AUSTRALIA Yeelirrie

Brisbane

SOUTH AUSTRALIA Beverley Olympic Dam ^ ^ NEW SOUTH WALES Honeymoon^. Radium Hill

Perth

Sydney Legend ^

VICTORIA

"Canberra

Uranium mines and treatment plants

X Historic mines

1000 Kilometers

Uranium Prospects Hobart Figure 10.2 Uranium mining in Australia (modified from www.uic.com.au, April 2000).

water or solution. At Honeymoon and Beverley in South Australia (Figure 10.2) uranium occurs about 110 m underground in sandstones which are suitable for in situ leaching (Howies 2000). Honeymoon contains about 4600 t and Beverley about 21 000 t recoverable uranium oxide. The groundwater in the ore zones is highly saline and contains radioactive elements—the water is not suitable for any use, not even the watering of stock. The drillholes are in a square pattern with injection holes at the corners of the squares and extraction holes in the centre (Townsend 1997; Howies 2000). More liquid is pumped out than is injected so that the groundwater flow in the orebody is towards the centre and leach solution does not escape outside the drill pattern. Monitoring wells surround the wellfield. The leaching solution is groundwater from the ore zone, with dilute sulfuric acid and oxidising agents added to dissolve the uranium. When the leaching solution containing uranium is pumped to the surface, the uranium is to be removed by solvent extraction (Honeymoon) or by ion exchange (Beverley), and converted to uranium oxide concentrate ('yellow cake'). In situ leaching produces very little waste either radioactive or non-radioactive. Gaseous radioactive waste, princi-

pally radon, is vented to the atmosphere. Liquid waste consists of spent process liquids and excess water pumped out of the extraction wells. Both will be neutralised and held temporarily in surface ponds. They are then pumped back underground into areas already mined out or into areas that will not be mined because they contain no ore. There are only small amounts of solid low-level radioactive wastes from the processing plant (e.g. filters, worn-out pumps, valves and piping). These may be contaminated with radium and would be buried on site, or taken to the national waste repository, probably at Billa Kalina (see below). In situ leaching is, or has been, conducted in numerous localities in the USA sometimes in potable aquifers with no adverse effects (Southern Cross Resources Australia 1997; Heathgate Resources 1998; Townsend 1997, 1998).

AUSTRALIA'S RADIOACTIVE WASTE Just like every production process, the use of radioactive materials in Australia results in the production of waste. The waste is mostly paper, plastic, laboratory glassware, protective clothing, laboratory equipment and soil which is

120

R. Major

slightly contaminated by radioactivity, and instruments and gauges such as electron tubes, industrial smoke detectors, luminescent signs, and radium-painted watch faces and compasses, which contain a radiation source but are no longer useful Unlike many other countries, Australia does not produce any high-level waste (which is the by-product of nuclear-power generation). Nor does Australia use nuclear power for electricity generation. Our only small, low-powered research reactor produces radioisotopes for research and for medical and industrial use. The used fuel plates are sent overseas for reprocessing.

Categories of radioactive waste All radioactive waste is classified into categories based on how much radiation it emits, and the length of time over which it will continue to emit radiation. The purpose of this classification system is to ensure that the waste is handled, stored and disposed of in a suitable way. The following four categories for Australia's radioactive waste have been defined by the National Health and Medical Research Council (NHMRC) and available as fact sheets produced by the Department of Primary Industries and Energy 1998). Category A wastes contain short-lived radioisotopes (mainly with half-lives of less than five years) that emit mainly beta or gamma radiation. Long-lived radioisotopes emitting alpha radiation are only present at very low concentrations. Typical Category A wastes are lightly contaminated items such as plastics, protective clothing, laboratory equipment, industrial tools and soil. These wastes are compacted into 200 L steel drums. Category B wastes emit considerably higher levels of beta or gamma radiation than Category A wastes, but alpha radiation levels are still relatively low. This category includes gauges and sealed radiation sources used in industry as well as in medical diagnosis and therapy. Category B wastes are enclosed in concrete prior to disposal. Category C wastes contain radioactivity levels similar to Category B. These wastes are generally bulk waste materials arising from the processing of radioactive materials, soils that have been significantly contaminated, or large items of contaminated equipment. Category S wastes emit higher levels of radiation than Categories A, B and C. The amount of Category S waste is very small, and consists mainly of sealed radium sources and concreted waste from the reprocessing of spent fuel plates from the Lucas Heights Research Laboratories of the Australian Nuclear Science and Technology Organisation (ANSTO) south of Sydney. Categories A, B and C are more commonly referred to as low-level and short-lived intermediate level waste, and the best management option for their safe disposal is a national repository. Category S waste is commonly referred to as long-lived intermediate-level waste and must be retained safely in storage until appropriate long-term disposal options are developed. This type of waste is not suitable for near-surface disposal, but can be safely stored in a purpose-built facility above the ground. The national repository will be for the disposal of Category A, B and C radioactive wastes. Colocation of a Category S waste stor-

age facility with the repository is being considered since a single national facility is the best management option for Australia.

How much waste? Australia has accumulated less than 3500 m 3 of radioactive waste from over 40 years of research, medical and industrial use of radioactive material. Over half of Australia's current waste—10 0 0 0 drums of lightly contaminated soil—resulted from CSIRO research into the processing of radioactive ores during the 1950s and 1960s. The combined waste of all four categories is very slowly accumulating at a rate of < 6 0 m3/y. A year's waste could fit into about three semitrailers. If all the accumulated radioactive waste were to be spread over a football field, it would be only 50 cm deep. By international comparisons, this annual accumulation in Australia is very small. Every year, Britain and France produce around 25 000 m 3 of low-level waste, which is about seven times as much as the total Australian waste over the last 40 years.

Present management responsibility Commonwealth, State and Territory Governments are responsible for the management and storage of radioactive wastes generated within their jurisdictions. Radioactive waste is now held safely at over 50 interim storage sites throughout Australia. Waste comes from Australia's research reactor at Lucas Heights, and from a range of medical, agricultural, industrial and research organisations that use radioactive materials. All Australian governments have agreed that a national repository is the best way of safely managing Australia's small quantities of low-level and short-lived intermediate-level radioactive waste.

Selection criteria for a radioactive-waste repository site It is important that the site has both long-term stability and particular attributes that will enable the radioactive waste to be isolated, so there will be no risk either to people or to the environment. The only way to find a suitable site is to apply a set of criteria that includes social, technical and environmental considerations, right across the Australian continent. The Code of Practice for the near-surface disposal of radioactive waste in Australia (Department of Primary Industries and Energy 1992) developed by the NHMRC, specifies general site characteristics and site-selection criteria to make sure that the waste, once buried, will remain isolated from the biosphere throughout the time that the waste has the potential to be harmful. The criteria in this Code of Practice were developed from criteria established by the International Atomic Energy Agency in Vienna. The siting of Australia's national radioactive waste repository will be in accordance with the clearly defined requirements of this code. The thirteen site-selection criteria (a-m in Table 10.7) are: (a) the site should be located in an area of low rainfall, should be free from flooding and have good surface drainage features, and generally be sta-

Natural radioactivity in Australia

121

Table 10.7 Potential Australian radioactive waste repository regions and selection criteria (Department of Primary Industries and Energy 1998 fact sheet 7). Criteria

Billa Kalina

Bloods Range

Everard

Jackson

Maralinga

Mt Isa

Olary

Tanami

NT + + + +

WA + + + _ + + +

SA _ + _

QJd

a b, c, g d e, f, h i ], k 1 m

SA + + + + + + + +

SA/NSW +

NT _

•

+ + + +

Figure 10.3 The eight regions in Australia considered for their suitability for a radioactive waste repository (after Department of Primary Industries and Energy 1997). Most likely investigative sites are near Woomera. The Great Artesian Basin is shown stippled, and the Murray-Darling Basin is shown by diagonal shading. ble with respect to its geomorphology (i.e. the lartdform and its underlying structure); (b) the water-table in the area should be at such a depth below the planned disposal structure that groundwater is not likely to rise to within 5 m of the waste, and the hydrological setting should be such that large changes in the water-table are not likely; (c) the geological structure and hydrogeological conditions should allow modelling of groundwater gradients and movement, and make it possible to predict radioisotope migration times and patterns; (d) the site should be away from any known or anticipated seismic (earthquake) or volcanic activity that could put into question the stability of the disposal structures and the integrity of the waste; (e) the groundwater in the region of the site ideally should not be suitable for human consumption, or for pastoral or agricultural use; (f) the site should be in an area of low population density, where the projected population growth or prospects for future development are also very low; (g) the site should have suitable geochemical and geotechnical

+ + +

_ + _ -

+ +

properties to restrict migration of radioisotopes and to assist repository operations; (h) the site should be in an area that has no known natural resources (such as potentially valuable mineral deposits) and has little or no potential for agricultural or outdoor recreational use; (i) the site should have reasonable access for the transport of materials and equipment during the construction and operation, and for the transport of wastes to the site; (j) the site should not be in an area that has special environmental attraction or appeal, that is of notable ecological significance, or that is the known habitat of rare fauna or flora; (k) the site should not be located in an area of special cultural or historical significance; (1) the site should not be located in reserves containing regional services such as electricity, gas, oil or water mains and (m) the site should not be located in an area where land ownership rights or control could put into question the retention of long-term control over the repository. A computer-based geographic information system has been used to analyse all the social and geographical/geological information to assess the suitability of areas throughout Australia against these criteria. In the analysis, all the selection criteria were considered equally. The result of the analysis was that eight broad regions in Australia contained sites that were considered to be suitable against the criteria. These eight broad regions are shown on Figure 10.3, and are listed in Table 10.7 against the 13 criteria. This table shows that only the Billa Kalina and Olary regions, both in South Australia, satisfy all criteria. However, the former was chosen because it has the largest area in which to determine the actual small repository site.

Billa Kalina as a national waste-repository site A 67 000 km 2 region in central South Australia called Billa Kalina has been identified as the most suitable area to site the proposed 2.5 ha National Radioactive Waste Repository. The towns of Andamooka, Roxby Downs and Woomera lie within this region which has a low population density, low rainfall, does not support intensive agriculture, and has excellent road and rail access. Its geological structure and geochemical and geotechnical properties are suitable for a repository (Department of Primary Industries and Energy 1998 fact sheets 3, 5, 7). The selection of the actual repository site in the Billa Kalina region will follow extensive field investigations and

122

R. Major

a community consultation phase. The site will be surrounded by a buffer zone, bringing the total area to 2.25 km 2 . The selected repository site would undergo environmental assessment under the Environment Protection (Impact of Proposals) Act 1974 (Parer 1998). As at May 2000, five potential disposal sites have been selected in this region west of Lake Torrens (Minchin 2000).

Radioactive-waste site facility The disposal area for the low-level and short-lived intermediate level radioactive waste will be about the size of a football field ( - 1 0 0 x 1 0 0 m). It will be on a 1.5x1.5 km site (2.25 km 2 ). The disposal area will contain one or more trenches less than 2 0 m deep. The safely packaged waste will be placed in the trench, and then an engineered cover will be placed over the buried waste to provide an effective barrier against radiation, to protect it from rain, to control erosion and to keep the waste secure. The repository site will also include a building for receipt, handling and short-term storage of the waste before it is disposed of in the trench. The site will have accommodation and amenities for staff, an access road, electrical power and freshwater. Security and surveillance equipment, as well as instruments for radiological surveillance and monitoring of groundwater will be on the site. The 2.25 km 2 site provides a large buffer zone. Groundwater, external gamma radiation and concentration of radioisotopes in the air, soil and vegetation within the buffer zone will be under regular surveillance. Once a suitable repository site has been selected, it will also be considered for the colocation of a storage facility for Category S waste (see above). The Category S storage facility would also cover less than 100x100 m and would fit comfortably on the 2.25 km 2 site. The Category S storage facility would be a secure building designed to provide total containment of radioactive material. Concrete vaults inside the building would provide shielding to the safely packaged material (Department of Primary Industries and Energy 1998 fact sheet 9).

Operation, management and closure of the repository Detailed plans are available for the operation and management of the repository. The NHMRC's Code of Practice has detailed guidelines for closure of the disposal facility. Disposal operations at the facility will cease when the authorised disposal space is filled or the limit on total site radioactivity is reached. At the end of the established institutional control period no further control of the repository site will be necessary because the radioactive materials will have decayed to safe levels. In the unlikely event of a human intrusion on the waste after this period, it would not result in significant human exposure or environmental impact above the prescribed radiological dose limits.

radiation and this chapter has reviewed our present understanding concerning the natural distribution of radioactivity to which humans are exposed. Cosmic radiation continually penetrates our bodies, and modern air travel, especially on polar routes and during times of increased solar activity, adds a considerable radiation dose. On Earth, the major elements of concern are radioactive potassium, uranium, radium, thorium, and especially the gas radon. The inhalation of dusts associated with these elements is far more important than ingestion (as food). Certain rocks like granites, some coals, phosphate (for fertiliser), some soils, groundwaters and brick buildings, may all be naturally radioactive to a significant degree. For the great majority of people, less than 1% of radiation exposure comes from coalfired and nuclear power stations, and from the testing of nuclear weapons. This chapter has also reviewed uranium mining and the disposal of radioactive wastes in Australia. Geological, hydrogeological and soil mapping, together with a good geochemical understanding provides a sound scientific basis for environmental planning and management.

ACKNOWLEDGEMENTS I wish to acknowledge the editorial help of Vic Gostin, Andrew Winkler and Sandra McLaren; the typing and editorial help of Liz Campbell; and the drafting by Sharon Proferes and Aaron Brown.

REFERENCES ASWATHANARAYANA U. 1986. Principles Balkema, Rotterdam.

of nuclear

geology.

BATTEY G . C . , MIEZITIS Y . & MCKAY A . D . 1 9 8 7 . A u s t r a l i a n

A. A. uranium

resources. Bureau of Mineral Resources Resource Report 1. CANADIAN NUCLEAR ASSOCIATION 1975. Nuclear power in Canadaquestions and answers. Canadian Nuclear Association, Toronto. CARD G. W. 1896. Mineralogical and petrological notes. No.2. Geological Survey of New South Wales Report 4 ( 1 ) . CEMBER H. 1983. Introduction to health physics (2nd edition). Pergamon Press, New York. CHOWN M. 1998. Quasars pack a punch. New Scientist 2 1 4 9 , 7. CLARK I. F. & COOK B. J. 1992. Perspectives of the Earth. Australian Academy of Science, Canberra. CLAY R. & DAWSON B. 1997. Cosmic bullets. Allen & Unwin, Melbourne. COATS R. P. & BLISSETT A. H. 1971. Regional and economic geology of the Mount Painter Province. Geological Survey of South Australia Bulletin 4 3 . DALE G. C. (Editor) 1979. The CEGB and nuclear power—questions and answers. Central Electricity Generating Board, London. DECKER R. & DECKER B. 1989. Volcanoes. Freeman and Co., New York. DEPARTMENT OF PRIMARY INDUSTRIES AND ENERGY 1 9 9 2 . A

radioactive

waste repository for Australia: methods for choosing the right site. A discussion paper. Department of Primary Industries and Energy, Canberra. DEPARTMENT OF PRIMARY INDUSTRIES AND ENERGY 1 9 9 7 . A

radioactive

waste repository for Australia: Site selection study—Phase 3: regional assessment: A public discussion paper. Bureau of Resource Sciences, Department of Primary Industries and Energy, Canberra. DEPARTMENT

PRIMARY

INDUSTRIES

AND

ENERGY

1998.

Our

Radioactive Waste. Fact Sheet Nos. 1-12. Department, of Primary Industries and Energy, Canberra. DICKINSON S . B . , SPRIGG R . C . , KING D . ET AL. 1 9 5 4 . U r a n i u m d e p o s i t s

SUMMARY Humans are continually subject to a variety of hazardous

in South Australia. Geological Survey of South Australia Bulletin 10. EICHOLZ G. G. 1976. Environmental aspects of nuclear power. Science Publishers Inc., Ann Arbor. EISENBUD M. & GESELLT. 1997. Environmental radioactivity from natural

Natural radioactivity in Australia industrial and military sources. Academic Press Inc., San Diego. ENGLAND P. 1992. Deformation of the continental crust. In: Brown G. C., Hawkesworth C. J. & Wilson R. C. L. eds. Understanding the Earth—A New Synthesis, pp. 275-300. Cambridge University Press, Cambridge. FISENNE I. M., PERRY P. M., DECKER K . M. & KELLER H. W. 1987. The daily intake of U234, U235 and U238, Th238, Th230, Th232, Ra226 and Ra228 by New York City residents. Health Physics 53, 357-363. FRIEDLANDER M. W. 1994. Cosmic rays. In: The New Encyclopaedia Britannica, Micropaedia 3, p.661. Encyclopaedia Britannica Inc., Chicago. GAINES M. 2000. Radiation and risk. New Scientist 2230 (Inside Science, 4 pp.). GILL R. C. O. 1996. Chemical fundamentals of geology (2nd edition). Chapman & Hall, London. HEATHGATE RESOURCES PTY LTD 1 9 9 8 . Beverley Uranium Mine— Environmental Impact Statement Finsbury Press, Adelaide. HOCKINGS C., HAIDER M. A. & HIGSON D. 1999. Radiation protection issues at a large commercial airline. Australasian Radiation Protection Society Newsletter 16, 4. HORE-LACY I. 1999. Nuclear Electricity (5th edition). Uranium Information Centre Ltd, Melbourne. H O W L E S S . R . 2000. Beverley uranium project—groundwater resources, management and monitoring. MESA Journal 17, 4-6. KRAUSE F. M . 1 8 9 6 . An Introduction to the Study of Mineralogy for Australian Readers. George Robertson & Co., Melbourne. LANDA E. R. 1987. Buried treasure to buried waste—the rise and fall of the radium industry. Colorado School of Mines Quarterly 82(2). LANGROO M . K . , W I S E K . N . , DUGGLEBY J . C . & KOTLER L . H . 1 9 9 0 . A

nation-wide survey of radon and gamma radiation levels in Australian homes. Australian Radiation Laboratory ARL/TR 090. 1979. The radiation controversy. Reddy Communications Inc., Greenwich, Connecticut. LOWRIE W. 1997. Fundamentals of Geophysics. Cambridge University Press, Cambridge. MINCHIN N. 2000. Five sites selected for further investigation for the National Radioactive Waste Repository. Media Release 00/155, 17 May 2000, Minister for Industry, Science and Resources, Canberra. LAPP R. E.

NATIONAL

HEALTH

AND

MEDICAL

RESEARCH

COUNCIL

1995.

Recommendations for limiting exposure to ionizing radiation ( 1 9 9 5 ) and National standard for limiting occupational exposure to ionizing radiation. Radiation Health Series 39. Australian Government Publishing Service, Canberra. NERO A . V . Jr. 1979. A guidebook to nuclear reactors. University of

123

California Press, Berkeley. NEWTON A. W . , W I L S O N M. A. & HARRIS J . 1 9 8 8 . Olympic Dam—the first decade. Mineral Resources Review South Australia 156, 5-26. OLYMPIC DAM MARKETING PTY LTD 1 9 8 7 . Olympic Dam, Issue 4. Finsbury Press, Adelaide. O ' N E I L B . 1 9 8 2 . In search of mineral wealth—the South Australian Geological Survey and Department of Mines to 1944. Department of Mines and Energy South Australia Special Publication 2. PARER W. 1998. Media release 98/276, 18 February 1998, Minister for Resources and Energy, Canberra. Rossi B. 1966. Cosmic Rays. George Allen and Unwin Ltd., London SOUTHERN CROSS RESOURCES AUSTRALIA PTY LTD 1 9 9 7 . Honeymoon Uranium Project—Declaration of Environmental Factors. Toowong, Queensland. TAYLOR S . R . 1 9 9 0 . Continental crust—not mere scum of the Earth. Nature 346, 6 0 8 - 6 0 9 . TOUSSAINT L . F. 1 9 9 8 . Radon concentrations in Western Australian homes. Radiation Protection in Australasia 1 5 , 1 5 - 1 9 . TOWNSEND I . 1 9 9 7 . M E S A inspection of U S A in situ leaching uranium operations. MESA Journal 7, 24-25. TOWNSEND J . 1 9 9 8 . Uranium in South Australia—an update. MESA Journal 9, 15. UNSCEAR 1993. Sources and effects of ionising radiation. United Nations Scientific Committee on the Effects of Atomic Radiation, United Nations, New York. WALKER P. M. B. (Editor) 1991. Chambers Science and Technology Dictionary. W & R Chambers Ltd, Edinburgh. W I L S O N R . & JONES W . J . 1 9 7 4 . Energy, ecology and the environment. Academic Press, New York. W M C OLYMPIC DAM CORPORATION PTY LTD 1 9 9 7 . Olympic Dam Expansion Project—Environmental Impact Statement. Volume 1. Finsbury Press, Adelaide. YEELES R . 1 9 9 9 . Olympic Dam expansion project opened by the Prime Minister. MESA Journal 13, 5.

Web sites CANDU reactors and nuclear power generation in Canada: http://www.ncf.carleton.ca/~cz725/ Uranium Information Centre, Melbourne: www.uic.com.au US Environmental Protection Agency, Radiation Protection Division: www.epa.gov/radiation/ US Nuclear Regulatory Commission: www.nrc.gov Received 5 October 1999; accepted 22 September 2000

124

R. Major

APPENDIX 10.1: GLOSSARY Units The following concepts of radioactivity are expressed in Systeme Internationale (SI) units. Bequerel (Bq) is the unit of measurement of radioactivity. It is one radioactive disintegration per second (Cember 1983 p.75). Gray (Gy) is the unit of measurement of absorbed dose. Radiation damage depends on the absorption of energy from radiation. One gray = an absorbed dose of one joule of energy from radiation per kilogram of substance (e.g. human tissue) (Cember 1983 p. 135). Sievert (Sv) is the unit indicating the biological damage caused by radiation. Different types of radiation damage (human) tissue at different rates, e.g. alpha radiation causes about 20 times the amount of damage than does gamma radiation for each unit of length of tissue pene-

trated. For instance, one gray of gamma or beta radiation has 1 Sv of biological effect, whereas 1 gray of alpha radiation has 20 Sv effect, and 1 gray of neutrons has 10 Sv effect (Cember 1983 pp.197,198). Radiation protection dose limits The following dose limits exclude exposure to natural background radiation and to medical exposure to radiation, i.e. these limits apply to industrial use of radiation. The limits have been set by the International Commission on Radiological Protection (ICRP) in 1991. The annual effective dose limit for members of the public is 1 mSv/y. However, a higher value could be allowed in a single year provided that the average over five years was 1 mSv/y. The annual effective dose limit for people employed in occupations involving exposure to radiation is 20mSv/y, averaged over a period of five consecutive calendar years, with a limit of 50 mSv in any one year (National Health and Medical Research Council 1995 Schedule A, Schedule 1)

THEME 4

WATER AND SEDIMENTARY BASINS

Geological Society of Australia Special Publication 21, 127-143

CHAPTER 11 —Hydrogeology and environmental geology of the Great Artesian Basin, Australia M. A. HABERMEHL Bureau of Rural Sciences; Land and Water Sciences Division, PO Box E 11 Kingston, ACT 2604, Australia. The Great Artesian Basin is a confined groundwater basin that underlies arid and semiarid regions across 1.7 million km2 or one-fifth of Australia. The basin's groundwater resources were discovered around 1880 and their development allowed the establishment of an important pastoral industry. Pastoral activity, town water supplies, mining and petroleum ventures are all totally dependent on artesian groundwater. The Great Artesian Basin is a multilayered confined aquifer system, with aquifers in Triassic, Jurassic and Cretaceous continental quartzose sandstones. Intervening confining beds consist of siltstone and mudstone; Cretaceous marine sedimentary rocks form the main confining unit. The basin is up to 3000 m thick and is a large synclinal structure, uplifted and exposed along its eastern margin and tilted southwest. Recharge occurs in the eastern margin, an area of relative high rainfall, and the western margin in the arid centre of the continent receives minor recharge. Regional groundwater movement is towards the southern, southwestern, western and northern margins, where springs discharge and produce carbonate mounds. Potentiometric surfaces of the Triassic, Jurassic and Lower Cretaceous aquifers are still above ground level, but pressure drawdowns of up to 100 m were recorded from 1880 to the 1990s in developed areas. Some waterbores ceased to flow, necessitating groundwater to be pumped. About 4700 flowing artesian waterbores were drilled in the main Lower Cretaceous -Jurassic aquifers at depths of up to 2000 m, but average 500 m. About 3100 controlled and uncontrolled artesian waterbores remain flowing with an accumulated discharge of 1500 ML/day. About 20 000 non-flowing artesian waterbores, generally using windmill-operated pumps, tap shallower Cretaceous aquifers. Groundwater quality of the Lower Cretaceous-Jurassic aquifers is good at 500 to 1500 mg/L total dissolved solids. Groundwater is suitable for domestic, town water supply and stock use, though unsuitable for irrigation in most areas. Groundwater temperatures at the boreheads range from 30° to 100°C, and are a potential geothermal energy source, but spring temperatures are 20-45°C. The basin comprises abundant hydrocarbon reservoir (and some source) rocks, and commercial and subcommercial oil and gas is produced from Jurassic and Cretaceous sandstones, contradicting earlier beliefs that the basin-wide groundwater throughflow had flushed out hydrocarbons. Environmental geology issues relate to the development of the artesian groundwater resources of the Great Artesian Basin and include aspects of sustainable groundwater use and groundwater and rangelands management. Extraction of artesian groundwater during the last 120 years has affected the basin to varying degrees through large-scale drawdowns, which reduced artesian pressures and reduced discharges from artesian waterbores. Recent programs aim to rehabilitate waterbores in poor condition and equip bores with control valves. The replacement of the inefficient open-earth drain-distribution system, which causes up to 95% wastage of the water, with a piping system is encouraged. These measures will benefit groundwater management and rangeland management and assist to alleviate land degradation and plant and animal pest problems. Groundwater use by the petroleum and mining industries during the last 20-35 years affects some parts of the basin. The South Australian part of the basin is an example where the combination of groundwater exploitation for the pastoral industry, town and homestead water supplies and petroleum and mining industries impact on the basin's groundwater conditions and on the natural artesian springs. KEY WORDS: artesian basins, environmental geology, groundwater assessment, groundwater management, hydrogeology, water chemistry.

INTRODUCTION The Great Artesian Basin is one of the larger artesian basins in the world and is Australia's largest and most important groundwater resource. The Great Artesian Basin extends across 1.7 million km 2 or 22% of Australia, and underlies parts of Queensland, New South Wales, South Australia and the Northern Territory (Figure 11.1). The basin underlies arid and semiarid regions and consist mainly of lowlying interior plains and is largely within Australia's rangelands. Highly variable and unreliable rainfall ranges

from an annual average of 600 mm near the eastern margins to less than 100 mm/y near the southwestern parts of the basin, where evaporation reaches almost 4000 mm/y. The Great Artesian Basin is a confined groundwater basin comprising aquifers in quartzose sandstones of continental origin and Triassic, Jurassic and Cretaceous ages (Habermehl 1980, 1996b; Habermehl & Lau 1997). The aquifers in the Great Artesian Basin are sheet-like sandstone deposits, which extend across parts or the entire basin. The Lower Cretaceous - Jurassic and Triassic aquifers alternate with confining beds of siltstone and mudstone with low per-

128

M. A. Habermehl regions, where surface water is sparse and unreliable. The discovery of the basin's artesian groundwater resources around 1880, made European-style settlement possible, and led to the establishment of an important pastoral industry. Pastoral activity and town water supplies are to a very large extent dependent on artesian groundwater in the basin area. In recent years artesian groundwater has been used increasingly in the mining (since the 1980s and 1990s) and petroleum (since the 1960s and 1970s) industries located both inside and outside the basin. Most of these industries are largely or totally dependent on the basin's artesian groundwater resources, including the Olympic Dam copper-uranium-gold-silver mine and the associated town of Roxby Downs in South Australia, several copper-gold and silver-lead-zinc mines in the northwest margin of the basin, and oil and gas production in northeast South Australia, southwest and southeast Queensland

Figure 11.1 Australia.

Location and extent of the Great Artesian Basin,

meability. Overlying the Lower Cretaceous - Jurassic aquifers is the main confining unit, a thick argillaceous sequence of sediments of marine origin and Early Cretaceous age, and these mudstones are overlain by confined aquifers of Early to Late Cretaceous age (Figure 11.2). The basin is up to 3 0 0 0 m thick and forms a large synclinal structure, uplifted and exposed along its eastern margin, and tilted southwest. Recharge occurs mainly in the eastern marginal zone, an area of relatively high rainfall, and large-scale regional groundwater movement is generally towards the southwestern, southern, western and northern margins. Recharge also occurs in the western margin of the basin, and groundwater flow directions are towards the southwestern discharge margin. Natural discharge occurs in those areas from flowing artesian springs, most of which have built up mound-shaped deposits of sediments or carbonates. Discharge from the artesian aquifers near the basin margins also occurs by diffuse leakage where the overlying confining beds are thin. Many springs are associated with structural impediments, such as faults, folds, monoclines and intersecting lineaments, and occur at abutment of aquifers against bedrock or where confining beds thin near the discharge margins. Abundant artesian groundwater supplies of good quality are obtained from flowing artesian waterbores and from pumped artesian waterbores in the basin. Groundwater in the most exploited aquifers in the Lower Cretaceous - Jurassic sequence generally contains about 5 0 0 - 1 5 0 0 mg/L total dissolved solids. It is of good quality, making it suitable for domestic and town water supply, stock use in the pastoral industry and water supplies for the mining and petroleum industries. Use of artesian groundwater in some areas for irrigation is also on the increase, though generally most of the artesian groundwater is unsuitable for irrigation because in much of the basin area it is chemically incompatible with the dominantly montmorillonitic clay soils. The Great Artesian Basin underlies arid and semiarid

Hydrocarbon source and reservoir rocks are abundant in the Mesozoic sedimentary sequence of the basin, and commercial and subcommercial oil and gas discoveries have been made in several Jurassic and Cretaceous sandstones (and in underlying Permian and Triassic basin sediments). These contradict earlier beliefs that the basin-wide groundwater flow had flushed hydrocarbons out of the system. Dissolved hydrocarbons in the artesian groundwater are generally dry gases and are useful petroleum-exploration indicators.

GEOLOGY The hydrogeological Great Artesian Basin comprises the sedimentary Eromanga, Surat and Carpentaria Basins and parts of the Bowen and Galilee Basins (Figure 11.3 on Plate 4) (Habermehl 1980). The geology of the basins has been reviewed in Habermehl (1980, 1986, 1996b). The constituent sedimentary basins are continuous across shallow ridges and platforms of older sedimentary, metamorphic and igneous rocks. The basin consists of several broad northeast-trending synclinal structures, overlying sedimentary, metamorphic and igneous rocks of pre-Jurassic or preTriassic ages. The Mesozoic sedimentary sequence in the central part of the basin reaches a maximum total thickness of about 3 0 0 0 m. Parts of the marginal areas of the basin have been eroded, in particular along the eastern border, which was uplifted during Cenozoic times. Sheet-like, conformable rock bodies extend relatively unchanged for hundreds of kilometres and are almost horizontal. The Great Artesian Basin is an asymmetrical basin elongated northeast-southwest and tilted towards the southwest. Four centres of basin subsidence are present, two coinciding with the Surat and Carpentaria Basins, and two within the Eromanga Basin, separated by the Birdsville Track Ridge and overlying the Permian and Triassic Cooper and Pedirka Basins (Habermehl 1980; Habermehl & Lau 1997). Cenozoic and earlier uplift along the eastern margin and subsidence in several parts of the basin, particularly in the central and southwestern parts, led to the basin's asymmetry. Many of the near-surface folds, particularly monoclinal features, grade downwards into faults and are the product of draping and differential compaction of the sediments over

Great Artesian Basin hydrology

UTHOSTRAT1GRAPHIC UNIT

UTHOLOGY

AGE

DEPTH AQUIFER _ _ (m) AQUIFERS MODEL r

jHS

129

Potentiometric surface "*ofL. Cretaceous, Jurassic and Triassic aquifers •

Potentiometric surface of Cretaceous aquifers Regional watertable

WINTON FM

'Cretaceous aquifers' Aquifers tapped generally - produce non-flowing artesian waterbores with poor quality groundwater

MACKUNDA FM ALLARU MDST

TOOLEBUC FM~ ig zldi < CD

COREENA MBR DONCASTER MBR

-uxnn[A-OWIE FM HOORAY SST

WESTBOURNE FM

UJ 0_ ZliJD DwO 'O CCCD ECBIRKHEAD FM

'L. Cretaceous, Jurassic and Triassic aquifers' of the Great Artesian Basin. Aquifers tapped generally produce flowing artesian waterbores with good quality groundwater

HUTTON SST EVERGREEN FM~

PRECIPICE SST I (0

MOOLAYEMBER FM

O 2

CLEMATIS SST

cr ,1 I- ClO M

C/3 JO

Schematic Not to scale Evl^'il

Sandstone

*»

Conglomerate

I'..", j Siltstone

| X X | Calcareous

}lZn|

Weathered/ chemically altered

Conformable

11 | | j Confining bed «/ww t

Mudstone

— j

Coal

| Confined

aquifer

Unconformable Artesian

contact (erosional)

contact

head

19/A/554

Figure 1 1 . 2 Lithostratigraphic column of the Great Artesian Basin, simplified aquifer groups and potentials of the main aquifers (after Habermehl 1980)

fault-bounded basement blocks. Several major fault and fold systems occur in the basin, in places forming en echelon structures. Throws of up to 300 m affect some niajor faults, though the displacement of Jurassic-Cretaceous sediments along normal faults is usually much less (Figure 11.4 on Plate 5). The stratigraphic succession and the distribution and correlation of the rock units and their equivalents of Middle Triassic to Late Cretaceous ages in the constituent sedimentary basins in the Great Artesian Basin are given in Figure 11.3 (on Plate 4) and Habermehl (1980, 1986) and Habermehl and Lau (1997). Figure 11.2 shows the sedimentary sequence and hydrogeological units in the central Eromanga Basin part of the Great Artesian Basin. The Jurassic sequence comprises continental quartzose sandstone, with lesser siltstone and mudstone. Siltstone, mudstone and lithic sandstone were deposited in shallow-marine environments during Early Cretaceous times. During the Late Cretaceous more sandy sediments were laid down in lacustrine and fluvial sedimentary environments. The Eromanga Basin is deepest where it overlies Palaeozoic and older Mesozoic sedimentary basins. Thinner sequences are present across the shallow ridges and platforms connecting the Eromanga Basin with the Surat and Carpentaria Basins (Figure 11.3 on Plate 4). The southeastern parts of the Great Artesian Basin includes the

sedimentary Surat Basin and the Coonamble Embayment, and consist of an alternation of Jurassic continental sandstone, siltstone, mudstone and some coal. The Cretaceous sediments are partly continental, but mainly shallowmarine lithic sandstone and mudstone. The Carpentaria Basin contains continental rocks of Jurassic age and marine sedimentary rocks of Cretaceous age. The deeply weathered erosional surface of the Cretaceous sedimentary rocks in these basins are overlain by Tertiary sedimentary rocks, which are also partly weathered and silicified, and by mostly unconsolidated Quaternary sediments. The latter overlie parts of the Great Artesian Basin and are usually up to several tens of metres in thickness, but form shallow basins as much as 150 m deep in some regions. Tertiary basalts cover some areas of Mesozoic rocks in the northeastern, eastern and southeastern parts of the basin.

HYDROGEOLOGY The confined aquifers of the Great Artesian Basin are present in a rock sequence which, where complete, is bounded by the Rewan Group at the bottom and the Winton Formation at the top (Habermehl 1980, Habermehl & Lau 1997) (Figure 11.2).

130

M. A. Habermehl

Aquifers are present in the Clematis, Precipice, Hutton, Adori and Hooray Sandstones, and the Cadna-owie Formation and their equivalents, and in the Mackunda and Winton Formations. Most of the individual aquifers are relatively uniform in their hydrogeological characteristics over large areas, and they are continuous and hydraulically connected across the constituent geological basins (Figures 11.2, and 11.3 on Plate 4, 11.4 on Plate 5). The major confining beds consist of the Rewan Group, Moolayember, Evergreen, Birkhead, Westbourne, Wallumbilla and Toolebuc Formations, and their equivalents, and the Allaru Mudstone and parts of the Mackunda and Winton Formations. The hydrogeological basement underlying the basin comprises impervious sedimentary, metamorphic or igneous Mesozoic, Palaeozoic and Proterozoic rocks, and this basement forms in part an aquiclude or aquifuge. The hydraulic characteristics of the confined aquifers have been determined since the early development of the basin from a large number of periodic tests carried out on the flowing artesian waterbores, and during the last 30 years from rock samples, and from wireline logs obtained from 1500 waterbores (Habermehl 2001), and from some of the 3500 petroleum exploration and production wells in the basin. Hydraulic conductivity values of the aquifers range from 0.1 to 10 m/day, the majority being in the lower part of that range, and mainly relate to the Cadna-owie Formation and Hooray Sandstone and their equivalents including the Algebuckina and Pilliga Sandstones (Habermehl 1980). Transmissivity values measured from periodic systematic tests by the State Water Authorities, range from 1 to 2000 m2/day. Storage coefficient values, as calculated from petroleum log data, range from 10~4 to 10~5. Intrinsic permeability ranges from several tens to several thousands of millidarcys. Porosity values range from 10 to 30%. Average vertical hydraulic conductivities of the leaky, very low permeability, confining beds range from 10 - 1 to 10~4 m/day. Hydraulic gradients of the potentiometric surfaces of the aquifers in the Lower Cretaceous - Jurassic sequence range from 1: 2000 in the eastern to 1: 4000 in the central-southwestern parts of the basin. Hydraulic gradients of the aquifers in the upper part of the Cretaceous sequence are about 1: 1800 (Habermehl 1980).

RECHARGE Recharge of the aquifers by infiltration of rainfall and through creeks into the outcropping aquifer sandstones and through unconsolidated sediments overlying the aquifers occurs mainly along the eastern, elevated, margins of the basin, located on the western slope of the Great Dividing Range. Recharge to the western margins of the basin, in the arid centre of the continent, where aquifers are exposed or overlain by sandy sediments, and through the beds of the Finke River, and possibly the Todd, Hale, Plenty and Hay Rivers, takes place when these rivers flow following highintensity short-duration cyclonic summer rainfall, which occurs once every few years or once a decade or less. Prior to development, the basin was in a natural steadystate condition, with an equilibrium between recharge and

natural discharge from springs and vertical upward leakage. Following development, natural discharge diminished. A visible effect has been the diminution in flow from springs in the south-central, southwestern and northern parts of the basin. Abstraction by waterbores caused a large-scale lowering of the potentiometric surface and a steepening of the hydraulic gradient, which allowed more recharge water to enter the system. Recharge has been estimated from modelling to have increased from 2200 ML/day to 3000 ML/day since the development of the basin (Habermehl 1980, Seidel 1980). At present a new approximate steadystate condition has been reached in which total recharge and discharge are approaching equilibrium again, and the sum of the discharges (Figure 11.5) and the vertical leakage are assumed to equal the recharge. The discharge from waterbores equals about half of the recharge to the basin, and based on rainfall and the areal extent of the aquifer outcrops, equals about 1% of the average annual amount of water available for recharge (Habermehl 1980). An approximate water balance for the Great Artesian Basin is: Recharge ( - 3 0 0 0 ML/day) = Discharge ( - 1 7 5 0 ML/day, including springs and waterbores: Figure 11.5) + Vertical Leakage ( - 1 2 5 0 ML/day, including some subsurface outflow) The directions of flow of the groundwater may be inferred from the potentiometric surface of the confined groundwater in the aquifers of the Lower Cretaceous Jurassic rock sequence (Figures 11.6, 11.7). Environmental isotope and hydrochemical studies of groundwater from drillholes in the recharge areas, and from waterbores located downgradient of the recharge areas and towards the centre of the basin, have confirmed the increase in residence times and determined the flow rates and flow patterns of the artesian groundwater predicted from the hydrogeological analyses and the potentiometric surfaces (Airey etal 1979, 1983; Calf & Habermehl 1984; Bentley et al 1986; Torgersen et al 1991; Herczeg et al 1991; Habermehl et al 1993; Cresswell et al 1996; Radke et al 2000; Love et al 2000). These studies showed that the artesian groundwater is of meteoric origin and also support an assumption of continuing recharge from geological to modern times. More detailed studies of the recharge areas and quantification of the recharge are required and hydrogeological, hydrochemical and isotope investigations have been carried out during the 1990s and are continuing.

DISCHARGE Discharge from the Great Artesian Basin aquifers takes place as natural discharge in the form of concentrated outflow from springs, vertical leakage from the Lower Cretaceous - Jurassic aquifers towards the Cretaceous aquifers and upwards to the regional water-table, subsurface outflow into neighbouring basins, and as artificial discharge by means of free or controlled artesian flow and pumped abstraction from waterbores drilled into the aquifers (Habermehl 1980). Diffuse discharge from the artesian aquifers through the confining beds towards the

Great Artesian Basin hydrology

131

Great Artesian Basin Groundwater Discharge and Use

54 ML/d (3%) Springs Dalhousie Springs 76 MLVd (4%) Springs - total other GAB springs

35 ML/d (2%) Mining (Olympic Dam/Roxby Downs)

1200 ML/d (70%) Pastoral bores flowing artesian waterbores

70ML/d (4%) Oil and gas

300 ML/d (17%) Pastoral bores non-flowing artesian waterbores

Total 1735 ML/day - 633 275 ML/yr

Figure 11.5 Groundwater discharge and use in the Great Artesian Basin. ground surface occurs in the marginal areas where the confining beds are relatively thin, potentials are high, and water-tables shallow (Woods et al 1990).

Natural discharge Springs and areas of seepage are abundant in the marginal areas of the basin, particularly in the southern, southwestern, northwestern and northern areas. Most springs are concentrated in groups, covering relatively small areas. Eleven groups have been identified in the main part of the basin (Habermehl 1982), and several springs exist in the far northern part of the basin on Cape York Peninsula. Rates of discharge from the springs are generally low, and range from less than 1 Us to about 150 Us (the latter from a spring at Dalhousie Springs, northern South Australia: Figure 11.12). The temperatures of the springwater range from about 20° to 45°C. Total spring discharge is estimated at 130 ML/day (Figure 11.5), with 54 ML/day being produced from the Dalhousie Springs. Springs are quite common in the recharge areas along the eastern margins, but most of these springs are the result of 'overflow' or the 'rejection' of recharge into the aquifers, or result from the intersection of the local topography and aquifers. Flowing artesian springs within the basin and in the discharge margins of the basin are generally associated with structural features, such as faults, folds, monoclines and intersecting lineaments. Upward groundwater flow along faults is the source of many springs, and also the abutment of aquifers against impervious bedrock, and pressure-water breaking through thin confining beds near the discharge margins of the basin. Many springs have built up conical mounds several metres to several tens of metres in diameter, and up to several metres high. The mounds are formed by deposition of particles brought up from the aquifers and

the confining beds, and by the chemical and biological precipitation of solids dissolved in the artesian groundwater. The artesian springs are terminal evaporitic systems, usually comprising a central carbonate mound, with outer zones dominated by sulfate and chloride salts. Mound morphology is controlled by several factors, including groundwater discharge rates, hydrochemistry, evaporation, influence of inorganic versus organic carbonate precipitation, local subsidence of the mound and microtectonics. In some areas the mounds consist mainly of particles brought up from the confined aquifers and the confining beds, and form mud mounds and mud volcanoes. Many mounds, particularly those built by springs in the western and southwestern margin of the Great Artesian Basin, consist of carbonate. The latter are dominated by calcite and dolomite, and occur as tufa, travertine and very fine-grained or crystalline carbonate that was deposited as a chemical precipitate out of the artesian groundwater and precipitated by a combination of chemical, algal and bacterial action. Terraced mounds and waterfall or cascade deposits produced by algae are common, though many accumulations consist of steeply sloping mounds. Artesian springs and their deposits in the Lake Eyre region in the southwestern part of the basin (Figure 11.12) range from topographically high springs to younger, topographically low springs as a result of the lowering of the land surface and spring outlet levels in Quaternary times (deposits of extinct pre-Quaternary springs occur more than 40 m above the present springs). This also indicates that the potentiometric surface in the Lake Eyre region has declined considerably during recent geological time (Habermehl 1982). Morphological diversity and lithofacies patterns indicate that the spring complexes have developed over several climatic cycles. The dated ages of spring deposits range up to 700 000 years, with some spring deposits probably being older. The ages of several basal spring deposits, as determined from 14 C, U/Th, thermoluminescence and palaeo-

CO h i C

ET o n> rt> sO

PNG

O n

X Q CT

CL o r? D

ai o

Gulf

0 3

o* o 3 ro

Carpentaria

cd u

o* CO c

3 ai cr. o D I

o o

ST o <T> vO o\ O

CORAL SEA

1960 potentiometric surface contours (m) Cadna-owle Formation / Hooray Sandstone aquifers —too— Contour (m) above MSL j

1970 potentiometric surface contours (m) Cadna-owle Formation / Hooray Sandstone (and other GAB) aquifers -100 - Contour (m) above MSL

n en ai tu CL D P- w O i ro rt> JU ^Tl o Do 3 su Ef <T> o ' "I Q o

Northern Territory

1 E o

> o

• ASceSprSigs

« Alice Springs

a> CO 5 ' so CL Cd C O co o D 5* O EU -O X> nT (75 on CO CO r?* <T> r? as CT*

n> CO :r K)

3 o

H-1 S / \Q

00 g O • -a o

3 CD

South Australia J Port Augusia

Great Artesian Basin hydrology

133

19/A/558

Figure 11.7 Regional groundwater flow directions in the Cadna-owie Formation - Hooray Sandstone aquifers, recharge and discharge areas, and locations of flowing artesian springs in the Great Artesian Basin (after Habermehl 1980; Habermehl & Lau 1997).

134

M. A. Habermehl

magnetic studies suggest that some springs might have been (re-) activated as a result of major climatic changes. Fossil carbonate spring deposits of Pleistocene age, consisting of very fine-grained carbonate with abundant reed casts, gastropod shells and algal structures, rise several tens of metres above the present land surface where the present active springs (and Holocene spring deposits) occur. The higher, older spring carbonates cap circular mesas and hills and overlie pedestals of Cretaceous mudstone. Vertical leakage from the aquifers upwards through the semipervious confining beds occurs throughout the basin, and despite the low percolation rates involves a considerable volume of water (approximately 1250 ML/day), which constitutes a major part of the groundwater throughflow of the basin. A deep phreatic surface, usually at several tens of metres (60-80 m) below the ground surface, generally conceals the vertical leakage.

Artificial discharge Groundwater in the Great Artesian Basin has been exploited from flowing artesian waterbores since artesian water was discovered in 1878, allowing an important pastoral industry to be established. Waterbores are up to 2000 m deep, but average about 500 m. Artesian flows from individual bores exceed 10 ML/day (more than 100 L/s), but the majority have much smaller flows. About 3100 of the 4700 flowing artesian waterbores drilled in the basin, remain flowing. The accumulated discharge of these waterbores (including water-supply bores in about 70 towns, as in most cases the artesian groundwater supply is the only source of water) is about 1200 ML/day, compared to the maximum flow rate of about 2000 ML/day from about 1500 flowing artesian waterbores around 1918. Flowing artesian bores obtain their groundwater from aquifers in the Lower Cretaceous and Jurassic sequence (mainly the aquifers in the Cadna-owie Formation, Hooray, Algebuckina and Pilliga Sandstones and their equivalents). The original non-flowing artesian waterbores generally tap the aquifers in the Winton and Mackunda Formations. These non-flowing bores, which number about 20 000, are generally shallow (i.e. several tens to hundreds of metres deep). It is estimated that these generally windmill-operated pumped waterbores supply on average 0.01 ML/day, and produce a total of about 300 ML/day (Figure 11.5). High initial flow rates and pressures of artesian waterbores have diminished as a result of the release of water from elastic storage in the groundwater reservoir, and approach a steady-state condition in many areas. Exploitation of the aquifers has caused significant changes in the rate of various discharges in time (Habermehl & Seidel 1979; Habermehl 1980; Seidel 1980). Spring discharges have declined as a result of waterbore development in many parts of the basin during the last 100 years and in some areas springs have ceased to flow.

GROUNDWATER MOVEMENT Regional lateral groundwater movement in the aquifers in the basin has been interpreted from the potentiometric surface maps of the aquifers in the Jurassic and Lower

Cretaceous sequences (Figures 11.6, 11.7). Flow directions are generally towards the south, southwest, west and north. In the western part of the basin regional groundwater movement is towards the southeast and south. Groundwater movement is slow, and based on hydraulic data probably around 1-5 m/y, as hydraulic conductivities and gradients are low and porosities high (Habermehl 1980). Groundwater flow rates based on carbon-14 and chlorine-36 studies range from <1 m/y to ~5 m/y (Calf & Habermehl 1984; Bentleyefa/. 1986; Torgersen etal 1991; Radke et al 2000). Groundwater residence times determined from carbon14 and chlorine-36 studies range from several thousands of years near the recharge areas to more than one million years near the centre of the basin. The flow rates and groundwater residence times in the Lower Cretaceous Jurassic Hooray Sandstone aquifer (and its equivalents) as calculated from hydraulic data are consistent with the residence times derived from environmental isotope studies carried out on artesian groundwater from flowing artesian waterbores tapping this aquifer throughout the basin (Airey et al 1979, 1983; Calf & Habermehl 1984; Bentley et al 1986; Torgersen et al 1991; Herczeg et al 1991; Habermehl et al 1993; Cresswell et al 1996; Radke et al 2000).

The potentiometric surfaces of the confined aquifers in the Lower Cretaceous - Jurassic sequence were above the ground surface over almost the whole of the basin before exploitation began around 1880. Since then the regional potentiometric surface of the exploited aquifers in the sequence has dropped by several tens of metres in many heavily developed areas (Figure 11.8). It is still above ground level in most of the basin, though in some areas flows from artesian waterbores ceased and water has to be pumped. The potentiometric surface of the confined aquifers in the upper part of the Cretaceous sequence (Winton and Mackunda Formations: Figure 11.2) has always been below the ground surface, consequently waterbores tapping these aquifers are non-flowing artesian and have to be pumped. The confined aquifers in the Lower Cretaceous Jurassic sedimentary sequence are sheet-like deposits, which are relatively uniform and extend for hundreds of kilometres. The aquifers are continuous across shallow ridges and platforms of older rocks. In some areas faults locally displace or disconnect aquifers, and obstruct part or all of the groundwater flow in the main Lower Cretaceous Jurassic aquifers, which is normally directed to these structures. These faults could act as permeable or impermeable barriers either to groundwater or to hydrocarbons migrating in the sandstones (Senior & Habermehl 1980). Other impermeable barriers could occur in the aquifers and be barriers of stratigraphic or diagenetic origin. Part of the north-south-striking Canaway Fault (east of Canaway 1 in Figure 11.4 on Plate 5) appears to be a preferential permeable zone along which groundwater moves from Jurassic aquifers upwards into Cretaceous aquifers (Habermehl 1986). Upward vertical leakage from the Lower Cretaceous - Jurassic aquifers into the aquifers of the Winton and Mackunda Formations probably accounts for a substantial part of the groundwater discharge from the former aquifers and influences groundwater movement.

Great Artesian Basin hydrology •

135

250 km

Gulf of Carpentaria CORAL SEA

Regional drawdown (m) of the potentiometric surface 1880-1970

0-10 10-20 20-30 30-40 1 1 1 40-60 tlltlti

18° —

60-80 80-100 100-120 1

Northern Territory • Alice Springs

South Australia

Broken Hill • Port Augusta

New South Wales 19/A/560

Figure 11.8 Regional drawdown of the potentiometric surface of the Cadna-owie Formation - Hooray Sandstone aquifers in the Great Artesian Basin for the period 1880-1970 (after Habermehl 1980; Habermehl & Lau 1997).

136

M. A. Habermehl

Great Artesian Basin S 2 H VS 5 18 0

0 -10

-20

5 o s C/5

X CM CO

-30 -40 -50 -60

-70 Figure 11.9 Deuterium - oxygen 18 stable isotope composition of groundwater from flowing artesian waterbores in the Great Artesian Basin.

Pressure changes Potentiometric maps showing the conditions during the early years of development and the 1970s for the main aquifers in the Lower Cretaceous - Jurassic sequence, which produce flowing artesian wells, are given in Habermehl (1980). These maps were produced from the results of the large number of periodic measurements carried out on the flowing artesian waterbores since the early development of the basin and computer simulation modelling of the hydrodynamics of the basin (Seidel 1980). Development of the artesian groundwater resources has led to considerable changes in the patterns of the potentiometric contours and the changes in water levels as a result of the regional drawdowns are shown in Figure 11.8. Predicted drawdowns, changes in discharges and predicted potentiometric surface maps as a result of possible future developments are shown in Habermehl and Seidel (1979) and Seidel (1980). Pressure drawdowns of up to 100 m were recorded between 1880 and the 1970s-1980s (Figure 11.8), and also reflect the location and density of the flowing artesian waterbores in those regions (Habermehl 1980; Habermehl & Lau 1997). Recent MODFLOW modelling of the 1960 steady-state condition for the Cadna-owie Formation - Hooray Sandstone aquifers across the entire basin has produced a new potentiometric surface map for these aquifers (Figure 11.6).

ISOTOPE HYDROLOGY Isotope hydrology studies on the artesian groundwater from the Lower Cretaceous - Jurassic aquifers in the Great Artesian Basin have confirmed recharge areas and regional groundwater flow patterns, and the sources and origin of the artesian groundwater.

Studies of stable isotope ratios 8 2 H and 8 l s O show that nearly all values plot on the meteoric-water line and that the artesian groundwater is meteoric in origin (Figure 11.9), which was a contentious issue during the early 1900s. Samples downgradient from the recharge areas show decreasing percentages of modern carbon, from high values in the exposed aquifers in the recharge areas to background levels further downgradient in the basin (Figure 11.10a), and carbon-14 derived isochrones have been determined (Calf & Habermehl 1984). The carbon isotope ratio 8 1 3 C increases basinwards, as does alkalinity, and the variation of S 1 3 C is closely related to the changes in HC0 3 (Calf & Habermehl 1984; Herczeg et al 1991). The age and residence times of the artesian groundwater and the groundwater movement rates and flow patterns in the Great Artesian Basin have also been determined with carbon-14 and chlorine-36 isotopes (Figure 11.10), and confirm the regional groundwater movement rates, directions and patterns determined from hydrogeological data (Figure 11.7). The data points include approximately 350 1 4 C and approximately 360 36 C1 analyses from more than 800 samples (analysed for stable isotope ratios 8 2 H, 8 l s O and 8 1 3 C and for detailed hydrochemistry results) collected from waterbores and springs in the basin by the author (listed in Radke et al 2000). Chlorine-36 has a longer half-life than carbon-14 (310 000 y against 5730 y), chloride and its chlorine isotopes are highly soluble in water, chloride has a relatively conservative behaviour in groundwater and has a simpler geochemistry than carbon, including 14 C, carbonates and carbon compounds. The application of 36C1 as a dating tool for the very old groundwater in the Great Artesian Basin has therefore many advantages and is well suited for the high ages of the slow moving artesian groundwater in this large groundwater basin, as it has a potential range of more than 2 million years.

Gulf

of Carpentaria

of Carpentaria CORAL SEA

CORAL SEA

Figure 11.10 (a) Contours of C pmC in the Great Artesian Basin, (b) Contours of C1/C1 ratios in the Great Artesian Basin.

Northern Territory

Northern Territory Queensland

New South Wales

Great Artesian Basin hydrology

South Australia

137

138

M. A. Habermehl

Groundwater ages from the Great Artesian Basin determined from the chlorine-36 results (Figure 11.10) are in general in good agreement with the calculated ages (Calf & Habermehl 1984) suggesting that the flow conditions of the artesian groundwater have remained largely unchanged during at least the last one million years (Bentley et al 1986; Torgersen et al 1991; Habermehl et al 1993; Cresswell et al 1996; Radke et al 2000). Alternative explanations for the 36Cl-derived age have been discussed by Andrews and Fontes (1993) and Torgersen and Phillips (1993). Bethke et al (1999) modelled groundwater flow and 4 He distribution in the basin. Fabryka-Martin et al (1985) determined iodine-129 levels in the basin.

HYDROCHEMISTRY Groundwater in the most widely exploited confined aquifers in the Lower Cretaceous - Jurassic sequence generally contains about 500-1500 mg/L total dissolved solids. Artesian groundwater has pH values which are almost always between 7.5 and 8.5. The artesian groundwater is chemically of the Na-HC0 3 -Cl type, and these ions contribute more than 90% of the total ionic strength of solutes in the main basin area (Figure 11.11 on Plate 6). Evolution of the groundwater chemistry along the flowlines is characterised by the removal of Na and K by reconstitution reactions involving kaolinite, a Na-smectite and illite (Herczeg et al 1991; Radke et al 2000). In the southwestern part of the basin the groundwater is characterised by Na-Cl-S0 4 -type water, and the two regional groundwater flow directions show different hydrochemical characteristics, with westward flowing water being of the Na-HC0 3 -Cl type and eastward flowing water being of the Na-Cl-S0 4 type. These flows within the same aquifer meet and mix and are directed towards the main discharge area near the southwestern margin of the basin (Habermehl 1986). Near the recharge areas Ca, Mg and S 0 4 concentrations are proportionally higher and these decrease basinwards. Variations of the major ion concentrations and ratios occur along the flowlines. Na and HC0 3 concentrations generally increase along the flowlines in most parts of the basin. CI and S 0 4 also increase in most areas, though an initial decrease occurs basinwards of the northeastern marginal area. Along the southerly directed flowpaths from the northeastern recharge area in Queensland to the New South Wales border, concentrations of all four ions decrease and then increase or remain approximately constant. In the centre of the basin some deep waterbores with high CI concentrations occur, and very low to semi-stagnant flows along the deep and long flowpaths might account for these high CI concentrations and high salinity values. In the northwestern part Na, HC0 3 and CI increase along the flowlines but S 0 4 concentrations are constant (Habermehl 1983, 1986; Radke et al 2000). Observed systematic variations in the chloride levels could reflect variations in the rate of recharge and infiltration of recycled salt throughout the Late Quaternary. Minima and maxima in chloride concentrations correlate with the last glacial and interglacial period respectively (Airey etal 1979).

Total dissolved solids values generally show an increase downgradient in the basin, and this is probably the result of mixing of the dilute recharge water with more saline groundwater in the deeper parts of the basin, diffusion of ions out of the mudstones of the confining beds containing higher salinity water, and dissolution of evaporites, carbonate minerals or incongruent dissolution of feldspars, micas or clay minerals (Herczeg et al 1987, 1991). The increase in alkalinity and decrease in S 0 4 concentrations in the basin might result from biochemical reduction of carbon dioxide to produce methane rather than the dissolution of carbonate minerals. The aquifer system is open to C 0 2 and the addition of C 0 2 is accomplished by fermentation processes occurring in situ. Some of the added C 0 2 is a by-product of methanogenesis. The addition of C 0 2 drives the carbonate dissolution reaction and so exerts an important control on the evolution of the Na-HC0 3 groundwater in the basin (Herczeg et al 1991). Fluoride values in many parts of the basin are high with values up to 10 mg/L and more, which is a problem for domestic and stock water supplies. High fluoride concentrations in the artesian groundwater have been attributed to groundwater being in contact with igneous (hydrogeological basement) rocks (Evans 1996; Habermehl & Lau 1993; Habermehl et al 1996). Contact of the artesian groundwater with igneous rocks has been interpreted to be the reason for the occurrence of surprisingly unradiogenic 87 Sr/ 86 Sr ratios (Collerson et al 1988). Cretaceous aquifers have higher salinities and CI values than the Lower Cretaceous - Jurassic aquifers, and the high Na-Cl values in the Cretaceous aquifers probably reflect the non-flushing characteristics of these mainly isolated and lenticular-shaped aquifers. The adjoining mudstones of marine origin in the Cretaceous sequence might also contribute salts to the system. The Lower Cretaceous - Jurassic aquifers can be distinguished from the Cretaceous aquifers based on their hydrochemical characteristics, but the distinction between individual aquifers in the Lower Cretaceous - Jurassic sequence is less obvious (Muller 1989; Quarantotto 1986, 1989). Mixing of groundwater from the latter aquifers occurs where intervening confining beds pinch out and the aquifers are in contact. The upper parts of the Cadna-owie Formation and Hooray Sandstone aquifers are influenced by the downward diffusion of ions from the marine mudstones of the Rolling Downs Group. Most artesian waterbores produce varying amounts of gases. The main constituents of the gases include N 2 , C0 2 , Ar and small amounts of H 2 and He. In addition many waterbores produce artesian groundwater containing small amounts of hydrocarbons. The hydrocarbons are mainly CH4 and lesser amounts of C 2 H 6 to C 7 H 1 6 but liquid hydrocarbon fractions have also been detected (Habermehl 1986, 1989). Hydrocarbon source and reservoir rocks are abundant in the sedimentary sequence of the basin, and commercial and subcommercial oil and gas discoveries have been made in several Jurassic and Cretaceous sandstones, and in particular in the underlying Permian and Triassic sedimentary rocks, contradicting earlier beliefs that the basin-wide groundwater flow had flushed hydrocarbons out of the system. Dissolved hydrocarbons in the artesian groundwater are generally dry gases and are useful petroleum exploration indicators.

Great Artesian Basin hydrology Water quality improves with depth in the aquifers in the Lower Cretaceous - Jurassic sequence, with groundwater obtained from aquifers in the older part of the Lower Cretaceous - Jurassic sequence having better quality water than the upper aquifer. The latter underlies the main Lower Cretaceous confining bed of marine origin and is probably in part affected by diffusion of salt from these Cretaceous mudstones. Groundwater from all of the aquifers in the Lower Cretaceous - Jurassic sequence is of good quality and suitable for domestic, town water supply and stock use, though it is generally unsuitable for irrigation because in much of the basin area it is chemically incompatible with the dominantly montmorillonitic swelling clay soils. Water from the upper, Upper Cretaceous, Winton and Mackunda Formation, aquifers has a higher salinity, though it is still acceptable as stock water. Reverse osmosis desalination plants upgrade the quality of groundwater used at the town of Roxby Downs and some of the mines in the northwestern margin of the basin. Other towns and homesteads use the artesian groundwater without any treatment, usually after cooling.

GROUNDWATER TEMPERATURES Groundwater surface temperatures of waterbores tapping aquifers in the Lower Cretaceous - Jurassic sequence range from about 30° to 100°C, and springs have temperatures from about 20° to 45°C. Geothermal gradients show a wide range and give a mean of about 39°C/km, and a range of about 15°C/km to 100°C/km (Polak & Horsfall 1979), as obtained from temperature wireline logs in waterbores (Habermehl 2001). References to earlier authors on geothermal aspects are given in Habermehl (1980, 1986). Waterbore values for the geothermal gradients, though internally consistent, are too high, based on data from deeper petroleum exploration wells in the central part of the basin (Cull & Conley 1983). Geothermal gradients derived from petroleum-exploration wells are given in Pitt (1986). Recent maps of groundwater temperatures of waterbores and geothermal gradients in the basin are included in Habermehl (2001). The heat flow in the basin is attributed to heat produced in the earth's crust by uranium and thorium, and by recent volcanic activity (Torgersen et al 1992). The effects of temperature variations on the hydrodynamics of the Great Artesian Basin have long been recognised, and have been incorporated in computer-based simulation modelling of groundwater flow.

ENVIRONMENTAL GEOLOGY Environmental geology aspects of the Great Artesian Basin include issues caused by the development of the artesian groundwater resources of the Great Artesian Basin for the pastoral industry and the resultant effects on the groundwater conditions in the form of large-scale drawdowns of the potentiometric surface (Figure 11.8). The reductions in pressures and flowing artesian discharges of the waterbores following about 120 years of exploitation has affected the pastoral industry, town water supplies and homesteads, but has also resulted in changes to the flowing artesian springs,

139

many of which have become extinct (Figure 11.7). The distribution and use of the artesian groundwater on the ground surface of the basin area by the pastoral industry has created an abundance of water in an arid and semiarid region where previously water was sparse or only occurred following irregular high level (cyclonic) rainfall events. The waterbores provide large numbers of watering points together with the open earth bore drains, which distribute the water from the flowing artesian waterbores. The bore drains are up to 100 km long, and have a total length of 34 000 km. The availability of water in the arid and semiarid landscape has had a significant impact on the flora and fauna in the region, with major changes to the biodiversity (Landsberg et al 1997). The inefficient bore-drain distribution system causes wastage of more than 95% of the groundwater produced, and has also resulted in land degradation, erosion, salinisation and the spread of introduced weeds, shrubs and trees, pest animals and large increases of feral and native animals. The Great Artesian Basin Bore Rehabilitation Program, a Federal and State Government-funded assistance program commenced in 1989 (Reyenga et al 1998) and aimed to rehabilitate waterbores in poor condition and equip leaking and free-flowing artesian waterbores without control valves. The program sought to eliminate some of the wastage of water and increase artesian pressures, and provide a basis for better management of the basin. The follow-up program, the Great Artesian Basin Sustainability Initiative started in 1999 and will accelerate bore rehabilitation and bore-drain replacement programs and aims to achieve partial recovery of artesian pressures in strategic areas of the basin. The installation of polyethylene piping to replace the open boredrain distribution system will cause the largest change to water wastage through a significant reduction in demand of water produced by the bores, and lead to increases in artesian pressure and will reactivate flow from some waterbores, which have ceased flowing. The above programs are intended to maintain options for future uses of the basin's artesian groundwater, to continue access to artesian supplies by existing users and achieve some recovery of artesian pressures. In addition they will allow for improved pastoral production through greater control of total grazing pressure and improved stock and vegetation management, and also allow for the reduction in the rate of land and water resource degradation associated with open bore drains and uncontrolled extraction of groundwater. The programs also provide opportunities for improved management of waterdependent ecosystems, largely the springs of the basin. Ultimately, the changes brought by the programs will provide for better management of the artesian groundwater resource and the reduction of the demand on the resource will alleviate fears of its sustainable use. Tangible benefits will be possible from enhanced rangelands and pasture management and assist to alleviate land degradation and to control the activities and numbers of animal pests. Land-use changes in the basin, such as land clearing through the removal of trees in the recharge areas have affected the groundwater recharge to the basin's aquifers. The disposal of domestic and industrial waste in municipal and other landfills in exposed aquifer sandstones in the (eastern) recharge areas in Queensland and New South Wales may pose a threat to groundwater quality. Expanding irrigation in and near the recharge areas in

140

M. A. Habermehl Betoota

•

Kco Outcrop areas

Regional groundwater flow direction

—2 —

Drawdown contours Olympic Dam Borefield B (metres)

Jua 1

Basement • — - Boundary of the Great Artesian Basin

Pipeline n +

Boundaries designated areas of Olympic Dam Borefields

Town Spring(s) Bores in Olympic Dam Borefield

Boundary Olympic Dam Borefields $$

Mine

Artesian boundary

Figure 11.12 Groundwater and environmental features in the South Australian part of the Great Artesian Basin.

northern New South Wales and feedlots also represent threats to groundwater quantity and quality. Enhanced understanding is required of the location and extent of the recharge areas and the recharge processes, and further studies are in progress. It is vital to have close cooperation of land and (ground)water managers, landholders and the wider community. It will be necessary to provide them with an understanding of the importance of the recharge areas for the sustained infiltration of good quality rain and surface water to replenish the Great Artesian Basin aquifers. Development of the basin's artesian groundwater resources during the last 120 years by the pastoral industry has created some of the problems shown above. However, development of the artesian groundwater for the petroleum and mining industries during the last 20-35 years has brought similar problems, including drawdowns caused by the development of borefields, which affect waterbores of other users and some springs, and also has had an effect on the biodiversity of some springs (Noble et al 1998; Niejalke

1998). Continued expansion is proposed for the spa-bath tourist facilities based on pumped warm artesian groundwater from bores in Moree, in the northeastern part of the basin in New South Wales. However, limits to the extraction might be imposed. The South Australian part of the Great Artesian Basin is a region (Figure 11.12) where the combination of groundwater development for pastoral use, town water supplies, petroleum production, mining and in situ leach mining impact on the groundwater conditions and on some of the artesian springs located near the margin of the basin. The effects of the groundwater extraction for the pastoral industry, and the activities of the Bore Rehabilitation Program in South Australia have been described by Sibenaler (1996) and Sampson (1996). The influence of the groundwater extraction (Figure 11.5) by the borefields for the Olympic Dam copperuranium-gold-silver mine has been discussed by KinhillStearns (1982, 1983, 1984), Armstrong and Rowan (1986),

Great Artesian Basin hydrology

Waterhouse and Armstrong (1990) and Kinhill Engineers (1997), and the drawdown contours of the Olympic Dam Borefield B are shown in Figure 11.12. The location of Borefield B in the deeper parts of the basin, away from the basin margin has advantages, as it allows for the extraction of larger groundwater quantities without the disadvantages experienced near the basin margin (at Borefield A, close to the springs) and also provides better quality groundwater. The groundwater obtained from the Olympic Dam Borefields A and B are pumped by pipeline outside the basin to the Olympic Dam mine, where it is used for the mining activities, mineral processing and in the refinery plant. A desalination plant processes part of the water for use in the town of Roxby Downs. Other towns, including Marree and Oodnadatta, and homesteads in the region also use artesian groundwater though no desalination is carried out. As part of the petroleum production, large amounts of artesian groundwater are brought to the surface in the Moomba oil and gasfields region in northeastern South Australia (Figure 11.12), and in oil and gasfields in southwest and southeast Queensland within the basin (Habermehl & Lau 1997). The water is disposed by surface discharge or more usually by evaporation in holding basins. The water production associated with the production of petroleum provides a challenge for the petroleum industry, as the aim is to produce minimal water, to increase the efficiency of the petroleum production. Alternatively, rather than surface disposal or evaporation, the water should be reinjected into the aquifers, avoiding any possible contamination and salinisation of the land surface, and depressurisation of the aquifers. The development of in situ leach mining in the basin is a new occurrence. However, the aquifer to be leached for uranium in the Beverley uranium mine project (Figure 11.12) is a Tertiary palaeochannel sand, which is confined and separated from the Cadna-owie Formation aquifer of the Great Artesian Basin by 50-85 m of low permeability clay (Heathgate Resources 1998; Habermehl 1999). The Beverley uranium mine leaching operation will operate at almost neutral hydraulic pressure and water balance, with the injection and extraction being almost equal. The extraction is slightly more than injection, to maintain an inflow of groundwater from the surrounding aquifer and prevent excursions of the mining solution away from the mining area. The artesian pressure of the Cadna-owie Formation aquifer is upwards, and above ground level, providing a barrier against any possible downwards leakage. The Honeymoon uranium mine is another uranium deposit in a Tertiary palaeochannel sand, which is proposed to be mined using in situ leach techniques. However, the Honeymoon deposit is south of the Great Artesian Basin margin (Figure 11.12) Australia, and is not underlain by the Cadna-owie Formation aquifer, which subcrops about 70 km north of the Honeymoon site (Southern Cross Resources Australia 2000). Other environmental geology aspects in this region include the increased tourist traffic and visits to the (mound) springs near Lake Eyre and at Dalhousie Springs since the mid to late 1980s. The increase in visitors threatens the fragile ecology and the carbonate mounds, platforms and terrace structures of some of the springs. The

141

installation of wooden boardwalks across the carbonate mound surface at some springs is a first step to protect these fragile geological features, and also protect indigenous cultural and heritage values.

ASSESSMENT AND MANAGEMENT A long history exists of interstate cooperation in the management and systematic investigation of the Great Artesian Basin since the early 1900s, though the states of Queensland, New South Wales and South Australia and the Northern Territory have separate and different legislation and management strategies. The states in Australia are responsible for water matters under the Australian Constitution, and this include water-resources management. Water, including groundwater is vested in the Crown. All bores in the Great Artesian Basin need to be licensed. The Great Artesian Basin Consultative Council, established in 1997 developed a strategic management plan for the whole of the Great Artesian Basin in 2000. The council, with representatives from federal, state and local governments, pastoral, petroleum and mining industries, traditional landholders, community and conservation groups, plans to address basin-wide management issues, sustainable development and use of the artesian groundwater. Management strategies for the allocation of water in the Great Artesian Basin were earlier discussed by Habermehl (1996a). Partial recovery and further decrease of the diminution of the artesian pressures and the reduction of the uncontrolled discharges are achieved since the introduction of the federal and state government-subsidised Great Artesian Basin Bore Rehabilitation Program in 1989. The program aims to rehabilitate waterbores in poor condition and place control valves on free-flowing artesian waterbores without control mechanisms (Hillier et al. 1995; Reyenga et al 1998). Flowing artesian waterbores are important for the pastoral industry in the basin area. The Great Artesian Basin Sustainability Initiative Program has followed the Great Artesian Basin Bore Rehabilitation Program since 1999, to accelerate bore rehabilitation and bore-drain replacement. About half of the 1200 uncontrolled or corroded waterbores have been rehabilitated (1999), at an average cost per bore of about $50 000 plus headworks (Reyenga et al 1998). Completion of the programs should result in better control and management of the flowing artesian waterbores and their discharges, particularly if a basinwide management program is implemented. Artesian pressures in most areas should thus increase significantly because of the increase in fully controlled flowing artesian waterbores and reduced outflows (reduction in water wastage or artesian groundwater 'saved'), and already important results of increased pressures and flows have been achieved in some areas (Reyenga et al 1998; Cox & Barron 1998). The past and present distribution of artesian groundwater by the main users, the pastoral industry, from the flowing artesian bores by open earth drains is extremely wasteful, owing to seepage, transpiration and evaporation of the water. Bore drains have lengths of many tens of kilometres. This causes wastage of more than 95% of the

142

M. A. Habermehl

groundwater produced. Introduction of (polythene) piping to replace the earth-drain reticulation system will significantly reduce the demand on flowing artesian waterbores for groundwater and could almost eliminate the wastage of water if piping is combined with float-valve-controlled tanks and trough systems. Piping of the water will also reduce the environmental effects caused by the introduction of large amounts of water and watering points in the semiarid and arid landscape. The availability of water in these areas has resulted in land degradation, the spread of introduced weeds, shrubs and trees, greatly increased numbers of feral and native animals attracted by the water, and affected the biodiversity around waterbores and bore drains, and near springs, with reduced outflows (Noble et al 1998). Reduced demand also provides resources for alternative industries, and in recent years mining and oil and gas production have become significant users and producers of artesian groundwater (Figure 11.5). Groundwater extraction by these industries has caused results similar to effects of the groundwater development for the pastoral industry, i.e. large drawdowns of the potentiometric surfaces, which affect other users and naturally occurring flowing artesian springs. Other environmental geology issues include the changes to the recharge areas caused by land-use changes, and the introduction of in situ leach mining techniques near Lake Frome in South Australia in the basin area, although the in situ leaching is carried out in a confined Tertiary aquifer on the surface of the Great Artesian Basin, which is hydraulically separated from the basin's aquifer.

ACKNOWLEDGMENTS This paper is published with the permission of the Executive Director, Bureau of Rural Sciences.

REFERENCES AIREY P. L . , CALF G . E . , CAMPBELL B . L . , HABERMEHL M . A . , HARTLEY P. E .

& ROMAN D. 1979. Aspects of the isotope hydrology of the Great Artesian Basin, Australia. In: Isotope Hydrology 1978. Proceedings International Symposium on Isotope Hydrology, Vol. 1, pp. 205-219. International Atomic Energy Agency, Vienna.

AIREY P. L . , BENTLEY H . , CALF G . E . , DAVIS S . N . , ELMORE D . , GOVE H . , HABERMEHL M . A . , PHILLIPS E ,

SMITH J . & TORGERSEN T .

1983.

Isotope hydrology of the Great Artesian Basin, Australia. In: Papers of the International Conference on Groundwater and Man, Vol. 1, pp. 1-11. Australian Water Resources Council Conference Series 8. ANDREWS J. N. & FONTES J. C. 1993. Comment on 'Chlorine 3 6 dating

of very old groundwater, 3 Further results on the Great Artesian Basin, Australia' by T. Torgersen et al. Water Resources Research 29,

1871-1874.

chemistry of the Great Artesian Basin, Australia. In: Isotope Hydrology 1983. Proceedings International Symposium on Isotope Hydrology in Water Resources Development, pp. 397-413. International Atomic Energy Agency, Vienna.

COLLERSON K. D . , ULLMAN W . J. & TORGERSEN T. 1 9 8 8 . G r o u n d waters

with unradiogenic 8 7 Sr/ 8 6 Sr ratios in the Great Artesian Basin, Australia. Geology 16, 5 9 - 6 3 . Cox R. & BARRON A. (Editors) 1998. Great Artesian Basin Resource Study. Great Artesian Basin Consultative Council, Brisbane. CRESSWELL R . G . , FIFIELD L . K . , KEYWOOD M . D . , HABERMEHL M .

strategy for the Olympic Dam Project water supply scheme. In: International Conference Groundwater Systems under Stress, pp. 363-372. Australian Water Resources Council Conference Series

13. BENTLEY H . W . , PHILLIPS F. M . , DAVIS S . N . , HABERMEHL M . A . , AIREY P. 1986.

Chlorine 36 dating of very old groundwater. The Great Artesian Basin, Australia. Water Resources Research 2 2 , 1 9 9 1 - 2 0 0 1 .

BETHKE C. M . , ZHAO X & TORGERSEN T. 1 9 9 9 . G r o u n d w a t e r flow a n d

the 4 He distribution in the Great Artesian Basin of Australia. Journal of Geophysical Research 104, 1 2 9 9 9 - 1 3 0 1 1 .

KELLETT J. R., JACOBSON G . & WISCHUSEN J. 1 9 9 6 . Theory, rainfall

and recharge: a chlorine-36 story. In: American Geophysical Union, 1996 Western Pacific Geophysics Meeting, p. W44. EOS Supplement 77(22). CULL J. P. & CONLEY D. 1983. Geothermal gradients and heat flow in Australian sedimentary basins. BMR Journal of Australian Geology & Geophysics 8, 329-337. EVANS P. A. 1996. Fluoride anomalies in aquifers of the Queensland section of the Great Artesian Basin and their significance. Geological Society of Australia Abstracts 43, 172-178. FABRYKA-MARTIN J . , BENTLEY H . , ELMORE D . & AIREY P. L. 1 9 8 5 . N a t u r a l

iodine-129 as an environmental Cosmochimica Acta 49, 337-347.

tracer.

Geochimica

2000. Great Artesian Great Artesian Basin

GREAT ARTESIAN BASIN CONSULTATIVE COUNCIL

Basin Strategic Management Plan. Consultative Council, Brisbane. HABERMEHL M. A. 1980. The Great Artesian Basin, Australia. BMR Journal of Australian Geology & Geophysics 5, 9-38. HABERMEHL M. A. 1982. Springs in the Great Artesian Basin, Australia—their origin and nature. Bureau of Mineral Resources Report 235. HABERMEHL M. A. 1983. Hydrogeology and hydrochemistry of the Great Artesian Basin, Australia. In: Papers of the International Conference on Groundwater and Man, Vol. 3, pp. 83-98. Australian Water Resources Council Conference Series 8. HABERMEHL M. A. 1986. Regional groundwater movement, hydrochemistry and hydrocarbon migration in the Eromanga Basin. In: Gravestock D. I., Moore P S. & Pitt G. M. eds. Contributions to the geology and hydrocarbon potential of the Eromanga Basin, pp. 353-376. Geological Society of Australia Special Publication 1 2 . HABERMEHL M. A. 1989. Hydrogeology of the Great Artesian Basin. In: BMR89 - Yearbook of the Bureau of Mineral Resources, Geology and Geophysics, pp. 115-117. Australian Government Publishing Service, Canberra. HABERMEHL M. A. 1996a. Water allocation in the Great Artesian Basin. In: Managing Australia's Inland Waters—Roles for Science and Technology. Prime Minister's Science and Engineering Council 14th Meeting, pp.46-53. Commonwealth Department of Industry, Science and Tourism, Canberra. HABERMEHL M. A. 1996b. Groundwater movement and hydrochemistry of the Great Artesian Basin, Australia. Geological Society of Australia Abstracts 43, 228-236. HABERMEHL M. A. 1999. Assessment of the Beverley Uranium Mine Proposal—Beverley Uranium Mine Environmental Impact Statement by Heathgate Resources Pty Ltd (for Environment Australia, the Environment Program of the Commonwealth Department of the Environment and Heritage). Bureau of Rural Sciences/Australian Geological Survey Organisation, Canberra. HABERMEHL M. A. 2001. Wire-line logged waterbores in the Great Artesian Basin, Australia—Digital data of logs and waterbore data acquired by AGSO. Bureau of Rural Sciences, Canberra, (3 maps at scale 1: 2 500 000 (+CD-ROM). HABERMEHL M . A . , ALLAN G . L . , FIFIELD L . K . & DAVIE R . F.

ARMSTRONG D. & ROWAN I. 1 9 8 6 . T h e development of a m a n a g e m e n t

L . , CALF G . E . , ELMORE D . , GOVE H . E . & TORGERSEN T .

CALF G. E. & HABERMEHL M. A. 1984. Isotope hydrology and hydro-

1993.

C136/C1 Ratios in the Great Artesian Basin, Australia. In: 6th International Conference on Accelerator Mass Spectrometry, Abstract Volume, p. 64. HABERMEHL M. A. & LAU J. E. 1993. Fluoride in groundwater of Cape York Peninsula. AGSO Journal of Australian Geology & Geophysics 14,

316-317.

HABERMEHL M . A. & LAU J . E . 1 9 9 7 . Hydrogeology

of the Great Artesian Basin (Map at scale 1: 2 500 000). Australian Geological Survey Organisation, Canberra.

HABERMEHL M . A . , LAU J . E . , MACKENZIE D . E . & WELLMAN P.

1996.

Sources of fluoride in groundwater in North Queensland, Australia. Geological Society of Australia Abstracts 41, 176.

Great Artesian Basin hydrology HABERMEHL M . A. & SEIDEL G . E . 1 9 7 9 .

Groundwater resources of the Great Artesian Basin. In: Hallsworth E. G. & Woodcock J. T. eds. Proceedings of the Second Invitation Symposium Land and Water Resources of Australia—Dynamics of Utilisation, pp. 7 1 - 9 3 . Australian Academy of Technological Sciences, Melbourne. HEATHGATE RESOURCES 1 9 9 8 . Beverley Uranium Mine Environmental Impact Statement. Heathgate Resources Pty Ltd, Adelaide. HERCZEG A . L . , TORGERSEN T . , CHIVAS A . R . & HABERMEHL M . A .

1987.

HERCZEG A . L . , TORGERSEN T . , CHIVAS A . R . & HABERMEHL M . A .

1991.

Geochemical evolution of groundwaters from the Great Artesian Basin, Australia. EOS 68, 1275-1276.

143

323-351. Geological Society of Australia Special Publication 12.

POLAK E. J . & HORSFALL C. L. 1979. Geothermal gradients in the Great

Artesian Basin, Australia. Australian Society of Exploration Geophysicists Bulletin 10, 144-148. QUARANTOTTO P. 1986. Hydrogeology of the southeastern Eromanga Basin, Queensland. Geological Survey of Queensland Record 1986/38. QUARANTOTTO P. 1989. Hydrogeology of the Surat Basin, Queensland. Geological Survey of Queensland Record 1989/26. RADKE

M.,

FERGUSON

J.,

CRESSWELL

G.,

RANSLEY T.

HABERMEHL M . A. 1995. The Great Artesian Basin-Technological Advances. In: Proceedings of the 16th Federal Convention of the Australian Water and Wastewater Association Inc.—Delivering the Vision for the Next Century, Vol. 2, pp. 245-251. KINHILL-STEARNS ROGER JOINT VENTURE 1 9 8 2 . Olympic Dam Project— Draft Environmental Impact Statement. Prepared for Roxby Management Services Pty Ltd, Adelaide. KINHILL STEARNS 1 9 8 3 . Olympic Dam Project—Supplement to the Draft Environmental Impact Statement. Prepared for Roxby Management Services Pty Ltd, Adelaide. KINHILL STEARNS 1984. Olympic Dam Project—Supplementary Environmental Studies, Mound Springs. Prepared for Roxby Management Services Pty Ltd, Adelaide. KINHILL ENGINEERS 1997. Olympic Dam Expansion Project Environmental Impact Statement. Prepared for WMC (Olympic Dam Corporation) Pty Ltd, Adelaide.

HABERMEHL M. A. 2000. The hydrochemistry and implied hydrodynamics of the Cadna-owie—Hooray Aquifer, Great Artesian Basin, Australia. Bureau of Rural Sciences, Canberra REYENGA P. J . , HABERMEHL M. A. & HOWDEN S. M. 1998. The Great Artesian Basin—bore rehabilitation, rangelands and groundwater management. Bureau of Resource Sciences, Canberra. SAMPSON L. 1996. The Great Artesian Basin well rehabilitation program 1977-1995. MESA Journal 3, 26-28. SEIDEL G. E. 1980. Application of the GABHYD groundwater model of the Great Artesian Basin, Australia. BMR Journal of Australian Geology & Geophysics 5, 39-45. SENIOR B. R. & HABERMEHL M. A. 1980. Structure, hydrodynamics and hydrocarbon potential of the Central Eromanga Basin, Queensland, Australia. BMR Journal of Australian Geology & Geophysics 5, 47-55. SIBENALER Z. 1996. The Great Artesian Basin—a 25 year water use scenario. MESA Journal 2, 18-19. SOUTHERN CROSS RESOURCES AUSTRALIA 2000. Honeymoon Uranium Project Environmental Impact Statement. Southern Cross Resources Australia Pty Ltd, Brisbane.

LANDSBERG J . , JAMES C . D . , MORTON S . R . , HOBBS T. J . , STOL J . , D R E W A . ,

TORGERSEN T . , HABERMEHL M . A . , PHILLIPS F. M . , ELMORE D . , KUBIK P.,

Geochemistry of ground waters from the Great Artesian Basin, Australia. Journal of Hydrology 126, 225-245.

HILLIER J . R . , HAZEL C . P., WILLIAMS M .

R . , HARRIS B . , JOLLY P.

& TONWAY H . 1 9 9 7 . The effects of artificial sources of water on rangeland biodiversity. Environment Australia & CSIRO Wildlife and Ecology, Canberra.

LOVE A . J . , HERCZEG A . L . , SAMPSON L . , CRESSWELL R . G . & FIFIELD L. K .

2 0 0 0 . Sources of chloride and implications for 3 6 C 1 dating of old groundwater, southwestern Great Artesian Basin, Australia. Water Resources Research 36, pp 1561-1574. MULLER P. J . 1 9 8 9 . Aspects of the hydrogeology of the southern Eromanga Basin, Queensland. In: O'Neil B. J. ed. The Cooper and Eromanga Basins, Australia. Proceedings of the Cooper and Eromanga Basins Conference of Petroleum Exploration Society of Australia, Society of Petroleum Engineers, Australian Society of Exploration Geophysicists (SA Branches), pp. 4 9 3 - 5 0 5 . NIEJALKE D . 1 9 9 8 . Mound Spring Researchers Forum and Spring Management Workshop. Mound Springs Researchers Group, Adelaide NOBLE J . C . , HABERMEHL M . A . , JAMES C . D . , LANDSBERG J . , LANGSTON A .

C. & MORTON S. R. 1998. Biodiversity implications of water management in the Great Artesian Basin. Rangelands Journal 20,

275-300.

PITT G. M. 1986. Geothermal gradients, geothermal histories and the

timing of thermal maturation in the Eromanga-Cooper Basins. In: Gravestock D. I., Moore P. S. & Pitt G. M. eds. Contributions to the Geology and Hydrocarbon Potential of the Eromanga Basin, pp.

JONES B. G., H E M M I C K T . & GOVE H . E .

1991. Chlorine-36 dating of very old groundwater 3. Further studies in the Great Artesian Basin, Australia. Water Resources Research 27, 3201-3213. TORGERSEN T . , HABERMEHL M. A. & CLARKE W. B. 1992. Crustal helium fluxes and heat flow in the Great Artesian Basin, Australia. Chemical Geology (Isotope Geoscience Section) 102, 139-152. TORGERSEN T. & PHILLIPS F. M. 1993. Reply. Water Resources Research 29, 1875-1877. WATERHOUSE J . D. & ARMSTRONG D. 1990. Operation and management of the Olympic Dam Project water supply scheme. In: International Conference on Groundwater in Large Sedimentary Basins, pp. 246-255. Australian Water Resources Council Conference Series 20. W E L S H , W . , 2000. GABFLOW: A steady state groundwater flow model of the Great Artesian Basin. Bureau of Rural Sciences, Canberra. W O O D S P. H., WALKER G. R. & ALLISON G. B. 1990. Estimating groundwater discharge at the southern margin of the Great Artesian Basin near Lake Eyre, South Australia. In: Proceedings of the International Conference on Groundwater in Large Sedimentary Basins, pp. 298-309. Australian Water Resources Council Conference Series 20. Received 28 February 2000; accepted 6 June 2000

Geological Society of Australia Special Publication 21, 145-151

CHAPTER 12—Water and mining in the Murray-Darling Basin: meeting the demands for a sustainable environment J. F. LOVERING1*, P. CRABB2 AND K. GOSS3 1 Environment

Conservation Council (Victoria), 66A Molesworth Street Kew, Vic. 3101, Australia. and Environmental Studies., Australian National University, ACT 0200I Australia. 3 Murray-Darling Basin Commission, GPO Box 409, Canberra, ACT 2601, Australia.

2 Centre for Resource

The Murray-Darling Basin covers some 14% of Australia's total area. It is a highly regulated river system with close to 12 000 GL being diverted each year, reducing the median annual flow out of the Murray mouth by 80%. Irrigation accounts for 95% of all water diversions, the remainder being used by industry and domestic consumers (including Adelaide). Because it is clear that continued growth in water consumption is unsustainable, a cap has been placed on the diversion of surface water for consumptive uses. While in some areas groundwater can provide additional resources, both surface water and groundwater must be managed on an integrated basis as one hydrological system. Water-quality problems include salinity, and related land salinisation, and nutrient pollution, which gives rise to toxic algal blooms. The newly emerging heavy-mineral sands mining industry has to address the issues of rehabilitation, infrastructure, community relations, and most importantly, water. Although the total quantity of water available is finite, this does not preclude new development. Water trading within strict environmental constraints is needed to achieve improvements in the efficiency of water use. KEY WORDS: community relations, groundwater, heavy minerals, mining, Murray-Darling Basin, natural-resources management, rehabilitation, water management.

INTRODUCTION The economy of the Murray-Darling Basin is dominated by agriculture, a situation that is unlikely to change. The annual farmgate value of its production is over $9 billion, which is equivalent to about 41% of the Australian total. The value of irrigated agriculture in the Basin is over $4 billion, approximately 75% of the Australian total. The value of production from the basin's manufacturing industry is more than $10 billion: food processing accounts for over 60% of this figure, while overall, between 70 and 75% of manufacturing is associated in some way with agriculture. Tourism and recreation, the other major resource-based industry, has an estimated annual value of $3.4 billion. These few figures provide a context for the basin's existing mining industry. In 1992-93, the only year for which data are available, the value of sales of goods and services by mining operations in the Murray-Darling Basin was $1.66 billion (Crabb 1997 p. 197). This came from over 170 establishments, large and small, producing a wide range of minerals. Mining may not be as large as the other major resource-based industries, but it is an industry that is set to expand. Though there have been a number of closures since 1992-93, such as at Cobar, Gunnedah and Woodlawn, new mines have been opened, as at NorthParkes and Cadia, New South Wales, and others are being developed. Much promising exploration work is being carried out, notably in the Lachlan Fold Belt, while the search continues for a major new find in the Cobar-Broken Hill region. To this must now be added the work on heavy-mineral sands in the western parts of the Murray geological basin, which covers

most of the southwest of the Murray-Darling Basin. Thus far, exploration for heavy-mineral sands indicates a potential value approaching at least $5 billion (Baker 1998; Bromby 1998; McManus 1998; Mason et al 1998). This would be a significant boost to an Australian industry which produces about one third of the world's output of high-grade rutile and zircon with export earnings of over $1.2 billion annually (Anon. 1998; Murphy 1999). Agriculture, manufacturing and tourism as well as the mining industry are all facing much the same challenges, because of their demands on the natural environment and growing community expectations regarding the ways in which they undertake their activities. This is especially true as regards their environmental impact; their infrastructure needs; for some, rehabilitation after the cessation of activities; and their demands for water on which all are dependent (together with a significant proportion of the South Australian population and economy). This is why the theme of this chapter is the future economic and environmental prosperity of the Murray-Darling Basin when its key resource—water—is scarce.

MURRAY-DARLING BASIN Located in inland southeastern Australia, the Murray-Darling Basin covers 1.06 x 10 6 km2, some 14% of Australia's total area. It is defined by its surface water resources. However, these are limited in quantity and * Past President, Murray-Darling Basin Commission.

146

J. F. Lovering et al.

unevenly distributed both seasonally and spatially. Of the basin's total mean annual runoff of 24 300 GL (which is a little over 6% of the national total), 45% comes from just three rivers, the upper Murray, Murrumbidgee and Goulburn. Overall, some 86% of the basin contributes virtually no runoff to the river systems, except during floods. Though one of the world's largest river systems in terms of the extent of its catchment, its surface runoff is among the smallest. In order to make maximum use of the limited surfacewater supplies, reservoirs and other control structures have been built on almost all of the basin's rivers—there are about 65 large dams, locks and weirs. As a result, the Murray-Darling, and especially the Murray, is a highly regulated river system (Figure 12.1 on Plate 7). Its flows are monitored and decisions on allocation are made on a daily basis. Each year, close to 12 000 GL are diverted from the rivers for use by agriculture, industry and domestic consumers within the basin and beyond it (for Adelaide and large areas of South Australia). The main consumer is irrigation which uses 95% of all diverted water. The diversions have had a significant impact on the flows of almost all of the basin's rivers, as well as reducing the median annual flow out of the Murray mouth by 80%. Although overall, the surface resources constitute the main source of water in the basin, they are only part of the total hydrological system and available water resources. Groundwater has long been the key resource in the arid and semiarid northern and western parts of the Murray-Darling Basin, especially those areas underlain by the Great Australian Basin and the Murray Basin. However, whilst it is increasingly evident that many of the Murray-Darling Basin's resource and environmental degradation problems are linked to its groundwater, in many areas groundwater is an underutilised resource.

Managing the basin THE MURRAY-DARLING BASIN INITIATIVE The Murray-Darling Basin Initiative brings together the Commonwealth, New South Wales, Victorian, South Australian, Queensland and Australian Capital Territory governments. Its purpose, as stated in Clause 1 of the Murray-Darling Basin Agreement, is 'to promote and coordinate effective planning and management for the equitable, efficient and sustainable use of the water, land and other environmental resources of the Murray-Darling Basin'. This task is undertaken by the Murray-Darling Basin Ministerial Council and the Murray-Darling Basin Commission. Although most resource management issues in the Murray-Darling Basin are matters of State jurisdiction, the agreement gives the Commission a key role in issues that have the potential to extend beyond an individual jurisdiction. Clause 46 requires that the Commission must be informed of any proposal which may significantly affect the flow, use, control or quality of water in the upper River Murray and in the River Murray in South Australia. The Commission must then assess the proposal's possible effects and make representations to the government or public authority concerned before any decision is made for the proposal to proceed. In other words, the Commission is to be informed and is to be able to inform. By way of illustra-

tion, the Commission and ministerial council played an important role in the commitment of the Australian Newsprint Mill near Albury to zero effluent discharge into the River Murray. It could have been argued that any decision with respect to the mill was solely a matter for the New South Wales government and its agencies. However, it was also an issue with implications for water quality in the upper River Murray (affecting Victoria as well as New South Wales) and in the River Murray in South Australia. THE NATURAL RESOURCES MANAGEMENT STRATEGY The philosophical underpinnings for the achievement of the initiative's objectives and for managing the natural resources of the Murray-Darling Basin are provided by the Natural Resources Management Strategy. It is the foundation document for the work of the Ministerial Council and the Commission (Murray-Darling Basin Ministerial Council 1990). The strategy is a broad charter for a government-community partnership to develop plans for the integrated management of the basin's natural resources on a catchment basis. The Murray-Darling Basin Commission underpins its policy and resource management functions with a major investment in research and knowledge generation. Through its work with federal, state and other agencies, the Commission is continually improving the knowledge basis and hence the understanding of the Murray-Darling Basin and its natural and environmental resources. WORKING IN A DYNAMIC ENVIRONMENT The dynamic natural and policy environments in which the Murray-Darling Basin Commission operates are particularly evident in terms of the basin's water resources. In little more than a decade, the policy environment for water resources has changed from virtually unconstrained development to 'managing for scarcity'. Knowledge generation has been of critical importance in this change. In 1995, an audit of water use in the Murray-Darling Basin clearly demonstrated that the continuing growth in the consumptive use of its surface water was unsustainable in economic, social and environmental terms (Murray-Darling Basin Ministerial Council 1995). As a result, the Ministerial Council placed a 'cap' on the diversion of surface water for consumptive uses. The cap is a recognition of the fact that there needs to be a balance between the consumptive uses of water and the instream needs of the basin's aquatic environments (Murray-Darling Basin Commission 1998). It means that there will be no further growth in diversions from surface-water sources, these being essentially limited to the 1993-94 levels of development. By preventing additional diversions from the basin's rivers, the cap will make an important contribution to constraining further deterioration of the riverine environments, especially in the lower reaches of the system, and in the longer term protect the security of water supply to all users. Under the cap, a range of specific policy objectives is being pursued, including assisting industries, especially irrigation, to improve water-use efficiency, setting objectives for environmental flows in individual rivers, and establishing a water-trading environment. For the Commission and

Water mining, Sustainability, Murray-Darling Basin responsible state agencies, the focus is now on managing for scarcity while contributing to sustainable development. To date, the focus of attention in managing the basin's water resources has primarily been on its surface waters. However, managing surface water for scarcity is inevitably leading to policy changes for the management and allocation of groundwater resources. Provision of the essential basic knowledge for such changes is a good illustration of the Commission's role in collaborative knowledge generation and its 'joint-action test', that is, involvement in issues that cross state boundaries. Working with the Murray-Darling Basin Commission, the Australian Geological Survey Organisation has completed the assessment and mapping of groundwater resources in the Murray Basin (Evans & Kellett 1989; Brown & Stephenson 1991) and the Darling River Basin (AGSO 1995). These hydrogeological studies, which have been undertaken without regard to state boundaries, have demonstrated two critically important facts regarding the Murray-Darling Basin's water resources. First, there are significant underutilised resources of groundwater. Second, and far more importantly, the surface waters and groundwaters must be managed on an integrated basis across the basin as one hydrological system. MANAGING WATER QUALITY The use of the Murray-Darling Basin's limited water resources is further constrained by a range of water-quality problems. These are being increasingly closely monitored and predictions made as to future hazards and impacts. Whereas some such issues are of local significance, two particular basin-wide problems are being addressed by Murray-Darling Basin Commission policies and actions. The massive toxic blue-green algal blooms that appeared along the Darling River in 1991-92 received dramatic media coverage. However, algal blooms are now an annual occurrence in most of the basin's waterways and many of its reservoirs. Their primary cause is elevated levels of phosphates and nitrates in the water. Human activities, including agriculture, certainly contribute to the quantities of nutrients in the waterways, but the main sources are the soils and rocks of the natural environment (Donnelly 1995). The Murray-Darling Basin Commission is addressing the problem through its Algal-Management Strategy, a critical component of which is the improvement of stream-flow regimes and flow management, in order to prevent or limit future blooms. The basin's most important water-quality issue is salinity, complemented by the major land-degradation problem—land salinisation (Lovering et al 1998). The recent salinity audit (Murray-Darling Basin Ministerial Council 1999) has indicated that the problems are far more serious and extensive than previous studies had indicated (Williamson et al 1997). It is predicted that 3 - 5 x 10 6 ha of the basin will be salt-affected in 5 0 - 1 0 0 years time and that the water quality of some rivers and streams will reach levels that will damage crops and the environment (Figure 12.2 on Plate 7). For example, as measured at Morgan, River Murray salinity will rise to an average 790 EC* within 50 years. This will have significant implications for domestic supplies for Adelaide. There will also be major impacts on infrastructure, including roads and buildings, and on

147

biological diversity. Many of these problems already exist, such as the loss of agricultural land, the impacts on buildings and infrastructure in such places as Wagga Wagga, and highly saline domestic water supplies, as in Yass and Boorowa. The costs are high now; they will be enormous in the future; and doing nothing will be even more costly. In particular, farming practices will have to change dramatically (Walker et al 1999). The Salinity and Drainage Strategy has made a valued contribution to offsetting rising salinity in the River Murray through the installation of saline groundwater interception schemes and drainage schemes in a number of irrigation areas along the Murray and Murrumbidgee Rivers (Murray-Darling Basin Commission 1999). These works were a response to what was seen as the main problems at the time the strategy was formulated, namely River Murray water salinity and land salinisation in irrigation areas. Now, the greatest hazard comes from the dryland areas. A new comprehensive salinity management strategy is now nearing completion.

KEY ISSUES FOR THE MINING INDUSTRY This chapter has thus far provided a setting for the sustainable management of the Murray-Darling Basin and its major economic activities and society. From what has been covered, four key issues can be identified for the basin's mining industry as a whole and for a future heavy-mineral sands industry in particular, namely rehabilitation, infrastructure, community relations and water. These are issues that the mining industry as a whole needs to get under control in order to be a good citizen of the Murray-Darling Basin. They are also key issues for the Murray-Darling Basin Commission. The Minerals Council of Australia has made a 'commitment to excellence in managing the environmental aspects of its operations...from initial exploration to closure and final rehabilitation' (Minerals Council of Australia 1996). Its Code for Environmental Management sets out nine principles that are all relevant to a heavy-mineral sands industry in the Murray-Darling Basin: sustainable development, environmentally responsible culture, community partnership, risk management, integrated environmental management, performance targets, continual improvement, rehabilitation and decommissioning and reporting (Minerals Council of Australia 1996). There is also an acknowledgment of the need to do better as needs and expectations change. As the managing director of BHP stated on the release of his company's 1998 environment report (produced to meet the requirements of the Minerals Council of Australia Code), 'the company's environmental performance was a key issue for its standing in the global community' (The Australian 12 December 1998). To this must be added the involvement of the local community of each particular operation. Of importance to the Murray-Darling Basin Commission is the fact that there are close parallels between the major aims of its Natural Resources Management Strategy and the principles con*Electrical Conductivity unit. 1 EC = 1 micro-Siemens per centimetre, measured at 25°C. It is used as a measure of water salinity.

148

J. F. Lovering etal.

tained in the Mineral Council's Code for Environmental Management. Before discussing each of the four key issues in more detail, two general points need to be made. First, all of the issues are matters of state jurisdiction. State and local governments and their agencies will impose their own conditions, especially in terms of environmental requirements. However, as discussed above, not only does accessing water resources have interstate implications that are of concern to the Murray-Darling Basin Commission, the wider and interstate ramifications of other issues are becoming increasingly evident. The fact that the Murray Basin and its heavy-mineral sands deposits take no account of state and other administrative boundaries provides ample illustration of this. The second general point is that all four key issues are interconnected; it is really not possible to consider one without the others.

Rehabilitation A mine has a finite life. At some point in time, mining will cease, either because the orebody has been worked out or because it is no longer economic to continue mining due to a lack of demand for the product or because the price is not sufficient to maintain a viable operation. Whereas in the past mining companies could simply close the gate and move on, this is no longer acceptable, as has been demonstrated by the response to such events as the closure of the CSA Cobar and Woodlawn mines. Removing the last truckload of ore is no longer the final task at a mine. Leaving the rehabilitation to a later date is also extremely costly as indicated by the $30 million cost placed on the cleanup of an old mine at Brukunga in the Adelaide Hills (on a tributary of the Bremmer River) that produced little more than $50 million worth of pyrites (MERSA 1997). In a discussion paper on mine closure, the Minerals Council has acknowledged that 'minesite decommissioning and rehabilitation standards are a significant issue for governments and the minerals industry' (Minerals Council of Australia 1997 p. 1). They are also an issue for the community. The paper also stated that '[i]t is evident the minerals industry lacks a coherent and strategic framework for dealing with mine closure issues and therefore remains vulnerable to reactive responses and negative opinion' (Minerals Council of Australia 1997 p. iii). The unfortunate Cobar, Woodlawn and other closures have confirmed this. The point will be particularly important for the heavy-mineral sands industry which will be working numerous deposits over the Murray Basin where the existing communities already have many issues in common. On the basis of the Commission's experience, the Murray-Darling Basin community will expect a clear commitment to the rehabilitation of mine sites. Rehabilitation must start at the beginning of a project, as is set out, for example, in the Environmental Effects Statement for the Wemen Project in northwest Victoria (NSR 1998). Can the requirements set out for 'Rehabilitation and Decommissioning' in the Minerals Council of Australia Code for Environmental Management be seen as the minimum standard? Can these be more than a voluntary code? Can the industry go further? Rather than just returning the landscape to as

close as possible to its former state, can it be improved upon, and perhaps be made more productive? Can the rehabilitation work contribute to the combating of dryland salinity and rising groundwater levels? There is much scope, as well as need, for innovative measures, not least because the heavy-mineral sands industry will be operating in a saline environment. Given this and the highly saline groundwater levels in many areas, the industry will have to learn to live with salt in both its mining and rehabilitation phases. For example, saline aquatic ecosystems can be established to provide habitat for native flora and fauna (after all, they are part of the existing environment) and there is some scope for aquiculture, which would provide landowners with another source of income. Successful progressive rehabilitation work is being undertaken at mineral-sands mining sites in Western Australia and, in some cases, this has been recognised with Landcare awards. Constantly improving industry best practice is expected in the Murray-Darling Basin.

Infrastructure Each mining operation requires its own particular infrastructure, facilities which are essential to its successful operation, such as roads, railways, water and electricity supplies, communications and service industries. Roads are a key issue when mining occurs at a number of locations which supply a central processing facility. The required infrastructure should be established in conjunction with the local communities. This will not only be in the best interests of the company, as many of the roads and associated facilities will be of immediate benefit to the local communities, but such facilities may well remain after the mining activities have ceased. Mining companies have therefore the potential to make an ongoing contribution to community health and well-being. Given that the mining operations will be located in rural areas, many of which suffer deprivation in terms of facilities and services most Australians take for granted, such contributions will be of significant value. Because of the locations of the prospective deposits, the heavy-mineral sands industry has the potential to make a significant positive contribution to many small outback communities. Such mining activities will clearly be supported by governments as being in keeping with their regional development policies. Just as there is a role for the communities to contribute to the industry, so there is the potential for the companies to invest in local communities, for the present and their future.

Community relations 'Community Partnership' is one of the principles of the Minerals Council of Australia's Code for Environmental Management which advocates '[c]onsulting the community on its concerns, aspirations and values regarding development and operational aspects of mineral projects, recognising that there are links between environmental, economic, social and cultural issues' (Minerals Council of Australia 1996). Especially in areas where there is little direct experience of living with a mining industry, it is critically impor-

Water mining, Sustainability, Murray-Darling Basin tant that the heavy-mineral sands industry develops strong relations with the local communities. Community-wide support is essential for the future and long-term stability of the mining activities. There will be economic and social impacts from increased employment and from any increases in the size of existing small communities due to an increased work force. The provision of jobs and infrastructure will be a plus for many communities, but accurate and transparent information programs will be essential to overcome inevitable and understandable community reservations. At a 1998 conference on mining and the environment, a distinction was drawn between 'information' and 'awareness'. In giving emphasis to both, it was pointed out that, in their absence, '[t]he community will fill information vacuums with innuendo and mistrust; and, perceptions become reality' (Jean 1998). Environmental issues will be a particular concern at all stages of the operation. Many of the deposits and prospects are located close to or on the margins of environmentally sensitive national parks and other reserves, especially the Bookmark Reserve and the Willandra Lakes World Heritage Area (see the map in Mason et al 1998). Concerns about mining in such areas are inevitable and understandable. As a first step, a detailed map is required that shows all the national parks and other reserves and all the known mineral deposits and prospects. No natural resource can be considered in isolation from others. There are concerns about disturbance of the natural landscapes and their flora and fauna, especially as the impacts of opencut mines are generally greater and more widespread than those of underground mines. Also, many deposits are in areas of shallow and commonly saline groundwater, where mining has the potential to disrupt groundwater supplies to existing users and add to salinity problems. There will be pollution concerns, of the land as well as surface and groundwaters, from dust, and in some cases, raised radioactivity levels, as in the mildly radioactive fine-grained sands of the WIM 150 deposit near Horsham (Hogan 1996).

Water Of the four key issues, water is perhaps the most significant. To begin with, there are two points that are relevant for the mining industry throughout the Murray-Darling Basin. First, although the industry is not a large user of water overall, this is not necessarily the case in particular locations. As has been stressed, the basin as a whole is a water-scarce environment, either because there is no water (as in the arid parts) or because it is already allocated and overallocated. Thus where mining does take place, demands may well be large in terms of available supplies. Second, there is no more 'freely' available water and so it will have to be bought. As a result of the cap, existing supplies will have to be shared among existing and new users. In order to meet the continuing increase in the demand for water and to achieve the environmental and economic sustainability of the basin, there has to be much greater efficiency in the allocation and use of the limited water resources. Water can only become available as a result of improved efficiency in its

149

use and by trading. With a limited available supply and increased demand, the price of water will inevitably increase. This will again contribute to greater efficiency of use, but it will also mean that water that does become available will go to the highest bidder. Water trading within strict environmental constraints is seen as one of the main mechanisms available to achieve desired improvements in the efficiency with which it is used. To date, virtually all of the water trading has been in the farm sector. Among the few exceptions has been the purchase of a number of water licences by the Forbes Shire Council to ensure the availability of water for the future growth of the town. This is clearly a far-sighted move, but irrigators in the Lachlan Valley have expressed their concerns at a town buying up 'their' water. A town has greater buying power than an irrigator in an open market. The purchase of water licences by a mining company, which will probably have even more buying power than a country town, will introduce a new player to the water market and thus create a new situation. Mining at the individual deposit level will be relatively short term and once it is concluded, the water licences can be placed on the market and so revert to agriculture. Alternatively, they can be made available for the environment and so contribute to environmental flows, an action which would be valued by the community. Where groundwater is used, as may be possible in some areas, a different set of issues will require consideration, not least to ensure that use does not exceed recharge capacity and that it is not to the detriment of other users. Of particular importance are the needs of the natural environment as a whole, the many small communities that depend on groundwater for community supplies, and the domestic and stock needs of the long-established pastoral industry. Great care will be needed in addressing these various needs given the location of the heavy-mineral sands in the complex Pliocene sands aquifer, the uppermost aquifer group of the Murray Basin, where the groundwater can be highly saline (Crabb 1999). As the Wemen Project Environmental Effects Statement illustrates (NSR 1998), the water requirements of the heavy-mineral sands industry are considerable. Whilst local conditions in terms of the nature of an orebody and water availability will determine whether wet or dry mining is undertaken, water is required for the initial processing which is undertaken at the mine site. It is understood that saline water can be used for this stage (though the upper salinity level is not known), but freshwater is required for later stages. The requirements for freshwater could well determine where processing takes place and whether or not it will be in the Murray-Darling Basin. Given existing water constraints, it is imperative to develop separation and beneficiation processes that use less water and overcome or reduce waste problems (Hines 1997). If implemented, there are indications that these would not only reduce water demand, but would also overcome any radioactivity issues and assist mine-site rehabilitation. From the policy issues that have been considered, it is clear that any potential development in the Murray-Darling Basin, particularly one that requires as much water as the heavy-mineral sands industry, has to engage the Murray-Darling Basin Commission and state

150

J. F. Lovering

etal.

jurisdictions and plan for sustainable use. Although the total quantity of water available is now finite, this does not preclude new development. As was indicated above, the cap on diversions and the development of a water market have already contributed to important changes in the irrigation industry, especially in terms of some expansion and improved water-use efficiency in the horticulture and viticulture industries. In addition, other industries are finding new ways for reusing water and improving efficiency of production. Finally, the smaller subeconomic irrigators who are selling their licences can exit the industry with cash-inhand and with dignity, and make water available for new demands.

CONCLUSION This chapter has focused on the extensive use of limited water resources and the likely demands on these resources by a new heavy-mineral sands industry. This industry will have to come to grips with managing water resources for scarcity. The Murray-Darling Basin Commission has policies in place to deliver on this. Although the current policies are focused on surface-water resources, it is inevitable that the principled approach applied to them will also be adopted for groundwater. In the first instance, good base data are available as a result of the Commission's contribution to knowledge generation, and steps have already been taken to control the use of groundwater in some regions. Other issues, such as environmental impact, infrastructure development, rehabilitation, and community relations, are more directly the responsibility of state and local governments, but the Murray-Darling Basin Commission also has a responsibility to develop strategic responses to these concerns. Given the objective of the Murray-Darling Basin Agreement and the charter of the Natural Resources Management Strategy, the Murray-Darling Basin Commission has a prime responsibility for the sustainable management of the basin's total natural and environmental resource base. The Commission will have a very important role to play in the development of a heavy-mineral sands industry, in terms of the Murray-Darling Basin Agreement's Clause 46, because of the location of many of the deposits and the interconnections between surface water and groundwater. The parallels between the Mineral Council's code and the aims of the commission's Natural Resources Management Strategy reinforce the application of principles and the expectations the Commission and the basin community have of an emerging heavy-mineral sands industry. Nothing but best practice—for which the code sets the minimum standards—is expected as the industry works in and with the Murray Basin and the wider Murray-Darling Basin. In terms of the total environment, and especially its water resources, the industry is investing in the present to ensure that it has a future.

chapter is available in Crabb (1997) and on the commission's web site <http://www.mdbc.gov.au>. The text has been adapted and revised from Lovering et al (1999).

REFERENCES ANON. 1998. Titanium minerals. Australian Commodities 5, 482-485. AGSO 1995. Hydrogeology of the Darling River Drainage Basin, Queensland-New South Wales (1:1 000 000 scale map on two sheets). Australian Geological Survey Organisation, Canberra. BAKER R. 1998. Rivers of sand. The Bulletin

19 M a y 1998, 5 4 - 5 5 .

BROMBY R. 1998. Deposits lure Perth junior to Basin float. The Australian, 27 April 1998. BROWN C. M. & STEPHENSON A. E. 1991. Geology of the Murray Basin,

Southeastern Australia. Bureau of Mineral Resources Bulletin 235. CRABB P. 1997. Murray-Darling Basin Resources. Murray-Darling Basin Commission, Canberra. CRABB P. 1999. Murray-Darling Basin Groundwater: a resource for the future. Murray-Darling Basin Commission, Canberra. DONNELLY T. H. 1995. Review and Scoping Study of Catchment Phosphorus Sources. CSIRO Division of Water Resources, Canberra, Consultancy Report 9 4 - 2 6 . EVANS W . R. & KELLETT J. R. 1 9 8 9 . T h e hydrogeology of the Murray

Basin, southeastern Australia. BMR Journal of Australian Geology & Geophysics 11, 144-166. HINES H. 1997. Impact of new technology on the mineral sands industry. In: Outlook '97: Volume 3 Minerals and Energy, pp. 445-449. Australian Bureau of Resource and Agricultural Economics, Canberra. HOGAN J. 1996. Outlook for Australia's titanium minerals industry. In: Outlook '96: Volume 3: Minerals and Energy, pp. 291-305. Bureau of Resource and Agricultural Economics, Canberra. JEAN R. 1998. Forum targets mining and environment. Australian Landcare 24-25 December 1998. LOVERING J .

F.,

CRABB

EVANS W .

1998.

Salinity

This text has been adapted from the Murray Basin Mineral Sands Conference, Mildura, Victoria, April 1999. Further information on almost all of the topics discussed in this

the

LOVERING J. F., CRABB P. & Goss K. 1999. Future growth of the

Murray-Darling Basin: meeting the demands for a sustainable environment. In: Stewart R. ed. Murray Basin Mineral Sands Conference, pp. 219-226. Australian Institute of Geoscientists Bulletin 26. MCMANUS P. 1998. Future potential of the Murray Basin as a titanium minerals producing region. In: Outlook '98: Volume 3 Minerals and Energy, pp. 3 8 7 - 3 9 1 . Australian Bureau of Resource and Agricultural Economics, Canberra. MASON A. J., TEAKLE M . & BLAMPAIN P. A. 1 9 9 8 . Heavy mineral sand

deposits, central Murray Basin. In: Berkman D. A. & Mackenzie D. H. eds. Geology of Australian and Papua New Guinean Mineral Deposits, pp. 647-650. Australasian Institute of Mining and Metallurgy Monograph 2 2 . MERSA 1997. Brukunga Mine Site: an environmental overview. Mines and Energy Resources South Australia, Adelaide. MINERALS COUNCIL OF AUSTRALIA 1 9 9 6 . Australian

Code for Environmental Australia, Canberra.

Management.

Minerals

Industry

Minerals Council of

MINERALS COUNCIL OF AUSTRALIA 1 9 9 7 . Mine Closure:

towards

a strat-

egy for the Australian minerals industry. Minerals Council of Australia, Canberra. MURPHY R. 1999. Mineral sands: industry survives mergers, closures and rationalisation. In: Bester G. ed. Register of Australian Mining 1999/2000, p. 333. Resource Information Unit, Subiaco. MURRAY-DARLING BASIN COMMISSION 1 9 9 8 . Murray-Darling

ACKNOWLEDGEMENTS

Murray-Darling Basin: a critical challenge for the 21st Century. In: Weaver T. R. & Lawrence C. R. eds. Proceedings of the International Association of Hydrogeologists International Groundwater Conference. Groundwater: Sustainable Solutions, pp. 215-230. International Association of Hydrogeologists (Australian National Chapter), Brisbane.

on Diversions: water year 1997/98: striking Murray-Darling Basin Commission, Canberra.

MURRAY-DARLING BASIN COMMISSION

1 9 9 9 . Salinity

the and

Basin

Cap

balance. Drainage

Strategy: ten years on, 1999. Murray-Darling Basin Commission, Canberra.

MURRAY-DARLING BASIN MINISTERIAL COUNCIL 1 9 9 0 .

Basin Natural Resources Management

Strategy.

Murray-Darling

Water mining, Sustainability, Murray-Darling Basin Basin Ministerial Council, Canberra.

MURRAY-DARLING BASIN MINISTERIAL COUNCIL 1 9 9 5 . An Audit of Water

Use in the Murray-Darling Ministerial Council, Canberra.

Basin.

Murray-Darling Basin

BASIN MINISTERIAL COUNCIL 1 9 9 8 . Review of Cap Implementation 1997/98: report of the Independent Audit Group. Murray-Darling Basin Ministerial Council, Canberra. MURRAY-DARLING BASIN MINISTERIAL COUNCIL 1 9 9 9 . The Salinity Audit of the Murray-Darling Basin: a 100-year perspective, 1999. Murray-Darling Basin Ministerial Council, Canberra. NSR 1998. Wemen Project: Environmental Effects Statement: main

MURRAY-DARLING

151

report. Report prepared for RZM Pty Limited by NSR Environmental Consultants Pty. Ltd., Melbourne. Effectiveness of Current Farming Systems in the Control of Dryland Salinity. CSIRO Land and Water, Canberra.

WALKER G . , GILFEDDER M . & WILLIAMS J . 1 9 9 9 .

WILLIAMSON D . R . , GATES G . W . B . , ROBINSON G . , LINKE G . K . , SEKER

M. P. & EVANS W . R . 1 9 9 7 . Salt Trends: historic trend in salt concentration and saltload of stream flow in the Murray-Darling Drainage Division. Murray-Darling Basin Commission, Canberra.

Received 23 March 2000; accepted 29 September 2000

Geological Society of Australia Special Publication 21, 153-158

CHAPTER 13—Salinisation and environmental geoscience: a case study in the Murray Basin 1 Bureau of Rural Sciences,

PO Box El I Kingston, ACT 2604, Australia.

The Murray Basin is a large shallow sedimentary basin in southeast Australia traversed by the lower reaches of the Murray and Darling Rivers and their tributaries. The basin underlies a strategic agricultural region with grazing, broad-acre cropping and irrigated enterprises. This development has come with a cost as regional water-tables have risen, mobilising salt that is inherent to the landscape and causing unproductive scalds and increased saline seepages to the River Murray. Land and water salinisation is a groundwater-related problem, so that its management requires an understanding of groundwater processes. The Lower Darling groundwater flow model is an example of how hydrogeological studies can provide valuable technical advice on the management of a major economic and environmental degradation issue. The model provided insights into the regional hydrogeology and how the groundwater system interacts with the Murray and Darling Rivers in the northwest quarter of the basin. It also predicted the adverse and unacceptable effects of further clearing of native vegetation in terms of elevated salt loads into the River Murray and increased land salinisation. KEY WORDS: groundwater, Murray Basin, Murray-Darling Basin, numerical models, salinisation, salinity.

INTRODUCTION The Murray Basin is a saucer-shaped sedimentary basin covering an area of 300 000 km 2 of western New South Wales, northwest Victoria and southeast South Australia (Figure 13.1). It is contained within the Murray-Darling drainage catchment, which has been termed Australia's 'food basket' generating $8.56 billion (in 1991-92) or, 41% of Australia's total gross value of agricultural production (Murray-Darling Basin Commission 1999a). The commodities produced in the basin are many and varied including wheat from the Mallee, rice from Coleambally, citrus and sultanas from Mildura, dairy produce from Kerang, fruits from Shepparton and vegetables from Griffith.

GEOLOGY OF THE MURRAY BASIN Despite its extent, the basin is only a veneer of Cenozoic sediments, typically less than 200 m thick. The stratigraphy and geological history of the basin is well-documented (Brown 1989; Brown & Stephenson 1991). With a technically stable basement of subdued topography, the Cenozoic sequence records deposition from ancestral rivers and lakes and by periodic flooding by shallow epicontinental seas. Fluvial sedimentation commenced in the Paleocene, about 60 Ma, with the medium to coarse sands of the Warina Sand (Figure 13.2). These deposits partially filled the major depocentres of the basin, the main one in the Renmark-Mildura area and a subsidiary centre near Hay, New South Wales. The carbonaceous silt, sand and clay of the fluviolacustrine Renmark Group, represent the thickest and most widely distributed sequence in the basin. In the

east, where fluvial deposition continued into the Miocene, a three-fold subdivision (Upper, Middle, Lower) of the Renmark Group has been made based on lithology, geophysics and palynology (Kellett 1989). Sea-level rises about 32 Ma resulted in the deposition of marine and marginal marine deposits over the western half of the basin. This commenced with the grey-green glauconitic and highly fossiliferous calcareous clay of the Ettrick Formation, progressing to the shallow-marine platform Murray Group limestones. These limestones are flanked on the landward side by sediments deposited in shallower marine and estuarine environments, including the fossiliferous calcareous clays of the Winnambool Formation and dark muds and silts of the Geera Clay. The clays and marls of the overlying Bookpurnong beds mark the onset of the last major marine incursion during the late Miocene to Pliocene, starting about 6 Ma. The Loxton-Parilla Sands is the extensive sand sheet deposited by the retreating sea, and characterised by northwesterly arcuate coastal ridge remnants and intervening swales. Meanwhile, fluviolacustrine conditions prevailed in the east and north, indicated by the kaolinised fine to coarse quartzose sand, silt and clay of the Calivil Formation, continuing with the mottled clay and polymictic sand of the Shepparton Formation, and finally the alluvial deposits of the modern Riverine Plain. Late Pliocene uplift resulted in the tectonic damming of the ancestral River Murray, producing the freshwater megalake, Lake Bungunnia, at about 2.5 Ma (Stephenson 1986). The Blanchetown Clay represents the silty clay and dolomitic limestone deposited on the megalake bed. The onset of aridity at about 0.5-0.7 Ma reduced Lake Bungunnia to a series of remnant hypersaline lakes and triggered the mobilising of Loxton-Parilla Sands by prevailing

154

R. S. Brodie

Figure 13.1 Location of the Murray Basin and the Lower Darling groundwater flow model.

westerly winds. This resulted in the extensive Quaternary dunefields that are characteristic of the western Mallee region of the basin.

THE SALINISATION ISSUE Salinity is a feature of the natural landscape in the Murray Basin. Through geological time, salt has been imported into the basin via rainfall, sea spray, dust fall, surface water from the Murray-Darling system and groundwater entering from adjacent valley fill. In comparison, export of salt out of the basin is limited to what can be removed by the River Murray as well as by salt-laden dust mobilised by prevailing winds. The low rainfall, flat topography and sluggish groundwater flow limits the ability of salts to be flushed. It has been estimated that the salt store in the Murray Basin aquifers exceeds 100 000 Mt and under the current hydrological regime, less than 0.004% of this is exported annually by the River Murray (Evans & Kellett 1989). This is compounded by the fact that water diversion from the river has led to median annual flows from the basin to the sea to be less than a quarter of pre-development flows (Murray-Darling Basin Commission 1999b). The basin is essentially a closed groundwater system due to its saucer-shaped geometry, so that groundwater is trapped and directed towards the main central-western depocentre near Renmark - Mildura (Brown 1989; Evans & Kellett 1989). Groundwater discharge has ebbed and flowed in local depressions through geological time, indicated by the scattering across the basin of salinas with their characteristic crescentic lunettes. Here salts accumulate by evapotranspiration, to be recycled into the underlying shallow aquifer. The groundwater system tends to be layered,

with highly saline waters in the shallow aquifers and salinity decreasing with depth. This process of salt accumulation has been particularly evident in the last 0.5 Ma when the climate has oscillated between arid and humid (Bowler 1990). As the stratigraphic sequence is thin and largely water-saturated so that it has limited additional groundwater storage capacity, the groundwater system is sensitive to changes in recharge brought about by climate change (Brown 1989). In particular, a major episode of groundwater discharge and salinisation occurred during the Last Glacial Maximum about 25-16 ka (Bowler 1988). This correlates with the drying out and deflation of Lake Eyre and the drying of Lake Keilambete and other maar lakes in western Victoria (Williams 2001). Agricultural development since European settlement has triggered another episode of salt mobilisation in the basin. The clearing of native vegetation to be replaced by shallow-rooted crops and pastures has significantly increased infiltration of water, resulting in water-table rise. Deep-rooted native vegetation such as Eucalypt mallee is a very efficient interceptor of rainfall, so that its removal can result in recharge increasing by several orders of magnitude (Cook et al 1996). The establishment of irrigation areas, particularly with limited drainage infrastructure, exacerbated this process. The consequences of water-table rise have been remobilisation and concentration of salt to the surface, reactivation of groundwater discharge sites and steepening hydraulic gradients near drainage lines causing increased saline groundwater discharge into the River Murray. This has impacts on water quality, in terms of both consumptive use and the environmental integrity of rivers and wetlands. Land-based salinisation seriously damages agriculturally productive soils, infrastructure such as roads and buildings, as well as local terrestrial ecosystems.

Salinisation, Murray Basin

155

Figure 13.2 Tertiary stratigraphy of the Murray Basin (from Brown & Stephenson 1991).

MANAGEMENT STRATEGIES Considering the national prominence of the River Murray and the fact that the Murray Basin straddles three states, the problems of land and water salinisation required a joint government action. This resulted in the signing of the Murray-Darling Basin Agreement in 1987 and the establishment of new institutions at a political (Murray-Darling Basin Ministerial Council), bureaucratic (Murray-Darling Basin Commission) and community (Community Advisory Committee) level. In 1989, the Salinity and Drainage Strategy was adopted to provide coordinated salinity management (Murray-Darling Basin Commission 1999a). The strategy sets out a specific reduction target against benchmark conditions, namely to reduce average river salinity at Morgan by 80 EC*. This would maintain salinity levels below the 800 EC threshold for desirable drinking water quality for 95% of the time. Morgan was chosen as it is immediately upstream of the pipeline offtakes for Adelaide's water supply. Engineering solutions such as salt interception and drainage diversion schemes have reduced salinity at Morgan by 75.9 EC. This has enabled the implementation of land and water management plans in irrigation areas to reduce waterlogging problems, but increasing salt loads to the river by 18.6 EC (Murray-Darling Basin Commission 1999a). These figures are derived from a register of schemes and plans that is maintained with their calculated impact as a salinity debit or credit, based on the estimated economic impact on River Murray water users. This assessment is done by the

Murray-Darling Basin Commission with numerical models that simulate water and salt balances in the River Murray, using the 1975-85 historic record as a baseline. The success of the strategy is tempered by the recognition of large salt loads emerging from dryland catchment sources with the average salinity at Morgan projected to exceed 800 EC in the next 5 0 - 1 0 0 years (Murray-Darling Basin Ministerial Council 1999). In this light, the Murray-Darling Basin Commission is currently reviewing the strategy.

THE LOWER DARLING GROUNDWATER FLOW MODEL Land and water salinisation in the Murray Basin is a groundwater-related problem. Amelioration requires an understanding of groundwater dynamics and how the watertable responds to man-made changes such as clearing, revegetation, irrigation development, groundwater-interception schemes and river regulation. Ideally, these developments should be translated into the likely EC impacts on the average salinity at the Morgan benchmark. To provide a predictive capacity in terms of long-term regional groundwater responses to management strategies, the Murray-Darling Basin Ministerial Council endorsed a basin-wide regional groundwater-modelling program. This involved subdividing * EC is a measure of electrical conductivity of water and used as an indicator of total dissolved solids. 1 EC is approximately equal to 0.6 mg/litre of total dissolved solids.

156

R. S. Brodie

Salinity KXKMSOO 1500-3000 ms& 3000-7000 mg/L yCKf 7000-14k mo/t "

14K-35kmg/L 35k* 100k mg/l

Figure 13.3 Cross-section along a regional groundwater flow line in the Lower Darling model area showing the fluvial dominant Calvil Formation (Tpc), Upper Renmark (Ter3), Middle Renmark (Ter2), Lower Renmark (Terl) and Warina Sands (Tew), the marginal marine Ettrick Formation (Toe), Winnmbool Formation (Tmw), Geera Clay (Tmg), Bookpurnong Beds (Tpb) and Parilla Sands (Tps) and the marine Murray Group limestones (Tml).

the basin into five regional model areas, with the Lower Darling model being one of these models. The Lower Darling groundwater flow model covers the northwestern quarter of the Murray Basin, comprising the Darling River downstream of Wilcannia, the Menindee Lakes and Lake Victoria storages, and the Murray River between Mildura and Morgan (Figure 13.1) (Brodie 1998). The area is typical of the Mallee province of the basin, with a semiarid climate, aeolian landforms and mallee eucalypt vegetation. This is the least-developed area of the basin, with limited clearing of the stands of mallee or belah-rosewood shrubland. Clearing of the native vegetation for pastoral or cultivation purposes is a contentious issue, and the main objective of the model was to assess the salinity impacts of any major clearing. The model area progresses from the northwestern basin margin to the main depocentre (Renmark Trough) that contains over 600 m of Cenozoic sediment fill (Figure 13.3). A basin structure of southwest-trending troughs and ridges directs the regional groundwater flow towards the River Murray. Near the basin margins, the vertical hydraulic gradient is downwards and leakage from streams and lakes lowers groundwater salinity in the shallow aquifers. The freshest groundwater of 600-900 mg/L is found in a 2 - 1 0 km corridor along the Darling River channel, with 1 3 0 0 - 2 5 0 0 mg/L typical further out in the floodplain. Groundwater salinity increases rapidly down gradient, exceeding seawater concentration (35 0 0 0 mg/L) near the marginal marine clays of the Geera Clay and Winnambool Formation. To the southwest, the deeper confined aquifers become increasingly pressurised causing upward groundwater flow. This pressurisation can be explained by a number of mechanisms. First, highly saline and dense groundwater stored in the overlying marginal marine clays induces a significant hydrostatic load. This loading was used to explain the concomitant rise in pressures in the Murray Group limestone aquifer and the Lower Renmark aquifer recorded in piezometers east of Lake Victoria (Barnett 1995). Second, the transgressive deposition of these clays dramatically thins the laterally equivalent Middle Renmark aquifer, diverting groundwater downwards. Third, the Hamley Fault significantly thins the Lower Renmark sequence towards the end of the regional groundwater flow path (Figure 13.3). Fourth, at the basin depocentre, flow in the basal aquifer

encounters groundwater flowing in the opposite direction that originated from the southern basin margin. The net result is that groundwater discharge conditions prevail near the River Murray, resulting in active salinas in local depressions and saline accessions to the river itself. The discharge of saline groundwater into the river is particularly evident in the Woolpunda Reach, requiring the construction and operation of the Woolpunda Salt Interception Scheme.

MODEL DESCRIPTION Numerical models simulate a groundwater system by representing physical processes by means of governing mathematical equations. For the project area, a steady state three-dimensional numerical model simulating groundwater conditions observed in 1988 was constructed using the MODFLOW (McDonald & Harbaugh 1988) model code. In this case, the three-dimensional movement of groundwater of constant density through a porous medium is described by the partial-differential equation: A | k x x — ) + J- ( k y y — ) + A ( k z z — \ - W = 0 dx v dx) 3y v 3y dz V dzJ 3y /

where Kxx, Kyy and Kzz are hydraulic conductivity values along the x, y and z coordinate axes, h is the potentiometric head and W is the volumetric flux per unit volume representing the sources and sinks of water. W h e n combined with boundary and initial head conditions, this is the mathematical representation of the aquifer system encapsulating the accumulated interpretation of its geometry, hydraulic properties, potentiometry, water inputs and outputs and relationship with surface water features and other aquifers. In the Lower Darling model, the five regional hydrogeological units in the Murray Basin sediments were separated into the five model layers, with each layer subdivided into 7.5 x 7.5 km cells. The Lower Cretaceous sediments that underlie the Tertiary sequence were indirectly represented in the model as boundary conditions. The objective of model calibration is to match the head distribution that is the output from the model with the observed heads. To this end, about 93% of model cells have simulated heads within 5 m of observed heads, and about

Salinisation, Murray Basin

157

Table 13.1 Modelled water and salt balance for the Lower Darling Area. Source Recharge from rainfall Recharge from irrigation Stream loss from Darling Stream loss from Tallyawalka Stream loss from Anabranch Stream loss from Murray Loss from Menindee Lakes Loss from Lake Victoria Murray base flow Darling base flow Inflow from Cretaceous Inflow from layers 1-4 Inflow from layer 5 Total source

Water flow (ML/d)

Salt flux (t/d)

26.2

0.2

12.5

3.3

4.0

1.2

0.3

0.1

0.3

0.1

4.1

1.0

10.7

4.5

2.6

1.1

0.4

12.6

9.7

139.8

0.3

3.2

0.1 71.2

1.1 168

half within 2 m. The standard deviation of the residual between simulated and observed heads was 2.25 m over the entire model domain. This is a good result considering the paucity of reliable water-level data in the area and the fact that the majority of the bores used in the calibration have not been levelled. In addition to simulating heads, the groundwater model also outputs the rate of water flow between the shallow aquifer and the streams within each model cell. W h e n multiplied by an estimate of groundwater salinity, this gives the salt loads for the various reaches of the Murray River, which can be compared with previous estimates. The salt loads predicted for most of the reaches are 3 0 - 7 0 % of values used in river salinity modelling, which is appropriate since the latter considers groundwater inflow from both sides of the river. Once calibrated, the model could be used as a predictive tool, investigating how changes in land and water management will effect groundwater levels and salt loads into the rivers.

MODEL OUTCOMES The modelled water and salt balance for the area is summarised in Table 13.1. In terms of input to the groundwater system, recharge from rainfall is the largest contributor. Even though recharge under mallee vegetation is less than 0.1% of rainfall, it adds over 2 6 ML/d into the groundwater system because of the large model area. Conversely, the irrigation areas cover less than 0.5% of the model area, but provide accessions of 12.5 ML/d. Leakage also occurs from the major storage systems of the Menindee Lakes (10.7 ML/d) and Lake Victoria (2.6 ML/d). Groundwater transfers from the underlying Lower Cretaceous accounts for 14% or 9.7 ML/d of input into the Cenozoic aquifers. This occurs in the southwest half of the model where upward flow conditions prevail. In this steady-state simulation, the River Murray loses about the same volume of water to the shallow aquifer (4.1 ML/d) as the Darling and its tributaries (4.5 ML/d). The difference is that stream losses in the Murray occur in a small

Sink Discharge from salinas Woolpunda North scheme Stream gain to Darling Stream gain to Tallyawalka Stream gain to Anabranch Stream gain to Murray Gain to Menindee Lakes Gain to Lake Victoria Murray base flow Darling base flow Outflow to Cretaceous Outflow to layers 1-4 Outflow to layer 5 Total sink

Water flow (ML/d)

Salt flux (t/d)

-12.2

-10.0

-190.0

<0.1

-0.6

<0.01

-27.8

-631.6

<0.1

-20.1

-278.1

-0.2

-3.0

-0.7

-10.6

-71.0

1114

number of model cells (9 out of 51) compared to the vast majority of Darling cells (102 out of 113). The stream losses in the Murray occur at or immediately upstream of the operational locks where the river levels are maintained artificially high. The river leakage of 0.4 ML/d predicted upstream of Lock 6 at Chowilla is a good example as it corresponds to a flushed zone of low salinity groundwater in the floodplain. Hence, river regulation has dramatically changed the dynamics between the river and the shallow aquifer and the model could simulate these regional changes. Although the Darling is a losing stream for most of its length, most of the leakage (94%) occurs upstream of Menindee. This reflects the extensive weir pool and the freshening of groundwater in the surrounding alluvial aquifer. As expected, the River Murray is the major sink for groundwater in the region, receiving seepages totalling 2 8 ML/d, resulting in a salt load of 6 3 2 t/d. The highest inflows occur in reaches adjacent to irrigation areas, particularly Berri, Renmark and Mildura. The Woolpunda scheme pumps 10 ML/d of saline groundwater flowing from the north, extracting 190 t/d of salt. The groundwater inflows into the Darling River are insignificant, totalling about 0.04 ML/d downstream of Burtundy. The underlying Lower Cretaceous is a significant sink for groundwater from the Tertiary aquifers in the northeast half of the model where downward leakage prevails. About a quarter of the 2 0 ML/d lost to the Lower Cretaceous occurs near the Menindee Lakes. Groundwater also exits out of the model where flow of 0.7 ML/d continues southward in the basal Lower Renmark aquifer (Layer 5) under the Woolpunda Reach. In addition, the active salt lakes are estimated to be evaporating over 12 ML/d out of the shallow aquifer.

MODEL PREDICTIONS The Lower Darling model was constructed to provide predictions of the likely salinity impact of land and water management practices in the region, in particular the effect of

158

R. S. Brodie

broad-scale native vegetation clearing. If post-clearing recharge is assumed to be 1% of rainfall, a figure commonly used in other models (Kellett 1997) then salt loads into the River Murray are predicted to increase from a baseline of 632 t/d to 1047 t/d. This equates to an 80 EC increase in average river salinity at Morgan, which is the entire reduction target for the Salinity and Drainage Strategy. Other streams also experience increased groundwater seepage with salt loads in the Darling River increasing from 0.6 t/d to 2.7 t/d. Parts of the Darling Anabranch switch from being a losing stream to gaining saline groundwater. This would have undesirable consequences for water quality in the small supply dams constructed in the channel that rely on a single annual replenishment from the Menindee Lakes. The scenario also suggests a similar problem for Lake Victoria, where saline groundwater seepage may develop along the lake margins. The model predicts that the water-table rise associated with the mallee clearing can exceed 2 0 m, particularly under the anticipated cleared areas north of the Murray in South Australia and the extreme west of New South Wales as well as east of the Darling. This has the effect of reactivating or expanding salinas and greatly increasing the risk of land salinisation in these areas. Other modelling scenarios were developed which dealt with the infrastructure used in river regulation and salinity mitigation. For example, removal of the Woolpunda Salt Interception Scheme bores north of the river resulted in predicted salt loads to the Woolpunda Reach increasing from 22 t/d to 145 t/d. The water-table in the Murray Group limestone recovered by 6 - 1 2 m in the vicinity of the bores. The effects of decommissioning the Menindee Lakes storages was also investigated. Downward leakage from the lakes is a significant contributor to the underlying aquifers, indicated by the water-table mound and associated dilution effect. Removal of the water storage resulted in a general decrease in heads in all aquifers, radially outward from the lakes. In the shallow aquifer, the water-table was lowered 7 - 1 1 m within a 2 0 km radius. Only a small decrease in salt loads to the Murray resulted. This probably relates to the vast distance along the flow line between the Menindee Lakes and the River Murray, so that the head decrease does not greatly effect the regional hydraulic gradient.

CONCLUSIONS Geoscience has a valuable role to play in the amelioration of one of Australia's most significant environment issues, that of land and water salinisation. As an example, the Lower Darling groundwater flow model was able to provide valuable technical advice towards appropriate management of the salinity in the River Murray. First, the model summarised the regional hydrogeological setting north of the river between Mildura and Morgan, including defining the regional aquifer systems, the major components of the water and salt balance and summarising the key hydrodynamic processes between the streams and the shallow aquifer. Second, the model was able to predict how such pressing issues as clearing of native vegetation would adversely effect salinities in the River Murray. Importantly, the predictions of increased salt loads could be readily

translated into the benchmark used in the overarching Salinity and Drainage Strategy, namely the average river salinity at Morgan. The model indicated that the increased salt load into the river due to vegetation clearing in the area would basically negate all of the salt diversion efforts of the Salinity and Drainage Strategy to date.

ACKNOWLEDGMENTS The Lower Darling groundwater model was work funded by the Murray-Darling Basin Commission when the author worked for the Australian Geological Survey Organisation (AGSO). The construction of a regional-scale groundwater model requires input from many people. Data, reports and advice were contributed from people in various state agencies in South Australia (Steve Barnett, Nick Watkins, Andrew Telfer, Phil Pfeiffer) and New South Wales (Michael Williams, Hugh Milner, David Salotti, Sarah Bish, David Harriss, Mike Erny, Peter Clarke, Phil Craven, Alan Hassett, John Hill, Karen Lawson). Information was also derived from other agencies such as the Bureau of Rural Sciences (Phil Tickle, Simon Veitch) and the Murray-Darling Basin Commission itself (Andy Close, Ben Dyer, Paul Nanninga). The massive effort of data collation and management was provided in AGSO by Evert Bleys, Stephen Hostetler, Rob Kingham, Estoban Lopez, Vicki Manson and Heather Rennie. The guidance given by Ray Evans, Jim Kellett and Jay Punthakey were crucial during the calibration phase. Jim Kellett and Gerry Jacobson reviewed the original manuscript of this chapter.

REFERENCES R. 1 9 9 5 . The rise and rise of confined aquifer pressuresWhy? In: Extended Abstracts Murray Darling 1995 Workshop, Wagga Wagga, pp. 2 4 - 2 5 . Australian Geological Survey Organisation Record 1995/61. BOWLER J . M. 1988. Environmental and salinity history of the Murray Basin in the last 500,000 years. In: Extended Abstracts, Murray Basin 88 Conference, p. 7. Bureau of Mineral Resources Record 1988/7. BOWLER J . M. 1990. The last 500,000 years. In: Mackay N . & Eastburn D. eds. The Murray, p. 95-111. Murray Darling Basin Commission. BRODIE R. S. 1998. The Lower Darling regional steady state groundwater flow model. Australian Geological Survey Organisation Record 1998/19. BROWN C. M. 1989. Structural and stratigraphic framework of groundwater occurrence and surface discharge in the subsurface of the Murray Basin, southeastern Australia. BMR Journal of Australian Geology & Geophysics 11, 367-385. BROWN C. M. & STEPHENSON A. E. 1991. Geology of the Murray Basin, southeastern Australia. Bureau of Mineral Resources Bulletin 235. BARNETT S .

COOK P. G . , KENNETT-SMITH A . K . , WALKER G . R . , B U D D G . R . , WILLIAMS

R . M . & ANDERSON R . 1 9 9 6 . Impact of dryland agriculture on land and river salinisation in the western lands, New South Wales. CSIRO Technical Memorandum 96/16. EVANS W . R. & KELLETT J. R. 1989. The hydrogeology of the Murray Basin, southeastern Australia. BMR Journal of Australian Geology & Geophysics 11, 147-166. KELLETT J. R. 1 9 8 9 . The Ivanhoe Block—its structure, hydrogeology and effect on groundwaters of the Riverine Plain of New South Wales. BMR Journal of Australian Geology & Geophysics 11, 333-353. KELLETT J. R. 1997. Lachlan Fan/Ivanhoe Block steady state groundwater model—model development, calibration, sensitivity analysis and predictions. Australian Geological Survey Organisation Record 97/029.

Salinisation, Murray Basin MCDONALD M . G . & HARBAUGH A. 1 9 8 8 . A modular three-dimensional

finite difference groundwater flow model. In: Techniques of Water Resources Investigations of the United States Geological Survey, Book 6 Chapter Al. US Department of the Interior. MURRAY-DARLING BASIN COMMISSION 1999a. Salinity and Drainage Strategy—ten years on. Murray-Darling Basin Commission. MURRAY-DARLING BASIN COMMISSION 1999b. The Cap: Providing security for water users and sustainable rivers. Murray-Darling Basin Commission MURRAY-DARLING BASIN MINISTERIAL COUNCIL 1 9 9 9 . The Salinity Audit

159

of the Murray-Darling Basin—a 100 year perspective. Murray-Darling Basin Ministerial Council. 1986. Lake Bungunnia—a Plio-Pleistocene megalake in southern Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 57, 137-156. WILLIAMS M . A. J . 2 0 0 1 . Quaternary climatic changes in Australia and their environmental effects. Geological Society of Australia Special Publication 21,3-11, STEPHENSON A. E.

Received 24 March 2000, accepted 23 May 2000

Geological Society of Australia Special Publication 21, 161-169

Chapter 14—Environmental geoscience issues in the Gippsland Basin, Victoria Department of Outdoor Education and Nature Tourism, La Trobe University, PO Box 199, Bendigo, Vic. 3552, Australia (I.hodgson@bendigo.latrobe. edu. au). The Gippsland Basin in southeastern Victoria is one of the major Mesozoic-Tertiary sedimentary basins in Victoria. Formed in the Early Cretaceous and influenced by tectonics, climate and sealevel change, up to 900 m of marine and terrestrial sediments have been preserved. These deposits contribute significantly to Victoria's resource needs. Brown coal, mined in large-scale opencut mines near Morwell, Loy Yang and Yallourn, supplies most of Melbourne's electricity demands. Offshore reserves of natural gas and petroleum have been developed since the 1960s with the offshore oilfields in Bass Strait being the largest in Australia. Groundwater plays a significant role in the supply of domestic water to Sale, Boisdale and Rosedale as well as water for irrigation and stock use. The coastal deposits around the Gippsland Lakes are important for tourism, and include wetland habitats recognised for their high conservation value under the international Ramsar Treaty. Studies of coastal Quaternary deposits have contributed to an understanding of environmental change. Since European settlement, exploitation of these resources has led to environmental changes including land subsidence to over 350 mm as a result of coal, oil and gas extraction. Rising groundwater levels following clearing and development of irrigation have resulted in waterlogging and salinity problems in the region, particularly west of Lake Wellington. Indirect impacts include river channel changes as a result of straightening and desnagging streams, and erosion of the Gippsland Lakes resulting from opening a permanent entrance to these lakes. Better management of the region in the future depends on an understanding of interactions between the natural processes that contribute to environmental problems and, importantly, on cooperation between the various organisations responsible for resource management. KEY WORDS: coal mining, environmental impact, geology, geomorphology, Gippsland Basin, groundwater, land subsidence, petroleum industry, salinity, tourism.

INTRODUCTION The Gippsland region of Victoria contributes significantly to Victoria's resource needs. Brown coal, mined in largescale opencut mines near Morwell, Loy Yang and Yallourn, supplies 85% of Melbourne's electricity demands. Since 1924 over 1000 Mt have been extracted, and for the foreseeable future, the industry is still economically viable (Barton et al 1992). Offshore reserves of natural gas and petroleum have been developed since the 1960s (Smith 1988), with the offshore oilfields in Bass Strait being the largest in Australia. Groundwater plays a significant role in the supply of domestic water to Sale, Boisdale and Rosedale as well as water for irrigation and stock use. The natural features, particularly along the coast, are important for the development of tourism (Tourism Victoria 1999), conservation of internationally recognised wetland habitats (Hodgson 1995), and understanding climate change during the Quaternary (Rosengren et al 1991). As a result of development of the region's resources, a wide range of environmental issues have arisen in Gippsland, many of which are related to its geological and geomorphological setting. These include land subsidence, salinity and waterlogging, nitrate pollution, stream-channel changes, coastal erosion, and erosion of the Gippsland

Lakes (Lawrence 1992). The aim of this chapter is to show the importance of geology and geomorphology in understanding contemporary environmental issues in Gippsland.

GEOLOGICAL SETTING OF THE GIPPSLAND BASIN The Gippsland Basin is one of the major Mesozoic-Tertiary sedimentary basins in Victoria. It covers an area of approximately 56 000 km 2 , although four-fifths of the basin lies offshore (Lawrence 1992). The basin is bounded by the Lower Cretaceous blocks of the Southern Uplands in the west; the Palaeozoic Eastern Highlands'to the north; the Bassian Rise in the south and southwest, while the eastern margin is open to the Tasman Sea (Jenkin 1981) (Figure 14.1, 14.2). The present-day onshore Gippsland Basin is a low-lying coastal plain gently sloping south to southeast. Six major rivers originating in the Eastern Highlands drain into the Gippsland Lakes, a series of coastal lagoons cut off from the ocean by an extensive Holocene barrier system. The Gippsland Basin was formed by continental rifting during the Early Cretaceous followed by subsidence during the Paleocene. This resulted in the formation of an extensive trough which enabled the onshore deposition of up to

162

L. Hodgson

Okm

15km

l i i ^ a

O n s h o r e Gippsland Basin

Figure 14.1 Gippsland Lakes Catchment Victoria (after Aldrick et al 1988).

Eastern Highlands

.......... Boundary of Gippsland Basin

Figure 14.2 Structural framework of the Gippsland Basin (after Walker & Mollica 1990).

900 m of unconsolidated continental and marine sediments overlying the bedrock of consolidated Cretaceous sandstone and mudstone (Jenkin 1981). The pattern of deposition in the basin since the Late Cretaceous has been influenced by a combination of tectonics, eustatic sea-level changes and the supply of sediment (Jenkin 1968). Tectonic movements dominated during the Late Tertiary and have influenced the large-scale morphology of east-west-trending anticlinal highs (the Baragwanath Anticline) and synclinal depressions (the Latrobe Valley Depression, Moe Sunkland, and the Lake Wellington and Seaspray Depressions) (Jenkin 1973) (Figure 14.2). Eustatic sea-level changes dominated during the Quaternary although the effects of tectonics could be masked by evidence of sealevel changes (Jenkin 1981). Continental sediments in the basin have been derived from the Eastern Highlands and Southern Uplands while marine carbonate sediments have formed in the offshore areas.

Environmental geoscience, Gippsland Basin

0.3

163

0.6 km

SHALLOW AQUIFER SYSTEM Tph Haunted Hills Formation and Quaternary Sediments

LATROBE VALLEY COAL MEASURES BALOOK/BOISDALE FORMATION AQUIFER SYSTEM Tpb Boisdale Formation Tob Balook Formation Tol Latrobe Valley Coal Measures

JEMMY'S POINT/SEASPRAYGROUP AQUIFER SYSTEM Tpj Jemmy's Point Formation Tml Gippsland Limestone Formation Tom Lakes Entrance Formation

Figure 14.3 Onshore Gippsland Basin stratigraphy (after Walker 1992).

LATROBE GROUP AQUIFER SYSTEM Tel Latrobe Group I I Basement Strzelecki Group Direction of groundwater flow within LVCM/Balook/Boisdale Formation Aquifer System

Downwarping in the Gippsland Basin from the Late Cretaceous to mid-Tertiary led to extensive deposition of sand, clay and coal of the Traralgon Formation and Latrobe Group (Figure 14.3). These deposits thin out towards the western margins of the basin. In the Late Eocene to Early Oligocene, basalt from volcanic eruptions in the Southern Uplands flowed into the western basin. This was followed by extensive downwarping in the coal basins, and deposition of the Latrobe Valley coal measures continued until the mid-Miocene (Bolger 1991). The coal measures were deposited during warm periods of low energy accompanied by the formation of meandering river systems, shallow lakes and swamp environments and the presence of plant growth conducive to coal development (Bolger 1991; Barton et al 1992). Accumulation of thick coal deposits was also facilitated by the presence of a stable coastline. These conditions have led to development of extremely thick coal beds in the Gippsland Basin, much thicker than in other Tertiary sedimentary basins in southeastern Australia (Bolger 1991; Barton et al 1992) In the mid-Miocene large-scale tectonic movement resulted in the formation of the Baragwanath Anticline which separated the Latrobe Valley from the coastal Seaspray Depression. Subsequent deposits of coal measures were confined to the northern Latrobe Valley Depression. The Baragwanath Anticline became an erosion surface supplying sediment to the Seaspray Depression and the Latrobe Valley. At the same time as the coal measures were being deposited, major Tertiary marine transgressions occurred

which reached a maximum in the early mid-Miocene. It was then that deposition of thick marl and limestone took place creating the Gippsland Limestone. Regression of the Tertiary sea began in the mid-Miocene and continued until the midPliocene, although there is some evidence of short transgressive phases in the Seaspray Depression in the Late Miocene and Pliocene (Jenkin 1981). At the end of the Pliocene the shoreline had approximately reached its present position. Downwarping in the Late Miocene and Pliocene in the central and eastern section of the Gippsland Basin resulted in the deposition of sand and gravel of the Boisdale Formation in the central and southern parts, interfingering with fine sediments of the Coongulmerang beds further east, associated with lacustrine conditions. Uplift during the Pliocene resulted in stream rejuvenation and erosion in the Eastern Highlands, and extensive deposition of sediments of the Haunted Hill Formation (Bolger 1991) in the Gippsland Basin. This formation is dominated by extensive beds of sandy clay and clayey sand but includes a wide range of particle sizes from boulders to fine clays (Jenkin 1968). It extends over most of the onshore part of the basin although it is thickest against its northern edge where alluvial fans formed as coarse gravel was deposited in response to change in gradient from the highlands to the coastal plain (Jenkin 1968). The marked change in composition from the coal-bearing Latrobe Group to the Haunted Hill Formation is indicative of a climate change from warm wet conditions to a colder arid phase, with higher stream discharge required to transport the boulders (Bolger 1991). Evidence of a global climate

164

L. Hodgson PRIOR BARRIER

£E

INNER BARRIER

OUTER BARRIER

- PLIOCENE

PLEISTOCENE

>> <

HOLOCENE

— >

Figure 1 4 . 4 Sea-level changes during the Quaternary (Hodgson 1995).

change at this time supports this (Bowler 1982). However Jenkin (1968) and Bolger (1991) considered the climatic influence secondary to the stream responses to tectonic uplift. Reworking of the unconsolidated Haunted Hill Formation during the Quaternary is largely responsible for many landforms present today. Although the Haunted Hill Formation is thin relative to other deposits in the basin, it is important from an environmental-management perspective because it comprises unconsolidated surface deposits which are most affected by changes in the environment. These may be long-term effects of tectonics and sea-level changes, or the more recent effects of human interference. Large-scale tectonic movements that characterised the Tertiary were not so pronounced during the Quaternary where the major influence on depositional patterns has been eustatic changes in sea-level (Jenkin 1973). Jenkin (1968) identified seven major phases of sea-level changes during the Quaternary. The marine and terrestrial events associated with these sea-level changes are summarised in Figure 14.4. During periods of low sea-level the shoreline extended seaward, rivers became entrenched in their valleys, which extended beyond the present coastline, and terrestrial geomorphic processes dominated (Jenkin 1968). During periods of high sea-level, the shoreline retreated landward, river valleys aggraded and extensive floodplains formed. At the coast, barrier and beach-ridge formation occurred as a result of shoreward movement of sediment during periods of sea-level rise. Bird (1978) identified three barrier systems in the Gippsland Basin: an early Pleistocene prior barrier system, a mid-Pleistocene inner barrier system, and a narrow outer barrier system which formed during the Holocene sea-level rise (Figure 14.5). Coastal lagoons, which became the Gippsland Lakes, formed landward of both the inner and outer barrier systems. River drainage was influenced by the position of the barrier systems. During the late Pleistocene marine regression, when sea-level fell to more than 140 m below present levels, rivers became deeply entrenched in their channels (Jenkin 1968) and flow was directed between the prior and inner barrier systems. The beach ridge and barrier systems of the Gippsland Lakes are considered to be of national significance (Rosengren et al

1991) because of the evidence they provide in understanding barrier system evolution during the Quaternary. High-level east-west-trending sand ridges and sand sheets that occur around Seaspray and north of Lake Wellington were thought to be relict coastal dune systems by Jenkin (1968) and Ward (1977). However, their alignment and sediment composition are more indicative of terrestrial deposition during the last glaciation (Jenkin 1981; Hill & Bowler 1995). In summary, the present-day Gippsland Basin is a region in which the terrestrial and coastal geomorphic processes that have influenced the geomorphology of the region during the Quaternary are still active and responsive to changes in environmental conditions.

GEOLOGICAL RESOURCES IN THE GIPPSLAND BASIN Energy resources Major brown-coal deposits occur in the Latrobe Valley Group sediments in the western section of the Latrobe Valley Depression near Morwell, Yallourn and Loy Yang. The coal is extracted by large-scale opencut mining. Mining commenced in 1924 and, for the foreseeable future is still economically viable (Barton et al 1992). The brown coal is converted into electricity and supplies 85% of Melbourne's electricity demands as well as briquettes for industrial and domestic use. The total production from 1924 to 1990 was 1113 Mt, with 30% coming from Morwell and 60% from Yallourn (Barton et al 1992) although, at present, the major tonnage comes from Loy Yang. Associated with mining of the coal seams is an extensive dewatering programme to maintain pressures within the seams and facilitate placement of overburden. Natural-gas reserves were first discovered offshore in 1964, and 254 development wells have been drilled between 1964 and 1987, with 11 commercial fields having been developed (Smith 1988). Gas is piped to the Longford Gas Plant where it is treated before being transhipped to other areas in Victoria. The discovery of oil onshore near Lakes Entrance in 1924 initiated petroleum exploration in Gippsland, but it was not until 1965 that large-scale offshore exploration and the development of extensive offshore petroleum reserves commenced. These offshore oil fields in Bass Strait are the largest in Australia although the reserves are currently in decline.

Groundwater resources Exploitable groundwater resources in Gippsland are associated with aquifer systems in the Latrobe Valley Group, the Boisdale Formation and the shallow Quaternary deposits. Groundwater flow in all these systems is regional with flow paths from the highlands towards the coast (Walker & Mollica 1990) although, particularly in the shallow aquifer system, localised groundwater flow systems are superimposed on the regional system (Hodgson 1995). Large volumes of groundwater are extracted from the Latrobe Valley Aquifer System to reduce pressure associ-

Environmental geoscience, Gippsland Basin

165

Figure 14.5 Barrier systems of the Gippsland Lakes (after Bird 1993). ated with opencut mining in the Morwell-Loy Yang area and for the Longford Gas Plant. Thermal groundwater (50-60°C) also occurs in the aquifers in the Latrobe Valley Group which offers a potential source of energy which has not yet been exploited (Leonard 1988). Good quality groundwater (300-600 mg/L) occurs in confined aquifers of the Boisdale Aquifer System (Walker & Mollica 1990) which extends from Rosedale to the east of Lake Wellington. The most significant aquifers occur west of Lake Wellington, at approximately 45, 90 and 120 m below the surface where artesian pressures occur. The groundwater is used as a water supply for the towns of Sale, Boisdale and Rosedale as well as providing irrigation water to landholders, in particular, around Lake Wellington. The Shallow Aquifer system (Harris 1976) refers to aquifers which occur in the Plio-Pleistocene and Holocene deposits with the most significant area of groundwater occurrence in the region between the Latrobe and Macalister Rivers. Groundwater in this area is associated with palaeochannels of Pleistocene streams (Figure 14.6) but it is difficult to locate high-yielding sections and yields are highly variable. Extracted groundwater is used predominantly for stock and irrigation. The quality of the water varies from approximately 700 mg/L in the recharge areas to over 15 000 mg/L (Walker & Mollica 1990) in the dis-

charge areas around Lake Wellington, making the groundwater here unsuitable for use. Since 1993, salinity management in the region has led to an increase in reuse of higher quality groundwater for irrigation (Lake Wellington Catchment Community Working Group 1993). Shallow groundwater on the Mitchell River floodplain is also exploited, primarily as a supplement for irrigation and stock water during drought and later summer months. The potential of artificial recharge of this aquifer has been investigated but never pursued (Thompson 1973).

ENVIRONMENTAL IMPACTS ARISING FROM RESOURCE USE Since European settlement of Gippsland, a little over 150 years ago, exploitation of the region's resources has led to changes to both the surface and subsurface environments. These effects arise from direct exploitation of resources including coal mining, development of offshore petroleum supplies, and onshore and offshore gas supplies. They include indirect impacts from such activities as subdivision and land clearing for agriculture, opening of a permanent entrance to the Gippsland Lakes, clearing of snags in rivers and, more recently, development of tourism in the region.

166

L. Hodgson

SALE 13

Figure 1 4 . 6 Map showing surface evidence of palaeochannels of the Thomson and Macalister Rivers. The dotted lines indicate areas where surface expression is obscure (after Jenkin 1968).

This section will discuss the major geological and geomorphological issues that arise as a result of exploitation of the region's resources: land subsidence, salinity, river changes and coastline changes.

Decline in groundwater levels: land subsidence It is well-established that the depressurisation of confined aquifer systems will eventually lead to compaction of associated deposits and to land subsidence (Freeze & Cherry 1979). Extraction of coal onshore, and petroleum and gas offshore from the Latrobe Valley Group necessitates largescale dewatering from the confined aquifers (Table 14.1).

1975

1977

1979

1981

1983

1985

1987

1989

1991

1993

Figure 1 4 . 7 Decline in head in Sale 13 Bore (1 km west of Lake Wellington) 1 9 8 3 - 1 9 9 2 , as a result of groundwater extraction (Walker 1992).

potential impact on sensitive low-lying areas around the coast as a result of subsidence include erosion of the outer barrier, destruction of the barrier island system around Corner Inlet, changes to the shallow groundwater system and associated impacts on the wetlands and lakes. Although production of oil and gas in the basin is declining, a lag effect suggests that the impact of subsidence in the coastal zone may be felt for some time. Walker (1992) considered that further monitoring is imperative as is the need to better understand the variability of the Latrobe Valley Group sediments in terms of thickness and compressibility of overlying sediments if the localised impacts of subsidence are to be identified.

Table 1 4 . 1 Groundwater extraction from coal, gas and petroleum mining in the Latrobe Valley Group aquifer, Gippsland, Victoria. Volume (ML/y) Coal (Morwell opencut) a Coal (Loy Yang) b Petroleum (offshore) 3

23 0 0 0 5859 75 0 0 0

Gas (Longford) 3

2640

Irrigation 3

9300

Walker 1992

LoyYang Power, pers. comm. 1999 (average for 1998/99)

Dewatering of the Latrobe Valley Aquifer System associated with the open-cut coal mines has led to lowering of the water-table near Morwell by more than 130 m over the last 40 years and the impact of this decline extends as far as Rosedale, approximately 30 km away (Walker 1992). Longterm monitoring of a bore on the western shore of Lake Wellington has shown a steady rate of decline of 1.6 m/y (Walker 1992) although the variation across the region is from 1.0 to 1.6 m/y (Figure 14.7). This decline in head is thought to be correlated to gas and oil production offshore (Lawrence 1992). Although there is clear evidence of decline in water levels from bore monitoring, the extent of the resulting land subsidence is not clear. Estimates range from 35 mm to more than 350 mm near the coast (Walker 1992) and a maximum of 1.8 m in the Moe Basin (Lawrence 1992). The

Rising groundwater levels: salinity and waterlogging While declining groundwater levels is a problem in deeper groundwater systems associated with the Latrobe Valley Group Aquifer System and the Boisdale Aquifer System, in the Shallow Aquifer System of central Gippsland, rising groundwater levels and associated waterlogging and salinity problems are the major concern. Groundwater flow in the shallow system is primarily associated with palaeochannels of early and late Pleistocene river systems. The regional flow is from the highlands towards the coast, with regional discharge occurring around Lake Wellington. Discharge also occurs north of the Snake Ridge Monocline where uplift of the monocline impedes groundwater flow. Before European settlement, saline areas occurred along river floodplains and particularly in low-lying wetland areas surrounding Lake Wellington (Hodgson 1995). While clearing has been responsible for changes in the groundwater hydrology of the basin, it was intensive irrigation in the 1950s that led to a rapid rise in groundwater levels occurring north of the Snake Ridge Monocline. This resulted in community pressure to install a detailed network of monitoring bores in the region, and regular monitoring of this system since the 1960s provides the most detailed record for the state (Mollica 1991). More recently, concern over increase in salt-affected land, in particular in sensitive wetlands around Lake Wellington, led to the development of a Lake Wellington Catchment Salinity Management Plan

Environmental geoscience, Gippsland Basin (Lake Wellington Catchment Community Working Group 1993) which proposed a wide variety of measures to address the problem. In particular, the focus has been on groundwater management in the Macalister Irrigation District to the west of Lake Wellington. While that catchment was perhaps the most severely affected, other significant salt affected areas in the basin occur near Bairnsdale and at Yarram.

River channel changes Seven major rivers flow from the Eastern Highlands to the unconsolidated alluvial plains of the Gippsland Basin (Figure 14.1). Channel changes on such floodplains occur naturally as the river adjusts to variations in flow regime, particularly in response to floods. The changes include within-channel alteration to sediment delivery, changes to the form of the meandering channel, including meander cutoffs, and channel avulsion in which the entire course of the channel is diverted (Summerfield 1991). Aerial photographs provide evidence of channel changes that have occurred in the Gippsland Basin during the Quaternary, in particular meander cutoffs and channel avulsion. Channel changes accelerated after European settlement (Erskine et al 1990), and have resulted in both changes to the channel form and also changes in the channel gradient, particularly downstream in response to increased sediment delivery. The changes are in part a response to direct changes to the channel, such as straightening and desnagging of the Thomson, Latrobe, and Mitchell Rivers, and to the building of levee banks along the Avon and Latrobe Rivers. Other changes include indirect impacts from mining (increased sediment), opening of a permanent entrance to the Gippsland Lakes (increased erosion of deltas of the Tambo and Mitchell Rivers), and agriculture (clearing and increased sediment). In spite of an increase in sediment transport following European settlement, evidence suggests that most of the sediment is deposited in the lower reaches of the river rather than the lakes. For example, the Avon River from its mouth to Redbank, 2 km upstream, was once a deep channel which boats would enter. Since European settlement the channel has filled, and the river is now a wide braided stream (Bird 1978). Grayson et al (1998) have shown that increased sediment in the Latrobe River has been deposited upstream from the mouth, not into Lake Wellington. This suggests that the source of increased deposition in the Lakes of 0.6 mm/y (Erskine et al 1991) is not dominantly of fluvial origin. Dams constructed on the major streams in the region have reduced the frequency of flooding, although large floods still occur and have lead to river avulsion, most notably the avulsion of the Avon River and the formation of Rainbow Creek in 1954 (Brizga 1984). While river avulsion in Gippsland is a natural occurrence (Jenkin 1968), under the current land use in the region it can cause considerable damage (Brizga 1984, 1990) and its recurrence poses significant challenges for land managers (Erskine et al 1990).

Coastline changes The low lying areas of the coastal plain encompass the barrier islands around Corner Inlet, the exposed coast along

167

the Ninety Mile Beach and the Gippsland Lakes lagoonal systems behind the Ninety Mile Beach. These areas are considered by Rosengren et al (1991) to be of national geomorphic significance because of the evidence they provide to help understand beach ridge and barrier development since the Pleistocene. OPENING OF A PERMANENT ENTRANCE TO THE GIPPSLAND LAKES Before European settlement, the outer barrier system fully enclosed the Gippsland Lakes except during severe storms when breaches in the sand barrier occurred allowing saltwater to enter the lakes system (Bird 1978). In order to facilitate the development of a port at Sale, a permanent entrance was cut into the barrier system in 1889 at Lakes Entrance. Following construction, changes occurred in the Gippsland Lakes, notably an increase in the salinity of the lakes, increased erosion of lake shore and delta of the Mitchell and Tambo Rivers (Bird 1978, 1983), and a fall in lake levels. The extent to which opening of the entrance has contributed to these changes has not been completely resolved. Variations in lake height are attributed to changed tidal conditions and more rapid escape of flood waters rather than a real change in elevation (Parliamentary Public Works Committee 1952). Skene (1952) suggested that lake levels have risen since the entrance was constructed due to increased sedimentation in the lake but recent work shows no evidence of increased sedimentation in Lake Wellington (Grayson et al 1998). A proposal to build a barrage across the narrow straits at the eastern end of Lake Wellington (Bird 1987) to maintain freshwater conditions in the lake has become a contentious issue. Uncertainty prevails about increased salinity that may be derived from land-based activities (clearing, irrigation and reduced flooding), and increased salinity which may be derived as a result of opening the permanent entrance. Understanding the fluvial, groundwater and coastal processes occurring at Lake Wellington is critical to effect appropriate integrated management. COASTLINE EROSION Erosion of the outer barrier has occurred in recent times with retreat from 50 to 100 m. However this rate of change is slow compared to world-wide trends (Bird 1993, 1978). The cause of this could be a world-wide rise in sea-level (Bird 1993) or more locally from the depressurisation of offshore petroleum reserves and resultant subsidence. Either way, the potential damage to the coastline from reduction in the width of the barrier system could be significant, ranging from damage to property, to increased breaches of the barrier into the lakes and changes to the ecology of the Gippsland Lakes System.

THE FUTURE The major forms of development that have had an impact on the geological resources in Gippsland are land clearance for agriculture (salinity and water logging, river-channel changes, changes to the Gippsland Lakes), onshore and

168

L. Hodgson

offshore mining (river-channel changes, subsidence, coastline changes) and potentially tourism (coastal erosion). With the energy resource extraction past its peak (Smith 1988) and agricultural development stabilised, the most significant impact is likely to come from tourism development in the area. It is a clear policy of Tourism Victoria (1999) to increase tourist numbers in the region with the emphasis on resort development as well as activitybased/adventure-style development focusing on exploitation of the region's natural resources, in particular its coastline. As tourist numbers increase so too does the potential impact on the land and water resources (Hinwood & McLean 1992). Recent housing development on the inner and outer barrier systems and the sinking of septic tanks has led to leakage from septic tanks. This is particularly significant where the water-table is rising (Lawrence 1992). Building and development on dunes could severely affect dune processes leading to erosion of the outer barrier system. As discussed before, the potential of land subsidence from offshore oil production and/or sea-level rise due to greenhouse gases could significantly alter the coastal system and therefore affect the extent of tourist development in the region. The management of Gippsland's geological resources may depend on an understanding not only of the cause and effect of each individual problem, but also of the degree of interaction between the natural processes. While the former may be well-understood, the effects these may have on any other activity are less well-known and require cooperation between various organisations responsible for resource management to ensure the long-term sustainable use of the region's bountiful resources.

REFERENCES ALDRICK J . M . , HOOK R . A . , VAN DE GRAAFF R . H . M . , NICHOLSON B . M . ,

O'BIERNE D . A . & SCHOKNECHT N . R . 1 9 8 8 . A Study of Land in the Catchment of the Gippsland Lakes. TC-17, Volume 1. Land Protection Division, Department of Conservation, Forests and Lands, Victoria.

BARTON C . M . , BOLGER P. F., HOLDGATE G . R . , THOMPSON B . R .

The brown coal geology of the Gippsland Basin. In: Barton C. M., Hill K., Abele C., Foster J. & Kempton N. eds. Energy, Economics and Environment: Gippsland Basin Symposium, pp. i-xiv. Australasian Institute of Mining and Metallurgy, Melbourne. BIRD E. C. F. 1978. Geomorphology of the Gippsland Lakes Region. Ministry for Conservation, Victoria., Environmental Studies Series 186. BIRD E. C. F. 1983. Shoreline changes in the Gippsland Lakes. Proceedings of the Royal Society of Victoria 95, 227-235. BIRD E . C . F. 1 9 8 7 . The Past, Present and Future of the Gippsland Lakes. In: Pit, M. W. & Synan T. P. eds. The Past, Present and Future of the Gippsland Lakes, pp. 1-8. Save the Gippsland Lakes Committee Symposium, Sale. BIRD E . C . F 1 9 9 3 . The Coast of Victoria: the Shaping of Scenery, University of Melbourne Press, Melbourne. BOLGER P. F. 1991. Lithofacies variations as a consequence of Late Cainozoic tectonic and palaeoclimatic events in the onshore Gippsland Basin. In: Williams M. A. J., De Deckker P. & Kershaw A. P. eds. The Cainozoic in Australia: a Re-appraisal of the Evidence, pp. 158-180. Geological Society of Australia Special Publication 18. BOWLER J . M . 1 9 8 2 . Aridity in the late Tertiary and Quaternary of Australia. In: Barker W. R. & Greenslade P. J. M. eds. Evolution of the Flora and Fauna of Arid Australia, pp. 35-45. Peacock Publications, Adelaide. WEBSTER R . L. 1 9 9 2 .

S. O. 1984. A Study of the Effects of the Development of Rainbow Creek on Land Use. BA (Hons) thesis, University of Melbourne, Melbourne (unpubl.). BRIZGA S. O. 1990. River Channel Changes in Gippsland, Victoria. PhD thesis, University of Melbourne, Melbourne (unpubl.). ERSKINE W . D., RUTHERFURD I . D. & TILLEARD J . W . 1 9 9 0 . Fluvial Geomorphology of Tributaries to the Gippsland Lakes. Ian Drummond and Associates P/L for Department of Conservation and Environment, Victoria. FREEZE R. A. & CHERRY J. A. 1979. Groundwater, Prentice-Hall, New York. BRIZGA

GRAYSON R . B . , KENYON C . , FINLAYSON B . L. & GIPPELL C . J.

1998.

Bathymetric and core analysis of the Latrobe River delta to assist in catchment management. Journal of Environmental Management 52,

361-372.

F. 1 9 7 6 . Groundwater Resources of the Gippsland Lakes Catchment. Ministry for Conservation, Victoria, Environmental Studies Series 113. HILL S. M . & BOWLER J . M . 1 9 9 5 . Linear dunes at Wilsons Promontory and South-East Gippsland Victoria: relict landforms from periods of past aridity, Proceedings of the Royal Society of Victoria 107,

HARRIS I .

73-81.

HINWOOD J. B. & MCLEAN

E. J. 1992. Development of a Parametric Estuarine Model for Gippsland Resource Planning. In: Barton C. M., Hill K., Abele C., Foster J. & Kempton N. eds. Energy, Economics and Environment: Gippsland Basin Symposium, pp. 249-256. Australasian Institute of Mining and Metallurgy, Melbourne. HODGSON L. 1995. Limited data, limited time: the use of groundwater models in salinity management. PhD thesis, University of Melbourne, Melbourne (unpubl.). JENKIN J. J. 1968. The geomorphology and Upper Cainozoic geology of south-east Gippsland, Victoria. Geological Survey of Victoria Memoir 27. JENKIN J . J. 1973. Late Cainozoic geology and geomorphology of south-east Gippsland In: McAndrew J. & Marsden M. A. H. eds. Regional Guide to Victorian Geology, pp. 145-157. School of Geology, University of Melbourne, Publication 1. JENKIN J . J . 1981. Evolution of the Victorian Coastline. Proceedings of the Royal Society of Victoria 92, 37-54. LAKE WELLINGTON CATCHMENT COMMUNITY WORKING GROUP

1993.

ROSENGREN N . ] . , MCCRAE WILLIAMS M . S . & KRAEMERS S . M .

1991.

Dropping the watertable: Lake Wellington Irrigation and Dryland Salinity Management Plan. Lake Wellington Catchment Salinity Management Plan Community Working Group, Victoria. LAWRENCE C. R . 1 9 9 2 . Hydrologic and Environmental Changes in the Gippsland Basin. In: Barton C. M., Hill K., Abele C., Foster J. & Kempton N. eds. Energy, Economics and Environment: Gippsland Basin Symposium, pp. 2 4 3 - 2 4 7 . Australasian Institute of Mining and Metallurgy, Melbourne. LEONARD J. 1988. Groundwater. In: Douglas J. G. & Ferguson J. A. eds. Geology of Victoria, pp. 547-557. Geological Society of Australia, Victorian Division, Melbourne. MOLLICA F. J. 1991. Nambrok-Denison groundwater pumping/monitoring review: central Gippsland, Victoria. Rural Water Commission of Victoria, Investigations Branch Report 1991/16. PARLIAMENTARY PUBLIC WORKS COMMITTEE 1 9 5 2 . Report on the Effects of Salinity in the Gippsland Lakes, Victorian Government Printer, Melbourne. Sites of geological and geomorphological significance in central Gippsland. Ministry for Conservation, Victoria. Environmental Studies Series 341. SKENE J . K . M . 1 9 5 2 . Report on soil investigations in the Gippsland Lakes area. Department of Agriculture Victoria, Soil Survey Report 14. SMITH G. C. 1988. Oil and gas. In: Douglas J. G. & Ferguson J. A. eds. Geology of Victoria, pp. 514-531. Geological Society of Australia, Victorian Division, Melbourne. SUMMERFIELD M. A. 1991. Global Geomorphology: An introduction to the study of landforms Longman, London. THOMPSON B. R. 1973. The geology and hydrogeology of the Mitchell River flats and a study of artificial recharge. Mines Department, Victoria Report (unpubl.). TOURISM VICTORIA 1999. Lakes and Wilderness: Regional Tourism Development Plan Summary. Government of Victoria, Melbourne. WALKER G . 1 9 9 2 . Effect of Petroleum Production on onshore ground-

Environmental geoscience, Gippsland Basin water aquifers in the Gippsland Basin. In: Barton C. M., Hill K., Abele C., Foster J. & Kempton N. eds. Energy, Economics and Environment: Gippsland Basin Symposium, pp. 235-242. Australasian Institute of Mining and Metallurgy, Melbourne. WALKER G. & MOLLICA F. J . 1 9 9 0 . Review of the groundwater resources of the south east region. Department of Water Resources Victoria.,

169

Water Resource Management Report 54. WARD W. T. 1977. Geomorphology and soils of the Stratford-Bairnsdale area, east Gippsland, Victoria. CSIRO Soils and Land Series 57. Received 3 August 1999; accepted 25 September 2000

Geological Society of Australia Special Publication 21, 171 -178

CHAPTER 15—Groundwater for Aboriginal communities in central Australia: the Western Water Study (Wiluraratja Kapi)*, Northern Territory G. JACOBSON' AND J. WISCHUSEN2

IRKilU .

.• •

% •_.'«....'

mmStSmiffP

W^Mmrnj^mmm

1 6 Hill

Corner, Yarralumla, ACT 2600, Australia. 2 50 Kingsclear Road, Alexandria, NSW 2015, Australia. Aboriginal communities in an arid region of central Australia have been consulted with a view to involving them fully in water-management issues. These consultations have developed a bicultural model incorporating the extensive Aboriginal knowledge of water. It is likely that the most effective programs for future development of water supplies in such groundwater-dependent regions will be based on this model or similar models. At present however, there is lack of organisation and representation of the communities with respect to water issues. In this region, the southwest Northern Territory, a geological information base has been developed to improve groundwater prospectivity. The integration of hydrogeological data into a regional conceptual groundwater model has also enhanced water prospectivity and enabled the analysis of regional water issues. The project GIS has been demonstrated in the field as a tool for groundwater assessment. Preliminary results of isotopic 'dating' of groundwaters indicate that the long-term sustainability of some of the community water supplies in this region is questionable. Analysis of regional groundwater quality data indicates that significant numbers of people are drinking water of marginal or poor quality: either salinity is above the Australian guideline values, or the water contains potentially harmful components such as fluoride, nitrate, uranium or boron. The methodology used for the study is transferable to other inland regions in Australia and to other arid and semiarid countries where groundwater resources need to be assessed and developed. KEY WORDS: Aboriginal communities, groundwater, Northern Territory, water supply.

INTRODUCTION In central Australia, the movement of Aboriginal people back to their ancestral lands, the homelands or outstation movement, has been supported by drilling for groundwater (Knott & MacDonald 1983). Here, and elsewhere in Australia, the availability of clean, dependable supplies of water is considered to be a vital step towards the improvement of health for Aboriginal communities (Hearn et al 1993; Federal Race Discrimination Commissioner 1994). The Western Water Study (Wiluraratja Kapi in the regional language Pintubi) is a groundwater-assessment program covering 68 000 km 2 of mainly Aboriginal land in the southwest of the Northern Territory (Figure 15.1). The study area comprises four 1:250 000 map sheets: Mt Doreen, Lake Mackay, Mt Liebig and Mt Rennie. This arid region (annual rainfall 250 mm, evaporation 3000 mm) has several major Aboriginal communities, numerous Aboriginal outstations and two pastoral homesteads. All these settlements rely on groundwater supplies. Previous groundwater-assessment studies have been undertaken at local sites such as community borefields or individual livestock or outstation water bores; this is the first regional assessment. A portion of this work has been funded by the Aboriginal and Torres Strait Islander Commission (ATSIC) on the understanding that this work will serve as a pilot project for further groundwater studies on Aboriginal and

Torres Strait Islander lands. The particular region (southwest Northern Territory) was selected for the pilot study because of perceptions of poor quality water by a high proportion of communities (ATSIC 1992). This collaborative work was undertaken between the Australian Geological Survey Organisation (AGSO), the Northern Territory Department of Lands, Planning and Environment (DLPE), the Central Land Council (CLC) and the Aboriginal communities in the region.

AIMS AND OBJECTIVES The Western Water Study aims to provide the necessary information for the assessment and management of groundwater resources beneath Aboriginal lands. Broader objectives are: to enhance environmental health; to ensure equity in access to acceptable safe water, especially in remote and arid areas; and to develop a rapid methodology that will provide these objectives. The main objective of this work is to improve access to groundwater information, currently held by various agencies, for Aboriginal people on their land. Other interested parties such as government planning agencies are also expected to benefit. * in the regional language, Pintubi

172

G. Jacobson and J. Wischusen

METHODOLOGY AND RESULTS Liaison and communication An initial round of consultations with Aboriginal communities in the region revealed serious concerns about water quality and about the provision of emergency roadside bores for travellers. Later, more extensive consultations with Aboriginal people on water issues were conducted for this project (Toyne et al 1997). These consultations showed that the most effective programs for development of water supplies will be based on the extensive Aboriginal knowledge of water. Such a bicultural approach will give a stronger emphasis to water use throughout Aboriginal lands rather than in population centres alone. It should lead naturally to a regional water-management structure. The Western Water Study has developed a traditional model for a prospective regional water program in central Australia that includes all activities connected with the supply and handling of water in both traditional and contemporary contexts, and all domestic, public and enterprise activities. This model is illustrated in Figure 15.2. The model has so far proved to be an effective vehicle for discussions of regional and community water issues. This approach to the consideration of water issues has resulted in the presentation of a number of specific proposals from Aboriginal people and their advisors for waterrelated developments in their communities and lands. These include new community infrastructure developments such as wood lots and swimming pools, strong calls for the provision of survival water supplies along isolated roads, and enterprise proposals in the context of both the community and the Aboriginal lands. This bicultural approach to water issues could be taken further through the setting up of regional water-management groups with predominantly Aboriginal membership, where Aboriginal knowledge and priorities can be applied to a regional water program with advice from appropriate non-Aboriginal people. The recommendations by Toyne et al (1997) provide ways in which the approaches to the handling of water issues that emerged from the consultations, can be facilitated in the study area (Appendix 15.1). They also provide possible pathways to the resources and support that will be needed to implement the various specific proposals.

Figure 15.1 Location map, Western Water Study, southwest Northern Territory.

Regional Water Group C (((©)))2>

ii)

Aboriginal Water Coordinator

iii)

ABORIGINAL KNOWLEDGE OF WATER Pina Ngapakurlu Yapakurlangu

(fi V

Technology / Training Coordinator iv)

J&fiQL jr£cu JsSfl,

Rock holes Soakages Water holes Springs

vi)

Aboriginal training and employment

Geology as a basis for groundwater assessment The existing regional geological maps of the southwestern Northern Territory were revised for this project by Lau (1997a), and included an assessment of previously unmapped Cenozoic deposits (Lau 1997b). New insights into regional geology and geomorphology are also apparent from an interpretation of multilayer digital imagery of the region (Woodcock et al 1997). Enhanced Landsat Thematic Mapper (TM) imagery combined with a Digital Elevation Model (DEM) were used to delineate an extensive network of palaeodrainage channels that probably dates back to the Mesozoic or earlier. This palaeodrainage system contains widespread calcrete deposits that contain important potential groundwater resources, and which overlie thick sand and gravel aquifers. Previously unrecognised alluvial-fan deposits were delineated with Airborne

NON-ABORIGINAL KNOWLEDGE OF WATER Pina Ngapakurlu Kardiyakurlangu

CS(|p3 Survival water in the country

Community water supplies

Water-related enterprises

i»mn«

Figure 15.2 Bicultural model for community water supplies (after Toyne et al 1997).

Gamma-Ray Spectrometry (AGS) overlain on the DEM, and these deposits are also of considerable hydrogeological significance. An extensive area of ferricrete east of Lake Mackay was identified by a combination of AGS and Landsat TM imagery; the groundwater potential of this area has not yet been evaluated. The integrated remote-sensing study is considered to be a cost-effective method of map-

Groundwater for Aboriginal communities, NT ping regolith materials and hydrogeological features rapidly over large, relatively unknown and inaccessible areas, such as the Western Water Study area. The study area covers three distinct geological provinces: the Proterozoic Arunta Complex; the Proterozoic-Palaeozoic intracratonic Amadeus and Ngalia Basins; and overlying Cenozoic sediments. Significant new information relates to the depth of Cenozoic cover and the age of the Cenozoic sediments (Lau 1997b), which were evaluated by correlation of existing drillhole information, regional geophysical surveys, and a groundwater drilling program. In the drilling program 20 additional investigation bores were drilled to depths of 200 m. Significant additional groundwater resources were also identified by this program. In order to further enhance the regional geological background, palynological determinations were undertaken on available drillcore and cuttings, including some from the special project drilling. This has led to a provisional stratigraphic framework for Tertiary basins in the southern half of the Northern Territory (Macphail 1996, 1997), including the project study area. The accumulation of organic-rich sediments in these basins has been continuous throughout Tertiary time. Three broad phases of deposition have been identified: during the Oligo-Miocene; during the Middle to Late Eocene; and during the Early Eocene. However, preservation of these organic facies has been haphazard within individual basins: organic facies of Paleocene age have not been identified and may not have been preserved. A significant finding is the presence of probable OligoMiocene spore-pollen in a water bore sited in a palaeodrainage system near Kintore on the Northern Territory-Western Australia border. Changes in the relative abundance of dryland pollen taxa may enable the correlation of organic units between the basins.

Groundwater prospectivity and sustainability The hydrogeology of the region was interpreted by JW (Figure 15.3). Groundwater flow directions were interpreted from pressure head measurements in water bores and from the elevations of known discharge zones. The regional groundwater system discharges to a linear system of playa lakes that includes Lake Bennett and Lake Mackay. The hydrogeology has been assessed and mapped at 1:500 000 scale using a GIS to interpret various spatial datasets and statistics of water-bore information (Wischusen 1998). Based on analyses of 850 water bores, the study area is divided into seven different aquifer systems to account for regional variations in hydrogeology. Palaeodrainage channels act as sinks along which groundwater drains to internal discharge playas. Statistical data for these aquifer systems and the individual geological formations drilled provide a starting point for further groundwater assessment in other geologically similar areas of central Australia. Only the Cenozoic fan sediments along a mountain range consistently provide low salinity (<1000 mg/L TDS) potable groundwater. Other aquifer systems have generally more saline groundwater. High individual bore yields of over 20 Us can be obtained in the Palaeozoic and deeper Cenozoic aquifers, but a relative lack of permeability storage precludes sustained high yield production from bores in the fractured Proterozoic rocks.

173

Modern groundwater recharge is known to occur in the vicinity of the Proterozoic basement ranges and near the margins of Cenozoic sediment cover (Wischusen 1998), but has not yet been observed elsewhere in the region. Stableisotope data for groundwaters from the Palaeozoic basins and Cenozoic aquifers show an evaporation effect, which may reflect a residual signal from a time when the climate was wetter and a direct recharge mechanism operated (i.e. an interglacial, pluvial period). Under such conditions, saline and evaporated vadose water stored during arid times may have entered the sedimentary aquifers. Thus groundwater in these systems may be considered to be a mix of waters recharged by various mechanisms at different times. The lack of observed recharge and the demonstrated slow groundwater movement within the Palaeozoic aquifers may mean that sustainable extraction is not possible, and that the groundwater resource is in effect being 'mined'. As hydrogeological and climatic data are limited, detailed studies of groundwater flow and the magnitude and frequency of recharge should accompany any major groundwater development in this arid area. Identification of fracture zones where increased permeability is likely to enhance hydraulic conductivity is important for groundwater assessment in this region (English 1997) and a lineament study was undertaken to enhance groundwater prospectivity of the fractured rock areas that underlie half of the study area. The permeability and groundwater storage capacity of these rocks is dependent upon their interconnected networks of fractures and fissures. In the sedimentary rocks of the intracratonic Amadeus and Ngalia Basins, fracturing produced by tectonic stress is as important to groundwater yields as primary porosity and permeability characteristics. A geological lineament analysis of the study area was carried out to aid the identification of potential aquifers and locate drill targets. Several thousand lineaments were interpreted from processed Landsat TM scenes. Band ratios and unmixed mineral processing were applied, along with interactive image enhancement of the processed imagery (Bierwirth 1990; English 1997). A significant proportion of the interpreted lineaments are in Cenozoic 'cover' material, their visibility attributable in part to minor Tertiary to Holocene reactivation of basement structures and partly to the processing applied to the imagery. The lineament array has been synthesised into regionalscale lineament zones, some of which represent previously unmapped major crustal structures (English 1997). The geometry of the lineament zones provides information about likely causative stress fields and the resultant deformation. This information has been used to predict areas of increased fluid flow. In this region, north-south-striking lineament zones and intersections of conjugate northwestand northeast-striking lineaments are hydrogeologically important. Favourable drill targets may include long fractures, and dense, intersecting networks of fractures. Relationships between lineaments, lithologies, topography and local water-table depths need to be accommodated in drill target selection. The sustainability of community water supplies drawn from shallow aquifers in the region has been evaluated using the radioisotope chlorine-36 (Cresswell et al 1999). Modern recharge in these central Australian groundwaters

174

G. Jacobson and J. Wischusen

129°

Ngalia Basin aquifers E'-TL-'d (fractured sandstone)

Figure 15.3 Regional hydrogeology and groundwater flow system (after Wischusen 1998).

is represented by a chlorine-36 ratio (36C1/C1) of 205 (±7) x 10~15 and this possibly represents dilution of the modern rainfall input ratio, 325 x 10~15 (Keywood 1995), by windblown salt from playa lakes. The two main sets of 36C1/C1 ratios for regional groundwaters are: -205 x 10~ 15 corresponding to groundwater <36C1 ages' of 10 ka or less (Holocene); and - 1 7 0 x 10~ 15 , corresponding to 36C1 ages of about 80 ka (last Interglacial). These are believed to be minimum ages or residence times for most of the shallow groundwaters in this region. Even lower values of the 36C1/C1 ratio have been measured in some groundwaters near playa lakes (Figure 15.3). These ratios may be affected by addition of remobilised salt from the playas, or they may represent older waters in slow-moving groundwater flow systems. Carbon-14 data for shallow groundwaters in this region are generally at variance with the 36C1 results, and the 14 C activity may be affected by diffusion. The 36C1 isotopic results have serious implications for groundwater management in this region. It seems that substantial recharge occurs only during favourable interglacial climatic regimes, and that most community water supplies depend on 'old', stored groundwaters.

Groundwater quality and health The project GIS includes groundwater quality information for 865 water bores in the Western Water Study region. About two-thirds of these bore waters are saline, beyond acceptable limits for drinking water according to the Australian Drinking Water Guidelines (1996). About onequarter of the bore waters have unacceptably high fluoride or nitrate concentrations, and some have high uranium or boron concentrations (Hostetler et al 1998) (Table 15.1). The generally high salinity and harmful element concentrations in these shallow regional aquifers are related to limited recharge in the modern climatic regime, high evaporation during the recharge process, and long residence times of groundwater in the flow system. The exceedance of the guideline value for sodium concentrations in 41 out of 120 bores used for drinking water (Table 15.1) may be a health concern because of the link between excess sodium intake and cardiovascular disease. The relative proportions of major ions, especially bicarbonate and chloride, reflect the position in the groundwater flow system. The bicarbonate-dominated groundwaters are noticeable in the recharge waters in the fractured basalt

Groundwater for Aboriginal communities, NT

175

Table 15,1 Bores exceeding Australian Drinking Water Guidelines (1996) for certain parameters, listed by aquifer. Total bores

Tertiary sediments 124 North Arunta Complex 88 South Arunta Complex 81 Ngalia Basin 111 Playa 1 Alluvial 52 Amadeus Basin 22 Total 479 Bores used for human consumption 120 ADWG, Australian Drinking Water Guidelines.

Bores within ADWG 6 1 12 6 0 9 5 39 26

41 44 27 16 0 5 1 134 31

63 33 17 43 1 1 5 163 17

Bores exceeding ADWG CI N0 Na 3

88 52 19 80 1 12 6 258 33

31 27 34 13 1 4 1 111 16

78 49 25 57 0 32 4 245 41

TDS

86 69 33 87 1 32 9 317 55

aquifer at Kintore (Wischusen 1994), and in the alluvialIt appears that for 20 years of the homelands/outstafan aquifer at Papunya. Other aquifers such as the Ngalia tions movement, Aboriginal people have needed to prove Basin and the fractured Arunta Complex rocks contain that 'potable' water is accessible at specific sites in order to groundwaters dominated by sulfate and chloride, reflecting reoccupy their ancestral lands. However, much of this longer residence times. water is marginal or unacceptable as drinking water under Nitrate concentrations beyond the drinking-water limit the increasingly stringent Australian Drinking Water of 50 mg/L are common in these shallow, mainly uncon- Guidelines (1996), although it may acceptable for hygiene fined aquifers (Table 15.1). In central Australia, nitrate-rich and other purposes. There is a case for the introduction of groundwaters were emplaced by episodic recharge events dual water systems in the larger communities like Kintore in a generally arid climatic regime. Nitrate has been flushed and Yuendumu which have populations of several hundred through the unsaturated zone which lacks denitrification people. Drinking water could be supplied by treatment, e.g. activity in this arid climate. The nitrate is formed by near- reverse osmosis, or by supplementary rainwater tanks. surface bacteria in termite mounds with the highest soil There is a need for water coolers to improve palatability of nitrate concentrations being found in the outer shells of ter- the drinking water in summer. There is a need for research mite mounds (Barnes et al 1992). Bacteria associated with and development into small-scale water-treatment techthe termites appear to fix nitrogen, which eventually nologies that are appropriate for outstations with populaappears in inorganic form principally as ammonia. Nitrate tions of tens of people. This would assist Aboriginal people is produced by bacterial oxidation of the ammonia, and is to increase their access to the land and to its available leached to the outside of the termite mound by capillary water resources. action. Recharge from extreme rainfall events then flushes this nitrate to the water-table. High fluoride concentrations (>1.5 mg/L) in these Groundwater information system regional groundwaters (Table 15.1) are particularly associ- Much of the data generated by this work has been held as ated with granites of the Arunta Complex (Hostetler et al paper records by the DLPE Water Resources Division. All 1998). High levels of fluoride in Tertiary sediments are relevant water bore data were entered into the DLPE dataprobably from those sediments that were derived from the base, HYDSIS, as part of this project. Incorporating the groundwater data into a Geographical Arunta Complex. Excess consumption of fluoride can induce fluorosis and affect dental health. High uranium Information System (GIS) was decided on at the outset to concentrations (>0.02 mg/L) in these regional groundwaters meet some of the project aims. Considerable effort has been are associated with sandstone units of the Ngalia Basin required to get relevant data into a suitable digital format sequence, the Tertiary sediments, or with granites of the (Wischusen et al 1997) and to link datasets held by the Arunta Complex. High boron concentrations (>0.3 mg/L) collaborating agencies. The GIS package was set up in that have been noted in these regional groundwaters ArcView, a commercial GIS program that is used by the appear to be associated with the more saline waters espe- CLC, the DLPE and AGSO. The datasets for this package are written to a CD ROM in an ArcView project format that cially in granites of the Arunta Complex. Several thousand people in this region may be affected can be run directly from the disc. The package has layers of by poor drinking-water quality. There is a high incidence of processed remotely sensed data, geological data (including kidney and urinary tract stones in young Aboriginal chil- lineament interpretation maps), topographical features, dren. The formation of stones is due to the supersaturation and cadastral data. These datasets complement the of urine with certain salts, and dehydration is believed to be groundwater database compiled from Northern Territory a major factor. In this region, Williams et al (1996) have water-bore record files. Additional work was undertaken by a consultant associated dehydration with the unpalatability of the available drinking water, especially under hot summer condi- (Gallagher 1996, 1998) to develop a method for visualising tions. More specific health effects from high fluoride, complex water-bore data lising a desktop GIS. The data can nitrate, etc., are not yet documented in the region, and may be stored in any tabular' database structure, and are displayed and analysed using new software modules added to be masked by a spectrum of other health problems.

176

G. Jacobson and J. Wischusen

ArcView. The borehole data are in a series of tables linked to borehole locations, in the traditional 'one-to-many' relationship. The model can be used in ArcView with some analysis tools implemented in Avenue script. A feedback session was undertaken for the Aboriginal communities in the region, and bore-siting and water-quality problems were addressed using the project GIS in regional council offices. The analytical capability of the project GIS was also tested using a regional issue of great concern to the people, the provision of emergency water bores for travellers along lonely roads. On a map of the region, this analysis superimposed a 2 0 km radius buffer zone around the existing bores to indicate 'safe' zones, and highlighted areas where bores need to be constructed, or where existing bores need to be equipped with hand pumps. The benefits of an integrative approach to groundwaterresources assessment are believed to be considerable. At a cost of about $1.2 million the study has provided a decision-support system for future water supplies in an arid region where water supplies are always going to be difficult to establish. This system is now available to facilitate planning, to support future drilling programs in the region, and to increase the success rate of both. As the water bore and tanks for one small outstation generally cost about $0.3 million, the future cost savings to the national infrastructure program are likely to be substantial. The methodology used for the study is readily transferable to other inland regions in Australia and to other arid and semi-arid countries where groundwater resources need to be assessed and developed.

CONCLUSIONS (1) Extensive consultation with Aboriginal communities in central Australia has developed a bicultural model that incorporates the extensive Aboriginal knowledge of water. It is likely that the most effective programs for development of water supplies will be based on this model or on similar models. (2) Recommendations have been made for specific actions to help the communities with better organisation and representation with respect to water issues (Appendix 15.1). (3) Geology is the basis of groundwater assessment, and an enhanced geological information base for this region has been developed to improve assessment of groundwater prospectivity. Updated geological maps, investigative drilling, and analysis of remotely sensed data have formed the basis for this. (4) Integration of hydrogeological data into a regional conceptual groundwater model has enabled the analysis of serious regional water issues. The project GIS has been demonstrated in remote communities as a tool for groundwater-resources assessment. (5) Preliminary results of radioisotopic 'dating' of regional groundwaters indicate that modern recharge is limited and that the long-term sustainability of some of the community groundwater supplies in this regipn.is questionable. (6) Evaluation of regional groundwater quality data indicates that significant numbers of people are drinking water of unacceptable salinity, or that contains deleterious

elements such as fluoride, nitrate, uranium and boron exceeding the Australian Drinking Water Guideline concentrations. (7) The integrative approach to groundwater-resources assessment is expected to result in substantial cost savings to the national infrastructure program. The methodology used for the study is transferable to other inland regions in Australia and to other arid and semiarid countries where groundwater resources need to be assessed and developed.

ACKNOWLEDGEMENTS The work was carried out while the authors worked for the Australian Geological Survey Organisation and was partly funded by ATSIC and by the Northern Territory Government. W e thank Graham Henderson, Toly Sawenko, Bruce Rose and Peter Jolly for their support of the project. W e thank our colleagues who contributed to the work: Tony Meixner, Tim Mackie, Tom Calvert and Phil Bierwirth for assistance with geophysical processing; Michael Jamieson, Libbie Lau, Lynne Woodcock, Stephen Hostetler and Sonya Lenz for assistance with GIS database development and the preparation of maps. We greatly appreciate the dedicated and expert consultancies from Peter Toyne and his associates, and from Michael Macphail, Pauline English, and Robyn Gallagher. We thank Doug Mackenzie for comments on the manuscript.

REFERENCES ATSIC 1992. National Housing and Community Infrastructure Needs Survey, preliminary report, stage 1, June 1992. Australian Construction Services, Brisbane. AUSTRALIAN DRINKING WATER GUIDELINES 1 9 9 6 . National Health and Medical Research Council & Agriculture and Resource Management Council of Australia and New Zealand. BARNES C . J . , JACOBSON G . & SMITH G . D . 1 9 9 2 . The origin of highnitrate groundwaters in the Australian arid zone. Journal of Hydrology 137, 1 8 1 - 1 9 7 . BIERWIRTH P. N. 1990. Mineral mapping and vegetation removal via data-calibrated pixel unmixing, using multispectral images. International Journal of Remote Sensing 11, 1999-2017. CRESSWELL

R.,

WISCHUSEN

J.,

JACOBSON

FIFIELD

1999.

Assessment of recharge to groundwater systems in the arid southwest Northern Territory, Australia, using the radioisotope Chlorine-36. Hydrogeology Journal 7, 3 9 3 - 4 0 4 . ENGLISH P. 1997. Lineament interpretation for groundwater assessment,Western Water Study (Wiluraratja Kapi), Northern Territory. Australian Geological Survey Organisation Record 1997/8. FEDERAL RACE DISCRIMINATION COMMISSIONER 1 9 9 4 . Water—a report on the provision of water and sanitation in remote Aboriginal and Torres Strait Islander communities. Australian Government Publishing Service, Canberra. GALLAGHER R. 1 9 9 6 . Visualising waterbore data with ARCVIEW2. Australian Geological Survey Organisation Record 1996/18. GALLAGHER R . 1 9 9 8 . A new G I S extension for visualising waterbore data and other tabular geological data. Australian Geological Survey Organisation Record 1998/12. HEARN W . , HENDERSON G . , HOUSTON S . , W A D E A . & WALKER B .

1993.

Water supply and Aboriginal and Torres Strait Islander health: an overview. AGSO Journal of Australian Geology & Geophysics 14, 135-146. HOSTETLER S . , W I S C H U S E N J . & JACOBSON G. 1 9 9 8 . Groundwater quality in the Papunya-Kintore Region, Northern Territory. Australian Geological Survey Organisation Record 1998/17. KEYWOOD M . D . 1 9 9 5 . Origins and sources of atmospheric precipitation in Australia: chlorine-36 and major-element chemistry. PhD

Groundwater for Aboriginal communities, NT thesis, Australian National University, Canberra (unpubl.). G. G. & M C D O N A L D P. S. 1983. Groundwater for Central Australian Aboriginal communities. In: Papers of the International Conference on Groundwater and Man, Sydney, pp. 141-150. Australian Water Resources Council, Conference Series 8. LAU J. E. 1997a. Geology of the Papunya-Yuendumu-Kintore region, Northern Territory, Western Water Study (Wiluraratja Kapi), 1:500 000 map. Australian Geological Survey Organisation, Canberra. LAU J. E. 1997b. Cainozoic Geology of the Papunya-Yuendumu-Kintore region, Northern Territory, Western Water Study (Wiluraratja Kapi), 1: 500 000 map. Australian Geological Survey Organisation, Canberra. MACPHAIL M . K. 1996. A provisional palynostratigraphic framework for Tertiary organic facies in the Burt Plain, Hale, Ngalia, Santa Teresa, Ti-tree & Waite Basins, Northern Territory. Australian Geological Survey Organisation Record 1996/58. MACPHAIL M . K. 1997. Palynostratigraphy of Late Cretaceous-Tertiary basins in the Alice Springs District, Northern Territory. Australian Geological Survey Organisation Record 1997/31. TOYNE P., GRANITES R . J . & ZIMRAN S. 1 9 9 7 . Towards a bicultural regional water program in central Australia—Consultations with Aboriginal community groups. Australian Geological Survey

KNOTT

177

Organisation, Canberra. WILLIAMS W . M . , NICHOLAS J . J . , NUNGURRAYI P. B . & NAPURRULA C . R .

1996. Paediatric urolithiasis in a remote Australian Aboriginal community. Journal of Paediatrics and Child Health 32, 344-346. J. D. H. 1994. Sustainability of a hard rock aquifer at Kintore, Gibson Desert, Central Australia. International Hydrology and Water Resources Symposium, Adelaide, Preprints, 343-349. WISCHUSEN J. 1998. Hydrogeology of the Yuendumu-Papunya-Kintore region, Northern Territory. Australian Geological Survey Organisation Record 1998/31. W I S C H U S E N J . , JAMIESON M . , ROSE B . ETAL. 1 9 9 7 . The Western Water Study GIS, an aid to groundwater assessment and prospecting on aboriginal land in central Australia. In: Proceedings of 3rd National Forum on GIS in the Geosciences, pp. 8 4 - 8 9 . Australian Geological Survey Organisation Record 1997/36. WOODCOCK L . G., BIERWIRTH P. N. & LAU J. E. 1997. An integrated remote sensing study for the Papunya-Kintore region, Northern Territory. Australian Geological Survey Organisation Record 1997/45. WISCHUSEN

Received 29 May 1998; accepted 7 May 2000

178

G. Jacobson and J. Wischusen

APPENDIX 15.1. WESTERN WATER STUDY: RECOMMENDATIONS FROM ABORIGINAL COMMUNITY CONSULTATIONS (AFTER TOYNE ET AL 1997) 1. Regional water programs It is recommended that regional water programs be developed and resourced in the Western Water Study area (Papunya-Yuendumu-Kintore, Northern Territory: Figure 15.1), under the auspices of regional water-management groups with majority membership and overall control vested with the Aboriginal people in the region. The aim of this initiative is to empower Aboriginal people, working through groups of their own choosing, in the handling of contemporary water issues. (la) It is recommended that the proposed regional water-management program adopt a holistic and bicultural approach as outlined in the model presented in this report (Figure 15.2). Such an approach involves the use of both Aboriginal and non-Aboriginal knowledge of water resources, and has a strong emphasis on the provision and use of water supplies in the Aboriginal lands as a whole, not just in the population centres. (lb) It is recommended that the proposed regional water management group be established, in the western section of the study area, through an extension of the functions of the Ananguku Yiwarra Aboriginal Corporation (AYAC) to include responsibility for a regional water program as well as their existing regional roads program. In the northeast section of the study area the water-management role could come under a Warlpiri regional group, whose formation has recently been proposed. 2. Water on Aboriginal land It is recommended that a systematic evaluation be undertaken of the potential for contemporary use of traditional water sources on the Aboriginal lands in the study area by suitable working parties of traditional owners, and that this be resourced appropriately. Such an evaluation would include: (i) recommendations as to which traditional sources are open to public use, rather than being restricted under traditional beliefs; (ii) recommendations as to how traditional ownership could be related to a program of contemporary wardenship which provides for the protection and upkeep of open water sources which are deemed to have continuing usefulness; (iii) recommendations as to the specific role which traditional water sources could play in providing survival water supplies along isolated roads, supporting small population centres, and supporting the development of enterprises; (iv) assessment of the relationship of traditional sources to other, technology based water supplies within the Aboriginal lands; and (v) advice from hydrologists as to the quality and supply of available water in traditional water sources in cases where they might be used for the regular supply of small population centres. 3. Community water supplies It is recommended that initiatives be undertaken aimed at: training Aboriginal people for work in community-based essential services; public education to improve the practices of community members as regards water use; and

public awareness programs on a regional, as well as community basis. This broader base would allow the establishment of specialised, local educational staff and resources, and provide continuity in the face of frequent changes in community essential services personnel (3a) It is recommended that the regional water management group provide an additional forum for the examination of essential service issues in Aboriginal communities, including the technological options for water supply and disposal, arrangements with training institutions for training programs, and social arrangements in support of employment. 4. Water-related enterprises (4a) It is recommended that the establishment of a jointventure plumbing enterprise be explored, which is capable of tendering for installation work in the new housing programs that are underway in the region, as well as for plumbing repair and maintenance. This possibility could be pursued through negotiations between AYAC and the national plumbers union (CEPU) as a means of creating local Aboriginal employment and participation, based on the substantial requirements for plumbing connected to the new and existing community housing stocks. Such a development could also be assisted by putting in place a DEET-funded traineeship or new work opportunity program, with CEPU assisting with the training framework, the selection of suitable personnel, and the development of a business plan and operational structure. Such a program would need to be suitably articulated with follow-up apprenticeship arrangements for the Aboriginal trainees. (4b) It is recommended that a business plan for a mixed, country-based enterprise be developed as part of a bush tucker study to be conducted by the Central Land Council. The enterprise would be based on the potential income which is available from camel mustering, ecotourism and the harvesting of bush tucker, and would be attached to the Ngurra Waltja outstation group. 5. AYAC resources It is recommended that a combination of research and DEET funding be used to establish two staff positions in AYAC with responsibility for water issues. It would be expected that the two positions would be supported eventually under the commercial plumbing development. An anangu program coordinator would be responsible for facilitating the input of Aboriginal and community knowledge and concepts into the water program on a regional basis. A technology and training coordinator would mediate the use of non-Aboriginal technologies and concepts and provide a suitable training framework. 6. A Warlpiri water group It is recommended that the establishment of a similar water program be incorporated into current attempts to establish a regional council (the Warnayaka Council) in the Warlpiri area. Such an initiative could mirror arrangements within the development of AYAC where this is seen as appropriate by the Warlpiri people. It could also be established by creating a regional water-management element in the existing housing and construction programs in the Warlpiri communities.

THEME 5 COASTAL AND NEARSHORE ENVIRONMENTS

Geological Society of Australia Special Publication 21, 181-199

CHAPTER 16—Human impacts on geoheritage features of the Swan Coastal Plain and coastal zone, southwestern Australia V. SEMENIUK AND C. A. SEMENIUK V&C Semeniuk Research Group, 21 Glenmere Road Warwick, WA 6024, Australia. In southwestern Australia a rapidly expanding human population has impacted on the natural history and the geoheritage of the Swan Coastal Plain and its coastal zone. To date, there have been losses of significant sites of geoheritage through re-landscaping of natural landforms, reclamation, urbanisation, alteration of soils, diverting of river channels, vegetation clearing, nutrient enrichment of waterways and groundwater, and groundwater abstraction, amongst others. The reasons for these impacts are complex and varied. For instance, at one level, environmental managers, government agency decision makers and politicians, frequently have not fully appreciated the significance of geologic processes and products, or the significance of geoheritage. At another level, there is a perception that within a technological culture, science, technology and engineering can fix all, without appreciating that systems can be irrevocably destroyed; and at a third level, decision makers are anthropocentric, placing development and wealth above all natural features. Many geoheritage features on the Swan Coastal Plain are of global to national significance. Examples include: (i) the suite of coastal forms showing the connection between tectonic setting, coastal processes and sedimentation in a basin-wide context, exemplified by the interrelationship of cuspate forelands along the northern coast and the dune barrier along the southern coast; (ii) the diversity and array of wetlands, illustrating a variety of geomorphology, hydrologic mechanisms and origins; and (iii) the occurrence and internal structures of star dunes, with their buried soils, cross-stratification and unconformities providing insight into Quaternary glacial/interglacial desert history and evolution of dune forms. This chapter provides an overview of the geoheritage features on the Swan Coastal Plain in order to alert the scientific community and land-use managers and decision makers to the important features therein and some of the developmental processes that have either degraded or destroyed them. An understanding of the significance of features of geoheritage in this region, and the history of their degradation or destruction will ensure that past errors in judgement may be averted in the future. KEY WORDS: anthropogenic factor, conservation, dunes, environmental geology, geoheritage, Quaternary, Swan Coastal Plain, wetlands.

INTRODUCTION Since European settlement of the Swan Coastal Plain, commencing over 170 years ago, centred initially on Perth on the shore of the Swan River, southwestern Australia (Figure 16.1), there has been an inexorable population increase (-1.5 million), bringing with it various types of impacts on the environment (Bolton & Hutchison 1973; Jarvis 1986). Many of these impacts have reduced the floral and faunal biodiversity of the region, resulting in local extinctions, and have contributed to the pollution, contamination, or lowering of the natural quality of water bodies, atmosphere and soil. These impacts occur within a region of southwestern Australia considered to be botanically one of the richest in the world (Marchant 1973). Much of the historical environmental perturbations, however, took place within a context of ignorance, culturalcentricity, anthropocentricity, an ethos of 'taming and developing a wild land', and a presumption of human dominance over the landscape, prior to the general raising of

environmental consciousness witnessed over the past two decades. In those times, there was generally not the information to appreciate the significance of the natural history environment nor the type of understanding of environmental issues as held today. In more recent times, spanning several decades, there has been an enormous increase in knowledge about the natural history of the Swan Coastal Plain, a better appreciation of its significance, and an awakening of an environmental consciousness to the extent that conservation of the natural environment is slowly becoming a more prominent part of the cultural ethos. However, population expansion and its concomitant exploitation of the natural environment continues (Ministry for Planning 1995a; Nagle 1998). Perth, for instance, is considered globally as a centre of hyperurbanisation. While acquisition of increased information on natural history is contributing both to more rational land-use planning, and to an informed public taking direct action on environmental issues, town planners, public service decision-makers, various government agencies and

182

V. Semeniuk and C. A. Semeniuk

LANCELIN

PERTH Study Area

|PERTH

ROTTNEST m ISLAND

KEY Swan Coastal Plain underlain by Quaternary sediments

MANDURAH

Scott Coastal Plain underlain by Quaternary sediments

= im=i ipii

Blackwood Plateau underlain by sand, laterite and Cretaceous sediments Dandaragan Plateau underlain by sand, laterite, and Cretaceous sediments

BUNBURY

Darling Plateau and Yilgarn Plateau/Plain underlain by sand, laterite, and Precambrian rock Leeuwin-Naturallste Ridge underlain by sand, laterite and Precambrian rock River

100 km 115 L_ Figure 16.1 Location of the Swan Coastal Plain in relationship to geological setting. Note that the northern third of the Swan Coastal Plain between Moore River and Dongara is not included in this map.

Human impact Swan Coastal Plain politicians appear still trapped in what may be termed the 'settler's ethos'. For example, between 1960 and 1980, studies of landform and vegetation on the Swan Coastal Plain generally took place within an agenda of land capability for silviculture, agriculture or urbanisation (Havel 1968; McArthur & Bettenay 1960; McArthur & Bartle 1980a, b). As recently as 1996, against a backdrop of global concern about increasing population and local concern that Perth has limited resources to sustain even its extant population without impacts on biodiversity, water resources and air quality, the Ministry for Planning (a government agency concerned with town planning) pursued a management philosophy that reflected a perception that there was a general lack of community support for any measures to restrict Perth's population growth (Ministry for Planning 1995a, b, 1996). Conservation is now an emerging paradigm with the establishment, for instance, of nature reserves at the local scale and the recognition of world-heritage sites at the global scale. These precepts have been applied to Western Australia, with the vesting of local Conservation Reserves, listing of sites on the Register of the National Estate by the Australian Heritage Commission, and recognition of some areas for World Heritage nomination (e.g. the Shark Bay region). To date, however, emphases on conservation have been in the sphere of biodiversity, with the conservation of natural history features associated with geology and geoprocesses not adequately addressed. In this context, this chapter attempts to deal with this problem by outlining some of the philosophy behind geoheritage, focusing on some of the features of the geoheritage of the Swan Coastal Plain, placing the significance of the region in a global, national to regional perspective, and documenting some of the destruction that has occurred over the past three decades.

CONCEPT OF GEOHERITAGE Conservation globally, to date, has been concerned with preservation of biota, in particular, rare and endangered species, and more recently, preservation of communities (Wyatt & Moss 1990; Blandin 1992; Withers & Horwitz 1996). Hence, the recognition of diversity (or biodiversity) as a basis to conservation (IUCN 1992; Ledec & Gooland 1988; Wyatt & Moss 1990) has become a relatively powerful focusing force in the conservation arena. Conservation, however, should be concerned with more than preservation of biodiversity, encompassing a range of natural history features. It should include: (i) purely biological phenomena of scientific and heritage value, from rare and endangered species, to representative communities (Soule & Wilcox, 1980), to biodiversity (McNeely et al 1990); (ii) features combining biota and geology, geomorphology, pedology and hydrology, essentially linking biodiversity with geoheritage (Hopkins 1994); and (iii) purely physical (i.e. non-biological) aspects of scientific and heritage value, such as unusual or representative rock and landscape formations, many of which may be subsumed under the term 'geoconservation' (Duff 1994; Creaser 1994; Markovics 1994). Where landscapes, geological formations or physical features are of outstanding importance, their conservation

183

is viewed as axiomatic. At the national level, examples are The Breadknife (The Warrambungals, New South Wales), Uluru (central Australia), the Twelve Apostles (coastal Victoria), or classic exposures of dinosaur footprints (Broome, Western Australia). Not so clear to the general public and decision-making non-scientists, is the case for geoheritage. Conservation of geoheritage sites generally has been made more acceptable where unusual geological features or a unique geological setting is coincident with or results in the occurrence of unusual or rare biota (Hopkins 1994) and caves as reservoirs of unusual or rare species (Osborne 1991; Humphrey 1993). The term geology, often used synonymously with earth sciences, is diverse and encompassing. This scientific discipline has related subsidiary disciplines of geomorphology, pedology, hydrology, and surface processes such as sedimentology. The latter have become more specific endeavours within the science. Additionally, the science of geology encompasses a diversity of interests. On the physical side, it includes at macro-regional scales, studies of global tectonics, mountain building and landscape evolution, at medium scales, studies of earth surface processes such as weathering, erosion and sedimentation, involving water, wind and ice, and at microscale, studies of crystal defects and deformation. On the chemical side, it involves studies of precipitation, cementation, solution and alteration at all scales. This broad definition and scope of geology becomes important when different types and expression of national or state-wide significant sites of geoheritage are discussed, and when the term geoheritage is used to encompass a wide variety of geological features. The term geology, in the context of geoheritage, encompasses, amongst others, the following: features of igneous, metamorphic and sedimentary rocks, and their relationships at all scales; features of craton/basin relationships; mineral locations; fossil locations; pollen locations; type stratigraphic locations; type igneous, metamorphic or pedogenic locations. It includes sites of importance in understanding geological processes; sites of importance for geomorphology, sedimentology, stratigraphy, pedology, and hydrology; as well as sites of profound aesthetic geological importance, or of intrinsic geological value. The term geoconservation is used here in the following sense: 'conservation, or preservation of earth science features (geologic, geomorphologic, pedologic, and hydrologic) of sufficient significance to warrant protection for purposes of heritage, science, or education'. The term geoheritage is used here as follows: 'global to national to state-wide important features of geology, including igneous, metamorphic, sedimentary, structural, geochemical, palaeontological, geomorphic, pedologic, and hydrologic attributes, that offer important information or insights into the formation or evolution of the continent; or that can be used for research, teaching, or reference sites'. From these perspectives, many features on the Swan Coastal Plain have geoheritage significance. They include textbook examples of geoprocesses, such as development of coastal cuspate forelands, occurrences of relict desertformed star dunes, and the geomorphology, stratigraphy, and hydrological functioning of wetland basins. Note should be taken that just as biologic systems are diverse, geological systems are also diverse. For instance,

184

V. Semeniuk and C. A. Semeniuk

Figure 16.3 Idealised block diagrams showing geomorphic units of the Swan Coastal Plain, their shore-parallel distribution, gross surface features, and stratigraphy, with details of landscape and vegetation habitats (from Semeniuk & Glassford 1989).

Human impact Swan Coastal Plain the terms wetland and dune carry implications of a large array of wetland types and dune forms. Thus, wetlands may be lakes, sumplands, damplands, paluslopes, palusmonts (Semeniuk & Semeniuk 1997), or in the terminology of the Ramsar Convention Bureau (1991), fens, marshes, ponds, oases, etc. Dunes may be linear, parabolic, hairpin, star or barchan, amongst others. Similar to the objective of nature conservation, to conserve the vast diversity of life forms, an objective of the conservation of sites of geoheritage value should be the conservation of the variety of geologic forms on the Earth. In this context, the conservation of a single lake as an example of a wetland system, or a single parabolic dune form in one area as representative of the full variety of dune types globally is insufficient. If wetlands, for instance, exhibit a large diversity of geometric and hydrologic types, stratigraphic fills and origins, then at the least their conservation should encompass an example of each of the types extant. Similarly, various types of dunes need to be conserved to capture the full range of shapes, types and origins that are possible as an example of the geoheritage of the Earth.

GLOBAL SETTING OF THE SWAN COASTAL PLAIN In the first instance, it is necessary to place the Swan Coastal Plain in a global to national perspective in order to appreciate some of the geological features occurring there and in this context the general pattern of coasts and nearcoastal plains are briefly reviewed. The case will be made that the Swan Coastal Plain, viewed globally, stands unique, whereas many of the coasts and coastal plains worldwide conform to recurring patterns of coastal and near-coastal landforms, with two or more examples representing the coastal form (Johnson 1919; Davies 1980; Bird & Schwartz 1985; Davis 1994). Coastal form and immediately inland coastal plains and uplands often can be related to their geologic and geomorphic setting within a global tectonic framework in terms of plate tectonics (Uyeda 1978). Examples are island-arc coasts (such as the Indonesian Archipelago), leading-plateedge coasts (the Andes-Rocky Mountains chain), trailingplate coasts (North American east coast), collision coasts (the northern Mediterranean coast). The Swan Coastal Plain, in this context, is a Quaternary surface developed on a subsiding linear basin developed along a passive-trailingplate margin. Globally, coastal forms, in generalised patterns, include: (i) erosional coasts cut into igneous, metamorphic, or sedimentary bedrock, or Quaternary aeolianites, resulting in cliffs and rocky shores, and where there is sediment supply, associated pocket beaches, local sandy bays, mud-lined embayments, spits, cuspate forelands, tombolos, and narrow alluvial plains, related to laterally equivalent rocky shores, such that cliff coasts grade to more diffuse coasts of alternating headlands and depositional pockets; (ii) ria coasts grading to archipelagos, where the post-glacial transgression has inundated fluvially dissected terrains, and within which sedimentation ranges from negligible to embayment-filling; (iii) inlet and estuary-dominated coasts, grading to deltaic coasts; (iv) low-gradient prograding coasts and their inland adjoining low coastal plains (these

185

coasts may be carbonate or terrigenous depositional systems); (v) sandy beach and dune coasts, often backed by prograded beach ridge plains; (vi) barrier island and barrier dune systems, and their adjoining leeward lagoons; and (vii) coral-reef coasts. The coastal zone of the Swan Coastal Plain varies from south to north, reflecting a climate gradient and the increasing effect of tectonic uplift. Coastal form changes from barriers-and-estuaries, to barrier dunes, to limestone rocky shores with intermittent cuspate forelands and pocket beaches (Searle & Semeniuk 1985). Western Australia exhibits a range of megascale basinand craton-related coastal and near-coastal landforms, crossing a variety of climate types (Woods 1980; Woods et a/., 1985; Semeniuk 1993, 1995a, b): the Kimberley ria/archipelago coast, the Canning Coast, the Pilbara Coast, the Carnarvon Coast, and the Perth Basin Coast. The latter coast is the context for the Swan Coastal Plain.

DESCRIPTION OF THE SWAN COASTAL PLAIN AND ITS COASTAL ZONE The Swan Coastal Plain and its coastal zone (Gentilli & Fairbridge 1952; Gentilli 1979; Searle & Semeniuk 1985) has a number of features that set it apart from the pattern of other coastal types in Western Australia and worldwide. (1) The Swan Coastal Plain, with elevations from near sea-level to 90 m above sea-level, is the Quaternary surface of the north-south linear Perth Basin (Figures 16.1, 16.2 on Plates 8 & 9, 16.3), a Phanerozoic basin that borders the linear north-oriented and extensive Darling Fault (Playford et al 1976; Geological Survey of Western Australia 1990). Sediments filling this basin range from Permian to Quaternary, although the Pliocene, Pleistocene and Holocene history has had the major role in determining the form of the coastal plain and coastal zone. (2) The coastal plain is bordered to the east by the retreating Darling Fault scarp, that defines the margin of an extensive plateau (Figures 16.2 on Plates 8 & 9, 16.3), situated some 300-600 m above sea-level. The northern part of the plain is bordered by the Dandaragan Plateau. Today the plain is located in a semi-arid climate, though at times during the Pleistocene it was in an arid climate (Glassford 1980; Glassford & Semeniuk 1995). (3) The Swan Coastal Plain Quaternary sediments form north-south-oriented belts parallel to the Darling Fault and to the coast. They are the product of the interplay of fluvial and alluvial-fan input, and alternating desert conditions and marine transgressions building a system of quartz-rich desert sands to the east and coastal limestone ridges to the west (Figure 16.3). (4) The onshore and nearshore coastal zone comprises shore-parallel limestone ridges forming ridge and wetland chains onshore, and chains of islands, rocky reefs and promontories in the marine nearshore. These aspects underpin why the coastal zone has developed its geomorphic character with the Swan Coastal Plain comprising generally north-south-oriented megascale geomorphic systems. In order from west to east, from marine environments to inland, these are: offshore chains of shoreparallel rocky reefs and islands, onshore limestone ridges, cuspate forelands, sandy beaches and dunes, beach ridge

186

V. Semeniuk and C. A. Semeniuk

Table 1 6 . 1 Summary of features of the various geomorphic units of the Swan Coastal Plain (all units are Quaternary in age). Geomorphic unit Pinjarra Plain: mainly eastern part of the Swan Coastal

Formation and description Guildford Formation: muddy and sandy deposits.

Plain, comprising alluvial fans and fluvial deposits. Bassendean Dunes: the eastern third of the Swan

Bassendean Sand: mainly quartz sand.

Coastal Plain, comprising low undulating sand hills, high sand hills, relict dune forms, and star dune. Spearwood Dunes: the western third of the Swan

Yellow sand overlying Tamala Limestone, i.e., quartz

Coastal Plain, in its central to northern parts, comprising

sand overlying calcarenite.

shore-parallel ridges. Mandurah-Eaton Ridge: long and moderately high

Yellow to white quartz sand and limestone lenses.

limestone and sand ridge in the southern part of the Swan Coastal Plain, in its southern part. Yalgorup Plain: the western third of the Swan Coastal

Yellow to white quartz sand overlying Tims Thicket

Plain, in its southern part, comprising limestone plains,

Limestone, Kooallup Limestone, and Myalup Sand,

former cuspate forelands, and shore-parallel ridges.

i.e., quartz sand overlying calcarenite.

Quindalup Dunes: coastal dunes of variable landforms.

Quartz/carbonate sand.

plains, and limestone rocky shores. Then there is a belt of limestone ridge and wetland chains, a belt of quartz sand hills and relict dunes, with associated wetland networks, and a zone of fluvial sediments and alluvial fans (Figures 16.2 on Plates 8 & 9, 16.3) (McArthur & Bettenay 1960; Searle & Semeniuk 1985; Semeniuk & Searle 1986a, b; Semeniuk 1985, 1997; Semeniuk et al 1988, 1989; C. A. Semeniuk 1988). The system of shore-parallel landform units have been formally named (Table 16.1) as Quindalup Dunes (Holocene coastal dunes), Spearwood Dunes (limestone ridge belt), Bassendean Dunes (quartz sand hills) and Pinjarra Plain (alluvial fan and fluvial terrain adjoining the scarp). The megascale landforms have developed soils and vegetation habitats specific to the geomorphic setting (Speck in Seddon 1972; Heddle et al 1980; Cresswell & Bridgewater 1985). For instance, the major soil associations overlie the following various fundamentally different parent materials (Figure 16.3): calcareous and quartz coastal dunes, limestone ridges, yellow quartz sand hills; white quartz sand hills, fluvial muddy sand and sand flats and plains; and ironstone and clay associations on slopes. At the local scale, these megascale settings for soil associations, in response to terrain aspect, slope, soil processes, depth to water table, and time, result in a plethora of small-scale mosaics of soils and vegetation habitats, each exemplifying a close landform-soil-vegetation relationship. The range of wetland types developed within a given geomorphic system and the vegetation habitats developed in coastal dune systems, as described by C A Semeniuk (1988) and Semeniuk et al (1989), well illustrate these patterns.

phic and stratigraphic setting. Over much of the coastal plain, the water-table is shallow, or perched, even if the terrain is relatively high and the groundwater mounded. Hence, there are numerous wetlands formed where watertables or perched groundwaters intersect or are close to the land surface. The variety of geomorphic/stratigraphic settings of these wetlands, and the range of different hydrologic settings in the various geomorphic/stratigraphic contexts has resulted in a wide range of wetland types, varying in size, shape, network patterns, hydrologic maintenance, stratigraphy, soils and origin (Figure 16.4). There are a variety of origins for the wetlands. Some are windows to the water-table formed as a rising post-glacial sea-level caused continental water-tables to rise, thus flooding ridgeand-basin systems, or basins within an undulating terrain (e.g. those in the Bassendean Dunes and some in the Spearwood Dunes). Some are perched, formed by sedimentary accumulation of muddy sediment which then acted as a perching layer (e.g. in Bassendean Dunes). Others are hydrologically coalesced basins (e.g. in Bassendean Dunes). Some are developed as karst solution features between two stratigraphic formations (e.g. contact between the Spearwood Dunes and the Bassendean Dunes), while others are expressions of karst within a formation (e.g. within the Spearwood Dunes). Some are deflation surfaces excavated by wind to the water-table (e.g. in the Quindalup Dunes) and some are interdune and beach ridge swales (also in the Quindalup Dunes). These wetlands, depending on setting, are filled with a variety of sediments.

Groundwaters and wetlands

Coastal zone and major coastal sectors

Unconfined to locally confined, mainly fresh groundwater underlies the coastal plain system, and its configuration, dynamics and hydrochemistry is determined by its geomor-

The coastal zone presents an interesting relationship between the existing oceanographic and coastal processes, megascale geologic setting, ancestral geomorphic systems,

Human impact Swan Coastal Plain

187

Figure 16.4 Idealised map showing the variety and distribution of the consanguineous wetlands, i.e. the natural wetland groupings or suites in the Darling Plateau and Swan Coastal Plain region centred on the Moore River to Bunbury region. In this region there are 42 such groups each related to geomorphic setting or interfaces between geomorphic units. Details of the suites are presented in C. A. Semeniuk 1988.

the source and availability of sediment, their sites of accumulation, and the resulting coastal and marine geomorphology. Swell from the Indian Ocean and that refracted from the Southern Ocean impinge on the shore from the southwest (Silvester 1963; Steedman & Craig 1983; Searle & Semeniuk 1985). Summer is characterised by a strong seabreeze/landbreeze system that generates local windwaves and onshore sand-mobilising winds. These derive from the west, to southwest, to north, in southern to middle to northern parts of the coast, respectively. As a result, the coastal zone experiences combined swell and wind-waves largely from southwest to south, and wind from similar directions, all of which shape and affect the coast profoundly (Searle & Semeniuk 1985). Southwesterly swell and wind waves impel a net northwards shoreline transport of sand, which locally is trapped behind energy shadows of

rocky reefs and islands to develop cuspate forelands and tombolos (Semeniuk et al 1988). Within a general tectonic setting in the Perth Basin of uplift to the north and subsidence to the south (expressed as the uplifted inliers and higher relief faulted splinters to the north, and the Bunbury Trough to the south), and longterm megascale protection from the Indian Ocean by a Precambrian inlier (the Leeuwin-Naturaliste Ridge), there is a dominance of limestone ridges, barriers and rocky reefs to the north of the Leschenault-Preston Sector, a general absence of such structures to the south, and a relatively low-energy coast to the far south (Figure 16.5). Consequently, the middle to northern part of the coast adjoining the Swan Coastal Plain is a bathymetrically complex, with cuspate forelands and local tombolos formed behind energy shadows developed leeward of the chains of

188

V. Semeniuk and C. A. Semeniuk

Table 16.2 Coastal sectors, their generalised form and origin (from Searle & Semeniuk 1985). Coastal sector

Generalised form and origin

Wedge Island to Dongara Sector (northernmost sector)

High energy, semi-protected coast, with cuspate forelands and tombolos developed behind the offshore chains of rock reefs and islands, and limestone cliff coasts developed along the more exposed parts of the coast; since the wind-wave regime has intensified, the wind waves erode the coast to develop a limestone rocky shore dominated coast.

Whitfords-Lancelin Sector

Cape Bouvard - Trigg Island Sector

High energy, semi-protected coast, with cuspate forelands and tombolos developed behind the offshore chains of rock reefs and islands, and limestone cliff coasts developed along the more exposed parts of the coast; the largest accumulation of sand occurs in the Rockingham-Becher cuspate foreland, which is developed as a result of sand transported from the eroding dune barrier to the south meeting the first major zone of shelter.

Leschenault-Preston Sector

High energy, unprotected coast, with development of dune barrier, barring a shoreparallel linear lagoon and estuary; dune barrier rapidly retrograding and eroding, supplying sand by northwards net shoreline transport to more northerly sectors.

Geographe Bay Sector (southernmost sector)

Relatively protected, prograding low barrier generating beach ridge plain, with dune swale wetlands; interaction with numerous short creeks and rivers; discharge to sea develops dynamic shore-parallel elongate estuarine complexes.

offshore barriers of limestone rocky reefs and islands (Searle & Semeniuk 1985). The southern third of the coast, free of offshore barriers, is relatively simple, with development of a barrier dune system (Semeniuk 1985, 1995c) and the southernmost part is a dynamic complex of low barriers and shore-parallel estuaries. In this context, the coastal zones can be divided into five broad sectors (Table 16.2). Key features from these sectors are described below. Within the Geographe Bay Sector, incorporating the onshore plain, the Vasse-Wonnerup estuary, and the shore, the system of Holocene to Pleistocene shore-parallel elongate estuarine complexes stands unique in Western Australia, because the semi-sheltered nature of the coast where beach-ridge barriers are forming and interacting dynamically with fluvial outlets. In the Leschenault-Preston Sector, characterised by an extensive shore-parallel barrier dune, there is a distinctive and important submarine geomorphic system: submerged beachrock bands and low ridges, marking the style of coastal retreat (Figure 16.6) (Semeniuk & Searle 1987). These beachrock bands/ridges have formed as a result of ongoing long-term coastal erosion (Semeniuk 1985), temporally marked by rapid coastal retreat induced by episodic storms, within a setting where beachrock is forming in the coastal zone, and in a context of more pronounced coastal erosion occurring to the south. Within the Cape Bouvard to Trigg Island Sector, predominantly south of Perth, there is the Rockingham-Becher Cuspate Foreland, a large accumulation of sand derived from the eroding Leschenault-Preston barrier to the south. The area is the location where the first series of offshore limestone barriers act to shelter the mainland shore, and

hence there has been a massive buildup of sediment over 7000 years to develop the largest (double) cuspate foreland system in Western Australia. Deriving from this situation are a range of landforms and wetlands that are nationally unique. This sedimentary accumulation is the largest, most continuous, and internally variable repository of Holocene archival material on Holocene coastal evolution, soil development, wetland development and stratigraphic evolution, climate changes (Semenuik 1986), and Holocene coastal process history in Western Australia. The next coastal sector, immediately north of Perth, contains a series of smaller, more discrete cuspate forelands, and there is a significant story in the development and natural destruction of these coastal landforms (Figure 16.7). The localised barriers of rocky reefs and small islands that helped form, shelter and sustain cuspate forelands, in time geomorphologically degrade and erode through biological, chemical and wave processes (Semeniuk & Johnson 1985). Additionally, cuspate forelands, formed initially behind a given offshore barrier or rocky reef, were in dynamic equilibrium under specific formative conditions of swell, wind strength and direction, and wind-waves. With changing net direction of coastal wind over the Holocene, and hence changing wind-wave and swell directions, the cuspate forelands were subject to progressively different oceanographic and wind conditions, and progressed to a state of disequilibrium. As a result, depending on their orientation relative to their protective barriers, cuspate forelands are in various stages of natural erosion. These stages represent important lessons of national significance in the story of coastal development.

Human impact Swan Coastal Plain

189

Figure 16.6 Occurrence and patterns of distribution of beachrock bands and ridges in the submarine nearshore of the Leschenault Peninsula (Semeniuk & Searle 1987). The bands and ridges record successive phases of coastal retreat. See Figure 16.5 for location.

Figure 16.5 Distribution of the five coastal sectors along the Western Australian coast adjoining the Swan Coastal Plain (after Searle & Semeniuk 1985). Note that the nearshore coastal zone becomes more bathymetrically complex from the southern part of the Cape Bouvard - Trigg Island Sector, northwards, as shown by the occurrence of offshore islands and reefs (stippled). From south to north, regionally, coastal wind changes in net direction and strength strongly influence the development of coastal dune form and orientation. Short parabolic dunes with axes oriented easterly progressively change to northeasterly inland ingressing, more attenuated parabolic dunes and fretted parabolic dunes, to northerly inland ingressing, highly attenuated parabolic (hairpin) dunes (Figure 16.8). Winds also influence the development of shore types along inland wetland margins (beach ridges, and lunettes).

Variety of estuaries Estuaries along the coast of the Swan Coastal Plain exhibit a regional variation, each with a distinct form related to their coastal setting (Hesp 1984): from north to south, the climate and the drainage catchment becomes more humid,

the rivers interact with varying ancestral coastal types, and the Holocene sea-level history has been different (Semeniuk & Searle 1986a). The main estuaries are: the Moore River estuary, the Swan-Canning River estuary, the Peel-Harvey estuary, the Leschenault Inlet estuary, and the Vasse-Wonnerup estuary. The Moore River estuary is a small estuary interacting with a mobile, dynamic dune barrier which at the coast has diverted the generally westerly course of the river. The Swan and Canning River estuary is a Pleistocene to Holocene shore-normal inundated river valley that traverses a range of Swan Coastal Plain geomorphic belts (the pre-Holocene age of the Swan River valley was documented by Churchill 1959). The Peel-Harvey estuary is a Pleistocene to Holocene, geomorphically compound and complex estuary (Semeniuk & Semeniuk 1990, 1991), formed in part by post-glacial inundation of both an area of confluence between three rivers (to form Peel Inlet), and the linear inter-ridge swale formed at the junction between the belt of limestone ridge and quartz sand hills (to form Harvey estuary). Deltas bordering the Peel-Harvey estuary, formed where the various rivers enter the relatively protected waters of the estuary, have responded to the environmental setting in relation to fetch, and dominant and prevailing wind directions to develop wave-dominated

190

V. Semeniuk and C. A. Semeniuk

CERVANTES STAGE

WHITFORDS STAGE

BURNS BEACH STAGE

SiS

mm. SSI

destruction of shelter

Figure 16.7 Long-term dynamics of cuspate forelands showing the three main stages related to the progressive loss of the shelter effect of the offshore rocky reefs. The first stage, exemplified by the Cervantes cuspate foreland, is the prograded fully developed cuspate foreland developed behind a large offshore rocky reef; the cuspate foreland has built out from a limestone cliff-line. The second stage, exemplified by the Whitfords Cusp (Semeniuk & Searle 1986b), following diminution of the offshore rocky reef shelter, is an eroding cuspate foreland, and through coastal retreat there is developed a series of inland ingressing parabolic dunes cross-cutting the primary geomorphology of the cuspate foreland. The third stage, exemplified by the Burns Beach system, is a nearly fully eroded cuspate foreland with an abundance of retrograded parabolic dunes; the coast is almost cut back to the buried limestone cliff-line.

deltas and fluvially dominated deltas (Semeniuk & Semeniuk 1990). Specific shores along the Peel-Harvey estuary, comprising spits, narrow beach-ridge plains, and bar-and-lagoons, are variably important from a geoheritage perspective in that they contain a record of Holocene estuarine geomorphic coastal history, and sea-level history. The tidal delta leeward of the oceanic entrance to Peel Inlet is a large, complex accumulation of sediment. The deposit resulted from shoreline sand drift, en route from the eroding Leschenault-Preston Barrier to the Rockingham-Becher Cuspate Foreland, being entrained by tides to form a floodtidal delta. The importance of the resulting tidal delta lies in its large size, complex history related to shifting centres of accumulation, and developmental history under conditions of a variable sea-level (Semeniuk & Semeniuk 1991). The Vasse-Wonnerup estuary system, located along the Geographe Bay coast, as noted earlier, is a dynamic complex resulting from the interactions of creeks/rivers and low coastal dune barriers. There is no defined delta, with most of the sediment accumulating under laterally migrating and meandering channel and point-bar systems, with local temporary outlets of the fluvial channel being periodically blocked by the low barrier dunes. Inland there are

Pleistocene equivalents of this system, with low limestone ridges (former low barriers) alternating with ribbons of estuarine sediment and fluvial mud, that provide textbook examples, illustrating the interrelationship of landforms, stratigraphy, soil types, groundwater and vegetation. They also provide Quaternary models of low dune barriers and estuarine sedimentation useful for interpreting the geologic record. In contrast to the estuaries above, the Leschenault Inlet estuary, formed behind a high dune barrier, is wholly Holocene in age (Semeniuk 1985). It once consisted of a narrow, elongate lagoon some 80 km long, similar to The Coorong in South Australia, with deltas to the south. Around 4.5 ka BP, the lagoon was segmented by parabolic dune encroachment, forming the now isolated Lake Preston lagoon to the north, and the Leschenault Inlet estuary to the south. The Leschenault Inlet estuary has a number of features that combine to make it significant: sedimentologically and stratigraphically it is simple at the large scale, but complex in detail because it formed under conditions of a variable sea-level (Semeniuk 1985, 2000) and parabolic dunes encroach on its western shore to form bar-andlagoon complexes. Two contrasting deltas occur in its

Human impact Swan Coastal Plain

Sector Wedge Island-Dongara

Dominant Coastal Process

Continuity of Coastal Dunes

Dominant Land Form (plan)

parabolic Limestone coast Discontinuous in Attenuated dunes^^^^^^ retreat and landward discrete accumulations dune encroachment connected by thin alternating with areas ribbons Bowls of coastal progradation in cusps Shore parallel ridges

Whltford-Lancelln

Limestone coast Discontinuous in Parabolic dunes retreat and landward discrete accumulations dune encroachment connected by thin alternating with areas ribbons of coastal progradation in cusps

Cape B o u v a r < i ~ J » f t Trigg I s l a n d ^ ^ B M l

Shoreline progradation and development of beachridge plain

Continuous extensive cuspate plain

Shore parallel ridges

Parabolic dunes

Leschenai j l t - P r e » t o n

Barrier retreat and development of blowouts and parabolic dunes

Continuous linear high-relief ridge

o o

Parabolic dunes

^^Blowouts Undulating plain

Geographe Bay

Shoreline progradation

Continuous arcuate Undulating plain low-relief ribbon Shore parallel ridges

Figure 16,8 The changes in morphology and dune types of the coastal dunes south to north along the various coastal sectors (after Semeniuk et al 1989). The illustration shows the changes in dominant coastal processes, the continuity of coastal dunes, and the idealised dominant dunal land form for each sector.

southern part, and it supports, as a population outlier, the most southern occurrence of the white mangrove, Avicennia marina, in Western Australia (Semeniuk 2000). These deltas are especially interesting in that one formerly was tide-dominated, with the delta outlet located where tidal processes dominated, and the other combines elements of wave-dominated and fluvial-dominated aspects within the same delta; both deltas are important estuarine deltaic features in Western Australia (Figure 16.9).

Sea-level history Along the southwestern Australian coast, there is also an important sedimentary archive and geomorphic record of a variable Holocene sea-level that Playford (1977, 1988), Semeniuk and Searle (1986a), Semeniuk and Semeniuk (1991), and Semeniuk (2000) related to Holocene tectonism. The sea-level history resides in the local sedimentary accumulations (coastal deposits and estuarine deposits), and limestone rocky shores, respectively (Fairbridge 1960; Playford 1977, 1988; Semeniuk 1985, Searle et al 1988, Semeniuk & Semeniuk 1991; Hamilton & Collins 1997). The variable sea-level history of this coast is a globally significant feature yet to be fully explored in terms of Quaternary tectonics and eustasy, coastal and continental shelf rheology in response to loading, the response of

191

coastal dynamics to the variable relative sea-levels, sedimentary accumulations, sediment budgets, and groundwater responses.

Limestone terrains Inland, in the Mandurah to Bunbury area, there is a limestone belt called the Yalgorup Plain [Semeniuk 1997: formerly referred to as the Spearwood Dunes (McArthur & Bettenay 1960) and Yoongarillup Plain (McArthur & Bartle 1980b)] that exhibits landforms and stratigraphy of former cuspate foreland, and quartz-sand-ribbon depositional systems. This system contains globally important Quaternary examples of Pleistocene beach ridge patterns, cuspate foreland geometry, classic seagrass bank to beach to dune shoaling stratigraphy, and wetlands intimately related to the post-depositional geomorphic modification of the plain.

Relict desert dunes and relationship to coastal dune deposits In the central to northern inland parts of the Swan Coastal Plain, the quartz-sand hills (Bassendean Dunes) exhibit features of star-dune morphology and complex linear megadune morphology (Figure 16.10) (Glassford 1980; Semeniuk & Glassford 1987). These are relict from earlier, more arid climates. Today, these dune forms are in a humid climate and vegetated mostly by a Banksia association. In addition, the western margin of the quartz-sand hills exhibits a complex stratigraphic relationship of intercalations of limestone and quartz sand, resulting in limestone lenses embedded in quartz-sand formations (Semeniuk & Glassford 1989), a feature of geological interest illustrating a type of geological contact between continental and coastal deposits.

GLOBALLY TO REGIONALLY SIGNIFICANT NATURAL HISTORY RESOURCES It is evident, from the above descriptions, that there is a range of features on the Swan Coastal Plain that are of geoheritage significance. The level of significance needs to be addressed. It should be noted that global or national significance can only be appreciated by people who value knowledge and heritage, and by those who know about, or can compare, similar overseas features. In this chapter, the notion of global significance applies to geologic features and areas that satisfy one or more of a number of criteria. First, they are not represented elsewhere globally (e.g. the linear wetlands formed by karst). Second, they exhibit features that illustrate to the geologic scientific community, important and exemplary processes and principles (e.g. Pleistocene cuspate foreland geomorphology, and seagrass to beach to dune stratigraphy therein; wetland development and processes in beach ridge swales in the Rockingham-Becher area). Third, they exhibit a gradation of forms that illustrate the principle behind their development (e.g. geomorphic types of cuspate forelands in the Whitfords-Lancelin Sector). Next, they represent globally the first description of a geologic principle or pattern (e.g. beachrock bands/ridge in the Leschenault Peninsula area).

192

V. Semeniuk and C. A. Semeniuk

1997

1966 J\ "FAN DELTA' K/ytldal \ / (shoals J/N

LANDFILL & ROADS

HARBOUR EXCAVAHON

u DELTA^ > ARTIFICIAL

Hr» SifiAdelta fiSALf*iIAiN| mm W^o Ito I

h River Preston £° delta 5'ver

llle

d e , t a

KJsj

s Pits & beachridges

sand

[ beachridge | barrier

i « P m u d - f i l l e d &S75EL h i i abandoned I delta plan ^ chamei

e~irrri . |i urbanised

Figure 16.9 The estuarine deltas in the Leschenault Inlet estuary (Semeniuk 2000). (a) Location map. (b) Location of the Preston River delta and the Collie River delta in southern Leschenault Inlet, (c) More detailed map of the Preston River delta in 1966 showing the subaerial delta plain, meandering river channels, tidal shoals, and the fan delta, (d) Aerial photograph centred on the delta in 1966, showing tidal shoals, fan delta, and meandering channels on the subaerial plain, (e) Same area shown in (c), but with four types of anthropogenic impacts (landfill and roads, harbour excavation, channel diversion, and redirected deltaic sedimentation) largely destroying the integrity of the delta by 1997. (f) Aerial photograph of the Collie River delta in 1978 showing main channel, mid-channel island, cheniers on the northeastern part of the delta, subaerial plain and abandoned distributary channels on the southern part of the delta, (g) Map of the Collie River delta in 1978, based on the aerial photograph shown in (f) showing the geomorphic components of the delta, (h) Map of the Collie River delta in 1999, showing the destruction of the southern part of the Collie River delta through urbanisation.

Finally, they contain important archival information in relation to geologic, biologic or climatic history that supplements the general picture that may be emerging globally. For example, the climate history recorded in the sediments of wetlands provides information on the functioning of climate during the Holocene in this part of the world, or that preserved in star dunes that provides information on Quaternary climate changes in relation to glacial and interglacial periods. In many situations, a feature of the Swan Coastal Plain provides information or a geological insight not represented elsewhere globally. Three examples are provided here. The first is the relationship of tectonic setting, coastal processes and sedimentation in a basin-wide context, as exemplified by the relationship of cuspate forelands along the northern coast and the barrier dune system along the southern coast. The second is the diversity and array of wetland types on the Swan Coastal Plain, illustrating a variety of geomorphology, hydrologic mechanisms and origins. The third is the internal structures of star dunes, with their buried soils, cross-stratification and unconformities related to former dune surfaces. The information from the star dunes provides insight into Quaternary glacial/interglacial desert history, and evolution of star-dune forms, as preserved on the Swan Coastal Plain. Such features, well preserved here,

have not been explored in modern star-dune environments elsewhere globally because of logistic constraints, hence the Swan Coastal Plain star dunes serve as an international model for understanding the development and evolution of such dune forms. National significance and state-wide significance apply to geologic features, essentially similar to those above, but important to the understanding of the geologic history of Australia and Western Australia, respectively, and representing perhaps the Australian examples of features present elsewhere globally. These features are important for research, education, illustrating principles, and natural outdoor museums. More detailed criteria for national significance are presented in Australian Heritage Commission (1990). Regional significance applies to features, similar to those outlined above, that may be present elsewhere globally and nationally, but that are well-developed in a given area to be of importance to local universities, schools and natural history students. For example, a well-exposed limestone cliff illustrating exceptionally well preserved aeolian cross-layering and buried soil sheets, though well represented elsewhere in Australia, may be important for education and research to the regional population. With these criteria, thirteen key areas and/or features of the Swan Coastal Plain and coastal zone that are of geo-

Human impact Swan Coastal Plain

triple-armed star dune

Figure 16,10 Star dune in the Ellenbrook area (see Figure 16.11 for location) showing trple-armed form of the dune, and the cuspid nature of the some of the arms. Left illustration shows the topographic contours above MSL, and right illustration shows interpretation of the star-dune form.

central peak

1 km

contour in 5 m Intervals

heritage significance are listed and described briefly below (Figure 16.11). (1) The Becher Cuspate foreland (Searle et al 1988): the most southerly and part of the the largest (double) cuspate foreland in Western Australia, with its suite of wetlands (the Becher Wetland Suite: C. A. Semeniuk 1988). This is a globally significant beach-ridge plain and associated Holocene wetlands that store information on climate change, Holocene sea-level history, former coastal processes, wetland evolution, wetland hydrochemical processes (the latter especially is of international significance). (2) Cuspate forelands in the Whitfords-Lancelin Sector (Searle & Semeniuk 1985; Semeniuk & Searle 1986b): a range of cuspate forelands of varying sizes and in different stages of natural geomorphic destruction that illustrate the principles of construction and natural destruction of such coastal landforms. The ensemble of these coastal forms is of global significance. (3) Yalgorup Plain (Semeniuk 1997): Pleistocene cuspate forelands showing a history of transgression and regression, Pleistocene climate and sea-temperature changes, beach-ridge history, and the relationship of the cusp axes to offshore barriers; also illustrates the development of wetlands along unconformity contacts, and the karst control of marine water recharge. This area is globally significant. (4) Swan Coastal Plain wetlands (excluding those formed within the Holocene beach ridges and dunes): the variety of geomorphically diverse wetland suites encompassing systems in the Bassendean Dunes, the Spearwood Dunes and Pinjarra Plain (C. A. Semeniuk 1988). The wetlands display a diversity of shape, size, association, stratigraphy, history, and margin types (e.g. lunettes that border them locally). These wetland ensembles, lying in parallel belts in the Swan Coastal Plain, in contrast to wetland types and assemblages in North and South America, Canada, Europe, Asia and Africa, are globally unique systems. (5) Leschenault Peninsula beach rock bands/ridges (Semeniuk & Searle 1987): comprising low bands and

193

KEY AREAS/FEATURES .'Y\ Becher cuspate J. foreland Whitfords-Lancelin \~-_J Sector cuspate forelands ( 3 ) Yalgorup Plain {4

Swan Coastal Plain wetlands Leschenault Peninsula Y ' beach rock bands/ridges

;

Forrestdale/Ellenbrook Star Dunes

( 7 • Peelhurst wetlands ; g :

Peel-Harvey estuary coastal landforms

: 9 • Preston River delta i ' n , Preston River delta ' M mangrove environments (11; Collie River delta , n n, MIrrabooka and Jandakot areas: linear megadune ,t

Busselton Plain and the Vasse-Wonnerup estuary

Figure 16.11 Location of 13 sites of geoheritage significance on the Swan Coastal Plain. ridges of submerged beachrock, formed where the coast retreated episodically; as the most southern beachrock location along an eroding coast that records extent and periodicity of coastal retreat. This feature is of national significance.

194

V. Semeniuk and C. A. Semeniuk

Table 1 6 . 3 Types of human impacts on the environment, Swan Coastal Plain. Human impact Urbanisation

Description with examples Extensive areas cleared for human habitation with levelling of coastal dunes and/or star dunes, filling of wetlands, excavation of peat from wetlands, nutrient enrichment of waterways and groundwater; alteration of groundwater regimes.

Marina and harbour construction, and general coastal modification

Excavation and landscape modification of coastal dunes, and peripheral estuarine landforms such as deltas; emplacement of breaker groynes and seawalls that fundamentally alter the wave character of the coast.

Clearing for industrial or recreational

Extensive areas cleared for development and exploitative landuse with levelling of

development (e.g., industry/golf),

coastal dunes and/or star dunes, filling of wetlands, nutrient enrichment of

or for pasture, horticulture, forestry

waterways and groundwater; alteration of groundwater regimes.

Mining and quarrying

Exploitation of carbonate sand from seagrass banks, mining of peat and diatomaceous earth from wetlands; quarrying of yellow sand for building purposes; mining/quarrying of mineral sands and use of the overburden as spoil dumps; quarrying of silica sand for glass-making.

Refuse disposal

Reclamation Groundwater contamination

Excavation and filling of dunes; filling of wetlands; nutrient enrichment of waterways and groundwater. Filling of wetlands; filling of estuarine deltas. Disposal of effluents that irrevocably modify the material of the natural environment; where groundwater contamination is of a nature that is not irrevocable (e.g., nutrient enrichment), in the medium term, it may be flushed by meteoric recharge.

Groundwater Abstraction

Alteration of groundwater regimes, and often extreme modification and destruction of wetlands.

Road Construction

Alteration of landscape and of groundwater regimes; fragmentation of landscapes; often modification and destruction of wetlands by infilling and peat excavation.

(6) Star Dunes in the Forrestdale/Ellenbrook area (Semeniuk & Glassford 1989): multi-arrayed to cuspate medium-sized star dunes and star-dune complexes, formed under conditions of a former arid climate. These features are of state-wide to national significance. (7) Peelhurst Wetlands (C. A. Semeniuk 1988): the wetland suite that was developed as dune slack systems along a line of natural blowouts on the southern coast of Becher Cuspate Foreland some 1000 years ago. This area represented a young example of a style of wetland development showing relation of wetland to dune-blowout history, and evolution of wetlands soils from ca 1000 a B.P. to present. When extant, the system was one of a kind in Western Australia, and was of state-wide significance. (8) Peel-Harvey estuary coastal landforms (Semeniuk & Semeniuk 1990): stranded channel shoal complexes, relict tidal delta system, modern tidal delta, spit-and-lagoon complexes, and beach-ridge ribbons. This area illustrates the development of estuarine peripheral coastal landforms, their soils and vegetation under conditions of a variable sea-level, changing climate, oceanic littoral sand drift, and coastal aspect. In terms of geoheritage, this system is of state-wide significance (9) Preston River delta, that occurred in the southern part of the Leschenault Inlet estuary (Semeniuk 2000): the combined fluvial and tidal dominated delta. This area is of state-wide significance. (10) Mangrove environments of the Preston River delta (Semeniuk et al in press): essentially mangrove fringed tidally oriented shoals. This feature is of state-wide significance.

(11) Collie River delta, that occurred in the southern part of the Leschenault Inlet estuary (Semeniuk in press): the combined fluvial dominated and storm-wave influenced delta. This area is of state-wide significance. (12) Linear megadune, Mirrabooka area and Jandakot area (Glassford 1980): two related linear sand-dune complexes formed under conditions of a former arid climate. This feature is of state-wide to national significance. (13) Busselton Plain and the Vasse-Wonnerup estuary: the low plain inland from Geographe Bay, comprising alluvial plains, geomorphically degraded dunes, fluvial courses, wetland complexes, and Pleistocene to Holocene estuarine complexes developed as a barrier and linear estuary complex. This area is of state-wide significance.

TYPES OF HUMAN IMPACTS With population growth, there has been a concomitant impact on the natural history features of the Swan Coastal Plain and coastal zone. In this region, as elsewhere globally, expanding human populations, through exploitation and modification of the environment, have contributed to a general lowering of the quality of the natural environment. This has occurred through deforestation, general vegetation clearing, alteration of soils, nutrient enrichment of waterways and groundwater, water abstraction, reclamation, diverting of river channels, and re-landscaping of natural landforms, amongst others. It is not implied here that these activities are wholly unacceptable, but rather that they

Human i m p a c t Swan Coastal Plain

195

Table 16.4 Key sites of geoheritage significance, Swan Coastal Plain and coastal zone, their destruction or modifications to date by human development. Key geoheritage area

Anthropogenic impacts

Becher Cuspate Foreland (i.e. remaining southern part of the double cuspate foreland: the RockinghamBecher Cuspate Foreland)

Urbanisation, industrial development, and golf course development to the extent that only 30% remains in Conservation Reserve Estate; urban and recreational development proceeded in the 1990s along the cusp axis, the most important part where wetlands were preferentially developed; in 1991 the Australian Heritage Commission recognised the geoheritage importance of the area, but this did not arrest its destruction and modification (Binnie & Partners 1988; Bowman Bishaw Gorham 1994; Environmental Protection Authority 1989, 1992a; Department of Planning and Urban Development 1993).

Cuspate forelands in WhitfordsLancelin Sector

Urbanisation during the 1970s to the 1990s has destroyed a number of these coastal landforms (LeProvost Environmental Consultants 1990; Department of Planning and Urban Development 1992a).

Yalgorup Plain

Though a portion of this system resides in the Yalgorup National Park, the majority was cleared for pasture in the past, and in the 1990s partly cleared for forestry; there has been urbanisation in the 1990s of large tracts of this plain.

Swan Coastal Plain wetlands (excluding those formed within the Quindalup Dunes

Urbanisation and road construction, clearing, infilling, and peat excavation, have largely destroyed wetlands of the Gnangara Suite and the Gwelup Suite, and wetlands in the type location of the Jandakot Suite (Department of Planning and Urban Development 1992b; Ministry for Planning 1995b); forestry and agriculture has largely destroyed wetlands of the Riverdale Suite and wholly destroyed wetlands of the Kooallup Suite.

Leschenault Peninsula beach rock bands/ridges

Disposal of acidic and iron-rich effluent on the Leschenault Peninsula during the 1970s and 1980s, resulted in seepage seawards, destroying this feature through gypsification and dissolution.

Star Dunes Forrestdale/ Ellenbrook

Urbanisation and development of golf courses in the Forrestdale and Ellenbrook area in the 1990s have resulted in the wholesale clearing and levelling of many such dunes (Department of Planning and Urban Development 1992b, 1994, 1996; Ministry for Planning 1995b).

Peelhurst Wetlands

Urbanisation and golf course development proceeded in the 1990s and fully destroyed this wetland suite. In 1991 the Australian Heritage Commission recognised the geoheritage importance of the area, but this did not arrest its destruction (Environmental Protection Authority 1992; Tingay & Associates 1992; Tingay & Associates and Warren F Johnson & Co 1992), Department Planning and Urban Development 1993).

Peel-Harvey estuary coastal landforms

Reclamation and marina development during the 1980s to 1990s destroyed many areas along the estuary shore, including the stranded shoal complex, the relict tidal delta, and locally the spit-and-lagoon complexes (BSD Consultants Pty Ltd 1992; Environmental Protection Authority 1992b); the last destruction was under way in 1999 with another urban/marina complex being developed on the remainder of the relict tidal delta complex, in spite of the fact that it was on the Register of the National Estate (Australian Heritage Commission), and that it adjoins and is part of a site nominated for conservation under the Ramsar Convention (Ramsar Convention Bureau 1991; Lane et al. 1996).

Preston River delta

Reclamation and harbour development, and diversion of the Preston River; these modifications took place in the late 1970s, and irrevocably altered the nature of the delta.

Mangroves of the Preston River delta

Though part of the Preston River delta, the mangrove habitats within the delta are singled out because of their State-wide significance; impacts derive from reclamation and harbour development, and road construction during the 1970s.

Collie River delta

Urbanisation of the southern subaerial plain in the late 1990s (LeProvost Environmental Consultants 1991) involving clearing, excavations, levelling; drainage of the northern subaerial plain along cheniers in the 1940s.

Linear megadune, Mirrabooka area and Jandakot area

Sand quarrying and adjoining urbanisation in the 1960s has resulted in the destruction and local modification through levelling, respectively, of this twin dune form.

Busselton Plain and the Vasse-Wonnerup estuary

Mineral sand mining from the 1960s to the 1990s has destroyed wetlands and landscapes of this region, and largely erased a unique and important stratigraphic record; coastal marina development has canalised the Holocene estuary system.

196

V. Semeniuk and C. A. Semeniuk

aO LU t— go

would not be so indiscriminately modified or destroyed. Indeed there has been a raising of consciousness in this matter (Carter 1987; Lemmon et al 1979; Joyce 1995): the concept of geoheritage is now embedded in the Australian conservation literature, criteria exist for Nomination for Selection for the Register of the National Estate by the Australian Heritage Commission ( 1 9 9 0 ) , and the Geological Society of Australia commissioned a study into the design of selection criteria for sites of geoheritage and geoconservation (Joyce 1995).

Zone of scattered \ smaller scale cuspate \ forelands In various >\J stages of construction \ and destruction V Major zone of sediment trapping and accumulation into a large cuspate foreland

13L 182

DISCUSSION

Zone of sediment trapping in an estuary entrance Zone of major coastal transport

Zed

Zone of net coastal retreat

HoJocene coastal , sand deposits: ! barriers, prograded ! plains. cuspate foretands, and Inland ingressing/ retrograding dunes

50 km

Figure 16.12 Regional processes and dynamics between Leschenault - Preston Sector and Whitfords - Lancelin Sector showing interrelationship of coastal types. should be undertaken in a rational manner, encompassing the issues of biodiversity, geoheritage, pollution/contamination, and wise environmental management. The notion is presented that, for example, urbanisation, quarrying, marina construction or mining, should not be undertaken in areas that are assessed as having high to moderate geoheritage significance. The types of human impacts on the natural environment that affect sites of geoheritage significance are described, with examples given in Table 16.3. The many and extensive areas impacted by human populations, but that are only of local to regional geoheritage significance (e.g. the Swan River estuary, the limestone rocky shores of Perth, the undulating sand hills of the Bassendean Dunes, and the catchment of the Brunswick and Collie Rivers) are excluded from consideration in this chapter. On the Swan Coastal plain, there already has been much destruction of areas of importance as geoheritage sites following activities listed above. Generally in Western Australia, in the past, and continuing to a large extent into the present, these modifications of the natural environment took place without any perception of geoheritage. Today, with the environmental controls imposed through the Environmental Protection Authority Act (1986) and some policies emplaced through the Department of Environment that attempt to protect wetlands and landforms (Department of Environment 1993), landform systems that were impacted decades ago would be critically examined and assessed as to their significance. Probably today they

Some of the areas on the Swan Coastal Plain that were of high geoheritage significance have been modified or destroyed. Key areas of geoheritage significance that were destroyed or modified are listed and described in Table 16.4. Some of these impacts occurred years before concepts such as geoheritage were conceived or in a context of ignorance (e.g. the Preston River delta of the Leschenault Inlet estuary). Some occurred in spite of the information about the geoheritage significance being readily available and that the information was part of an assessment process (e.g. the southern part of the Collie River delta of the Leschenault Inlet estuary; the Peelhurst Wetlands; the cuspate foreland of Point Becher with its wetlands arrayed along the cusp axis; see discussion below). Current perceptions about conservation and geoheritage now generally held by scientists and the general community, probably would not have been acceptable two hundred years ago at the time of the European settlement of Australia, and even 5 0 years ago, immediately after World War II—a factor indicating the changes in community attitudes to environmental and conservation issues that have occurred over the past few decades. Projecting this pattern into the future, with increasing awareness of environmental issues, education and personal wealth, one may hope that there will be a greater consciousness of the significance of geoheritage issues and a more concerted effort to preserve features of geoheritage significance. Future land managers and decision makers cannot plead naivety given increasing numbers of publications dealing with geoheritage issues, the documentation of the destruction of globally significant sites to date, and the focus on issues of geoheritage. An important issue in geoheritage in Western Australia is the notion of interconnectedness or linkage. Sites of significance can be viewed as site-specific, such as a stratigraphic type section, or a specific wetland, but in fact there is a regional interconnectedness of many features of the Western Australian geologic systems, especially in the Quaternary, and it is the integrated system that stands as a globally significant model. For instance, consider the coastal zone between Geographe Bay and Lancelin (Figure 16.12): there is a south to north interrelated pattern in sediment transport, sedimentary accumulations, and geomorphic evolution of the coast related to the regional system of tectonics (resulting in a submergence of the offshore barrier limestone chains and the south to north changes in tectonic response as reflected in a variable sea-level history), the gradient in climate during the Holocene (resulting in a change in resultant wind direction), the prevailing net northward littoral sand drifts, style of biological, chemical

Human impact Swan Coastal Plain and physical erosion of the limestone reefs, and the changes in maintenance of coastal form due to changes in wind patterns (resulting from Holocene climate changes). The total integrated picture is a model of global significance, and preservation of sites of geoheritage must address the integrated whole picture and not deal only with site-specific areas in a piecemeal fashion. To date, there have been significant losses of sites of geoheritage in Western Australia. This may be related to a range of complex issues. Environmental managers, government agency decision makers and politicians, despite being in positions of adjudication, frequently have not fully appreciated the significance of geologic processes and products, or the significance of geoheritage, i.e. naturalgeoscience information. In fact, they frequently do not understand the significance of heritage outside the realm of cultural heritage. There is also an unsubstantiated opinion that in a technological culture, science, technology and engineering can fix all, without appreciating that systems can be irrevocably destroyed. Examples here include star dunes with their stratigraphy and buried soils, for instance, that contain irreplaceable and valuable geohistorical information, which have taken hundreds of thousands of years to develop, or wetland systems that have taken ten thousand years to develop, and that are still very poorly understood in terms of their functioning, and indeed cannot be restored or reconstructed. This chapter has provided an overview of the significance of the Swan Coastal Plain, southwestern Australia, from a view of geoheritage and outlined some of the human impacts on globally, nationally and regionally significant sites. From this perspective, the objective of this chapter is to alert the scientific community and land-use managers and decisions makers to the important features of the Swan Coastal Plain, and to the destruction of the Australian geoheritage that has occurred to date, so that an understanding of such history will ensure that past errors in judgement are not repeated.

ACKNOWLEDGEMENTS We thank Vic Gostin for the opportunity to contribute to this volume. We also thank M. Brocx, P. Clifford and D. K. Glassford for critically reading earlier drafts of the manuscript and providing helpful comments, J. Unno for drafting several of the illustrations, and M. Freeman for assistance with the Landsat Image illustrated in Figure 16.2. The Landsat imagery was provided by Australian Centre for Remote Sensing, AUSLIG, Canberra and digitally enhanced and produced by Satellite Remote Sensing Services, Department of Land Administration, Western Australia.

REFERENCES 1990. Future directions in assessing National Estate Significance. Australian Heritage Commission, Canberra. BINNIE & PARTNERS PTY LTD 1988. Environmental Review and Management Program: Vol. 1: Summary Report for Port Kennedy Joint Venture. Binnie & Partners Pty Ltd, Perth. BIRD E. C. & SCHWARTZ M. L. (Editors) 1985. The World's Coastline. Van Nostrand Reinhold, New York. AUSTRALIAN HERITAGE COMMISSION

BLANDIN P. 1992. La nature en Europe.

197

Bordas, Paris.

BOLTON G. C. & HUTCHISON D. 1973. European Man in southwestern

Australia. Journal of the Royal Society of Western Australia 56, 56-64. BOWMAN BISHAW GORHAM 1994. Geomorphic Site Management Plan: Port Kennedy Regional Recreation Centre. Compliance and Progress Report Stage 1.2, Vol. II. Report prepared for Fleuris Pty Ltd by Bowman Bishaw Gorham, West Perth. BSD CONSULTANTS PTY LTD 1992. Amendment No. 183 City of Mandurah District Zoning Scheme No. 1A, Harbour City Canal Estate, with supporting Planning and Environmental Reports, Esplanade (Mandurah) Pty. Ltd. Perth, Western Australia. CARTER J. D. 1987. Important Geological Localities in Western Australia. Geological Society of Australia, Western Australian Division, Perth. CHURCHILL D. M. 1959. Late Quaternary eustatic changes in the Swan River district. Journal of the Royal Society of Western Australia 42, 53-55 CREASER P. P. 1994. An international earth sciences convention framework for the future. The Australian Geologist 90, 45-46. CRESSWELL I . D. & BRIDGEWATER P. 1 9 8 5 . Dune vegetation of the Swan Coastal Plain, Western Australia. Journal of the Royal Society of Western Australia 67, 1 3 7 - 2 4 8 . DAVIES J . L. 1 9 8 0 . Geographical Variation in Coastal Development (2nd edition). Longman, London. DAVIS R . A. 1 9 9 4 . The evolving coast. Scientific American Library. DEPARTMENT OF ENVIRONMENT 1 9 9 3 . A guide to wetland management in the Perth and near Perth Swan Coastal Plain area. Department of Environment, Perth, Bulletin 685. DEPARTMENT OF PLANNING AND URBAN DEVELOPMENT 1992a. Northwest Corridor Structure Plan—Final. Department of Planning and Urban Development, Perth. DEPARTMENT OF PLANNING AND URBAN DEVELOPMENT 1992b. Draft Jandakot Land Use and Water Management Strategy. Department of Planning and Urban Development, Perth. DEPARTMENT OF PLANNING AND URBAN DEVELOPMENT 1993. South-West Corridor Structure Plan. Department of Planning and Urban Development, Perth. DEPARTMENT OF PLANNING AND URBAN DEVELOPMENT 1994. North-East Corridor Structure Plan. Department of Planning and Urban Development, Perth. DEPARTMENT OF PLANNING AND URBAN DEVELOPMENT 1996. South-East Corridor Structure Plan (South of Armadale)—Final. Department of Planning and Urban Development, Perth. D U F F K. 1994. Natural areas: an holistic approach to conservation based on geology. In: O'Halloran D., Green C., Harley M., Stanley M. & Knill J. eds. Geological and Landscape Conservation, pp. 121-126. Geological Society of London, London. ENVIRONMENTAL PROTECTION AUTHORITY ACT 1986. Government Printer, Perth. ENVIRONMENTAL PROTECTION AUTHORITY 1989. Port Kennedy Regional Recreation Centre. Report and Recommendations of the Environmental Protection Authority to Port Kennedy Joint Venture. Environmental Protection Authority, Perth, Bulletin 398. ENVIRONMENTAL PROTECTION AUTHORITY 1992a. Urban development of lot 1001 (affecting part of System Six Recommendation M107), Singleton. Report and recommendations of the Environmental Protection Authority to Temwood Holdings Pty. Ltd. Environmental Protection Authority, Perth, Bulletin 649. ENVIRONMENTAL PROTECTION AUTHORITY 1992b. Harbour City canal estate, Mandurah. Report and recommendations of the Environmental Protection Authority to Esplanade (Mandurah) Pty. Ltd. Environmental Protection Authority, Perth, Bulletin 656. FAIRBRIDGE R. W. 1960. Eustatic changes in sea-level. In: Ahrens L. H., Press F., Rankama K. & Runcorn S. K. eds. Physics and Chemistry of the Earth, Vol. 4, pp. 99-185. Pergamon, Oxford. FRITH H . J . 1 9 7 3 . Wildlife Conservation. Australian Natural Science Library. Angus & Robertson, Sydney. GENTILLI J. 1979. Regions and Landscapes. In: Gentilli J. ed. Western Landscapes, pp. 3-45. University of Western Australia Press, Perth. GENTILLI J . & FAIRBRIDGE R . W . 1 9 5 2 . Physiographic diagram of Australia. Geographic Press, Columbia University, New York. GEOLOGICAL SURVEY OF W E S T E R N AUSTRALIA 1 9 9 0 . Geology and Mineral Resources of Western Australia. Geological Survey of Western Australia Memoir 3.

198

V. Semeniuk a n d C. A. Semeniuk

GLASSFORD D. K. 1980. Late Cainozoic desert aeolian sedimentation in Western Australia. PhD thesis, University of Western Australia, Perth (unpubl.). GLASSFORD D. K. & SEMENIUK V. 1995. Desert-aeolian origin of late

Cenozoic regolith in arid and semi-arid Southwestern Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 114, 131-166.

HAMILTON N . T. M . & COLLINS L . B .

1997. Morphostratigraphy

and

HEDDLE E . M . , LONERAGAN O . W . & HAVEL J . J . 1 9 8 0 . V e g e t a t i o n

com-

evolution of a Holocene composite barrier at Minninup, southwestern Australia. Australian Journal of Earth Sciences 44, 113-124. HAVEL J. J. 1968. The potential of the northern Swan Coastal Plain for Pinus pinaster Ait. plantations. Western Australia Forestry Department Bulletin 76. plexes of the Darling System, Western Australia. In: Atlas of Natural Resources, Darling System, Western Australia, pp. 37-72. Department Conservation & Environment, Perth. HESP P. A. 1984. Aspects of the geomorphology of south western Australian estuaries. In: Estuarine Environments of the Southern Hemisphere, pp. 6 1 - 8 3 . Department of Conservation and Environment Perth Bulletin 161. HOPKINS J . 1994. Geology, geomorphology & biodiversity. Earth Heritage 2, 3-6. HUMPHREY W. F. (Editor) 1993. Biogeography of the Cape Range, Western Australia. Proceedings of Symposium, W.A. Museum, Perth, 21 November 1992. Western Australian Museum Records Supplement 45. IUCN 1992. Regional Reviews. IVth World Congress on National Parks and Protected Areas, Caracas, Venezuela, 10-21 February 1992. The World Conservation Union., Gland, Switzerland. JARVIS N. (Editor) 1986. Western Australia: an Atlas of Human Endeavour (2nd edition). Department of Lands and Surveys and Education Department of Western Australia, Perth. JOHNSON D. W. 1919. Shore processes & shoreline development. Wiley, New York. JOYCE E. B. 1995. Assessing the significance of geological heritage. A methodology study for the Australian Heritage Commission by the Standing Committee for Geological Heritage of the Geological Society of Australia. LANE J., JAENSCH R. & LYNCH R. 1 9 9 6 . W e s t e r n Australia. In: Blackley

R., Usback S. & Langford K. eds. A Directory of Important Wetlands in Australia (2nd edition), pp. 762-943. Australian Nature Conservation Agency, Canberra. LEDEC G . & GOOLAND R . 1 9 8 8 . Wildlands: their protection, and management in economic development. The World Bank, Washington. LEMMON T. C . , G E E R . D . , MORGAN W . R . & ELKINGTON C . R .

1979.

Important Geological Sites in Perth and the southwest area of Western Australia. Geological Society of Australia, Western Australian Division, Perth. LEPROVOST ENVIRONMENTAL CONSULTANTS 1 9 9 0 . Eglinton Beach Resort Public Environmental Review Volumes 1 & 2. Report to Eglinton Resort Development Pty. Ltd. September 1990. LeProvost Environmental Consultants Report (unpubl). LEPROVOST ENVIRONMENTAL CONSULTANTS 1991. Pelican Point Public Environmental Review, Vols I and II. LeProvost Environmental Consultants Report (unpubl.). MARCHANT N. G. 1973. Species diversity in the southwestern flora. Journal of the Royal Society of Western Australia 56, 23-30. MARCHANT N . G . , WHEELER J . R . , RYE B . L . , BENNETT E . M . , LANDER N .

S . & MACFARLANE T. D . 1 9 8 7 . Flora of the Perth Region. Western Australian Herbarium, Perth. MARKOVICS G. 1994. National Parks and geological heritage—examples from North America, and applications to Australia. In: O'Halloran D., Green C., Harley M., Stanley M. & Knill J. eds. Geological and Landscape Conservation, pp. 413-416. Geological Society of London, London. MCARTHUR W. M. & BARTLE. G. A. 1980a. Landforms and soils as a basis for Urban Planning in the Perth Metropolitan North-West Corridor, Western Australia. CSIRO Land Resource Management Series 5. MCARTHUR W. M. & BARTLE G. A. 1980b. Soils and land-use planning in the Mandurah-Bunbury coastal zone, Western Australia. CSIRO Land Resource Management Series 6.

MCARTHUR W . M . & BETTENAY E. 1 9 6 0 . T h e development a n d distrib-

ution of soils of the Swan Coastal Plain. ,Western Australia. CSIRO Soil Publication 16.

M C N E E L Y J . A . , MILLER K . R . , REID W . V . , MITTERMEIE R . A . & WERNER

T. B. 1 9 9 0 . Conserving the World's Biological Diversity. I U C N , Gland, Switzerland. of National Parks to the community. Proceedings of 2nd National Wilderness Conference, University of Sydney, 23-25 November 1979. Australian Conservation Foundation, Sydney. MINISTRY FOR PLANNING 1995a. State Planning Strategy Discussion Paper: Population. Western Australian Planning Commission, Perth. MINISTRY FOR PLANNING 1995b. State Planning Strategy Discussion Paper: Managing Growth. Western Australian Planning Commission, Perth. MINISTRY FOR PLANNING 1996. State Planning Strategy Discussion Paper: Population. Western Australian Planning Commission, Perth. NAGLE C . 1998. Western Australia Year Book. Australian Bureau of Statistics, Western Australian Office, Perth. OSBORNE R. A. L. 1991. Caves as geological heritage. The Australian Geologist 81, 6-7. PLAYFORD P. E. 1977. Geology and groundwater potential. In: Geology and Hydrology of Rottnest Island, pp. 1-53. Geological Survey of Western Australia Report 6. PLAYFORD P. E. 1988. Guidebook to the geology of Rottnest Island. Excursion Guidebook. Geological Society of Australia, Western Australian Division, Perth.

M E S S E R J . & MOSELY G. (Editors) 1980. The value

PLAYFORD P. E . , COCKBAIN A . E . & L o w G . H .

1 9 7 6 . G e o l o g y of the

Perth Basin, Western Australia. Geological Survey of Western Australia Bulletin 124. RAMSAR CONVENTION BUREAU 1991. Proceedings of the 4 th Meeting of the Conference of Contracting Parties, Montreux, Switzerland. Ramsar Convention Bureau, Gland, Switzerland. SEARLE D. J. & SEMENIUK V. 1985. The natural sectors of the Rottnest

Shelf Coast adjoining the Swan Coastal Plain. Journal of the Royal Society of Western Australia 67, 1 1 6 - 1 3 6 .

SEARLE D . J . , SEMENIUK V . & W O O D S P. J . 1 9 8 8 . T h e

geomorphology,

stratigraphy and Holocene history of the Rockingham - Becher Plain. Journal of the Royal Society of Western Australia 70, 8 9 - 1 0 9 . SEDDON G. 1972. Sense of Place. University of Western Australia Press, Perth. SEMENIUK C. A. 1988. Consanguineous wetlands and their distribution in the Darling System. Journal of the Royal Society of Western Australia 70, 95-111. SEMENIUK C. A. & SEMENIUK V. 1 9 9 0 . T h e coastal landforms and

peripheral wetlands of the Peel-Harvey Estuarine System. Journal of the Royal Society of Western Australia 73, 9-21. SEMENIUK V. 1985. The age structure of a Holocene barrier dune system and its implication for sea-level history reconstructions in southwestern Australia. Marine Geology 67, 1 9 7 - 2 1 2 . SEMENIUK V. 1986. Holocene climate history of coastal southwestern Australia using calcrete as an indicator. Palaeogeography, Palaeoclimatology, Palaeoecology 53, 2 8 9 - 3 0 8 . SEMENIUK V. 1993. The mangrove systems of Western Australia— 1993 Presidential Address. Journal of the Royal Society of Western Australia 76, 9 9 - 1 2 2 . SEMENIUK V. 1995a. Coastal form, processes, and products along the arid Pilbara coast of northwestern Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 123, 49-84. SEMENIUK V. 1995b. An early Holocene record of rising sea level along a bathymetrically complex coast in southwestern Australia. Marine Geology 131, 177-193. SEMENIUK V. 1995c. The Holocene record of climatic, eustatic, and tectonic events along the coastal zone of Western Australia—a review. Journal of Coastal Research Special Issue 17, 247-259. SEMENIUK V. 1997. Pleistocene coastal palaeogeography in southwestern Australia—carbonate and quartz sand sedimentation in cuspate forelands, barriers and ribbon shoreline deposits. Journal of Coastal Research Research 13, 4 6 8 - 4 8 9 . SEMENIUK V. Sedimentology and Holocene stratigraphy of Leschenault Inlet. Journal of the Royal Society of Western Australia (Leschenault Inlet Special Issue) 8 3 (in press). SEMENIUK V . , CRESSWELL I . D .

& WURM

P. A . S .

1989.

Quindalup

Dunes: the regional system, physical framework and vegetation habitats. Journal of the Royal Society of Western Australia 71, 23-47

SEMENIUK V. & GLASSFORD D. K. 1 9 8 7 . Origin of limestone lenses in

Human i m p a c t Swan Coastal Plain Perth Basin yellow sand, southwestern Australia. Journal of the Royal Society of Western Australia 70, 35-47. SEMENIUK V . & GLASSFORD D. K . 1 9 8 9 . Bassendean and Spearwood Dunes: their geomorphology, stratigraphy and soils as a basis for habitats of Banksia woodlands. Journal of the Royal Society of Western Australia 71, 8 7 - 8 8 . SEMENIUK V . & JOHNSON D . P. 1 9 8 5 . Modern and Pleistocene rocky shore sequences along carbonate coastlines, southwestern Australia. Sedimentary Geology 44, 2 2 5 - 2 6 1 . SEMENIUK V. & SEARLE D . J . 1986a. Variability of Holocene sea-level history along the southwestern coast of Australia—evidence for the effect of significant local tectonism. Marine Geology 72, 47-58. SEMENIUK V . & SEARLE D. J. 1986b. The Whitfords Cusp—its geomorphology, stratigraphy, and age structure. Journal of the Royal Society of Western Australia 68, 29-36. SEMENIUK V . & SEARLE D . J . 1 9 8 7 . Beach rock ridges/bands along a high-energy coast in southwestern Australia—their significance and use in coastal history. Journal of Coastal Research 3, 331-342. SEMENIUK V., SEARLE D. J. & W O O D S P. J. 1988. The sedimentology and stratigraphy of a cuspate foreland, southwestern Australia. Journal of Coastal Research 4, 551-564. SEMENIUK V. & SEMENIUK C. A. 1991. Radiocarbon ages of some coastal landforms in the Peel-Harvey estuary, south-western Australia. Journal of the Royal Society of Western Australia 73, 61-71. SEMENIUK V. & SEMENIUK C. A. 1997. A geomorphic approach to global classification for natural inland wetlands and rationalisation of the system used by the Ramsar Convention—a discussion. Wetlands Ecology & Management 5, 145-158. SEMENIUK V., TAUSS C. & U N N O J. Avicennia marina in the Leschenault

199

Inlet area. Journal of the Royal Society of Western Australia (Leschenault Inlet Special Issue) 83 (in press). SILVESTER R. 1963. Coastal Engineering. Elsevier, Amsterdam. SOULE M. E. & W I L C O X B. A. (Editors) 1980. Conservation Biology: an Evolutionary-Ecological Perspective. Sinauer Associates Inc., Sunderland. STEEDMAN R. K . & CRAIG P. D . 1 9 8 3 . Wind-driven circulation in Cockburn Sound. Australian Journal Marine & Freshwater Research 3 4 , 1 8 7 - 2 1 2 . TINGAY & ASSOCIATES 1992. System 6 Area Ml 07 Discussion Paper. Alan Tingay & Associates, Perth, Report 92/6. TINGAY & ASSOCIATES & WARREN F. JOHNSTON & Co. 1 9 9 2 . Singleton Amendment to the Metropolitan Region Scheme (affecting part of System 6 area Ml 07), Consultative Environment Review Report 91/29. UYEDA S. 1 9 7 8 . The New View of the Earth. W. H . Freeman & Co., San Francisco. W I T H E R S P. C. & HORWITZ P. (Editors) 1996. Symposium on the design of reserves for Nature Conservation in southwestern Australia. Journal Royal Society Western Australia 79, 225-304. W O O D S P. G . 1 9 8 0 . Coastal management in Western Australia. Department Conservation & Environment Bulletin 49. W O O D S P. G., W E B B M. J. & ELIOT I. G. 1985. Western Australia In: Bird E., Schwartz C. & Schwartz M. L. eds. The World's Coastline, pp. 929-947. Van Nostrand Reinhold, New York. WYATT B. & Moss D. 1990. CORINE biotopes: the design, compilation, and use of an inventory of sites of major importance for nature conservation in the European Community. The Commission of the European Communities, Luxembourg. Received 2 February 2000; accepted 29 May 2000

Geological Society of Australia Special Publication 21, 201-213

CHAPTER 17—Late Quaternary sea-levels, climate change and South Australian coastal geology N HARVEY1, A. P. BELPERIO2 AND R. P. BOURMAN1 1 Department of Geographical and Environmental Studies,

University of Adelaide, SA 5005, Australia.

2 Minotaur Resources Ltd, 3 Boskenna Avenue, Norwood, SA 5067, Australia.

South Australian geological data provide evidence of Late Quaternary climatic and associated sea-level fluctuations. There is compelling evidence of Late Quaternary high sea-level stands associated with Oxygen Isotope stages 1 to 19, preserved by progressive uplift of the Southeastern coastal plain. The most comprehensive data are available for the last interglacial high sea-level stand, which provides an excellent datum from which the variable tectonic activity of the South Australian coastline may be determined. Well-defined geological data also exist for interstadial events during the last glacial (Oxygen Isotope stages 3, 5a and 5c) and for the Holocene postglacial sea-level rise, primarily from the marine and intertidal sediments of the South Australian gulfs. Sequences of river terraces and dunefields have developed in response to climatic and sea-level fluctuations during and since the Last Interglacial. Regional variations in the Holocene sea-level curve were derived from peritidal coastal environments adjacent to historic tide-gauge sites. Palaeo-sea-level indicators were identified in the sediments and accurately levelled to determine their positions relative to modern counterparts. The dominant indicators were the boundaries between the subtidal seagrass (Posidonia) fades and the intertidal sand-flat fades; between the sand-flat facies and the mangrove (Avicennia marina) fades; and between the mangrove facies and the samphire (Halosarcia-Sarcocornia and Sclerostegia-Halosarcia) marsh facies. Radiocarbon dating of these indicators facilitated the production of sea-level curves. These data reveal the influence of both post-glacial hydroisostatic warping of the shelf and anthropogenic influences near the tide-gauge sites, suggesting that there are many overestimates of the rates of global sea-level rise. The adjusted sea-level trend data have implications for South Australia's coastal policy which, although it is based on early more rapid global sea-level rise estimates, does make allowance for localised variation in uplift or subsidence. It is shown that geological studies can reduce some of the uncertainties of current sea-level measurements by correcting sea-level data for neotectonic and anthropogenic influences. These data are important in the context of the global predictions of a greenhouse-accelerated sea-level rise and country-based assessments of coastal vulnerability. KEY WORDS: climate change, coastal management, coastal sediments, greenhouse effect, Holocene, neotectonics, Quaternary, sea-level rise.

INTRODUCTION Numerous geological influences have helped to shape the South Australian coast but it is those of the Quaternary Period that have had the most dramatic effect on the appearance of the modern coastline, creating extensive coastal barrier and dune facies, back-barrier lagoon facies and peritidal facies (Figure 17.1). The Quaternary is of particular significance because of the impacts of oscillating sea-levels and alternation between pluvial and arid phases in the landscape. Throughout the past 1.7 million years of the Quaternary, periodic global climate changes have caused repetitive build-up and decay of continental-scale icesheets, resulting in alternate flooding and exposure of continental margins, advances and retreats of deserts and alternating incision and aggradation in stream valleys, associated with periods of erosion and sedimentation. Emiliani (1955) was able to outline the major features of the glacial-interglacial fluctuations by measuring oxygen isotope ( 1 8 0 / 1 6 0 ) ratios of planktonic foraminifers extracted from deep ocean sediments. This work estab-

lished a quasi-periodic cycling of climate that had recurred many more times than the four major glaciations recognised from previous North American and continental European stratigraphic studies. In Australia, these global climatic changes were manifested as repeated transgressions and regressions across the continental shelf, leaving an intermittent sedimentary record along the coast. Much of this sedimentary evidence occurs below present sea-level because the highstands associated with each transgression were rarely higher than the current highstand. Importantly, records of many of these fluctuations have been exposed above the present level of the sea in areas of neotectonic uplift. Of approximately 27 major glacial-interglacial cycles recognised in the Quaternary, at least 13 sea-level highstands are preserved as elevated shorelines (high-energy coastal barriers) on the Southeast Coastal Plain of South Australia (Figure 17.2), one of the best onshore preserved sequences in the world. Generally former shorelines are only sporadically exposed and preserved elsewhere around Australia, but the last interglacial shoreline, which stood higher than the pre-

202

N.Harvey etal.

Figure 1 7 . 1 Quaternary coastal sedimentation in South Australia.

sent sea-level, is sufficiently consistently preserved for documenting continent-wide neotectonic deformation. Within protected embayments such as the South Australian gulfs (Spencer Gulf and Gulf St Vincent), the repeated transgressions and regressions are preserved in the substrate as layers of marine sediment separated by palaeosols (Figure 17.3). At the landward limit of marine inundation, the marine strata comprise a complex of lowgradient, low wave-energy, tide-dominated peritidal facies. Like their Holocene equivalents (Belperio et al 1988), these tidal sediment facies contain clear evidence of the location and relative elevation of former sea-levels. The present South Australian coastline, of approximately 3700 km, embodies a variety of coastal landforms that include cliffed coasts, rocky outcrops and shore platforms, mangroves, mudflats, estuaries, extensive sandy beaches, coastal dunes and barrier systems, and a number of nearshore reefs and islands. The coastline has been close to its present position for only the last 7000 years* and is superimposed on the pre-Holocene geological structure and sedimentary environments. The various coastal land-

forms have also developed in association with a range of tidal conditions (micro, meso and macro) and are associated with wave climates varying from high energy on exposed open ocean coasts through to low energy in the protected shorelines of the upper gulfs. In some locations the present shoreline has inherited features of earlier shorelines which occurred at similar elevations. The rapid rise in sea-level between 17 and 7 ka BP brought the sea into contact with a variety of geological structures, rocks and sediments that have had a profound influence on the modern coast. For example, the spectacular cliffed coastline of the Great Australian Bight has developed on level-bedded Tertiary limestones, whereas exposed coasts on the Gawler Craton have produced rocky headlands, Old Hat shore platforms with the stripping of palaeoweathering zones, and embayments containing sandy beaches derived from coastal aeolianite. The technically depressed region of the gulfs in which Quaternary alluvial * Holocene ages used in this paper refer to Radiocarbon Years Before Present (BP)

Late Quaternary coastal geology, SA

203

/MW//SW/

Figure 17.3 Layer-cake stratigraphy developed in the protected gulfs with coastal and peritidal complexes developed at the 'pinch-out' limits of sedimentation.

Figure 17.2 High-energy coastal barrier shorelines preserved as a staircase of raised shoreline deposits along the Mt Gambier Coastal Plain. Numbers 1, 5c, etc. are Oxygen Isotope stages.

deposits are common, is dominantly protected, producing coastlines of aggradation, although locally, erosion may occur. Kangaroo Island and the Mt Lofty Ranges, comprising old, uplifted, resistant rocks typically produce rugged rocky coasts. In some cases, such as Backstairs Passage and Christmas Cove, resurrected Permian glacial topography was drowned by the postglacial sea-level rise, producing very distinctive coastal features. The coastline of the Murray Basin and the Southeast Coastal Plain contains the estuary of the River Murray (Harvey 1996) which is the coastal outlet for Australia's largest, but highly regulated, river system. The estuary is protected from high-energy waves by the extensive dune coastal barrier system of Sir Richard and Younghusband Peninsulas, with the latter enclosing the long narrow, backbarrier lagoonal system of the Coorong. South of the Coorong, rocky shorelines have sporadically developed on Quaternary coastal aeolianite and some older rocks. To emphasise the geological youth of the coastal zone, only some 20 000 years ago, at the height of the last glacial maximum, the shoreline stood between 120 and 130 m below present sea-level; the gulfs and Backstairs Passage were dry land and the South Australian shoreline stood near to the edge of the continental shelf. Another significant point with respect to sea-level change during the Quaternary is that during the last 1.7 million years sea-level has not been markedly higher than at present. For example, the earliest Pleistocene Burnham Limestone (Pillans & Bourman 1996) and its equivalents vary in elevation along the tectonically controlled eastern shore of Gulf St Vincent from -82 m in the Port Adelaide area to +50 m asl at Cape Jervis; Marino (+17 m asl), near Hallett Cove (+30 m asl), Maslins Bay (+ 29 m asl), Port Willunga (+15 m asl), Sellicks (+8 m asl) (Firman 1976; Ludbrook 1983; May & Bourman 1984; Belperio 1995). This elevation distribution reflects warping and faulting resulting from Pleistocene

reactivation of Early Tertiary tectonism (Ludbrook 1983) and none of the levels can be considered to reflect the absolute level of the sea during the earliest Pleistocene. In the succeeding period, up until the Last Interglaciation (132-118 ka BP) (Eisenhauer et al 1996) sea-level did not reach much above its present level. As with the earliest Pleistocene Burnham Limestone, the Last Interglacial shoreline has also been dislocated by earth movements (Murray-Wallace & Belperio 1991), but sedimentological evidence obtained from more stable sections of the South Australian coastline demonstrate that sealevel during the Last Interglacial may have only stood up to 2 m higher than at present.

LATE QUATERNARY COASTAL DEPOSITS IN SOUTH AUSTRALIA For the purposes of this chapter, South Australian Quaternary coastal deposits are discussed according to four broad groupings: (i) the high-energy coastal barriers of the Southeast Coastal Plain; (ii) other coastal deposits; (iii) deposits of the Last Interglacial shoreline as a regional datum for land/sea-level changes; and (iv) some terrestrial equivalents of the Quaternary coastal environments.

High-energy coastal barriers of the Southeast Coastal Plain The stranded barrier shorelines of the Southeast Coastal Plain represent an important global record of successive marine transgressions. Their setting, chronology and interpretation have been documented by Belperio et al (1996), Murray-Wallace et al (1996) and Huntley et al (1993). Sediment facies associated with the coastal barriers are exposed in various drains and cuttings through the ranges (Figure 17.4) and include skeletal calcarenite and marl deposited in shallow subtidal, littoral, dune, lagoon and ephemeral lacustrine environments associated with the former shorelines. The spatial association of facies resembles that of the contemporary Coorong barrier-lagoon system. Transgressive dune or aeolianite facies form the bulk of the visible ridges. They consist of weakly cemented, fine- to medium-grained skeletal sand composed of well-sorted and rounded molluscan, foraminiferal and algal grains.

204

N.Harvey etal.

Metres

Figure 17.4 Internal sediment facies of coastal barriers as revealed in a drainage cut through the Woakwine Range, Mt Gambier Coastal Plain. Reactivation surfaces, marked by weakly developed palaeosols and rhizolith concentrations, are common. These separate cosets of very large-scale cross-beds that dip consistently landward at 3-30°. High-energy littoral sediments occur in the core of the ranges as distinctive bidirectional trough cross-bedded strata. The dune facies toeout in their lee where they interfinger with, or overlie, sediments of the lagoon facies. The backbarrier lagoon facies is particularly significant in the reconstruction of palaeo-sea-levels. These sediments are typically very shelly and muddy calcarenites with numerous gastropods and large articulated intertidal to subtidal bivalves including Anadara, Katelysia and Ostrea. They grade up into lacustrine calcitic and dolomitic mudstones which contain remains of a variety of fresh- to brackishwater molluscs, foraminifers, ostracods and charophytes. The intertidal fauna provide a key datum for measuring shoreline elevation and reconstructing former sea-levels. Since the earliest studies of Sprigg (1952), it has been recognised that there is good agreement between the uplifted coastal plain sequence and de facto global climatic and Milankovitch insolation signatures (Cook et al 1977; Sprigg 1979; Idnurm & Cook 1980; Schwebel 1983). In the absence of more complete chronological data, reasonable age estimates can be made by matching individual barriers to the oxygen isotope record and utilising the indicated palaeomagnetic and geochronological constraints (Belperio 1995). Thermoluminescence dating of the stranded barrier systems is generally in accord with this interpretation (Huntley et al 1993). A relatively complete sequence of stranded high sea-level deposits during the past 800 000 years is suggested, indicating coastal sedimentation in phase with global climatic change. An average rate of uplift of 0.07 mm/y for the past 800 000 years is indicated for the Robe area by the highly coherent plot of shoreline elevation versus inferred age for a Robe-Naracoorte line of section (Figure 17.5). Alongshore spatial variability is demonstrated by mapping shoreline features from north to south along the coastal plain. The Last Interglacial Woakwine Range for example (see Figure 17.1 for location) rises from + 10 m at Robe to +18 m at its southernmost extreme near Mt Gambier. Consistency of uplift over time is demonstrated by coherent plots of shoreline elevation versus age for different sectors of the coastal plain (Belperio 1995).

200

400 600 THOUSAND YEARS BP

800

Figure 17.5 Shoreline elevation vs age for the stranded barriers of the Mt Gambier Coastal Plain on a section line from Robe to Naracoorte using data from Belperio et al (1996) and Huntley et al (1993,1994).

Other coastal deposits Along other high wave energy sections of the South Australian coastline, such as the south coast of Kangaroo Island, western Eyre Peninsula and southern Yorke Peninsula, there has not been the progressive uplift to preserve the alternating sequences of transgression and regression. Instead, as each successive transgression reached the level of the previous one, there was complex erosion, reworking and redeposition of pre-existing beach/dune systems (Milnes & Ludbrook 1986). Miocene marine sediments were noted to be conspicuous amongst the source materials for the aeolianite deposits and Milnes and Ludbrook (1986) suggested that Miocene microfossils had been cycled through various stages of sedimentary and pedological reworking. Complex sequences of palaeosols separate aeolianite packages which probably reflect successive sea-level lowstands, but these have yet to be successfully dated. Within the gulfs, the protected tide-dominated basinal environment has resulted in sea-floor aggradation and coastal progradation as the dominant style of sediment accumulation. The repeated Quaternary transgressions and regressions are therefore preserved in a thick subsea pile of marine sediments with numerous superimposed palaeosols separating successively younger marine units (Hails et al 1984).

Last Interglacial shoreline as a regional datum for relative sea-level changes The Last Interglacial shoreline is the most widely preserved of the Quaternary highstands and allows a continent-wide and regional perspective of sea-level migration to be developed. In South Australia, coastal sediments deposited during the Last Interglacial are referred to as the Glanville Formation. They are characterised by some elements of a warmer water fauna that no longer live in local waters. These include the Sydney blood cockle Anadara trapezia, Shark Bay pearl oyster Pinctada carchariarum, conicalfusiform gastropod Euplica bidentata, and the large foraminifer Marginopora vertebralis (Howchin 1888, 1909, 1912, 1924, 1935; Ludbrook 1976). However, the bulk of the foraminifer-mollusc detritus is similar to that found in

Late Quaternary coastal geology, SA

Figure 17.6 Elevation of the Last Interglacial shoreline in South Australia (after Murray-Wallace & Belperio 1991; Bourman et al 1999). modern peritidal environments or backbarrier environments. Along western Eyre Peninsula, Last Interglacial coastal sediments are well-preserved as high-energy beach-dune aeolianites with low-energy lagoonal to protected intertidal facies present behind the barriers. Former extensive lagoons and sheltered embayment systems, such as at Fowlers Bay and Lake MacDonnell, with their rich cockle, oyster and scallop assemblages, include typical intertidal sand-flat facies at their landward limits. This allows confident estimates of sea-level to be made at numerous sites along some 2000 km of coastline (Figure 17.6). The west coast of Eyre Peninsula consistently indicates a Last Interglacial sea-level 2 m higher than present, coincident with the technically stable Precambrian Gawler Craton. This level is an important datum for coastal deformation studies in Australia (Murray-Wallace & Belperio 1991), being the local benchmark to which other Australian sites are compared. Within and beneath Spencer Gulf and Gulf St Vincent, marine sediments of the Last Interglacial occur extensively as a blanket deposit in the subsurface and subcrop at low elevations around the gulf margins. In northern Gulf St Vincent, the calcreted and karstified surface of the Glanville Formation crops out discontinuously in the contemporary supratidal zone and records a sea-level 3 m above present. In the Port Gawler to Port Adelaide region, the extent and landward limit of the Glanville Formation is largely derived from borehole information and does not outcrop except in excavations. Belperio (1985) ascribed the decrease in elevation of the intertidal facies of the Glanville Formation from Port Wakefield to Port Adelaide to differential subsidence between these. Subsequently, Belperio (1993) was able to quantify significant human-induced subsidence in the Port Adelaide area. Around the coastal margins of the Mt Lofty Ranges, occurrences of the Last Interglacial at Port Stanvac (4 m), Sellicks Beach (4-5 m), Normanville (12 m) and Victor Harbor (6 m) suggest post-Last Interglacial tectonism resulting in differential uplift of the Fleurieu Peninsula, with a maximum in the Normanville area, and submergence in the Adelaide and Murray Lakes areas (Murray-Wallace & Belperio 1991; Bourman et al 1999). The Last Interglacial shoreline on the Southeast Coastal Plain is well-defined, mapped and dated by amino-acid

205

racemisation, U-series and thermoluminescence dating methods (Belperio 1995). The coastal barrier consists predominantly of sand-sized biogenic carbonate with occasional thin conglomerate and shell beds. It interfingers on the lee or landward side with cockle, scallop and oyster-rich lagoonal to intertidal shell beds (Glanville Formation) and dolomitic lacustrine facies. Internal complexities of the Woakwine Range are revealed by good exposures in the walls of deep drains (Figure 17.4), where littoral and backbarrier facies indicate a single major depositional event with a sea-level that fell 10-8 m. In the lower Murray lakes and Coorong area, the Last Interglacial shoreline has now been mapped from Goolwa, across Hindmarsh Island along the site of the barrage system on which it is built, along the landward side of the Coorong lagoon, as well as to the north of Salt Creek. The elevation of this shoreline rises progressively from <2 m AHD on Hindmarsh Island to about 5 m AHD near Salt Creek. At the continental scale, the spatial variability of the elevation of the Last Interglacial shoreline (Murray-Wallace & Belperio 1991) reveals interesting large deformational domains such as a general subsidence of northern Australia, contrasting with relative uplift of southern Australia. At the regional scale (Figure 17.6), more complex regions of uplift are evident within these broader domains, such as regions of neotectonic upwarp of the Southeast Coastal Plain and the Mt Lofty Ranges.

Some terrestrial Quaternary equivalents By investigating river terraces produced by sea-level and climatic changes, it is possible to correlate parts of the Late Quaternary terrestrial record with marine events. Alternating phases of incision and aggradation occur in river valleys of Fleurieu Peninsula and reflect migrations of sealevel. For example, during lower sea-levels prior to Last Interglacial times the Inman and Hindmarsh Rivers excavated valleys up to 20 m below sea-level. During the higher sea-level of the Last Interglacial the valleys became estuaries, which infilled with interfingering terrestrial (Pooraka Formation) and marine (Glanville Formation) sediments, demonstrating their time equivalence at these locations. Subsequent thermoluminescence dating of Pooraka Formation sediments well inland from the coast confirmed their Last Interglacial age (Bourman et al 1997), when the climatic conditions were warmer and wetter than at present. Most high river terraces on Fleurieu Peninsula are developed on valley-fill materials of the Pooraka Formation. Incision into these deposits during glacial and interstadial times produced valleys which extend below present sea-level, and which were subsequently infilled with grey-black sandy clays, dated as mid-Holocene using radiocarbon and amino-acid racemisation dating techniques on in situ estuarine molluscs. These sediments form the lowest terraces and floodplains. During interstadials and glacial low-sea-levels, former coastal and shelf sediments were exposed to remobilisation. Thermoluminescence data derived from aeolian sediments resting above Last Interglacial molluscs suggest that there have been several episodes of aeolian activity, at approximately 30, 50 and 70 ka BP, along the eastern shoreline of Gulf St Vincent during times when sea-level

206

N.Harvey

etal.

YORKE PENINSULA

<f$f ffi/S iff/

Jftzwmu r

pmmmuk/

Sea level rise 24 mm/yr (8000-6700 years BP) KANGAROO ISLAND

-20

i-25

-30

Sea level rise 9 mm/yr (10,000-8000 years BP)

O in _i -35

• 1 4 C date from submarine sediment core -40

was lower than at present (Bourman et al 1999). Much of this dune material originated as Last Interglacial coastal sediments. It was multiply reworked, and contains worn and fragmented bioclastic fragments indicative of its original source. Elsewhere, such as the west coast of Eyre Peninsula and the Southeast Coastal Plain, calcretisation of Quaternary coastal landforms was achieved much more rapidly, thereby limiting the degree of subsequent aeolian reworking. During the last glacial maximum when the shoreline stood near the edge of the continental shelf (20 ka BP), the climate was colder, more arid and windier than at present (Dutkiewicz & Prescott 1997). Extensive continental dunefields formed across the continental shelf, covering far greater areas than the modern inland dunefields on the Eyre Peninsula, Yorke Peninsula and the Murray Basin. Erosion and reworking of these dunes occurred during the post-last glacial rise in sea-level, providing a ready source of sediment for the rapidly transgressing shoreline. Thus multiple reworking of marine sediments has occurred following deposition, firstly as a component of soil development, then in continental aeolian dunefields, and finally their incorporation into subsequent coastal deposits as a relict component.

2 14

Figure 17.7 Approximate position of shoreline advance at different stages during the Holocene postglacial sea-level rise (after Lampert 1981).

J_ 4 6 8 C YEARS BP (x 1000)

Northern Spencer Gulf (4.5m) OE

Port Wakefield (3.0m) Port Gawler (2.1 mK

LUCK:

Port Pine (2.2m) Coffin Bay (0.8m)

-ib lu s <o LU — ' CO I—

o2 LU

Ceduna (0.4m)

Present sea level

<UJ a

4 6 8 ™C YEARS BP (x 1000)

Figure 17.8 Holocene sea-level rise for southern Australia. The enlargement shows that variable but predictable sea-level regression has occurred due to differential isostatic response (after Belperio 1995).

Late Quaternary coastal geology, SA

HOLOCENE SEA-LEVEL RISE AND SHORELINE FLUCTUATIONS Global sea-levels rose rapidly after the last glacial maximum. In areas such as Australia, far away from the effects of continental deglaciation, the seas transgressed the inner continental shelf about 10 to 8 ka BP, reaching the present coastline ca 1 ka BP. The rapidity of this transgression is illustrated for Gulf St Vincent (Figure 17.7). Since then, coastal and shelf processes and coastal sedimentation have progressively stabilised, producing a range of depositional styles that relate increasingly to the contemporary hydrodynamic regime. Hydroisostatic loading of the continental shelf by the rising sea also resulted in different degrees of coastal warping and a geographically variable apparent sea-level regression around the State's coastline over the past 6000 years (Figure 17.8). These relative mid-Holocene highstands reached a maximum of 4.5 m in the northern Spencer Gulf, 3.0 m in northern Gulf St Vincent and 1.0 m along the Southeast and Western Eyre Peninsula coastlines. The gulf and shallow-shelf waters of Southern Australia have an outstanding richness of marine plants. Larger algae and seagrasses provide shelter and a substrate for smaller plants and animals including carbonate-secreting biota such as coralline algae, foraminifers, molluscs and bryozoans. Growth of Posidonia species is particularly vigorous, with high leaf blade production and turnover. Sedimentologically, the seagrass meadows act as major carbonate factories, with high rates of in situ carbonate production and bioclastic skeletal accumulation. Decaying leaf sheaths of P. australis leave behind a residue of fine, pale, lignin-impregnated cellulose fibre. Sediments accumulating in P. australis seagrass meadows are therefore characteristically bound in the subsurface by the fibres and create organically bound carbonate banks. In both Spencer Gulf and Gulf St Vincent, many seagrass banks have accumulated to low-water level, and are now mantled by a variety of intertidal facies. Intertidal environments regularly inundated by neap to spring tides include bare or seagrass-colonised (Zostera) intertidal sand flats, and mangrove-algal and halophyte-algal marshes. Areas of bare, saline flats with isolated stranded beach ridges characterise the supratidal zone. Small beaches with a wide frontal sand flat are developed along shorelines exposed to southwesterly winds and wind-generated waves. There is a well-developed surficial zonation of environments and sediments in the northern gulf. Monospecific stands of the mangrove Avicennia marina form a low woodland from about mean sea-level to spring high-tide level. Progressively higher vegetation zones are a Halosarcia-Sarcocornia samphire marsh, a Sclerostegia-Halosarcia samphire marsh, and an Atriplex sp. saltbush community. Cyanobacterial mats occur throughout the mangrove and halophyte zones, with best growth occurring 2.5-3.5 m above low water. The evaporative supratidal flats are gypsiferous and locally, dolomite nucleates within a fenestral calcitic mud. The surface of the flats represents the upper supratidal limit of sediment accumulation in the prograding tide-dominated coastal environment. Each of the distinctive intertidal facies is related to inundation frequency or tidal elevation and hence can be

207

used to decipher past relative sea-level change. At Port Pirie, Barnett et al (1997) documented specific palaeo-sealevel indicators that could be used in tide-dominated coastal plains of southern Australia and discussed local and regional differences in the reliability of the indicators. Dated material from these horizons have been used to link the neotectonic data to anomalies in the historic tide-gauge data from Port Pirie, which indicates a relative sea-level fall (Harvey et al 1999). Coastal progradation and shoreline regression in the gulfs have been enhanced by a relative sea-level fall of up to 4.5 m over the past 6000 years (Figure 17.8), attributed to hydroisostatic upwarping of the distal parts of the gulf relative to the continental shelf (Nakada & Lambeck 1989). This has resulted in extensive areas of stranded or inactive supratidal flats along the eastern shoreline and stranded shingle and cobble beach ridges along the western shore. Elsewhere, stranded beach ridges relate to the onshore movement of local sediment immediately following the Holocene sea-level transgression. For example, a detailed geochronological study of a Holocene beach ridge system for the Lefevre Peninsula near Port Adelaide (Bowman & Harvey 1986; Harvey & Bowman 1987), demonstrated that the large volumes of sediment being moved onshore in the early Holocene have progressively declined to the extent where beaches have to be artificially supplied by modern management. At Port Augusta, the sediment record beneath extensive intertidal flats contains a detailed chronology of local relative land/sea-level changes (Belperio et al 1984) consistent with the hydroisostatic warping model. Compilation of a number of sea-level curves around the margins of both gulfs suggests maximum warping at the apex of Spencer Gulf of 4-5 m, decreasing down the gulf (Figure 17.8) to 1 m at the entrance at Port Lincoln (Belperio 1995). The observations from field data are consistent with theoretical calculations of Nakada and Lambeck (1989) from rheological modelling of an elastic earth. Contrasting tide gauge anomalies at Port Adelaide, showing a relative sea-level rise, were linked with geological investigations by Belperio (1993) to indicate significant localised land subsidence caused by human activities of port development, wetland reclamation and groundwater extraction. Belperio (1993) concluded that three quarters of the secular sea-level rise signal at the Port Adelaide and Outer Harbor tide gauges could be attributed to ground subsidence.

CURRENT SEA-LEVEL TRENDS, GREENHOUSE ACCELERATED SEA-LEVEL RISE PREDICTIONS AND COASTAL IMPLICATIONS The most recent sea-level rise following the last glacial has had a profound influence on the coast but it is the historic, current and projected rates of sea-level rise that are of significance for coastal managers. Geological studies are useful in detecting local changes in relative land/sea-level, particularly near tide gauges where there may be an influence on the historic record, and also in providing data on coastal erosion and general coastal response scenarios for sea-level rise predictions associated with global warming.

208

N.Harvey

etal.

Table 1 7 . 1 Sea-level trends calculated from South Australian tide gauge data. Years of

Sea-level

Isostatic adjustment

Anthropogenic

Adjusted sea-level

record

trend (mm/y)

(mm/y)

(highstands)

effects (mm/y)

trend (mm/y)

Thevenard

30.0 b

0.44^

0.06

(0.4 m)

assumed 0

Port Pirie

58.3 54.3^

-0.02 2.25^

0.33

(2.2 m)

assumed 0

0.31

0.35

(2.1 m)

- 2 . 2 subsidencec

0.40

Tide gauge

Outer Harbor

0.50

Inner Harbor

41.lb

2.06k

0.35

(2.1 m)

- 1 . 8 subsidencec

0.61

Port Lincoln

31.4k

0.85k

0.13

(0.8 m)

assumed 0

0.98

Victor Harbor

26.7

0.58

0.17

(1.0 m)

assumed 0

0.75

Port MacDonnell

21.7d

0.49d

0.14

(0.84 m)

silting of gauge

<0.63

National Tidal Facility records up to 1996

National Tidal Facility records up to 1994

Belperio (1993)

National Tidal Facility records, gauge no longer operable

South Australian coast: secular sea-level rise Records of sea-level in Australia, from the period 1897 to present, have been monitored and analysed by the National Tidal Facility at Flinders University in South Australia. There is significant spatial inhomogeneity in the secular sea-level trends resulting from the myriad of factors affecting relative sea-level behaviour at each tide-gauge site. Although there is a variation in the quality of tidal records available, the National Tidal Facility analysis of tidal data from gauges with an acceptable datum stability indicate an Australian average rate of rise of 0.68 ±0.08 mm y 1 at the 95% confidence level. This figure is lower than, but not inconsistent with global rates of 1.0-2.0 mm y 1 (Gornitz 1993) or the Intergovernmental Panel on Climate Change best estimates for global sea-level rise of 1-2.5 mm y 1 , notwithstanding the caution expressed about the validity of some of these figures (Aubrey & Emery 1993). South Australian sea-level trends based on tide-gauge data have been presented by Mitchell (1991) (Table 17.1). As noted by Harvey and Belperio (1994), Australia is frequently, but incorrectly, quoted as a stable continent from which absolute sea-levels may be measured. Neotectonic movements caused by structural, geoidal and isostatic processes, together with factors such as sediment compaction, affect Australia at various spatial and temporal scales. This complexity of underlying factors that control relative sea-level change has been demonstrated by the reconstruction of palaeo-sea-level histories from numerous sites around Australia. The sea reached its present level around the South Australian coast between 7 and 6 ka BP. South Australia is far away from the immediate effects of global deglaciation, but is affected by subtle, ongoing isostatic adjustment of shelf and coast. This is manifested as an relative highstand of the 6 ka BP shoreline, the height of which varies systematically and predictably around the coast (Figure 17.8). In particular, the elevation of the highstand increases up the two gulfs with increasing distance from the continental margin. Sea-level change over the past few thousand years has been dominated by this regression, which increases in magnitude from 1.0 m or less along Eyre Peninsula, to 3.0 m at the head of Gulf St Vincent and 4.5 m in Upper Spencer Gulf. Isostatic adjustments thus vary from 0.1 mm y 1 to 0.8 mm y 1 averaged over these time-scales. This has

caused slow but obvious coastal regression, particularly at the heads of both gulfs. Geological studies confirm Holocene variations in relative land/sea-level, which need to be taken into account by coastal managers. For example, a study of sea-level indicators in a mesotidal coastal environment at Port Pirie, South Australia, derived a long-term rate of sea-level fall of 0.33 mm y 1 from a Holocene highstand at 2.2 m 6700 years ago to the present, in response to isostatic adjustment of the continental shelf and coast (Harvey et al 1999). Adjustments with present sea-level trends, derived from adjacent tidal data, reveal that present sea-level, when corrected for Holocene relative land/sea movement, is rising at a rate of around 0.31 mm y 1 , less than the global sea-level rise estimates of 1-2.5 mm y 1 . Superimposed on geographically variable Holocene isostatic warping are longer term tectonic movements. Tectonic effects are most noticeable along the South East coastal plain, between Lake Alexandrina and Mt Gambier, where Quaternary volcanism has resulted in ongoing uplift and tilting of the coastal plain. The scale and variability of this upwarp can be illustrated by the changing elevation of the Last Interglacial shoreline (Figure 17.6). This 125 000 year old shoreline rises progressively southwards, from 4 m above present sea-level at Salt Creek, to in excess of 18 m near Port MacDonnell. Uplift rates in the Port MacDonnell region are 0.2 mm y 1 if averaged out over this entire time period. One of the most underrated effects associated with cities is sediment compaction and land subsidence associated with coastal reclamation and withdrawal of underground water (Bird 1993). Such effects are local, but are sufficiently frequently associated with harbours and tide-gauge sites to raise serious questions on the validity of global averages obtained from their secular trends (Davis 1987; Pirazzoli 1986). The local record from Port Adelaide clearly illustrates these effects and the inherent danger of using tide gauge data without adequate neotectonic correction (Belperio 1989; 1993). Data from tide gauges, mangrove migration patterns and from dated subsurface strata all indicate a contemporary relative rise in sea-level in the Port Adelaide estuary. The geographic restriction of these effects to the Port Adelaide region, together with preliminary geodetic evidence, indicate that the apparent rise in sea-level is a local phenomenon, resulting chiefly from land subsidence. Belperio (1993) concluded that up to 1.0 m of sub-

Late Quaternary coastal geology, SA

sidence had occurred associated with wetland reclamation, increasing urban and industrial utilisation and groundwater withdrawal Highly variable rates of land subsidence, between 1.8 and 10 mm y~l , were estimated to be occurring over different parts of this region. More significantly, some three-quarters of the secular rise of sea-level indicated by the Port Adelaide and Outer Harbor tide-gauge data could be attributed to land subsidence over the last 50 years at this location. The tide-gauge data from Port Adelaide and Outer Harbor have been used in calculations of global and Australian sea-level rise averages without adequate local neotectonic correction. Many of the world's tide gauges are similarly biased by land subsidence effects, indicating the importance of making such local neotectonic corrections to all tide-gauge data before inferring local or global sea-level changes. Estimates of global sea-level rise using tide-gauge data are often dominated by neotectonic and anthropogenic effects resulting in an overestimation of global sealevel rise by two to three times (Pirazzoli 1989). Greenhouse-accelerated sea-level rise predictions The recent Greenhouse debate has focused attention on the effects of contemporary climatic change on sea-level (Warrick et al 1993) and also the effects of a rising sealevel on coastal environments (Bird 1993; Tooley & Jelgersma 1992). In addition, the work of the Intergovernmental Panel on Climate Change (IPCC) has produced scientific assessments of climatic change (Houghton etal 1991, 1992, 1996). Before discussing the implications of this recent research for the South Australian coast, it is important to note that there is a great variation in physical and geological processes which are responsible for sea-level change. These variations, which have been categorised by Pugh (1993) in terms of their spatial and temporal influence, can vary considerably from short term wind waves with periods of about 10 s and an extent of tens of metres, up to global changes in sea-level related to sea-floor spreading with time periods of hundreds of millions of years. In addition, the process of redistribution of mass over the earth resulting from deglaciation, addition of meltwater to the oceans and transgression and regression over the continental shelves, itself results in a variety of isostatic responses of the crust to the changing loads. The resultant lack of uniformity in global sea-level change is often overlooked or underestimated. A major problem related to the identification of the current rate of eustatic sea-level change from tide-gauge data are the influences of neotectonic, isostatic, and anthropogenic effects (Warrick 1993; Gornitz 1993; Aubrey & Emery 1993; Woodworth 1993) as well as larger scale crustal-induced vertical land movements affecting relative sea-level change (Houghton et al 1996). Other difficulties in determining accurate estimates for sea-level trends result from the unequal geographical distribution of historical tide-gauge data (Houghton et al 1996; Warrick 1993; Gornitz 1993; Aubrey & Emery 1993; Woodworth 1993). These problems create uncertainty in extrapolating the eustatic component of sea-level change and has led authors such as Aubrey and Emery (1993) to express caution in attempting to extrapolate actual sea-level changes from the

209

data. They suggested that the apparent post-1930 accelerated sea-level rise may be related to factors other than human-induced factors. For example, it may relate to a delayed response to climatic warming following the Little Ice Age or oceanographic factors; or the accelerated rise may not even be statistically significant. Other authors (see Douglas 1991), however, have attempted to determine these changes by methods such as accounting for 'post-glacial rebound' in their estimates. Gornitz (1993) has also suggested that after extraction of long-term trends and data averaging, it is possible to obtain a true picture of sea-level rise. Gornitz (1993) presented evidence based on 16 tidegauge data studies suggesting that estimates of global sealevel rise over the last 100 years have been between 0.5 and 3 mm y~l, with most estimates in the range of 1-2 mm y 1 . The best estimate based on recent analyses is that over the last 100 years sea-level has risen by about 18 cm with a range of uncertainty of 10-25 cm (Houghton et al 1996) although some authors (Pirazzoli 1993) do not believe that a global figure for sea-level rise can be accurately estimated. Houghton et al (1996) also concluded that there is no evidence for any acceleration of sea-level rise this century. Sea-level rise over the last 100 years has had major contributions from both thermal expansion and glacial melt but according to Houghton et al (1996) there are many uncertainties regarding the role of ice sheets and other hydrological factors. In addition to studies attempting to identify eustatic sealevel changes based on analysis of tide-gauge data, there are also a number of studies on sea-level rise projections related to climate change. The key study upon which the South Australian coastal policy has been based, was the earlier IPCC sea-level rise predictions with a best estimate of a 0.65 m rise to the year 2100 (Houghton et al 1991). This 1991 IPCC report provided a significant downwards revision of earlier sea-level rise predictions but more recent calculations have continued to produce similar best-estimate figures, either by more qualitative expert analysis (0.61 m by the year 2087: Woodworth 1993), or by detailed recalculation (0.46 m by the year 2100: Wigley & Raper 1993). Since the 1991 IPCC report there have been many estimates of both the degree of sea-level rise over the last 100 years (and whether it has in fact been accelerating) as well as the predicted future sea-level rise. In Houghton et al's (1996) opinion the major conclusions reached in the 1991 IPCC report remain qualitatively unchanged. Despite this, all subsequent estimates of predicted sea-level rise are significantly lower than that estimated by the 1991 IPCC report, although Houghton et al (1996) warned that direct comparisons can not be made due to differences in factors such as emission scenarios and radiative forcing changes. The best estimate of Houghton et al (1996) is that sealevel will rise by 49 cm by the year 2100, with a range of uncertainty of 20-86 cm. This projection is lower than that presented by the 1991 IPCC report due primarily to the lower temperature projections, the inclusion of a slowdown of the thermohaline circulation and changes to the glacier model (Houghton et al 1996). However, the authors stressed that the understanding of climate-sea-level relationships has not changed. Hence, if future temperature change is higher than expected, sea-level rise will also be higher.

210

N. Harvey

etal.

Vulnerability of the South Australian coast and government response to sea-level rise

The immediate impact of any sea-level rise will be to increase the magnitude and frequency of extreme tides and levels of storm erosion, although the effects of this will vary greatly around the coast. The least vulnerable areas will be the resistant rocky coasts of the Fleurieu Peninsula, Kangaroo Island and Eyre Peninsula but there is likely to be greater erosion on the more predominant softer aeolianite and Tertiary limestone rocky coasts. The actual rate of cliff or bluff retreat on these coasts will vary with factors such as rock resistance, structure, the presence or absence of shore platforms or nearshore reefs, exposure to wave action and tidal range (Bird 1993). The sandy coasts, which represent about half (1900 km) of the South Australian coast, are likely to have a variable response because of differing rates of littoral drift, onshore-offshore sediment movement and sediment size variability. The high-energy beaches of the Younghusband and Sir Richard Peninsulas, for example, are backed by an extensive dune coastal-barrier system which would be vulnerable to increased storm attack with elevated sea-levels. Coastline retreat and the development of dune blowouts are likely to cause a migration of the barrier towards the Coorong. As this occurs underlying calcrete and backbarrier muds would become exposed causing variable rates of retreat. In addition, raised water levels would impact on the Coorong. In contrast, the metropolitan sandy coast lacks the extensive backing dune barriers of the South East. Urban encroachment across the frontal dune, together with extensive protective works, have necessitated the establishment of a sand-replenishment programme to maintain the beaches. Elevated sea-levels are likely to have greatest financial impact in this area where storm protection will need to be upgraded, together with an increased sandreplenishment program, if the metropolitan beaches are to be maintained. To the north of metropolitan Adelaide, land subsidence has already been noted for the Port Adelaide area. Elevated sea-levels will exacerbate the rate of relative sea-level rise causing mangroves to advance further inland and induce changes to ecological zonations of the intertidal and supratidal biota. In some places mangrove advance may be restricted by artificial embankments resulting in mangrove dieback. In the gulf regions, similar displacement of ecological communities such as seagrasses, mangroves and samphires would be pronounced along the low-gradient coasts. This rapid coastal retreat would be associated with reactivation of tidal swamps, localised flooding of the coastal plains and erosion of beach-ridge systems. In other parts of the coast, an elevated sea-level is likely to cause flooding of low-lying land, enlargement of coastal lakes and/or connection of some lakes to the sea, raised groundwater levels and alteration to estuarine environments. South Australia has few estuaries although there could be major implications for urban development adjacent to estuaries such as the Onkaparinga and the Port Rivers. In the case of the Port River estuary, approximately 25% of the nearby urban development is currently below high water mark. Elsewhere, potential impacts on the River

Murray residual estuary (artificially constrained by the construction of barrages) may be less significant for urban areas but could have major implications for the operating levels of the barrages and affect the management of the Murray Mouth and lower Murray Lakes region (Harvey 1988). The South Australian Government responded to the threat of sea-level rise by endorsing (in May 1991) a Coast Protection Board policy on coast protection and new coastal development. The policy, which is described in detail elsewhere (Coast Protection Board 1992), relies in part on local records of coastal erosion, flooding and sealevel rise but more importantly has incorporated the earlier 1991 IPCC estimates of greenhouse-induced eustatic sealevel rise. These estimates predict a sea-level rise to the year 2100 of approximately 0.65 m (range 0.33-1.10 m) for a 'business as usual' scenario (Houghton et al 1991). Given these estimates the Coast Protection Board used the 'precautionary principle' in preparing its policy. The precautionary principle, which was adopted by all Australian governments, states that where there are threats of serious or irreversible environmental damage, lack of full scientific certainty should not be used as a reason for postponing measures to prevent environmental degradation (Anon 1992, para 3.5.1). In accordance with this principle the Coast Protection Board has adopted the policy that any new coastal development should be capable of being reasonably protected from a i m sea-level rise by the year 2100. The policy establishes the 100 year average return interval (ARI) water level as a standard for coastal development in South Australia. It recommends that site and building levels should be determined by adding 0.3 m to the 100 year ARI water level and, where appropriate, making an adjustment for localised subsidence or uplift. Floor levels of buildings should be an additional 0.25 m above this level, and buildings should not be approved unless they are capable of being protected or raised to withstand a further 0.7 m of sea-level rise (e.g. by means of a bund wall). In the case of flood-protected sites, the calculation of the 100 year ARI design flood level must incorporate the extreme tide (plus surge) and stormwater events, together with wave effects, in the development. The policy also makes a general recommendation for an erosion setback distance. This is to be based upon 100 years of erosion at a site, allowing for local coastal processes and a sea-level rise of 0.3 m to the year 2050, and taking account of storm erosion from a series of severe storms. For major coastal development it is suggested that calculations are based upon 200 years of erosion. The policy is less specific about the protection of existing property; it reaffirms an earlier government policy not to protect private property. Although part of the Coast Protection Board's duties are to protect the coast, most coast protection works are carried out by local councils on a cost sharing basis between State and local government. The underlying question for any protection works for existing property is essentially a decision whether to protect or relocate, which is complicated by the level of public or private involvement and the relative responsibilities of State and local governments.

Late Quaternary coastal geology, SA

RELEVANCE OF COASTAL GEOLOGY AND COASTAL PALAEOENVIRONMENTAL RECONSTRUCTION FOR MANAGEMENT PURPOSES Initial estimates of the current rate of sea-level rise obtained by averaging tide-gauge data from around the world have been reducing. The principle reason for this reduction is the recognition that land level changes need to be removed from the tide-gauge data before they can be used for this purpose. Underlying, unrecognised, neotectonic effects remain the main reason for the geographic variability in secular tide-gauge trends. A global sea-level rise signal cannot be expected to be detected until adequate corrections are made for these effects at each tidegauge site. With consensus estimates of predicted sea-level rise now down to 0.46 m over the next 100 years, neotectonic effects will be determining factors in overall local sealevel behaviour. In the gulfs, any rise of sea-level will be mitigated by ongoing isostatic upwarp of up to 0.8 mm y - 1 . In the South East, tectonic uplift will also be a mitigating circumstance. In major towns, particularly where wetland reclamation has occurred, or excessive groundwater withdrawal is taking place, any accelerated greenhouse sealevel rise will be exacerbated by land-subsidence effects. It is also important to note that the sea-level rise predictions and the rate of rise have to be considered in the geological context. This shows that although sea-level has not been more than 2 m higher than present during the last 1.7 million years, sea-level has at times risen at a faster rate (14 mm y^1) than some of the current human-enhanced greenhouse-related sea-level rise predictions. In the absence of accurate geodetic or altimetric data, the geologic record can be used to obtain first-order estimates of land-level changes. In South Australia, the coastal record has produced some useful results that go some way towards explaining the variability of sea-level change documented by tide gauges. However, there remains a pressing need for local and global crustal-scale geodetic control of all tide-gauge sites. In addition, a geographically suitable spread of tide-gauge sites is required to account for the scale and variability of neotectonic processes. In South Australia, these tide-gauge sites are required within the gulfs as well as along the oceanic coastline and nearshore islands. The implications of this for South Australia's coastal policy are twofold. First, it highlights the need for accurate local sea-level data to be corrected for neotectonic and anthropogenic factors. Second, this South Australian case study provides support for Pirazzoli's concerns that the global averages for current eustatic sea-level rise may also be overestimated. The compounding effect of uncorrected global sea-level rise trends added to uncorrected local trends will produce inaccuracies in the sea-level trend data upon which the policies are based. In addition, the uncertainties surrounding climatechange predictions and the associated sea-level response necessitate the adoption of the precautionary principle allowing safety margins for building levels and erosion setbacks. Although this may have major cost implications for coastal development, it is unlikely that a greater precision for climate-change models and sea-level response will be reached in the near future. However, it is possible to reduce

211

some of the uncertainties of current sea-level measurements by correcting sea-level data for neotectonic and anthropogenic influences.

CONCLUSIONS Sediments of the coastal zone contain records of past sealevels, climates and rates of change over time-scales of hundreds to millions of years. Understanding and quantifying these changes are important for understanding present patterns of sea-level behaviour, as displayed by tide gauges, and predictions of sea-level rise and inundation under various 'greenhouse' scenarios. From age and height attributes of palaeoshorelines, it is possible to determine how much the sea-level has risen or fallen relative to the adjacent land surface, and how much of this is due to global climatic change, crustal deformation or local subsidence of the land surface. Neotectonic deformation, calculated from the geological record, and human-induced land-level changes derived from geodetic surveys, are necessary correction factors for the reliable interpretation of secular sea-level trends from tide gauges (Belperio 1989, 1993; Pirazzoli 1986, 1989). The Last Interglacial shoreline of 125 000 years ago provides a particularly useful, laterally persistent shoreline datum from which the magnitude and spatial variability of coastal uplift or subsidence can be calculated. Around South Australia, this shoreline occurs in places at or below present sea-level, rising to extremes of up to 18 m above present sea-level at other sites. Critical to the utility of this evidence is the correct and accurate determination of a palaeo-sea-level datum from evidence preserved in the sediments. The most recent time when climatic conditions and sealevel were similar to the present, was during the Last Interglacial period at around 125 ka BP when sea-level in the Australian region was between 2 and 8 m higher than today (Chappell 1987). Since then, sea-level fluctuations have always been lower than present with evidence from the Australian region of a low sea-level of between 130 and 165 m lower (Chappell 1987) at 18 ka BP, after which it rose at a rate of between 6 and 12 mm y - 1 prior to reaching its present level between 7 and 6 ka BP. The Holocene coastal record contains a somewhat higher resolution record of relative land/sea-level change over the past 6000-8000 years compared to the Last Interglacial. In South Australia, two styles of sedimentation dominate the preserved Quaternary record and the modern or Holocene coastal facies: (i) high-energy, swell-dominated barrier/lagoon settings; and (ii) protected, low-gradient, prograding intertidal environments. Both settings generate key sediment facies that contain palaeo-sea-level information as discussed further below. A third style of sedimentation that preserves information about the relative land/sea relationships in the Holocene involves estuarine environments such as terminal lakes of the River Murray and narrower estuaries such as those of the Onkaparinga, Inman and Hindmarsh Rivers of Fleurieu Peninsula. It is in the record of the last 6000 years that scientists have focused attention on finding evidence for any longterm trends in sea-level either by direct sea-level indicators

212

N.Harvey

etal.

in the geological record or by analogous palaeoclimatic change evidence. Careful field studies from many coastal localities have supported geophysical models which indicate that subtle differences in sea-level behaviour are the norm even for the South Australian coast, long regarded as stable and uniform (Lambeck & Nakada 1990). There has also been direct measurement of sea-level and detailed analysis of recent tide-gauge records to extrapolate relative trends (Gornitz 1993; Pirazzoli 1991). As noted by Bird (1993) there are numerous factors affecting relative sea-level change. Apart from eustatic sea-level change there is the tectonic response of the land and the isostatic response of the continental margins relating to changing volumes of ice, water or sediment. In addition, human activities such as groundwater or hydrocarbon extraction, land reclamation, artificial coastal structures, dredging and pumping of sediment can affect local sea-level change. Pirazzoli (1989) suggested that local secular tide-gauge data are dominated by neotectonic and anthropogenic effects, resulting in an overestimation of global sea-level rise by two to three times when these factors are ignored. The data presented in this chapter demonstrate the importance of Quaternary sediments in shaping the present coastline. These sediments also provide valuable data on broader Quaternary climatic change and sea-level variations such as in the South East where a long-term coastal sedimentary record has been preserved across a steadily uplifted landscape. In addition, extensive Holocene sedimentation in the gulfs provides detailed data on the rate of the postglacial sea-level rise and demonstrates the necessity for taking neotectonic and anthropogenic factors into account for calculating current rates of sea-level rise. Thus the Quaternary coastal sediments not only hold the key to understanding the evolution of most of the South Australian coast but they also provide useful pointers for interpreting historic data with direct relevance for current issues such as sea-level rise predictions.

ACKNOWLEDGEMENT W e acknowledge funding support from the National Greenhouse Advisory Committee, the Australian Research Committee, Mines and Energy South Australia and the University of South Australia. W e also wish to thank Kris James for her assistance in editing the manuscript and Colin Murray-Wallace for some useful comments on an earlier version of this chapter.

REFERENCES 1992. Intergovernmental Agreement on the Environment. Australian Government Publishing Service, Canberra.

ANON

AUBREY D. G. & EMERY K. O. 1993. Recent global sea levels and land

levels. In: Warrick R. A., Barrow E. M. & Wigley T. M. L. eds. Climate and Sea Level Change: Observations, Projections and Implications, pp. 45-56. Cambridge University Press, Cambridge.

BARNETT E . J . , HARVEY N . , BELPERIO A . P. & BOURMAN R . P. 1 9 9 7 . S e a -

level indicators from a Holocene, tide-dominated coastal succession, Port Pirie, South Australia. Transactions of the Royal Society of South Australia 121, 125-135. BELPERIO A. P. 1985. Quaternary geology of the Sandy Point and Outer Harbor-St. Kilda areas, Gulf St. Vincent. South Australia. Geological Survey of South Australia Quarterly Geological Notes

65, 2-4.

BELPERIO A. P. 1989. The Greenhouse debate and rising sea levels:

geological and other factors controlling sea level change. In: Dendy T. ed. Greenhouse '88: Planning for Climate Change, pp. 77-80. Department of Environment and Planning, South Australia, Adelaide. BELPERIO A. P. 1993. Land subsidence and sea level rise in the Port Adelaide estuary: implications for monitoring the Greenhouse Effect. Australian Journal of Earth Sciences 40, 359-368. BELPERIO A. P. 1 9 9 5 . The Quaternary. In: Drexel J . F. & Preiss W . V. eds. The Geology of South Australia, Vol. 2 , pp. 2 1 8 - 2 8 1 . Geological Survey of South Australia Bulletin 54. BELPERIO A. P., CANN J . H . & MURRAY-WALLACE C . V. 1 9 9 6 . Quaternary evolution, sea level change and neotectonics: the Coorong to Mount Gambier Coastal Plain. An Excursion Guide. IGCP Project 367 Excursion Guide, 3rd annual meeting, Sydney, November 1996. BELPERIO A . P., GOSTIN V. A . , CANN J . H . & MURRAY-WALLACE C . V.

1988. Sediment-organism zonation and the evolution of Holocene tidal sequences in Southern Australia. In: De Boer P. L., van Gelder A. & Nio S. D. eds. Tide-influenced sedimentary environments and fades, pp. 475-497. Reidel Publishing Co., Dordrecht. BELPERIO A . P., SMITH B . W . , POLACH H . A . ETAL. 1 9 8 4 . Chronological studies of Quaternary marine sediments of northern Spencer Gulf. Marine Geology 6 1 , 2 6 5 - 2 9 6 . BIRD E. C. F. 1993. Submerging Coasts: The Effects of a Rising Sea Level on Coastal Environments. John Wiley and Sons, Chichester. BOURMAN R . P., BELPERIO A . P., MURRAY-WALLACE C . V. & CANN J. H . A

1999. last interglacial embayment fill at Normanville, South Australia, and its neotectonic implications. Transactions of the Royal Society of South Australia. 123, 1-15.

BOURMAN R . P., MARTINAITIS P., PRESCOTT J . R . & BELPERIO A . P. 1 9 9 7 .

The age of the Pooraka Formation and its implications, with some preliminary results from luminescence dating. Transactions of the Royal Society of South Australia 1 2 1 , 8 3 - 9 4 . BOWMAN G. M . & HARVEY N . 1 9 8 6 . Geomorphic evolution of a Holocene beach ridge complex, LeFevre Peninsula, South Australia. Journal of Coastal Research 2 , 3 4 5 - 3 6 2 . CHAPPELL J. 1987. Late Quaternary sea-level changes in the Australian Region. In: Tooley M. J. & Shennan I. eds. Sea-Level Changes, pp. 296-331. Blackwell, New York. COOK P. J . , COLWELL J . B . , FIRMAN J . B . , LINDSAY J . M . , SCHWEBEL D . A

& VON DER BORCH C. C. 1977. Late Cainozoic sequences of southeast South Australia and Pleistocene sea level changes. BMR Journal of Australian Geology & Geophysics 12, 81-88. COASTAL PROTECTION BOARD 1992. Coastal Erosion, Flooding and Sea Level Rise Standards and Protection Policy. Coastline 26, 1-6. DAVIS G . H . 1 9 8 7 . Land subsidence and sea level rise on the Atlantic Coast Plain. Environmental Geology and Water Science 10, 6 7 - 8 0 . DOUGLAS B . C . 1 9 9 1 . Global sea level rise. Journal of Geophysical Research 9 6 ( C 4 ) , 6 9 8 1 - 6 9 9 2 . DUTKIEWICZ A. & PRESCOTT J . R . 1 9 9 7 . Thermoluminescence ages and palaeoclimate from the Lake Malata-Lake Greenly Complex, Eyre Peninsula, South Australia. Quaternary Science Reviews (Quaternary Geochronology) 16, 3 6 7 - 3 8 5 . EISENHAUER A . , ZHU Z . R . , COLLINS L. B . , WYROLL K - H . & EICHSTATTER

R. 1996. The last interglacial sea level change: new evidence from the Abrolhos islands, West Australia. Geologische Rundschau 85, 606-614. EMILIANI C. 1955. Pleistocene temperatures. Journal of Geology 63, 539-578. FIRMAN J. B. 1976. Limestone at the base of the Pleistocene sequence in South Australia. Geological Survey of South Australia Quarterly Geological Notes 58, 2-5. GORNITZ V. 1993. Mean sea level changes in the recent past. In: Warrick R. A., Barrow E. M. & Wigley T. M. L. eds. Climate and Sea Level Change: Observations, Projections and Implications, pp. 25-44. Cambridge University Press, Cambridge. HAILS J . R . , BELPERIO A . P., GOSTIN V. A . & SARGENT G . E . G . 1 9 8 4 . T h e

submarine Quaternary stratigraphy of northern Spencer Gulf, South Australia. Marine Geology 61, 3 4 5 - 3 7 2 . 1988. Coastal management issues for the mouth of the River Murray, South Australia. Coastal Management 16, 139-149. HARVEY N. 1996. The significance of coastal processes for the management of the River Murray Estuary. Australian Geographical Studies 34, 45-57. HARVEY N.

Late Quaternary coastal geology, SA HARVEY N . , BARNETT E . J . , BOURMAN R . P. & BELPERIO A . P.

1999.

Holocene sea-level change at Port Pirie, South Australia: a contribution to global sea-level rise estimates from tide gauges. Journal of Coastal Research 15, 607-615. HARVEY N. & BELPERIO A. P. 1994. Implications of climatic change for the South Australian coast. Transactions of the Royal Society of South Australia 118, 45-52. HARVEY N . & BOWMAN G . M . 1 9 8 7 . Coastal management implications of a Holocene sediment budget : Le Fevre Peninsula, South Australia. Journal of Shoreline Management 3, 77-93. HOUGHTON J . T . , CALLENDER B . A . & VARNEY S . K . (Editors). 1 9 9 2 . Climate Change 1992. The Supplementary Report to the IPCC Scientific Assessment Published for the IPCC by Cambridge University Press, Cambridge. HOUGHTON J . T., JENKINS G. J . & EPHRAUMS J . J . (Editors) 1991. Climate Change. The IPCC Scientific Assessment. Published for the IPCC by Cambridge University Press, Cambridge. HOUGHTON J . T . , MEIRA FILHO L. G . , CALLANDER B . A . , HARRIS N . ,

(Editors) 1996. Climate Change 1995: The Science of Climate Change. Published for the IPCC by Cambridge University Press, Cambridge. HOWCHIN W. 1 8 8 8 . Remarks on a geological section at the new graving dock, Glanville, with special reference to a supposed old land surface now below sea level. Transactions of the Royal Society of South Australia 1 0 , 3 1 - 3 5 . HOWCHIN W. 1909. Notes on the discovery of a large mass of living coral in Gulf St Vincent, with bibliographic references to the recent corals of South Australia. Transactions of the Royal Society of South Australia 33, 242-252. HOWCHIN W. 1912. Notes on recurrent transgressions of the sea at Dry Creek. Transactions of the Royal Society of South Australia 36, 34-39. HOWCHIN W . 1 9 2 4 . The Recent extinction of certain marine animals of the southern coast of Australia, together with other facts that are suggestive of a change in climate. Australasian Association for the Advancement of Science Report 1 6 , 9 4 - 1 0 1 . HOWCHIN W. 1935. Notes on the geological sections obtained by several borings situated on the plain between Adelaide and Gulf St Vincent. Transactions of the Royal Society of South Australia 59, KATTENBURG A . & MASKELL K .

68-102.

]. T. & PRESCOTT J . R . 1 9 9 3 . The stranded beach-dune sequence of south-east South Australia: a test of thermoluminescence dating, 0-800 ka. Quaternary Science Reviews 12, 1-20. HUNTLEY D . J . , HUTTON J . T. & PRESCOTT J . R . 1 9 9 4 . Further thermoluminescence dates for the dune sequence in the south-east of South Australia. Quaternary Science Reviews 1 3 , 2 0 1 - 2 0 7 . IDNURM M. & COOK P. J. 1980. Palaeomagnetism of beach ridges in South Australia and the Milankovitch theory of ice ages. Nature 286, 699-702. LAMBECK K. & NAKADA M. 1990. Late Pleistocene and Holocene sealevel change along the Australian coast. Palaeogeography, Palaeoclimatology, Palaeoecology (Global and Planetary Change Section) 89, 143-176. LAMPERT R. 1981. The Great Kartan Mystery. Terra Australis 5. Department of Prehistory, Research School of Pacific Studies, Australian National University, Canberra. LUDBROOK N . H . 1 9 7 6 . The Glanville Formation at Port Adelaide. South Australia. Geological Survey of South Australia Quarterly Geological Notes 5 7 , 4-7. LUDBROOK N . H . 1 9 8 3 . Molluscan faunas of the Early Pleistocene Point Ellen Formation and Burnham Limestone, South Australia. Transactions of the Royal Society of South Australia 107, 3 7 - 4 9 . MAY R. M . & BOURMAN R. P. 1984. Coastal landslumping in Pleistocene sediments at Sellicks Beach, South Australia. Transactions of the Royal Society of South Australia 108, 85-94. MILNES A. R. & LUDBROOK N. H. 1986. Provenance of microfossils in HUNTLEY D . J . , HUTTON

213

aeolian calcarenites and calcretes in southern South Australia. Australian Journal of Earth Sciences 33, 145-159. MITCHELL B. 1991. Sea level and climate change. In: Proceedings of the 2nd Australasian Hydrographic Symposium, pp. 353-360. MURRAY-WALLACE C . V. & BELPERIO A. P. 1 9 9 1 . The last interglacial shoreline in Australia: a review. Quaternary Science Reviews 10, 441-461. MURRAY-WALLACE C . V., BELPERIO A . P., CANN ]. H . , HUNTLEY D . J . &

PRESCOTT J . R . 1 9 9 6 . Late Quaternary uplift history, Mount Gambier region, South Australia. Zeitschrift fur Geomorphologie 106, 4 1 - 5 6 .

NAKADA M. & LAMBECK K. 1989. Late Pleistocene and Holocene sea-

level change in the Australian region and mantle rheology. Geophysical Journal 96, 497-517. PILLANS B. & BOURMAN R. P. 1996. The Brunhes/Matuyama polarity transition (0.78 ka) as a chronostratigraphic marker in Australian regolith studies. AGSO Journal of Australian Geology & Geophysics 16, 289-294. PIRAZZOLI P. A. 1986. Secular trends of relative sea level (RSL) changes indicated bt tide-guage records. Journal of Coastal Research 1, 1-26.

PIRAZZOLI P. A. 1989. Present and near-future global sea level change.

Palaeogeography, Palaeoclimatology, Palaeoecology (Global and Planetary Change) 75, 241-258. PIRAZZOLI P. A. 1991. Possible defences against a sea level rise in the Venice area, Italy. Journal of Coastal Research 7, 231-248. PIRAZZOLI P. A. 1993. Global sea-level changes and their measurement. Global and Planetary Change 8, 135-148. PUGH D . T. 1 9 9 3 . Improving sea level data. In: Warrick R . A . , Barrow E. M. & Wigley T. M. L. eds. Climate and Sea Level Change: Observations, Projections and Implications, pp. 5 7 - 7 1 . Cambridge University Press, Cambridge. SCHWEBEL D. A. 1983. Quaternary dune systems. In: Tyler M. J., Twidale C. R., Ling J. K. & Holmes J. W. eds. Natural History of the South East, pp. 15-24. Royal Society of South Australia Occasional Publications 3. SPRIGG R. C. 1979. Stranded and submerged sea-beach systems of Southeast South Australia and the aeolian desert cycles. Sedimentary Geology 22, 53-96. TOOLEY M . J . & JELGERSMA S . 1 9 9 2 . Impacts of Sea-Level rise on European Coastal Lowland. Blackwell Publishers, Oxford. WARRICK R. A. 1993. Climate and sea level change: a synthesis. In: Warrick R. A., Barrow E. M. & Wigley T. M. L. eds. Climate and Sea Level Change: Observations, Projections and Implications, pp. 3-21. Cambridge University Press, Cambridge. WARRICK R. A., BARROW E. M. & WIGLEY T. M. L. 1993 (Editors). Climate and Sea Level Change: Observations, Projections and Implications. Cambridge University Press, Cambridge. WATSON R . T., ZINYOWERA M. C. & Moss R . H . 1 9 9 6 (Editors). Climate Change 1995: Impacts, adaptations and Mitigation of Climate Change: Scientific - Technical Analyses. Published for the IPCC by Cambridge University Press. WIGLEY T. L. M. & RAPER S. C. B. 1993. Future changes in global mean temperature and sea level. In: Warrick R. A., Barrow E. M. & Wigley T. M. L eds. Climate and Sea Level Change: Observations, Projections and Implications, pp. 111-133. Cambridge University Press, Cambridge. WOODROFFE C. D. & MCLEAN R. F. 1993. Cocos (Keeling) Islands: Vulnerability to Sea-Level Rise. Report to Climate Change and Environmental Liaison Branch, Department of the Arts, Sport the Environment, and Territories, Canberra. WOODWORTH P. L. 1993. Sea level changes. In: Warrick R. A., Barrow E. M. & Wigley T. M. L. eds. Climate and Sea Level Change: Observations, Projections and Implications, pp. 379-391. Cambridge University Press, Cambridge. Received 10 March 1998; accepted 26 August 2000

Geological Society of Australia Special Publication 21, 215-225

CHAPTER 18—Value of estuarine sediments for understanding natural environmental change in coastal regions S, L NiCHOL Department of Geography, University of Auckland, Private Bag 92019, Auckland, New Zealand (s. nichol@auckland. ac. nz). Estuaries along the south-central coast of New South Wales hold a detailed history of environmental change that spans most of the Holocene epoch. This record is captured in sediments that have been supplied from land and sea during post-glacial sea-level rise and subsequent stillstand under a microtidal, wave-dominated regime. New South Wales estuaries comprise three distinct sediment zones: a sandy barrier at the mouth, a muddy central basin, and a coarse-grained fluvial delta at the landward end, with the degree of estuary infilling largely determined by fluvial sediment supply. This chapter examines the evidence for and assesses the importance of, three natural processes that have impacted on New South Wales estuaries: sea-level rise, river floods, and acidification. Each process also has the potential to impact upon human uses of estuaries. Sea-level rise in the early to mid-Holocene forced a landward migration of all estuarine depositional environments, as shown by the stratigraphy of the Hawkesbury River estuary. A similar response can be modelled for future sea-level rise in a Greenhouse world. The impacts of river floods throughout the Holocene are also recorded in estuarine deposits with documented examples from the textural and diatom record in the Hawkesbury River and Tuross River estuaries. These records highlight the importance of flood events to estuarine ecology and also inform us that the impact of floods changes through time as an estuary matures. An understanding of the type and distribution of major facies within an estuary is also vital to avoid acidification of estuarine waters through release of acid sulfates from sediments, which can result from engineering works such as canal and drain dredging. The threat of acidification is potentially severe for south-central coast estuaries because of the relatively large area occupied by sulfide-bearing sediments. Knowledge of estuarine geomorphology, stratigraphy and evolution is an essential prerequisite to effective management of the impact of the natural processes on human uses of estuaries. KEY WORDS: acid sulfate soils, coastal sediments, environmental management, estuaries, floods, Holocene, New South Wales, sea-level.

INTRODUCTION Estuaries straddle the boundary between land and sea and are influenced by a mix of oceanic and terrestrial processes. Because of this interaction of marine (tides and waves) and non-marine (rivers) processes, estuaries are recognised as highly complex natural systems. They are also regarded as a valuable resource for a variety of human activities, with estuarine waters and surrounds used for urban settlement, industry, agriculture, aquaculture and recreation. The importance of an awareness and understanding of the complexity of estuarine systems cannot be understated, particularly as estuaries come under increasing pressure from human use. This chapter concentrates on the geological character of estuaries with particular emphasis on the sedimentary record. The primary aim is to demonstrate the value of estuarine sediments for understanding natural environmental change in coastal regions. Along the southeast coast of Australia, the record of environmental change observed in estuarine deposits spans most of the last 10 000 years of the Holocene, and in most estuaries the record is very detailed because estuaries

are efficient sediment traps of material washed in from land and sea. However, the study of estuarine deposits involves more than reconstructing the past. We can use our understanding of past change to model and plan for future change. Thus, the relevance of knowledge regarding past environmental change to planning for future change will also be discussed here. The geographic focus of this chapter is the central to southern sector of the New South Wales coast (Figure 18.1), where a range of estuary types has evolved to varying stages of geomorphic development. Coastal evolution along the New South Wales coast is governed by a wavedominated, microtidal process regime. Modal wave height is 1-2.5 m and the tidal range is <2 m. Freshwater flow into estuaries is highly variable, ranging from near zero during drought conditions to large floods and accompanying storms. This variability in fluvial energy will influence processes of sediment transport and deposition in estuaries with the potential to leave its mark on the sediment record, as will be shown later in this chapter. Similarly, sedimentary evidence may be left by variations in wave energy associated with changes in prevailing ocean swell, extreme

216

S. L. Nichol

events such as storm surge or tsunami, and by gradual changes to tidal conditions as an estuary fills with sediment over geological time. The challenge for the sedimentary geologist is to learn to recognise and understand these signatures in the sediment record. Thus, one of the goals of this chapter is to convey some of the information that has been collected from estuaries in the central to south coast of New South Wales, relevant to our understanding of environmental change in these systems. DEFINITIONS AND KEY CONCEPTS

The problem of establishing a definition for estuaries that is both relevant and applicable to real-world situations is long-standing. Early definitions employed criteria that recognised the unique hydrological conditions of estuaries by reference to salinity levels (Pritchard 1967) and tidal limits (Fairbridge 1980). As argued by Dalrymple et al (1992), these definitions are of little value to the earth scientist studying the estuarine sedimentary record, either modern or ancient. With the sedimentologist in mind, Dalrymple et al (1992) proposed the following definition for an estuary: 'the seaward portion of a drowned river valley which receives sediment from both fluvial and marine sources, and which contains facies influenced by tide, wave and fluvial processes; the estuary is considered to extend from the inner limit of tidal facies at its head to the outer limit of coastal facies at its mouth.' Several important characteristics of estuaries are recognised by this definition. First, that sediment is sourced from both terrestrial and marine environments. Second, sediment transport in an estuary is influenced by tidal, waveinduced and river currents. Because the relative importance of marine (waves + tides) and fluvial processes varies along the length of an estuary, the physical character of estuarine facies is also spatially varied. In this regard, there is considerable complexity within estuaries and even greater variation between estuaries. To simplify this complexity, a broad classification of estuarine systems has been developed by Dalrymple et al (1992) with the intention that it be applicable globally. This classification scheme is underpinned by the concept that estuaries lie along a continuum defined according to the relative importance of fluvial, tidal and wave energy. At one end of the continuum are tidedominated estuaries which develop a distinctive facies association that is an expression of tidal processes overwhelming wave and river depositional processes (Dalrymple et al 1992). The inner Bay of Fundy in eastern Canada is an example of this estuary type (Dalrymple et al 1990). At the other extreme, a wave-dominated estuary will develop a facies zonation that reflects the dominance of ocean wave energy at the estuary mouth. The estuaries along the central to south coast of New South Wales are excellent examples of the latter type of estuary. Estuary facies zonation in New South Wales

The presence of distinct zones in estuaries along the southeast coast of Australia was first recognised by Bird (1967 a, b) who defined three zones in terms of salinity properties and associated each zone with a range of ecological and

Figure 18.1 Map of the central to south coast of New South Wales, showing the location of estuaries mentioned in the text. geomorphic characteristics (Jennings & Bird 1967). Bird also drew attention to the fact that estuaries along the coast of southeast Australia differ in terms of the surface area occupied by each depositional zone, suggesting this was a function of estuary size and shape, river flow and volume of tidal exchange. Subsequent workers have documented the physical and biological properties of surface sediments in select south-central coast estuaries (Roy & Peat 1973, 1975, 1976; Reinson 1977; Kidd 1978; Eliot 1979; Nichol & Murray-Wallace 1992; Bradshaw 1988; Hunter 1989; Nichol 1991, in press; Buman 1995). These surveys document a consistent tripartite zonation pattern of estuarine facies along the New South Wales south-central coast, termed here the seaward zone, central zone and landward zone.

Environmental change in estuaries

Barrier

Figure 18,2 Oblique aerial photograph of Murrah Lagoon on the New South Wales south coast, showing the barrier complex, central basin and fluvial delta zones typical of New South Wales south-central coast estuaries. The seaward zone of estuaries along the New South Wales south-central coast is characterised by a sandy barrier complex (Figure 18.2). The development of this barrier is dependent upon adequate sediment supply from marine sources, such as the shoreface and inner continental shelf, and sufficient wave, wind and tidal energy for shoreward sediment transport. The dominant sediment type is wellsorted, medium-sized quartzose sand, with variable carbonate content (Roy 1984b, 1994). On the New South Wales south-central coast, barriers comprise beach, nearshore and dune depositional environments, and are contiguous with a tidal inlet that connects the estuary to the sea. In some estuaries on the south-central coast, the tidal inlet is an ephemeral feature and the barrier completely blocks the central and landward zones from regular tidal exchange. Reopening of an ephemeral inlet typically requires a river flood to scour sediment from the lowest part of the barrier. At the landward side of the tidal inlet, incoming tidal currents transport marine sand to form a floodtidal delta. Ebb-tidal deltas, which form on the seaward side of a tidal inlet, are rare due to the high wave energy along this coast which efficiently reworks any sediment that may be transported seaward from an estuary mouth. Backbarrier flats are situated landward of beach and dune environments, lie within intertidal to low supratidal elevations and are usually vegetated by mangrove and saltmarsh plants. The second facies zone in New South Wales estuaries, located landward of the barrier complex, is the central basin (Figure 18.2). Depositional environments in this zone include a subtidal basin, an intertidal shoreline and a supratidal shoreline. It is important to note that the formation of a central basin requires the presence of a barrier and tidal delta to seaward, which shelter the basin from ocean waves. Consequently, the central basin is a low-energy depositional environment with wave action restricted to locally generated wind waves. In New South Wales southcentral coast estuaries, water depth in the central basin ranges from < 2 m to - 1 0 m and sediments comprise fine sand, silt and clay washed in from the catchment. In addition to clastic terrigenous sediment, organic debris accu-

217

mulates on the floor of the central basin. Organic matter is derived from a variety of local, or autochthonous sources, such as mangrove forests, saltmarsh and Zostera seagrass beds, plus from in situ production of faecal material of benthic fauna. Sediment deposited along the shoreline of estuarine basins is a mix of silt and sand derived from the catchment, plus coarse sand and gravel washed in from adjacent slopes. Shoreline deposits are usually colonised by mangrove (Avicennia marina) in intertidal areas, and saltmarsh on supratidal surfaces. The third and most landward facies in New South Wales south-central coast estuaries includes the fluvial delta, channel and alluvial plain depositional environments (Figure 18.2). Each of these environments comprises subenvironments, reflecting the relative complexity of this zone. Thus, the fluvial delta may have supratidal, intertidal and subtidal sub-environments, and the alluvial plain can include levee, overbank, crevasse splay and cutoff channel sub-environments. Sediments in this zone originate in river catchments and range in texture from poorly sorted, coarse sand and gravel to well-sorted sand and silt.

Estuary evolution The geological evolution of estuaries involves progressive infilling of an estuarine valley through changes in the surficial extent (and volume) of each of the three facies zones described above. Roy (1984b, 1994) presented a threestage model for this process, based on New South Wales estuaries. Thus, an estuary at stage 1 is characterised by a relatively large central basin, small fluvial delta - alluvial plain zone and an established barrier complex. On the New South Wales coast, barriers formed with the post-glacial rise in sea-level and had developed their present form by about 3000 years before present (Thom et al 1978; Thom & Roy 1985). Progression to stage 2 involves progradation of the fluvial delta into the central basin, usually with little change to the size of the barrier. A stage 3 estuary is characterised by an extensive fluvial delta and alluvial plain that occupy the former basin area and are contiguous with the barrier complex. Tidal flow in a stage 3 estuary is confined to a channel network that connects with the tidal inlet. Estuaries along the New South Wales south-central coast represent the full continuum of evolutionary stages, due to local and regional controls on infilling related to factors such as geology, catchment size and topography, fluvial competence and partial preservation of Pleistocene valley fill (Nichol & Murray-Wallace 1992). This range of infill states is summarised in Figure 18.3, showing the proportional area of each facies zone for estuaries along the New South Wales south-central coast. Two estuary types are defined here on the basis of the area occupied by the barrier complex (Nichol 1991). Thus, Type 1 estuaries (n = 49) have a barrier area which occupies about 20% of the total estuary area, whereas Type 2 estuaries have a barrier area approximately 60% of the total area. There are fewer examples of the latter group (n = 19) and these are dominated by broad prograded barriers, such as the Moruya River. Within each estuary type, subtypes are defined in terms of the area occupied by the fluvial delta - alluvial plain (Figure 18.3). For Type 1 estuaries, three subtypes

218

S. L. Nichol

1A | Alluvial plain - fluvial delta

Middle Lagoon - Type 1A

• Central basin

Wapengo Lagoon - Type 18

2B Barrier complex

Pambula River - Type 1C

exist which together describe the progression from partially filled (Type 1A) to infilled estuary valleys (Type 1C). The same progression exists in Type 2 estuaries, between two subtypes. This grouping of south-central coast estuaries highlights the importance of the development of an alluvial plain and fluvial delta in the infilling process of all estuaries, a process that is ongoing. In contrast, onshore delivery of marine sand appears to have ceased along the southcentral coast of New South Wales, as demonstrated by investigations of prograded barrier systems, which report negligible additional barrier progradation after 3000 a BP (Thorn et al 1978, 1981; Roy & Thorn 1981; Thorn & Roy 1985). Thus, it is the estuarine basin and fluvial delta alluvial plain zones of estuaries along the New South Wales south-central coast that are most likely to hold a detailed record of environmental change spanning the full period of estuary development. Even though estuaries act as highly efficient sediment traps, the sediment record is rarely if ever fully preserved. Erosion of some estuarine deposits is common, particularly in situations where advanced infilling of an estuary valley leads to increased channelisation of river and tidal flows, and in turn, incision into pre-existing sediments. Erosion within an estuary is typically marked in the sediment column by a surface of unconformity at a facies boundary. A common example found in mature estuaries is a sharp contact between silt-clay facies and overlying sand-gravel facies, representing channel scour into central basin deposits and subsequent infilling of that channel (Nichol 1991 figure 7). This unconformity surface may also represent a significant hiatus in the sediment record (hundreds to thousands of years) and loss of the local environmental history.

ESTUARINE ENVIRONMENTAL RECORD Previous studies of estuaries along the southern and central coast of New South Wales have documented the stratigraphy and facies character of valley-fill deposits at a number of sites (Roy 1984a,b, 1994; Roy et al 1980; Roy & Boyd

Figure 18.3 Cumulative percentage chart showing the proportion of estuary area occupied by each facies zone for type 1 (small barrier) and type 2 (large barrier) estuaries along the south-central coast of New South Wales, with oblique aerial photographs of representative examples. Subtypes (1A, IB, 1C, 2A, 2B) highlight the importance of fluvial delta - alluvial plain development in the process of estuary filling. The number (n) of estuaries in each subtype is also shown. This classification was produced using the ENTROPY statistical grouping procedure (Johnston & Semple 1983; Nichol 1991).

1996; Nichol 1991; Nichol et al 1997). These studies provide a basis upon which the importance of the estuarine environmental record can be evaluated. This section focuses on three natural processes that may be registered in estuarine deposits and which, in human terms, are perceived to be environmental threats: sea-level rise, river floods, and acidification of estuarine waters. An understanding of each process is not only relevant from a scientific perspective, but also for management of estuarine systems as they come under increased pressure from human activity.

Sea-level rise Estuarine sediments have the potential to record the effects of sea-level rise, as well as the shoreline transgression (landward retreat) that typically accompanies sea-level rise. Because estuarine sediments lie within a valley, their preservation potential is relatively high. However, preservation requires that deposits are not removed by rivers or scoured by wave and tide action. Consequently, our understanding of sea-level rise and the coastal response is strongly based on description and interpretation of estuarine sediments preserved on the modern inner continental shelf and in modern estuaries. The coast and inner continental shelf of southeast Australia have yielded a wealth of data that provides for reconstruction of sea-level rise during post-glacial time (Thorn & Roy 1983). More than 60 radiocarbon ages on estuarine shell, wood and in situ mangrove stumps have been used to establish a sea-level curve on the premise that these organisms are a proxy indicator of past sea-levels. The sea-level curve for southeast Australia shows a rise in relative sea-level of 1.3 m per century between 12 and 7.7 ka BP, then decreasing to 0.4 m per century until the present level was reached about 6.4 ka BP (Thorn & Roy 1983). The following case study of the Hawkesbury River estuary provides an example of the impact of this sea-level rise on estuaries. The lower reaches of the Hawkesbury River valley are occupied by a succession of deposits that record the transi-

Environmental change in estuaries

219

Figure 18.4 Hawkesbury River estuary, (a) Generalised stratigraphy of the seaward zone in Broken Bay (after Roy 1984a). (b) Plot of sediment texture for core showing facies transition from silty sand (fluvial delta) to silt-clay (central basin) that accompanied sea-level rise in the early to mid-Holocene. (c) Plot of sediment texture and positions of peaks in freshwater diatoms from a core taken from the upper Hawkesbury valley (after Devoy et al 1994; Nichol et al 1997).

220

S. L. Nichol

tion from a fluvial to an estuarine environment, as a result of post-glacial sea-level rise along the east Australian coast. The Hawkesbury River incised valley is among the deepest along the south-central New South Wales coast, with a maximum bedrock depth of 125 m below present sea-level in Broken Bay (Albani & Johnson 1974). This depth has been sufficient to accommodate both Pleistocene and Holocene sediments and store them below present sealevel. A subaerial barrier has not formed at the entrance to the Hawkesbury estuary, because all marine-derived sediment is accommodated subtidally in a large flood-tidal delta in Broken Bay. Stratigraphic investigations in Broken Bay reveal this tidal delta is a landward-thinning wedge of sand with a maximum thickness of 30 m (Roy 1984a, 1994) (Figure 18.4). This wedge extends 8 km landward in the subsurface and interfingers with, and overlies estuarine basin mud (Roy 1984a). In turn, this estuarine mud overlies fluvial silt and sand of Late Pleistocene age. This stratigraphic arrangement of marine-sourced sand overlying mud and sand deposited by the Hawkesbury River is key evidence that the seaward facies zone in the Hawkesbury estuary has migrated, or transgressed, at least 10 km landward under the influence of rising sea-level. Radiocarbon dating of shells recovered from estuarine mud suggests this process was complete by ca 7 ka BP (Roy 1984a). Further upstream in the Hawkesbury River estuary, there is additional evidence for sea-level rise causing major environmental change. A core collected from the modern floodplain of the Hawkesbury River approximately 40 km from the estuary entrance (Figure 18.4) provides a 23 m record of estuarine sedimentation dating to earlier than 8 ka BP (Nichol et al 1997). Sediments buried 19 m below present mean sea-level consist of fine- to medium-grained laminated sand (Figure 18.4). In other cores from the upper Hawkesbury, the same sand facies preserves bi-directional cross-bedding and flaser bedding (Nichol et al 1997). Bidirectional cross-bedding is interpreted as evidence for reversing flow direction and flaser bedding as evidence for fluctuations in flow velocity. In an estuarine context, this association of sedimentary structures is diagnostic of tidal processes. Detrital organic material is also preserved as concentrated beds, and radiocarbon dating of this material from two cores provide an age range of 8-6.8 ka BP for the interval 18.2-6.8 m below present mean sea-level. Based on our knowledge of modern estuarine facies and using these age-depth results as a guide, this sandy deposit most likely formed in a shallow subtidal to intertidal fluvial delta environment when sea level was 5 - 1 0 m lower than present (Thorn & Roy 1983; Nichol et al 1997 figure 2). With continued sea-level rise, however, this ancestral river delta was rapidly drowned, leading to a major change in depositional conditions at this site as evidenced by the facies transition from delta sand to a mix of fine-grained sand and mud (Nichol et al 1997). The muddy sediments are also characterised by fine horizontal laminations, suggesting low energy to quiescent depositional conditions. In an estuarine context, fine-grained sediment is deposited preferentially in the central basin where wave, tide and river energy is low and water depth sufficient to allow silt and clay to settle out of suspension. Thus, the muddy facies in the Hawkesbury River core is interpreted as evidence for a former central basin in the upper Hawkesbury valley.

In summary, for both the marine and fluvial portions of the modern Hawkesbury River there is a clear record of a landward, or transgressive, shift in the position of the major depositional zones due to the influence of sea-level rise. At the seaward end of the estuary, the flood-tidal delta migrated into the central basin as tidal currents transported marine sand landward. Along the landward reaches of the estuary, an ancestral fluvial delta was drowned by rising sea-level allowing the estuarine basin to expand landward and causing the river delta to retrograde upvalley (Nichol et al 1997). This response is not unique to the Hawkesbury River estuary, nor indeed the south-central coast, with similar stratigraphy and evolutionary histories documented for other New South Wales estuaries (Roy 1994; Roy et al 1995; Roy & Boyd 1996). Furthermore, the scale of sealevel rise that caused the landward retreat of estuarine systems in the early Holocene was of a similar magnitude to that forecast for future sea-level rise in a Greenhouse world (Gornitz 1995). Current estimates range from 0.4-1.0 m rise in global sea level by 2100 (Gornitz 1995). Assuming the estuarine response to the post-glacial rise in sea-level will be repeated, it is also possible to forecast major landward shifts in the position of depositional zones in all estuaries along the southeast coast of Australia.

Floods There are few published examples of flood deposits in estuarine sediments of the New South Wales south-central coast, partly because of the difficulty associated with establishing criteria for their recognition. It is evident, nonetheless, from the limited available research findings that these criteria must include physical and biological indicators of floods. A simple change in sediment grainsize, for example, is insufficient evidence because it is open to a variety of interpretations, especially in estuaries where flood currents are one of several possible causes for a grainsize variation. In terms of biological indicators, diatoms are particularly sensitive to the changes in salinity of estuarine waters which accompany floods (Battarbee 1986), with the fossil diatom record offering promise for this avenue of research (Devoy et al 1994). Diatoms are a microscopic unicellular algae (Class Bacillariophyceae) that commonly live in marine and freshwater environments in benthic and pelagic colonies (Werner 1977; Round et al 1990). All diatom species are sensitive to the pH, salinity, and other chemical properties of water, with certain species only able to develop under particular conditions. There is now an established diatom ecology which allows assemblages of diatom species to be interpreted in terms of particular water salinity, pH and chemistry through the use of statistical techniques (Birks 1995). The microscopic cell wall (termed the frustule) of diatoms is composed of silica and is generally resistant to decay. Consequently, where diatom frustules are preserved in sediments local to their habitat, they can provide a detailed record of environmental change at that site, including floods (Stoermer & Smol 1999). It is important to note that diatoms are readily transported and that a proportion of a fossil diatom assemblage may include diatoms from a variety of aquatic environments. In estuaries, however, the presence of these transported, or allochthonous,

Environmental change in estuaries diatom species is important information that may be used to recognise floods. TWo estuaries of the New South Wales south-central coast have been the focus of palaeoenvironmental research using combined physical and biological analytical techniques, with both sites yielding promising records of varied river flow, among other changes. A 16 m core taken 90 km upstream from the mouth of the Hawkesbury River records an association between concentrations of freshwater to brackishwater diatom species and increased concentrations of silt and clay, in a dominantly sandy facies (Figure 18.4) (Devoy et al 1994). Within a 10 m section of that core, spanning the time interval 7.9-6 ka BP, the concentration of freshwater and brackishwater diatoms exceeds 70% at seven intervals (Devoy et al 1994). These concentrations are interpreted as evidence for short-lived fluctuations in the salinity of the Hawkesbury River, either due to a single large flood, or a period (perhaps a decade) of above-average discharge, or both. Certainly, freshwater discharge in the modern Hawkesbury River is known to range from 38 to 1250 m 3 s _ 1 (Kjerfve et al 1992), within decadal trends of flood-dominated and drought-dominated flow regimes (Erskine & Warner 1988). The concentration of siltand clay-sized sediment in the core is also consistent with increased suspended sediment load during a flood. One interesting and important trend in the Hawkesbury core is the upward decrease in sand content (Figure 18.4). This is a good example of the type of changes in depositional style that can occur through time, noted earlier. In this instance, the decrease in sand may be interpreted as a function of weaker river currents as this site became infilled and was transformed from an estuarine basin to channel and ultimately to alluvial plain. The progressive increase in silt and clay content associated with freshwater diatom concentrations supports this interpretation, with the greater proportion of silt more likely to be deposited in an overbank environment than in the river channel. The Tuross Lakes estuary is the second south-central coast site to have been studied using palaeoecological and sedimentological techniques (Tibby 1996). The mid- to late Holocene sediment record in the modern fluvial delta zone of the Hiross estuary contains a history of major changes in the salinity of the upper estuary, as evidenced by the fossil diatom record. Two intervals of the Holocene record, as reconstructed by Tibby (1996), are of particular interest here. The first spans the period 5.5-4.9 ka BP and is characterised by a strongly marine diatom assemblage preserved in estuarine muds below 6.5 m depth (Tibby 1996). The core site from which this material was taken is today a cutoff embayment of the Tuross delta system located approximately 12 km from the sea and with mesosaline waters ( ~ 7 % o : Tibby 1996). However, the diatom record indicates that this site was formerly polysaline with a salinity range of 2 0 - 3 0 % o . The implication of this finding is that about 5 0 0 0 years ago, the fluvial delta to the Tuross River (and perhaps the seaward barrier) had not yet prograded into the central basin zone and marine waters regularly penetrated to the upper estuary in higher concentrations than today. Fluvial influences on sedimentation were restricted to the most landward reaches of the Tuross estuary system during this time, with no apparent sedimentary evidence of floods, although floods must have occurred.

221

The second interval of interest in the Tuross sediment record features a marked increase in fluvial influence that Tibby ( 1 9 9 6 ) estimated prevailed for 5 5 0 years, ending about 4.4 ka BP. The argument for higher river discharge is based on a concentration of freshwater ( 0 - 0 . 5 % o ) to oligosaline ( 0 . 5 - 5 % o ) diatom species in muddy sediments that also contain local concentrations of broken shell. Unlike the Hawkesbury River case, there does not appear to be a major change in sediment grainsize. The preservation of shell fragments, however, supports the hypothesis of a shift to higher energy depositional conditions. Tibby ( 1 9 9 6 ) also reported pollen evidence for an increase in wet forest taxa during this interval, which is consistent with higher rainfall and runoff. Whether this ecological change in the upper Tuross estuary was due to flooding or the progradation of the fluvial delta is unclear from available information. However, this example does illustrate the variability in the style of response to long-term changes in river discharge. These differences in the style of flood-related deposition between the Hawkesbury and Tuross Rivers highlight the importance of understanding flood behaviour in estuaries. Our present level of knowledge for south-central coast estuaries is meagre. Yet at the very least we need an appreciation of the range of possible flood responses within estuaries, and there is scope for realising this on the New South Wales south-central coast by examining the flood record in estuaries representative of the range of infill states. From this understanding will emerge a scientific framework for flexible policy regarding flood-hazard management of coastal environments. Acidification of estuarine waters Acidification of estuarine waters is a contemporary environmental problem faced by users and managers of lowland coastal areas in Australia that has the potential to seriously impact upon the ecology and overall amenity value of many estuaries (Sammut et al 1 9 9 6 ) . Acid estuarine waters may result from discharge of polluted water from industrial, urban or agricultural sources (Chenhall et al 1995) or from human disturbance of naturally occurring acid-bearing estuarine sediments (White & Melville 1 9 9 6 ) . In this section, the focus is on the latter form of acidification. The acid released from estuarine sediments into waterways is sulfuric acid that is derived from acid sulfate soils formed in organic-rich sedimentary environments (White & Melville 1 9 9 6 ) . It is because acid is leached from sediments that an understanding of estuarine depositional systems is important to proper assessment and management of acid sulfate soil risks in the coastal area. In an estuarine context, acid sulfate soils are sedimentary deposits which contain concentrations of iron sulfides, typically in the form of iron pyrite (Melville et al 1991; Sammut et al 1 9 9 6 ) . Optimum conditions for the formation of iron sulfide are provided in low-energy, brackish to saline estuarine environments where reduction of dissolved sulfate held in seawater is promoted by anaerobic (oxygendepleted) conditions in iron-bearing muddy sediments. A local supply of organic detritus is also necessary to promote sulfide reduction via microbial processes (Sammut et al 1 9 9 6 ) . Regular tidal exchange is a further important ingredient because tides flush out alkaline by-products of the

222

S. L. Nichol

waters. In several documented cases for the New South Wales north coast, sulfuric acid levels following high rainfall events have been sufficiently lethal to kill large numbers of fish, crustaceans and other benthic fauna (Sammut et al 1996; White & Melville 1996). For estuaries of the New South Wales south-central coast, it is the fluvial delta zone and estuarine basin zone that provide the most suitable environments for acid sulfate soil formation, with the sandy tidal delta and barrier zone presenting less of a risk of releasing sulfuric acid into estuarine waters. This risk has been recognised by state government authorities and addressed through the publication of the Acid Sulphate Soil Risk Map series for the state of New South Wales (Soil Conservation Service 1995) and accompanying management guidelines (Blunden 1995). This map series shows the alluvial plain, fluvial delta, estuarine basin and shoreline environments of New South Wales estuaries to have a 'high probability of occurrence' of acid sulfate within the soil profile (Soil Conservation Service 1995). This classification, based largely on aerial photo surveys, presents interpretations of the surficial extent of the acid sulfate soil risk and does not extend to (h FfwsaS MM* subsurface information. Subsurface data must necessarily gstuarifttetey come from site-specific investigations and indeed, such surtoveys form part of the recommended guidelines for acid sulll fate soil risk assessment (Blunden 1995). Using two Mtfim examples from the New South Wales south-central coast, ij mi Wapengo Lagoon and Narrawallee Inlet, the importance of Pbmtemm mtuarim-titoy stratigraphic investigations to a better understanding of the acid sulfate soil risk, is highlighted below. Wapengo Lagoon, situated on the far south coast of New South Wales (Figure 18.1), is a partially filled estuary with a wide variety of depositional environments, including an alluvial plain and delta complex that is prograding into a central mud basin. The seaward zone is dominated by a large flood-tidal delta that extends 3 km into the lower estuary (Figure 18.5a). In contrast, Narrawallee Inlet is an infilled estuary characterised by a supratidal alluvial plain contiguous with a barrier complex. Tidal waters are confined to a narrow (<50 m), meandering channel network that conveys seawater almost to the landward limit of the valley (Figure 18.5b). In terms of the risk of sulfuric acid Figure 18.5 (a) Map of facies zones in Wapengo Lagoon and release from estuarine deposits, both Wapengo Lagoon and generalised stratigraphy in cores from the fluvial delta, (b) Map Narrawallee Inlet present varied risk levels (Table 18.1). of facies zones in Narrawallee Inlet with generalised strtigraphy Low-risk depositional environments include the barrier from three drillcores. complex of both estuaries, with the exception of flood-tidal areas which are assigned a high to severe risk level (Soil Conservation Service 1995). More important, however, are diagenetic process of iron sulfide formation (White & the alluvial plain, fluvial delta and central basin environMelville 1996). ments that cover a much larger area than the tidal delta and Provided iron sulfide-bearing sediments remain subtidal are rated as posing a high to severe acidification risk (Table or buried, either by alluvial or aeolian deposits, they pose 18.1). It is important to note that in both cases, the acid sulminimal hazard to the quality of estuarine waters. fate soil risk is potential with no documented instances of actual acid sulfate soil-related pollution in either estuary. In However, acid sulfate soils become a problem when they addition, central basin deposits in Wapengo Lagoon are are exposed to air by activities such as canal dredging and subtidal and therefore unlikely to be exposed to oxidation. excavating for drains, or dams. When exposure to oxygen Acid sulfate soil risk maps show the surface extent of occurs, sulfidic sediments dry out and rapidly oxidise. A deposits of a given risk with an added classification based common signal of this process is the formation of jarosite, on the depth interval at which sulfidic soils are expected to a distinctly yellow mineral, in exposed sediments (Sammut occur. As noted earlier, this classification requires site-spe& Lines-Kelly 1996). The oxidation process involves concific subsurface investigations to adequately evaluate the version of iron sulfides to sulfuric acid, which is free to acid sulfate soil risk. Subsurface data for Wapengo Lagoon leach from the exposed sediment and enter estuarine

Environmental change in estuaries

223

Table 18.1 Present surface area, volume and overall acid sulfate soil (ASS) risk of major facies zones in Wapengo Lagoon and Narrawallee Inlet, New South Wales south coast.

Barrier complex Estuarine basin Alluvial complex

Surface area (km2) 2.3 1.4 ^2

Wapengo Lagoon Volume (m3 x 106) 13.7 4.2 16J

ASS risk Low Severe High

Surface area (km2) 1.4 0 9.5

Narrawallee Inlet Volume ASS risk (m3 x 106) 64.7 Low 10.9 N/A 57.2 Severe

Depositional environments for each zone as follows: Barrier complex, beach, dune, tidal inlet, flood-tidal delta & backbarrier flat; Estuarine basin, central mud basin and shoreline; Alluvial complex, fluvial delta, alluvial channel and alluvial plain (Nichol 1991). ASS risk assessment from 1:25 000 Acid Sulphate Soil Risk Map Series, Murrah/Brogo Sheet (Wapengo) and Milton-Cunjurong Sheet (Narrawallee) (Soil Conservation Service 1995).

and Narrawallee Inlet serve to illustrate this point. Two cores from the alluvial plain-fluvial delta zone of Wapengo Lagoon, an area rated as posing a high to severe acid sulfate soil risk if disturbed, reveal Holocene estuarine deposits are 3.3-3.5 m thick and overlying weathered (oxidised) deposits of presumed Pleistocene age (Figure 18.5a). A calibrated radiocarbon age of 6885 a BP on wood buried immediately above the weathered material confirms the Holocene age of non-oxidised deposits. Even though the Pleistocene sediments are oxidised they pose minimal acid sulfate soil risk because it is assumed sulfuric acid has already been leached when the host deposit was exposed prior to Holocene sea-level rise. The texture of the overlying Holocene sediments is highly variable, ranging from sandy silt and clay at the base to silty, coarse sand and gravel toward the surface. Hence the potential for iron sulfide formation will also vary through the sediment column, being greater in finer grained sediments below 1-1.5 m depth. This information can be used to set limits on drain depths, for example, if ever proposed for Wapengo Lagoon. The risk of acid sulfate soil pollution from deposits in the landward part of Narrawallee Inlet is rated as high with potential acid sulfate soil buried within 3 m of the ground surface (Soil Conservation Service 1995). Cores collected from this area reveal the potential acid sulfate soil horizon is actually 8-12 m thick, comprising alluvial sand and silt (Figure 18.5b). This alluvial material overlies orange-brown estuarine mud, indicating oxidation has already occurred. However, the acid sulfate soil risk posed by this oxidised mud facies is assumed to be negligible because dating of fossil shells suggest deposition occurred during the Last Interglacial sea-level highstand of 125 ka BP (Nichol & Murray-Wallace 1992). Therefore, oxidation and acid leaching most probably occurred during sea-level lowstand of the last glacial period. However, local variations to the stratigraphy of Narrawallee valley and acid sulfate soil risk do exist. For example, a core taken midway along the Narrawallee valley recovered dark-brown to black estuarine mud at 4.5-7.6 m depth below alluvial sand (Figure 18.5b). A calibrated radiocarbon age of 7250 a BP on wood from -6.5 m suggests this shallower mud deposit formed in a mid-Holocene central basin. Moreover, the dark colour of this deposit indicates anaerobic conditions have been maintained since deposition and that the risk of acid sulphate release remains high if this material is ever disturbed. Similar spatial variations in estuary stratigraphy are expected to exist in all estuaries at an advanced stage of

infill. Because these infilled estuaries offer large areas of 'land' that if drained, can be utilised for a variety of agricultural, industrial or urban activities, they are at greatest risk of realising the acid sulfate soil potential. Hence, the need for site-specific subsurface surveys is greatest for these more complex systems. Effective management of the acid sulfate soil risk in these areas must embody an understanding of the style and character of estuarine facies development.

CONCLUDING REMARKS The purpose of this chapter was to demonstrate the utility of the estuarine sedimentary record as an indicator for natural environmental change. The examples of sea-level rise, floods and acidification of estuarine waters are natural processes that have and continue to produce change in all estuaries of southeast Australia, not only along the southcentral coast of New South Wales. It is important that the sediment record of these processes is documented and understood because the resultant information can be used to assist management of coastal regions, as outlined below. With regard to sea-level rise, our knowledge of how estuaries have responded in the past to sea-level rise is vital to formulation of predictive models for estuarine responses to future sea-level rise. These models already exist (Cowell et al 1992, 1995), and have been used to forecast deepening and widening of existing tidal inlets and central basins, flooding of fluvial deltas and alluvial plains by marine waters and landward reworking of major facies zones. The next step is to apply these models to individual estuaries, so that the present distribution of agricultural, aquacultural, urban and industrial activities in and around estuaries can be assessed and possibly reorganised if sea-level rise poses a risk to their viability. Extreme events such as river floods are known to have significant impacts on the geomorphology of coastal rivers and estuaries in southeast Australia (Erskine & Warner 1988; Nanson & Erskine 1988). However, our understanding of the sediment record of these events in New South Wales estuaries is quite limited. Available research indicates that the sediment record may hold a detailed history of floods that includes information on hydrological and ecological impacts (Devoy et al 1994; Tibby 1996). As argued by P. S. Roy (pers. comm. 1999), the link between the geomorphology, hydrology and ecology of estuaries must be a

224

S. L. Nichol

subject for future research. In terms of coastal management, this link is important to understand because it can provide the basis for informed decisions regarding human activities in estuaries and catchments. For instance, in the case of estuaries which have only an ephemeral connection to the sea, understanding the changes to estuarine ecology and water chemistry that occur when a flood causes barrier breaching can be used to decide whether artificially opening and maintaining an permanent entrance is desirable. Finally, proper management of estuarine soils with the potential for release of sulfuric acid into estuarine waters demands an understanding of facies organisation and character in estuaries. The phenomenon of acid sulfate soil has only emerged in recent years to become an issue for water quality in estuaries, and many estuaries have yet to develop the problem. By applying our understanding of estuarine stratigraphy to management of land and waterway use we have an opportunity to minimise the extent of future sulfate pollution. Such an outcome would be clear demonstration of the value of applying scientific knowledge to management of human activity in the coastal zone.

comparison of terrestrial and marine records. Quaternary Science Reviews 13, 241-256. Lake: Report on its condition, use and development potential, pp. 4.2-4.13. University of Wollongong Study Group (unpubl.). ERSKINE W . D. & W A R N E R R. F. 1988. Geomorphic effects of alternating flood- and drought-dominated regimes on NSW coastal rivers. In: Warner R. F. ed. Fluvial Geomorphology of Australia, pp. 223-245. Academic Press, Sydney. FAIRBRIDGE R. W. 1980. The estuary: its definition and geodynamic cycle. In: Olausson E. & Cato I. eds. Chemistry and Biogeochemistry of Estuaries, pp. 1-37. John Wiley, New York. GORNITZ V. 1995. Sea-level rise: a review of recent past and nearfuture trends. Earth Surface Processes and Landforms 20, 7-20. HUNTER T. 1989. Late Quaternary development of two estuaries, southern New South Wales, Australia. BA (Hons) thesis, University of Sydney, Sydney (unpubl.). JENNINGS J . N. & BIRD E. C. F. 1967. Regional geomorphological characteristics of some Australian estuaries. In: Lauff G. H. ed. Estuaries, pp. 121-128. Academic Press, New York. JOHNSTON R . J . & SEMPLE R . K . 1 9 8 3 . Classification using information statistics. Concepts and Techniques in Modern Geography 37. Geo Books, Norwich. KIDD R . W. 1 9 7 8 . Estuarine sediment regimes, far south coast, New South Wales. PhD thesis, Macquarie University, Sydney (unpubl.). ELIOT I. 1979. Lake sediments. In: Wallaga

KJERFVE

REFERENCES ALBANI A . D. & JOHNSON B. D. 1974. The bedrock topography and ori-

gin of Broken Bay, NSW. Journal of the Geological Society of Australia 21, 209-214. Handbook of Holocene Palaeoecology and Palaeohydrology, pp. 527-570. Wiley, Chichester. BIRKS H . J . B . 1 9 9 5 . Quantitative palaeoenvironmental reconstructions. In: Maddy D. & Brew J. S. eds. Statistical Modelling of Quaternary Science Data, pp. 1 6 1 - 2 5 4 . Quaternary Research Association, Cambridge. BIRD E . C. F. 1967a. Coastal lagoons in southeastern Australia. In: Jennings J. N. & Mabbutt J. A. eds. Landform Studies from Australia and New Guinea, pp. 365-386. Australian National University Press, Canberra. BIRD E . C . F. 1967b. Depositional features in estuaries and lagoons on the south coast of New South Wales. Australian Geographical Studies 5, 113-125. BLUNDEN B . 1 9 9 5 . Assessing and managing acid sulfate soils— Guidelines for land management in New South Wales coastal areas. Environmental Protection Authority, New South Wales, Sydney. BRADSHAW B. E. 1988. St. Georges Basin: morphology and Late Quaternary deposits. BSc (Hons) thesis, University of Sydney, Sydney (unpubl.). BUMAN M. 1995. Molluscs and the Holocene evolution of the Shoalhaven estuary. BSc (Hons) thesis, University of Wollongong, Wollongong (unpubl.). CHENHALL B. E., YASSINI I., DEPERS A. M. ET AL. 1995. Anthropogenic marker evidence for accelerated sedimentation in Lake Illawarra, New South Wales, Australia. Environmental Geology 26, 124-135. COWELL P. J . , ROY P. S. & JONES R . A . 1 9 9 2 . Shoreface translation model—computer simulation of coastal sand-body response to sea-level rise. Mathematics and Computers in Simulation 33, 603-608. COWELL P. J . , ROY P. S. & JONES R . A . 1 9 9 5 . Simulation of large-scale coastal change using a morphological behaviour model. Marine Geology 126, 4 5 - 6 1 . BATTARBEE R. 1986. Diatom analysis. In: Berglund B. E. ed.

DALRYMPLE R . W . , KNIGHT R . J . , ZAITLIN B . A . & MIDDLETON G . V . 1 9 9 0 .

Dynamics and facies model of a macrotidal sand bar complex, Cobequid Bay-Salmon River estuary (Bay of Fundy). Sedimentology 37, 5 7 7 - 6 1 3 . DALRYMPLE R. W., ZAITLIN B . A. & BOYD R. 1992. Estuarine facies models: conceptual basis and stratigraphic implications. Journal of Sedimentary Petrology 62, 1030-1146. DEVOY R. J., DODSON J. R., THOM B. G. & NICHOL S. L. 1994. Holocene environments in the Hawkesbury valley, New South Wales: a

B.,

SEIM

E.,

BLUMBERG A .

WRIGHT

1992.

Modelling of the residual circulation in Broken Bay and the lower Hawkesbury River, NSW. Australian Journal of Marine and Freshwater Research 4 3 , 1 3 3 9 - 1 3 5 7 . MELVILLE M . D., W H I T E I. & WILLETT I. R. 1991. Problems of acid sulphate soils and water degradation in Holocene pyritic systems. In: Brierley G. & Chappell J. eds. Applied Quaternary Studies, pp. 89-94. Department of Biogeography and Geomorphology, Australian National University, Canberra. NANSON G. C. & ERSKINE W. D. 1988. Episodic changes of channels and floodplains on coastal rivers in New South Wales. In: Warner R. F. ed. Fluvial Geomorphology of Australia, pp. 201-222. Academic Press, Sydney. NICHOL S. L. 1991. Zonation and sedimentology of estuarine facies in an incised valley, wave-dominated, microtidal setting, New South Wales, Australia. In: Smith D. G., Reinson G. E., Zaitlin B. A. & Rahmani R. A. eds. Clastic Tidal Sedimentology, pp. 41-58. Canadian Society of Petroleum Geologists Memoir 16. NICHOL S. L. Estuarine sedimentary environments of the New South Wales south coast. 1:25 000 Map Series. New South Wales Department of Land & Water Conservation, Sydney (in press). NICHOL S. L . & MURRAY-WALLACE C . V . 1992. A partially preserved last interglacial estuarine fill: Narrawallee Inlet, NSW. Australian Journal of Earth Sciences 39, 545-553. NICHOL S. L., ZAITLIN B. A. & THOM B. G. 1997. The upper Hawkesbury River, New South Wales, Australia: a Holocene example of an estuarine bayhead delta. Sedimentology 44, 263-286. PRITCHARD D. W. 1967. What is an estuary—Physical viewpoint. In: Lauff G. H. ed. Estuaries, pp. 3-5. Academic Press, New York. REINSON G. E. 1977. Hydrology and sediments of a temperate estuary—Mallacoota Inlet, Victoria. Bureau of Mineral Resources Bulletin 178. ROUND F. E . , CRAWFORD R . M . & MANN D . G. 1 9 9 0 . The Diatom: Biology and morphology of the genera. Cambridge University Press, Cambridge. ROY P. S. 1984a. Holocene sedimentation histories of estuaries in southeastern Australia. In: Hodgkin E. P. ed. Estuarine Environments of the Southern Hemisphere, pp. 23-60., Department of Conservation and Environment, Western Australia, Bulletin 161. ROY P. S. 1984b. New South Wales estuaries: their origin and evolution. In: Thom B. G. ed. Coastal Geomorphology in Australia, pp. 99-122. Academic Press, Sydney. ROY P. S. 1994. Holocene estuary evolution—stratigraphic studies from southeastern Australia. In: Dalrymple R. W., Boyd R. & Zaitlin B. A. eds. Incised-valley systems: Origin and sedimentary sequences, pp. 241-264. Society for Sedimentary Petrology Special Publication 51. ROY P. S. & BOYD R . 1996. Quaternary geology of southeast Australia: a technically stable, wave-dominated, sediment-deficient margin.

Environmental change in estuaries Field Guide to the Central New South Wales coast, International Geological Correlation Program: Project 367, Sydney. ROY P. S . COWELL P. J., FERLAND M . A. & THOM B. G . 1 9 9 5 . W a v e - d o m -

inated Coasts. In: Carter R. W. G. & Woodroffe C. D. eds. Coastal Evolution, pp. 121-186. Cambridge University Press, Cambridge. ROY P. S. & PEAT C. 1973. The bathymetry and bottom sediments of Tuggerah, Budgewoi and Munmorah Lakes and the subsurface stratigraphy of Tuggerah Lake. Geological Survey of New South Wales Report GS 1973/285. ROY P. S. & PEAT C. 1975. Bathymetry and bottom sediments of Lake Illawarra. Records of the Geological Survey of New South Wales 17, 65-79.

ROY P. S. & PEAT C. 1976. Bathymetry and bottom sediments of Tuross estuary and Coila Lake. Records of the Geological Survey of New South Wales 18, 103-134. ROY P. S. & THOM B. G. 1981. Late Quaternary marine deposition in New South Wales and southern Queensland—an evolutionary model. Journal of the Geological Society of Australia 28, 471-489. ROY P. S., THOM B. G. & WRIGHT L. D. 1980. Holocene sequences on

an embayed high-energy coast: an evolutionary model. Sedimentary Geology 26, 1-19. SAMMUT J. & LINES-KELLY R. 1996. An introduction to acid sulphate soils. Department of the Environment, Sport and Territories, Canberra.

SAMMUT J., WHITE I. & MELVILLE M . D . 1 9 9 6 . Acidification of a n estu-

arine tributary in eastern Australia due to drainage of acid sulfate soils. Australian Journal of Marine and Freshwater Research 47, 669-684.

SOIL CONSERVATION SERVICE 1995. Acid Sulphate Soil Risk Maps (1:25000 maps plus booklet). Soil Conservation Service of New

225

South Wales, Department of Land and Water Conservation, Sydney. STOERMER E. F. & SMOL J. P. (Editors) 1999. The Diatoms: Applications for the Environmental and Earth Sciences. Cambridge University Press, Cambridge. THOM B . G., BOWMAN G . M . , GILLESPIE R., TEMPLE R. & BARBETTI M .

1981. Radiocarbon dating of Holocene beach-ridge sequences in south east Australia. Department of Geography, University of New South Wales (Royal Military College, Duntroon) Monograph 11.

THOM B. G., POLACH H. & BOWMAN G . 1 9 7 8 . H o l o c e n e age structure

of coastal sand barriers in New South Wales, Australia. Department of Geography, University of New South Wales (Royal Military College, Duntroon) Report (unpubl.). THOM B. G. & ROY P. S. 1983. Sea-level change in New South Wales over the past 15 000 years. In: Hopley D. ed. Australian Sea Levels in the last 15 000 Years: A Review, pp. 64-85. James Cook University Monograph Series Occasional Paper 3. THOM B. G. & ROY P. S. 1985. Relative sea levels and coastal sedimentation in southeast Australia in the Holocene. Journal of Sedimentary Petrology 55, 257-264. TIBBY J. 1996. A mid- to late-Holocene diatom and pollen palaeoecology of the Tuross Lake system, south coast, New South Wales. Department of Geography and Environmental Science, Monash University, Monash Publications in Geography 46. WERNER D. 1977. The Biology of Diatoms. University of California Press, Berkeley. WHITE I. & MELVILLE M. D. 1996. Acid Sulfate Soils—Facing the Challenges. Earth Foundation Australia, Sydney, Monograph 1. Received 30 September 1999; accepted 1 June 2000

Geological Society of Australia Special Publication 21, 227-233

CHAPTER 19—Trace-metal pollution and sedimentation in coastal lagoons: an example from Lake lllawarra, New South Wales B. E. CHENHALL, B. G. JONES AND A. M. DEPERS School of Geosciences, University of Wollongong, NSW 2522, Australia. The low concentrations of toxic metals in Lake lllawarra appear to present few short-term, serious environmental problems. The redox and pH conditions in the lagoonal sediment together with its high organic matter and pyrite contents favour metal retention. Sedimentation rates in the lagoon since European settlement have increased greatly over the average Holocene rate of <1 mm y 1 . In the last decade, Wollongong City Council has employed an active policy of revegetation of portions of the catchment and has constructed sediment-retention ponds with wetland filters on a number of the smaller creeks. This program should reduce the rates of sediment erosion and transportation, and reduce the quantity of nutrients and trace metals entering the lagoonal system. However, sediment retention in this lagoonal system remains the greatest challenge facing scientists, engineers and statutory authorities in the future management of the water body. Disturbance of the sediment by dredging and foreshore reclamation work should be carefully monitored because of the potential for acid sulfate soil formation, with consequent release of trace metals back into the aquatic realm. KEY WORDS: anthropogenic factor, environmental management, Lake lllawarra, pollution, sedimentation, trace metals

INTRODUCTION Pollution of lagoonal systems is an inevitable consequence of urban, rural and industrial development in and adjacent to their catchments. Pollutant types and sources are diverse, including phosphorus and nitrogen from rural and agricultural practices, trace metals derived from industrial sources such as base-metal refining and metallurgical processing (Wittman 1981; Roy & Crawford 1984; Batley 1987; Chenhall et al 1992) and sediment influx in excess of natural levels (Kilby & Batley 1993). Throughout the world, one key adverse environmental impact of trace-metal contamination of the sediment and water column, not only in lagoonal, but also in terrestrial, riverine and marine settings, is the potential for transference of toxic metals (e.g. Pb, As, Hg) into the food chain. Indeed, many scientists will be familiar with the Minamata Bay disaster in Japan (Wittman 1981), where Hg contamination and subsequent assimilation of mercury by aquatic organisms resulted in extensive mercury poisoning including the death of many local people, mainly fisherfolk. This example represents an extreme case of assimilation of toxic trace metals, however, other very serious incidents of Hg assimilation have been reported in the Gulf of St. Lawrence, Canada (Laliberte et al 1992) and Cr and Cd poisoning have been documented in Japan. Arsenic poisoning, facilitated by the combustion of arsenic-rich coal, has been recorded from Czechoslovakia (Wittman 1981). In contrast to the relatively recent recognition of the consequences of Hg and Cd assimilation, health disorders in humans, including brain damage and mental deficiency,

promoted by the direct bioaccumulation of Pb have been recognised for millennia (Keller 1992). Both the Greeks and Romans made extensive use of Pb in the form of water pipelines, storage vessels and cooking and drinking utensils (Keller 1992). Some scientists have maintained that Pb assimilation resulted in the decline of the ruling classes of both empires. Increased rates of sediment supply to estuaries can result from natural flood cycles. However, accelerated sedimentation can also be promoted by anthropogenic practices causing increased rainfall runoff. Such practices include initial deforestation and clearing of the catchment, and increased rates of erosion and transportation of sediment as a consequence of urban expansion, for example the construction of roads, freeways and housing developments. Among the adverse environmental impacts of increased sediment influx into lagoonal systems are degradation of fringing seagrass beds and the siltation of navigation routes. In addition, increased turbidity associated with hydrological disturbance of sediment in shallow, peripheral bays can reduce the amount of light available for photosynthesis by benthic flora. Siltation of waterways can be ameliorated by dredging, although dredging programs need to be carefully designed to avoid potentially adverse environmental impacts of dredge-spoil disposal. Numerous studies in Europe and the USA have documented the environmental impacts of dredge-spoil disposal on land (Van Driel & Nijssen 1988), including salinisation and decline in groundwater quality, release of heavy metals and trace-metal assimilation by the food chain.

228

B. E. Chenhall etal.

Figure 1 9 . 1 Location of Lake Illawarra, lagoon bathymetry and the distribution of marine and deltaic sand deposits.

This chapter synthesises a number of research papers on Lake Illawarra which focus on trace-metal contamination and retention by the lagoon sediments, the rates of sediment accumulation prior to and during European settlement, and management strategies for this waterbody. The role of nutrients, such as N and P, in facilitating the growth of algal blooms and the environmental consequences of algal proliferation are not covered in this chapter. Lake Illawarra (Figure 19.1) is an almost landlocked coastal lagoon located 80 km south of Sydney, near Wollongong, in New South Wales. The lagoon is a shallow (<3.7 m), elongate waterbody, normally connected with the ocean by a single, shallow (<2 m), narrow (20 m), sinuous tidal channel 2.4 km in length. Perennial creeks, sourced in the steep Illawarra escarpment 9-14 km west of the lagoon, plus numerous smaller, local creeks deposit sediment into the lagoon. Prograding deltas are associated with all creeks, the most conspicuous of which is the active birdsfoot delta of Macquarie Rivulet in the southwest corner of the lagoon. Lake Illawarra originated by the renewed growth of a Pleistocene coastal sand barrier (Windang Peninsula: Figure 19.1) across the eastern side of a shallow embayment and subsequent flooding of the captured river plain (Yassini & Jones 1987, 1995). The lagoon would have had its maximum extent about 6500 years ago during the peak of the Holocene sea-level rise. The tidal range in this waterbody is -3.3 cm (Eliot et al 1976), apart from locations near the lagoon entrance (Figure 19.1). Thus the lagoon exhibits a microtidal environment with diurnal fluctuations in the order of 2% of spring tides in the adjacent open ocean (Eliot et al 1976). The hydrodynamic characteristics of this system are restricted lagoon-ocean water exchange and periodic freshwater flushing of the surficial layer during major rainfall events. The entrance channel shows seasonal positional switching and in the last 50 years shoaling at the mouth has resulted in the complete closure of the lagoon on a few occasions; such an event occurred most recently

in 1994 during a period of extended drought. The physical and hydrological characteristics of the lagoon result in effective sediment retention except during the periods of exceptionally high rainfall and flooding when suspended sediment load is flushed out to sea. The catchment was first settled in 1817, and in the period 1817-1993 approximately 47% and 20% of the catchment were converted to agricultural and urban purposes, respectively, by removal of native vegetation. Between 1896 and 1906, the Smelting Company of Australia operated a custom base-metal processing plant on a site near Kanahooka (Figure 19.1). Currently, the northern hinterland of the lagoon is host to the diversified Port Kembla industrial complex, encompassing the largest steel works in the southern hemisphere (BHP Steel International) which commenced operations in 1928, a base-metal refining plant (Southern Copper) from 1910 to 1995, grain and coal handling facilities and chemical fertiliser works. A pulverised coal-fired power station (Tallawarra Power Station), located on the western foreshore of the lagoon (Figure 19.1) commenced operations in 1954 and was decommissioned in 1989. The western and southwestern sections of the lagoon foreshore and catchment are experiencing rapid urban expansion; the lagoon is now almost entirely ringed by urban and industrial development.

METHODS Lagoon bottom samples were collected using 10% HC1leached, distilled water rinsed, with 90 mm diameter PVC pipe as the coring device. Cores recovered by this technique generally ranged from 60 to 70 cm in length, although core lengths of up to 1.3 m were obtained in muddominated substrates. Compaction during the coring process varied, depending on the substrate grainsize. Sanddominated substrates exhibited a compaction of approximately 6%, whereas mud-dominated cores from the deeper parts of the lagoon showed a mean compaction of 11%. Immediately after extraction of the core barrel from the sediment, each core was air extruded on board the vessel onto a 100 mm PVC hemicylinder lined with polyethylene. Any disadvantages, for example further core distortion, during the core-extraction procedure were considered of secondary importance, since this method allowed determination of pH and redox potential on freshly extracted samples. This concept is most important to meaningful studies of estuarine materials, because laboratory measurements confirmed the 'reactive' nature of the sediment when exposed to the atmosphere, with pH and redox potential rapidly shifted towards more acidic and oxic values respectively. After measurement of pH and redox potential, the outer sediment in contact with the core barrel was removed and the cores were subdivided into 5 cm intervals using a metal-free spatula. This important step was necessary to ensure that the possibility of sample contamination by core drag or contact with the barrel is greatly reduced. The sediment samples were then placed in sterile plastic vials on ice (essential for preserving organic matter) and transported to the laboratory where the measurements of grainsize, ash

Trace metal pollution, Lake lllawarra NSW

229

deposition of coarser particulate matter, probably represent domains where the rate of sediment accretion exceeds 1 cm y^ (Chenhall et al 1994). Below 50 cm depth, typical ranges for Zn, Pb and Cu in sand-dominated sediments are 14-33 |xg g" , 2-6 fig gr arid 18-23 |xg g , respectively. Enrichment factors (EF) for the mud-dominated sediment can be derived using the expression: 1

£P _

Figure 19.2 Distribution of average values for zinc in the top 20 cm of the sediment column. The stippled area represents sanddominated substrate; the clear portion mud-dominated substrate. Contour values are in pg g" on a dry weight basis. 1

content, combustible organic matter (COM) and trace-element content were made. More detailed information concerning the laboratory procedures used in these determinations can be found in Payne et al (1997), and ash speciation and content in Chenhall et al (1994). Valves of the mollusc Notospisula trigonella were retained for radiocarbon dating. TRACE-ELEMENT ABUNDANCES IN LAKE ILLAWARRA

Trace-element abundances in Lake lllawarra sediments have previously been documented by Roy and Peat (1974), Jones et al (1976), Ellis and Kanamori (1977) and more recently by Chenhall et al (1994), Batley and Chenhall (1995) and Chenhall et al (1995). Batley and Chenhall (1995) concluded that the most abundant trace elements of environmental significance were Zn, Pb and Cu. Other toxic trace metals, for example Cd and Hg, are present in very low concentrations of less than 3 |mg g ; the natural concentrations of other trace metals, for example As and Cr exceed 10 |xg g" although no significant enrichment of these elements is present in the near-surface sediments. Payne et al (1997) carried out a comprehensive study (40 core sites) of the spatial distribution of Zn, Pb and Cu in this lagoon. The following summarises their findings. Established background (i.e. natural) concentrations in the muddy lagoonal sediments (with standard deviations) are 68 ±13 |xg g" , 17 ±5 |jig g" and 33 ±8 fig g" for Zn, Pb and Cu, respectively. Background concentrations for sanddominated sediments could not be established with any confidence because of the high sedimentation rates. For example, sites such as Macquarie Rivulet, characterised by -1

-1

mean concentration of trace metal in the upper 20 cm of sediment background concentration (below 45 cm)

With the exception of the southern portion of Griffins Bay (close to the Port Kembla industrial complex: Figure 19.1), where the EF for Zn, Pb and Cu were greater than 5.9, 3.5 and 1.8, respectively, sediment in the lagoon exhibited low enrichment factors. Indeed Zn concentration was enriched by a factor of between 2 and 2.5, Pb by a factor of up to 2.3, and Cu by 1.6 times background. These factors contrast strongly with data for some more highly polluted east coast Australian lagoons. For example, at the northern end of Lake Macquarie, the continuous operation of the Cockle Creek Smelter has resulted in Zn and Pb contamination of up to 1-2 orders of magnitude greater than in Lake lllawarra (Roy & Crawford 1984; Batley 1987). The southern portion of Lake Macquarie exhibits much lower EF, reflecting greater distance from the Cockle Creek Smelter point source. In contrast the lightly urbanised catchment of Burrill Lake (Chenhall et al, 1992) is characterised by lower EF compared with Lake lllawarra. ANTHROPOGENIC TRACE-METAL LOADING AND RETENTION IN LAKE ILLAWARRA

Figure 19.2 illustrates several important concepts concerning the spatial distribution of trace elements in the lagoon. Mud-dominated sediments in the southern portion of Griffins Bay contain the greatest concentrations of Zn, and Payne et al (1997) established a similar distribution pattern for Pb and Cu. Trace-metal loading in Griffins Bay reflects the proximity of this site to point-source inputs from the industrial complex. Atmospheric dispersal of particulate-bound trace metals from current and past industry, for example fly-ash emissions from Tallawarra Power Station, probably account for the relatively uniform distribution of Zn in the central, deeper, mud-dominated portion of the lagoon. Zinc distribution additionally is related to sediment grainsize, the coarser sediment fraction being characterised by lower Zn concentrations. A statistical evaluation of the more significant relationships between some of the measured variables is provided in Table 19.1. These data are interpreted: (i) positive, moderately strong correlations between Zn, Pb and Cu indicate that the trace elements have a common, though not necessarily unique, origin and mode of occurrence; (ii) weak, negative correlations between depth and each of Zn, Pb and Cu are a reflection of low trace-metal enrichment factors (EF < 2.5) and between-site variation in the depth to which trace-metal enrichment extends; (iii) organic matter content is negatively correlated with sand and positively correlated with Rb and the finer sediment fractions; and (iv)

230

B. E. Chenhall et al.

Table 1 9 . 1 Significant product moment correlation coefficients at the 99% probability level for Lake Illawarra sediment data. Depth

% sand >63 pm

% coarse silt % fine silt 10-63 pm

4-10 pm

% clay < 4 pm

-0.28

% coarse silt

-0.21

% fine silt

-0.62

-0.31

% clay

0.22

-0.74

-0.31

0.38

-0.38

-0.3

0.23

0.20

-0.27

-0.33

0.21

% sand

Zn Rb

0.72

0.69

0.73

Redox

0.26

-0.90

0.20

0.60

0.62

0.35

0.36

0.26

Redox

0.39

-0.21

-0.27

-0.34

-0.31

%COM

-0.96

0.69

0.78

0.53

0.69

0.56

0.94

COM, combustible organic matter.

organic matter content is also positively correlated with the trace metals. These interrelationships, in general, lend support to the concept that organic-rich, fine-grained substrates are favourable for trace-metal accumulation and retention (Kersten 1988). Furthermore, this analysis demonstrates the potential use of 'conservative' elements (naturally occurring Rb; not introduced by anthropogenic activities) as monitors of grainsize-related variations in trace-metal abundances. According to Bourg (1988), pH and redox potential together with microbial activity (Kersten 1988) are important physicochemical parameters influencing metal mobility versus fixation in estuarine environments. A well-developed redox boundary is present in the upper 2-4 cm of the sediment column in Lake Illawarra substrates and sedimentbound water exhibits a uniform pH regime, with pH >7.6. Under oxic, alkaline conditions (above the redox boundary), Fe and Mn are present in the trivalent (hence the redbrown colour of the oxic sediment) and tetravalent states, respectively. Trace elements, can be bound (surfaceadsorbed: Bourg 1988) to Fe and Mn oxyhydroxides formed under such oxic conditions where pH >7. Below the redox boundary, reduction of Fe 3 + to Fe 2 + and Mn 4 + to Mn 2 + (with both forms being soluble) potentially can result in release of surface-adsorbed trace metals (Bourg 1988). However, trace metals and Fe in anoxic sediments generally undergo diagenesis via sulfate-reducing bacteria (Beveridge 1989; Morse 1994) with the consequent formation of metal sulfides (biotic fixation), including framboidal pyrite. The development of abundant framboidal pyrite (Yassini et al 1995) in Lake Illawarra sediments is probably a manifestation of such activity. Trace-metal retention in the sediment is thus favoured by both the physicochemical and biological processes operating in Lake Illawarra. Although the trace-element concentrations of the Lake Illawarra sediments are generally below Dutch Guideline criteria for the requirement of either continuous monitoring or intervention by dredging and secure disposal of the dredge spoil, transfer of these sediments from an anoxic to an oxic environment can result in detrimental, relatively short-term environmental effects. Dredging and dumping of lagoonal sediments onto the foreshore of northern Griffins Bay to create a wetland filter system resulted in the development of acid sulfate soils (pH <4) due to the production of sulfuric

acid through framboidal pyrite decomposition during oxidation. These conditions are not favourable to the establishment of native flora and the potential adverse environmental impact of consequent trace-element release back into the aquatic system remains unresolved at the present time.

SOURCES OF TRACE METALS IN LAKE ILLAWARRA Although trace metals derived from diverse sources, for example Zn from galvanised iron, Pb from the combustion of petrol-Pb additives, and trace metals affiliated with domestic effluents (Wittman 1981), can contribute to tracemetal contamination in estuarine systems, potentially a prime source of trace metals in Lake Illawarra resides in the dispersal of point-sourced, industrially derived particulate matter (ash) from present and past base-metal refining sites, iron and steel making and coal-fired electric-power generation. The State Pollution Control Commission (Anon. 1986) estimated that total particulate emissions from the Port Kembla industrial complex exceeded 40 000 t in 1986. Trace-element concentrations in ceiling dusts from dwellings located close to industry lie in the ranges 0.09-6.9%, 0.03-0.5% and 0.01-2.8% for Zn, Pb and Cu, respectively (Anon. 1993). Chenhall et al (1994) and Yassini et al (1995) analysed ash extracted from the lagoon sediments. These workers concluded that four major ash types, siliceous (Si-Al), carbonaceous, sulfide and iron oxide ash were present in the sediment. Some of the carbonaceous components, such as charwood and coal, together with iron oxides can be derived from both natural and anthropogenic sources (Chenhall et al 1994). However, other ash types have unique sources. Much of the siliceous ash (fly ash) present in the sediment in the western embayments of the lagoon was generated during the operation of the pulverised coal-firing utility (Tallawarra Power Station). Similarly, pyrometallurgical graphite fume is generated exclusively during iron and steel manufacture. Chenhall et al (1994) demonstrated that the trace-element concentrations of the sediments could be correlated with ash abundance in the sediment. Potentially, as indicated below, the depths to which uniquely sourced forms of ash extend in the sediment can provide valuable insight into sediment age and the rates of sediment accumulation.

Trace metal pollution, Lake lllawarra NSW

(a)

Concentration 1 ^gg"

100

200

Ash % 5 10 1 I V /V / / / / //////////A

15 L_

zzzzzzszzzm

3 Z

Figure 19.3 (a) The vertical distribution of zinc (total length of upper bar), lead (black portion of bar) and copper (lower bar with diagonal hatching) concentrations in a mud-dominated core from the northern end of Lake lllawarra. The concentration of the 'conservative' element rubidium is depicted by the heavy line. All concentration data expressed on a dry weight basis. Arrowhead marks the position of the redox boundary, (b) Ash percentage in the >63 mm sediment fraction from Yallah Bay showing the proportion of siliceous (close hatching) to carbonaceous ash. Depth scale as in (a)

SEDIMENT CHRONOLOGY Rates of sedimentation in Lake lllawarra prior to European development were estimated by radiocarbon dating methods on shells (Notospisula trigonella) preserved in the sediment (Chenhall et al 1995). Chenhall et al (1995) summarised the limitations attached to this dating technique including its limited applicability to sediments less than 200 years in age and the potential for erroneous 14C ages if older carbon, in the form of reworked, redeposited shells are incorporated in the sediment. These limitations notwithstanding, a relatively consistent picture of preEuropean sediment accumulation emerged from this work indicating that sedimentation rates in the deeper mud zone of the lagoon were of the order of <1 mm y 1 . Significantly, Modern (<200 years) radiocarbon ages are obtained from shells at depths of 1 m around the sandy deltaic margins of the lagoon, implying that the rates of sediment accumulation since European settlement are >5 mm y~l. Commonly the chronology of sediments deposited in the last 120 years can be determined by construction and modelling of 2 1 0 Pb (Smith 1982) and 137 Cs (Loughran & Campbell 1983) depth-decay curves. 137 Cs dating was dialed for Lake lllawarra deltaic sediments in 1990; the results of these trials were somewhat disappointing, primarily due to non-measurable 137 Cs activity in some cores, possibly due to low clay content and very rapid sediment accumulation rates. 2 1 0 Pb and 137 Cs dating techniques are

231

currently being re-evaluated for the deeper portion of this lagoon. Recent sediment chronology in Lake lllawarra has been assessed by interpretation of trace-metal concentration-depth and ash concentration-depth profiles. Figure 19.3a shows a typical trace-metal profile for mud-dominated sediment collected in 1994 from the northern end of the lagoon. Above background concentrations of Zn and Pb in the upper 45 cm of sediment can be equated with sediment contamination by adjacent industry, indicating a sedimentation rate of between 5 and 6.5 mm y 1 depending on selection of a 1910 (Southern Copper) or 1928 (BHP Steel) date for calculation. The lower concentrations of Zn and Pb below 45 cm are not related to a change in sediment type, for example increase in sand and consequent decrease in trace-metal values, for the Rb profile, a valuable monitor of clay sediment grainsize, is relatively constant. Using this approach it is estimated that typical rates of sedimentation of between 3 and 5 mm y~l over the last 70-90 years characterise the northern and north-central portions of the lagoon. An obvious limitation of this approach is that the introduction of trace metals (Zn and Pb) into the sediment can be facilitated by more than one source operating over different time frames. Readers should consult Batley (1987), Kilby and Batley (1993) and Chenhall et al (1994) for a more rigorous evaluation of the limitations of this method, including the potential for biological and physical disturbance of the sediment, diagenetic remobilisation of metals in the sediment column and the difficulties attached to interpreting trace metal-depth profiles in sand-dominated sediment. Ash, uniquely sourced from Tallawarra Power Station, is relatively abundant in sediment in the western and southwestern embayments of Lake lllawarra. The vertical distribution of ash in the lagoon appears to be relatively unaffected by physical (e.g. wind waves), and biological processes (e.g. significant redistribution by burrowing organisms). Thus the presence of fly ash indicates deposition of this sediment after 1954. Figure 19.3b shows the distribution of ash in a sediment core collected in 1990 from Yallah Bay, adjacent to the power station. Ash extending to a depth of 40 cm in this core indicates a sedimentation rate of 11 mm y - 1 . Sedimentation rates in Koona Bay are similar and adjacent to the active birdsfoot delta of Macquarie Rivulet, the sedimentation rate exceeds 16 mm y - 1 . Very recent sedimentation rates in excess of 1 cm y^1 thus appear to be typical of the western and southwestern embayments where active development is occurring.

CONCLUSIONS AND MANAGEMENT STRATEGIES FOR LAKE ILLAWARRA In Lake lllawarra, the low concentrations of toxic metals appear to present few short-term, serious environmental problems, although recent data for trace-metal accumulation by benthic flora and fauna are sparse (Chenhall et al 1992). Redox and pH conditions, high organic matter content, together with biologically mediated processes such as the formation of metal sulfides (predominantly pyrite) by bacterial reduction of seawater sulfate, favour metal retention in the lagoonal sediment. Disturbance of the sediment by dredging and foreshore reclamation work should be

232

B. E. Chenhall

etal.

carefully monitored because of the potential for acid sulfate soil formation, with consequent release of trace metals back into the aquatic realm. Information presented in this chapter indicates that very recent sedimentation rates are much greater than the rate of <1 mm y 1 , for the period prior to European settlement. In some portions of Lake Illawarra at least 1 m of sediment has accumulated since the catchment was modified by Europeans. Should this rate of sedimentation continue unchecked, then adverse environmental impacts, for example continued heavy shoaling of shallow bays adjacent to urban centres will contribute to further decline in water quality and reduction in aesthetic appeal of the lagoon margin. In the last decade, Wollongong City Council has employed an active policy of revegetation of portions of the catchment. This program should have direct beneficial effects, since it has the potential to reduce the rates of sediment erosion and transportation. Sediment retention in the terrestrial portions of the catchment has recently been facilitated by the construction of sediment retention ponds with a wetland filter component. These filters were mostly sited on small creeks at the northern end of the lagoon; no sediment retention structures have been developed on the larger rivers providing most of the sediment influx into Lake Illawarra. Wetland filters allow both sediment retention and the extraction of nutrients, although their efficiency in trapping the finer, clay-sized particles should be monitored, especially during flood cycles. One logical way to improve water quality is to dredge the highly silted portions of the lagoon or areas where sediment is rapidly accumulating, although it is important to realise that dredging alone does not solve the problem of siltation, it is merely a short- to medium-term remedial action. Any program of extensive dredging should be very carefully planned to ensure that key environmental issues, such as the destruction of benthic floral and faunal habitat, the potential development of acid sulfate soils and tracemetal release, are fully considered. This chapter has focused on trace-metal accumulation and sediment retention in Lake Illawarra. Trace-metal contamination and the nutrient budget of many other east coast Australian lagoons needs to be addressed by future studies, since data are sparse. It seems fitting to conclude that many east coast New South Wales lagoons possess hydrodynamic characteristics similar to Lake Illawarra whereby natural processes tend to fill these temporary sedimentary basins. Sediment retention in these systems is probably the greatest challenge facing scientists, engineers and statutory authorities in the future management of these water bodies.

REFERENCES ANON. 1986. Wollongong/Port Kembla Pollution Control Study. State Pollution Control Commission of New South Wales, Sydney. ANON. 1993. Metals content of dust in the roofs of houses around the Port Kembla industrial area. Illawarra Public Health Unit, University of Wollongong and Illawarra Area Health Service, Wollongong. BATLEY G. E. 1987. Heavy metal speciation in waters, sediments and biota from Lake Macquarie, New South Wales. Australian Journal of Marine and Freshwater Research 38, 591-606. BATLEY G. E. & CHENHALL B. E. 1995. Trace element content of the lake

sediments. In: Yassini I. & Depers A. M. eds. Recent Sediments in Lake Illawarra: Implications for Management, pp. 57-77. Department of Geology, University of Wollongong and Illawarra Catchment Management Committee, Wollongong. BEVERIDGE T. J. 1989. Role of cellular design in bacterial metal accumulation and mineralisation. Annual Review of Microbiology 43, 147-171. BOURG A. C. M. 1988. Metals in aquatic and terrestrial systems: sorption, speciation and mobilization. In: Salomons W. & Forstner U. eds. Chemistry and Biology of Solid Waste, Dredged Material and Mine Tailings, pp. 3-32. Springer-Verlag, Berlin. CHENHALL B. E., BATLEY G . E., YASSINI I., DEPERS A. M . & JONES B. G.

1994. Ash distribution and metal contents of Lake Illawarra bottom sediments. Australian Journal of Marine and Freshwater Research 45, 997-992.

CHENHALL B. E., YASSINI I., DEPERS A. M . ET AL 1 9 9 5 . Anthropogenic

marker evidence for accelerated sedimentation in Lake Illawarra, New South Wales, Australia. Environmental Geology 26, 124-135.

CHENHALL B. E., YASSINI I. & JONES B. G. 1992. Heavy metal concen-

trations in lagoonal saltmarsh species, Illawarra region, southeastern Australia. The Science of the Total Environment 125,

203-225. ELIOT I. G . , YOUNG R. W . & CLARKE D . J. 1 9 7 6 . Lake hydrology. In:

Young R. W., Eliot I. G, Jones B. G., Harris M. & Turnbull E. S. eds. Illawarra Lake—An Environmental Assessment Project, pp. 41-50. Wollongong City Council and University of Wollongong, Wollongong. ELLIS J. & KANAMORI S. 1977. Water pollution studies on Lake Illawarra. III. Distribution of heavy metals in sediments. Australian Journal of Marine and Freshwater Research 28, 485-496.

JONES B. G . , ELIOT I. G . & DEPERS A. M . 1 9 7 6 . S e d i m e n t a t i o n in Lake

Illawarra. In: Young R. W., Eliot I. G., Jones B. G , Harris M. & Turnbull E. S. eds. Illawarra Lake—An Environmental Assessment Project, pp. 20-40. Wollongong City Council and University of Wollongong, Wollongong. KERSTEN M. 1988. Geochemistry of priority pollutants in anoxic sludges: cadmium, arsenic, methyl mercury, and chlorinated organics. In: Salomons W. & Forstner U. eds. Chemistry and Biology of Solid Waste, Dredged Material and Mine Tailings, pp. 170-213. Springer-Verlag, Berlin. KELLER E. A. 1992. Environmental Geology. MacMillan Publishing Company, New York. KILBY G. W. & BATLEY G. E. 1993. Chemical indicators of sediment chronology. Australian Journal of Marine and Freshwater Research 44, 635-647. LALIBERTE C., DEWAILLY E., GUGRAS S . ET AL. 1 9 9 2 . M e r c u r y contami-

nation in fishermen of the lower north shore of the Gulf of St. Lawrence (Quebec, Canada). In: Vernet J P. ed. Impact of Heavy Metals in the Environment, pp. 15-28. Elsevier, Amsterdam.

LOUGHRAN R. J. & CAMPBELL B. L. 1983. The determination of sedi-

mentation depth by caesium 137. Search 14, 157-158. MORSE J. W. 1994. Interactions of trace metals with authigenic sulfide minerals: implications for their bioavailability. Marine Chemistry 46, 1-6. PAYNE M . , CHENHALL B . E., MURRIE M . & JONES B . G . 1 9 9 7 . Spatial

variation of sediment-bound zinc, lead, copper and rubidium in Lake Illawarra, a coastal lagoon in eastern Australia. Journal of Coastal Research 13, 1181-1191. ROY P. S. & CRAWFORD E. A. 1984. Heavy metals in a contaminated Australian estuary—dispersion and accumulation trend. Estuarine, Coastal and Shelf Science 19, 341-358. ROY P. S. & PEAT C. 1974. Trace metals in Lake Illawarra. Geological Survey of New South Wales Report GS 1974/319. SMITH J. D. 1982. Lead 210 dating of sediments. In: Ambrose W. & Duerden P. eds. Archaeometry: An Australian Perspective, pp. 303-309. ANU Press, Canberra. VAN DRIEL W. & NIJSSEN J. P. J. 1988. Development of dredged material disposal sites: implications for soil, flora and food quality. In: Salomons W. & Forstner U. eds. Chemistry and Biology of Solid Waste, Dredged Material and Mine Tailings, pp. 101-126. Springer-Verlag, Berlin. WITTMAN G. T. W. 1981. Toxic metals. In: Forstner U. & Wittman G. T. W. eds. Metal Pollution in the Aquatic Environment, pp. 3-68. Springer-Verlag, Berlin. YASSINI I. & JONES B. G. 1987. Ostracoda in Lake Illawarra: environ-

Trace metal pollution, Lake lllawarra NSW mental factors, assemblages and systematics. Australian Journal of Marine and Freshwater Research 38, 795-843. YASSINI I. & JONES B. G. 1995. Foraminiferida and Ostracoda from Estuarine and Shelf Environments on the Southeastern Coast of Australia. University of Wollongong Press, Wollongong. YASSINI I., DEPERS A. M . & CAITCHEON G . 1 9 9 5 . Ash in t h e s e d i m e n t .

233

In: Yassini I. & Depers A. M. eds. Recent Sediments in Lake lllawarra: Implications for Management, pp. 34-56. Department of Geology, University of Wollongong and lllawarra Catchment Management Committee, Wollongong. Received 2 September 1997; accepted 9 June 2000

Geological Society of Australia Special Publication 21, 235-242

CHAPTER 20—Stable lead isotopes: application to sourcing of lead in the environment B. E. CHENHALL1, M, CHIARADIA4, B. L GULSON2 3, B. G. JONES1 AND A. M. DEPERS1 1 School of Geosciences,

University of Wollongong, NSW 2522, Australia.

2 CSIRO, Division of Exploration and Mining Geoscience, PO Box 136, North Ryde, NSW 2113, Australia. 3 Graduate School of the Environment

Macquarie University, NSW 2109, Australia. 13, CH-1211 Geneve 4, Switzerland.

4 Departement de Mineralogie, Universite de Geneve, Rue de MaraTchers

The stable isotopes of lead provide an insight into possible sources of natural and human introduced Pb to the surface environment. Limitations to resolution include variation in Pb isotopic composition with time from petrol Pb additives, isotopic mixing of different Pb sources, from the blending of ores, and the introduction of Pb from two or more sources having the same Pb isotopic signature. Two case studies from the lllawarra region show that stable Pb isotopes can be used to constrain possible sources of Pb pollution. Lead isotopic compositions preserved in slag at the Kanahooka smelter (1896-1906) site indicate initially blended ore from the Orange-Bathurst-Young province, Mt Morgan and Broken Hill with an increase in the treatment of Broken Hill ore over time. The 8 km2 halo of Pb contamination around the smelter site is best explained by the atmospheric dispersal of Pb sourced from the Proterozoic Broken Hill deposits. In contrast, sediments in Lake lllawarra contain a mixture of Pb that can be modelled from one natural source (eroded Palaeozoic rocks) and four anthropogenic sources: leaded petroleum products, the Kanahooka smelter, the Southern Copper smelter and coal-utilising industries (BHP Steel and Tallawarra Power Station). KEY WORDS: anthropogenic factor, Lake lllawarra, lead isotopes, pollution, smelters.

INTRODUCTION In recent years, scientific and public awareness concerning the impact of the toxic heavy metal Pb on the natural environment has increased significantly (Keller 1992). This awareness is reflected in recent literature dealing with the health hazards posed by Pb to humans (Gulson & Wilson 1994; Gulson et al 1994, 1995b). Anthropogenic and natural addition of Pb to ecosystems has been recognised on a global basis from studies which determine the Pb distribution in remote, relatively 'pristine' regions, including Greenland (Rosman et al 1993, 1994a), Antarctica (Rosman et al 1994b) and the Arctic (Rosman & Chisholm 1994). It is generally accepted that Pb loading in these remote areas is derived from atmospheric addition of aerosols generated by the combustion of Pb-alkyl additives in petroleum-derived products (Sturges & Harrison 1986). This contention is reinforced by considerations involving temporal variation in both abundance and isotopic composition (i.e. variation in isotopic mass ratios) of Pb in ice cores (Rosman et al 1994a) and in oceanic waters (Veron et al 1994). On a regional or local scale, urban and industrial development and expansion can make an important and significant contribution to Pb loading in the environment (Mukai et al 1993; Graney et al 1995). Naturally occurring stable Pb isotopes (non-radiogenic 204pb; 206pb originating from the decay of 2 3 8 U; 2 0 7 Pb from the decay of 2 3 5 U; and 2 0 8 Pb from the decay of 232 Th), are routinely measured as ratios by mass spectrometry (Gulson 1986). Stable Pb isotopes have been

used by geoscientists to place age constraints on ore deposits and as a geochemical exploration tool (Gulson 1986). The potential use of Pb isotopes as a 'fingerprinting tool' for the identification of sources of Pb pollution in ecosystems, has been recognised for a few decades (Chow & Johnstone 1965). This approach is feasible if naturally occurring and anthropogenically introduced Pb have significantly different, homogeneous isotopic ratios that are known with confidence. An attractive feature of this approach is that Pb isotopes are generally unaffected by industrial processing or biological recycling and therefore retain the isotopic composition of their source (Ault et al 1970). The power of Pb isotopes in resolving the origin of Pb pollution is substantially reduced if (i) the Pb ore deposit source has a heterogeneous isotopic signature, for example those with high U/Pb; (ii) Pb from different sources is mixed either during industrial processing or by earlier geological processes, such as hydrothermal transport; and (iii) two or more different lead sources introduced over different time frames into an ecosystem have similar isotopic compositions. Numerous studies in the northern hemisphere have documented Pb pollution histories provided by the isotopic record preserved in lake and estuarine sediments (Flegal et al 1987; Ritson et al 1994; Gobeil et al 1995; Graney et al 1995). Equivalent studies in the southern hemisphere, particularly in Australia, are largely lacking. This chapter focuses on two case studies in the lllawarra region, New South Wales, which attempt to decipher the historic record of Pb pollution.

236

B. E. Chenhall etol.

Pacific Ocean

Figure 2 0 . 1 Location map of the Lake Illawarra area, approximately 80 km south of Sydney, showing the position of major pollution sources. SC, Southern Copper; cross hatched area, industrial complex; diagonal hatching, urban development; arrow, location of slag deposit at Kanahooka smelter (Figure 20.2); star, location of sediment core LI6 (Figures 20.5, 20.6).

KANAHOOKA (DAPTO) SMELTER In 1895, the Smelting Company of Australia established a custom base-metal smelter on the western foreshore of Lake Illawarra, adjacent to Mullet Creek (Figure 20.1). This enterprise heralded the initiation of metallurgical processing in the Illawarra region and, in the past, the smelter has been referred to either as the Kanahooka or the Dapto smelter. During its relatively short lifetime (1896-1906) the plant treated ores from diverse sources including the company's own mines at Orange and New England, together with ores from Broken Hill, Cobar and Captains Flat in New South Wales, Chillagoe and Mt Morgan in Queensland, Kalgoorlie in Western Australia and Zeehan in Tasmania (O'Malley 1968). During the peak of production ca 1901-1902, the smelter employed between 400 and 500 persons (O'Malley 1968). The closure of the smelter was swift, and promoted by both decline in metal values and the sudden withdrawal of the supply of Pb concentrates from Broken Hill and gold concentrates from

Kalgoorlie (O'Malley 1968). In 1907, most of the plant infrastructure was acquired by the Electrolytic Refining and Smelting Company of Australia Pty Ltd and was transported to Port Kembla. The smelting works was taken over by Southern Copper which ceased production in 1995. The plant is due to reopen in 2000 as Port Kembla Copper. Today the abandoned Kanahooka smelter site is partially occupied by residential development. Unpublished consultant reports have documented the removal of topsoils with toxic trace metals, including Pb, As, Cd and Hg, from the western portion of the site for containment behind earthen and stone barriers as part of the development scheme. Slag deposits, up to 5 m thick are extensively developed in the eastern, non-residential section of the smelter site. Approximately one half of the slag has been removed for road base and fill. The remaining slag has been placed under a heritage order. Research into the potential usage of stable Pb isotopes to monitor the sources of Pb pollution, slags from soils and carbonaceous ash residues (fume) has been concentrated on eastern portion of the smelter site. These data are presented in Tables 20.1 and 20.2. Elemental concentrations were determined by X-ray fluorescence spectrometry (XRFS) and instrumental neutron activation analysis (INAA). Pb isotopic data were obtained by high-precision thermal ionisation mass spectrometry (TIMS: Gulson 1986). Major base-metal concentrations in the slags have ranges of 2.6-5.1%, 0.07-0.4% and 0.25-1.84% for Zn, Cu and Pb respectively. The high concentrations of Zn in the slags were a reflection of the 'relative inefficiency' of early pyrometallurgical smelting processes involving the treatment of mixed Pb-Zn ores until alternative methods, for example ore concentration by selective flotation, were developed. Additionally, the slags contain significant concentrations of potentially toxic trace elements, including As and Sb. It is important to realise that these metals are 'locked' in the slag, unless they are released into the environment either by slag degradation through natural processes such as weathering, or by human intervention, for example, comminution of the slag (crushing) that provides finer grained, 'reactive' particles for decomposition. Isolated deposits of carbonaceous ash (Table 20.1) containing high concentrations of Sb, As (3.3%) and Hg are still evident in the demolished processing section of the plant. All contaminated soils contain Pb ranging from 76 jig g"1 to in excess of 3500 pg g" 1 , with Pb concentration generally

Figure 2 0 . 2 Cross-section through portion of the slag deposit at the Kanahooka smelter showing sample locations and 2 0 6 P b / 2 0 4 P b of the slags. Sample S9 is from the surface of the slag.

Sourcing of lead in the environment

237

Table 20.1 Chemical composition of selected soils, fume and slags from the Kanahooka smelter site. Sample

Soils and fume KFG8 479 KFG9a 636 KFG11 2322 KFG12b 641 KFG14 294 KFG15 93 KFG17 2722 KFG18 94 KFG19 353 KFG25* 4.52% KFG26 290 Slags S3 8069 S4 9836 4996 S5 1.15% S6 S7 8268 1.84% S8 6026 S9

131 191 263 56 172 135 n.a. 48 170 808 146

166 308 375 43 202 136 560 119 187 983 170

8 11 37 2 10 1 173 2 7 1890 5

91 100 333 13 304 14 1120 13 79 3.30% 42

11 10 10 3 24 24 9 5 18 11 25

56 131 333

2781 4077 2760 3584 3102 715 1192

5.09% 4.93% 4.51% 2.66% 3.53% 3.50% 4.06%

127 54 120 136 75 115 348

91 51 66 59 168 235 157

51 62 62 67 204 82 39

46 46 424 24 69 5320 33 -

12 17 13 -

44 -

74 -

17 14 -

6 -

* indicates fume sample containing 611 pg g Hg Analysis by Instrumental Neutron Activation Analysis at Becquerel Laboratories, ANSTO facility, Lucas Heights, NSW, except for Cu, Pb and Zn, determined by X-ray fluorescence analysis. Trace-element data expressed in pg g"1 (ppm); Au values in pg g - 1 (ppb); - indicates below analytical detection limits. -1

greater than Cu and Zn. The halo of Pb contaminated soil is extensive, encompassing an area of 8 km 2 . Given the diverse sources of processed ore, the fact that some treated ores originated from unspecified sources (O'Malley 1968) and the probability that ores were blended (mixed prior to processing), it would appear at first sight, that minimal useful information could be obtained from Pb isotope studies (Table 20.2) seeking to 'fingerprint' the sources of Pb at the smelting site. Figure 20.2 is an illustration showing the disposition of slag deposits together with analysed Pb isotopic composition. These deposits can be interpreted from a geological perspective to become younger from west to east since the layering in the 'slag stratigraphy' indicates the timing of slag dumping. Furthermore, near-horizontal slag layers on top of the dipping sequence post-date the underlying dipping sequence. Lead isotopic data indicate that there is a progressive shift with time to slags containing a geologically 'old' Pb component with lower 2 0 6 P b / 2 0 4 P b ; the only known utilised source of this Pb is from Broken Hill (O'Malley 1968). A more rigorous appraisal of the Pb isotopic data is presented in Figure 20.3. It is apparent from this diagram that the Pb isotopic composition of the slags plot in two compositional fields: 'older' slags tend to cluster towards the Palaeozoic deposits of the Lachlan and New England Fold Belts whereas the 'younger' slags are dominated by the introduction of geologically older Broken Hill Pb. The isotopic composition of the slags plot between two mixing lines. This is indicative of a blending of ore sources between Broken Hill Pb with relatively Pb-rich deposits of the Lachlan Fold Belt (Cobar and Captains Flat) and relatively Pb-impoverished deposits including Mt Morgan and Zeehan. It is apparent that ore blending can

obscure identification of ore source, however it is also apparent that: (i) slags S3 and S4 plot close to the Pb isotopic composition of Orange-Bathurst-Young deposits with which the smelting company had direct links; these slags do not contain Pb from Broken Hill, Cobar, Captains Flat or Mt Morgan because they plot outside the mixing triangle defined by these deposits; (ii) slags SI and S2 plot

0.80-|—— 2.05

2.10

2.15

208Pb/206Pb

2.20

T—

2.25

Figure 20.3 208 Pb/ 206 Pb vs 207 Pb/ 206 Pb diagram showing the isotopic compositions of the slags compared with isotopic compositions of Australian ore deposits. BH, Broken Hill; CF, Captains Flat; CH, Chillagoe; CO, Cobar; HP, Hall's Peak; KS, Kanahooka smelter; MM, Mt Morgan; OBY, Orange-Bathurst-Young type deposits; OL, Orange-Lucknow; W, Woodlawn; ZO, Zeehan Oceana; ZS, Zeehan Spray (Chiaradia et al 1997 table 1).

238

B. E. Chenhall et al.

Table 2 0 . 2 Lead isotope data for slag samples from the Kanahooka smelter site. Sample

208/206

207/206

206/204

Pb (pgg- 1 )

2.1199

0.8785

17.680

2500

2.1182

0.8781

17.644

2.0995

0.8619

18.083

8069

2.1006

0.8646

17.981

9836

2.1382

0.8925

17.376

4996

2.1431

0.8954

17.337

1.15%

2.1992

0.9394

16.419

8268

2.2224

0.9573

16.073

1.84%

2.2005

0.9402

16.397

6026

Pb determined by X-ray fluorescence analysis; na, not analysed.

close to the Pb isotopic signature of Mt Morgan; although it cannot be definitely established from Figure 20.3 whether the Pb is solely from this deposit or from blending of Zeehan and Broken Hill Pb, the 2 0 7 Pb/ 2 0 4 Pb and 2 0 8 Pb/ 2 0 4 Pb indicate that Pb in these slags has a very similar overall isotopic composition to Mt Morgan Pb; (iii) younger slags contain an increasing proportion of Broken Hill Pb; and (iv) Chillagoe Pb was not detectable in the analysed slags, which may simply indicate that slags containing Pb from this deposit were not sampled. The isotopic composition of lead preserved in the slags is essentially consistent with the cursory historical records of plant production (O'Malley 1968). These records indicate that during its zenith in the first six months of 1902, the plant produced 564 t of lead during the processing of Broken Hill lead concentrates and 26 t of copper. Complete historical records of smelter ore/concentrate receival and processing would allow development of a more comprehensive analysis of the slags. Residual, carbonaceous ash (fume: Table 20.2) preserved on the site has a characteristic Broken Hill Pb isotopic composition, indicating that during the latter stages of production most of the Pb ore came from this single source. Topsoils developed on the Permian (Palaeozoic) Dapto Latite (Carr et al 1996) at the site yield consistent 2 0 6 Pb/ 2 0 4 Pb isotopic values of 16.2-16.3, indicative of their contamination by geologically old (Proterozoic) Broken Hill lead. No isotopic data are available for background soils; however, the geological exposures are Permian in age and rocks of similar affinity cropping out to the south of the area (Milton Monzonite) are characterised by 2 0 6 Pb/ 2 0 4 Pb of 18.8. The extent of Pb contamination together with the isotopic uniformity of the contaminated soils is consistent with atmospheric dispersal of Pb, either in the form of ash or as fine particles from ore stockpiles during the processing of Pb derived principally from Broken Hill.

LEAD ISOTOPIC RECORDS IN LAKE ILLAWARRA Coastal lagoons, such as Lake Illawarra, dominated by restricted, terrestrial-marine exchange (Yassini & Jones 1995) are potentially fertile ground for detailed scientific studies involving Pb pollution of both the water and sediment column. These environments, together with land-

Figure 2 0 . 4 2 0 8 P b / 2 0 6 P b V5 2 0 7 P b / 2 0 6 P b diagram showing the isotopic compositions of sediments compared with isotopic compositions of potential Pb sources. A, air; BH, Broken Hill; BHP1, BHP Steel; CF, coal users (BHP Steel and Tallawarra power station); G, petrol; KS, Kanahooka smelter (soils and fume); MC, Mullet Creek sediment sample; PK, Port Kembla Harbour sediment; SC, Southern Copper; TPS, Tallawarra power station (Chiaradia et al 1997 table 1).

locked aquatic systems, provide instructive information on both the isotopic record of Pb pollution and sedimentation history (Batley 1987; Chenhall et al 1995).

Sources of lead Apart from naturally introduced Pb, derived from the weathering, erosion and transportation of rocks cropping out in the catchment, other sources including biomass combustion (Graney et al 1995), fossil-fuel burning and the combustion of leaded petrol (Corrin & Natusch 1979) can contribute to Pb loading of lagoons and estuaries. Specific past and current industries which potentially contribute to the lead burden in Lake Illawarra are the abandoned Kanahooka smelter referred to in the case study presented above, industries utilising coal as an energy source (Tallawarra Power Station, 1954-1989, and BHP Steel) and the Southern Copper smelter. These sources have been reviewed in more detail by Chiaradia et al (1997). Figure 20.4 displays the isotopic composition of the potential Pb sources together with compositional data for some Lake Illawarra sediments. The complete table of isotopic data is to be found in Chiaradia et al (1997). A bulk sediment sample from the inner harbour at Port Kembla is included for comparison. Limitations to completely resolving the sources of Pb in the sediments arise from the following: (1) Pb was introduced into petroleum products in Australia in 1948. Long-term isotopic records are not available for the composition of petroleum products in Australia. Gulson et al (1983) and Rosman et al (1994b) have established that petrol and aerosols had a relatively constant 2 0 6 Pb/ 2 0 4 Pb of -16.5 in the early 1980s and a range of 16.2-16.6 for the period 1991-1996 (Gulson et al 1995). However, Gulson et al (1983) indicated that the iso-

Sourcing of lead in the environment topic compositions of Pb in Australian petrol may have been more radiogenic (higher 2 0 6 Pb/ 2 0 4 Pb), at least in the period between 1980 and 1990. Pb in petrol additives sourced from geologically old (Proterozoic) Australian deposits (Mt Isa and Broken Hill) will have an isotopic signature similar to soils and some slags from the Kanahooka smelter (Figure 20.4). (2) Similarly, long-term isotopic records are not available for the Southern Copper smelter, although the isotopic homogeneity of smelter bag house dusts over a three year sampling period is demonstrable (Chiaradia et al 1997). Complete records of processing and production are not available, however, the measured isotopic compositions are compatible with the predominant use of isotopically similar, homogeneous, Palaeozoic Australian ore deposits during the life of the plant. This is reflected in the shift of 2 0 6 Pb/ 2 0 4 Pb towards either 'older' Pb from the recycling of geologically older Pb in scrap materials or the introduction of a minor component of more radiogenic (i.e. high U/Pb) from other sources. (3) Non-polluted (i.e. pre-industrial age) lagoonal sediments (herein termed background sediments) are characterised by 2 0 6 Pb/ 2 0 4 Pb close to 18.7 (Figure 20.4). The isotopic composition of this lead is similar to lead generated by industries which have utilised local Permian coals as an energy source. Clearly, some caution must be exercised in the rigorous evaluation of these data. However, it is apparent that all the lagoon sediments (and the sediment sample from Mullet Creek) lie inside a mixing triangle defined by geologically 'old' Pb of Broken Hill affinity, Southern Copper smelter Pb and background sediment Pb. This indicates that all of these sources contribute to the Pb burden of the lagoon sediments. Sediment from Mullet Creek (MC in Figure 20.4) has the lowest 2 0 6 Pb/ 2 0 4 Pb of 17.22. The most plausible explanation for the lowering of 206pjy204pb \s contamination of this sediment by Pb sourced from the Kanahooka smelter. By way of contrast, the isotopic composition of Pb in sediment from Port Kembla Harbour is consistent with the introduction of a predominant component of Pb sourced from coal utilisation such as BHP Steel. Later in this chapter, an example will be presented of isotopic modelling, leading to resolution of the fractional contributions of Pb from potential sources.

Lead isotope sediment-depth profiles The isotopic profile of a sediment core from Lake Illawarra (Figures 20.1,20.5) shows isotopic variations which are mirrored by changes in Pb concentrations (Chiaradia et al 1997 table 1). The Pb isotopic composition of all sediment below 45 cm in this core is fairly constant with a 2 0 6 Pb/ 2 0 4 Pb of -18.7. The constancy of both the concentrations of Pb in these sediments (-10 \xg g"1) and their isotopic compositions indicates that the Pb below 40 cm is from a natural source equating with Permian rocks cropping out in the catchment of the lagoon (Chiaradia et al 1997). Between 25 and 40 cm the 2 0 6 Pb/ 2 0 4 Pb of the Pb present in the sediments displays a linear decrease with decreasing depth. Assuming that the background source of Pb remains the same, in terms of both isotopic composition and rate of Pb supply, the decreasing 2 0 6 Pb/ 2 0 4 Pb indicates

16.8

17.0 17.2

17.4

206Pb/204Pb 17.6 17.8

18.0

18.2 18.4 18.6

239

18.8

Pb(|ig g-0 Figure 2 0 . 5 Pb concentration and variations of 206pb/204pb ratio with depth showing the amount and isotopic composition of anthropogenic lead (Pbant) added to the sediment in the Lake Illawarra core LI6. The sediment age in the upper part of the profile was determined by 1 3 7 Cs dating. A constant sedimentation rate of 5 mm y"1 was applied, yielding the ages shown. Pb contamination towards the base of the core (indicated by a shift in 2 0 6 Pb/ 2 0 4 Pb) is chronologically consistent with initiation of the earliest base-metal industries. S, G and P indicate the commencement of smelting, introduction of leaded petrol and commencement of Tallawarra power station, respectively.

an increasing contribution to the sediment Pb burden of one (or more) anthropogenic source(s) of Pb with a lower 2 0 6 Pb/ 2 0 4 Pb upwards in the sediment column. This isotopic shift is accompanied by a steady increase in the Pb content in the sediment. From a depth of 25 cm upwards, the 2 0 6 Pb/ 2 0 4 Pb is uniform at a value of 17.6, while Pb concentration is variable but always >30 pg g"1. Chiaradia et al (1997) have established that other nearsurface sediments in Lake Illawarra were isotopically homogeneous with 2 0 6 Pb/ 2 0 4 Pb ranging from 17.6 to 17.7. The most plausible explanation for the observed widespread isotopic uniformity in the near-surface sediments is contamination by an atmospherically deposited, isotopically homogeneous, though not necessarily unique source of Pb.

Modelling and identification of lead sources in a Lake Illawarra sediment core Calculation of the isotopic composition of anthropogenically introduced Pb is contingent upon the assumptions that the natural Pb background throughout the sediment profile has a constant content and isotopic composition. If these

240

B. E. Chenhall etol.

assumptions are reasonably valid, as suggested by Chiaradia et al (1997), then the Pb concentration and Pb isotopic composition distributions similar to that in Figure 20.5 can be modelled by a mass-balance equation to resolve the isotopic composition of the introduced Pb, ( 206 Pb/ 204 Pb) 1 : (

206p b /204p b ) iFi = (206pb/204pb)2p2 _ (206pb/204pb)

where F 2 and F 3 are respectively the total and background concentrations of Pb in the sediment, and Fx is estimated by subtracting the natural Pb component provided by nonpolluted sediments below 40 cm. Figure 20.5 summarises these calculations. Below a depth of 27 cm (prior to 1930) anthropogenic addition of Pb reflects the environmental impact of early base metal refining. Other industries, including BHP Steel and electric-power generation (Tallawarra power station), had yet to commence operation. Pb petrol additives had yet to be introduced in Australia. The isotopic composition of the added Pb is not unique to any early base-metal industry. Reduction of the introduced Pb to 2 0 6 Pb/ 2 0 4 Pb of 17.2 probably indicates mixing of Kanahooka smelter Pb with Pb sourced from copper refining (Southern Copper). The uppermost 20 cm contains added Pb with an uniform 206pb/204pb 0 f 17.0-17.2 (Figure 20.5). This isotopic composition again does not closely correspond with any single past or current industry. Chiaradia et al (1997), in an attempt to resolve the sources of this Pb, analysed ceiling dusts, containing significant amounts of industry-derived particulate matter, from dwellings in the Illawarra region. The isotopic composition of these dusts is also essentially uniform and closely corresponds to the isotopic composition of Pb added to near-surface Lake Illawarra sediments. On this basis, Chiaradia et al (1997) suggested that in the last five decades, atmospheric deposition of Pb-bearing particulate matter appeared to be the dominant pathway for Pb addition to this lagoon. These workers, using isotopic modelling similar to that outlined below, were also able to resolve that Pb in the ceiling dusts was a homogenised mixture containing (in addition to minor, natural background Pb) significant proportions of Pb from copper refining, coal utilisation and the combustion of leaded petrol. Isotopic information presented in Figure 20.4 can be modelled to solve the relative contributions of sources to the Pb burden of the lagoon. The following equations allow resolution of the relative contributions Fv F2 and F 3 from three sources (a three-source mixing model). F2 + F 2 + F 3 = 1 (

(2)

2 0 8 p b / 2 0 6 p b ) i F i + (208p b / 206p b ) 2 p 2 + ( 2 0 8 p b / 2 0 6 p b ) ^ = (

208p b/ 206p b)measured

(3)

(207p h / 206p b ) i p i + (207p b / 206p b ) 2 p 2 + ( 2 0 7 p b / 2 0 6 p b ) ^

= ( 2 0 7 Pb/ 2 0 6 Pb) m e a s u r e d

Pb(|ig g-0

(4)

Resolution of a fourth source is possible if a third measured isotopic ratio is added and the relative contribution of a fourth source, F 4 is accommodated. Figure 20.6 pre-

Figure 2 0 . 6 Modelled Pb contributions from natural background (Pbnat), air (Pbair) and the copper smelter (PbSC) in core LI6 from Lake Illawarra. Note that each point represents the mean value for a 5 cm sediment slice.

sents the results obtained for the Lake Illawarra core using a three-source model with end members being background sediments, Southern Copper and Pb in air. These calculations are only relevant to the three named sources. If another source, such as Broken Hill, is selected then the calculations would require revision. As indicated above, caution is advised with data interpretation due to the potential for Pb isotopic variation in sources over time. Nevertheless, some very useful information can be drawn from Figure 20.6. (1) A component of Pb in air (with isotopic composition approaching that of geologically old Pb), present below 22 cm depth (i.e. prior to 1948), represents Pb probably sourced from the Kanahooka site. This Pb component exhibits a pronounced increase post-1948. This trend is consistent with the introduction into Australia of Pb-bearing petroleum products. Owing to their potential isotopic similarity, the relative contributions of Kanahooka smelter and Pb in air cannot be assessed in the upper part of the core, unless a constant rate of supply in the form of sediment from the Kanahooka site is assumed. (2) A Pb component from the Southern Copper source is present at 40 cm. The presence of Pb with a Southern

Sourcing of lead in the environment Copper isotopic signature at this depth closely corresponds with establishment of the plant in 1910. Lead contribution from the copper smelter exhibits a progressive increase prior to World War II and a decline post-1960. The decline in Pb contribution from this source in the last three decades is probably attributable to the progressively more stringent pollution control measures introduced at the plant. (3) The background Pb contribution is relatively constant in the lower portion of the core, although there is a trend towards increasing background Pb from about 1900. This is possibly due to land clearing causing the release of larger quantities of clay-sized material. Decreased background contributions between 20 and 25 cm depth are probably attributable to grainsize variation in the sediment. The more sandy intervals in this part of the core contain a high proportion of industrial ash and consequently lower concentrations of natural lead. Post-1960, the background Pb contribution exhibits a pronounced increase, unrelated to grainsize variation in the sediment. This pattern could be attributed to natural variation in the input of Pb, although it is very probable that other sources, utilising materials of similar Pb isotopic composition to the background sediments (e.g. coal), are a factor in sediment contamination. These sources include BHP Steel and Tallawarra power station since fly ash attributable to these sources is present in the lagoon sediments from this interval (Chenhall et al 1995). Southern Copper utilises coal as an energy source, however, the abundant Pb present in most base-metal ore concentrates masks the low concentrations of coal-derived Pb from this source. It is significant to note that industry, particularly BHP Steel experienced rapid expansion post1960. Thus, plausible explanation for the higher background contributions of Pb can be found from isotopic modelling, together with historical records of industrial expansion.

CONCLUSIONS Unlike many metals, Pb has four stable isotopes, providing, not only a 'window' into the age of Planet Earth, but insight into the introduction of Pb via both natural and humaninduced change to the surface environment. The two case studies have demonstrated that it is not always possible in a scientific context to resolve the absolute contribution of all potential sources of the toxic metal lead in the environment. The limitations to resolution are generated by factors such as variation in Pb isotopic composition with time from petrol Pb additives, isotopic mixing of different Pb sources, from the blending of ores, and the introduction of Pb from two or more sources having the same Pb isotopic signature. Nevertheless, the two case studies presented in this chapter indicate that stable Pb isotopes can be utilised to place many constraints on the sources of Pb pollution. The 8 km 2 halo of Pb contamination at the Kanahooka smelter site is best explained by the atmospheric dispersal of Pb sourced from geologically old (Proterozoic) Broken Hill deposits. Lead isotopic compositions preserved in the slags, although indicative of the practice of ore blending, are consistent with cursory historical records indicating an increase

241

in the treatment of Broken Hill ore over time. Isotopic signatures of the Orange-Bathurst-Young province and Mt Morgan can be recognised in some older slag deposits. Sediments in Lake Illawarra indicate that one natural source (eroded Palaeozoic rocks) and four anthropogenic sources [leaded petroleum products, the Kanahooka smelter, the Southern Copper smelter and coal utilising industries (BHP Steel and Tallawarra power station)] have contributed to the Pb burden of the lagoon. Although the absolute contributions of these sources cannot be completely assessed, isotopic modelling of the sediment record is consistent with historical records of industrial development in the catchment of Lake Illawarra. Complete records of ore sources and production output would enable finer resolution of Pb sourcing in the Illawarra region.

REFERENCES AULT W . A., SENECHAL R. G. & ERLEBACH W . E. 1 9 7 0 . Isotopic c o m p o -

sition as a natural tracer of lead in the environment. Environmental Science and Technology 4, 305-313. BATLEY G. E. 1987. Heavy metal speciation in waters, sediments and biota from Lake Macquarie, New South Wales. Australian Journal of Marine and Freshwater Research 38, 591-606.

CARR G. R., DEAN J. A., SUPPEL D . W . & HEITHERSAY P. S. 1 9 9 5 . Precise

lead isotope fingerprinting of hydrothermal activity associated with Ordovician to Carboniferous metallogenic events in the Lachlan Fold belt of New South Wales. Economic Geology 90,

1467-1505. CARR P. F., PERKINS M . G., CHENHALL B . E . , PEMBERTON J. W .

MIDDLETON R. G. 1996. Heulandite from the Illawarra district, NSW. Australian Journal of Mineralogy 2, 71-75.

CHENHALL B. E., YASSINI I., DEPERS A. M . ET AL. 1 9 9 5 . Anthropogenic

marker evidence for accelerated sedimentation in Lake Illawarra, New South Wales, Australia. Environmental Geology 26, 124-135.

CHIARADIA M . , CHENHALL B. E., DEPERS A. M . , GULSON B. L. & JONES B.

G. 1997. Identification of historical lead sources in roof dusts and recent lake sediments from an industrialized area: indications from lead isotopes. The Science of the Total Environment 205, 107-128. CHOW T. J. & JOHNSTONE M. S. 1965. Lead isotopes in gasoline and aerosols of the Los Angeles Basin, California. Science 147, 502-503. CORRIN M. L. & NATUSCH D. F. S. 1979. Physical and chemical characteristics of environmental lead. In: Boggess W. R. & Wixson B. G. eds. Lead in the Environment, pp. 7-31. Castlehouse, Washington. FLEGAL A. R., ROSMAN K. J. R. & STEPHENSON M . D . 1 9 8 7 . Isotope sys-

tematics of contaminant leads in Monterey Bay. Environmental Science and Technology 21, 1075-1079.

GOBEIL C., JOHNSON W . K., MACDONALD R. W . & WONG C. S. 1 9 9 5 .

Sources and burden of lead in St Lawrence Estuary sediments: isotopic evidence. Environmental Science and Technology 29, 193-201.

GRANEY J. R., HALLIDAY A. N., KEELER G . J., NRIAGU J. O., ROBBINS J. A.

& NORTON S. A. 1995. Isotopic record of lead pollution in lake sediments from the northeastern United States. Geochimica et Cosmochimica Acta 59, 1715-1728. GULSON B. L. 1986. Lead Isotopes in Mineral Exploration. Developments in Economic Geology 23. Elsevier, Amsterdam. GULSON B. L., DAVIS J. J., MIZON K. J., KORSCH M . J., LAW A. J. &

HOWARTH D. 1995. Lead bioavailability in the environment of children: blood lead levels in children can be elevated in a mining community. Archives of Environmental Health 49, 326-331.

GULSON B. L., MIZON K. J. & KORSCH M . J. 1 9 8 3 . Fingerprinting the

source of lead in Sydney air using lead isotopes. In: Carras J. N. & Johnson G. M. eds. The Urban Atmosphere of Sydney—a Case Study, pp. 233-244. CSIRO, Melbourne.

GULSON B. L., MIZON K. J., LAW A. J., KORSCH M . J., DAVIS J. J. &

HOWARTH D. 1994. Source and pathways of lead in humans from the Broken Hill mining community—an alternative use of explo-

242

B. E. Chenhall

etal.

ration methods. Economic Geology 89, 889-908. History of lead exposure in children revealed from isotopic analyses of teeth. Archives of Environmental Health 49, 2 7 9 - 2 8 3 . KELLER E. A. 1992. Environmental Geology. MacMillan Publishing Company, New York.

GULSON B . L . & W I L S O N D . 1 9 9 4 .

MUKAI H . , FURUTA N . , FUJII T . , AMBE Y . , SAKAMOTO K . & HASHIMOTO T.

1993. Characterization of sources of lead in the urban air of Asia using ratios of stable lead isotopes. Environmental Science and Technology 27, 1347-1356. O'MALLEY J. P. 1968. The Old Dapto Smelting Works (2nd edition). Illawarra Historical Society, Wollongong. RITSON P. I., ESSER B. K., NIEMEYER S. & FLEGAL A. R . 1994. Lead isotopic determination of historical sources of lead to Lake Erie, North America. Geochimica et Cosmochimica Acta 58, 3297-3305. ROSMAN K . J . R . & CHISHOLM W . 1 9 9 4 . Determination of lead isotopes in Arctic and Antarctic ice and snow. Analysis 22, M 5 1 - 5 3 . ROSMAN K . J . R . , CHISHOLM W . , BOUTRON C . F., CANDELONE J . P.

U. 1993. Isotopic evidence for the source of lead in Greenland snows since the late 1960s. Nature 362, 333-335. GORLACH

ROSMAN K . J . R . , CHISHOLM W . , BOUTRON C . F., CANDELONE J . P. & HONG S . 1994a. Isotopic evidence to account for changes in the concentration of lead in Greenland snow between 1960 and 1988. Geochimica et Cosmochimica Acta 58, 3265-3269.

ROSMAN K . J . R . , CHISHOLM W . , BOUTRON C . F., CANDELONE J . P. & PATTERSON C . C . 1994b. Anthropogenic lead isotopes in Antarctica. Geophysical Research Letters 21, 2669-2672. W. T. & HARRISON R. M. 1986. The use of Br/Pb ratios in atmospheric particles to discriminate between vehicular and industrial lead sources in the vicinity of a lead works—I. Thorpe, West Yorkshire. Atmospheric Environment 20, 833-843.

STURGES

V £ R O N A . J . , CHURCH T . M . , PATTERSON C . C . & FLEGAL A . R . 1 9 9 4 . U s e

of stable lead isotopes to characterize the sources of anthropogenic lead in North Atlantic surface waters. Geochimica et Cosmochimica Acta 58, 3 1 9 9 - 3 2 0 6 . YASSINI I . & JONES B . G . 1 9 9 5 . Foraminiferida and Ostracoda from Estuarine and Shelf Environments on the Southeastern Coast of Australia. University of Wollongong Press, Wollongong. Received 2 September 1997; accepted 15 June 2000

Geological Society of Australia Special Publication 21, 243-54

CHAPTER 21—Role of geology in environmental science: an example of pollution assessment of the estuarine and marine environment in central New South Wales G. F. BIRCH, S. E. TAYLOR AND C. MATTHAI Environmental Geology Group, Department of Geology and Geophysics, University of Sydney, NSW, 2006, Australia. Geology is making an important contribution to environmental science, mainly in understanding the processes controlling natural hazards and in solving problems associated with adverse human impact on the Earth. The source, fate and effects of particle-bound contaminants are controlled by physical, chemical and biological processes that are well understood by environmental geologists. Research being conducted on Sydney's four major estuaries and on the adjacent continental shelf is providing environmental geologists with new insights on how contaminants behave in aquatic environments. Although poor past industrial practices have contributed greatly to the pollution of these environments, present-day discharges continue to add toxicants to the estuaries and to the offshore. The extent of adverse biological impact caused by these contaminants in the bottom sediment is difficult to judge, but initial studies indicate that impact may be substantial. Remediation of the estuaries will be difficult due to the extent of contamination and the large mixture of different types of toxicant. KEY WORDS: environmental geology, estuary, heavy metals, sedimentology, New South Wales, pollution, remediation, Sydney Harbour.

INTRODUCTION Aspects of geology, which exert an influence on pollution and waste disposal in the aquatic environment, are first reviewed in this chapter. Research carried out in the estuaries and on the continental margin of central New South Wales is subsequently used to demonstrate the important role of the geologist, and especially the sedimentologist, in contaminant science. Geology is an interdisciplinary science with interactions between the biosphere, hydrosphere and atmosphere and it is the training and the capacity of geologists to interact between disciplines that leads to the successful application of geology in environmental science. It is also the extent of this interdisciplinary endeavour which often determines success in environmental problem solving. Ultimately, it is human health that must be maintained and protected and this necessitates coordination between many disciplines, e.g. biology, ecology, geology, chemistry, economics, medicine, etc. Integration of earth systems science and an understanding of the constraints imposed by physical, chemical and biological processes is essential to constructing a fundamental framework for solving environmental problems (Thompson & Turk 1993). There is a new and growing public interest in issues related to environmental geology, e. g. volcanoes, earthquakes, climate, including waste disposal and pollution. It is the growing concern over the future environment, which is in turn encouraging earth scientists to see their contribution to society other than in the traditional involvement in resource industries. As environmental consciousness grows in the new millennium, the

physical, chemical and biological constraints exerted by geological processes will ensure that geology plays an increasingly important role in monitoring, understanding and solving environmental problems. Geologists have the dual role of harvesting the earth's rich resources, as well as protecting the environment. We must therefore ensure that we encourage the economy to grow while minimising environmental degradation. It is ironic therefore that geology has not been more quickly and effectively integrated into environmental science. Until recently the thrust of the environmental movement came from the biological sciences, but a new acceptance that environmental problems cannot be solved without a multidisciplinary approach has brought a universal recognition that geology provides an essential framework for environmental understanding. In this work, we emphasise the important role geology, and especially sedimentology and geochemistry, make to the understanding of processes that profoundly influence the nature and extent of contamination in modern aquatic systems.

SEDIMENTOLOGICAL CONTROLS ON THE NATURE AND DISTRIBUTION OF CONTAMINANTS IN THE AQUATIC ENVIRONMENT Use of sediments in aquatic environmental science Sediments are being used increasingly to assess environmental impact and to detect contaminant sources instead of utilising water, as has been done previously (Bubb &

244

G. F. Birch et al.

Lester 1994). Sediments have been found to more faithfully record and time-integrate the environmental status of an aquatic system, in contrast to water data which are highly variable in the short term, difficult to interpret and costly to acquire. Sediments also have contaminant concentrations, which are generally orders of magnitude higher than that of surface water, and even more so for interstitial water and therefore their contaminant concentrations are cheaper and more accurately determined. However, the most important aspect regarding sediments is that they are the major carrier phase for pollutants and provide the largest storage for both metallic and organic contaminants, as well as nutrients. Because sediments play a major role in the transport and storage of contaminants, they are important in determining temporal and spatial variance of pollutants in the surface and interstitial water phases. Sediments are economically attractive as an initial survey tool and are being increasingly employed in the early phases of environmental assessment in aquatic systems. As pollutants are mainly transported in association with sedimentary particles, the manner in which geological factors and processes influence the nature of contaminants in the aquatic environment is extensive. In physical terms, geology influences the nature of contaminants from the catchment to the final repository, while also constraining the physical, chemical and biological processes that control the ultimate fate of pollutants. Regional geology determines the topography of the catchment and to a large extent the rate of denudation and sediment supply. It also influences the rate at which sediments are moved through the catchment and the proportion of material stored there. Catchment geology also determines the texture and mineralogy of the particles being transported through the system and thereby exerts a great influence on the nature of the contaminants being transported and deposited. On reaching the sea, the fate of sediment-bound contaminants is determined by the nature of the coastline. Drowned, indented coastlines would tend to trap contaminants in extensive deep estuaries, whereas emergent coastlines would ensure that pollutants are not trapped in intermediate storage areas, but are transported directly onto the continental shelf. Here pollutants could accumulate, or be dispersed, depending on the depth of the continental shelf and the ambient energy impinging on the sea floor.

Physical processes The manner in which sediments influence the nature and concentration of contaminants through physical processes is mainly related to particle size and the behaviour of particles in water. Sediment size is the main determining factor controlling the capacity of sediments to bind pollutants. Size fractionation of contaminants is due to the affinity of contaminants to adsorb to the fine fraction of aquatic sediments. The reason for this relationship is well-understood and extensively documented in the literature (Forstner & Wittmann 1979; Sakai et al 1986; Horowitz & Elrick 1988; Brook & Moore 1988; Bubb et al 1990; Barbanti & Bothner 1993). This association is related to the exponential increase in surface area with decreasing particle size and increased surface charge. Also important are the common association of

organic matter with the fine fraction and the ability of organic matter to adsorb contaminants. Moreover, mineralogy exerts control on heavy-metal concentrations by its influence on particle size (Dossis & Warren 1980) and because the composition and crystal structure of minerals provide the exchange sites for adsorbing/desorbing ions. The ion-exchange capacity of smectite is ten times that of kaolinite, with other major clay mineral groups having intermediate values (Lewis & McConchie 1994). This size relationship dictates that the majority of contaminants be transported in association with the suspended population, and not as bedload. Because of this natural size fractionation, the majority of contaminants will accumulate in lowenergy, fine-grained depositional environments, rather than in high-energy, coarse-grained (erosive) environments, irrespective of proximity to source. It is difficult, if not impossible, to make comparisons of sedimentary trace-metal concentrations without compensating for grainsize, especially for multiple sedimentological environments with variable ambient energies such as in fluvial, estuarine and marine environments. Thus, some form of normalisation to minimise the confounding effects of variable grainsize is required in order to interpret contaminant data. The spatial distribution of contaminants cannot be used for identification of source, determination of dispersion pathways, or locating areas of contaminant deposition and accumulation without some form of adjustment to account for variable textural characteristics. Over limited areas, where grainsize is uniform, total sediment contaminant data may be comparable, but for the vast majority of cases the effect of grainsize on contaminant concentrations is acute and overwhelming. A procedure for minimising this texturally induced confounding effect should therefore be an integral part of the protocol for aquatic environmental assessment. Currently, the most common techniques used to reduce the effects of textural variance are size normalisation and elemental normalisation. Many 'conservative' elements (Fe, Al, Li, Sc, Cs, Rb, Ti) have been used as a proxy for the fine-sediment fraction in elemental normalisation, but none is used universally due to local and regional variance in mineralogy and chemistry. The advantages of the size-normalisation method are that it is simple, relatively quick, produces universally comparable data and provides valuable information on dispersion. Data produced in this manner are also important for fingerprinting and source identification, and can be readily related to pre-anthropogenic background contaminant levels. As contaminants are mainly bound to particles, they are controlled by the same physical factors that control sediments in aquatic environments. Contaminants can therefore be remobilised by reworking in the more energetic parts of the fluvial and marine environment, or be resuspended in the quieter, mainly depositional areas. Reworking of temporary contaminant storage, e.g. in flood plains or bank deposits, is common and plays an important part in remediation strategies as eradicating primary sources may not reduce contaminant supply in the short or medium term. Likewise, the amount of particulates that can be resuspended into the overlying water column by turbulent diffusive events by biological, or physical processes, is important from two points of view. Particle-sorbed contaminant forms can be ingested by pelagic species, but more

Role of geology in environmental science

Figure 2 1 . 1 Location of Sydney's four estuaries and their catchments.

important is the partitioning of both organic and metallic contaminants between the particulate and dissolved forms. It has been shown that when contaminated sediments are suspended in to the water column, dissolved contaminant concentrations increase concurrently (O'Connor & Connolly 1980; Di Toro et al 1986.). The reason for this occurrence is not well-understood and may be associated with increases in colloids, macromolecules and dissolved ligands, or to kinetic interactions (Burgess & Scott 1992). This increase in particulate and dissolved or colloidal forms greatly increases the bioavailability of contaminants in this quiescent aquatic environment and, as we will see, it has important associated chemical repercussions as well.

Chemical processes First-order chemical reactions that commonly control contaminant transfer in bed and suspended sediments are acidb a s e d a n d redox reactions, as well as precipitation processes (Arakel & McConchie 1995). Under high-energy conditions (fluvial or marine) the redox potential at or near the sediment-water boundary can be strongly influenced by dynamic physical processes such as tidal or wave action in estuaries, or stream erosion in the fluvial system. Disturbance at the sediment-water boundary can remobilise anoxic sediment into an oxic environment thereby

245

markedly changing the chemical characteristics of the sediment. Such major changes in redox conditions result in repartitioning of contaminants between the particulate and dissolved chemical phases, which substantially changes the availability of contaminants to biota. Complex adsorption and desorption reactions are linked to redox changes at, or near, the sediment-water interface. Freshwater fluvial material undergoes substantial physical and chemical modification on entering the estuarine environment (Pierce & Nichols 1986). Flocculation is the aggregation of small particles and is initiated at very low (<l%o) salinities close to the point of discharge. The presence of organic matter and biota can further facilitate the formation of aggregates (Gibbs et al 1989). These processes not only affect the transport and fate of contaminants, but also influence the bioavailability of toxicants (Hart & Birkett 1986). By affecting adsorption and sedimentation, organic matter plays an important role in the transport of contaminants in estuaries. Rapid flocculation and sedimentation increases the ability of an estuary to act as a sink for fine sediments and more importantly for contaminants as well. Decomposition of organic carbon in bottom sediment facilitates reduction reactions and largely controls the partitioning behaviour of ionic organic compounds in sediments. In addition, because the toxicity of contaminants is considered associated with the free ion concentration, organic carbon content of sediment substantially affects the bioavailability of contaminants in the estuarine environment. Depending on grainsize and the porosity of the bed sediment, more oxygenated water may be forced into the subsurface by changes in capillary pressure during tidal exchanges. 'Tidal pumping' has been demonstrated to markedly change the redox gradient in mangrove sediments on a diurnal time-scale (D. McConchie pers. comm. 1998). Bioavailability and therefore the toxicity of estuarine sediment may be highly dynamic and the monitoring and assessment of sediment quality may be affected by temporal changes on a time-scale of hours. Early authigenic reactions, e.g. replacement, precipitation and recrystallisation, can result in the selective uptake of contaminants by carbonates, sulfides, oxides and organic complexes in surficial sediment. Such processes can markedly affect the solid and dissolved phase chemistry and thereby influence the free ion concentration of interstitial solutions and therefore the availability and toxicity of the system.

Biological processes Biological processes can accelerate abiotic reactions at the micro- and macroscale (Burgess & Scott 1992). Microbenthic processes include the formation of low molecular weight organic compounds, or colloids, to which contaminants partition, and the metabolism of complexes, including contaminants. Benthic bacteria not only metabolise organic carbon complexes, but also induce changes to redox and pH conditions. Bacteria metabolise metal sulfides, releasing metal ions and producing excess hydrogen ions. This reduction in pH causes metal hydrous oxides, carbonates and sulfides as well as Fe and Mn coprecipitates to dissolve, releasing metals to the interstitial environment. Possibly, the most important macrobenthic biological process is bioturbation, or bioirrigation. Bioturbation

246

G. F. Birch et al.

Cu (ug/g) >100

•

SO-too

<50

nomud

TNT Figure 2 1 . 2 Copper concentrations in fine (<62.5 pm) surficial sediments in Port Hacking.

effects turbidity, sediment oxygen content, pH and the depth of the reduction horizon, which influences the form of many of the contaminants in the interstitial water and hence their bioavailability (Burgess & Scott 1992). Bioturbation causes aeration, sediment mixing, fluid transport and resuspension of sediment into the overlying water column. The other important macrobenthic process is migration, which involves the physical transfer of contaminants that have bioaccumulated in infauna and epifauna.

HEAVY-METAL CONTAMINATION IN CENTRAL NEW SOUTH WALES ESTUARIES AND ADJACENT CONTINENTAL MARGIN Much of the character and beauty of Sydney can be attributed to its estuaries. Four deeply incised flooded valleys dissect the raised coastal margin producing one of the longest coastlines of any capital city in the world (Figure 21.1). These beautiful waterways have injected a maritime character deep into the heartland of suburban Sydney and have provided an extensive shoreline along which many beautiful historic homesteads and parklands have been built. These physical attributes have been vital in the development of Sydney into one of the most beautiful cities in the world. These waterways have also provided excellent harbours for trade and the city is at the intersection of many national and international transport routes. Because of its locality, Sydney has grown rapidly and its population now stands at four million—almost one-quarter of the total Australian population. The waterways have provided easy access for raw materials and export of produce, supplied coolant for machinery, and constitute a convenient receptacle for waste disposal. All sewage and industrial waste was dis-

charged directly into the estuaries until 1898 when three major coastal outfalls were constructed to dispose of effluent into the surf zone off Sydney. Direct discharge into the estuaries was controlled by the Clean Waterways Act only in 1972 and coastal sewage disposal was upgraded to a deep-water disposal system only in 1990. Over the last ten years the Environmental Geology Group in the Department of Geology and Geophysics at the University of Sydney has been conducting geochemical investigations of bottom sediment in the four major Sydney estuaries (Parramatta River/Port Jackson, Port Hacking, Georges River/Botany Bay, and Hawkesbury River) as well as in the Hunter River (Birch 1995; Birch & Taylor 1995; Birch et al 1996b, 1997, 1998, 1999). Consistent sample collection, storage, pretreatment and analysis throughout the programme have insured compatibility of data. This extensive, GIS-based heavy-metals inventory has now been compiled into a regional contaminant framework for central New South Wales (Birch & Hennecke 1996; Birch & Taylor 1996). The database can be interrogated to provide information regarding potential sources and to 'fingerprint' different geochemical environments. It is also being used to better understand the nature of urban and industrial impact on the estuarine and marine environments of Sydney.

Methods Approximately 3700 estuarine samples have been collected using a polycarbonate corer, operated either by a diver geologist or from a boat, or with a stainless-steel box corer. Approximately 1000 samples have been obtained from creeks and rivers draining surrounding catchments and over 300 samples have been recovered from the adjacent continental margin using a stainless steel Smith-Mclntyre

Role of geology in environmental science

247

Figure 2 1 . 3 Copper concentrations in fine (<62.5 pm) surficial sediments in the Georges River/Botany Bay estuary. grab. This set of samples has evolved into the largest regionally compatible database for any aquatic environment in the country. A suite of nine elements (Cd, Co, Cr, Cu, Fe, Mn, Ni, Pb and Zn) are routinely analysed using a 2:1 H C 1 0 4 : H N 0 3 digestion and a Perkin Elmer (model 3 0 0 0 ) Flame Atomic Absorption Spectrometer. All samples are size-normalised ( < 6 2 . 5 pm) and both fine ( < 6 2 . 5 pm) and coarse ( 6 2 . 5 - 2 0 0 0 pm) fractions are analysed for heavy metals. Normalised results minimise the confounding effects of variable grainsize, whereas total sediment data (fine plus coarse fractions) enable the assessment of sediment quality. Precision, determined by repeated analyses of several estuarine standard reference materials, is better than 10% relative standard deviation for most elements. Internal standards, international reference materials and blanks were run on each sample batch. Complete consistency in field, laboratory and analytical methodology has been maintained throughout the program.

Estuarine heavy-metal contaminants Although nine elements are routinely analysed, only finefraction Cu will be used in this work to describe the distri-

bution of heavy-metal contaminants in both the estuarine and marine environments of central New South Wales. Sediment Cu concentrations in Port Hacking are low ( < 4 0 jig gr 1 ) the seaward end of the main channel, but increase in the upper reaches of the channel (75 pg g~l) and in the northern ends of offchannel embayments (140 pg gr 1 C u ) (Figure 21.2). The main channel of the lower Georges River/Botany Bay estuary also contains sediments with minor Cu (generally less than 5 0 pg g - 1 Cu), whereas concentrations increase steadily upstream (Figure 21.3). Three distinct point sources of heavy metals are evident in the Georges River/Botany Bay estuary. Cooks River and Salt Pan Creek have Cu concentrations of consistently over 2 0 0 pg gr 1 , whereas Prospect Creek generally exhibits values between 70 and 120 pg g" 1 (McCleod & Birch 1995; Birch & Taylor 1995). The highest Cu concentrations ( > 3 0 0 pg g - 1 ) in the Parramatta River/Port Jackson estuary are located in embayments in the south Central Harbour region (Figure 21.4 on Plate 10). Elevated values ( > 2 0 0 pg g - 1 ) also occur in parts of the Parramatta and Duck Rivers, the main channel of the Central Harbour, and in restricted parts of the Lower Harbour, e.g. Rose Bay (Irvine & Birch 1998; Birch & Taylor 1999). All heavy metals display strong declining con-

248

G. F. Birch etal.

Figure 2 1 . 5 Copper concentrations in fine (< 62.5 pm) surficial sediments in the Hawkesbury River (see Figure 21.1 for location).

centrations down the estuary. Middle Harbour and Lane Cove River have variable, but moderate to low Cu concentrations (50-150 jig g"1). The Cu content of sediments mantling the bed of the main Hawkesbury River channel is regular, generally varying only between 20 and 30 pg g - 1 (Figure 21.5) (Shotter et al 1995; Birch et al 1999). The upper reaches of many tributaries exhibit elevated Cu concentrations. Pittwater, which opens into Broken Bay at the mouth of the Hawkesbury River, has Cu values which increase markedly towards to the southeast (>100 pg g"1). Cu concentrations also increase towards the headwaters of Cowan Creek near Bobbin Head where recreational and boating facilities are well-established. Maximum Cu concentrations (>100 pg g~l) in Berowra Creek occur adjacent to Berowra Waters, a small township and ferry crossing. Environmental status of Sydney's estuaries There are various methods to compare the environmental status of Sydney's four estuaries in terms of heavy-metal impact, i.e. maximum and average concentrations, as well as baseline and enrichment over background. Maximum Cu concentrations in Port Hacking, Georges River/Botany Bay, Port Jackson/Parramatta River and Hawkesbury River are 140, 457, 1178 and 203 pg g - 1 , respectively, whereas average concentrations are 62, 70, 124, and 47 pg g - 1 , respectively (Table 21.1, Figure 21.6). Baseline is the most common concentration range, or a range around the modal concentration for a given area. Often baseline is related to the level of diffuse, or non-point source for a region. Baseline for Port Hacking, Georges River/Botany Bay, Parramatta River/Port Jackson and Hawkesbury River are 35-50, 35-50, 100-150 and 30-40 pg g-1 Cu, respectively. Background Cu concentrations, as determined from core samples taken from below the anthropogenic zone, are consistent within the four estuaries (e.g. 9, 10, 10 and 14 Pg g~l Cu > respectively). This provides a valid tool for comparing maximum enrichment over background for the four Sydney estuaries of about 15, 45, 110, and 17 pg g"1 for Cu, respectively (Birch et al 1999) (Table 21.2).

Figure 2 1 . 6 Mean Cu concentrations in surficial sediments in Sydney's four estuaries and adjacent continental margin.

Land use, source and estuarine impact Eighty-six percent of the Parramatta River/Port Jackson catchment is urbanised and/or industrialised, whereas similar figures for the Port Hacking, Georges River/Botany Bay, and Hawkesbury River are 10%, 42% and <5%, respectively (Figures 21.7, 21.8) (Birch et al 1998). Generally, there is a close relationship between catchment land use and estuarine impact—the greater the proportion of industrialisation/urbanisation, the higher the estuarine contaminant levels (Figure 21.8). The highest heavy-metal concentrations in central New South Wales estuaries occur at the headwaters of shallow, offchannel embayments and at the ends of tributaries. Sizenormalised heavy-metal concentrations exhibit strong declining gradients away from stormwater outlets at the ends of some of these embayments and tributaries. Sediment and water in canals and creeks draining heavily urbanised/industrialised catchments have highly elevated metal contents (Birch et al 1996a). Stormwater from these catchments is thought to make an important contribution to the general contamination of the Sydney waterways. Extensive areas of intertidal flats, especially in Port Jackson (>7.5 km2), have been reclaimed using waste (Liu 1988). The juxtaposition of the most impacted parts of Port Jackson with the location of reclamation may indicate that leachates from these areas are affecting the upper reaches of these waterways. Leachates from a waste dump at the headwaters of Salt Pan Creek have been shown to be at least partly responsible for the high heavy-metal concentrations in this tributary (Birch et al 1996b). High heavy-metal concentrations in surficial sediments in embayments of the central Port Jackson and Georges River/Botany Bay estuaries is testament to the intense industrialisation of these catchments (Taylor & Birch 1995a). Cooks River and Prospect Creek drain extensive industrial areas. Substantial heavy-metal discharge from industry into Alexandra Canal, a tributary of Cooks River,

249

Role of geology in environmental science Table 21.1 Heavy-metal concentrations in sediments of central New South Wales estuaries. Cu

Port Hacking3

33-140 62

72-291 124

111-367 183

bd-1 <1

27-72 50

bd-8 5

59-76 44

1.1-4.3 29

61-133 98

Georges River/Botany Bay3

17-457 70

29-924 155

76-2641 393

bd-12

17-185

<0.1-6.4

7-28 16

13-122

Port Jacksonb

4-1078 124

92-1319 268

-18-2246 548

<1-10 3

17-1472 118

25-60 19

12-86 38

1.0-12.2 3.8

38-919 144

Hawkesbury Riverc

17-203 47

19-174 55

68-272 135

Hunter Rivei^1

35-193

48-777 172

31-1638

bd-8

14-22 17

4.0-10.0

37-78 56

58-94

289-851 445

Hawkesbury River3

17-122 42

24-174 82-297 47 140 Birch et al 2000; b Irvine & Birch 1998; c Birch et al 1998; d Birch et al 1997 (Port Hunter only); All values in jig g"1 except Fe in %. Single values are mean concentrations, bd, below detection limit. a

Table 21.2 Heavy-metal enrichment in sediments of central New South Wales estuaries.

Port Hacking3 Georges River3 Port Jacksonb Hawkesbury Riverc Hunter Rivei^

mean max. mean max. mean max. mean max. mean max.

7 15 6 45 12 110 2 17 nd 9

4 9 5 12 8 40 3 8 nd 20

5 9 9 29 12 48 2 4 nd 32

Birch et al 1996d; Irvine & Birch 1998; Birch et al 1998; All values are times background, nd, no data. a

1 9 1.5 10

1 1 1 2 2 29

3 4 2 4 1.1 4

2 3 2 5 1.5 3

nd 8

nd 20

nd 3

Birch et al 1997.

has had a marked influence on the adjacent sediments of Botany Bay. Here a contaminant 'footprint' is clearly visible on size-normalised data extending to the southeast of the bay (Birch et al 1996b) (Figure 21.3). Sediments in the north arm of Sugarloaf Bay in Middle Harbour have high heavy-metal concentrations due to the occurrence of past poor industrial practice in the adjacent catchment. Other possible sources include, shipping activities, sewage overflows, and historical and modern illegal dumping and atmospheric deposition.

Biosignificance of estuarine contaminants Although total recoverable heavy-metal concentrations are useful in assessing the environmental status of aquatic environments, the critical issue is whether toxicants are available to biota and if they are entering the food chain. No sediment-quality criteria are available for Australia, but toxicity can be estimated using the biological effectsbased criteria of Long and Morgan (1990) and Long et al (1995). Although these empirically derived relationships between contaminant levels and biological effect are based on North American environments, they probably give an indication of possible effects on biota in New South Wales

estuaries. Sediment-quality guidelines are based on total sediment contaminant concentrations, hence the need for fine as well as coarse fraction heavy-metal analyses. Areas (in km 2 ) where (total) sediment exceeds the Effects Range-Medium (ER-M) concentration value, i.e. where 'biological effects are frequently, or always observed, or predicted among most species' (Long & Morgan 1991) are very minor in all estuaries except Port Jackson for Pb, and Zn (Figure 21.9). The majority of the Upper and Middle Harbour exceed the ER-M values for Pb and Zn and approximately 38%, 36% and 2% of the harbour area exceeds the ER-M concentrations for Pb, Zn and Cu, respectively. The areas of Sydney Harbour where some biological effects can be expected for certain sensitive species [Effects Range-Low (ER-L) values] cover the majority of the entire estuary. Another estimate of the biologically available fraction is by selective (Weimin et al 1992) and sequential extractions (Tessier et al 1979). Weimin et al (1992) have shown that dilute HC1 and ethylenediaminetetraacetic acid (EDTA) closely predict the uptake of Zn, Pb and Cd by deposit feeders (Batley 1987). Selective extractions have been carried out in two studies on sediments from Sydney's estuaries. The proportion of metal (Cu, Mn, Ni, Pb and Zn) released

250

G. F. Birch et al.

Land Use

m* e

— 33'S

CATCHMENT BOUNDARIES 1 Northern lagoons 2 Port Hacking 3 Port Jackson Parramatta River 4 Georges River Botany Bay 5 Hawkesbury ~ Nepean River

Figure 21.7 Land use in the catchments of Sydney's four estuaries.

by EDTA and HC1 was variable and low (10-20%) compared to total extractions in a study of freshwater sediments in the Parramatta River basin, a subcatchment of Port Jackson (Birch et al 1998). In Middle Harbour, another subcatchment of Port Jackson, 57%, 33% and 8% were extracted by EDTA and 11%, 38% and 1% were extracted by HC1 for Pb, Zn and Cu, respectively. The tendency for elements to accumulate in biota depends partly on the ability for the element to be removed from the solid phase into solution. Thus information on the chemical form in which the trace element is present in the sediment is more informative than total element chemistry (Tessier et al 1979). Only limited information is available on chemical speciation for these estuarine sediments. Some Zn and a small proportion of the Pb are related to the exchangeable/adsorbed phase in the Parramatta River catchment (Birch et al 2000). Heavy metals in the estuarine sediment of the Hawkesbury River are associated more with the organic/sulfidic phase than with the easily exchangeable phase (Birch et al 2000). Arakel (1995), however, found a large proportion (>50%) of the total metal load to be chemically reactive in the same river. Hanna (1992) determined that the majority of the metals in sediments of the Cooks River, which drains into the Botany Bay/Georges River estuary, are associated with the more stable chemical phases. The exchangeable/adsorbed phase, being the most reactive, is generally regarded as being potentially the most mobile and bioavailable fraction (Kersten & Forstner 1995). The above sequential extraction data suggest that some Zn and, to a smaller extent, Cd and Pb may be available to aquatic fauna and flora in these estuaries. A direct assessment of bioavailability can be made from heavy-metal analyses of tissue from biota in these estuaries. Oysters from a contaminated region of the Hawkesbury River have substantially higher Ag, Cu, Cd, Co, Hg, Ni and Zn concentrations than elsewhere on the river (Hardiman & Pearson 1996). In addition, Cu concentrations exceeded the National Food Authority maximum permitted concentration of 70 pg g b y 20%. A close relationship between heavy metal concentrations in oyster tissue and in ambient surficial sediments has also been found in Pittwater, an embayment of the Hawkesbury River (Birch 1996a). These investigations indicate that heavy-metal contaminants may be entering the food chain in some localities.

Readily resuspendable material in the estuaries

Figure 21.8 A close relationship exists between the intensity of urbanisation/industrialisation in the catchment and the concentration of heavy metals in estuarine sediments (pg g'1).

Extensive parts of Sydney's estuaries are mantled in a thin (1-30 mm) deposit of fine (mean size approximately 15 pm), oxidised, hydrous sediment of low density that attains a maximum thickness in shallow offchannel embayments (Figure 21.10). This material is often highly enriched in metallic contaminants and in most areas concentrations decline away from source. Detailed speciation work carried out on cores suggests that metals may be scavenged from the water column by adsorption onto fine surface-active particles and by coprecipitation with oxides and oxyhydroxides (Taylor & Birch 1995b, 1996a; Birch et al 1996c; Taylor et al 1996). Deeper (>15 cm) in the sediment the metals undergo early diagenetic speciation changes so that below the redox transition zone most metals are more per-

Role of geology in environmental science

isnce

MIDDLE

>£R-M

Sugarloerf Bay

<ER-M >ER-L <ER-L CENTRAL

251

HARBOUR

HARBOUR LOWER HARBOUR

UPPER HARBOUR

33'50'S .

Parramatta

Homebush Bay ISfOCE

Hen & < Chicken Bay ;

lron Cove

Blackwattle Bay

{

MIDDLE

HARBOUR

>ER-M <ER-M >ER-L <ER-L CENTRAL

LO WER HARBOUR

UPPER HARBOUR

Homebush Bay

Hen & J Chicken A Bay

Blackwattle Bay i

Zn Sugarloqf* Bay

>ER-M <ER-M >ER-L <ER-L CENTRAL

Homebush Bay

Hen A Chicken Bay i

HARBOUR

LOWER HARBOUR

Blackwattle Bay

Figure 21.9 It is important to establish the extent to which sedimentary heavy metals are affecting biota in these estuarine environments. The empirical effects-based technique of Long and Morgan (1990) is used to give an indication of likely adverse effects on estuarine biota. ER-M, Effects Range-Medium; ER-L, Effects Range-Low.

252

G. F. Birch et ol. water column has previously unrecognised environmental consequences for the pelagic flora and fauna. This is particularly important as contaminants were previously believed to be trapped in benign sulfidic phases. Furthermore, because this hydrous material is readily resuspended and transported, it may play an important role in the processes of estuarine contaminant transfer.

Heavy-metal contaminants on the adjacent continental margin

Figure 2 1 . 1 0 Distribution of readily resuspendable material (dark shading) in Sydney's four estuaries.

manently bound as metal sulfides where they react with sulfide produced by sulfate-reducing bacteria in anoxic sediment (Taylor & Birch 1997, 1998). Data derived from sediment traps, turbidity and direct measurement of suspended solids, indicate that the hydrous material is being continuously recycled into the water column. Winter sedimentation rates are substantially lower than in summer and a strong cyclical pattern observed in summer is subdued during the winter months. Many potential physical (tidal, wind, rain, etc.), biogeochemical (bioturbation, precipitation, etc.) and anthropogenic (trawling, ferry transport, etc.) processes may be involved in such behaviour. Large differences in suspended particulate loads in summer and winter may indicate an important biological influence. Biopumping may be responsible for initially suspending the material into the water column, after which tidal currents are probably sufficiently strong to keep the material in suspension. Resuspension by wind stress in shallow areas during strong northeasterly gales and by disturbance of the estuary floor during prawn trawling have also been identified. The continual recycling of very large volumes of highly enriched, oxic contaminants in a bioavailable form in the

The central New South Wales continental margin is narrow and deep with an average width of 25 km and approximately 80% of the shelf area deeper than 50 m (Yassini & Jones 1995). The shelf can be divided into an inner shelf of quartzose sands (<60-80 m water depth), a middle shelf of quartzose carbonate-rich muds (10-60% mud) to muddy sands (60-80 to 120 m water depth), and an outer shelf of calcareous, iron-stained gravelly sands (>120 m water depth). The middle shelf mud belt is continuous between Jervis Bay and Port Stephens, except adjacent to Sydney where surficial sediments are slightly sandier. The central New South Wales continental shelf is a high-energy storm-dominated shelf, with local strong windforcing affecting longshore currents to a depth of 100 m (Cresswell 1974). Sediment resuspension, reworking and winnowing beyond 80 m water depth during severe storm events results in highly mobile shelf sediment and contaminant transport (Birch & Davey 1995). Anthropogenic sources affect the distributions of all analysed heavy metals in the fine fraction of surficial sediments, with the exception of Ni. Metal concentrations are elevated on the inner and middle shelf adjacent to the urban centres of Sydney, Wollongong/Port Kembla and Newcastle (Matthai & Birch 1995, 1997a, c, 2000). In contrast, total sediment trace-metal distributions are similar to the distribution of fine sediment with maximum concentrations occurring on the middle shelf (Figure 21.11 on Plate 11). These results emphasise the importance of reducing the confounding effects of sediment size when attempting to determine anthropogenic contributions to the aquatic environment (Matthai & Birch 1996). Had only total sediment been analysed, the main source of heavy metals may have been interpreted as being the muds located on the middle shelf, which in fact are acting as a sink for these contaminants. The concentration of heavy metals in the fine fraction of surficial sediment adjacent to Newcastle is marginally elevated due to long-term dumping of harbour dredge spoil. The advantages of size-normalisation are again well-illustrated in this region as no enrichment of heavy metals can be detected in the distribution of total sediment. Mud and associated contaminants are prevented from accumulating to levels that would be harmful to benthic biota due to efficient dispersal by high-energy shelf processes (Long et al 1995). A number of anthropogenic sources contribute trace metals to inner and mid shelf sediments adjacent to Sydney. An enrichment of total organic carbon and tracemetals in the fine fraction of these shelfal sediments suggests that an important source could be that of the coastal and deep-water sewage outfalls. Elevated trace-metal con-

Role of geology in environmental science

centrations in sediment fine fractions in the vicinity of the Malabar deep-water ocean outfall (Matthai et al 1996a; Matthai & Birch 1997b) tend to support this possibility. However, the total mass of heavy metals above preanthropogenic background is only between 2 and 3% of sediment-bound Cu, Pb and Zn (Matthai & Birch 1997b). The total mass of anthropogenic Cu, Pb and Zn on the central New South Wales continental margin calculated from vertical and spatial information is approximately 259 t, 437 t and 1089 t, respectively, but the proportion of this mass which is related to sewage is yet to be determined (Matthai et al 1996b, c). Generally, the anthropogenic trace-metal content of surficial sediments on the central New South Wales continental margin is low due to a highly dynamic oceanographic climate. Moreover, the concentrations are low compared to adjacent estuarine sediments where fine sediments and contaminants have been trapped in quiescent sedimentary environments.

Remediation of the central New South Wales estuaries Extensive areas of Sydney's four estuaries are mantled by sediment containing very high concentrations of toxic heavy metals. Remediation options for these enormous volumes of polluted material are to: (i) leave it undisturbed; (ii) cover it with a 'blanket' of clean sand; (iii) remove it to the deep sea; (iv) remove the heavy metals and return the material to the estuaries; or (v) dispose of the polluted sediments on land. Until recently, the consensus was to leave the contaminated sediments undisturbed on the estuary floor, the rationale being that the toxicants were locked up in anoxic sediments in relatively insoluble and stable organic and sulfidic chemical phases. However, because it has now been shown that these surficial sediments are readily resuspended into the overlying water column, and because the toxicants are in a bioavailable form, they may be readily consumed not only by benthic fauna but also by pelagic species (Taylor & Birch 1995a, b, 1996; Birch et al 1996c). These polluted sediments can now no longer be regarded as toxicologically benign. Moreover, recent work indicates that the contaminants are entering the food chain and therefore some form of remediation is probably required (Taylor & Birch 1997, 1998; Birch 1996b). Biological (bioturbation) and physical (resuspension) turnover of surficial sediment would make the long-term success of a clean blanket approach questionable and the isolation of enormous quantities of contaminated estuarine sediment onland is problematical. The chemical removal of such large quantities of heavy metals from these estuarine sediments would probably be financially prohibitive. Very efficient dispersal by shelf currents and substantial dilution off the central New South Wales coast requires that deep sea removal, at least, be investigated as cost-effective remedial option. The long-term success of remediation is doubtful if the contaminant source is not controlled. Recent work on contemporary sediments being supplied to the receiving waters indicates that contaminants continue to enter these estuaries (Birch et al 1996a). Although absolute fluxes are not yet known, concentrations of metallic contaminants are very

253

high, especially from some central-city catchments. The sources of these toxicants will have to be established and strategies will have to be put in place to control them before any long-term solutions to the degradation of these New South Wales estuaries can be made.

CONCLUSIONS This study of the nature and distribution of heavy-metal contaminants in surficial sediments mantling the estuaries and adjacent continental margin of central New South Wales has convincingly demonstrated the importance of geological processes in controlling toxicants in these aquatic environments. The coastal morphology of central New South Wales of drowned, incised river valleys has produced a dendritic coastline comprising numerous small embayments with low energy conditions. This morphology has promoted the accumulation of extensive areas of fine-grained tidal sediments which provide ideal adsorption sites for hydrophobic contaminants. Although the Hawkesbury Sandstone of the hinterland comprise only minor fine-grained sediment, this coastal morphology has produced an ideal receptacle for the efficient entrapment of fine material and sedimentbound contaminants. The morphology and oceanographic conditions of the adjacent continental margin, by contrast, have resulted in a high-energy environment, which is not conducive to the accumulation of copious fine material and contaminants. Sediment size is the most important parameter determining the nature and concentration of contaminants in the aquatic environment, irrespective of the proximity to source. This work has demonstrated the critical requirement to reduce the confounding effects of variable grainsize to produce compatible data and to identify source and to establish dispersion pathways. The common association of organic matter and fine-grained sediment often results in anoxic bottom sediment which subsequently influences chemical speciation and bioavailability of contaminants. Resuspension plays a dominant role both in the estuarine and the marine environment. The extent and importance of resuspension in the estuaries of New South Wales has not been previously recognised. The highly contaminated surficial sediments of these estuaries were believed to be trapped in stable sulfidic and organic chemical phases and to be largely benign. Geologists with their understanding of particle behaviour in aquatic media have now shown that the uppermost part of these surficial sediments is regularly resuspended into the overlying water column. Resuspension changes the chemical nature of the contaminants to a more bioavailable form, which is then accessible to both infauna and pelagic species. Resuspension on the continental margin, by contrast, plays a powerful role in dispersing and removing contaminants from this environment. Although a considerable amount of contaminated material is disposed of offshore, only minor accumulations of toxicants occur on this continental shelf. This efficient dispersion and removal of contaminants from the New South Wales continental shelf provides a possible opportunity for cost-effective remediation of the highly polluted adjacent estuaries.

254

G. F. Birch et at.

REFERENCES Towards developing sediment quality assessment guidelines for aquatic systems: an Australian perspective. Australian Journal of Earth Sciences 4 2 , 3 3 5 - 3 6 9 . ARAKEL A. V . & M C C O N C H I E D . M . 1 9 9 5 . Geochemistry of aquatic sediments in Australia: a status report. Australian Marine Data Collection and Management Guidelines Workshop. CSIRO Marine Laboratories, Hobart, Tasmania. BARBANTI A. & B O T H N E R M. H . 1993. A procedure for partitioning bulk sediments into distinct grain-size fractions for geochemical analysis. Environmental Geology 2 1 , 3 - 1 3 . BATLEY G . E . 1 9 8 7 . Heavy metal speciation in waters, sediments and biota from Lake Macquarie, New South Wales. Australian Journal of Marine and Freshwater Research 38, 5 9 1 - 6 0 6 . BIRCH G . F. 1 9 9 5 . Sediment-bound metallic contaminants in Sydney's estuaries and adjacent offshore, Australia. Estuarine, Coastal and Shelf Science 4 2 , 3 1 - 4 4 . B I R C H G. F. 1996a. Anthropogenic impact on sediments and biota in Pittwater, NSW. 30th Newcastle Symposium on Advances in the Study of the Sydney Basin, p. 26. Department of Geology, University of Newcastle, Newcastle. BIRCH G. F. 1996b. The environmental status of Sydney's estuaries. In: Fraser R. A. ed. 9th Edgeworth-David Day Symposium, University of Sydney, pp. 4 1 - 4 9 . BIRCH G . F. 2 0 0 0 . Marine pollution in Australia, with special emphasis on central New South Wales estuaries and adjacent continental margin. International Journal of Environment and Pollution 13

ARAKEL A . V . 1 9 9 5 .

573-600. B I R C H G . F. & DAVEY S. 1 9 9 5 .

Accumulation of metallic contaminants in surficial sediments on a high energy continental shelf, Sydney, Australia. The Science of the Total Environment 1 7 0 , 8 1 - 9 3 .

BIRCH G .

F., DAVIS-MCCONCHIE

F., M C C O N C H I E

TAYLOR S .

1996c The significance of hydrous material in estuarine contaminant transfer. International Symposium on Environmental Chemistry and Ecotoxicology, Sydney, 14-18 July, 1996 (unpaginated). B I R C H G. F., E V E R D E N D. & T E U T S C H M. 1996b. Dominance of point source in heavy metal distributions in sediments of a major Sydney estuary (Australia). Environmental Geology 2 8 , 1-6. B I R C H G. F. & H E N N E C K E W. 1 9 9 6 . An interactive GIS-controlled database for contaminant information on sediments in New South Wales estuaries. Geological Society of Australia Abstracts 4 1 , 36. B I R C H G . F . , E Y R E B . D . & TAYLOR S . E . 1 9 9 9 . The use of sediments to assess environmental impact on a large coastal catchment—The Hawkesbury River system. AGSO Journal of Australian Geology & Geophysics 1 7 , 1 7 5 - 1 9 1 . BIRCH G. F., INGLETON T. & TAYLOR S. E. 1997. Environmental status of the world's largest coal exporting harbour, Port Hunter, Newcastle, Australia. Journal of Marine Environmental Engineering 4, 133-145. B I R C H G . F . , R O B E R T S O N E . , TAYLOR S . E . & M C C O N C H I E D . 2 0 0 0 .

Use

of sediments to determine environmental impact on the fluvial system. Environmental Geology in press. status of Hawkesbury River sediments. Australian Geographical Studies

B I R C H G. F., S H O T T E R N. & S T E E T S E L P. 1998. The environmental 36,

37-57.

Towards a regional contaminant framework for central NSW estuaries. Australian Marine Sciences Association Conference Proceedings, University of Technology, Sydney (unpaginated). B I R C H G. F. & TAYLOR S. E. 1996. A contaminant framework for sediment in Sydney's four main estuaries. International Symposium on Environmental Chemistry and Ecotoxicology, Sydney, 14-18 July, 1996 (unpaginated). B I R C H G. F. & TAYLOR S. E. 1999. Source of heavy metals in sediments of Port Jackson estuary, Australia. The Science of the Total Environment 2 2 7 , 123-138. B I R C H G . E , TAYLOR S . E . , MATTHAI C . & F O N G M . 1996a. Preliminary results of contaminant dynamic studies in Port Jackson estuary. Geological Society of Australia Abstracts 41, 38. B R O O K E . J . & M O O R E J . N. 1 9 8 8 . Particle-size and chemical control of As, Cd, Cu, Fe, Mn, Ni, Pb, and Zn in bed sediment from the Clark Fork river, Montana (U.S.A.). The Science of the Total Environment B I R C H G . F. & TAYLOR S . 1 9 9 5 .

76,

247-266.

B U B B J . M . & LESTER J . N . 1 9 9 4 .

Anthropogenic heavy metal inputs to

lowland river systems, a case study. The River Stour U.K. Water, Air and Soil Pollution 78, 2 7 9 - 2 9 6 1990. Distribution of heavy metals in the River Yare and its associated broads. III. Lead and Zinc. The Science of the Total Environment 1 0 2 , 189-208. B U R G E S S R. M. & S C O T T K. J . 1992. The significance of in-place contaminated marine sediments on the water column: processes and effects. In: Burton G. A. ed. Sediment Toxicity Assessment, pp. 129-165. Lewis Publishers, London. C R E S S W E L L G. R. 1974. Ocean currents measured concurrently on and off the Sydney area continental shelf. Australian Journal of Marine and Freshwater Research 25, 4 2 7 - 4 3 8 . B U B B J . M, R U D D T . & L E S T E R J . N.

TORO D .

M.,

MAHONY J.

D.,

KIRCHGRABER P . R . ,

O'BRYNE L.

R.,

PASQUALE L . R . & PICCIRILLI D . C. 1986. Effects of nonreversibility, particle concentration and ionic strength on heavy metal sorption. Environmental Science and Technology 20, 55-61. Dossis P. & W A R R E N L J. 1980. Distribution of heavy metals between the minerals and organic debris in a contaminated marine sediment. Contaminants and Sediments 1, 120-139. F O R S T N E R U. & W I T T M A N N G. T. W . 1979. Metal Pollution in the Aquatic Environment.Springer-Verlag, New York. GIBBS

J.,

SHUDY

M.,

KONWAR

MARTIN

1989.

Coagulation and transport of sediments in the Gironde estuary. Sedimentology 36, 9 8 7 - 9 9 9 . HANNA R. G. 1992. Sequential extraction of metals in Cooks River sediments. In: Miskiewicz A. G. ed. Proceedings of a bioaccumulation Workshop, Sydney, pp. 177-186. Sydney Water Board and Australian Marine Sciences Association, Sydney. HARDIMAN S. & P E A R S O N B. 1996. Heavy metals, TBT, DDT in the Sydney rock oyster (Saccostrea commercialism sampled from the Hawkesbury River estuary, NSW, Australia. Marine Pollution Bulletin 30, 5 6 3 - 5 6 5 . H A R T B . T. & B I R K E T T R. 1986. The composition of suspended particulate matter from the Magela Creek system, Northern Australia. Environmental Technology Letters 7, 613-624. H O R O W I T Z A. J. & ELRICK K. A. 1986. An evaluation of air elutriation for sediment particle size separation and subsequent chemical analysis. Environment Technology Letters 7, 17-26. IRVINE I . & B I R C H G. F. 1998. Distribution of heavy metals in surficial sediments of Port Jackson, Sydney, Australia. Australian Journal of Earth Sciences 4 5 , 169-174. K E R S T E N M . & F O R S T N E R U. 1 9 9 5 Speciation of trace elements in sediments. In: Batley G. ed. Trace Elements Speciation: Analytical Methods and Problems, p. 157. CRC Press, Boca Raton. L E W I S D . W . & M C C O N C H I E D . M . 1 9 9 4 . Analytical Sedimentology, Chapman Hall, New York. Liu K. 1988 Evolution of the Sediments in Sydney Harbour. MSc thesis, University of Sydney, Sydney (unpubl.). LONG

R.,

MACDONALD

D.,

SMITH

CALDER

1995

Incidence of adverse biological effects within ranges of chemical concentrations in marine and estuarine sediments. Environmental Management 19, 8 1 - 9 7 . L O N G E. R . & M O R G A N L . G. 1990 The potential for biological effects of sediment-sorbed contaminants tested in the National Status and Trends Program. NOAA Technical Memorandum NOS OMA 52. US National Oceanic and Atmospheric Administration. Seattle, Washington. MATTHAI C. & B I R C H G. F . 1995. Anthropogenic and natural accumulations of heavy metals on the N S W continental shelf. 29th Newcastle Symposium on Advances in the Study of the Sydney Basin, pp. 7 4 - 8 1 . Department of Geology, University of Newcastle. MATTHAI C., B I C K F O R D G. P. & B I R C H G. F. 1996a. Spatial and temporal trends of heavy metals in settling organic particulate matter near a deep sea ocean outfall off Sydney, N S W Geological Society of Australia Abstracts 41,271. MATTHAI C. & B I R C H G. F. 1996. Trace metals in surficial marine sediments near a large ocean outfall off Malabar, Sydney, Australia. In: Davies P. J. ed. IGCP 396 Conference, University of Sydney, p. 70. MATTHAI C. & B I R C H G. F. 1997a. Trace metal tracers in continental shelf sediments—a regional approach from the central NSW continental shelf. 3rd Australian Marine Geoscience Conference, Adelaide, October, 1997, p. 42. MATTHAI C. & B I R C H G . F. 1997b. The application of the equilibrium partitioning method in surficial sediments near a deep water out-

Role of geology in environmental science fall off Malabar, Sydney, Australia. 4 th International Symposium on Environmental Geochemistry, Vail, October, 1997, p. 61. The effects of Newcastle Harbour dredge spoil dumping on surficial marine sediments. 21st Newcastle Symposium on Advances in the Study of the Sydney Basin, pp. 8 3 - 8 9 . Department of Geology, University of Newcastle, Newcastle. MATTHAI C. & BIRCH G. F. 2000. Effects of coastal cities on surficial sediments mantling an adjacent high-energy continental margin— central New South Wales, Australia. Marine and Freshwater Research 51, 565-576. MATTHAI C., BIRCH G. F. & BICKFORD G. 1996b. Heavy metal contamination in central NSW inner continental shelf sediments adjacent Sydney, Newcastle and Wollongong. International Symposium on Environmental Chemistry and Ecotoxicology, Sydney, 14-18 July, 1996 (unpaginated). MAITHAI C., BIRCH G. F. & BICKFORD G. 1996C. Silver as sewage effluent tracer in settling particulate matter from a deep ocean outfall off Malabar, Sydney. Australian Marine Science Association Conference, Hobart, Tasmania, July, 1996, p. 121. MCCLEOD K . & BIRCH G. F. 1 9 9 5 . Understanding urban and industrial impact by comparing two major NSW estuaries. 29th Newcastle Symposium on Advances in the Study of the Sydney Basin, pp. 66-73. Department of Geology, University of Newcastle, Newcastle. O'CONNOR D. J. & CONNOLLY J. P. 1980. The effect of concentration of adsorbing solids on the partition coefficient. Water Research 14, 1517-1523. PIERCE J. W . & NICHOLS M . M . 1 9 8 6 . change in particle composition from fluvial into an estuarine environment: Rappahannock River: Virginia. Journal of Coastal Research 2 , 4 1 9 - 4 2 5 . SAKAI H., KOJIMA Y . & SAITO K . 1 9 8 6 . Distribution of heavy metals in water and sieved sediment in the Toyohira river. Water Research

MAITHAI C . & BIRCH G . F. 1 9 9 7 C .

20,

559-567.

SHOITER N . , O'DONNELL M., STEETSEL P. & BIRCH G. F. 1995. The envi-

ronmental status of a large NSW estuary under threat—the

255

Hawkesbury River system. 29th Newcastle Symposium on Advances in the Study of the Sydney Basin, pp. 59-66. Department of Geology, University of Newcastle, Newcastle. TAYLOR S. & BIRCH G. F. 1995a. Contaminant dynamics in offchannel embayments of Port Jackson, NSW. Australian Marine Sciences Association Conference, University of Technology, Sydney (unpaginated). TAYLOR S. E. & BIRCH G. F. 1995b. Environmental implications of readily resuspended contaminated estuarine sediments. 5th Annual New South Wales Coastal Conference, Murwillumbah (unpaginated). TAYLOR S . E . & BIRCH G . F. 1 9 9 6 . The environmental implications of readily suspended contaminant estuarine sediments. International Geological Congress, Beijing, p. 381. TAYLOR S. E. & BIRCH G. F. 1997. Contaminant dynamics in Port Jackson estuary, Sydney, Australia. 4th International Symposium on Environmental Geochemistry, Vail, October, 1997, p. 90. TAYLOR S. E. & BIRCH G. F. 1998. The environmental implications of readily resuspendable contaminated estuarine sediments. AGSO Journal of Australian Geology & Geophysics 1 7 , 2 3 3 - 2 3 7 . TAYLOR S. E., BIRCH G . F., COX R. & GROSKOPS M. 1996. Resuspension of contaminated sediments from Port Jackson, Australia. International Symposium on Environmental Chemistry and Ecotoxicology, Sydney, July 14-18, 1996 (unpaginated). TESSIER A., CAMPBELL P. G. C . & BISSON M. 1979. Sequential extraction procedure for the speciation of particulate trace metals. Analytical Chemistry 51, 844-850 THOMPSON G . R . & TURK J . 1 9 9 3 . Earth Science and the Environment. Saunders College Publishing. Fort Worth. WEIMIN Y . , BATLEY G . E . & AHSANULLAH M . 1 9 9 2 The ability of sediment extractants to measure the bioavailability of metals to three marine invertebrates. The Science of the Total Environment 125, 7-84 YASSINI I. & JONES B . G . 1 9 9 5 . Foraminifera and Ostracoda from estuarine and shelf environments on the southeastern coast of Australia. University of Wollongong, Wollongong. Received 12 February 1999; accepted 1 December 2000

Geological Society of Australia Special Publication 21, 257-265

CHAPTER 22—Contribution of geological process studies to environmental management: case studies from northeast New South Wales and southeast Queensland W. | BOYD, R. t BUSH, M. W. CLARK, J . V . SMITH AND L A, SULLIVAN School of Resource Science & Management Southern Cross University, Lismore, NSW 2480, Australia. Four case studies are presented in this chapter which, together with the case studies in Chapter 23, provide a selection of examples dealing with issues typical of environmental management issues encountered in the east Australia coastal zone. All contain a geological component within the broader environmental focus, and it is the geological components that are described here. Central to each case study is the role that examination of physical processes plays in understanding key elements of the environment, and thus guiding appropriate management activities. The first two case studies concern aspects of water management, particularly in relation to sediment-water interactions as they bear upon water quality issues and pollution. While these introduce concepts of conflicting land use, the following two case studies look explicitly at issues of conflicting land use, quality control and remediation of mine and quarry sites. KEY WORDS: geological processes, pollution, quarries, remediation, sulfides, water management, water quality.

INTRODUCTION Any environmental science textbook dealing with natural systems clearly indicates the complex nature of the processes that allow natural environments to operate (White et al 1984; Briggs & Smithson 1985). The natural environment represents an intimate system of linkages between biological, hydrological, sedimentary and soil processes, within which human activity occurs. Environmental management is to a large extent concerned with mediating between the processes of human activity and those of the natural environment which, both in the past and at present, have tended to be put out of balance by the human activity. This imbalance is generally indicated by an acceleration of already existing processes or a change in dominant processes. Consequently environmental management requires an understanding of natural processes, an understanding that allows evaluation of the human impact, the rates and directions in which both natural and induced changes are occurring, and the development of credible and effective strategies of remediation or preventative management. The current debate, for example, regarding global climate change and predicted sea-level rise is confounded by difficulties in separating human-induced global warming effects from those of the natural longer term climatic changes that form an essential part of the world's natural system. This need for examination of an historic perspective is examined in Chapter 23. Regardless of the historical context of global warming and sea-level rise, the predominant global scientific response has largely focused on the modelling of climatic and oceanographic processes (CSIRO 1992) and their impacts upon, in particular, coastal environments (Parslow & Jernakoff 1992; Zann 1996). This scientific response is enshrined in major international

collaborative programmes, extensively funded and resourced by international agencies and national governments (Zann 1996). This global response emphasises the need, if such programmes are to work effectively (Pearman 1988), for a firm understanding of the underlying processes in any environmental system. There is a current trend of formal inquiry and review amongst environmental managers, which has resulted in a plethora of overview documents (over 60 such reports, for example, for the coastal zone between 1960 and 1993: Resource Assessment Commission 1993). While these frequently focus on the administrative and bureaucratic elements of environmental management, there is a developing interest in environmental processes, despite the apparent gap that exists between science and management (Zann 1996). This gap represents a belief that scientific research is too 'pure' to adequately address 'real-world' management issues. However, there are many examples emerging in which scientific research has clear and readily applied implications for environmental management. A few will suffice here. Flood's (1988) review of Holocene sea-level history and consequential coastal sedimentation processes, for example, provides important input to the development and management of residential canal estates in southeast Queensland. This type of research and the geological knowledge which emerges from it is readily extended to other management situations in the region (Holmes 1995). In the area of nutrient balance and apparent imbalance—an important interface between geological processes of nutrient supply and cycling and the use of natural resources in food production—McConchie's Shark Bay (Western Australia) research (McConchie & Lawrence 1991; McConchie et al 1988) on the processes of geological concentration and release of cadmium and subsequent uptake in molluscs, is

258

W. E. Boyd et al.

CASE STUDY 1, WATER QUALITY AND SOIL CHANGES: GEOLOGICAL METHODS FOR IDENTIFYING FACTORS AFFECTING THE WATER QUALITY HAZARD POSED BY SULFIDIC SEDIMENTS

a particularly useful example of a highly applicable scientific process study. By way of a final example, Blong and McKee's (1995) evaluation of the impact of the Rabaul volcanic eruption in east New Britain (Papua New Guinea) illustrates the role that even minimal process studies can play in developing appropriate environmental management strategies (see also Lentfer & Boyd, in press). In this latter case, no systematic study was made of the effects of ash fall over the town of Rabaul; this town was to a significant extent demolished. However, careful examination of the stages of this geological event allow a reasonable reconstruction of the eruption, ash fall and, most importantly, the subsequent reworking and replacement of ash during later wet weather. The tangible consequences of identifying and elucidating these geological processes lie in, for example, the development of appropriate building codes, insurance strategies and post-disaster remediation plans. This chapter pursues this theme, although taking a local view of the northeast of New South Wales and the southeast of Queensland, by describing studies concerning representative local land-use issues in the management of hydrological, geological and soils resources. The four case studies presented here (Figure 22.1) provide a selection of studies from the wide range of day-to-day management concerns of public and private natural resource managers working in the region: soil/sediment - water relationships, especially in relation to water quality issues and pollution (Case study 1, water quality and soil changes; and Case study 2, polluted water, sediments and vegetation) and coastal sediment extraction and the remediation of mining sites (Case study 3, geological resource extraction; and Case study 4, mining and soils).

The eastern Australian coastline contains extensive coastal floodplains containing Holocene sulfidic sediments that normally lie below the water-table. Disturbance of these sulfidic sediments by excavation or drainage can lead to exposure of these sediments to the air resulting in a reaction between oxygen gas and the iron sulfides that leads to the production of essentially sulfuric acid. When this acid leaches into waterways the pH (i.e. acidity) of the water can fall below 2. Such oxidation can also lead to the degradation of water quality in streams by increasing the concentrations of dissolved aluminium and iron, both of which can damage riverine and estuarine environments (Sammut et al 1996). Sulfidic sediments affected by oxidation are called 'acid sulfate soils' and have caused stream acidification leading to mass kills of fish and other marine organisms as well as damage to engineering structures and crops (Bloomfield & Coulter 1973). Sulfidic sediments are widely distributed around Australia: it has been estimated that potential sulfidic sediments are adjacent to more than 50% of the coastline from tropical Western Australia to southern New South Wales (Crossland 1996). However, some areas of disturbed sulfidic sediments pose a more severe hazard to water quality than others and the need to develop predictive capacities for the sulfide oxidation rate and the rate at which acidified water is entering the waterways has been recognised (Bush & Sullivan 1996; White et al 1996). Factors that can affect sulfide oxidation rates are numerous and include: sulfide mineralogy; the presence of coatings on sulfide crystals; the size of the sulfide crystals; sediment buffering capacity; sediment permeability to both air and water; and depth of burial of the sulfidic sediments. This case study will describe some geological methods that have been used to investigate sulfide mineralogy in these sediments and of the effect of this factor on oxidation rates. The site chosen for this study was in the Tuckean Swamp, an embayment infilled by estuarine sediments during the Holocene (i.e. some 6000 to 3000 years before present). The swamp is part of the Richmond River floodplain on the north coast of New South Wales. Undisturbed samples of unoxidised sediment were collected from a depth of 0.9-1.0 m from an area of the swamp that is known to be producing acidification at rates which severely damage water quality in the adjacent waterway. The general properties for this soil layer were described in Bush and Sullivan (1996) and are: fine sandy clay loam texture, dark bluish-grey colour, massive structure with frequent channels lined with root remnants, pH 6.2, electrical conductivity 1.4 dSirr 1 , organic carbon 1.01%, total sulfur 2.41% and carbonates 0.06%. These samples were snap frozen using liquid nitrogen and freeze-dried. The mineralogy of these samples was examined by Xray diffraction, light microscopy and scanning electron microscopy with energy dispersive X-ray analysis SEMEDAX using quantitative elemental analysis (Sullivan & Bush 1997; Bush & Sullivan 1997). This examination

Geological process studies

259

CASE STUDY 2, POLLUTED WATER, SEDIMENTS AND VEGETATION: MANAGEMENT IMPLICATIONS OF METAL-TRANSFER PATHWAYS FROM A REFUSE TIP TO MANGROVE SEDIMENTS

Figure 22.2 Platy crystals of greigite mixed with several pyrite octahedra. The scale bar represents 1 pm.

demonstrated that the magnetic fraction of the sulfides contained appreciable quantities of the monosulfide mineral greigite (Fe 3 S 4 ) as well as pyrite (FeS 2 ) (Figure 22.2). This mineral had not previously been identified in these sediments in Australia. An investigation of the behaviour of moistened samples of the main sulfide minerals extracted from these sediments (Bush & Sullivan 1997) demonstrated that greigite is highly unstable and will readily oxidise at room temperature whenever there is sufficient moisture to enable chemical reactions. On the other hand pyrite was stable in this investigation, and other investigations at Southern Cross University indicate that pure pyrite buried in acidic, well-drained soil conditions in humid coastal regions can remain unaffected for many months (Gilmore 1996). Although pyrite is the most abundant sulfide in these soils and is the largest source of potential acidity, the slow chemical oxidation of pyrite results in the rate of pyrite oxidation dramatically increasing only after the soil becomes more acidic than pH 2.5, when the activity of necessary microbial catalysts is favoured (Singer & Stumm 1970). Given the reactivity of greigite and the alkaline nature of most unoxidised sulfidic sediments, sediments containing greigite would be expected to acidify far more rapidly and thus pose a greater environmental hazard than would sediments containing pyrite alone. Indeed, it has long been speculated that chemically reactive monosulfides such as greigite may be a necessary precursor to enable pyrite oxidation in these sediments (Dent 1986; Evangelou 1995). Even small concentrations of greigite in sulfidic sediments could considerably decrease the time (after exposure to oxidising conditions) when such sulfidic sediments become sufficiently acidic to enable rapid acid production from biologically catalysed pyrite oxidation. The geological techniques described here have identified variations in sulfide mineralogy that are likely reasons for the observed variations in the degree of hazard posed to water quality in streams by disturbance of sulfidic sediments in adjacent areas. These techniques offer the capability to readily assess these (and other) previously overlooked factors.

Mangroves have often been thought of as mosquito-infested wastelands and, because of this attitude, many mangrove forests have been used as sites for refuse tips, sewage outfalls and as illegal dumping grounds. This case study describes the investigation of heavy-metal transfer from a refuse tip to mangrove sediments. In Australia mangrove ecosystems can be found in all but the most southerly estuaries and harbours, where they provide nursery and breeding grounds for many species of marine fauna (Hutchings & Saenger 1987). The recognition of the environmental and economic benefits of mangrove protection is reflected in the legislative protection accorded to them by Australian State governments. However, despite this protection, many mangrove forests in Australia remain polluted by metallic and non-metallic anthropogenic wastes. At some sites, effluent is discharged directly into mangrove forests, at others mangrove forests act as a buffer between pollutant sources and open-marine waters and, in many places, pollutants are illegally dumped in the mangrove forest. The case for the protection of mangrove ecosystems as a buffer between sources of metallic pollutants and nearby aquatic ecosystems has been made previously (Saenger et al 1991). Although some work on the reaction of mature mangrove plants and seedlings to elevated heavy-metal loads has been published (Augusto et al 1990), there are few papers (Nye 1990) in the scientific literature on the response of mangrove ecosystems to heavy metals. Furthermore, little is known about how heavy metals behave in mangrove sediments and how they are affected by the trees and infauna. To investigate the behaviour of trace metals in mangrove ecosystems, a study was initiated at several sites, including the Wynnum (near Brisbane) site, where a domestic garbage tip is separated from Moreton Bay by a narrow ( - 2 0 0 m wide) Avicennia-dominated mangrove forest (Clark et al 1997, 1998). The mangrove forest and the tip face are separated by a slight depression, devoid of any macroflora, and characterised by highly reducing muds containing an abundance of metallic and non-metallic refuse. The methods of investigation included the geochemical analysis of sediment core and surface samples, leachate samples, and surface water and groundwater samples. Sediment samples were acid and sequentially extracted (Clark et al 1997, 1998) and analysed by anodic-stripping voltametry for Cu, Cd and Pb on a Chemtronics PDV 2000 using the method outlined in McConchie (1987). Zn, Ni, Cr and Fe were analysed by atomic adsorption and Hg was analysed on the PDV 2000 using the method outlined by Jaya et al (1985). Piezometer tubes on the site were sampled fortnightly to observe changes in groundwater chemistry and were monitored over tidal cycles for salinity. Leachate, surface water and groundwater samples from throughout the site were chemically concentrated (Clark 1992) and analysed for the above heavy metals using the outlined methods. Stratigraphic analysis provided a sediment sequence from an underlying basaltic bedrock through shelly lag deposits, grading upwards into sands, muddy sands and the root-bound muds of the mangrove-forest sediments; this

260

W. E. Boyd et al.

Figure 22.3 Schematic diagram of the chemical fluxes to, from and within the sediments at Wynnum. The depths of the upper and lower redoxclines depend on the location within the forest. On the mudflats and within the area of dead mangroves only the upper redoxcline is applicable. Sulfate is able to migrate to lower parts of the sediment by infiltrating burrows. In the upper reduced zone, other metals (M) may substitute for Fe in the sulfide lattice, or form their own discrete sulfides.

sequence is consistent with a prograding, tidally dominated estuary (Clark 1996). Analysis of sediment metal content revealed a high metal concentration at the top of the basalt, which is thought to reflect migration of tip leachates across the top of the impermeable basement. The variations in metal concentrations across the weathered basalt basement are thought to reflect mineralogical changes associated with microenvironments related to surface relief variations present during weathering (Clark 1992). Analysis of surface water and groundwater shows that metals are strongly partitioned between the sediment and the groundwater phases. While the surface waters and groundwaters show similar metal concentration trends to the sediments, the concentration of metals in these waters was three orders of magnitude lower than in the sediment. This strong partitioning of metals to the sediment indicates that the transfer processes from groundwater to the sediment are very efficient. Lateral distributions of metals in the sediment indicate that, despite low metal concentrations in the groundwater, the groundwater plays an important role in the transport of metals through the area. Comparison of data taken at the site prior to 1991 (Saenger et al 1991) and data from the present study indicates that the seasonal changes in groundwater chemistry and stratigraphic position remobilises metals from the sediment back to the groundwaters for redistribution down the hydrological slope (Clark et al 1997). Partitioning of the metals to the sediment may be from direct adsorption onto fine-grained sediment, complexing with organic matter, and/or the formation of insoluble sulfides by reaction with bacterially generated sulfide. However, the vertical distribution of metals in the sediment is not uniform and there are geochemically distinct hori-

zons that favour metal trapping (neutral to high pH, low Eh horizons) or metal mobilisation (low pH, high Eh horizons). The sequential data indicates that heavy-metal speciation (e.g. exchangeable, carbonate bound, oxide bound and sulfide bound) and hence mobility is strongly influenced by the presence of the mangrove roots and the position of the water-table (Clark et al 1998) (Figure 22.3). These geochemical results in conjunction with the site investigation indicate four pathways for metal migration from tip cell to mangrove forest: (i) direct seepage across the tip cell floor to deep groundwater; (ii) tidal overtopping of the bund wall and capillary suction of leachates to shallow groundwaters; (iii) direct seepage through the cell wall of leachates to surficial sediments; and (iv) surface runoff during rainfall events (Figure 22.4). The effects of these processes are complicated by the effects of interacting tidal waters and the presence of the mangroves. Calculations of metal inputs via surficial runoff indicate that significant quantities of heavy metals are transferred to the mangrove forest by this mechanism, although metal transfer by surface runoff is probably small in comparison to the other three mechanisms discussed. This study has significant implications for environmental management because of the variable nature of the transfer mechanisms operating, the variability in sediment texture both vertically and laterally, the variability in geochemical conditions both vertically and laterally (Figure 22.4), and the number of transfer pathways (Figure 22.4) makes potential management of the site difficult. Any attempt at management of the tip site (i.e. reduction of metal leaching rates) must consider both the feedback mechanisms that operate (e.g. decreasing runoff from the

Geological process studies

261

Surficial runoff Top soil cover Direct seepage through the tip cell wall Refuse Material

Combined surficial runoff and direct seepage to surficial sediments

Capillary suction by tidal waters to the shallow groundwater Direct seepage across the bund floor to all levels of the groundwater

Figure 22.4 Metal-transfer pathways into, through and out of the mangrove sediments at Wynnum. tip cell increases infiltration and groundwater leaching) and that the whole of the sediment column is involved in metal transfer. The nature of the transfer mechanisms and the hydrological and geochemical regimes operating at the Wynnum site makes remediation difficult and costly. The Brisbane City Council, under pressure from environmental groups and the State government, has installed a leachate drain along the eastern face of the northern tip cell (Figure 22.4). However, there are serious doubts about the drain's effectiveness because the drain does not intersect many of the leachate seeps or runoff that flow directly to the surficial sediments, nor does the drain intersect the deep groundwater flow across the basalt. Furthermore, the drain is permeable to both saline tidal waters coming up the hydrological slope and groundwater flow moving down the hydrological slope, hence metals entering the drain from the tip are capable of being removed to the groundwater and down the hydrological slope, through diffusion and/or tidal pumping (Figure 22.4). Finally, this investigation indicates that metal concentration data in waters, are by themselves, insufficient for sound environmental management, and that both water volume and metal concentration data are required to make valued judgments on contamination risks. CASE STUDY 3, GEOLOGICAL RESOURCE EXTRACTION AND CONFLICTING LAND USE AND EFFICIENCY: QUARRYING OF SAND AND GRAVEL DEPOSITS AND BASALT BODIES

In a region of growing population such as the north coast of New South Wales there is a continuing need for materials

for the development of infrastructure (Brownlow 1994). The need for extraction of these construction materials can be in direct opposition to other needs of the growing community—most notably, the need for residential areas that are not affected by industrial activity and the need for protection of areas of natural environment. The notion that construction materials of adequate quality are ubiquitous (Kellet 1995) implies that planning for resource extraction can be left to sites which have little competition for land use. This is not the case. A clear understanding of the geological processes controlling the distribution and quality of potential construction material shows that optimum materials need to be identified early, allowing extractive land uses to compete for attention on the planning desk. To illustrate this important idea, the distribution and planning implications of sand and gravel deposits and hard-rock basalt deposits on the north coast of New South Wales are discussed for two examples: a sand and gravel quarry at Suffolk Park and basalt quarrying at Lismore. Batson's Quarry at Suffolk Park, 6 km southwest of Byron Bay (Figure 22.5), is a major regional sand and gravel resource. Expansion of the quarry has been stopped by surrounding residential developments. Stubbs and Smith (1997) investigated the geology of the deposit and discussed the implications for resource planning. Previous studies had suggested that the poorly consolidated sand and gravel deposit worked at Batson's Quarry was a Cenozoic sedimentary deposit overlying consolidated Mesozoic sediments in the area. Lithological analysis, including particle-size analysis and mineralogical determinations, indicated that the quarry materials were indistinguishable from samples of

262

W. E. Boyd et ol. 300-

Lamington Volcanics E-yfel rhyolite f M basalt Jg Grafton Formation Kangaroo Creek Sandstone Jw Walloon Coal Measures Rjb Bundamba Group Rc Chillingham Volcanics p z Palaeozoic metamorphics

280-

lava

260-

240220

(/) 05 _Q CO >

Jtf

0 >

O -Q Ctf

0—» ••

200

180 160 140 120

1001 i i

ash

80 60 40

lil-dlmi'oui'cl HitoS' U/ '

Ballina

•Cuadm-D

20 0

10km

29°S—

Quarries

10 11

Figure 22.5 Geological map of northeast New South Wales showing Tertiary Lamington Volcanics and outcrop and subcrop of pre-Tertiary formations. The distribution of basalt quarries (•), some of which are located on basalt outliers too small for the map scale, and the boundaries of local government areas are also shown.

Figure 2 2 . 6 Occurrence of quarries relative to elevation in the Tertiary lava pile. Large volume, high quality deposits at Teven (T), North Lismore (abandoned) (N), Blakebrook (B) and Chadburn's Quarry (C) are labelled. Deposits are concentrated by channels on the palaeolandsurface at the lava pile base and at a palaeolandsurface inferred at approximately 100 m above the lava base which coincides with the eruption of rhyolite ash and lavas.

weathered Mesozoic sediments outside the quarry. Detailed mapping in the area also demonstrated that structures such as cross-bedding and its interpreted palaeocurrent data were consistent between the quarry and known localities of Mesozoic sedimentary rocks in the area. It was therefore, determined that the quarry material was weathered sedimentary bedrock rather than more recent sedimentary deposits as had been previously suggested. The different modes of occurrence of unconsolidated sedimentary deposits have implications for the distribution and planning implications of construction materials. Relatively modern sedimentary deposits have the advantage of being identifiable partly by geomorphic analysis without dependence on extensive bedrock investigations. However, extraction of such resources inevitably impacts on active processes and environments and is often ruled out on environmental grounds. The recognition of Batson's Quarry deposit as weathered bedrock material indicates that extraction of this sand is less likely to impact on active environments, but that occurrences of similar material will be restricted to the extent of similar bedrock, that is, an area 6 km long by 2 km wide (Figure 22.5). This area has been developed to the degree that extension of the quarry or development of other quarries is virtually impossible.

tial developments in the vicinity have undergone extensive landscape modification that would have been compatible with quarry-site rehabilitation. Hard-rock deposits suitable for producing crushed stone for concrete aggregate and road making are also sometimes considered to be readily available. In reality, the quality of potentially suitable rock varies greatly and must be comprehensively tested for compliance with quality specifications. For most quarrying purposes economies of scale and environmental requirements dictate that localised large quarries are preferable to numerous smaller operations. However, in the case of low- to moderate-quality road-making materials there are benefits in having resources located close to the roads on which they will be used. Thus, studies of the distribution of hard-rock resources must consider quality, quantity and location of resources. Geological mapping and analysis of samples is required to determine the regional distribution of suitable materials. Houston and Smith (1997) have described the quality of basalt rock in the vicinity of one of the region's largest hardrock basalt deposits. The study used X-ray diffraction analysis of samples from percussion and diamond drilling to show that basalt below the floor of the existing production level is of significantly lower quality making it unsuitable for many engineering uses. The high-quality material is interpreted as channel-filling lava which formed to a sufficient thickness to produce rock with more widely spaced joints and less glass than found in other settings. A survey of the elevation of quarried basalt deposits within the lava

If the unique characteristics of the material had been recognised earlier it may have been possible to delay development in the vicinity of the quarry to allow greater extraction of the resource. This could have permitted planning for sequential land use of the area as the dominantly residen-

Geological process studies

263

Figure 22.7 (a) A soil profile in a mined area near Hawks Nest, New South Wales, showing the development of an A2 horizon in the mine spoil beneath the base of the replaced topsoil. Scale bar 10 cm. (b) Iron-impregnated coatings of kaolin and illite clay particles embedded on a quartz beach sand grain. Scale bar 10 pm. (c) Detail of (b). Scale bar 10 pm. (d) Coatings of organic matter, iron and aluminium on a mine spoil sand grain. Scale bar 10 pm. pile (Figure 22.6) shows that they are dominantly found at the base of the lava pile where erosion produced topography to promote the formation of such thick flows. Most other significant deposits occur at approximately 100 m elevation above the lava pile base, suggesting the occurrence of an eruptive hiatus allowing further incision of the landscape. This model of the distribution of high-quality basalt indicates a far more restricted resource than inspection of the available geological maps would imply. The history of exploitation of the basalt resource in the region also has geological reasons. As the map in Figure 22.5 shows, basalt exists over much of the region, but it has been dominantly exploited within the Lismore and Kyogle local government areas. Figure 22.5 also shows the outcrop and subcrop of pre-Tertiary formations based on previously published maps and geological mapping and interpretation (Smith et al 1998). The lack of exploitation of basalt in the Ballina and Byron local government areas can be explained by the availability of Palaeozoic quartzite and argillite, which are used preferentially because of their superior mineralogical stability. The resources of Palaeozoic rocks are under extreme pressure from development in those areas and it is important that the high quality of these resources should be considered in development planning in the region. These examples indicate that geological studies involving detailed and regional mapping, lithology and interpretation of geological environments are no less important for delineating potential construction materials than in any other field of mineral-resource evaluation. Comprehensive and accurate geological input is especially important given both the increasing quality-control requirements in the industry for products such as concrete and road-making aggregates, and the close proximity of resources to actual or potential conflicting land uses.

CASE STUDY 4, MINING AND SOILS: GEOLOGICAL CHARACTERISTICS AND REHABILITATION SUCCESS AFTER SAND MINING The eastern Australian coastline contains extensive coastal barrier formations and plains derived from marine sand

(Chapman et al 1982). The soil types formed on these marine sand deposits are usually podsols which are characterised by thick, dark-coloured and impermeable B horizons enriched in organic matter, aluminium and iron, underlying thick bleached white sandy A2 horizons. Many of these sand deposits are enriched in heavy minerals and are being mined. The successful long-term rehabilitation of sand-mined areas back to near their original state is likely to be dependent on the reformation of the podsol soil profile. This is because in the undisturbed pre-mining situation both the structure and floristics of the overlying vegetation in these environments is almost totally dependent on nature of underlying soil (especially degree of development and depth at which the B horizon occurs) (Walker et al 1981). Therefore, in mined areas where the aim is to revert the environment back to near its original state, the reformation of podsols is essential. However, thermoluminescence dating studies of marine sand deposits in eastern Australia suggest that podsols take many thousands of years to develop on marine sand (Tejan-Kelle et al 1990). There is evidence to suggest that reformation of podsols on sand materials resulting from mining activities may be occurring relatively quickly. For example, Paton et al (1976) examined a sand-mined area on the coastal plain of New South Wales that had been revegetated nine years previously. At this site and at others it is apparent that immediately underneath the replaced dark-grey topsoil layer, a bleached layer (A2 horizon) including bleached pipes is forming in the yellowish spoil material. Around these bleached areas thin dark-brown zones (new B horizon) are evident. This apparent podsol reformation has been taken to indicate that podsol reformation is occurring at these sites at relatively rapid rates. However, not all mined sites exhibit such podsol reformation. Thus given the need for this process to occur to facilitate successful coastal dune rehabilitation, two relevant questions are: why is the rate of podsol development on some mined sands much quicker that in natural marine sand deposits; and why do some mined podsols rehabilitate quickly whereas others do not? In this study two sand-mine rehabilitation sites were chosen for study: one near Hawks Nest on Holocene marine deposits (age around 6000 a), and the other in the

264

W. E. Boyd et al.

Bundjalung National Park south of Evans Head on Pleistocene inner barrier deposits (around 120 ka). The Hawks Nest site was close to that studied by Paton et al (1976) and shows clear apparent podsol reformation with the clear development of a bleached A2 horizon immediately underneath the topsoil layer (arrowed in Figure 22.7a). The Bundjalung site does not exhibit such podsol reformation. Although the sand in natural marine deposits (e.g. beaches) and in mine spoils is both golden and optically similar, geological analysis and analytical scanning electron microscopy of the coatings around the sand grains shows clear mineralogical differences that account for their different observed weathering behaviours. The colour of the marine sand is due to iron oxide-impregnated coatings of illite and kaolinite clay embedded in the surface of the quartz sand grains (Figure 22.7b, c) (Sullivan & Koppi 1998). However the colour of the mine-spoil sand is mainly due to coatings of organic matter, iron and aluminium (Figure 22.7d) that derive from the B horizon of the podsols prior to the mining operations. Kaolinite and illite are both minerals that show high resistance to weathering, whereas the organic matter-rich coatings can easily be leached by water percolating at first through the litter of the local native plant species, and subsequently through the sand (Paton et al 1976). Effectively, the weathering of mine-spoil sand in some areas is quicker because the weathering is occurring on pre-weathered materials. As noted above rapid podsol reformation is apparent on only some mined areas and this will restrict successful rehabilitation. A geological approach also is useful in understanding these different behaviours. The mined sands that have been observed to successfully reform after mining operations derive from podsols formed on Holocene marine deposits (e.g. age around 6000 a) whereas the mined sands that are not reforming post-mining derive from podsols formed on Pleistocene inner barrier deposits over 120 000 years old. The younger podsols formed on the outer barriers are less well-developed, are more freely draining, and have B horizons that are less-cemented than those on the inner barriers (Roy & Thorn 1981; Thompson & Bowman 1984). The chemical characteristics of these soils that explain the remobilisation of the organic coatings on the outer barrier mine spoil but not the inner barrier mine spoil are the carbon:aluminium ratio, the soil pH and the nature and amount of organic matter inputs (Skjemstad 1992). For example, experiments by Barlow (1995) have indicated that plant extracts are capable of leaching the mine spoil from a Holocene deposit but not from a Pleistocene deposit. As seen from the above discussion an understanding the geological processes involved in the reformation of podsols allows a prediction of which marine sand deposits will rehabilitate unaided as well as promising to allow the development of management practices such as amendment of organic matter contents and soil pH that will predispose mine spoil from inner barrier deposits to timely podsol reformation post mining.

CONCLUSIONS The case studies presented here describe the geological process study contributions to issues of management and

planning within a range of environmental conditions. They are representative of the situation in any part of the coastal zone, and reflect issues across the full spectrum of resource planning, exploration, management, use and remediation. The common theme here is the role that geological methods and concepts may play in contributing to successful management. In these case studies, the focus is on the dominant geological processes that either explain the consequence of a particular combination of environmental conditions, or can be harnessed to enhance the process of management decision making or site remediation.

REFERENCES Metal reservoirs in a red mangrove forest. Biotropica 22, 3 3 9 - 3 4 5 .

AUGUSTO C . , SUVA R . , LACERDA L . D . & REZENDE C . E . 1 9 9 0 .

BARLOW N.1995. An investigation of the factors influencing the refor-

mation of podzol soils after disturbance from mineral sand mining. Integrated project thesis, Southern Cross University, Lismore (unpubl.). BLONG R . & M C K E E C . 1 9 9 5 . The Rabaul Eruption 1994: Destruction of a Town. Natural Hazards Research Centre, Macquarie University, Sydney. BLOOMFIELD C . & COULTER J. K. 1973. Genesis and management of acid sulphate soils. Advances in Agronomy 25, 265-326. BRIGGS D . & SMITHSON P. 1 9 8 5 . Fundamentals of Physical Geography. Hutchinson, London. BROWNLOW J. W. 1994. Construction materials in the North Coast Region: demand assessment and proposed resource management strategy. Geological Survey of New South Wales Report GS 1994/199. BUSH R. T. & SULLIVAN L. A. 1996. What soil analyses do not tell you about acid sulphate soil behaviour. In: Smith R. J. ed. Second National Acid Sulphate Soil Conference, pp. 32-39. Coffs Harbour, September 1996. BUSH R. T. & SULLIVAN L. A. 1997. Morphology and behaviour of greigite from a Holocene sediment in eastern Australia. Australian Journal of Soil Research 35, 853-861. CHAPMAN D . M . , GEARY M . , ROY P. S. & THOM B . G . 1 9 8 2 . Coastal Evolution and Coastal Erosion in New South Wales. Coastal Council of New South Wales, Sydney. CLARK M. W. 1 9 9 2 . Physical and geochemical controls on heavy metal cycling in mangal sediments, Wynnum, Brisbane. MSc thesis, University of Canterbury, Christchurch (unpubl.). CLARK M. W. 1996. The nature of Quaternary sediments at Wynnum, Brisbane. Quaternary Australasia 14, 21-28. CLARK M . W . , MCCONCHIE D . , LEWIS D . W . & SAENGER P. 1 9 9 8 . Redox stratification and heavy metal partitioning in Avicennia dominated mangrove sediments: a geochemical model. Chemical Geology 149,

147-171.

CLARK M . W . , MCCONCHIE D . , SAENGER P. & PILLSWORTH M .

1997.

Hydrological controls on copper, cadmium, lead and zinc concentrations in an anthropogenically polluted mangrove ecosystem, Wynnum, Brisbane. Journal of Coastal Research 13, 1050-1058.

1996. A perspective on ASS as a coastal management issue. In: Smith R. J. ed. Second National Conference on Acid Sulphate Soils, pp. 15-19. Coffs Harbour, September 1996. CSIRO 1992. Climate Change Scenarios for the Australian Region. CSIRO Division of Atmospheric Research, Aspendale. D E N T D . 1986. Acid sulphate soil: a baseline for research. International Institute for Land Development and Reclamation, Wageningen, Publication 39. EVANGELOU V. P. 1995. Pyrite oxidation and its control. CRC Press, New York. FLOOD P. G. 1 9 8 8 . Subsidence of organic-rich sediments deposited in river deltas, estuaries, salt marshes and tidal flats: the need for management guidelines in canal-estate development. In: Arakel A. V. ed. Proceedings of 1988 Workshop on Coastal Zone Management, pp. 62-68. Centre for Sedimentary & Environmental Geology, Queensland University of Technology, Brisbane.

CROSSLAND C .

Geological process studies GILMORE M.

1996. An evaluation of the in situ bag technique for soil weathering studies. BSc (Hons) thesis, Southern Cross University, Lismore (unpubl.). HOLMES N. 1995. Coastal dynamics and global; change: implications for coastal managers. In: Hotta K. & Dutton I. M. eds. Coastal Management in the Asia-Pacific Region: Issues and Approaches, pp. 81-93. Japan International Marine Science & Technology Federation, Tokyo. HOUSTON E. C. & SMITH J. V. 1997. Assessment of rock quality variability due to smectitic alteration in basalt using X-ray diffraction analysis. Engineering Geology 46, 19-32. HUTCHINGS P. & SAENGER P. 1987. Ecology of Mangroves. University of Queensland Press, Brisbane. JAYA S., RAO T . P. & RAO G . P. 1 9 8 5 . Simplified anodic-stripping voltametric determination of mercury in sea water using a glassy electrode and acetate-chloride buffer. Analyst 1 1 0 , 1361-1364. KELLET J. E. 1995. The elements of a sustainable aggregates policy. Journal of Environmental Planning and Management 38, 569-579. LENTFER C. & BOYD W . E . The effects of volcanic eruptions on people and the environment: The 1994 Rabaul eruptions, a case study. Southern Cross Press, Lismore (in press). MCCONCHIE D. M . 1987. Anodic stripping voltametry in environmental analysis. In: Environmental and Exploration Geochemistry, pp. 4-6. Queensland Institute of Technology, Brisbane. MCCONCHIE D . M . & LAWRENCE L . M . 1 9 9 1 . The origin of high cadmium loads in some bivalve molluscs from Shark Bay, Western Australia: a new mechanism for cadmium uptake by filter feeding organisms. Archives of Environmental Contamination & Toxicology 21,

303-310.

MCCONCHIE D . M . , M A N N A . W . , LINTERN M . J . , LONGMAN D . , TALBOT V . , GABELISH A. J . & GABELISH M . J . 1 9 8 8 .

Heavy metals in marine biota, sediments and waters from the Shark Bay area, Western Australia. Journal of Coastal Research 4, 3 7 - 5 8 . NYE L. B. 1990. Trace metal accumulation under differing sediment conditions in the mangrove, Rhizophora mangal L., in Key Largo, Florida. MS thesis, University of Miami, Miami (unpubl). PARSLOW J . & JERNAKOFF P. (Editors) 1992. Managing Australia's Fisheries under the Threat of Climate Change Impacts. Australian Government Publishing Service, Canberra. PATON T . R , MITCHELL P. B . , ADAMSON D . , BUCHANAN R . A . , F o x M . D . &

BOWMAN G. 1976. Speed of podsolisation. Nature 2 6 0 , 601-602. G. I . (Editor) 1988. Planning for Climate Change. CSIRO, Australia. RESOURCE ASSESSMENT COMMISSION 1 9 9 3 . Coastal Zone Inquiry: Final Report. Australian Government Publishing Service, Canberra. ROY P. S. & THOM B. G. 1981. Late Quaternary marine deposition in New South Wales and southern Queensland—an evolutionary model. Journal of the Geological Society of Australia 28, 471-489. SAENGER P., M C C O N C H I E D . M. & CLARK M. W . 1 9 9 1 . Mangrove forests PEARMAN

265

as a buffer between anthropogenically polluted areas and the sea. In: Arakel A. V. ed. 1990 Workshop on Coastal Zone Management, Vol. 1, pp. 280-300. Geo-Processors, Yeppoon. SAMMUT J . , W H I T E I . & MELVILLE M . D . 1 9 9 6 . Acidification of an estuarine tributary in eastern Australia due to drainage of acid sulphate soils. Marine and Freshwater Research 4 7 , 6 6 9 - 6 8 4 . SINGER P. C. & STUMM W. 1970. Acid mine drainage: the rate limiting step. Science 167, 1211-1215. SKEMSTAD J. O. 1992. Genesis of podzols in coastal dunes in southern Queensland. III. The role of aluminium-organic complexes in profile development. Australian Journal of Soil Research 30, 645-665. SMITH J . V . , MIYAKE Y . & HOUSTON E. C . 1 9 9 8 . Mesozoic age for volcanic rocks at Evans Head, northeast New South Wales. Australian Journal of Earth Sciences 45, 9 5 5 - 9 6 1 . STUBBS B. J. & SMITH J. V. 1997. Weathered bedrock as a source of sand and gravel aggregate on north-eastern New South Wales, Australia. Environmental Geology 32, 64-70. SULLIVAN L. A. & B U S H R . T. 1997. Quantitative elemental microanalysis of rough-surfaced soil specimens in the scanning electron microscope using a peak-to-background method. Soil Science 162, 749-757. SULLIVAN L . A. & KOPPI T . 1 9 9 8 . Iron staining of quartz beach sand in southeastern Australia. Journal of Coastal Research 14, 9 9 2 - 9 9 9 . TEJAN-KELLA M . S . , CHITTLEBOROUGH D . , FITZPATRICK R . W . , THOMPSON

C. H . , PRESCOTT J . R . & HUTTON J . T. 1 9 9 0 . Thermoluminescence dating of coastal sand dunes at Cooloola and North Stradbroke Island, Australia. Australian Journal of Soil Research 2 8 , 4 6 5 - 4 8 1 . THOMPSON C. H. & BOWMAN G. 1984. Subaerial denudation and weathering of vegetated coastal dunes in eastern Australia. In: Thom B. G. ed. Coastal Geomorphology in Australia, pp. 263-290. Academic Press, Sydney. WALKER J . , THOMPSON C. H., FERGUS I . F. & TUNSTAL B. R . 1 9 8 1 . Plant succession and soil development in coastal dunes and subtropical eastern Australia. In: West D. C., Shugert H. H. & Botkin D. B. eds. Forest Succession: Concepts and Application, pp. 1 0 7 - 1 3 1 . Springer-Verlag, New York. WHITE

D.,

MOTTERSHEAD

HARRISON

1984.

W H I T E I . , W I L S O N B . P., MELVILLE M . D . , SAMMUT J . & LIN C .

1996.

Environmental Systems: An introductory Text. George Allen & Unwin, London.

Hydrology and drainage of acid sulphate soils. In: Smith R. I. ed. Second National Conference on Acid Sulphate Soils, pp. 103-108. Coffs Harbour, September 1996. ZANN L . P. (Compiler) 1 9 9 6 . The State of the Marine Environment: Report for Australia: Technical Summary. Department of the Environment, Sport and Territories, Canberra. Received 30 October 1997; accepted 14 November 2000

Geological Society of Australia Special Publication 21, 267-280

CHAPTER 23—Study of historical change as a contribution to environmental management issues: case studies from northeast New South Wales and southeast Queensland W. E. BOYD, M. M. COTTER, R. McGRATH, S. PATHIRANA AND A. SPECHT School of Resource Science & Management Southern Cross University, Lismore, NSW 2480, Australia. This chapter presents five case studies which, together with those in Chapter 22, provide a cross-section of studies dealing with issues typical of the range of environmental management issues encountered in the coastal zone of eastern Australia. In all, there is to a greater or lesser extent a geological component within the broader environmental focus. It is the geological component in each case that is described here. The central issue in these case studies is that an understanding of the geologically historic perspective of the relevant environments contributes to the management of that environment; this applies to the management of both natural and human landscapes. The first two case studies deal directly with the management of plant and vegetational resources, in particular in relation to vegetation and fire management, and the management of ti-tree plantations. The next study concerns one aspect of water management, the vegetational consequences of water extraction. The final two case studies reflect the common element to all of these studies, that is that while the geological focus of each study is on the natural environmental processes, a core element of any environmental management issue is human activity. These final case studies deal with coastal sediments as cultural heritage resources, and the modelling of river mouth sediments as a contribution to planning processes. These latter case studies therefore draw this theme to a conclusion, with illustrations of the important contribution that geological research can make to an emerging area in environmental management that concerns itself with issues of cultural heritage. KEY WORDS: coastal sediments, cultural heritage, environmental management, fire management, geological history, groundwater mining, vegetation management.

INTRODUCTION The natural environment is a complex system in which biological, hydrological, sedimentary and soil processes are intimately linked. Human activities, both in the past and at present, have a tendency to create imbalance in natural systems, usually by either accelerating already existing processes or by diverting the direction of such changes. Contemporary environmental management requires an understanding of both the processes and the history of these processes together with their effects. This is in order to identify: (i) the truly detrimental effects of human interference in the natural system; (ii) the rates and directions in which both natural and induced changes are occurring; and (iii) the most credible and effective strategies to be adopted in remedial or preventative management. The current debate, for example, regarding global climate change and predicted sea-level rise is confounded by the difficulties inherent in separating the effects of human-induced global warming from these of the longer term cyclical climatic changes which form an essential part of the world's natural system. This global issue emphasises the absolute need for a firm geological understanding of the underlying process and the history of any natural system. This chapter, however, takes a more local view, describing studies from northeast New South Wales and southeast Queensland, all of which focus on the historical development of environments

and highlight the contribution of these environmental histories to contemporary environmental management. The focus in these case studies is largely on the relatively recent geological period, the Quaternary era, during which many of the contemporary landscape parameters were established, and from which much of the modern landscape derives its immediate character. By way of example of the importance of considering this geological period and the impacts of processes operating during it, one may consider the importance of Quaternary palaeoecological and stratigraphic research as used to address issues of management in the contemporary environment. Palaeoecology is the geological study of fossil plant and animal parts which has as a principle aim the elucidation of past biological environments (Roberts 1989; Williams et al 1991, 1993). Palaeoecology has an inherent interest in understanding the evolution of both individual groups of plants (families, genera and species) and vegetation types, the past distribution of specific plant species, and by using the evidence as proxy data, understanding past climate and climate change (Lamb 1995). Palaeoecological studies incorporate a wide range of techniques, including sedimentological analysis, geochemical analysis, radiometric dating, and the study of fossil plant and animal parts (Berglund 1986), and these are firmly set within the traditions of modern geological research. While palaeoenvironmental and palaeoclimatic reconstruction are the major

268

W. E. Boyd et al.

activity. Building on themes developed in Case study 3, the final case studies (Case study 4, coastal sediments as cultural heritage resources; Case study 5, environmental management and buried cultural heritage) draw this theme to a natural conclusion, with illustrations of the important contribution that geological research can make to the emerging area in environmental management which concerns itself with issues of cultural heritage management.

CASE STUDY 1, VEGETATION AND FIRE MANAGEMENT: COASTAL HEATH, BUNDJALUNG NATIONAL PARK

NP, national park.

interests of palaeoecologists, palaeoecological studies play an important role in contemporary biological management, especially in determining and evaluating the processes resulting in the present landscape, processes which in turn may be modified in the future. Flannery (1994) provides a most cogent argument that our understanding of present and future environments is dependent on a thorough understanding of past biological environments. There is an emerging literature documenting the contribution palaeoecology can make to environmental management. A few examples illustrate the potential: modelling and predicting future change from historical data (Wasson & Clark 1985); studying the effects of acid rain and history of acidification in lakes (Flower et al 1988); the management of rare and endangered tree species (Boyd et al 1988); and contributions to monitoring and managing mining tailings (Pickup et al 1987). Indeed, given the close relationship between the past and present, it is not surprising that many palaeoecological studies are conducted in national parks and thus contribute to the general understanding of the modern vegetation (Dodson et al 1994; Boyd 1990). Palaeoecological data is also increasingly being used to support submissions in the planning processes (Gell 1996). The five case studies presented in this chapter (Figure 23.1) provide examples of studies dealing largely with issues of plant and vegetational resource management (Case study 1, vegetation and fire management; Case Study 2, biological resources: ti-tree plantation management) and water and vegetational resources (Case study 3, vegetational response to freshwater extraction). A common element to these studies is that, while the geological focus on each study is on natural environment processes, the root cause lies in human

Bundjalung National Park (northern New South Wales) is one of the few remaining extensive areas of near-natural coastal land in the region, and is characterised by a wide diversity of habitats and vegetation types, and a rich fauna and flora, including populations of rare and vulnerable species. The research described here addresses several environmental issues, notably those regarding the management of natural vegetation and fire regimes within and beyond the park, and the development of long-term management responses to changes in the vegetation of the region. In particular, the new data contribute to park management practices, practices traditionally based on the presumed long establishment of present patterns of biodiversity. Should this presumption be flawed (earlier evidence from the region suggested that this may be so), management practices may need to be reassessed and thus monitoring strategies altered. The research described here is a test of that presumption. The biological nature of this area is well-recorded, and has more recently been the focus of research into fire management. From this more recent research emerged the recognition that maintaining high levels of habitat diversity was important in meeting the needs of rare and vulnerable species. It is generally assumed that such biodiversity developed over the 6000 years of the post-Holocene marine-transgression period, i.e. since sealevel stabilised at approximately the present position (Hopley 1983), allowing coastal terrestrial conditions to stabilise and thus levels of habitat diversity to remain relatively static. However, if local environmental diversity has developed as recently as geological evidence elsewhere in the region suggests (W. E. Boyd unpubl. data; Flood et al 1986; Boyd 1993), then inferences regarding the established nature, the length of maintenance, and rates of change of biodiversity patterns may need to be reviewed. Several geological methods have been applied to this issue (McGrath 1995; McGrath & Boyd 1998; Boyd et al 1997). These predominantly focus on characterisation of swamp sediments from the area, sediments which contain fossil palaeoenvironmental evidence largely as plant macro- and microfossils (leaves, seeds and pollen). Additionally, sediment types and sequences were logged, and samples of the organic sediments submitted for radiocarbon dating. The study used organic sediment sequences from several localities, four in the national park area and one beyond the park at Bungawalbin Creek. These organic sediments are important, since they both preserve fossil pollen well and are readily dated using radiocarbon dating techniques. At all of these, the lowest portion of each

Historical change studies

269

- Tr«« and S**U*

(a)

i vm* BP

i t . h

.iii

nil

Smimm

III!

40 60 Percent oi pollen count

| BackhiMcnnl

20 40 60

ss.

50 100 Summary of major taxa

E3.

200 400 xlOOOOO

1000 2000 atOOOOOG

200 400 Pol i en car*

Anafysi* fey Roger McGfMh

(b)

Years BP.

Sediment / / / / /

S X ^ v

500

•V v \ \ \ \ 4 704 ±

1000

W V " -

1500

70A ' V / /

2000 2500

ggg

3000 3500

/A/Ys

4000

4500

6 630 ± 60 •

5000 400 x1000000

5500 Black silt with some orange mottling

Black silty sand

200 400 Pollen court

Analysis by Roger McGrath

Figure 23.2 Pollen diagrams: (a) Bundjalung National Park and (b) Bungawalbin Creek, northeast New South Wales. The pollen diagrams express the quantities of pollen as percentages of total count per sample, and illustrate the changing proportions of plants in the various vegetation communities surrounding the study sites. sequence was dated to determine the age of onset of the swamps, and thus since the swamps represent important habitat components of this coastal region, the date of their establishment provides an indication of the onset of present patterns of biodiversity. Two sequences provided: (i) limited geochemical data used to characterise the sediments; (ii) fossil pollen from stratified samples used to give a proxy vegetation history; and (iii) fossil charcoal used as an indication of fire history.

The radiocarbon dating programme adopted indicated that the onset of swamp conditions in the major swamp areas of Bundjalung National Park dates back to the mid to early Holocene, except in one area where the effects of later Holocene high sea-level caused a relatively late reversion to freshwater swamp conditions. Local differences in onset date reflect topography and the effects, principally waterlogging, of enhanced spring flow during the mid-Holocene. Both palynological and geochemical analyses contributed

270

W. E. Boyd et ol.

to establishing the history of the sites (Figure 23.2). While there is some doubt about the correctness of a basal radiocarbon date of 8700 ±60 a BP obtained for one of the national park cores, an age of 6630 ±60 a BP for the Bungawalbin Creek site is credible; both records indicate a history for at least the last six and a half thousand years. Fossil pollen evidence from the sites indicates a dynamic environment with an overall tendency towards drier conditions. This evidence is supported by charcoal analyses indicating an increase in the occurrence of fire over time. Chemical analyses of the sediments indicate that the Bungawalbin site was strongly influenced by estuarine conditions, whereas the Bundjalung National Park site was only influenced indirectly by estuarine conditions. This study starts to provide palaeoenvironmental data that contributes to the understanding of the evolution of the vegetation of this coastal region. Importantly, the research offers an important opportunity to extend the Quaternary vegetation history of eastern Australia, especially given that until recently (Peters 1990) the nearest published and substantive long vegetation history is from the high altitude sites at Barrington Tops to the south (Dodson et al. 1986) and in Fraser Island to the north (Longmore 1997). Neither of these areas are representative of this subtropical region, and the need for a regionally appropriate locality has been recognised. Specifically, this study indicates that the broad patterns of habitat diversity in this national park have been established for at least 6000 years, although there have been notable dynamism in the vegetation. Details of that dynamism, in terms of climatic and environmental moisture levels, vegetational composition and the role that fire has played during the last several thousand years are now recorded. This record provides a baseline from which contemporary vegetational management may be undertaken. In particular, it provides some insight into the historical processes which have resulted in the present pattern of vegetation in the park. Although the temporal resolution of the records is still quite coarse, they do provide an insight into the rate and scale of changes in vegetation and fire regime over this long period of time. This information has not previously been available and should be invaluable to the development of vegetation and fire management plans for the region. However, most management plans are designed to last for 5 - 1 0 years. This study provides a much longer term perspective and indicates that major environmental change has operated in this region at centurial and millennial scales. The results of this study indicate that the vegetation has not remained static for any great length of time and has been in a state of constant flux. Information pertaining to such processes and changes occurring in the natural environment over long time periods is therefore important if the short-term (annual to decadal) management plans are to conserve natural conditions and processes effectively. By being able to demonstrate that change is part of the natural environmental history of the region, future changes need not necessarily be viewed, as they tend to be at present, as being unacceptable and thus to be avoided. Furthermore, attempts to preserve what exists now may prove futile in the long run. One issue, for example, is that the fire management plan of the park sets out guidelines, based on an

understanding of recent fire regimes, which recommend that most vegetation communities (excepting rainforests) should be burnt if there have been no fires for 30 years. This study, however, suggests that there have been much longer periods (possibly longer than one thousand years) in the not too distant past when the occurrence of fire has been very low at the Bundjalung site. This changes the perspective of current recommendations.

CASE STUDY 2, BIOLOGICAL RESOURCES: TI-TREE PLANTATION MANAGEMENT, BUNGAWALBIN This study was undertaken at the request of an essentialoils production company based in northern New South Wales. The company was interested in establishing the historical presence of Melaleuca alternifolia (used in ti-tree oil production) in the core region of oil production. The company was interested in the former geographical distribution and population sizes of this economically important plant, and whether severe fluctuations of population and distributions had occurred in the past, since such changes may indicate possible environmental stresses. The company considered that such information could be used by the industry in the management of present M. alternifolia stands, which occur mostly in plantations, and as an aid in assessing the need to protect and conserve the remaining natural stands particularly as gene-pool reservoirs, so as to insure protection against potential future loss of genetic diversity of the species. The research approach adopted to address this issue involved the analysis of stratified pollen sequences taken from sediments in the core ti-tree growing area (McGrath 1996). The intention was to identify the general characteristics of the vegetation over the last several thousand years, and specifically the presence and abundance of ti-tree. To this end, the scanning electron microscope (SEM) was used exclusively in an attempt to reliably identify M. alternifolia from other Myrtaceae species, an acknowledged difficult task (Erdtman 1952; Pike 1956; Mclntyre 1963; Chalson & Martin 1995; Zhou & Heusser 1996). A reference collection of extant Myrtaceae pollen was obtained from twenty species of Myrtaceae present in the region. Following standard preparation for examination under the SEM (the Southern Cross University Leica Scanning 440 SEM), various morphological characteristics (in particular, surface texture, dimensions, shape and arrangements of apertures: Chalson & Martin 1995) were recorded for groups of 50 grains per species. For the historic vegetation study a 3.5 m-deep sediment sequence was obtained for pollen analysis from an extant M. alternifolia swamp on the Bungawalbin Creek floodplain. An initial study of 9 subsamples at 40 cm intervals were analysed for pollen. Examination of the modern reference pollen indicated a large variation in many of the morphological characteristics studied (Figure 23.3). Although some differentiation between species was possible, there was considerable overlap between species making specific identification difficult. In particular, whereas the apex-to-base measurements for Melaleuca species suggests a general genus-wide conformity of size, some species such as M. quinquenervia could be dis-

Historical c h a n g e studies

271

Figure 2 3 . 3 Scanning electron microscope photomicrographs of pollen of Melaleuca alternifolia. This is a northern New South Wales species of ti-tree which forms the basis of the emerging ti-tree oil industry. This pollen type is ubiquitous throughout the Myrtaceae family, and is difficult to identify to genus and species level.

272

W. E. Boyd et al.

tinguished on the basis of size. Based on external morphology it was not possible to positively separate the pollen of M. alternifolia (the target species) from that of other species in the modern pollen samples. The level of identification possible in the modern pollen samples was reduced further in the fossil record. Damaged grains (i.e. collapsed, crushed or corroded) with similar morphological features could not positively identified beyond genus level. As a consequence M. alternifolia was placed in a Melaleuca group. The grouping of M. alternifolia with other Melaleuca species made it impossible to determine the history of its occurrence in the fossil record as was hoped. However, within the limitations of the coarse temporal sampling of the sediment sequence, it is possible to record that the site has been part of a stable freshwater swamp/floodplain for the duration of the record. If a basal AMS (accelerator mass spectrometry) radiocarbon date is correct, this could be as long as 34 260 years. Given that M. alternifolia currently occurs at the site and the Melaleuca group to which this species was assigned in the pollen analysis has been constantly represented throughout the fossil pollen record, there is reasonable evidence for the probable presence of M. alternifolia as part of the vegetation for this time.

CASE STUDY 3, VEGETATIONAL RESPONSE TO FRESHWATER EXTRACTION: EIGHTEEN MILE SWAMP, NORTH STRADBROKE ISLAND Eighteen Mile Swamp is a back barrier coastal swamp on the seaward side of North Stradbroke Island, southeast Queensland. It is some 25 km long and 1 km wide, and is protected from the Pacific Ocean by Holocene coastal dunes. North Stradbroke Island is one of several coastal sand islands protecting the seaward edge of Moreton Bay, a subtropical marine bay fringed to the east by relatively uninhabited islands and to the west by the city of Brisbane. Pressures on the environmental resources of this region are recognised, and planning zonation is under review. While this area's natural history is moderately well documented (Coleman et al 1984), a program to monitor ecological effects of freshwater extraction from North Stradbroke Island has revealed that ecological understanding here is only partial (Specht 1992). The swamp provides the only large and reliable freshwater supply for Redlands Shire, a rapidly growing satellite town of the greater Brisbane area. The project has attracted attention as a model for large-scale water extraction from the water-table of large coastal sand islands fringing the coast of this region. There is concern that such extraction may have an effect on the natural vegetation of the swamp, and consequently a monitoring program utilising systematic on-site vegetation recording and remote-sensing analysis has been running since 1988 (Dutton et al 1993). The principal focus of this monitoring program is on the dependence of vegetation on specific water conditions, in particular: (i) whether positions of plant communities can be correlated with water depth; (ii) whether vegetation activity registered in satellite imagery provides useful early warning; (iii) whether any species signal significant vegetation change; and (iv) what vegetation change is tolerable. From this program, it is clear that there is a strong rela-

tionship between vegetation and the hydrological dynamics of the swamp. So far, the ecological results indicate characteristic patterns of vegetation and hydrological dynamics on the swamp: low community species richness, sharply delineated communities correlating with water depth, and a reduction in community structure as waterlogging increases. The swamp vegetation is adapted to some variation in water level, although critical maximum or minimum water levels and flooding periodicities are unclear. It is, therefore, difficult to gauge critical stages in water extraction. Furthermore, against the background of past longterm natural variation, the long-term future effects of water extraction on the vegetation may be quite insignificant. A key indicator species of vegetation-hydrology relationships is the woody species Melaleuca quinquenervia, which becomes established during periods of lowered water level and enhanced surface drying. This species is dormant during higher water level periods. Present studies of the vegetation dynamics and ecology of the swamp indicate, however, that there is currently a major gap in our understanding of forested freshwater wetland dynamics, particularly in regard to effective regeneration and replacement of existing populations of tree species. Woody species recruit and regenerate in wetlands very rarely, and when they do it is in large even-aged cohorts, which eventually die out en masse. The recruitment patterns appear to be influenced by seedling-establishment criteria and periodicity of flooding events, although the details of environmental preconditions and possible causes are unknown. Since many wetlands, such as Eighteen Mile Swamp, are presently used in several ways, some of which may conflict with conservation requirements, it is critical to understand these preconditions and causes: effective conservation is presently impeded by a lack of knowledge. The methodological approaches adopted in this study include a mixture of ecological mapping and recording of the present and recent vegetational patterns on the swamp and a palaeogeographical study of the history of the swamp (Boyd et al 1999). These methods reflect steps identified to examine the environmental preconditions and causes: (i) the establishment of the date(s) of even-aged recruitment event(s) in the recent past; and (ii) an analysis of the environmental conditions prevailing at that or these dates. From the geological (palaeogeographical) perspective, this project addressed these issues in several ways. Estimating the establishment date of present even-aged stands is currently under way using Pb-210 dating of superficial sediments. The examination of longer term change is by examination of a swamp sediment core from Eighteen Mile Swamp to determine from the fossil pollen record and radiocarbon dating program the vegetational history over the last several thousand years. This provided an indication not only of the biological history, but also of the hydrological and sedimentological conditions. The nexus between the various interests of the palaeogeographer and the modern ecologist working on this swamp is aimed to provide an understanding of the fundamental processes of vegetation response to water-regime fluctuations. In particular, the long-term history of vegetation of the swamp complements the present-day data collection to provide the baseline for effective predictive modelling of the vegetation of the swamp and its response to continuing water extraction.

Historical change studies The palaeogeographic study of the swamp indicates that most of this large swamp probably formed only around 6 0 0 years ago following the closure of a coastal sand spit (Boyd 1993). Earlier, since at least 2 4 0 0 years ago, the area was a typical low-energy tidal bay. During these 2 4 0 0 years, changes in dominant vegetation types in the area largely reflect former water-table fluctuations. This long-term history of hydrological change and associated vegetational change provides a key in effective predictive modelling of swamp vegetation and its response to continuing water extraction. The effect of water-table changes on vegetation can be assessed against the longer term variations only provided by historical information. While the palaeogeographical record of the magnitude and periodicity of long-term water-level fluctuation is still moderately crude, order-ofmagnitude changes and their effects are certainly registered. Furthermore, the palaeogeographic model supports the archaeological record for this region [cf. Case study 4 below: this record is not discussed here, but it should be noted that this palaeogeographic study also contributes to the study of the regional archaeology (Hall 1983; Hall & Hiscock 1988) by (i) defining available past environmental resources, (ii) placing time limits on the use of particular resources, (iii) placing geographical limits on locations of occupation, and (iv) providing details of the nature of human occupation in this area] which provides a sensitive indicator of past resource availability and exploitation related to the state of the swamp, and thus serves as a proxy palaeohydrology-palaeovegetation record.

CASE STUDY 4, COASTAL SEDIMENTS AS CULTURAL HERITAGE RESOURCES: HOLOCENE COASTAL EVOLUTION OF DECEPTION BAY Deception Bay, in southeast Queensland, forms the most northern mainland element of the larger, more widely known Moreton Bay. It is a low-energy tidal embayment with a shoreline characterised by a relatively flat coastal plain itself dominated by estuarine mudflats and prograded beach ridges. The local prehistoric archaeological record of Deception Bay features a diverse range of site types including a ceremonial bora ground, a possible stone-fish trap, the well-documented Sandstone Point midden complex, a probable swamp-fern processing site, and numerous undated shell middens (Cotter 1996). The chronological evidence suggests that prehistoric occupation and resource exploitation of Deception Bay has occurred only within the last 2 0 0 0 years in contrast to regional evidence indicating occupation of the southeast Queensland coastal lowlands for the past 2 0 0 0 0 years (Hall 1987). This rich cultural record is considered worthy, both at State and local government level, of conservation and thus there are cultural considerations within the decision-making process of planning for competing land uses. However, it is often apparent that anything other than a superficial understanding of the nature of the cultural artefactual evidence is available, in part since much of it is buried, and in part since much of it is only recorded by accidental finds rather than by analytical research. In an analytical study, it becomes immediately apparent that a firm understanding of past environments provides the essential context for the artefactual evidence

273

and the story that it tells. Such an understanding is of urgent importance in this region, given the proximity of the greater Brisbane metropolis and the demands upon land and geological resources that the growth of such a metropolitan area imparts upon relatively undeveloped land (Tarte 1993). There are pressures both to expand residential development and to use the area for sediment extraction; both of these will impact upon the cultural heritage of the area, and until that heritage is fully understood, appropriate planning decisions about these potential destructive land uses cannot be made. The essential link between the geological study of the area and its cultural heritage is that: (i) the artefactual evidence for past human occupation of the area is embedded in the geological matrix; and (ii) the geological matrix provides detailed evidence for the nature and distribution of the past environmental conditions which in part control the location, distribution, type and survival of artefactual evidence. In the context of an environmental archaeological study this case study describes summary details of palaeobotanical and geochemical evidence for the mid- to late Holocene coastal evolution of Deception Bay. Radiocarbon and thermoluminescence age determinations of several sediment facies in the bay also provide a chronometric framework for this coastal evolution, a framework directly comparable to the archaeological record (Cotter 1996). In light of this comparison, discussion is directed towards an assessment of the likely impact of this Holocene coastal evolution on the formation, preservation and current visibility of the archaeological record of the Bay. Finally, an assessment is made of: (i) the importance of understanding coastal geomorphic processes for the effective management of prehistoric cultural heritage within the coastal zone; and (ii) the specific implications of the Holocene evolution of Deception Bay on the management of the prehistoric cultural heritage of northern Moreton Bay. The evidence suggests that coastal progradation, windblown sand transport, and the intermittent formation of tidal lagoons adjacent to the shoreline are features of the geomorphic history of the bay (Figure 23.4). In particular evidence for influxes and subsequent declines in mangrove pollen types (i.e. Rhizophoraceae, Avicennia and Aegiceras) through time, in sediments situated 1.75 km landward of the present coastline, provides clear evidence both for sealevel fluctuations in the region during the mid-Holocene as well as for subsequent progradational events. Geochemical evidence of acid-sulfate soil conditions concurs with this palaeobotanical evidence for a past marine transgressive event in the bay. Importantly AMS radiocarbon dating of associated sediments indicates that the most significant positive change in sea-level/tidal regime occurred at ca 4 6 0 0 a BP. Later, as sea-level began to drop to present levels, shoreline progradation was facilitated by a relatively stable supply of marine sediments so that a characteristic ridge and swale system has developed parallel to the present coastline. In addition, particle-size analyses indicate that aeolian sands have accumulated over the earliest prograding dunes obscuring much of the microtopography. This geomorphic history provides some explanation of both the spatial and temporal patterns observed in the local archaeological record of Deception Bay, and it has several implications with regard to the formation, preservation and

274

W. E. Boyd et al.

Figure 23.4 Aspects of the physical evolution of the north Deception Bay (southeast Queensland) coastal landscape during the last 6000 years, based on field mapping and stratigraphic sediment analysis, (a) ca 6000 a BP. Sea-level was about stable with present levels but water encroached further landward than at present. This westerly extension of marine conditions resulted because Deception Bay was at this time a flat coastal plain with no previous coastal features of Holocene age to act as impediments to inundation, (b) ca 4600 a BP. Sealevel, or at least the tidal regime, was higher than at ca 6000 a BP. The coastal 'barrier dune' that began forming -800 years earlier was breached at its northern section adjacent to a probable palaeochannel. The newly inundated area behind this coastal barrier dune, with egress to the tide, became a coastal lagoon dominated by mangroves, (c) ca 2500 a BP. A rapid decline in sea-level at ca 3000 a BP completely isolated the former tidal lagoon eliminating mangrove taxa from the area. In this still predominantly low-lying area, Melaleuca swamp conditions developed and expanded. The sea-level decline exposed marine sand of the coastal plain to wind transport and these sands were deposited in the swales between the barrier dunes and on top of these sand-ridge features. The shoreline began to prograde seawards as sediment supply to shoreline was maintained and sea-level continued to decline to present levels, (d) Present day. Today, Bayside Drive Lagoon is in a phase of retreat as siltation of the inlet zone restricts seawater influx to major spring tides. From the dating

Historical change studies

275

evidence it appears that this lagoon formed sometime in the last 2000 years under similar conditions to a previous lagoon [see (b)]. The continued deposition of sand as spits seaward of the present coastline suggests that a continuation of the present tidal, sedimentary and climatic regimes will result in the further production of a 'barrier dune' behind which a future lagoonal system may develop.

276

W. E. Boyd et al.

interpretation of that record. First, it is unlikely that evidence of early Holocene or late Pleistocene occupation of the coastal lowlands will be found in this region. Moreover, if Pleistocene occupation occurred in the area, all evidence would now be overlain by several metres of Holocene sands. Secondly, the palaeobotanical evidence suggests that the tidal lagoon system presently featured along the Deception Bay shoreline is a modern consequence of an intermittent but repeated geomorphic process operating in the area throughout the mid- to late Holocene. It is probable, therefore, that the resource zones presently observable in Deception Bay (including estuarine mudflats, tidal lagoons and Melaleuca swamps) existed in some form throughout the last 5000 years, and hence were available throughout this period for prehistoric human exploitation despite the lack of archaeological sites of this antiquity. Thirdly, artefactual material noted eroding out of dunes dated to ca 5000 a BP suggest that elements of the midHolocene component of the regional archaeological record have been overlain with aeolian sands, rendering them invisible in the present landscape. This provides at least one environmental reason for the lack of observable sites of more than 2500 years old in the area. Finally, the unconsolidated sediments of the estuarine mudflats and prograded dunes of the coastal plain do not provide the raw material resources for many of the lithic artefacts found in Deception Bay. As a consequence artefacts found in these dunal systems must have been transported to the area, perhaps through trading networks. It is likely therefore that a lack of suitable geological outcrops for stone tool manufacture in northern Deception Bay impinged upon the nature and type of social interactions in which the prehistoric occupants of the area engaged. In terms of the implications for prehistoric cultural heritage management in Moreton Bay, it is clear that a full appreciation of the spatial and temporal patterns observed in the archaeological record of Deception Bay is dependent upon a detailed understanding of the nature, timing and extent of coastal geomorphic change in the bay. This finding has three implications for cultural heritage management initiatives and practice in the coastal zone, and specifically for northern Moreton Bay. First, prehistoric cultural heritage management in the coastal zone requires the adoption of interdisciplinary studies for the effective determination of the significance and vulnerability of the local and regional archaeological record. Of paramount importance is the determination of the effect of coastal processes on the form, preservation and visibility of the known archaeological record (Boyd 1982, 1999; Boyd et al 1996a). Second, in Deception Bay it is apparent that there exists a subsurface component to the prehistoric archaeological record. Cultural heritage management strategies that are designed to ascertain and protect surface exposures of cultural material only, are therefore likely to be inadequate. Finally, northern Moreton Bay exists within one of the fastest growing regional areas of Australia. To maximise the effectiveness of cultural heritage management in this rapid urban growth area, prior to new development subsurface assessments for cultural heritage material should be carried out. This risk assessment of buried archaeological material would thereby allow for the most appropriate development strategies to be adopted for the coastal zone of northern Moreton Bay.

The effective management of a rich and diverse prehistoric archaeological record in northern Moreton Bay is clearly dependent on a thorough knowledge of landscape evolution and the dynamic geomorphic processes in operation within this coastal zone. It has been demonstrated here that the geomorphic evolution of Deception Bay has impinged on the preservation and current visibility of the known archaeological record. Moreover, by indicating the existence of a subsurface component to the archaeological heritage of the area, the need for cultural heritage management strategies which incorporate buried cultural material risk assessments in their design and implementation has been highlighted. With a thorough knowledge of such risks informed and appropriate strategies for development can be formulated for this rapid urban growth area.

CASE STUDY 5, ENVIRONMENTAL MANAGEMENT AND BURIED CULTURAL HERITAGE: RICHMOND RIVER MOUTH The aim of the case study described here is to assess coastal and fluvial erosion threats, both past and present, to the historic shipwreck heritage of the north coast of New South Wales. The project described addresses concerns regarding the conservation of these heritage sites by identifying site-specific coastal and fluvial geomorphological characteristics, and by providing essential data on real or potential past and present erosion or exposure threats to known shipwreck sites. The shipwreck site characteristic identified as being most important, in terms of heritage conservation and management, is the geological and geomorphological history of these sites. Reconstruction of this history provides geographical limits to the original locations of wreck sites, indicates possibilities of wreck preservation, and defines the geographical limits on probable present wreck location (Boyd et al 1995). Such palaeogeographical heritage modelling has been extended by examining the potential risk of damage to possible buried shipwrecks in the Richmond River mouth by developing a Geographical Information System (GIS) in which layers of data, including data regarding geomorphological evolution, interact to model the likelihood of disturbance to a particular site (Boyd et al 1996b). Such a model exemplifies the possible input of geological mapping to heritage management decision making. To form a basis for management of the historical shipwreck heritage of any part of the coast, it is necessary to appreciate the geological and geomorphological history of that part of the coast. This enhances understanding of both the original distribution of wreck sites, and the subsequent spatial distribution and probability of preservation of the wrecks. This study focuses on the mouth of the River Richmond at Ballina, northern New South Wales, where there are 94 recorded historical shipwrecks dating from prior to 1845 through to 1908 (New South Wales Department of Urban Affairs and Planning 1995). The exact locations of many of these are unknown, although they can be estimated. To provide appropriate conservation or preservation management of this rich shipwreck heritage it is important to know the following: (i) the identification of shipwrecks in the area; (ii) the location of the shipwrecks;

Historical change studies

277

Figure 23.5 The evolution of the river mouth of the Richmond River, northern New South Wales. This evolutionary model is based on an assessment of historical maps and documents that provide evidence for periodic changes in the geomorphology of the river mouth during the 19th century.

278

W. E. Boyd et al.

Figure 23.6 Outputs from GIS modelling of the risk of damage by disturbance to buried heritage objects (in this case shipwrecks) in the area of the mouth of the Richmond River, northern New South Wales. The outputs are based on numerical models in which the relative influences of natural and human characteristics are varied. The natural characteristics provided the basic physical model, and included land surface elevation, surface geology, slope, soil type, land cover, terrain, shore proximity and type and degree of erosion. An important physical contribution to this model is the understanding of the historical geomorphological evolution of the river mouth (Figure 23.5). While some human characteristics (shipwreck location and landuse patterns) provided input as model elements, these were more important as inputs of potential land-modification processes. The relative importance of each model element was varied by weighting it relative to estimates of its contribution to the risk under various forms of potential land-use practice. Models 1 to 3iii represent various combinations of these variables. These provide input to local government planning decisions about potentially disruptive activities.

Historical change studies arid (iii) the environmental circumstances or context of shipwreck locations. These three items of information then allow management options to be developed which take full account of both the historical value of the individual wreck and the degree of environmental threat of exposure either by natural erosion or human induced damage. This approach differs slightly from the process of assessment used by the New South Wales Department of Planning (1990, 1994), whose criteria for assessment focus only on the nature of significance (historical, archaeological, etc.) and the degree of significance (inclusive of rarity and representivity). This case study represents an attempt to extend the assessment criteria by gathering and integrating both historical and geomorphological data for the Richmond River mouth, in other words by examining the recent geology of the area. A survey of historical documents, maps and charts, and aerial photographs provides evidence for major changes in the geometry of the river mouth at Ballina during the historical period (Figure 23.5). These changes can be dated reasonably accurately and provide evidence for a changing sequence of geographical conditions throughout the 19th century. Examination of this sequence allows certain areas of the present mouth of the Richmond River estuary to be identified as remnant features of either former river mouths or sand bars. These features, when placed in chronological context, may then be associated with shipwrecks of the corresponding period. From this information it is also possible to identify areas in which, due to subsequent sedimentation, the preservation potential of particular shipwrecks is high (Figure 23.6). In some of these areas, recent land-use activities (for example, the development of a residential housing estate on a former tidal sand bank and subsequently a supratidal dune field) ensure no further disturbance to possible underlying wrecks, and thus provide one form of heritage protection in this area. This study, therefore, identifies sites which naturally conform to one of the conservation methods advocated by the New South Wales Department of Planning, namely the 'protection by covering the heritage item with sand or a synthetic material...' (New South Wales Department of Planning 1994 p. 8). The ultimate purpose of such a study is to contribute to the decision-making process, usually vested at local government level, associated with land management and planning permission. To this end, the geological study of the river mouth provides only one, albeit important, input to the database required to support planning decisions. Where there is a conflict of interests between various landuse issues, and where these land-use issues may have an environmental consequence, it is necessary to model for all interactions between environmental parameters. Of particular interest in this case was the need for cultural heritage and other environmental managers to be able to develop appropriate management strategies for the protection and conservation of such material remains, especially in the light of potential changes in land use and the continuing impacts of erosive environmental processes in the coastal zone. To that end, the geological data described above were incorporated into a management support system in the form of a risk assessment GIS model. The risk being evaluated is that of the possible disturbance or exposure of potential buried historical shipwrecks in this coastal area.

279

The model does not attempt to predict the actual location of possible buried shipwreck remains, but focuses on the degree and, to a lesser extent, nature of the risk of disturbance to any buried materials that realistically may be present. At this stage in the development of the model, it appears that the model is moderately robust, although there is scope for some fine tuning. The model appears to be applicable to different situations (e.g. other areas of coastline) and to be usable as a model for proposed future changes in, for example, land-use zonation.

CONCLUSIONS The case studies presented here describe the contribution of geological-history perspectives to issues of environmental management and planning, and are representative of the situation in many parts of the coastal zone. The common theme here is the geological context—that the knowledge gained about the history and development of an environmental through geological methods may play an integral part in contributing to the successful assessment of natural and human resources and the subsequent management of these resources.

REFERENCES B . E . 1 9 8 6 . Handbook of Holocene Palaeoecology Palaeohydrology. John Wiley, Chichester.

BERGLUND

and

BOYD W. E. 1982. Archaeological implications of a new palaeo-envi-

ronmental model for part of the Ayrshire coast. Glasgow Archaeological Journal 9, 15-17. W. E. 1990. Quaternary pollen analysis in the arid zone of Australia: Dalhousie Springs, Central Australia. Review of Palaeobotany and Palynology 64, 331-341. BOYD W. E. 1993. Proto-Swan Bay: a palaeogeographical model of the Late Holocene Eighteen Mile Swamp, North Stradbroke Island, S.E. Queensland. Proceedings of the Royal Society of Queensland 103, 75-83. BOYD W. E. 1999. Environmental archaeology of the coasts of S.E. Asia and P.N.G. In: Hall J. & McNiven I. J. eds. Australian Coastal Archaeology: Current Research and Future Directions, pp. 307-315. ANH Publications, Canberra. BOYD W. E., CHARTER C . J. & COTTER M. M. 1995. A contribution to the management of the historic shipwreck heritage, northern New South Wales: the environmental context of the Richmond River mouth shipwreck heritage at Ballina, New South Wales. Bulletin of the Australian Institute for Maritime Archaeology 19, 37-43. BOYD W . E . , HICHAM C . F. W . & THOSARAT R . 1996a. The Holocene palaeogeography of the S.E. margin of the Bangkok Plain, Thailand and its archaeological implications. Asian Perspectives 35,139-162. BOYD W . E . , LAING A. I . , DICKSON J . H . & STEVENS G . 1 9 8 8 . The history and present management of the rare endemic trees on the Island of Arran, Scotland. Environmental Conservation 15, 65-66. BOYD W. E . , MCGRATH R . J . , COTTER M . M . & PETERS R . 1 9 9 7 . Long term vegetation change in the coastal wallum of northeast New South Wales and southeast Queensland. In: Proceedings, I 9 9 7 Coastal Conference, Ballina, pp. 3 7 9 - 3 9 0 . Ballina Shire Council, Ballina. BOYD W. E., PATHIRANA S. & BELL G. 1996b. A contribution to the management of buried historical shipwreck heritage: a GIS risk assessment model for the River Richmond mouth, northern New South Wales. Bulletin of the Australian Institute for Maritime Archaeology 20, 57-70 BOYD W . E . , SPECHT A . & COTTER M . M . 1 9 9 9 . Integration of archaeological. palaeogeographical and ecological investigations on Stradbroke Island, S.E. Queensland. In: Hall J. & McNiven I. J. eds. Australian Coastal Archaeology: Current Research and Future BOYD

280

W. E. Boyd et al.

Directions, pp. 185-189. ANH Publications, Canberra. CHALSON J. M . & MARTIN H. A . 1 9 9 5 . The pollen morphology of some co-occurring species of the Family Myrtaceae from the Sydney region. Proceedings of the Linnean Society of New South Wales 115,

163-191.

R. J., COVECEVICH J. & DAVIES P. (Editors) 1984. Focus on Stradbroke: New information on North Stradbroke Island and surrounding areas, 1974-1984. Boolarong Publications, Brisbane. COTTER M . M . 1 9 9 6 . Holocene environmental change in Deception Bay, Southeast Queensland: a palaeogeographical contribution to MRAP Stage II. Tempus 6, 1 9 3 - 2 0 5 . DODSON J. R . , GREENWOOD P. W . & JONES R . L. 1 9 8 6 . Holocene forest and wetland vegetation dynamics at Barrington Tops, New South Wales. Journal of Biogeography 1 3 , 5 6 1 - 5 8 5 . DODSON J. R . , DE SALIS T., MYERS C . A . & SHARP A . J. 1 9 9 4 . A thousand years of environmental change and human impact in the alpine zone at Mt Kosciusko, New South Wales. Australian Geographer

COLEMAN

25,

77-93.

DUTTON I . M . , JENNER G . , SPECHT A . & THOMSON G . 1 9 9 3 . T h e u s e o f

vegetation and water resource monitoring as an input to environmental management—A case study from North Stradbroke Island. In: Proceedings of the Australian Water and Wastewater Association 15th Federal Convention, pp. 4 1 3 - 4 1 8 . Australian Water and Wastewater Association, Brisbane. ERDTMAN G . 1 9 5 2 . Pollen Morphology and Plant Taxonomy: Angiosperms (An Introduction to Palynology. I). Almqvist and Wiksell, Stockholm. FLANNERY T. F. 1994. The Future Eaters. An ecological history of the Australasian lands and people. Reed, Port Melbourne. FLOOD P. G., GRANT B. P. & FRANKEL E. 1986. The origin of Eighteen mile Swamp: a modern back-barrier peat-forming environment, North Stradbroke Island, southeast Queensland. In: Frankel E., Keene J. & Waltho A. E. eds. Recent Sediments in Eastern Australia—Marine through Terrestrial, pp. 131-140. Geological Society of Australia, New South Wales Division, Sydney. FLOWER R . F., BATTERBEE R . W . , NATKANSKI J . , RIPPEY B . & APPLEBY P. G .

1988. The recent acidification of a large Scottish loch located partly within a National Nature Reserve and Site of Special Scientific Interest. Journal of Applied Ecology 25, 715-724. GELL P. 1 9 9 6 . Basin Lakes Quarry loses appeal. Quaternary Australasia 14, 4. HALL J. 1 9 8 7 . A short prehistory of the Moreton Region. In: Fisher R. ed. Brisbane: Aboriginal, Alien, Ethnic, pp. 1 4 - 2 2 . Department of History, University of Queensland, Brisbane History Group Papers 5. HALL J . & HISCOCK P. 1 9 8 8 . The Moreton Region Archaeological Project (MRAP) Stage II: an outline of objectives and methods. Queensland Archaeological Research 5, 4-24. HALL I. W . 1 9 8 3 . Archaeological research project concerning North Stradbroke Island and adjacent isles. In: Cameron MacNamara & Partners eds. North Stradbroke Development Strategy, Appendix B., pp. 1-10. Report to the Queensland Premier's Department, Brisbane. HOPLEY D. (Editor) 1983. Australian sea levels in the last 15,000 years: A review. Department of Geography, James Cook University, Townsville. LAMB H. H. 1995. Climate, History and the Modern World (2nd edition). Routledge, London.

M. E. 1997. Quaternary palynological records from perched lake sediments, Fraser Island, Queensland, Australia: rainforest, forest history and climatic control. Australian Journal of Botany 45, 507-526. MCGRATH R. J. 1995. Environmental reconstruction at Bundjalung National Park and Bungawalbin Creek on the New South Wales north coast. Integrated project thesis, Southern Cross University, Lismore (unpubl.). MCGRATH R. J. 1996. An S E M examination of the palaeoecological history of Melaleuca alternifolia in the Bungawalbin Creek catchment, northeastern New South Wales. BSc (Hons) thesis, Southern Cross University, Lismore (unpubl.). MCGRATH R. J. & BOYD W. E. 1998. Holocene vegetation history of Bundjalung National Park and Bungawalbin Creek, northeastern New South Wales. Australian Geographer 29, 205-221. MCINTYRE D. J. 1963. Pollen morphology of New Zealand species of Myrtaceae. Transactions of the Royal Society of New Zealand Botany 2, 83-107. N E W SOUTH W A L E S DEPARTMENT OF PLANNING 1990. Heritage Assessment Guidelines. New South Wales Department of Planning, Sydney. N E W SOUTH WALES DEPARTMENT OF PLANNING 1994. Underwater Heritage. New South Wales Department of Planning, Sydney. LONGMORE

N E W SOUTH WALES

DEPARTMENT OF URBAN AFFAIRS AND PLANNING

1995. Shipwreck Atlas of New South Wales (2nd edition). New South Wales Department of Urban Affairs and Planning, Sydney. PETERS R. 1990. Emu Swamp—From glaciation to 'grassed plains': vegetation history and landuse conflict at Emu Swamp, sunshine Coast, Queensland. Integrated project thesis, University of New England—Northern Rivers, Lismore (unpubl.). PICKUP G , WASSON R . J . , WARNER R . F., TONGWAY D . & CLARK R . L.

1987. A feasibility study of geomorphic research for the long term management of uranium mill tailings. CSIRO Divisional Report 87/2, Canberra. PIKE K. M. 1956. Pollen morphology of Myrtaceae from the southwest Pacific area. Australian Journal of Botany 4, 13-53. ROBERTS N. 1989. The Holocene: An Environmental History. Basil Blackwell, Oxford. SPECHT A. 1992. Vegetation monitoring of Eighteen Mile Swamp, North Stradbroke Island 1992. Centre for Coastal Management, Lismore. TARTE L. 1993. Moreton Bay: coastal resource management perspectives. In: Greenwood J. G. & Hall N. J. eds. Future Marine Science in Moreton Bay, pp. 63-66. School of Marine Science, University of Queensland, Brisbane. WASSON R. J. & CLARK R. L. 1985 Environmental history for explanation and prediction. Search 16, 258-263. WILLIAMS M . A. J . , DE DECCKER P. & KERSHAW A. P. (Editors) 1 9 9 1 . The Cainozoic in Australia: a re-appraisal of the evidence. Geological Society of Australia Special Publication 18. WILLIAMS M . A . J . , DUNKERLEY D . L . , DE DECCKER P., KERSHAW A . P. &

T. 1993. Quaternary Environments. Edward Arnold, London. M. & HEUSSER C. J . 1 9 9 6 . Late-glacial palynology of the Myrtaceae of Southern Chile. Review of Paleobotany and Palynology 91, 2 8 3 - 3 1 5 . STOKES

ZHOU

Received 3 October 1997; accepted 18 October 2000

Geological Society of Australia Special Publication 21, 281-294

CHAPTER 24—Holocene Great Barrier Reef: sedimentary controls and implications for environmental management Marine Geophysical Laboratory, School of Earth Sciences, James Cook University., Townsville, Qld 4811 Australia (piers.larcombe@jcu.edu.au) The Great Barrier Reef is one of the most visible biological structures on the Earth's surface visible from space. In the last half-million years, the corals that form the basis of this massive ecosystem have died as sea-levels have fallen during ice ages (down to over 100 m below modern levels) and been reinitiated as they rose again. For the last 8000 years we have been in the most recent phase where corals flourish, and in the future the reef system as we know it will inevitably die and corals will once more be confined to a series of geographically restricted refugia. Maximum rates of coral growth and accumulation of calcium carbonate tends to occur during stages where suitable reef substrates are submerged and able to grow upwards unrestricted by water depth—this is likely to occur during the latter stages of each phase of sea-level rise. Coral growth is affected by a range of environmental factors other than sea-level change, including the presence of mobile sediments and water turbidity, but the precise nature of the relationships is yet to be determined for the evolving Holocene reef system or the modern reef. Recent measurements of turbidity and sediment transport at coral reefs have documented far greater levels of turbidity than previously known, and the presence of coral reefs long having occurred in turbid environments can also be interpreted from the geological record. Geological data, by virtue of providing a temporal record of change, is invaluable in assisting environmental management of the reef system by helping to identify potential human impacts on the reef system. Two examples are the historical occurrence of the crown-of-thorns starfish, and the potential threat from increased sediment input. KEY WORDS: coral reefs, dredging, environmental management, Great Barrier Reef, Holocene, sealevel, sedimentation.

HOLOCENE SEA-LEVEL RISE AND CORAL GROWTH IN NORTHEAST AUSTRALIA The Great Barrier Reef (Figure 24.1) is a relatively young feature on the Earth. Drilling through the edge of the modern Great Barrier Reef near Cairns has demonstrated that the age of the first Great Barrier Reef is probably less than 500 000 years (Davies & McKenzie 1993). Since then, the reef has undergone cycles of reinitiation and growth. The reef flourishes during highstands of sea-level, when the polar ice caps are relatively small, only to die again as the sea-level retreats towards glacial lowstands. Major fluctuations in sea-level occur over time periods of -125 000 years (Figure 24.2a). We are presently in the latter stages of the most recent sea-level highstand, with a reef tract which is only 8-9000 years old. A few thousand years into the future, sea-level will again fall, exposing the Great Barrier Reef shelf to the atmosphere along with thousands of carbonate reefs as limestone crags on a wide, low gradient fluvial/coastal plain. Global changes in sea-level (i.e. eustatic sea-level) may be inferred from the oxygen isotope curves derived from analyses of the skeletons (tests) of microscopic marine organisms, which, upon their death, are deposited on the sea floor, as biopelagic sediments (Shackleton & Opdyke 1973; Linsley & Thunnel 1990; Hodell et al 1991). The most common such sediments are Globigerina ooze (foraminifers, carbonate-rich, typical of warm surface waters) and Radiolarian ooze (silica-rich, typical of cold, polar waters). Approximately 70% of the isotope signal in the tests of

foraminifers is caused by melting or freezing of polar ice and the consequent change produced in the 1 8 0 / 1 6 0 ratio of seawater. Changes in temperature or salinity through time can also influence the curve obtained (Nelson et al 1993). For the Great Barrier Reef region, Peerdeman and Davies (1993) provided a curve from a core taken by the Ocean Drilling Project on the upper continental slope off Cairns (core location shown in Figure 24.1). Covering the last 20 000 years, the curve does not swing monotonically from less positive to more positive isotope values, but is rather punctuated by a number of short intervals when the isotope signature becomes more positive (Figure 24.2b). Some of the excursions this curve displays must be due to changes in ocean temperature or salinity, because the major (0.4%o) shift towards the top of the core is younger than 4.9 ka BP and would indicate 40 m of equivalent sea-level fall, in stark contrast to the depositional evidence (outlined below). Many such curves show such features globally. As the precision of oxygen isotope curves increases, it is important to understand which isotope shifts indicate regional changes (e.g. changes in ocean currents) and which mark global changes in ice volume and hence global sea-level.

Evidence for sea-level changes on the Great Barrier Reef A wealth of information exists on Holocene relative sealevel for the Great Barrier Reef shelf obtained from a range of types of sea-level indicator (Figure 24.3). Many sea-level

282

P. Larcombe

tss

10 S

17 S

18$

19 S

20 S

Figure 24.1 Map of the cen-

150 6 145 6

146 E

147 B

148 E

indicators are sedimentary materials sampled from cores (e.g. mangrove muds, palaeosols and peats), while others are biological remains deposited in situ (e.g. coral reef cores, coral microatolls and oyster beds). Ages of these indicators are commonly determined using radiocarbon dating, and the subsequent potential position of sea-level plotted on an age-elevation diagram. Overall, since the last glacial maximum at ca 18 ka BP, the sea has risen by -120 m. The shelf was flooded and the modern Great Barrier Reef system was initiated in the last 8-9000 years. Studies using drilling and seismic profiling reveal that the modern shelf reefs occur as caps on top of raised platforms of Pleistocene limestone (Davies & Hopley 1983; Johnson et al 1984; Walbran 1994) of thickness between 4 and 24 m (see review by Harvey & Searle 1983 and references therein). The precise nature of the sea-level rise is discussed below. The data used to determine ancient sea-levels are rarely precise, resulting in opportunity for disagreement on both the broad nature and finer detail of the post-glacial sealevel rise. The primary mechanism of relative sea-level change has been eustatic (Chappell et al 1982, 1983; Nakiboglu et al 1983). The sea-level envelope of Thorn and Roy (1983) is perhaps the most widely quoted post-glacial sea-level curve for eastern Australia and is the basis for the best-known interpretations of reef growth (Hopley 1982; Davies & Hopley 1983). The sea-level envelope shows an overall continuous rising trend, though Thorn and Roy were careful to note that rates of rise may have varied, and that the sea-level may even have fallen at some stage. Other workers have suggested explicitly that post-glacial sea-level rose in a series of rapid rises, with intervening stillstands or even minor falls in sea-level (Carter & Johnson 1986; Carter et al 1986), citing as evidence the presence of drowned shorelines and reefs on the shelf shown on seismic profiles

149 E

tral Great Barrier Reef shelf showing the locations of some places mentioned in the text.

and sampled in cores. Harris and Davies (1989) have criticised this episodic model because some submarine features of the type described by Carter and his coworkers are not laterally continuous. This debate continues, with recent stratigraphic and other work again presenting evidence for episodic sea-level rise (Gagan et al 1994; Larcombe et al 1995b; Larcombe & Carter 1998) prompting discussion (Harris 1999; Larcombe & Carter 1999). The particular significance of the debate about rates of sea-level change for the Great Barrier Reef shelf is the implications for the rate of vertical accretion of coral reefs. Coral accretion is the sum of constructional and erosional processes which result in overall accumulation of reef material. This is distinguished from coral growth, which is the growth of the exoskeletons of individual coral colonies (Buddemeier et al 1974). In general, the maximum rate for sustained accumulation of coral reefs (i.e. coral accretion) is ~5 mm/y (Smith & Kinsey 1976; Smith 1983), although in shallow water (<5 m) rates of up to 8-15 mm/y may occur (Hopley & Kinsey 1988; Chappell & Polach 1991; Collins et al 1993). If sea-level rise was episodic, there would have been periods of sea-level rise faster than 10 mm/y, during which coral reefs would have been unable to keep up with sea-level. At the end of each rise, they would have been in deeper water than when they started, and if the rise was of a large magnitude, then environmental conditions in the deeper water may not have permitted coral survival. Coral reefs may thereby have been drowned during phases of rapid sea-level rise (the 'give-up' reefs of Neumann & Mclntyre 1985), whereas they may have flourished during phases of slower sea-level rise Ckeep up' reefs), or during stillstands, when some reefs might 'catch-up' with sea-level. Some reefs may have been preferentially initiated at certain times, and thus in a certain range of water depths and locations across the shelf.

Holocene Great Barrier Reef

283

Age (Ka) 140

120

100

0 -10 -20

-30 -40 -50 -60

-70

-80 -90 -100 -110 -120 -130

-2.5

0) Q) -Q 3 C£ a: < D (D Figure 24.2 (a) Proxy sea-level curve for the last 140 000 years derived from oxygen isotope data from deep sea cores (•) and coral terraces at Huon Peninsular, Papua New Guinea (•) (after Chappell 1994). (b) Post-glacial oxygen isotope curve of Peerdeman and Davies (1993) from a core taken near Cairns, with potential matches to the curve of Larcombe et al (1995b) for the last 9000 years (after Larcombe et al 1995).

CD oo Q o aQ.

-1.5

y\/' p

O "O

Sea-level Curve

k /

-1.0 Lowstand

Post-glacial Transgression

Regression

-0.5

Highstand

Recent analysis of coral reef d a t a on the Great Barrier Reef shelf, including d a t a from drillcores, indicates that modern coral reefs that occur in shallow turbid water (the 'coastal turbid-zone reefs' of Larcombe & Woolfe 1999a) probably o c c u p i e d similar environments during the late stages of the Holocene marine transgression (Figure 24.4). The modern d a t a indicate that such reefs tend to occur in waters shallower than 4 m below m e a n sea-level. In slightly deeper water, agitation by w a v e s m a y be insufficient to prevent long-term net sediment accumulation, and/or light levels at the b e d m a y b e too low (see below). It is thus possible to distinguish 'coastal turbid-zone reefs' from those reefs in deeper water ('inner a n d mid-shelf reefs') on the Holocene central Great Barrier Reef, a n d to infer the potential growth p h a s e s for e a c h group in the light of sea-level curves. At present, only a small proportion of reefs in shallow water live in turbid conditions, b e c a u s e there h a s been sufficient time for the m a n y reefs initiated on the mid- a n d outer shelf to reach sea-level. The first-order fit to these groupings is

Depth (m below sea floor) clear, but much research is needed to analyse these inferred relationships further.

CONTROLS ON CORAL GROWTH: EFFECTS OF SEDIMENTS, AND THE DISTINCTION BETWEEN TURBIDITY AND SEDIMENT ACCUMULATION Water turbidity is the relative optical clarity of water a n d is typically m e a s u r e d using optical backscatter s e n s o r s . Turbidity m a y be c a u s e d by sediment grains, organic particles, dissolved c o m p o u n d s , or even air bubbles in the water. It is distinct from s u s p e n d e d sediment concentration which is the weight of s u s p e n d e d sediment contained in a unit volume of water (typically in units of mg/L or g/L). From these simple definitions, it is clear that although water m a y be turbid, it may have a low s u s p e n d e d sediment concentration a n d if so, its capacity to c a u s e sediment accumulation on the s e a b e d may b e very small. In a

284

P. Larcombe

c o

m m

Figure 24.3 Elevation ranges of various sea-level indicators used on the central Great Barrier Reef shelf (after Larcombe et al 1995b). Directional sea-level indicators can only indicate an upper or lower limit to the elevation of mean sea-level, whereas the finite indicators can place an upper and lower limit. The modern elevation range of the finite sea-level indicators is shown by the the length of the vertical bars and the uncertainty in the location of ancient sea-levels derived from their dating is up to twice this length. geological context, sediment accumulation is the vertical accumulation of a body of sediment due to deposition of sediment from an overlying water body. Many people assume that turbid water always results in sediment deposition, and that turbid water observed near a coral reef results in the deposition of sediment upon that reef. But there is a problem here. We should recognise instead that turbid water may be simply carrying suspended sediment past a point, without necessarily depositing it. Finally, there are currents which actually erode the bed and become turbid in the process. Therefore, all depositional currents are turbid, but not all turbid waters are depositional! In considering the relationships between coral reefs, turbid waters and sediment accumulation, it is therefore very important to distinguish between water turbidity and sediment accumulation. Sediment affects the ability of corals to grow and survive, through three main physical effects: (i) sediment accumulation onto a coral may reduce coral performance through excessive use of energy in activating sediment rejection mechanisms (Stafford-Smith & Ormond 1992)—higher levels of accumulation can completely smother the coral polyps; (ii) high levels of suspended sediment in the water column will decrease light levels, and inhibit or halt photosynthesis of the algae which live within

the coral tissues, from which the corals gain their energy; and (iii) the soft tissues of the coral may be damaged through abrasion or impact by sediment particles (especially sandy grains). Other potentially adverse effects of suspended or deposited sediments can be caused through the exchange and/or supply of nutrients, chemical compounds and elements. Corals display a range of features indicative of sublethal stress, including sediment stress, such as bleaching, extrusion of mesenteries, and mucus production (Brown & Howard 1985; Pastorok & Bilyard 1985; Rogers 1983, 1990; Stafford-Smith 1992; Stafford-Smith & Ormond 1992). Continued high levels of stress may eventually result in coral death. After intense rainfall, much publicity often accompanies the discharge of large muddy plumes of freshwater from rivers onto the Great Barrier Reef shelf. Are these plumes 'a threat to the reef? Will coral reefs be smothered by mud? Is the land-use change in the hinterlands associated with European settlement a significant factor in changing sediment input to the Great Barrier Reef lagoon? In terms of sedimentary processes, the combination of a very high sediment discharge and a low-energy hydrodynamic regime at the coast (e.g. weak tidal currents and no waves) is favourable to

Holocene Great Barrier Reef

Inferred turbid-zone reefs of late transgression

5 <

285

Field for potential — c o a s t a l turbid-zone reefs -

-10

TLI

-20

Sample elevations of coral data only.

"Coastal turbidzone reefs" PHASE OF REEF GROWTH Inner- and mid-shelf reefs

Drowning of some pre-9.5 ky reefs \ w

Episodic initiation, catch-up, keep-up

Lateral growth, some erosion

Lateral growth, little initiation

-N Few reefe initiated

Reef initiation

Continuous catch-up

Lateral growth, little initiation

C14age (kyrBP) Figure 24.4 Sample elevations and radiocarbon ages of coral data from the central Great Barrier Reef shelf (size of crosses show error bars), the relative sea-level curve of Larcombe et al (1995b) for the central Great Barrier Reef shelf (upper line) and a lower curve denoting the likely boundary (at -4.5 m depth) of the existence field for potential 'coastal turbid-zone reefs'. The shaded area denotes the upper half of this zone for the latest phases of transgression, in which all data are derived from reefs which today are in turbid zones. These are South Myall and Rykers reefs at Cape Tribulation, and the fringing reefs at Lindquist, Dunk, Fantome, Orpheus, Rattlesnake, Stone and Cockermouth islands, all of which lie within or close to the modern inner shelf sediment wedge (Done 1982; Johnson & Carter 1987; Johnson & Risk 1987; Hopley 1982, 1995; Mapstone et al 1989; Partain & Hopley 1989; Kleypas & Hopley 1992; Larcombe et al 1995a; Veron 1995). Note that many mid- and outer shelf reefs may also plot in this field, but are not adjacent to a sediment wedge. The figure also delineates the inferred growth phases for 'coastal turbid-zone reefs' and for those reefs in deeper water ('inner- and midshelf reefs') on the Holocene central Great Barrier Reef (from Larcombe & Woolfe 1999a).

coastal sediment accumulation. This may impact existing local coral reefs. The answer is thus that local reefs might be threatened or smothered, but this is uncertain. On a regional scale and on a time-scale of years to a few centuries we can be fairly certain that most reefs are not at significant risk, because most existing coral reefs occur in locations unfavourable to sediment accumulation and because the quantity of 'new' sediment introduced into the Great Barrier Reef lagoon is rather small (Larcombe & Woolfe 1999b). From a geological perspective, turbid river plumes are certainly the main method by which muddy sediment is introduced on to the Great Barrier Reef shelf. These plumes occur primarily during the wet season (November-March) when episodic flow events follow major rainfall events in large river catchments. Whilst these plumes may be visually very spectacular, they often contain very little sediment. For 9-10 months of the year, many north Queensland rivers may have zero freshwater flow and sediment load. Only during major flood events do they supply much fine sediment and contribute to turbidity in the nearshore turbid zone. In fact, the suspended sediment concentrations of the river plumes are generally an order of magnitude (or more) smaller than that produced by resuspension of mud from the shelf sea floor, and their sediment loads are thus too small to form a significant component of the turbidity regimes on the inner shelf (Larcombe & Woolfe 1999b).

Recent data from a plume of the Barron River, Cairns (350 km north of Townsville) (Taylor 1995) and other oceanographic evidence indicate that these 'turbid plumes' are confined to the upper 2 m of the water column and contain suspended sediment concentration of only 3-10 mg/L (see also Wolanski 1994). If such a plume extended for 45 km along the coast and for 10 km out from the coastline, it would contain 2700-9000 t of sediment, far less than is held is suspension by swell waves many times per year in some Great Barrier Reef coastal embayments (such as Cleveland Bay: see below). We therefore need to carefully distinguish sediment input to the Great Barrier Reef shelf from sediment resuspension, and from net sediment transport. When you read about such 'major' events, some quick calculations like this are very useful to help in assessing their sedimentary significance. Water quality, although not discussed in detail here, is also an important factor in coral growth. Major river floods may rapidly bring large volumes of freshwater, sediment and nutrients into the Great Barrier Reef lagoon (Queensland Department of Primary Industries 1993: Wasson 1997). The freshwater itself may constitute a problem in the marine environment, because exposure of corals to low salinity water for excessive periods may contribute to 'coral bleaching', where the corals expel their zooxanthellae and may die. The impacts of any associated nutrients

286

P. Larcombe Rate of accumulation of terrigenous material

Rate of removal of reefal carbonate

Rate of accumulation of reefal carbonate

Rate of removal of terrigenous material

Limiting physical factors

Major causes of turbidity

Rate of accumulation of terrigenous material

Reef Initiation -

stress Ideal conditions Increasing stress

Rate of removal of terrigenous material

Figure 2 4 , 5 (a) Stability-field diagram showing conceptual existence fields for coral reefs in terms of net carbonate production and terrigenous sedimentation. Fields of erosion (lower half of plot) and sediment accumulation (upper) are defined along with zones of terrigenous- and carbonate-dominance are defined by a plot of T (terrigenous accumulation) against R (reefal carbonate accumulation). Coral reefs are restricted to the central-right portion of the diagram. In general, turbidity increases away from the x-axis through sediment accumulation from turbid water (for T>0), or from erosion of existing substrate (T<0), so that 'clear-water reefs' plot near the xaxis and 'turbid-zone reefs' plot further away. Continued divergence from the x-axis leads to the formation of terrigenous-dominated shelf sediments or carbonate lags. With an increasing absolute magnitude of sedimentation, reef death occurs through a combination of light attenuation, burial or erosion. Sediment transport through an area may also lead to turbid conditions (T~0). To aid understanding of the diagram, some examples of typical sedimentary deposits and environments are given for each field, but these should not be regarded as definitive or restrictive, (b) Summary of the physical factors acting to limit the existence field for coral reefs and major physical causes of turbidity in the turbid-zone reef field (after Woolfe & Larcombe 1999).

Holocene Great Barrier Reef

287

Figure 2 4 . 6 The location of dredging activities and sedimentdispersal monitoring sites in Cleveland Bay. The fringing reefs of Magnetic Island are within 1.4 km of some parts of the dredged channel. Seagrass is mostly confined to the sheltered southeast portion of the bay in depths of 2 m or less (after Larcombe et al. 1995b). on coral growth are unknown, largely because understanding of the nature of the dynamics of sediments in the Great Barrier Reef lagoon is relatively poor, and even less wellknown are the nature of the interactions between corals, sediments and water in the field. Although early papers promoted the potential for longterm damage to coral reefs induced by raised nutrient concentrations (Bell & Gabric 1990, 1991), the long-term record of nutrient status of the waters of the Great Barrier Reef lagoon (Furnas et al. 1995) reveals no measurable change though time. However, many nutrients are directly associated with fine-grained sediments, and periods when muds are resuspended from the bed will correspond with greatly increased nutrient levels in the water column, so short-term variation in nutrient levels in Great Barrier Reef waters is likely to be high. Regionally, we could thus predict that there will be greater temporal variation and higher time-averaged nutrient levels in the muddy inner shelf of the central Great Barrier Reef shelf, say between Cairns and Bowen, than in those areas north of Princess Charlotte Bay, simply because of the presence of more mud on the inner shelf. Offshore, where less mud occurs, less regional variation is likely. Existing data do not yet allow firm conclusions, but initial indications are that there are no regional stresses caused by (increased concentrations of) nutrients. The spatial and temporal variability in the nutrient chemistry and the detailed interactions of the sediments and waters of the Great Barrier Reef lagoon should be a focus for future research. The geological record can provide useful background information here. As well as the documented Holocene patterns and mean rates of sediment accumulation in coastal embayments (Woolfe & Larcombe 1998), corals themselves can help. The radial growth patterns of some coral skeletons means that some aspects of the physical environment may be recorded in 'growth bands' within them. When sliced into sections, these banding patterns can be viewed under a fluorescent light, and the nature of the bands

assessed. Mid-Holocene coral skeletons appears to document some occurrences of major palaeo-river discharges (see review of Gagan et al. 1997), and perhaps even periods of increased resuspension of sediments from the bed. Ultimately, these might be compared with similar data from modern corals where measurements of modern sedimentary processes will be available to improve interpretation.

INTEGRATING OUR UNDERSTANDING OF CORAL REEF DEVELOPMENT WITH TERRIGENOUS SEDIMENTATION As explained above, the position of sea-level is clearly a fundamental control on growth of coral reefs (they need to be underwater!). The scientific literature reflects this fact, and most models of reef growth and evolution use sea-level as the main factor under consideration (Hopley 1982; Davies & Hopley 1983; Larcombe et al. 1995). However, there are other important environmental controls on reef growth which are integral to the story, as summarised by Hopley (1995). Those of special relevance to environmental geology relate to the presence or absence of sediments. Many coral species require clear water for their survival and a relative lack of terrigenous sediment input. Most reefs on the middle and outer shelf fall into this category (Done 1982). However, some corals on the Great Barrier Reef shelf live in conditions of turbid water and close to sediment sources. Examples of this occur on the inner shelf of the central Great Barrier Reef shelf near Townsville (Done 1982; Mapstone et al. 1989; Veron 1995; Woolfe & Larcombe 1998), and the Cape Tribulation reefs (Partain & Hopley 1989; Hopley 1995). Characteristically, these reefs occupy shallow turbid waters, are close to sources of fine terrigenous sediment, and are subject to hydrodynamic processes that tend to prevent the accumulation of the fine sediment on the corals. (Two examples of such modern

288

P. Larcombe

Days into 1993 Figure 2 4 . 7 Time series of wave data and suspended sediment concentration (SSC) data (suspended sediment concentration measured 0.3 m above bed), from outer Cleveland Bay ( - 1 2 km offshore) with tidal curve from Townsville harbour. Waves were measured with a Waverider Buoy and the period used to separate swell waves from wind-waves is 7 s (Patterson 1994). Note that near-bed turbidities are controlled largely by the height of the swell waves and are unrelated to the tide (modified from Larcombe & Woolfe 1999a).

reefs are discussed later in this chapter.) Thus, for the modern inner shelf, where terrigenous sediments are most abundant, there are probably two different types of reefs: (i) those inhabiting relatively turbid nearshore areas, within the influence of muddy inner shelf terrigenous sediments, where sedimentary processes act to prevent significant active sediment accumulation (e.g. Paluma Shoals, Cape Tribulation); and (ii) those occurring seawards of the zone of inner shelf terrigenous sediment, perhaps inhabiting offshore islands or recolonised reefs (e.g. Magnetic Island). Using basic concepts of sedimentation, a diagram can summarise the range of potential relationships between coral reefs and terrigenous sediments (Figure 24.5). In many places, inner shelf waters of the Great Barrier Reef are muddy and the sea bed is soft. In other places, the sea bed of the inner shelf is coarse grained, hard and relatively stable, and waters may be less turbid. A hard stable substrate is required for coral larvae to settle, establish and grow into a coral reef. Ideal substrates include pre-existing coral reefs and rocky outcrops. Successful settlement of larvae is not likely upon soft sediments which are actively

accumulating or are subject to episodes of resuspension and redeposition. Likewise, successful settlement is unlikely on sediments that are being actively eroded by currents. We can therefore infer that, coral reef 'initiation' must plot near the centre on Figure 24.5a. Assuming that the coral larvae have successfully settled and begun to grow on a suitable substrate, sediments may then continue to accumulate, be eroded or neither. If sediment continues to accumulate the rate may be high enough to swamp the young polyp. Continued and indefinite accumulation of terrigenous sediment results in the death of the polyp. If erosion occurs this may disturb the newly fastened coral polyp and either damage it or tip it over. Thus, in relatively simple terms, we can relate sedimentary processes to the likely survival and growth of a reef. The magnitude of any parameter that puts stress on corals is important, but must be assessed together with the duration for which the stress is imposed. The turbidity of water overlying a reef may be high, but if only imposed for a few hours may cause no detectable long-lasting damage. The thickness of a mantle of settled sediment on a coral

Holocene Great Barrier Reef

Figure 24.8 Location map of Cleveland Bay and Halifax Bay showing locations of coral reefs. Fringing coral reefs (marked in black) occur on most of the continental islands of the inner shelf, such as Magnetic, Great Palm, and Orpheus Islands. Coastal turbidzone reefs (such as Paluma Shoals) and unnamed coral accumulations also occur close to much of the Halifax Bay shoreline in turbid water of the shallow subtidal zone influenced by resuspension of the inner shelf sedimentary wedge. The whole shoreline and nearshore zone has not yet been surveyed in detail, and it is likely that more coral reefs and coral accumulations occur than shown here (after Larcombe & Woolfe 1999a).

289

ir4s*s

10*00*$

and duration of that cover need to be compared to the time for which a particular coral may be able to withstand being swamped by sediment. In some reefs, the period between repeated times of accumulation and resuspension may be a function of natural rainfall, tidal or wind regimes. Such variation may lead to severe damage or death of individual or entire reefs. The consequence for individual corals may be severe, but must also be seen in the context of variation of the natural environment. Geoscientists have an important role to play in providing a balanced view of natural coral reef environments over the long term. Two examples are presented below where the distinction between turbidity and sediment accumulation is essential for understanding the geological and environmental processes. Such an understanding contributes to improved assessment and management of the Great Barrier Reef.

CASE STUDY 1: MAGNETIC ISLAND CORALS AND IMPACTS OF DREDGING Cleveland Bay, Townsville, is a north-facing embayment in which occurs a voluminous muddy shore-attached sediment wedge (Carter etal 1993). Relatively sheltered from the dominant southeast trade winds and the associated waves, the bay is a zone of natural sediment accumulation. In January 1993, a major developmental dredging program commenced at the port designed to deepen the existing dredged shipping channel and extend it to the outer fringes of Cleveland Bay (Figure 24.6). Due to both the muddy nature of the sediments to be dredged and the ecology of many parts of Cleveland Bay (particularly the presence of fringing coral reefs of nearby Magnetic Island and extensive seagrass beds of the southern bay) environmental regulations required that a monitoring

program be undertaken (Benson et al 1994). Biologists and coral ecologists monitored the status of the coral and seagrass during dredging. The program included an intensive study of turbidity and sediment transport in the bay. Over a four-month period, up to 14 self-logging nephelometers were deployed on the muddy bed in the bay and at selected reef sites (Figure 24.6), together with three Interocean S4 current meters to record currents and a Waverider buoy to measure waves. The nephelometers are devices which incorporate an optical backscatter sensor and a small on-board computer to control data recording and data storage. The instruments detect reflective particles in the water column, which in Cleveland Bay were mostly silt-sized sediment grains. Samples of water were taken at various times next to the deployed instruments and their sediment concentration was measured by filtration. These data were used to calibrate the sensors, allowing conversion of the raw output from the sensors into suspended sediment concentrations (calibration errors may be ±30%). The results show a close correspondence between periods of high swell waves and raised turbidity, indicating that the major cause of sediment resuspension (and thus turbidity) from the muddy floor of Cleveland Bay was swell waves, produced by the regional southeast winds (Figure 24.7). By themselves, tidal currents were a very minor cause of turbidity. In contrast, at the shallow fringing reefs, locally induced wind-waves, such as those produced by the strong afternoon sea breeze, were the major cause of resuspension. The material resuspended at the reefs comprised mud derived from the main bay together with any locally derived fine carbonate sediment. The fringing reefs were probably flushed of suspended sediment at spring tides by the tide sweeping across the reef flats and expelling a relatively turbid plume of water into the bay. Thus, the natural mechanisms which generate turbidity

290

P. Larcombe

in the bay are different to those which generate turbidity at the reefs, and the two mechanisms are generally unrelated. Further, current data and modelling showed that the main current in the bay flows weakly into the bay from the northeast, and ultimately out of West Channel into adjacent Halifax Bay. This flow is associated with periods of medium or strong regional southeast winds, and probably forms part of a wind-driven coast-parallel current of regional scale. Importantly, there were no currents (measured or inferred) that might carry significant quantities of suspended sediment towards the reefs from the bay, either naturally, or from the dredge sites. Suspended sediment concentration data taken simultaneously at a similar reef in the next embayment northwards (Rattlesnake Island in Halifax Bay: Figure 24.8) where there was little fine sediment on the bed, were much lower than data at Magnetic Island, indicating that the reefs of Magnetic Island are indeed affected to a limited extent by the presence of Cleveland Bay. We can make some initial assessment of the likely impact of such a dredging exercise by comparing the volume of dredged sediment with that transported by natural processes. Historically, the average weight of material dredged each year has been 170 000 m 3 , i.e. the equivalent of -250 0001 (at a specific gravity of 1.5). Following the latest developmental dredging program, it is estimated that annual dredging will increase to -300 0001 (Sinclair Knight Merz 1996). We can estimate that the bay has an area of 200 km2 and an average depth of 5 m. Thus a depth-mean suspended sediment concentration of 50 mg/L results in a total mass of sediment in suspension of 50 000 t. The suspended sediment concentration of 50 mg/L is a low estimate, and real concentrations are probably twice as high during many swell-wave events. Thus, the total amount of material dredged annually is only five or six times the amount resuspended in the bay by natural resuspension processes many times per year. In this case the amount of dredged material deposited in the outer bay is not significant, at least when viewed simply in terms of total volumes. Whether the sum effect of 10 or 20 years of such activity could be viewed in the same way is another question. In conclusion, the monitoring exercise indicated that: (i) suspended sediment concentrations at the fringing reefs were quite variable (from 0 to 50 mg/L), both spatially and temporally; (ii) suspended sediment concentrations in the bay itself were perhaps half an order of magnitude higher than at the fringing reefs, where suspended sediment concentrations are controlled by different hydrodynamic processes; and (iii) fringing reefs occur in much more turbid environments than previously thought (indeed at an early point in the study it was suggested that the nephelometer data had been presented with the decimal point in the wrong place!). At all reef sites, natural variation in suspended sediment concentrations was greater than the variation caused by dredging. Indeed, no effects of dredging were detectable in the turbidity data at the fringing reefs. It is thus unlikely that similar dredging in the future will harm the fringing reefs of Magnetic Island, at least in terms of increased sedimentation. Some environmental impacts remain largely unknown, including the long-term impacts of the dumped sediment on the benthic fauna of Cleveland Bay (particularly at the dump site itself), and the rate and impacts of remobilisation of the sediment dump.

CASE STUDY 2: PALUMA SHOALS, A CORAL REEF AT THE BASE OF A TERRIGENOUS BEACH Some coral reefs occur in shallow turbid water on the inner shelf north of Townsville (Woolfe & Larcombe 1998; Larcombe & Woolfe 1999a; Larcombe et al. in press). The Paluma Shoals are a set of coral reefs - 2 km offshore from Two Mile Creek, in Halifax Bay and 25 km north of Townsville (Figure 24.8). The coastline here is gently curved, formed of a chenier/beach-ridge plain with generally sandy beaches. Muddy mangrove swamps occur in sheltered intertidal conditions behind the beach. Much of the coastline show signs of long-term erosion, such as fossil mangrove muds outcropping on the beachface, and granitic boulders and cobbles (probably representing Pleistocene alluvial fans) exposed on the beach and in places below the low water mark. Like most of the Great Barrier Reef inner shelf, Halifax Bay has accumulated a significant inner-shelf mud wedge over the Holocene. Wave action has resulted in little accumulation of mud at the shoreline itself, and most accumulation occurs beyond the low water mark out to a depth of - 1 5 m and to a maximum thickness of - 2 m. In places between the low tide mark and the main body of the sediment wedge are coral reefs, founded upon boulders and cobbles. One such reef, Paluma Shoals (Figure 24.9a on Plate 12), has nearly 150 species of coral (Veron 1995), with the main corals being faviids and mussids, small Acropora, and large stands of Galaxea. Similar inner shelf reef communities have been described by Done (1982). Sand is present in spaces between individual coral heads. In most places, these corals have no muddy sediments on them, probably because the near-constant wave action does not allow settling and accumulation of mud greater than the corals themselves can remove. This is well shown on many coral microatolls, where there is an accumulation of mud in places many centimetres thick in the dead portions of the microatoll, immediately next to areas with clean live coral (Figure 24.9b on Plate 12). Waves are the most important factor in resuspending muddy sediments around the reef. This is shown by current and wave measurements at Paluma Shoals, along with measurements of water turbidity (Figure 24.10). Regionally, water turbidity is generally highest at low water, when wave action has the greatest effect upon the bed. These waves are caused by the persistent southeast trade winds, which blow shore-parallel along Halifax Bay, suppress the tidal currents, and drive a slow current (mostly less than 10 cm/s) flowing northwards along the coast. Hence, as well as waves not allowing mud to settle on the reef, we can infer that mud in suspension is transported alongshore and away from the reefs, by the windinduced current.

SIGNIFICANCE FOR ENVIRONMENTAL MANAGEMENT OF THE GREAT BARRIER REEF By identifying and documenting the environmental and geological characteristics of these reefs and understanding the hydrodynamic controls of sedimentation, we can improve the assessment and management of the coral reefs

Holocene Great Barrier Reef

291

200 Turbidity (NTU)

100 A.,

Current meter depth

(m)

Period of increased variation in the depth record, caused by waves 15

Figure 24.10 Time series of turbidity (measured 0.3 m above bed), current meter depth, current speed and current direction (measured 1 m above bed) at Paluma Shoals, and wind direction measured at Townsville Airport for days 1-8. Note that turbidity increases dramatically following the commencement of southeast winds (day 3 onwards) due to sediment resuspension by waves. The southeast winds also suppress the normal tidal currents to produce a constant coast-parallel current, towards the northwest (modified from Larcombe & Woolfe 1999a).

Current speed (cm/s)

Wind-driven coast-parallel current (flow to N W ) , suppressing tidal currents

10 5

O Current directioni (degs)

100 o

Wind directionl (degs)

ifl

Flood

A VM

300

Ebb w

300

100

Period of persistent S E winds (speeds up to 18.5 knots)

N winds 10-15 knots

• 1— •

of the Great Barrier Reef. The two case studies above are important because they demonstrate that: (i) some coral reefs naturally occupy environments with high levels of turbidity; (ii) high turbidity does not necessarily result in sediment accumulation and by inference may not lead to coral damage; and (iii) the presence of nearby muddy sediment is not necessarily detrimental to corals. As an example of the direct application of this knowledge, it is unlikely that detailed monitoring of the coral reefs at Magnetic Island would be required for further developmental dredging (or for maintenance dredging) of the channel at the Port of Townsville. This saves a great deal of money and effort for the Port Authority and the various environmental management authorities. We can look for similar geomorphic, geological and hydrodynamic characteristics in other unstudied places on the shelf and better infer the sedimentary processes at work in those environments. We can also improve the design of future studies on the shelf that involve interactions between sediment transport and important shelf ecosystems. For example, if an outfall was planned to discharge into Halifax Bay near Paluma Shoals or another similar reef, we might first calculate whether the maximum likely discharge of fine sediments to the shelf was able to contribute significantly to the turbidity regime at the reefs. If a monitoring program was undertaken, instruments may be deployed to measure natural conditions and potential increases in turbidity and sedimentation caused by the outfall. Such instruments might be arranged in an array parallel to the coast and along the likely sediment transport path, rather than shore-normal. If the outfall was approved, discharges might be best timed to coincide with periods of southeast winds and waves.

• •• • • • •

••• • • •

• • • • • •

r 4

Days

Finally, as noted above, we can also place into fuller context the sedimentary impacts of the highly publicised 'big events', such as turbid freshwater plumes which result from rainfall associated with cyclones.

'OUTBREAKS' OF THE CROWN-OF-THORNS STARFISH: ARE THEY ANTHROPOGENIC? Acanthaster planci, the crown-of-thorns starfish, is an echinoderm which grazes on corals. Over the last 35 years, these starfish have destroyed corals at many reefs in the Indo-Pacific region. The life history of the crown-of-thorns starfish, and many other echinoderms, is one of 'boom and bust'. Episodes of high populations ('plagues') alternate with intervening periods with much smaller populations. Where observed by humans, these periods of high population are often called 'outbreaks'. The Great Barrier Reef has been the site of two major outbreaks in the last 35 years (Engelhardt & Lassig 1997), in 1962-1976 and 1979-1991 (Moran 1986), both of which apparently originated at Green Island off Cairns. High density populations of crown-of-thorns starfish have effects upon coral reefs at regional and local scales. Regionally, -17% of reefs were impacted at some level by crown-of-thorns starfish in the 1979-1991 outbreak (Moran et al 1992). The concentration occurred in the central portion of the Great Barrier Reef between Townsville and Lizard Island. Coral cover was patchy, and took about 12-15 years to return to pre-impact status. Regionally, recovery progressed southwards through the main region of impact (Done et al 1992). The effects at individual reefs were highly variable. Moderate to high coral mortality

292

P. Larcombe

0 5 10

Number of elements

Figure 24.11 Downcore distribution of skeletal elements of A. planci at sites on John Brewer and Green Island reefs (modified from Walbran et al 1989a). (30->50% of corals destroyed) occurred at -57% of reefs in the 1979-1991 outbreak (Moran et al 1992). At Green Island detailed monitoring showed that, following an outbreak in 1980, coral recovery had produced coral cover averaging -17% by 1990 with the recovering coral community dominated by arborescent and platy Acropora species (Fisk 1992). It is important to determine whether the crown-ofthorns starfish 'outbreaks' are caused by a human activity that degrades environmental conditions. It is widely acknowledged that crown-of-thorns starfish outbreaks may be caused by many factors, but the primary cause, and indeed focus for appropriate further research, is much argued. Historically, the bone of contention lies between two main groups. Many (although by no means all) marine biologists hold that a human cause of outbreaks is likely (with major implications for environmental management). The mechanisms by which human activity are supposed to have contributed to this phenomenon are broadly either (i) coastal development and river catchment modifications, increasing nutrient inputs into the Great Barrier Reef lagoon (Birkeland 1982), which may increase the development and or survival of crown-of-thorns starfish larvae or increase settlement success, or (ii) fishing or other human activity reducing numbers of crown-of-thorns starfish predators, leading to increasing numbers and reproductive success of crown-of-thorns starfish. Most geologists have tended to consider that the record of Acanthaster skeletal parts in the geological record indicates that outbreaks are intrinsically a natural phenomenon. The arguments were brought to the forefront a decade ago, following the publication of papers which presented downcore distributions of A. planci skeletal material in cores taken from three reefs between Cairns and Gladstone (Walbran et al. 1989a, b). The stratigraphic record reveals that large concentrations of crown-of-thorns starfish skeletal elements occur in places in the cores, presumably from mass mortality of starfish following individual outbreaks. Radiocarbon dating showed that A. planci

skeletal parts have occurred throughout the last 8000 years (Figure 24.11). In 1992, a special issue of the journal Coral Reefs (Wilkinson & Mclntyre 1992) addressed the crown-of-thorns starfish theme. Papers by Keesing et al (1992) and Pandolfi et al (1992) criticised the geological interpretations on a range of themes from the temporal and spatial nature of crown-of-thorns starfish, their mortality patterns and its signal in surface sediments, and the interpretation of the stratigraphic sequence shown in cores. Henderson and Walbran's (1992) paper answered each of these criticisms comprehensively and their conclusion was based upon the principle of simplicity: the geological explanation was the simplest required to fit the facts— alternative explanations for the geological observations were not appropriate because they require processes that have not been substantiated. In their conclusions, Henderson and Walbran (1992) stated that 'with respect to present knowledge of sedimentology, stratigraphy, carbon-14 chronology, taphonomy, ecology, population dynamics and population densities of contemporary A. planci in the Great Barrier Reef, there are no substantive observations that are inconsistent with the observed fossil record'. Since then, there has been no serious challenge published to the geological interpretation, and yet significant amounts of money remain committed to studies of crown-of-thorns starfish. One might enquire whether some money would be better spent on researching and informing the public visiting the Great Barrier Reef of the role which crown-of-thorns starfish and other predators on corals play in the natural variability of the reef system. Well, perhaps some positive news is worthy of attention. Over the last few decades, many artificial surfaces in the Great Barrier Reef, such as shipwrecks, or small jetties, have been colonised and support a healthy growth of corals. Perhaps we should see the headline 'Coral reef outbreak' more often? Finally, if in some way you see yourself as an Environmental Geoscientist, you should realise the value of your background and training to the community. At a recent meeting in Ireland, the conference was reminded by a scientist from the Netherlands that 'we only see what we understand'. There is currently little geoscience (or other physical science) input into many of our major environmental management agencies at both state or federal levels. Significant management or research groups with a geoscience focus are largely absent. This can lead to unfortunate outcomes. To use the crown-of-thorns starfish as an example, a recent published review of the crown-of-thorns starfish phenomenon concluded that 'To date, scientific studies into the most commonly suggested possible causes of the outbreaks have been unable to produce adequate evidence implicating human activities in this periodic (sic) phenomenon'. This conclusion is based upon work performed over around 20 years, and appears unarguable. However, the geological perspective of the crown-of-thorns starfish was given only a single paragraph, and did not include the important contribution of Henderson and Walbran (1992). Another example of where a geological background can assist in management has been described above. The issue of sediment input into the Great Barrier Reef lagoon, enhanced by human impacts in the last 50 years or so, has been a focus for environmental managers

293

Holocene Great Barrier Reef for o v e r a d e c a d e , regarding t h e p o t e n t i a l i m p a c t s u p o n various o r g a n i s m s or e c o s y s t e m s w i t h i n t h e l a g o o n . S e a g r a s s b e d s a n d c o r a l reefs h a v e b e e n p e r c e i v e d a s p o t e n t i a l l y b e c o m i n g s w a m p e d b y m u d , or killed slowly b y i n c r e a s e d turbidity in G r e a t B a r r i e r R e e f w a t e r s . H o w e v e r , t h e r e is n o e v i d e n c e of a r e g i o n a l i n c r e a s e in t h e rate of s e d i m e n t a c c u m u l a t i o n or i n c r e a s e in turbidity a l o n g t h e G r e a t Barrier R e e f shelf, a n d a r g u m e n t s b a s e d o n s h e l f h y d r o d y n a m i c s and sedimentary processes have also concluded that no threat e x i s t s from s e d i m e n t a t i o n a l o n e to m o s t c o r a l reefs in the Great Barrier Reef. G e o s c i e n c e r e s e a r c h c a n h a v e g r e a t v a l u e in b e i n g p a r t of t h e p r o c e s s b y w h i c h t h r e a t s to t h e G r e a t B a r r i e r R e e f are

assessed

and

management

plans

developed.

E n v i r o n m e n t a l G e o s c i e n t i s t s , get o u t t h e r e a n d s p r e a d t h e word! Geological knowledge a n d principles are f u n d a m e n tal to b e t t e r e v a l u a t i o n a n d m a n a g e m e n t of c o a s t l i n e s a n d shallow shelf seas, including the Great Barrier Reef.

isostacy and the sea-level isobase of 5 5 0 0 B.P. in north Queensland, Australia. Marine Geology 49, 81-90. COLLINS L . B . , Z H U Z . R , W Y R O L L K - H . , HATCHER B . G . , PLAYFORD P. E . , EISENHAUSER A . , C H E N J . H . , W A S S E R B U R G G . J . & BONANI G .

1993.

Holocene growth history of a reef complex on a cool-water carbonate margin: Easter Group of the Houtman Abrolhos, Eastern Indian Ocean. Marine Geology 115, 29-46. DAVIES P. J. & HOPLEY D. 1983. Growth fabrics and growth rates of Holocene reefs in the Great Barrier Reef. BMR Journal of Australian Geology & Geophysics 8, 237-151. DAVIES P. J. & MCKENZIE J. A. 1 9 9 3 . Controls on P l i o c e n e - P l e i s t o c e n e

evolution of the northeastern Australian continental margin. In: McKenzie J. A., Davies P. J., Palmer-Julson A. et al eds. Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 133, pp. 755-762. College Station, TX. DONE T. J. 1982. Patterns in the distribution of corals communities across the central Great Barrier Reef. Coral Reefs 1, 95-107. DONE T. J.,

DEVANTIER L . M . ,

FISK D .

VAN W O E S I K

1992.

Regional and local patterns of coral recovery on reefs affected by crown-of-thorns starfish: central Great Barrier Reef. Final Report to the Great Barrier Reef Marine Park Authority, Townsville.

ENGELHARDT U. & LASSIG B. R. 1997. A review of the possible causes

ACKNOWLEDGMENTS B o b C a r t e r is t h a n k e d for c o m m e n t s o n a n early draft of this contribution, a n d reviewers Nick Harvey a n d B o b B o u r m a n are t h a n k e d for their u s e f u l c o m m e n t s . T h e Townsville Port

and consequence of outbreaks of the crown-of-thorns starfish (Acanthaster planci) on the Great Barrier Reef—an Australian perspective. In: Proceedings of the Great Barrier Reef Conference, Townsville, November 1996, pp. 243-259. FISK D. A. 1992. Recruitment of Acanthaster planci over a five-year period at Green Island reef. Australian Journal of Marine and Freshwater Research 4 3 , 629-633. FURNAS M . J., MITCHELL A. W . & SKUZA M . 1 9 9 5 . Nitrogen a n d p h o s -

Authority g a v e p e r m i s s i o n to p u b l i s h i n f o r m a t i o n from their

phorous budgets for the central Great Barrier Reef shelf. Great Barrier Reef Marine Park Authority Research Report 36.

environmental monitoring program. Colour plates courtesy

GAGAN M . K . , ANKER S . , AYLIFFE L . K . , BARNES D . J . , CHAPPELL J . M . A . , HOPLEY D . , ISDALE P. J . , LOUGH J . M . & MCCULLOCH M . T .

of A n d r e w C o s t e n .

November

REFERENCES remote sensing in determining the extent and causes of eutrophication in the Great Barrier Reef lagoon. In: Proceedings of the 4th Pacific Conference on Marine Science and Technology, pp. 2 5 - 3 2 . BELL R R. F. & GABRIC A. J. 1991. Must GBR pollution become chronic before management reacts. Search 22, 117-119. BENSON L . J . , GOLDSWORTHY R

M.,

BUTLER I . R .

OLIVER J .

1994.

Townsville Port Authority Capital Dredging Works 1993: Environmental Monitoring Program. Townsville Port Authority. BIRKELAND C. 1982. Terrestrial runoff as a cause of outbreaks of Acanthaster planci. Marine Biology 69, 175-185. BROWN B. E. & HOWARD L. S. 1985. Assessing the effects of 'stress' on

reef corals. Advances R.

W.,

in Marine Biology 22, 1-63.

MARAGOS

KNUTSON

1974.

Radiographic studies of reef coral skeletons: rates and patterns of coral growth. Journal of Experimental Marine Biology and Ecology 14,

179-200.

CARTER R. M . , CARTER L. & JOHNSON D . P. 1 9 8 6 . S u b m e r g e n t s h o r e l i n e s

in the S W Pacific: evidence for an episodic post glacial transgression. Sedimentology 3 3 , 6 2 9 - 6 4 9 .

CARTER R. M. & JOHNSON D. P. 1 9 8 6 . Sea level controls on the post

glacial development of the Great Barrier Reef, Queensland. Marine Geology 71, 137-164.

CARTER R. M . JOHNSON D . P. & HOOPER K. G . 1 9 9 3 . E p i s o d i c post-

glacial sea-level rise and the sedimentary evolution of a tropical embayment (Cleveland Bay, Great Barrier Reef shelf, Australia). Australian Journal of Earth Sciences 40, 229-255. CHAPPELL J. 1994. Upper Quaternary sea levels, coral terraces, oxygen isotopes and deep-sea temperatures. Journal of Geography 103, 828-840. CHAPPELL J . , CHIVAS A . , WALLENSKY E . , POLACH H . A .

& AHARON P.

1983. Holocene palaeoenvironmental changes, central to north Great Barrier Reef inner zone. BMR Journal of Australian Geology & Geophysics 8, 2 2 3 - 2 3 5 . . CHAPPELL J & POLACH H. 1991. Post-glacial sea-level rise from a coral record at Huon Peninsula, Papua New Guinea. Nature 3 4 9 , 147. CHAPPELL J

1996,

pp. 1 9 5 - 2 0 9 .

GAGAN M . K., JOHNSON D. P. & CROWLEY G . M . 1 9 9 4 . Sea-level control

BELL P. R. F. & GABRIC A. J. 1990. The use of field survey and satellite

BUDDEMEIER

1997.

Massive corals: grand archives of Australian palaeoclimate. In: Proceedings of the Great Barrier Reef Conference, Townsville,

R H O D E S E . G . , T H O M B . G . & WALLENSKY E . 1 9 8 2 .

Hydro-

of stacked late Quaternary coastal sequences, central Great Barrier Reef. Sedimentology 41, 329-351. HARRIS P. T. 1999. Sequence architecture during the Holocene transgression: an example from the Great Barrier Reef shelf, Australia—Comment. Sedimentary Geology 125, 2 3 5 - 2 3 9 HARRIS P. T. & DAVIES P. J. 1989. Submerged reefs and terraces on the

shelf edge of the Great Barrier Reef, Australia: morphology, occurrence and implications for reef evolution. Coral Reefs 8, 97-98.

HARVEY N. & SEARLE D. E. 1983. Seismic investigations of Late

Quaternary reefal and inter-reefal sediments of the Great Barrier Reef province. In: Barnes D. J. ed. Perspectives on Coral Reefs, pp. 154-163. Australian Institute of Marine Science, Brian Clouston, Manuka.

HENDERSON R. A. & WALBRAN P. D. 1 9 9 2 . Interpretation of the fossil

record of Acanthaster planci from the Great Barrier Reef: a reply to criticism. Coral Reefs 11, 95-101.

HODELL D . A . , CURTIS J . H . , JONES G . A , H I G U E R A - G U N D Y A . , BRENNER M.,

BINFORD

DORSEY

1991.

Reconstruction

Caribbean climate change over the past 10 500 years. Nature 352, 790-793. HOPLEY D. 1982. The Geomorphology of the Great Barrier Reef: Quaternary Development of Coral Reefs. Wiley-Interscience, New York. HOPLEY D. 1995. Continental shelf reef systems. In: Carter R. W. G. & Woodroffe C. D. eds. Coastal Evolution: Late Quaternary Shoreline Morphodynamics, pp. 303-340. Cambridge University Press, Cambridge. HOPLEY D. & KINSEY D. W. 1988. The effects of a rapid short-term sea-

level rise on the Great Barrier Reef. In: Pearman G. I. ed. Greenhouse: Planning for Climate Change, pp. 189-201. CSIRO, Melbourne

JOHNSON D. P. & CARTER R. M.. 1987. Sedimentary framework of main-

land fringing reef development, Cape Tribulation area. Great Barrier Reef Marine Park Authority, Technical Memorandum 14. E. 1 9 8 4 . Holocene reef sequences and geochemistry, Britomart Reef, central Great Barrier Reef, Australia. Sedimentology 31, 515-529. JOHNSON D . P. & RISK M. J. 1 9 8 7 . Fringing reef growth on a terrigenous JOHNSON D . P., C U F F C . & RHODES

294

P. Larcombe

mud foundation, Fantome Island, central Great Barrier Reef, Australia. Sedimentology 34, 275-287. KEESING J . K . , BRADBURY R . H . , DEVANTIER L. M . , RIDDLE M . J . & DE'ATH

G. 1992. The geological evidence for recurring outbreaks of the crown-of-thorns starfish: a reassessment from an ecological perspective. Coral Reefs 11, 79-85. KLEYPAS J. A. & HOPLEY D. 1992. Reef development across a broad continental shelf, southern Great Barrier Reef, Australia. In: Proceedings of the Seventh International Coral Reef Symposium, Guam, pp. 1129-1141. LARCOMBE P. & CARTER R. M . 1 9 9 8 . Sequence architecture during the Holocene transgression: an example from the Great Barrier Reef shelf, Australia. Sedimentary Geology 117, 9 7 - 1 2 1 . LARCOMBE P. & CARTER R. M. 1999. Sequence architecture during the Holocene transgression: an example from the Great Barrier Reef shelf, Australia—Reply. Sedimentary Geology 125, 241-241. LARCOMBE P., CARTER R . M . , D Y E J . , GAGAN M . K . & JOHNSON D .

1995b. New evidence for episodic post glacial sea level rise, central Great Barrier Reef, Australia. Marine Geology 127, 1-44. LARCOMBE P., COSTEN A. & W O O L F E K . The hydrodynamic and sedimentary setting of nearshore coral reefs, central Great Barrier Reef, Australia: Paluma Shoals, a case study. Sedimentology (in press). LARCOMBE P., RIDD P. V . , PRYTZ A. & W I L S O N B . 1995a. Factors controlling suspended sediment on inner-shelf coral reefs, Townsville, Australia. Coral Reefs 14, 163-171. LARCOMBE P. & W O O L F E K . J . 1999a. Terrigenous sediments as influences upon Holocene nearshore coral reefs, central Great Barrier Reef, Australia. Australian Journal of Earth Sciences 46, 141-154. LARCOMBE P. & W O O L F E K. J. 1999b. Increased sediment supply to the Great Barrier Reef will not increase sediment accumulation at most coral reefs. Coral Reefs 18, 163-169. LINSLEY B. K. & THUNNEL R. C. 1990. The record of deglaciation in the Sulu Sea: evidence for the Younger Dryas event in the tropical Western Pacific. Palaeoceanography 5, 1025-1039. MAPSTONE B . D . , CHOAT J . H . , CUMMING R . L. & OXLEY W . G . 1 9 8 9 . T h e

Fringing Reefs of Magnetic Island: Benthic Biota and Sedimentation—a baseline study. Great Barrier Reef Marine Park Authority Research Publication 13. MORAN P. J. 1986. The Acanthaster phenomenon. Oceanography and Marine Biology 24, 379-480. MORAN P. J . , DE'ATH G . , BAKER V. J . , BASS D . K . , CHRISTIE C . A . , MILLER

I . R., MILLER-SMITH B. A . & THOMPSON A . A . 1992. Patterns of outbreaks of crown-of-thorns starfish (Acanthaster planci L.) along the Great Barrier Reef since 1966. Australian Journal of Marine and Freshwater Research 43, 555-568. NAKIBOGLU S . M . , LAMBECK K . & AHARON P. 1 9 8 3 . Postglacial sea-levels in the Pacific: implications with respect to deglaciation regime and local tectonics. Tectonophysics 91, 3 3 5 - 3 5 8 .

NELSON C . S . , COOKE P. J . , HENDY C . H . & CUTHBERTSON A . M .

1993.

Oceanographic and climatic changes over the past 160,00 years at Deep Sea Drilling Project Site 594 off southeastern New Zealand, southwest Pacific Ocean. Paleoceanography 8, 4 3 5 - 4 5 8 . NEUMANN C . & MACINTYRE I . G . 1 9 8 5 . Reef response to rising sea level: keep up, catch up or give up. In: Proceedings of the 5th International Coral Reef Symposium, Tahiti, Vol. 3 , pp. 1 0 5 - 1 1 0 . PANDOLFI J . M . , WILKINSON C . R . & MACINTYRE I . G . 1992. A palaeobiological examination of the geological evidence for recurring outbreaks of the crown-of-thorns starfish, Acanthaster planci (L.). Coral Reefs 11, 87-93. PARTAIN B. R. & HOPLEY D . 1 9 8 9 . Morphology and development of the Cape Tribulation fringing reefs, Great Barrier Reef, Australia. Great Barrier Reef Marine Park Authority Technical Memorandum 21.

PASTORAK R. A. & BILYARD G. R. 1985. Effects of sewage pollution on

coral reef communities. Marine Ecology Progress Series 21, 175-189. PATTERSON D . 1 9 9 4 . Oceanographic data collection. In: Benson L . J . , Goldsworthy P. M., Butler I. R. & Oliver J. eds. Townsville Port Authority Capital Dredging Works 1993: Environmental Monitoring Program, pp. 1 2 5 - 1 4 7 . Townsville Port Authority, Townsville.

F. & DAVIES P. J. 1993. Sedimentological response of an outer-shelf, upper-slope sequence to rapid changes in Pleistocene eustatic sea level: Hole 820A, northeastern Australian margin. In: McKenzie J. A., Davies P. J., Palmer-Julson A. et al. eds. Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 133, pp. 303-313. College Station, TX. QUEENSLAND DEPARTMENT OF PRIMARY INDUSTRIES 1993. The condition of river catchments in Queensland. Queensland Department of Primary Industries, Brisbane. ROGERS C. S. 1983. Sublethal and lethal effects of sediments applied to common Caribbean reef corals in the field. Marine Pollution Bulletin 14, 378-382. ROGERS C. S. 1990. Responses of coral reefs and reef organisms to sedimentation. Marine Ecology Progress Series 62, 185-202. SHACKLETON N . J . & OPDYKE N . D . 1 9 7 3 . Oxygen isotope and palaeomagnetic stratigraphy of equatorial Pacific core V28 238: oxygen isotope temperatures and ice volumes in a 105 and 106 time scale. Quaternary Research 3 , 3 9 - 5 5 . SINCLAIR KNIGHT M E R Z 1 9 9 6 . Long term dredge spoil disposal strategy for the Port of Townsville. Report to the Townsville Port Authority (unpubl.). SMITH S . V. 1983. Coral reef calcification. In: Barnes D. J. ed. Perspectives on Coral Reefs, pp. 240-247. Australian Institute of Marine Science, Brian Clouston, Manuka. SMITH S . V . & KINSEY D . W . 1 9 7 6 . Calcium carbonate production, coral reefs growth and sea-level change. Science 194, 9 3 7 - 9 3 9 . STAFFORD-SMITH M. 1992. Mortality of the hard coral Leptoria phrygia under persistent sediment influx. In: Proceedings of the 7th International Coral Reef Symposium, Guam, Vol. 1, pp. 289-299. STAFFORD-SMITH M. G. & ORMOND R. F. G. 1992. Sediment-rejection mechanisms of 42 species of Australian scleractinian corals. Australian Journal of Marine and Freshwater Research 43, 683-705. TAYLOR J . 1 9 9 5 . Sediment input to the Great Barrier Reef lagoon via river discharge: the Barron River. In: Larcombe P. & Woolfe K. eds. Great Barrier Reef: Terrigenous Sediment Flux and Human Impacts, pp. 9 2 - 9 4 . CRC Reef Research Centre, Research Symposium Proceedings, Townsville. THOM B. G. & ROY P. S. 1983. Sea level change in New South Wales over the past 15,000 years. In: Hopley D. & Thom B. G. eds. Australian Sea Levels in the Last 15000 Years: a Review, pp. 64-84. James Cook University Geography Department, Monograph Series, Occasional Paper 3. VERON J. E . N . 1 9 9 5 . Corals in Space and Time: the Biogeography and Evolution of the Scleractinia. University of New South Wales Press, Sydney. WALBRAN P. D. 1994. The nature of the pre-Holocene surface, John Brewer Reef, with implications for the interpretation of Holocene reef development. Marine Geology 122, 63-79. PEERDEMAN

WALBRAN P. D . , HENDERSON R . A . , FAITHFUL J . W . , POLACH H . A . , SPARKS

R. J., WALLACE G. & LOWE D. C. 1989a. Crown-of-thorns starfish outbreaks on the Great Barrier Reef: a geological perspective based on the geological record. Coral Reefs 8, 67-78. WALBRAN P. D., HENDERSON R. A., JULL A. J . & HEAD J . M. 1989b. Evidence from sediments of long-term Acanthaster planci predation on corals of the Great Barrier Reef. Science 245, 847-850. W A S S O N R. J. 1997. Run-off from the land to the rivers and the sea. In: Proceedings of the Great Barrier Reef Conference, Townsville, November 1996, pp. 23-41. WILKINSON C. R. & MACINTYRE I . G. (Editors) 1992. The Acanthaster debate. Coral Reefs 11. WOLANSKI E . 1 9 9 4 . Physical Oceanographic Processes of the Great Barrier Reef. CRC Press, Boca Raton. W O O L F E K . J. & LARCOMBE P. 1 9 9 8 . Terrigenous sediment accumulation as a regional control upon the distribution of reef carbonates. In: Camoin G. F. & Davies P. J. eds. Reefs and Carbonate Platforms in the Pacific and Indian Oceans, pp. 2 9 5 - 3 1 0 . International Association of Sedimentologists Special Publication 25. W O O L F E K. J. & LARCOMBE P. 1999. Terrigenous sedimentation and coral reef growth: a conceptual framework. Marine Geology 155, 331-345. Received 4 July 1997; accepted 19 September 2000

Geological Society of Australia Special Publication 21, 295-302

CHAPTER 25—Sediment dispersal along the inner shelf of the central Great Barrier Reef H j j j H K. J. WOOLFE* AND P. LARCOMBEt mmm^^^^^mmm^mmmm

<• i11 M H i

-1 - H 1 . H U M HKlgBI'

H K

;

Marine Geophysical Laboratory, School of Earth Sciences., J a m e s Coo/c University, Townsville, Qld 4811 Australia The Great Barrier Reef is the worlds largest and most complex system of coral reefs, located on a continental shelf into which the Holocene input of sediment has been, in global terms, relatively low. Since European settlement it is possible that there has been an increase in river sediment supplied to the Great Barrier Reef coastline, and understanding the likely fate and impacts of this sediment is important to environmental managers of the Great Barrier Reef region. Further, the Great Barrier Reef shelf represents a continental shelf where carbonate and land-derived sediments are mixed, and also a shelf which is impacted by cyclones. This contribution briefly reviews the nature and processes of sediment movement pertinent to the Great Barrier Reef coastline. KEY WORDS: cyclones, Great Barrier Reef, Holocene, sediment dispersal.

INTRODUCTION The Great Barrier Reef shelf (Figure 25.1) is a mixed siliciclastic-carbonate province (Belperio 1983) famous for its colourful coral reefs, clear waters and abundant marine life. The shelf is typically 80-120 km wide in the central section of the Great Barrier Reef, and has an average depth of -40 m. Recently it has become widely accepted (Zann & Sutton 1995) that the amount of sediment reaching the Great Barrier Reef coastline has increased over the last 200 years as a direct result of agricultural and urban development associated with settlement of Europeans. This post-European increase in sediment supply down rivers has the status of 'common knowledge' and is logical in many respects, particularly when considering catchment modifications such as those associated with grazing, deforestation and tilling of grassland and savanna. We should note, however, that at present the increase is heavily inferred rather than proven fact, and the magnitude of any increase is certainly unquantified (Larcombe et al 1996; Larcombe & Woolfe 1999). The fate of river sediments once in the marine environment is of concern to environmental managers, for reasons including: (i) suspended sediment increases turbidity, reducing water clarity and thus tourist enjoyment of the marine life; (ii) increased turbidity may reduce light levels on the seabed and hence inhibit coral growth; (iii) enhanced sedimentation may occur over time periods of hours to years, resulting in the physical burial and perhaps the subsequent death of corals or other benthic organisms; and (iv) agricultural and industrial contaminants may be introduced to the marine environment and transported via sediment particles. These concerns have heightened awareness amongst the scientific community of the importance of river sediment accumulation in coastal environments and considerable effort is now being directed towards generating an understanding of the rates of sediment supply, the mechanisms of sediment dispersal, the temporary sediment stores (on a

time-scale of centuries to a few thousand years) and the ultimate locations of sediment repositories on the Great Barrier Reef shelf (Larcombe et al 1996; Wasson 1997; Orpin et al 1999; Larcombe & Woolfe 1999). A good qualitative understanding of the physical processes operating on the inner shelf of the Great Barrier Reef has developed, and in this chapter we present a working model for the dispersal of sediment along the central Great Barrier Reef coastline.

PROCESSES CONTROLLING SEDIMENT SUPPLY TO THE GREAT BARRIER REEF COAST The Queensland Department of Primary Industries maintains, or has previously maintained, flow-gauging stations on many of the catchments bordering the Great Barrier Reef lagoon (Moss et al 1993; Horn 1995). Based on all available gauging records, catchment areas, land-use and rainfall patterns, estimates of the rate of supply of terrigenous sediment to the coastal zone have been prepared. While estimates of total sediment load vary greatly, from 13 Mt y^*1 (Moss et al 1993) to 28 Mt yl (Belperio 1983), it is clear that sediment supply is largely controlled by short-duration, high-volume flows associated with individual meteorological events (e.g. tropical cyclones, rain depressions, convergences). It is widely asserted that the terrigenous sediment flux to the Great Barrier Reef has increased as a result of human impacts on catchments, particularly in the last 200 years (Belperio 1983; Belperio & Searle 1988; Moss et al 1993; Neil & Yu 1995). However, most of the rivers entering the Great Barrier Reef lagoon are 'flashy' (i.e. have extremely large, high-flow/low-flow ratios) and at high-flow stages, the river is able to transport cobbles and larger clasts, which in * Ken Woolfe died tragically on 1 December 1999. His passion and enthusiasm for earth science will be missed, f Corresponding author: piers.larcombe@jcu.edu.au

296

K. J. Woolfe and P. Larcombe 146*30*

wfw

147® 30

Figure 25.1

Map of the central

Great Barrier Reef near Townsville showing the three-fold sedimentary division of the shelf and the principle locations discussed in the text.

Table 2 5 . 1 Comparison of long-term accumulation rates of sedimentary environments between the mouth of the Burdekin River and Halifax Bay (arranged in order of distance from source) (from Woolfe & Larcombe 1998). Total sediment accumulation (mm V'1)

Terrigenous fraction of sediment

(%)

Terrigenous sediment flux to the bed (BTF)

0.7 0.45 0.5-8.0 <0.2 <0.25 <0.1 0.03-0.2

70 85 95 85 85 75 10-30

20 15.8 19-300 <6.8 <8.5 <3 0 . 1 2 - 2.4

Upstart Bay 2500 y mean Upstart Bay 6500 y mean Upstart Bay & Bowling Green Bay, intertidal sediments Cleveland Bay 30 y mean Cleveland Bay 6000 y mean Halifax Bay 7000 y mean Mid-shelf off Townsville 6000 y mean

BTF is the volume of sediment supplied by all the world's rivers distributed over the areas of world's oceans.

practice makes in situ measurements of sediment transport difficult. Further, the rivers overtop their banks, inundating the floodplain for many kilometres on either side of the main channel, making the gauging of flows and sediment fluxes even more problematic. In addition, the hydrology and vegetation dynamics of many catchments are not well-known. Quantification of the sediment increase associated with land-use change is thus extremely difficult (Belperio 1983; Horn 1995; Mitchell et al 1995). Current estimates of the rates of terrigenous sediment supply to the central Great Barrier Reef shelf derive from: (i) estimation of the volume of Holocene sediment on the shelf, to give a long-term average of sediment accumulation rate (Belperio 1983; Harris et al 1990; Carter et al 1993; Woolfe & Larcombe 1998); (ii)

extrapolation of measurements of sediment transport in rivers and estuaries to annual or longer time-scales (Belperio 1978, 1983; Belperio & Searle 1988); or (iii) modelling sediment yields from soil erosion and runoff in the river catchments (Moss et al 1993; Neil & Yu 1996). Most terrigenous Holocene deposits are confined to the inner shelf and are well-documented, so that back-calculating long-term sediment input from them is fairly reliable although this long-term average (few thousand years) provides background data only for studies of recent sediment supply and accumulation (Table 25.1). Due to the nature of many rivers of the central Great Barrier Reef coastline, whose characteristics include relatively low mean-annual runoff, great inter-annual variability of discharge, hetero-

Sediment dispersal G r e a t Barrier Reef

In deep and moderately d e e p water-wind driven currents decrease with depth. Wind-driven current

4 Current direction (degs)

Flood

I*!

100

Wind direction (degs)

300 N winds 10-15 knots

|tt,t

•• h

Period of persistent S E winds (speeds up to 18.5 knots)

,t Days

Figure 25.2 Current meter data from an inshore mooring in Halifax Bay together with wind data from Townsville. Note how the regular ebb and flood tidal signature is disrupted by the southeast trade winds for most of the record. geneous vegetation patterns in their catchments (including altered catchments), and unusual channel morphologies, the latter two methods [(ii) and (iii) above] can result in considerable uncertainty in the calculated rate of sediment supply. Using their catchment model, based on catchment-scale data, Neil and Yu (1996) indicated that since European settlement, sediment fluxes to the coast from individual catchments have increased between 1 and 30 times (depending on land-use changes), and the total increase in sediment flux to the Great Barrier Reef coastline since European settlement is - 2 - 4 times. Methods of addressing sediment delivery to the coast are being refined progressively, and although its magnitude is undocumented, the occurrence of an increase is unchallenged. Thousands of dams have been constructed in the Great Barrier Reef hinterland. These range in scale from small (<1 ML) farm dams to large (>10 6 ML) domestic and agricultural supply dams. As a consequence it is likely that over a few decades to a century or so, there will be a decreased supply of coarse-grained sediment to the coast, due to the stabilisation of bars in rivers below the dams and because some sediment will be trapped in the dams. Long-term fluxes of fine-grained suspended sediment may be relatively less affected, because the bulk of this material is supplied to the coast during large-scale runoff events, where dams will be overtopped, and because altered catchments may yield increased amounts of fine-grained sediment. The detailed effects of dam construction on coastal sedimentation and inner shelf coral reefs remain unknown.

MECHANISMS OF SEDIMENT DISPERSAL In terms of the dispersal of sediment along the coast and inner shelf, two types of sediment need to be distinguished. (1) Suspended Sediment. Some fine-grained sediment is transported by being suspended in the water, maintained above the bed by turbulence. The weight of suspended material in the water column will decrease with lower flow speeds and smaller waves, but will rarely approach zero. (2) Bedload Sediment. Coarser material is moved along or closely associated with the bed. Sediment movement

297

requires a threshold shear stress to be exceeded, so that at low flow speeds, no sediment may be moved at all. For most practical purposes, suspended load can be considered to consist of 'muddy' material (fine silt and clay) which travels approximately at the speed of the water, whereas bedload sediment consists largely of coarse silt, sand and gravel, which travels significantly slower than the water column above. Suspended load is more significant in the transport of contaminants because finer particles have larger specific surface areas (surface area per unit mass) and hence a greater ability to transport surface-attached chemical species. Sedimentary material found in suspension in the Great Barrier Reef lagoon will have been derived directly from the rivers and/or from existing (reworked and resuspended) shelf sediments. Some suspended sediment is supplied in muddy plumes of rivers that discharge into the Great Barrier Reef lagoon. Because they are freshwater, these plumes tend to be buoyant and are thus generally confined to the upper few metres of the water column. As such, they have relatively low suspended sediment concentrations (<100 mg/L) and overlie marine waters which, during calm conditions, will have still lower suspended sediment concentrations. Another 'source' of suspended sediment is the sea bed itself, where fine sediments may be resuspended upwards from the bed into the water column. Waves, often concurrent with tidal or wind-driven currents, are generally the dominant mechanism in causing resuspension. Muddy waters produced by this mechanism will tend to decrease in suspended sediment concentration upwards away from the bed, and may reach suspended sediment concentrations much greater than those found in river plumes (up to 300 mg/L or more) (Larcombe et al 1995a). Regardless of its mechanism of generation, the rate at which suspended sediment will be moved is the product of flow depth, with depth-averaged suspended sediment concentration and speed. Generally, the greater thickness and sediment concentration of resuspension plumes result in a higher rate of sediment transport than surface flood plumes. Furthermore, resuspension of sediment from the bed is a process which occurs throughout the year ( - 1 0 3 h) over large areas of the inner shelf ( - 1 0 4 km 2 ) and to depths of - 1 0 m, whereas flood plumes are generally small (>10 3 km 2 ), exist for only brief periods ( - 1 0 2 h) and are relatively thin ( - 1 m). Thus sediment resuspension is likely to be about three orders of magnitude more important than river plumes in terms of coastal sediment transport. Bedload transport only occurs once threshold velocities have been exceeded, and even then, the speed of the moving sediment grains is significantly slower than water movement. Transport will occur near the coast in rivers and estuaries under the influence of freshwater flood and tidal currents. It will also occur on the inner shelf as a result of waves alone, or waves in combination with other currents. There are a number of mechanisms that might control the distribution and dispersal of river sediment which reaches the coast. Below, we describe some of the principle mechanisms that influence sediment dispersal in the coastal zone and comment on their potential significance on the Great Barrier Reef shelf.

298

K. J. Woolfe and P. Larcombe

Approaching wave train

Deep-water, non-breaking waves produce no net translation of- the water mass. However, water is driven forwards by shallow-water and/or breaking waves.

Wind- and 1 wave-driven current with turbid coastal boundary layer

B Approaching wave train

V^nd- and \ wave-driven current

The Rossby Radius varies inversely with latitude. Low latitudes 4 J , . , , High latitudes

Sediment-rich density plume

Equatorial

Density-driven current moving downsJope (

Wind- and wave-driven boundary current Tidal currents introduceturbfd water into high tidal flats and mangrove swamps,where sediment accumulation occurs.

Strong onshore winds

Wind-driven surface current

low atmospheric pressure Storm surge

Eddy currents transport turbid water into sheltered bays where sediment trapping occurs

Po-sf-cyd'one tdaxatiart How

Wind- and wave-generatec^ onshore current

High-energy waves resu spend mud, send e^d sheBy grovel and transport it shoreward

Coarse^ained material is left deposited as wave junup loses energy, whie Cyclone-generated chenier ridge finer rnatencd is returned to the oc ean

Sand/, gaveltyfbiogenic} mud substrate.

Figure 2 5 . 3 Principle processes responsible for the resuspension, transport and deposition of sediment in the coastal zone. It should be noted that some of the illustrated processes (i.e. b, e, f, and to a lesser extent d) are not considered to be important in the Great Barrier Reef lagoon.

Sediment dispersal Great Barrier Reef

Oceanographic currents Major oceanic currents are largely prevented from entering the Great Barrier Reef lagoon (see Larcombe 2001 figure 24.1 for a map) by the presence of the barrier reef, but the Northeast Australian Current can generate a significant net flow in the lagoon during certain times of year and under certain wind conditions (Wolanski 1994). However, these currents do not directly impinge on the coast, and they are not considered important in the distribution of sediment. Wind-driven currents Wind-driven currents result from friction between the atmosphere (wind) and the water surface, and thus tend to be surface currents. Even in the relatively shallow waters of the Great Barrier Reef shelf, these currents do not necessarily propagate to the bed. Friction between the wind and the water is proportional to the roughness of the water surface, so that for any given wind speed, the magnitude of the wind-driven current is increased by the presence of surface waves. The Great Barrier Reef lagoon lies in the southeast trade wind belt, and for approximately 9 months of the year (March-November) persistent southeast winds of 5-8 m/s (10-15 knots) occur. This results in a strong northwardflowing wind-driven surface current (Figures 25.2, 25.3) which has a significant impact on the dispersal of suspended sediment. Wind-driven currents rarely exceed the thresholds for bedload transport, but in conjunction with waves and tidal currents may locally play a role in the redistribution of bedload sediment. During the summer months (December-March) winds are generally lighter and may even blow offshore. Exceptions to the light winds of summer months in tropical regions (south of ~8°S) are cyclones, which are discussed later. Wave-driven currents Particles in open water and experiencing non-breaking waves move in an orbital pattern with no net translation (Figure 25.3a). However, as these waves shoal, begin to steepen and ultimately break, a net translation of water occurs in the direction of wave propagation. Consequently, waves transport water towards the coastline, where they raise water levels and also generate a return flow away from the coast. When waves break at an angle to the coastline (Figure 25.3c, d) they generate a coast-parallel boundary current (see below). Breaking waves are very effective resuspension agents. Observations show that wave-driven currents of this type may be strong enough to cause bedload transport and may also overwhelm tidal currents along the Great Barrier Reef coastline. The normal waves parallel the wind, so wind- and wave-driven currents generally tend to reinforce each other in the coastal zone. Rip currents Rip currents occur along coasts where the orientation of wave fronts approaching the coastline is nearly coast-parallel (Figure 25.3b). Under these conditions, excess water is unable to escape laterally along the coast and strong offshore-directed rip cells are generated. On an open coast

299

these may be strong enough to transport coarse-grained bedload (gravel) seawards to beyond the surf zone. Consequently, on open coastlines they may represent a significant process of offshore-directed sediment transport. However, in the Great Barrier Reef lagoon, these currents are minimal because of the general lack of long-period waves with significant amplitude, the low-gradient dissipative shelf and shoreface, and the tendency for the wave fields to have a substantial longshore component. Consequently, rip currents do not provide a major mechanism of shore-normal sediment transfer. Longshore drift The expression 'longshore drift' is widely, and incorrectly, used to describe the transport of sediment and water along a coastline regardless of the mechanism. Longshore drift per se occurs when wave fronts arrive at an angle to the coast, when particles (bedload) are transported obliquely up the beach during wave run-up and return seaward (shore-normal) in the backwash. Overall this results in the transport of sand particles in one direction along the coastline (Figure 25.3d). Longshore drift is a significant mechanism of bedload transport along open coasts (e.g. the New South Wales coastline). While this process occurs along the Great Barrier Reef coastline, and mainly under the influence of the southeast trade winds (causing northward transport), the relatively low-energy wave climate (caused by limited fetch) means that the active sandy beaches are effectively partitioned by headlands (cf. pocket beaches). Consequently, longshore drift is probably only locally significant in the Great Barrier Reef. Density currents Density-driven currents occur when high-density water (cold, hypersaline or turbid water) moves downslope under the influence of gravity. Gravity currents are probably best known from the deep ocean where they are associated with the redistribution of slope sediments giving rise to extensive marine turbidites. However, density flows driven by thermal differentials, salinity contrasts and suspended sediment are common off major river mouths (Mulder & Syvitski 1995). However, in the Great Barrier Reef, both the shallow nature of the shelf and a relatively persistent wind regime result in effective vertical mixing of the water column in the coastal zone. This mixing, together with a lowgradient shelf, reduces the likely significance of density-driven flows. Tidal currents Tidal currents are the dominant factor causing movement of sediment in many tidal creeks and estuaries (Wolanski 1994; Larcombe & Ridd 1995, 1996) and are important in transporting fine-grained sediment and coarse materials suspended by waves in the coastal embayments of the central Great Barrier Reef coast (Larcombe et al 1995a). During the dry season, asymmetry of tidal currents in some creeks and estuaries may cause long-term landward movement of suspended and/or bedload sediment into and up the estuaries where it may be stored (Larcombe & Ridd

300

K. J. Woolfe and P. Larcombe

the landward part of the zone of wave-induced sediment resuspension and is evident along the Great Barrier Reef coastline as a coastally trapped zone of turbid water (Figures 25.3c, h).

Coriolis Force

20" S

Figure 25.4 Map of the Townsville sector of the Great Barrier Reef (same area as Figure 25.1) showing areas of erosion and deposition together with typical coastal sedimentary deposits. Shaded area depicts an idealised configuration of the turbid coastal boundary layer, where sediment is kept in suspension by wave action and is advected north by a wind- and wave-driven current.

1995, 1996; Furukawa et al 1997; Bryce et al 1998). Thus, tides are important as factors contributing to the partitioning of sediment between the shelf and the intertidal/subaerial portions of the coastal sediment wedge. Tidal currents may be strong enough to transport bedload material in some channels between inner shelf islands (e.g. Whitsunday Islands, 200 km south of Townsville). However, along the open Great Barrier Reef coastline tidal currents are generally not strong enough to cause resuspension (Larcombe et al 1995a). The role of tidal currents in sediment transport along the coast is subordinate to that of wind- and wave-driven currents.

Eddies Eddies may occur where currents flow past islands and/or headlands. The presence of large-scale eddies (kilometrescale radii) combined with flow expansion and reduced wave influence (Figure 25.3h) is significant in the sedimenttrapping capacity of north-facing bays along the Great Barrier Reef coastline (discussed below).

Boundary currents Boundary currents occur wherever a current impinges on an impermeable or immovable boundary. In the context of this discussion, the coastline represents such a boundary to wind, wave and tidally driven currents, consequently, under some conditions, strong shore-parallel currents are developed. Moreover, as the prevailing southeast winds and associated currents have both longshore and onshore components, a coastal boundary layer is developed, in which suspended particles are moved both shoreward and northwards. This coastal boundary layer corresponds with

In the absence of other overwhelming processes, buoyant surface plumes such as the plumes resulting from freshwater floods behave like surface jets when they enter the Great Barrier Reef lagoon, and are thus affected by the spinning of the earth. In the southern hemisphere such plumes tend to be deflected to the right, with respect to their direction of movement (the Coriolis Effect) and the radius of this curvature is known as the Rossby radius (Pond & Pickard 1983) (Figure 25.3f). The Rossby radius is a function of the cosine of latitude, so that within the tropics, the radius is large. As a result, in the Great Barrier Reef lagoon, both tidal and wind-driven flows tend to overwhelm the Coriolis effect in controlling the direction of flow of river plumes.

MINOR ROLE OF CYCLONES IN LONG-TERM COASTAL SEDIMENTATION Waves are the dominant mechanism of resuspending sediment on the inner shelf of the central Great Barrier Reef (Belperio 1983; Larcombe et al 1995b; Orpin et al 1999), and once in suspension, tides and wind-driven currents are dominant in its transport. Belperio (1978) and Larcombe et al (1995a) have demonstrated that persistent southeast winds of only 5-8 m/s (10-15 knots) produce a northdirected current that completely overwhelms the southdirected ebb tide on the inner shelf near Townsville (Figure 25.2). Thus, wave- and wind-driven currents probably form a major long-term regional control on coastal sedimentation patterns. These patterns are a function of waves transporting sand northwards in the intertidal and shallow subtidal zone, and wind-driven currents carrying resuspended fine-grained sediment northwards in a turbid, coastally trapped boundary layer (Woolfe & Larcombe 1998). Most major geomorphic features along the Great Barrier Reef coastline, such as north-directed sandy spits and bars, relatively exposed linear sandy coastlines, and mud-dominated sediment accumulations in north-facing bays, attest to strong northward longshore transport. Data from the midshelf off Cape Cleveland and within Cleveland Bay itself (both near Townsville) indicate the dominance of the southeast trade winds (dry season) in the production of significant long (>7 s) period waves (Patterson 1994). While the largest waves are produced by cyclones, these are relatively infrequent episodic events (Putinen et al 1997). The patterns of sediment accumulation together with the orientation of bars and spits along the Great Barrier Reef coast are evidence that the major long-term sedimentary processes are dominated by the trade winds of the dry season. The apparent subordinate role of cyclones in long-term coastal sediment transport along the Great Barrier Reef coastline is in part counterintuitive, because cyclones unquestionably create the biggest waves and will produce the most intense sediment reworking (Gagan et al 1990). In coastal regions, storm surges and relaxation flows (Figure

Sediment dispersal Great Barrier Reef 2531 j) may enhance cyclone effects. However, Holocene coastal deposits indicative of cyclonic activity (e.g. cheniers and storm ridges) are almost entirely restricted to the southern sheltered margins of north-facing bays and the prograding regions of the coast that are sheltered by headlands, islands or reefs. This indicates that along the exposed coastlines, waves and wind-driven currents associated with the southeast trades are, with time, able to rework most cyclone deposits. On the shelf itself, the sediments resulting from Cyclone Winifred (Gagan et al 1990) were clearly marked when sampled days after the cyclone, but had been completely reworked by bioturbation a few months later.

COAST-PARALLEL DIVISION OF THE SHELF The present-day distribution of sediment (and hence the late Holocene 'average') is well-established (Maxwell 1968; Orme et al 1978; Belperio 1983; Johnson & Searle 1984). In general terms, terrigenous sediment is partitioned into a coastal wedge (Belperio 1978, 1983), whereas the mid-shelf (20-40 m water depth) is essentially starved of terrigenous sediment (Harris et al 1990; Gagan et al 1990; Ohlenbusch 1991; Carter et al 1993). In many places (net) starvation is so extreme that the pre-Holocene land surface is exposed at the sea floor. This also occurs in places on the inner shelf, for example in Cleveland Bay (Carter et al 1993) and Halifax Bay (Woolfe & Larcombe 1998). The outer shelf (40-80 m water depth) is also starved of terrigenous sediment. Carbonate-dominated sediment accumulations occur near reefs. Coral debris is dominant in the southern and central sectors, whereas accumulations of the coralline algae Halimeda form large bioherms in many northern areas (Maxwell 1968; Roberts & Macintyre 1988; Harris et al 1990; Woolfe et al 1998). This shore-parallel, three-fold division of the shelf is evident along the entire Great Barrier Reef shelf and is largely maintained by a combination of coastal boundary currents inhibiting seaward migration of terrigenous sediment. While it may appear that the mid- and outer shelf are dominated by high rates of carbonate production, cross-shelf differences in the rate of net carbonate production (i.e. from corals, algae, foraminifers, molluscs etc.) are as yet unquantified. We may view the occurrence of the carbonate province as a result of a relative absence of terrigenous sediment rather than an excessive production of carbonate. The abundance of terrigenous sediment decreases northwards and the coastal sediment wedge becomes thinner, narrower and more calcareous. Along many open and straight portions of the northern Great Barrier Reef coastline the nearshore sediment wedge is completely absent and fringing reefs pass directly into a carbonate-dominated middle shelf.

NORTHWARD-FACING EMBAYMENTS: NATURAL SEDIMENT TRAPS Fine-grained sediment accumulates where wave energies are insufficient to maintain particles in suspension and where flushing is insufficient to remove resuspended material. Consequently, north-facing bays (e.g. Upstart Bay,

301

Bowling Green Bay, Cleveland Bay) are prime sites for the accumulation of such sediment. In these embayments, wave energy is reduced because the bays are sheltered from the prevailing southeast trade winds. Wind-driven coastal boundary currents tend to be poorly formed or absent. Moreover, eddies associated with the regional northdirected coastal boundary current introduce turbid water into the generally calmer north-facing bays, where sedimentation occurs. Sediment transfer zones occur between the natural sediment traps. Along exposed straight reaches of the coast, wave- and wind-generated currents drive a strong northdirected coastal boundary current that transports nearshore sediments northwards. This current prevents the accumulation of permanent muddy deposits but in transporting finegrained material, it facilitates the production of sandy beach ridges, spits and bars. These sections of the coast represent zones where sediment may be stored temporarily while overall being moved slowly northwards. Muddy sediment is generally prevented from settling on the beachface because of wave activity. Hence, a shore-detached sediment wedge develops, with the muddier sediment generally confined to the zone below the level of the lowest astronomical tide (Woolfe & Larcombe 1998, 1999). In these transfer zones the lower beachface may be erosional and the pre-Holocene surface ('Reflector A' of Johnson & Searle 1984) may become exposed as a hard substrate. Where this occurs, colonisation by opportunistic corals is possible, and small reefal accumulations may occur within the turbid coastal boundary layer (Woolfe & Larcombe 1998, 1999). In places, such colonies may be only short-lived as they may be overwhelmed by solitary dunes migrating northwards along the intertidal and shallow subtidal zone.

CONCLUSIONS The distribution of terrigenous sediment along the central Great Barrier Reef coastline is largely controlled by the effects of southeast trade winds. These winds produce an along-shelf, north-flowing, wind-driven coastal current that is reinforced by a north-flowing wave-driven current (Figure 25.4). The resultant coastal boundary current transports suspended sediment, while the concentration of suspended sediment in the current is largely controlled by wave energy. Limited bedload is transported northwards under the influence of longshore drift. Some export of suspended sediment to the mid- or outer shelf is possible when buoyant (low-concentration) surface flood plumes spread across the lagoon during calm conditions or when offshore winds prevail. Bottom return currents may occur but, to date, their occurrence is only inferred. The absence of significant deposits of terrigenous mud on the mid-shelf shows that any such sediment does not remain there. Suspended sediment is carried into mangrove swamps and saltflats by tidal processes where some of it may be trapped. Sediment trapping also occurs where eddies cause the coastal current to carry sedimentladen water into sheltered north-facing bays, where settling and accumulation take place. Cyclones play an important role in the delivery of sediment to the coast but appear to be less important in its evolution over centuries and millenia.

302

K. J. Woolfe and P. Larcombe

ACKNOWLEDGMENTS

Amsterdam. MITCHELL A . W . , REGHENZANI J . , HUNTER H . M . & BRAMLEY R . G . V.

Vic Gostin is thanked for his review of this contribution, and, as always, our colleagues are t h a n k e d for their c o m ments a n d discussions.

M o s s A. J., RAYMENT G . E., REILLY N . & BEST E . K. 1 9 9 3 . A preliminary

REFERENCES BELPERIO A. P. 1978. An inner-shelf sedimentation model for the Townsville region, Great Barrier Reef province. PhD thesis, James Cook University, Townsville (unpubl.). BELPERIO A. P. 1983. Late Quaternary terrigenous sedimentation in the Great Barrier Reef lagoon. In: Baker J. T., Carter R. M., Sammario P. W. & Stark K. P. eds. Proceedings of the Great Barrier Reef Conference, pp. 71-76. James Cook University, Townsville. BELPERIO A. P. & SEARLE D. E. 1988. Terrigenous and carbonate sedi-

mentation in the Great Barrier Reef province. In: Doyle L. J. & Roberts H. H. eds. Carbonate-Clastic Transitions, pp. 143-174. Developments in Sedimentology 42. Elsevier, Amsterdam.

BRYCE S., LARCOMBE P. & RIDD P. V. 1 9 9 8 . T h e relative i m p o r t a n c e of

landward-directed tidal sediment transport versus freshwater flood events in the Normanby River estuary, Cape York Peninsula, Australia. Marine Geology 149, 55-78.

CARTER R. M . JOHNSON D. P. & HOOPER K. G . 1 9 9 3 . Episodic post-

glacial sea-level rise and the sedimentary evolution of a tropical embayment Cleveland Bay, Great Barrier Reef shelf, Australia. Australian Journal of Earth Sciences 40, 229-255.

FURUKAWA K., WOLANSKI E. & MUELLER H. 1 9 9 7 . Currents a n d sedi-

ment transport in mangrove forests. Estuarine, Coastal and Shelf Science 44, 3 0 1 - 3 1 0 .

GAGAN M . K., CHIVAS A. R. & HERCZEG A. L. 1 9 9 0 . Shelf-wide erosion,

deposition, and suspended sediment transport during cyclone Winifred, central Great Barrier Reef, Australia. Journal of Sedimentary Petrology 60, 456-470.

HARRIS P. T., DAVIES P. J. & MARSHALL J. F. 1 9 9 0 . Late Q u a t e r n a r y sed-

imentation on the Great Barrier Reef continental shelf and slope east of Townsville, Australia. Marine Geology 94, 55-77. HORN A. M. 1995. The surface water resources of Cape York. Cape York Peninsular Land Use Strategy Program, Department of Primary Industries, Brisbane.

JOHNSON D. P. & SEARLE D. E. 1984. Post-glacial seismic stratigraphy,

central Great Barrier Reef, Australia. Sedimentology 31, 335-352. LARCOMBE P. 2001. Chapter 24—Holocene Great Barrier Reef: sedimentary controls and implications for environmental management. Geological Society of Australia Special Publication 21, 281-294.

LARCOMBE P., CARTER R . M . , D Y E J . , GAGAN M . K . & JOHNSON D .

1995b. New evidence for episodic post-glacial sea-level rise, central Great Barrier Reef, Australia. Marine Geology 127, 1-44.

LARCOMBE P. & RIDD P. V. 1995. Megaripple dynamics and sediment

transport in a mesotidal mangrove creek—implications for palaeoflow reconstruction. Sedimentology 42, 593-606.

LARCOMBE P. & RIDD P. V. 1996. Dry season hydrodynamics and sedi-

ment transport in mangrove creeks. In: Pattariatchi C. ed. Mixing Processes in Estuaries and Coastal Seas, pp.409-425. Coastal and Estuarine Studies 46. American Geophysical Union.

LARCOMBE P., RIDD P. V . , W I L S O N B . & PRYTZ A . 1 9 9 5 a . F a c t o r s

1996. Water quality and nutrient fluxes from river systems draining into the Great Barrier reef. In: Eyles A. G., Hunter H. M. & Rayment G. eds. Proceedings of the National Conference on Downstream Effects of Land Use, pp. 23-34. Central Queensland University, Rockhampton.

con-

trolling suspended sediment on inner-shelf coral reefs, Townsville, Australia. Coral Reefs 14, 163-171.

LARCOMBE P. & WOOLFE K. J. 1999. Increased sediment supply to the

Great Barrier Reef will not increase sediment accumulation at most coral reefs. Coral Reefs 18, 163-169. LARCOMBE P., WOOLFE K. & PURDON R. G. (Editors) 1996. Great Barrier Reef: Terrigenous Sediment Flux and Human Impacts (2nd edition). CRC Reef Research Centre, Current Research, Townsville. MAXWELL W . G. H. 1 9 6 8 . Atlas of the Great Barrier Reef Elsevier,

assessment of sediment and nutrient exports from Queensland coastal catchments. Queensland Department of Environment and Heritage, Environment Technical Report 5.

MULDER T. & SYVITSKI J. P. M . 1995. Turbidity currents generated at

river mouths during exceptional discharges to the world oceans. Journal of Geology 103, 285-299. NEIL D. T. & Yu B. 1996. Simple climate-driven models for estimating sediment input to the Great Barrier Reef lagoon. In: Larcombe P., Woolfe K. J. & Purdon R. G. eds. Great Barrier Reef: Terrigenous Sediment Flux and Human Impacts (2nd edition), pp. 122-127. CRC Reef Research Centre, Townsville. OHLENBUSCH R. 1991. Post-glacial sequence stratigraphy and sedimentary development of the continental shelf off Townsville, central Great Barrier Reef province. BSc (Hons) thesis, James Cook University, Townsville (unpubl.). ORME G . R . , W E B B J . D . , KELLAND N . C . & SARGENT G . E . C .

1978.

Aspects of the geological history and structure of the northern Great Barrier Reef. Philosophical Transactions of the Royal Society of London A291, 23-35.

ORPIN A. R., RIDD P. V. & STEWART L. K. 1 9 9 9 . A s s e s s m e n t of the rela-

tive importance of major sediment transport mechanisms in the central Great Barrier Reef lagoon. Australian Journal of Earth Sciences 46, 8 8 3 - 8 9 6 . PATTERSON D. 1994. Oceanographic data collection. In: Benson L. J., Goldsworthy P. M., Butler I. R. & Oliver J. eds. Townsville Port Authority Capital Dredging Works 1993: Environmental Monitoring Program, pp. 125-147. Townsville Port Authority. Oceanography. POND S. & PICKARD G. L. 1983. Introductory Dynamical Pergamon Press, Oxford.

PUTINEN M . L., DORE T. J. & SKELLY W . C. 1 9 9 7 An a t l a s of tropical

cyclones in the Great Barrier Reef region. CRC Reef Research Centre Technical Report 19. CRC Reef Research Centre, Townsville. ROBERTS H . H . & MACINTYRE I . G . (Editors) 1988. Halimeda. Coral Reefs 6, 1 2 1 - 2 7 9 . WASSON R. J. 1997. Run-off from the land to the rivers and the sea. In: Turia N. & Dalliston C. eds. The Great Barrier Reef, Science, Use and Management: a National Conference, pp. 23-41. Great Barrier Reef Marine Park Authority, Townsville. WOLANSKI E . 1 9 9 4 . Physical Oceanographic Processes of the Great Barrier Reef, CRC Press, Boca Raton. WOOLFE K. J. & LARCOMBE P. 1998. Terrigenous sediment accumula-

tion as a regional control upon the distribution of reef carbonates. In: Camoin G. E, Bergersen D. D. & Davies P. J. eds. Reefs and Carbonate Platforms in the Pacific and Indian Oceans, pp. 295-310. International Association of Sedimentologists Special Publication 25.

WOOLFE K. J. & LARCOMBE P. 1999. Terrigenous sedimentation and coral

reef growth: a conceptual framework. Marine Geology 155,331-345.

W O O L F E K . J . , LARCOMBE P., ORPIN A . R . , PURDON R . G . , MICHAELSEN P.,

MCINTYRE C. M. & AMJAD N. 1998. Controls upon inner-shelf sed-

imentation, Cape York Peninsula, in the region of 12°S. Australian Journal of Earth Sciences 45, 611-621. P. & SUTTON D . C. 1 9 9 5 . State of the Marine Environment Report for Australia Technical Annex: 2, Pollution. Great Barrier Reef Marine Park Authority, Commonwealth of Australia, Townsville.

ZANN L .

Received 11 March 1999; accepted 25 September

2000

Geological Society of Australia Special Publication 21, 303-314

CHAPTER 26—Geomorphology and Holocene geology of coastal and estuarine plains of northern Australia Research School of Earth Sciences, Australian National University, ACT 0200, Australia. The coastal lowlands of northern Australia include wide coastal plains, estuarine floodplains and biologically rich wetlands. Prized for their natural environmental values as well as their agricultural potential the lowlands are vulnerable to saltwater invasion and soil degradation, if mismanaged. The lowland realms are far from homogeneous, however. Their underlying sediments, which largely were deposited under rising sea-level between 8000 and 6000 years ago, vary considerably in character and strongly influence present-day soils, shallow-groundwater chemistry and flow, and potential for salt invasion or acid sulfate degradation. Wide coastal plains with isolated chenier ridges are common throughout northern Australia but the sedimentary geology of lowland estuarine or deltaic basins varies with local tidal regime and river catchment size, which together govern past and present sedimentary processes. Macrotidal basins typically have wide tidal rivers that pass inland through a series of estuarine meanders, with floodplains close to river level and freshwater backswamps between the backplains and bedrock high ground. Their Holocene sediment tracts include extensive mangrove mud deposits, which accumulated under rising sea-level. Mesotidal rivers are narrower, tidal currents are weaker, their floodplains rise progressively inland, and their Holocene sediments are dominantly subtidal estuarine and nearshore deposits. Differences between sediment tracts reflect sediment transport, which is dominantly upstream under flood-tide flows in the macrotidal realm, but is downstream in the ebb-dominated mesotidal realm. Wet season fluvial floods partly counteract upstream macrotidal transport, depending on river catchment size. This chapter reviews the sedimentary geology, geomorphology and present processes of macrotidal and mesotidal lowlands in northern Australia, from the southern Gulf of Carpentaria to far northern examples including the South Alligator, Adelaide and Daly Rivers, and aims to show why future management of the present landscapes should be intimately linked to a knowledge of their past. KEY WORDS: acid sulfate soils, coasts, estuaries, geomorphology, Holocene, sea-level, wetlands.

INTRODUCTION Holocene coastal and estuarine sediment bodies are a key to interpreting similar deposits in ancient stratigraphic sequences; they are a guide to the effects of future sea level changes, and they include deposits at high risk to acid sulfate soil degradation. Furthermore, many of the estuaries and tidal rivers are not efficiently flushed and the risk of pollution is high. With these factors in view, this chapter describes the geomorphology and Holocene sedimentary geology of coastal lowlands in far northern Australia, where mangroves fringe the estuaries, tidal creeks and much of the open coast. North Australia's continental shelves are wide; nearshore seabeds tend to be muddy, and wave energies at the coast normally are low except during tropical cyclones, when storm waves and surges drive shell detritus or coral shingle inshore, building beach ridges up to several metres above highest tide level. Rivers pass through the lowlands in wide, meandering tidal channels, which vary with tidal regime and sediment supply. The far northern region from Arnhem Land to Bonaparte Gulf is macrotidal, with tides ranging from 4 m to over 6 m. In this region, tidal rivers are wide near the mouth and extend well inland, with floodplains close to river level and freshwater backswamps between the backplains and bedrock high ground. In con-

trast, in mesotidal regions where tides range to 2-3 m or are diurnal, as in northeast Queensland and the Gulf of Carpentaria, currents in tidal rivers are weaker, the channels are less wide and their banks typically rise progressively inland. In areas where runoff from the land is meagre, mangrove swamps with tidal creeks trap what little sediment is available. Figure 26.1 shows examples of these three types (described in more detail later): the South Alligator River—a typical long, funnelling macrotidal river in the far north; the Norman River—a mesotidal river of the Gulf of Carpentaria, with a very wide coastal plain; and mangrove swamp with tidal creeks but no fluvial catchment, in Missionary Bay, Hinchinbrook Island. Today's coastal and estuarine lowlands of northern Australia were formed when rising sea-level invaded prior lowlands and valleys, owing to retreat of continental ice sheets in the northern hemisphere at the end of the last ice age. At the ice age climax about 20 000 years ago, sea level was 115-130 m lower than today; flooding of the continental shelves began soon afterwards and Australia's wide northern shelves were rapidly transgressed by the rising sea. Indeed, around 11-14 000 years ago, sea-level was rising at about 2 m per century and the shorelines of northern Australia would have migrated at rates of several hundred metres every decade, where the shelves are widest. However, not until about 9000 years ago did the sea begin

304

J. Chappell

Figure 26.1 Three contrasting coastal lowlands in northern Australia. Left: the macrotidal South Alligator River and its estuarine plain, which fills a prior valley drowned by rising sea-level 6-8000 years ago. Centre: the wide coastal chenier plain of the mesotidal Norman River; chenier ridges in black. Top right: tidal creeks and mangrove swamp in Missionary Bay, Hinchinbrook Island [mangrove (shown by Vegetation' pattern) is enclosed by coastal dunes on eastern side (hook symbol) and bedrock uplands to north and south (diagonal hachure)]. Bottom right: locality map.

to invade prior valleys near the present north Australian coast. Sea-level relative to the region stabilised 6-7000 years ago and has since fallen by 1-2 m, relative to the Australian mainland coast. Field evidence shows that the actual amount of sea-level fall varies with location (Figure 26.2), which Lambeck and Nakada (1990) explain in terms of regional isostatic effects. Except on the eastern side of the coastal ranges of north Queensland, where orographic effects are strong, the annual rainfall near the coast is about 8 0 0 - 1 3 0 0 mm and declines inland. Rainfall is strongly seasonal: December through March are the wettest months and peak floods tend to occur towards the end of this season. Geomorphologic and sedimentary processes in tidal rivers of the macrotidal region are dominated by tidal flows that tend to trap sediment in the estuarine systems, whereas the mesotidal coasts are more strongly influenced by wet season floods, which move sediment seawards. Sediment yield from Australia is very low by world standards, owing

to the low relief and hard-surfaced terrain of the continent (Gale 1991) and is insufficient to account for the quantities of sediment deposited in the coastal and estuarine lowlands in Holocene times. Evidence described later indicates that large amounts of sediment moved inshore during rising sea-level, and accumulated in widespread mangrove swamps. As a legacy of their mangrove-swamp past, today's estuarine plains and wetlands are underlain by large tracts of salty, organic mud. This substrate is prone to sulfuric acidification if the water-table is lowered, through oxidation of authigenic pyrite in the buried mangrove sediments. The plains also are vulnerable if sea-level rises, because tidal invasion would lead to replacement of their rich freshwater wetlands and paperbark swamps by samphire, mangrove and salt desert. As shown below, responses to these hazards are different in the meso- and macrotidal realms but to some degree can be anticipated from a knowledge of the Holocene sedimentary geology.

Holocene coastal and estuarine plains

(a) 1.5-1 Y

Microatoll data: +1m, 5500 B.P. isobase

3000 Years B.P.

Figure 26.2 Sea-level changes relative to two different regions of north Queensland, (a) Inner zone of the northern Great Barrier Reef, Townsville to Princess Charlotte Bay (from Chappell et al 1983). (b) Southern Gulf of Carpentaria (from Rhodes 1980).

HOLOCENE SEDIMENTARY GEOLOGY Methods The Holocene sedimentary geology of coastal lowlands in northern Australia has been determined largely by drilling and trenching, coupled with survey and sampling of modern sedimentary environments, and radiocarbon dating has been used extensively to determine age structures of the deposits (Rhodes 1980, 1982;. Chappell & Grindrod 1984; Woodroffe et al 1986, 1989, 1993; Chappell 1993). Rhodes (1980, 1982) introduced modern morphostratigraphic and facies model concepts into north Australian Holocene sedimentary geology, and he also used shallow seismology to identify the Holocene basal surface. Depositional environments of Holocene sediments in northern Australia have been based on comparisons with present-day sediments and biota. Several methods have been used, in addition to conventional analyses of sediment structures, composition and particle-size parameters. Many palaeoenvironmental reconstructions from the region

305

refer to macrofauna, mostly intertidal and subtidal molluscs, but their transport and taphonomy generally has received little attention, other than in studies concerned to distinguish prehistoric human shell middens from beach sediment and other natural shell deposits (Bailey et al 1994). Microfossils have been used more systematically, however. For example, tropical foraminifers include many species with different estuarine adaptations (Wang & Murray 1983). Thus, coastal and estuarine sub- and intertidal deposits were identified in Holocene deposits from the South Alligator River, referenced to foraminifers in plankton and surface sediment samples (Wang 1990). Similarly, pollen analysis was used to identify ancient mangrove communities and intertidal deposits in Holocene sediments at Princess Charlotte Bay, Missionary Bay and the Alligator Rivers region, supported by analyses of samples from pollen traps and surface sediments (Grindrod 1988a, b; Chappell & Grindrod 1984; Grindrod & Rhodes 1984; Woodroffe et al 1986). Age structures of Holocene sediment bodies in northern Australia have been almost entirely reported in terms of conventional radiocarbon ages (i.e. in radiocarbon years Before Present or BP). This leads to errors when rates of sedimentation or other processes are calculated, because the radiocarbon time-scale diverges from calendar reckoning (Stuiver & Becker 1993). Thus, some recent studies report chronologic data both as conventional and as calibrated radiocarbon ages, where calibration has been done with an accepted procedure such as the OxCal package (Bronk Ramsey 1994). However, uncalibrated radiocarbon dates continue to be the common base for Australian Holocene studies and, unless otherwise stated, dates mentioned in this chapter are conventional ages in radiocarbon years BP, corrected for carbon reservoir effects where appropriate according to details given by the primary authors.

Coastal sediment tracts Sedimentary coasts in northern Australia are dominated by mangrove-fringed prograded coastal plains. Chenier plains such as those of the southern Gulf of Carpentaria (Figure 26.1) and beach-ridges are common. Described examples include coastal plains of Broad Sound north of Rockhampton (Cook & Mayo 1977), Princess Charlotte Bay on the eastern side of Cape York Peninsula (Chappell & Grindrod 1984), the southern Gulf of Carpentaria (Rhodes 1980, 1982), and van Diemen Gulf (Woodroffe et al 1989, 1993). Drillhole transects show that the sediments typically wedge out in the landward direction and overlie clayey palaeosols, river sand or compacted freshwater swamp sediments, which were drowned during the postglacial sea-level rise. The South Alligator coastal plain is an example (Figure 26.3). Typically, a basal transgressive zone of littoral shelly mud (often with mangrove sediment), underlies a wedge of subtidal mud and shelly sand, which in turn is capped by regressive littoral sediments including muddy mangrove deposits. Cheniers built of shell hash and/or sand punctuate the coastal plains and vary in numbers from a few distantly separated ridges (Figure 26.3) to series of multiple ridges (e.g. the coastal plain of the Norman River: Figure 26.1); at

306

J. Chappell

2120»sC 25(5"

Modernj

I I I I I I

A L6380

\ \ K \ W \ W

\ V U

Oxidised and soil zone Terrestrial sand Mangrove muds

UdUli

Marine sand and mud

| y \ \ | Pre-Holocene C

Chenier ridge

Figure 2 6 . 3 Stratigraphic section across the eastern coastal plain of the South Alligator River (after Woodroffe et al 1986) (see Figure 26.1 for location). Top, drillhole locations and radiocarbon dates; bottom, sedimentary facies. The basal mangrove sediment was deposited as rising sea transgressed the prior land surface; dates show that this occurred between 6700 and 5590 a BP. Subsequent progradation of the coast was rapid from about 5000 to 2000 a BP and has been relatively slow since then. Chenier ridge-building events occurred about 5000 and 2000 a BP; dates of 940 and 480 a BP from shell middens on chenier crests are evidence of recent human occupation.

their base, the coarse chenier sediments overlie fine-textured intertidal sediments that may include mangrove muds. Series of beach ridges overlying sandy intertidal sediment occur at some localities, such as the deltaic coastal plain of the Edward River on western Cape York Peninsula (Rhodes 1980, 1982). Shell middens and shell mounds commonly occur on the cheniers and beach ridges (two are shown in Figure 26.3). In some areas these are very large. Near Weipa (western Cape York Peninsula), for example, there are a large number of mounds over 5 m in height and the largest reported is 13 m high composed almost entirely of large, edible mollusc shells: the mounds are considered to have been built by humans (Bailey et al 1994). Radiocarbon ages of mangrove and shelly beach deposits at the landward edge of the coastal sediment wedge show that in most cases coastal progradation (regression) commenced around 6-5000 radiocarbon years BP (Figure 26.3). The regressive deposits become younger seawards although stratigraphic evidence shows that progradation often but not invariably was interrupted when chenier ridges were built, apparently during intense storms (Rhodes 1980; Chappell & Grindrod 1984; Woodroffe & Grime 1999). In the longer term, radiocarbon ages show that progradation rates have varied over the last 6000 years. In the southern Gulf of Carpentaria, for example,

progradation initially was rapid and later slowed, becoming almost zero at the present day (Rhodes 1982). A similar pattern is seen at the South Alligator River, according to dates shown in Figure 26.3, but a different pattern occurred at Princess Charlotte Bay, where progradation initially was slow but accelerated in the last 2000 years (Chappell & Grindrod 1984). Trends doubtless are affected by the supply and nearshore transport of sediment, but the geometry of sediment accumulation is important. If the sediment is supplied at a constant rate from the land, progradation of the sedimentary wedge is likely to decelerate if the nearshore water depth increases as the coast moves seawards; on the contrary, progradation may accelerate where sedimentation leads to progressive shallowing of the basin (Chappell & Grindrod 1984). Fluctuations in rates of chenier plain progradation and chenier ridge accretion on thousand-year time-scales, recognised from radiocarbon dating studies on several long coastal reaches, including the Gulf of Carpentaria, Princess Charlotte Bay and van Diemen Gulf, have been interpreted as evidence for variations of sediment yield from the hinterland (Chappell & Thorn 1986) or changes of wave climate (Lees & Clements 1987). However, as T. Graham (pers comm. 1998) has pointed out, the effects of changes to coastal geometry and longshore transport, that are liable to

Holocene coastal and estuarine plains

•

Big s w a m p mangrove in core

Riverbank exposure of big s w a m p Track of meandering river

UH Mangrove area (proveh) Ijijijijijl Mangrove area (inferred) Estuarine sedimentation jjyyvyi Shallow marine sedimentation over drowned mangrove forest

Figure 26.4 Estuarine plains of the South Alligator River, showing the extent of mangrove sediments deposited between 7000 and 5000 a BP (after Woodroffe et al 1985, 1986 who refer to the widespread mid-Holocene mangrove deposits as the 'big swamp'). Numbers are radiocarbon dates for mangrove sediments in drillholes at locations as marked; most dates are from 1.5-6.0 m depth. Stratigraphically higher dates are printed above stratigraphically lower dates, at sites with more than one date. occur when a coast progrades, should be considered before climatic explanations are firmly accepted. Evidence for variations in the past occurrence of cyclonic storms also has been sought in north Australian sequences of chenier and beach-ridge deposits. On the basis of radiocarbon ages of events represented by coral shingle ridges at the Palm Islands and Lady Elliot Island, Chivas et al (1986) found no statistically compelling evidence for any changes in storm frequency over the last 4000 years. Subsequently, more detailed analyses of ridge sequences at the Palm Islands and Princesses Charlotte Bay found no evidence for any change over the last 5000 years (Hayne 1997; Hayne & Chappell in press).

Estuarine sediment tracts Tracts of Holocene estuarine sediments are associated with tidal rivers in northern Australia. Filling prior valleys and bordered by bedrock uplands, estuarine sediments underlie

307

wetlands and floodplains inland from the coastal plains, and pass upstream into fluvial sediments. The estuarine tracts of larger rivers vary in length from about 20 to 100 km inland, depending on the size and gradient of the prior valley, and tend to be most extensive in macrotidal systems. In the macrotidal Alligator Rivers, for example, estuarine sediments underlie the floodplains and broad tributary wetlands up to or beyond the tidal limits, about 90 km inland (Woodroffe et al 1986; Nanson et al 1993), and extend well beyond the tidal limit in the case of the Glyde River-Arafura Swamp in Arnhem Land. In many mesotidal systems such as the McArthur River, described by Woodroffe and Chappell (1993), the transition from estuarine to fluvial sediments is well downstream of the tidal limit. Thickness of the Holocene sediments is controlled by the underlying topography and varies from a few metres to over 20 m. Typically overlying a basal surface of clayey palaeosols, fluvial sand or thin, compacted peat, the estuarine sediments are dominated by three facies groups: (i) channel sediments of rippled or cross-bedded sand and mud; (ii) mangrove mud and muddy peat; and (iii) shelly estuarine sand. Macro- and mesotidal systems differ in terms of proportions of these groups. The macrotidal case is illustrated by the South Alligator system (Figure 26.4): mangrove mud and muddy peat comprise the bulk of the sediments underlying the estuarine plains except for areas occupied by channel sediments. The organic content of the mangrove facies commonly is higher below depths of 4 - 6 m and is higher in the central and upstream parts of a basin, than in the region of the wide estuarine funnel, which typically includes shelly sand and sandy mud, similar to subtidal deposits that underlie the coastal plains. In the mesotidal realm (e.g. the Norman River: Figure 26.1), estuarine deposits are dominated by shelly marine sand and mud, and mangrove sediments usually are restricted to the basal transgressive deposits and regressive intertidal sediments, if they occur at all. Both macro- and mesotidal systems include belts of channel sediments, principally following the path of the tidal river channel (or channels, in the case of deltas such as those of the McArthur River and Gilbert River deltas: Woodroffe & Chappell 1993; Jones et al 1993).

Age structure and origin of the sediments Broadly speaking, the Holocene tracts in both the macroand mesotidal realms are bodies of littoral and estuarine sediment, threaded with channel deposits. However, their age structures, documented from drilling and radiocarbon dating, are different. In the mesotidal case, the shoreface prograded after sea-level stabilised and estuarine sediment bodies expanded progressively seaward, in the same manner as their adjacent coastal plains; thus, time-surfaces dip gently offshore and become younger seawards. The pattern is well-illustrated by radiocarbon dating results from the deltaic plains of rivers entering the Gulf of Carpentaria, such as the Norman and its neighbours (Rhodes 1980, 1982) and the McArthur (Woodroffe & Chappell 1993). In contrast, the macrotidal estuarine tracts filled vertically during the late stages of rising sea-level and the sediment surface effectively kept pace with sea-level. This was demonstrated by radiocarbon dates from drillholes

308

J. Chappell

Figure 26.5 Estuarine plains of the (a) Daly and (b) Adelaide Rivers, Northern Territory. Both featured very extensive mangrove swamps ('mangrove fades') in midHolocene times but have contrasting tidal river channels that reflect differences of catchment area and Holocene sediment thicknesses. The Daly tidal river (catchment 49 000 km2) has highly mobile estuarine meanders in a wide scroll plain set in Early to midHolocene estuarine and mangrove sediments up to 30 m thick; the Adelaide River (catchment about 5000 km2) upstream of a short estuarine funnel follows a series of stable, sinuous meanders inherited from the prior fluvial river that flowed in this course before sea-level rose; the midHolocene mangrove sediments are only 4-7 m thick near the river and are thinner away from it. For a period before 2300 a BP the Adelaide River flowed into Chambers Bay via a funnelled tidal river (upper left arrow symbol).

throughout the estuarine plains of the South Alligator; all drillholes show similar ages (7000-5500 a BP) at shallow depths, throughout the plains (Figure 26.4). Pollen and sediment analysis show that the accreting surface was an intertidal mangrove swamp (Woodroffe et al 1985, 1986, 1989). Drilling and dating from estuarine plains of the Daly and Adelaide Rivers (Figure 26.5) showed similar results (Chappell 1993; Woodroffe et al 1993). By keeping pace with rising sea-level, sedimentation in the mid-Holocene macrotidal mangrove basins of northern Australia differed very significantly from Holocene mud basins impounded by sand barriers on the high-energy coasts of southeastern Australia described by Roy (1984) and Roy et al (1994), many of which have not yet filled despite that sedimentation has been sustained for the last 6000 years. It is unlikely that 'keep-up' sedimentation in the macrotidal systems was supported by fluvial sediment, particularly as regional rates of denudation in far northern Australia appear to be even lower than rates for the high-

lands of southeastern Australia (Gale 1991). Data from 'keep-up' mangrove sediments that accumulated between 7500 and 5500 a BP in the South Alligator estuarine tract illustrate the problem: the minimum sediment input required is at least 1.3xl0 6 t/y, which is about 25 times the catchment yield indicated by historical gauging in the region (<5xl0 4 t/y, i.e., approximately 3-10 t km - 2 yh Woodroffe et al 1986, 1993). Furthermore, not all sediment from the catchment reaches the estuarine system: studies from Magela Creek in the East Alligator system (Nanson et al 1993) and the Arafura Swamp in Arnhem Land (author's unpubl. data) show that a substantial amount is trapped in broad freshwater wetlands upstream of the estuarine systems. Woodroffe et al (1986, 1993) considered that the difference between Holocene sediment volume and fluvial supply is far too large to be accounted for by climate variability in the last 8000 years and argued that most of the 'missing' sediment was transported landwards from the nearshore region,

Holocene coastal and estuarine plains

309

Distance from the mouth km Figure 2 6 . 6 Evolution of salinity through the dry season in the South Alligator River, shown as vertically and tidally averaged longitudinal profiles. Numbers at right indicate days elapsed after peak of typical wet season flood. Heavy lines, observations; light lines, model profiles at 20-day intervals computed with equation 1.

during the culminating stages of sea-level rise (transport mechanisms are examined under 'geomorphology and sedimentary processes', below). Similar conclusions have been reached for other north Australian macrotidal sediment tracts including the Daly River, which has a catchment area and terrigenous sediment input about 5 times larger than the South Alligator (Woodroffe et al 1993; Chappell 1993). Consistent with sediment input from the seaward direction, the percentage of clastic sediment in the mid-Holocene 'keep-up' mangrove deposits has been found to increase seawards in the South Alligator, Daly and Arafura-Glyde River systems. In the mesotidal realm there may be similar shortfalls of fluvial sediment but this is more difficult to determine, because the Holocene sediments extend further offshore and are less clearly trapped within prior valleys. Sedimentation rates in the macrotidal systems diminished abruptly when the widespread mangroves gave way to freshwater floodplains and sediment from tidal waters no longer moved overbank. In the case of the South Alligator, for example, sedimentation on the plains in the last 5000 years amounts to less than a metre, i.e. <2 cm per century, commensurate with the present-day catchment yield, cited above. It is arguable that this major environmental change was fostered by the slow fall of sea-level that occurred during the last 6000 years, illustrated in Figure 26.2.

GEOMORPHOLOGY AND SEDIMENTARY PROCESSES Tidal rivers are the key geomorphologic element in the coastal and estuarine sediment tracts of northern Australia. With floodwaters and sediment from their inland catchments, tidal rivers mould their sedimentary landscapes. In contrast, tidal creeks, which carry tidewater but have no significant upland catchment (cf. the Missionary Bay man-

grove: Figure 26.1), are but secondary features of the estuarine tracts. Channel morphology and sediment transport in the tidal rivers of northern Australia vary with tidal range and tidal period. Macrotidal rivers typically occur where tides are semidiurnal with a spring-tide range greater than 4 m (e.g. Bonaparte Gulf, the Darwin region, Kakadu, the Blyth River-Glyde River region in Arnhem Land, and Broad Sound); mesotidal types occur where the largest tides are dominantly diurnal (e.g. the southern Gulf of Carpentaria) or where semidiurnal tides have a spring-tide range of 2-3 m, as on most of the north Queensland coast. Wet season rainfall and runoff strongly influence estuarine processes in the region. Floodwaters from heavy rainfall events, which often stem from deep rain depressions derived from tropical cyclones, typically flush the tidal rivers and form nearshore plumes superimposed on salt-wedge structures at or within estuary mouths. During a flood, tidal reversal ceases throughout most of a tidal river, although tidal rise and fall continue to act and affect floodwater velocity. Flood recession rates vary with catchment size: during recession of a Daly River flood, for example (catchment 49 000 km2), halving of discharge takes about two days for surface runoff and about 20 days for baseflow (Chappell & Bardsley 1985). Recession is quicker for smaller catchments. From the end of the wet season, seawater diffuses upstream as reversing tidal behaviour is re-established and salinity increases throughout the tidal rivers as the dry season advances; final salinity levels near the tide limit are highest in macrotidal rivers with no dry season freshwater inflow. Salinity is also raised by evaporation, and can exceed seawater salinity within the lower reaches, creating a reverse salt wedge against the sea (Wolanski 1986). Water-balance calculations show that these systems are dominated by marine inflow through the dry season. Seawater intrusion can be modelled by the diffusion equation:

310

2s at

J. Chappell

ds+ d_Lds\ ax axlDaxJ

where S is salinity at distance x upstream from the mouth, D is the coefficient of turbulent diffusion and U is the net discharge or inflow (freshwater out or seawater in). Seawater inflow is enhanced by evaporation from the river surface. To illustrate, Figure 26.6 shows dry season evolution of salinity computed by equation (1) for the South Alligator River, matched to observed salinity profiles: a net inflow of 30 m 3 s _ 1 (U-0.002 m s _ 1 at the entrance) throughout the dry season was required to match the observations. Summed over the dry season (250 days) this is sufficient to fill the tidal river (5.2x10 8 m 3 ) and meet the evaporative loss (about 2000 mm) from its surface. Similar results are obtained for other systems, including those with large catchment and a sustained freshwater inflow through the dry season (such as the Daly: 10 m 3 s _ 1 dry season freshwater inflow; 25 m 3 s _ 1 seawater inflow to match observed salinity profiles). In short, these systems do not seem to export water or solutes to the sea in the dry season: the reverse salt wedge acts as a barrier and net inflow is upstream.

Tidal river morphology Tidal river morphology reflects the interactions between tidal and fluvial flows, and bank erosion and sedimentation. In common with all tide-dominated sedimentary estuaries, discharge of the tidal prism determines the cross-sectional area at the mouth (O'Brien 1980). If fluvial flow had no influence, the funnel should taper upstream in proportion to the decrease of the tidal prism. If the mean depth is constant (approximately true in north Australian examples), channel width W decreases negative-exponentially: W - W n exp(-kx); k 0

— HtU

(2)

where W 0 is width at the mouth, x is distance upstream from the mouth, v and t are tidal amplitude and period, and U is the average peak tidal-current velocity (Chappell & Woodroffe 1994). This equation rests on the important assumption that bank erosion is governed by shear stresses developed by peak tidal currents Macrotidal rivers in north Australia are strongly influenced by tidal forces. Peak flood-tide currents reach 1.4-1.8 m s _1 throughout much of their length. Their lower reaches are wide, tapering estuarine funnels, typically 20-40 km long, that conform with equation (2) in terms of width versus distance with values of k around 5 - 8xl0~ 5 rrr 1 (Vertessy 1990; Chappell & Woodroffe 1994). Funnels lead inland to long tidal channels that pass through wide, seasonally flooded plains. At the point where tidally reversing flows become negligible, channels are governed by floods from their freshwater catchments. Except for low levees in the upstream reaches, the floodplain surface near the river typically lies less than 0.5 m above high spring-tide level and slopes gently downwards away from it. Wetlands and backswamps between the plains and the surrounding bedrock uplands generally lie below high-tide level but are isolated from tidal invasion as long as the plains and tidal river banks are not

breached. When banks are breached, saltwater invasion has drastic effects on the freshwater plains and backswamps of the macrotidal and mesotidal wetlands alike. Knighton et al (1991) described examples of conspicuous degradation that has occurred this century in northern Australia, largely owing to breaches caused by feral buffalo. Between their estuarine funnels and upper tidal reaches, macrotidal river channels typically include estuarine meanders of the type described by Ahnert (1960), which have a sharp cusp at the apex of a bend between broad pools and mid-channel shoals. Ebb- and flood-tide currents typically follow paths on opposite sides of the shoals (Chappell & Woodroffe 1994). Averaged over decades, cuspate meanders tend to be stationary although the shoals change position from wet season to dry season: for example, airphotos indicate that cuspate meanders in the South Alligator River have not shifted significantly in the last 40 years (Woodroffe et al 1986). Migrating sinuous estuarine meanders also occur, particularly in macrotidal rivers with a large fluvial catchment. A notable example is the lower Daly River (Figure 26.5a), in which sinuous meanders migrate, cut off, and regenerate rapidly, in response to seaward transport of sediment during wet season floods from the large catchment: airphotos over 40 years show an average migration rate of 20 m y 1 and maximum rates >50 m y""1 (Chappell 1993). In mesotidal realms of northern Australia, tidal rivers typically are approached from seaward by long, curving channels, flanked by muddy sand bars or tidal levees that emerge at low tide. The mesotidal channels diminish slowly in width as they pass through their coastal plains into the fluvial floodplains, and the fit of width-distance data to equation (2) is poor. The banks rise gently inland and often stand 5-10 m above the river at its tidal limit. Dry season tidal currents diminish upstream from the mouth and maximum bank shear occurs during wet season floods. In the southern Gulf of Carpentaria, flow in the tidal reaches is enhanced by tidal runoff from the coastal plains, which flood widely because nearshore sea-level in the region rises by up to 1 m, during the wet season (December-March).

Sedimentary processes Tidal flows dominate sediment transport in macrotidal rivers of northern Australia, although freshwater floods play a role in the wet season; in contrast, transport in the mesotidal rivers is dominated by wet season floods. Tides in the macrotidal rivers generate high current velocities throughout most of their length but the tidal curve becomes increasingly asymmetrical with distance up-river, so that flood tides are of shorter duration and have higher current velocities than ebb tides, and produce flood-tide bores in shallow reaches (Woodroffe et al 1986; Vertessy 1990). The pattern reverses during wet season floods, when seaward flow dominates, supported by flow from both the catchment and the floodplain. In contrast, tidal currents in the mesotidal rivers are weaker and diminish upstream. Where tide water spreads significantly beyond the channel across tidal flats and mangrove, the flow pattern is opposite to that in the macrotidal rivers, with the ebb having shorter duration and higher velocities than the flood tide. Sediment transport in the tidal rivers and sediment

Holocene coastal and estuarine plains exchanges between the rivers and their estuarine plains are separate processes. Within the river channels, sand transport increases with current velocity (u) and water slope (s); when reduced purely in terms of u, most transport formulae contain a kernel of the form u z where z is equal to or greater than 3 (a range of formulae is given by TACPSM 1971). Integrated through a tide cycle, net transport is directed upstream in macrotidal rivers during the dry season because the peak velocity of flood-tide currents exceeds the ebb peak velocity. Suspended sediment behaves similarly. Turbulence from high velocities and shallow water in early flood tide generates high suspended-sediment concentration through the water column, but sediment settling at high-tide slack water leads to formation of a lutocline (a water layer with a strong suspended-sediment concentration gradient), with relatively low current velocities in the lower, sediment-rich layer (Wolanski et al 1988). The vertically integrated velocity-density average tends to be greater for the flood than for ebb tide, inducing upstream drift of suspended sediment. Furthermore, upstream drift of suspended sediment is probably enhanced by diffusive and evaporative invasion of seawater, reviewed above. The drift reverses during wet season floods but, on an annual basis, there may be little or no net export of sediment from the river. In mesotidal channels, on the other hand, highest velocities and net transport are directed downstream throughout the year but principally in the wet season. Morphologically, this is manifest by lobate and elongate muddy sandbars that flank their nearshore approach channels. The estuarine floodplains and mangroves accrete sediment carried by tidal and fluvial floodwater. This is particularly important during rising sea-level when the ability of the intertidal surface to keep pace with the rising sea is determined by the flux of sediment from the channel. In the case of mangrove, sediment moves in and is trapped during high-tide flooding (Wolanski 1995), whereas it reaches freshwater wetlands only during wet season floods. Suspended-sediment concentration governs the potential sedimentation rates. In macrotidal rivers, high turbulence generates high suspended-sediment concentration (typically 1 - 1 0 g L - 1 ), and sediment lost overbank is replenished by upstream transport, provided that there is a source to seaward. The combination of high suspendedsediment concentration and a nearshore supply from sediment reworked from the drowned landscape during rising sea-level appears to account for keep-up mangrove sedimentation in the macrotidal systems in Holocene times. In contrast, suspended-sediment concentration in wet season floodwaters is low (values measured for South Alligator and Daly River floodwaters entering the tidal system range from 0.01 to 0.1 g L - 1 ), leading to low potential sedimentation rates in the freshwater wetlands that succeeded the mid-Holocene mangroves, as described above. Furthermore, the actual sedimentation rate may be less than the potential rate, depending on the rate of sediment loss from the floodplain. This can be very low in mangrove or sedge and grass plains, but where the plain is bare, sediment is efficiently deflated until the surface accords with the capillary fringe of the water table (at Princess Charlotte Bay, older chenier ridges are capped by dunes built of pelletised clay deflated from adjacent mudflats:

311

Chappell & Grindrod 1984). This mechanism has led to the development of broad, barren surfaces at about high spring-tide level in the southern Gulf of Carpentaria, where remnants of earlier cheniers, perched on clay pedestals, show that deflation has lowered the plain by up to 2 m near its landward margin (Rhodes 1980).

Channel inheritance and extinction Cuspate and active sinuous meanders are dynamic forms in macrotidal estuaries that lie entirely within Holocene sediment bodies and so too are the sinuous meanders of many mesotidal systems. Palaeochannels on the estuarine plains are evidence for migration, growth and cutoff of tidal meander loops; indeed, evidence for repeated switching between cuspate and sinuous forms is well-preserved in the active meander belt of the Daly River (Chappell & Woodroffe 1994). However, there are also reaches of tidal river that follow fluvial channels inherited from preHolocene rivers, that existed before their valleys were inundated by post-glacial rising seas. Such rivers persisted in their ancient courses while mangrove sedimentation proceeded at their sides, during the last stages of sea-level rise, and have remained in place through the subsequent period of stable sea-level. The Adelaide River is the most striking example: a 40 km reach of smoothly sinuous meanders (with mangroves on the convex banks) flows through in a wide plain with mid-Holocene mangrove sediments only 1 m beneath its surface; drilling and sounding shows that these are 4 - 7 m-thick near the tidal river (Figure 26.5b). The river itself follows a channel incised several metres into the underlying lateritic bedrock (Chappell & Woodroffe 1994). Similar though shorter and less spectacular reaches occur in other tidal rivers in both their present channels and their palaeochannels. There is evidence that some tidal rivers experienced major changes of course or even were extinguished after their estuarine floodplains were established, owing to blockage by coastal sediment. The Adelaide River is a notable example that followed its inherited meandering course during rising sea-level and 'big swamp' phase, but later switched eastward to flow into Chambers Bay where it developed a typical macrotidal estuarine funnel (Figure 26.5b). This apparently was blocked during an episode of beach-ridge accretion, about 2500 years ago, when the river switched back and reoccupied its sinuous channel (Woodroffe et al 1993). During the same period, the nearby Mary River estuary developed two major, meandering funnels, which later were blocked with sediment and died as tidal systems, so that the fluvial discharge of the Mary River largely dissipated in broad swamplands and evaporated with only minor outflow to the sea (Woodroffe et al 1993).

Summary: geomorphologic groups and boundary conditions Tidal river systems reflect their hydrologic and sedimentary boundary conditions. Chappell and Woodroffe (1994) classified tidal rivers in terms of two indices: (i) the ratio Q/Q^, where is the mean flood-tide inflow across the entrance and Qf is the freshwater bankfull flood at the tide limit; and

312

J. Chappell

(ii) the ratio V/V e , where V e is the volume of Holocene estuarine sediment and V f is the mean annual sediment yield from the catchment, multiplied by the age of the Holocene sediment body. The first index represents a spectrum from tide dominance (high Q/Qf) to fluvial dominance (low Q/Qf), which also was used by Wright and Coleman (1973) to discriminate between deltas. The second index represents a spectrum from basins starved of fluvial sediment (low V/V e ) to systems where sediment passes through an active delta to the marine realm (high V/V e ). In these terms, macrotidal, sediment-starved systems in northern Australia, such as the Alligator Rivers, are readily separated from macrotidal, sediment-bypassing systems such as the Fly River in Papua New Guinea. Equally, starved basins are separated into 'catch-up' micro- and mesotidal basins (low Q/Q^) such as sand-barrier mud basins of southeastern Australia, and macrotidal 'keep-up' basins (high Q/Qf) (Chappell & Woodroffe 1994). A weakness of the schema is that it does not indicate whether sedimentation in an estuarine basin will keep up with or lag behind rising sea-level. As shown earlier, this does not depend on the tidal range but on whether the highest tidal velocities occur during flood or ebb tide, which is determined by the area of tidal flooding relative to the channels. For example, tidal creeks incised into supratidal mudflats have higher flood-tide velocities, whereas those that drain broad intertidal flats or mangroves are ebb-dominated. The threshold between these conditions depends both on the tidal range and the extent of tidal flooding. This was explored experimentally by Wolanski and Chappell (1996), who used a numerical tide model (non-linear terms were included) to determine the amount of sea-level rise required to change a macrotidal river from flood-tide dominance to ebb dominance. When calibrated with field measurements of channel profiles and tides for the South Alligator River, the model showed that ebb velocities increase throughout the river as sea-level rises but floodtide velocities are almost unaffected. The model South Alligator became ebb-dominated with 1 m of sea-level rise. The result indicates that 'keep-up' behaviour is impossible if sea-level rises indefinitely unless (i) sedimentation allows the floodplain to keep up, and (ii) the depth of tidal flooding is less than a critical value.

FUTURE SHOCKS Coastal and estuarine plains in tropical Australia are vulnerable to sea-level rise and to inappropriate land use. Lying close to or even below high tide level, wetland plains in the macrotidal realm obviously are susceptible to saltwater invasion, leading to destruction of their freshwater ecosystems, if tidal river banks are breached or regularly overtopped. Vegetated plains lying above high spring-tide level, such as coastal plains and mesotidal estuarine floodplains, are liable to deflate to bare flats if the vegetation is destroyed, either by marine encroachment or by overgrazing. The tidal invasion and substantial damage that has followed buffalo in some parts of Northern Territory (Knighton et al 1991) is a foretaste of the sort of transformation that will occur if the global greenhouse effect leads to sea-level rise, in the next 50-100 years.

Sea-level rise Freshwater wetlands will change to mangroves, salt marshes and tidal mud flats if sea-level rises in future: the question is how much sea-level rise will lead to widespread and irreversible changes. In principal, the plains and wetlands could survive if sedimentation from wet season floods kept pace with rising sea-level (dry season tidal sediment is carried by saline tidewater) but this is most unlikely if sea-level rises by more than 1-2 cm in the next century, because the sediment yield of north Australian catchments is very small. In the case of the South Alligator, as shown above, sedimentation on the plains in the last 5000 years amounts to less than 1 m, i.e. <2 cm per century, which is commensurate with the catchment sediment yield today. The sea-level rise forecast for the next hundred years could be between 10 and 120 cm (Wigley & Raper 1992; Gregory 1993). Thus, keep-up behaviour—if it happens at all—must be supported by sediment transported tidally from the seaward direction. This process, irrespective of its rate, must lead to salinisation of the freshwater wetlands. There are two parts to the question of whether tidally supported sedimentation will keep pace with rising sealevel: (i) maintenance of flood-tide dominance while sealevel rises; and (ii) the origin of sediment derived from the seaward direction. The question of tidal behaviour was outlined in the previous section. As for the sediment source, Wolanski and Chappell (1996) argued that rising sea-level would cause sediment to be eroded near the entrance of a tidal river, owing to the increase in tidal prism caused by overbank flooding upstream. Calculations for the South Alligator River showed that bank shear stresses increase with rising sea-level (ebb velocities increase more rapidly than flood velocities and eventually exceed them, as noted earlier), and indicated that bank shear stresses exceed present values with a sea-level rise of 0.5 m. At this point, calculated tidal asymmetry and net transport were still directed upstream but less strongly than today. Wolanski and Chappell (1996) noted that 0.5 m depth of inundation at high tide is consistent with keep-up sedimentation under mangrove that prevailed in the South Alligator when sealevel was rising 7-6000 a BP, and concluded that sedimentation can keep up with sea-level rise, where sediment is derived from coastal retreat and widening of the estuary entrance. Wolanski and Chappell (1996) also showed that similar calculations for the mesotidal Norman River lead to quite different results: the erosion threshold will be reached with a rise of only 0.1 m and net sediment transport is entirely downstream. Again, their conclusions were supported by the Holocene sedimentary record.

Inappropriate land use Acid sulfate soil degradation potentially is a future hazard for the estuarine plains of northern Australia. The tracts of Holocene mangrove sediment, which underlie large areas of the estuarine plains, contain pyrite produced in situ under the reducing conditions that prevail today below about 1 m depth. Sulfuric acidification of this type of sediment is caused by oxidation, which occurs if the watertable is lowered by artificial drainage. Acid sulfate degradation of drained saltmarsh and mangrove soil, is a widely recognised problem globally (Dent 1986). In south-

Holocene coastal a n d estuarine plains eastern Australia, drainage of estuarine floodplains for agriculture has caused both acid sulfate soil problems and acid-toxic conditions in drainage and runoff waters (Bush 1993; Lin et al 1998; Sammut et al 1996). The potential is high for similar problems to occur where mangrove sediments underlie the estuarine plains of far northern Australia if these lands are inappropriately drained in future for agricultural use. Considering that the acid sulfate hazard is much lower in Holocene freshwater deposits than in estuarine sediments, facies maps of the Holocene sediment bodies should be a guide when land use intensifies in these regions. Finally, the risk of pollution is high in the estuaries and tidal rivers of far northern Australia if potentially hazardous substances find their way into waterways. This is particularly so in the dry season, when runoff ceases and seawater invades the estuaries and a reverse salt-wedge creates a barrier to exchange across their mouths. Indeed, estuaries are not always flushed in the wet season. For example, comparison of observed flooding of the the Daly in March-April 1984 (Chappell & Ward 1985) with previous records (Chappell & Bardsley 1985) suggested that the estuary may not have flushed during the poor wet season of 1 9 6 9 - 7 0 . Engineered reduction of flow by damming would both enhance upstream sediment drift and further reduce the ability of these systems to flush any pollutants that may enter them in future.

REFERENCES AHNERT F. 1960. Estuarine meanders Geographical Review 50, 390-401.

in Chesapeake

Bay.

BAILEY G., CHAPPELL J. & CRIBB R. 1994. The origin of Anadara

shell

mounds at Weipa, North Queensland, Australia. Archaeology in Oceania 29, 6 9 - 8 0 .

BRONK RAMSEY C. 1994. OXCAL V2, a Radiocarbon Calibration and Analysis Program. Radiocarbon Accelerator Unit, Research Lab for Archaeology, Oxford. BUSH R. (Editor) 1993. Proceedings of a National Conference on Acid Sulphate Soils, Coolangatta June 1993. New South Wales Department of Agriculture, Wollongbar. CHAPPELL J. 1993. Contrasting Holocene sedimentary geologies of lower Daly River, northern Australia, and lower Sepik-Ramu, Papua New Guinea. Sedimentary Geology 83, 339-358. CHAPPELL J. & BARDSLEY K. 1985. Hydrology of the lower Daly River, Northern Territory. Australian National University, North Australia Research Unit, Monograph. CHAPPELL J., CHIVAS A., WALLENSKY E . , POLACH H . A. & AHARON P.

1983. Holocene palaeo-environmental changes, central to north Great Barrier Reef inner zone. BMR Journal of Australian Geology & Geophysics 8, 223-235. CHAPPELL J. & GRINDROD J. 1984. Chenier plain formation in northern Australia. In: Thorn B. G. ed. Coastal Geomorphology in Australia, pp. 197-231. Academic Press, Sydney. CHAPPELL J. & THOM B. G. 1986. Coastal morphodynamics in northern Australia: review and prospect. Australian Geographical Studies 24, 110-127. CHAPPELL J. & WARD P. 1985. Seasonal tidal and freshwater chemistry of the South Alligator and Daly Rivers. In: Bardsley K., Davie J. & Woodroffe C. eds. Coasts & Tidal Wetlands of the Australian Monsoon Region, 97-108. Australian National University, North Australia Research Unit, Mangrove Monograph 1. CHAPPELL J. & WOODROFFE C. D. 1994. Macrotidal estuaries. In: Carter R. W. G. & Woodroffe C. D. eds. Coastal Evolution, pp. 187-218. Cambridge University Press, Cambridge. CHIVAS A., CHAPPELL J., POLACH H . A., PILLANS B . & FLOOD P. 1 9 8 6 . R a d i o c a r b o n evidence for the timing and rate of island develop-

ment, beach-rock formation and phosphatisation at Lady Eliot

313

island, Queensland Australia, Marine Geology 69, 273-287. COOK P. J. & MAYO W. 1977. Sedimentology and Holocene history of a tropical estuary (Broad Sound, Queensland). Bureau of Mineral Resources Bulletin 170. DENT D. 1986. Acid Sulphate Soils. International Institute for Land Reclamation and Improvement, Wageningen, The Netherlands, Publication 39. GALE S. J. 1991. Long-term landscape evolution in Australia. Earth Surface Processes and Landforms 17, 323-343. GREGORY J. M. 1993. Sea level changes under increasing atmospheric C 0 2 in a transient coupled ocean-atmosphere experiment. Journal of Climate 6, 2247-2262. GRINDROD J. 1988a. The palynology of Holocene and mangrove saltmarsh sediments, particularly in northern Australia. Review of Palaeobotany and Palynology 55, 229-245. GRINDROD J. 1988b. Holocene mangrove history of the South Alligator River estuary, Northern Territory, Australia. PhD thesis, Australian National University, Canberra (unpubl.). GRINDROD J. & RHODES E. G. 1984. Holocene sea level history of a tropical estuary: Missionary Bay, north Queensland. In: Thom B. G. ed. Coastal Geomorphology in Australia, pp. 151-178. Academic Press, Sydney. HAYNE M. 1997. Cyclonic frequency during the Holocene in northeastern Australia. PhD thesis, Australian National University, Canberra (unpubl.). HAYNE M. & CHAPPELL J. Cyclone frequency during the last 5000 years at Curacoa Island, North Queensland, Australia. Palaeogeography, Palaeoclimatology, Palaeoecology (in press). JONES B . G . , MARTIN G . R. & SENAPATI N . 1 9 9 3 : R i v e r i n e - t i d a l i n t e r a c -

tions in the monsoonal Gilbert River fandelta, northern Australia. Sedimentary Geology 83, 319-338.

KNIGHTON A. D . , MILLS K. & WOODROFFE C. D . 1 9 9 1 . Tidal c r e e k

extension and saline intrusion in northern Australia. Geology 19, 831-834.

LAMBECK K. & NAKADA M. 1990. Late Pleistocene and Holocene sea level change along the Australian coast. Palaeogeography, Palaeoclimatology, Palaeoecology (Global and Planetary Change) 89, 143-176. LEES B. G. & CLEMENTS A. M. 1987. Climatic implications of chenier dates in northern Australia. Radiocarbon 29, 311-317. LIN C., MELVILLE M . D . , WHITE I., WILSON B . P. & Y u D .

1998.

Sedimentary environment and acid potential of the pyritic sediments in a sand barrier estuarine system, eastern Australia. Geochemical Journal 32, 91-101.

NANSON G . C., EAST T. J. & ROBERTS R. G . 1 9 9 3 . Q u a t e r n a r y stratigra-

phy, geochronology and evolution of the Magela Creek catchment in the monsoon tropics of northern Australia. Sedimentary Geology 8 3 , 277-302.

O'BRIEN M. P. 1980. Comments on tidal entrances on sandy coasts. In: 17th International Conference on Coastal Engineering, Sydney 1980, pp. 242-243. American Society of Civil Engineers. RHODES E. G. 1980. Modes of coastal progradation, Gulf of Carpentaria, Australia. PhD thesis, Australian National University, Canberra (unpubl.). RHODES E. G. 1982. Depositional model for a chenier plain, Gulf of Carpentaria, Australia. Sedimentology 29, 201-221. ROY P. S. 1984. New South Wales estuaries: their origin and evolution. In: Thom B. G. ed. Coastal Geomorphology in Australia, pp. 99-121. Academic Press, Sydney. ROY P. S . , COWELL P. J., FERLAND M . A. & THOM B . G . 1 9 9 4 . W a v e d o m -

inated coasts. In: Carter R. W. G. & Woodroffe C. D. eds. Coastal Evolution, pp. 121186. Cambridge University Press, Cambridge.

SAMMUT J., WHITE I. & MELVILLE M . D . 1 9 9 6 . Acidification of a n e s t u -

arine tributary due to drainage of acid sulphate soils. Marine and Freshwater Research 45, 669-684. STUIVER M. & BECKER B. 1993. High precision decadal calibration of the radiocarbon time-scale, AD 1950-6000 BC. Radiocarbon 35, 35-65.

TACPSM (Task Force Committee for Preparation of Sediment Manual) 1971. Sediment transportation mechanics: H. Sediment discharge formulas. Journal of the Hydraulics Division, ASCE 97, 523-567. VERTESSY R. 1990. Morphodynamics of macrotidal rivers in far northern Australia. PhD thesis, Australian National University, Canberra (unpubl.). WANG P. 1990. Distribution of foraminifera in estuarine deposits: a

314

J. Chappell

comparison between Asia, Europe and Australia. In: Saito T. & Ishizaki K. eds. The Centenary of Japanese Micropaleontology, pp. 71-83. Tohoku University, Sendai. WANG P. & MURRAY J. W . 1 9 8 3 . T h e use of Foraminifera a s indicators

of tidal effects in estuarine deposits. Marine Geology 51, 239-250

WIGLEY T. M . L. & RAPER S. C. B. 1992. Implications for climate and sea

level of revised IPCC emissions scenarios. Nature 357, 293-300. WOLANSKI E. 1986. An evaporation-driven salinity-maximum zone in Australian tropical estuaries. Estuarine Coastal and Shelf Science 22, 415-424. WOLANSKI E. 1995: Transport of sediment in mangrove swamps. Hydrobiologia 295, 31-42. WOLANSKI E. & CHAPPELL J. 1996. The response of tropical Australian

estuaries to a sea level rise. Journal of Marine Systems 7, 267-279.

WOLANSKI E . , CHAPPELL J . , RIDD P. & VERTESSY R . 1 9 8 8 . F l u i d i s a t i o n o f

mud in estuaries. Journal of Geophysical Research 93, 2351-2361.

WOODROFFE C. D. & CHAPPELL J. 1 9 9 3 . H o l o c e n e emergence a n d evo-

lution of the McArthur River delta, southeastern Gulf of Carpentaria, Australia, Sedimentary Geology 83, 303-318.

WOODROFFE C . D . , CHAPPELL J . M . A . , THOM B . G . & WALLENSKY E . 1 9 8 6 .

Geomorphological Dynamics and Evolution of the South Alligator Tidal River and Plains Northern Territory. Australian National University, North Australia Research Unit, Mangrove Monograph 3. WOODROFFE C . D . , CHAPPELL J . , THOM B . G . & WALLENSKY E .

1989.

Depositional model of a macrotidal estuary and floodplain, South Alligator River, northern Australia. Sedimentology 36, 737-756.

WOODROFFE C. D. & GRIME D. 1999. Storm impact and evolution of a

mangrove-fringed chenier plain, Shoal Bay, Darwin, Australia. Marine Geology 159, 303-321.

WOODROFFE C. D . , MULRENNAN M . & CHAPPELL J. 1 9 9 3 . E s t u a r i n e infill

and coastal progradation, southern van Diemen Gulf, Northern Australia. Sedimentary Geology 83, 257-275.

WOODROFFE C . D , THOM B . G . & CHAPPELL J . 1 9 8 5 : D e v e l o p m e n t

widespread mangrove swamps in mid-Holocene times in northern Australia. Nature 317, 711-713.

WRIGHT L. D. & COLEMAN J. M. 1973. Variations in morphology of major

river deltas as functions of ocean wave and river discharge regimes. American Association of Petroleum Geologists Bulletin 57, 370-398.

Received 14 February 2000; accepted 11 November

2000

THEME 6 MARINE GEOSCIENCE

Geological Society of Australia Special Publication 21,317-328

CHAPTER 27—Environmental management of Torres Strait: a marine geologist's perspective P.T.HARRIS

HMHHHHHHHNHMNHfMtJMMwmmshMW I VV^M

^n^sSS^'^r ^js i'-*";".w*

-,J

- jy -

Antarctic CRC, Australian Geological Survey Organisation, GPO Box 252-80, Hobart, Tas. 7001 Australia. Environmental management of marine systems requires basic descriptive data comprising: an information system; an understanding of the physical, chemical and biological processes controlling sedimentation; and a facies model that can be used to predict the character and order of succession of sedimentary environments in a spatial and stratigraphic context. This paper describes these three levels of marine geological knowledge for an epicontinental seaway in northeastern Australia, Torres Strait and applies them to current management issues relating to the marine environment. In Torres Strait basic descriptive data are available for most areas, although there are significant gaps in the dataset, particularly in northern Torres Strait and along the southern coastline of Papua New Guinea. Our knowledge of sedimentary processes is based on only a few observations, and the processes controlling the supply of (suspended) terrigenous sediments to northern Torres Strait in particular are only poorly understood. Mining waste derived from the Fly River has not been detected yet in Torres Strait; however, terrigenous mud is deposited in the northern subtidal zone, particularly along the southern Papua New Guinea coastline between Daru and Saibai Island. Here, the deltaic mud facies extends from the Fly Delta into Torres Strait as a continuous belt; this mud is transported towards the west in low suspended sediment concentrations (<50 mg/L) as measured at two different locations, in Missionary Passage and in the channel north of Saibai island. Monitoring of this northern Torres Strait area for variations in heavy-metal levels and terrigenous sedimentation patterns is necessary, given the limitations of our present understanding of this complex sedimentary system. The proposed Torres Strait gas pipeline should avoid the deeply incised section of Missionary Passage, because unconsolidated Pleistocene clays may outcrop near the surface in this area. Erosion of the clay is likely under the strong currents, and the potential for environmental impact is high given the close proximity of coral reefs. An alternative route through a lower energy depositional area west of Missionary Passage is suggested. KEY WORDS: environmental management, Fly River, gas pipeline, mine wastes, sedimentary facies, sedimentology, Torres Strait, tidal currents, turbidity.

INTRODUCTION According to one specialist 'environmental management' is best viewed as 'a process that begins with goal setting and extends through the functions of information systems, research, planning, development, regulation and financing (MacNeill 1971 p. 5). Whether or not this definition is accepted, it is clear that sound environmental management practices must be based upon a broad foundation of knowledge, a pyramid-shaped structure (de Pablo et al 1994) having a descriptive-observational information system at the base and above which are linkages to specific environmental processes that can be modelled. Before the planning stage of the management process is reached (and well before the development stage) basic descriptive data must be collected and research must be conducted to understand the environmental processes and linkages to the earth's environmental systems (NASA 1988). It is perhaps one of the major stumbling blocks of marine environmental management in Australia that so much of the essential basic data is yet to be collected: things as elementary as water depth, sea-floor sediment type and dominant biota are quite simply not known for large parts of the Australian Exclusive Economic Zone.

From a marine geologist's perspective, the fundamental descriptive variables that are needed, at the 'information system' level, are the bed morphology, grain properties, sediment composition and biogenic constituents (Figure 27.1). These may be linked to specific physical, chemical and biological processes, which in various combinations are characteristic of different sedimentary environments. Facies and stratigraphic models are placed at the top of the pyramid (Figure 27.1) because they are based upon a synthesis of the descriptive data and of the processes that are specific to a given sedimentary environment (Walker & James 1992). To demonstrate how these levels of marine geological knowledge may be applied to the management of the marine environment, this paper examines a geographic location which has been the centre of controversy and environmental debate in recent years, Torres Strait. The first part of the paper will be an outline of the different levels of knowledge (Figure 27.1) available for Torres Strait, followed by a discussion of how this knowledge relates to two environmental management issues: the dispersal of mining waste and the installation of a submarine gas pipeline.

318

P.T.Harris Environmental Management Fades mode! Stratigraphy Physical

Bed morphology bedforms bed thickness outcrop erosion features level 1

Grain properties size sorting shape density porosity

Chemical Processes

Sediment composition carbonates trace metais other minerals

Biological Processes

Biogenic constituents calcareous siliceous (palaeontology)

Figure 27.1 Levels of knowledge related to the work required for different sedimentology studies (after Harris 1995b). Studies may be mainly descriptive (level 1), they may examine one or more sedimentary processes (level 2), or they may use this information to derive a facies (spatial) or temporal (stratigraphic) model (level 3). Environmental management requires knowledge derived from all three levels.

see a pipeline laid across Torres Strait, probably along one of the routes shown in Figure 27.1 (NSR 1997). The challenge for the management of this environment is to apply the existing marine geological information on sediment and heavy-metal distributions and their natural transport pathways in order to assess potential environmental impacts. Information on the physical processes controlling sedimentation is needed for planning the installation of the proposed gas pipeline in a safe and cost-effective manner. Hence, the key marine geological questions are: (i) what are the main dispersal pathways and depositional sites for sediments derived from the Fly River in Torres Strait; and (ii) what are the main physical processes controlling sediment erosion, transport and deposition in Torres Strait? The available information on terrigenous sediment input and dispersal in Torres Strait will be reviewed with the aim of addressing these questions.

Level 1: Torres Strait marine geological database CASE STUDY: TORRES STRAIT

BATHYMETRY A N D CURRENTS

Torres Strait is located at the northern end of the Great Barrier Reef in northeastern Australia (Figure 27.2). The strait is a major shipping seaway and it contains valuable fisheries resources (Williams & Staples 1990). It is at once the oceanographic boundary between the Coral Sea and Gulf of Carpentaria, a biological barrier correlating with the northern limit of the Great Barrier Reef, a political boundary between Australia, Papua New Guinea and Indonesia (Irian Jaya) and a geological mixing zone of terrigenous and calcareous sediments derived from various sources. Of particular importance is the local community's perceptions of environmental contamination of the strait resulting from mining activities in Papua New Guinea (Lawrence & Cansfield-Smith 1990). Further, a new gas pipeline project is currently under way, which (if fully implemented) would

Bathymetric surveys have concentrated on the main shipping channels of Torres Strait which are well charted, but in other areas such as along the southern coast of Papua New Guinea, little or no data have been collected. The available data indicate a shallow, low-relief seabed in the eastern Gulf of Carpentaria, Torres Strait and Great North East Channel areas which have typical depths of about 15-25 m; Torres Strait has a sill depth of 12 m (Harris 1988) (Figure 27.2). To the north, the steep prodelta slope of the Fly Delta extends onto the adjacent continental shelf and is delineated by the closely spaced 10-30 m isobaths. Submarine channels (Figure 27.2) include a southeast-trending valley up to 120 m deep, which dissects the shelf at the northern end of the Great Barrier Reef (Harris et al 1996). Channels with localised 'deeps' of up to 60 m extend through Missionary

Figure 27.2 Bathymetry of Torres Strait, based on published nautical charts and unpublished data. The area where peak tidal current speeds exceed 0.8 m/s is shown by the horizontal shading (from Harris 1995a). The locations of the proposed routes of the Torres Strait gas pipeline and of seismic sections shown in Figure 27.12 are indicated.

Environmental management Torres Strait

North

319

Current meter | Stn. 21

Oriomo R.

Bristow

<5m

>20m

Saibai Island

Figure 27.3 Bathymetry of northern Torres Strait (from Harris et al. 1991) showing locations of seismic survey track lines, sediment sampling stations and current meter mooring sites as referred to in the text. Inset, shows a bathymetric profile across Missionary Pass, with local deep incisions and the location of the current meter station 21. The locations of seismic sections shown in Figure 27.6 are indicated. Passage and continue westwards for over 100 km north of Saibai and Boigu Islands near to the coast of Papua New Guinea. These channels were mapped by Harris (1994) and shown to be mainly of relict, fluvial origin, although scour by tidal currents has caused some local overdeepening (e.g. in Missionary Passage: Figure 27.3). The strong tidal currents which characterise nearly all of Torres Strait are forced by two separate and dissimilar tidal systems, one located in the Gulf of Carpentaria and the other in the Coral Sea (Bode & Mason 1995). The tidal amplitude is reduced by coral reef complexes in the strait, and the maximum spring tidal range is about 3.5 m at Daru (Figure 27.3). Tidal currents, however, are accelerated by hydraulic effects in narrow reef passes and in the constrictions between islands; this is particularly important for the Torres Islands in the south (where a maximum tidal current speed of 4 m/s is given on nautical charts) and for the narrow passes between Saibai and Boigu Islands and mainland Papua New Guinea in the north. In general, a zone of relatively strong (>0.8 m/s maximum speed) tidal flow extends north from Cape York over the Warrior Reefs and the Torres Islands to the Papua New Guinea coast (Harris 1995a; Bode & Mason 1995) (Figure 27.2). A zone of weaker tidal flow is located in central Torres Strait (Figure 27.2). Superimposed on this tidal regime are less important ocean currents, storm surges and wind-driven circulation patterns (Amin 1978; Wolanski et al 1988). Wind-driven currents reverse from a westward flow during the southeast

trade wind season (April to November) to an eastward flow during the northwesterly monsoon season (December to March). Oceanic currents are manifest as a small (-0.2 m) difference in sea-level across Torres Strait. Although the magnitude of wind-driven currents is typically less than 0.2 m/s, it is enough to induce a seasonal reversal in tidally averaged bedload transport vectors, which causes 4 m-high bedforms to reverse their asymmetry and migration direction (Harris 1989, 1991). SURFICIAL SEDIMENT DISTRIBUTION The nature of the surficial sediments varies as a function both of the proximity of fluvial sediment sources (the Fly River) and of tidal current regime. The strong tidal currents described above have produced areally extensive bedforms which are visible in aerial photographs, LANDSAT images and sidescan sonar images (Harris 1988, 1995a) (Figure 27.4). The occurrence of bedforms implies that the bed sediments are regularly mobilised, and that the associated surficial sediments comprise well-sorted sands and gravels, as compared with the surrounding sediments. The most useful basic descriptors of surficial sediment are grainsize distribution and percentage content of calcium carbonate. Grainsize distribution may be related to the relative current energy and the relative position of the sediment along the transport path, from source to depositional environments. A simple, yet effective measure is mud content, defined as the percentage by dry weight of grains

320

P. T. Harris

Figure 27.4 Part of a processed LANDSAT image (after Rennie et al. 1997) showing the distribution of subtidal bedforms in northwestern Torres Strait. Although some of these dunes were observed in aerial photographs described by Harris (1988) their true extent was not known prior to the release of this image. <63pm is diameter. A map of mud in surficial sediments of Torres Strait (Figure 27.5a) shows two areas of relatively high mud content: one forms a belt of >80% mud trending across the front of the Fly Delta; the other (>60% mud) is in the eastern section of Torres Strait among the numerous patch reefs. These are both relatively low-energy depositional areas. The areas of low mud content (<20% mud: Figure 27.5a) are mainly sand and gravel, and the better sorted areas (<10% mud) are associated with bedforms (cf. Figures 27.4 and 27.5a) (Harris 1988). The calcium carbonate content of sediments is derived primarily from benthic foraminifers, supplemented with bryozoans, molluscs, the alga Halimeda and small amounts of corals (Harris 1988). The percentage carbonate content map of Torres Strait (Figure 27.5b) demonstrates that the sands and muds deposited in the Fly Delta are low carbonate (terrigenous) sediments, whereas most of the Torres Strait is high (>80%) in carbonate. A large field of quartzose

sandbanks is located in western Endeavour Strait near Cape York (Figures 27.2, 27.5b), which explains the low carbonate values here, a fact first noted by Maxwell (1968). These quartz sands were derived probably from the reworking of Pleistocene subaerial dunes that were submerged during the Holocene (Harris 1988). Finally, another zone of relatively low carbonate content occurs along the southern coast of Papua New Guinea and two, southward-pointing tongues of relatively low carbonate (60-80%) are in the Great North East Channel and in central Torres Strait (Figures 2, 5b).

SEISMIC PROFILING AND CORE DATA High-resolution (boomer) seismic surveys and coring programs in Torres Strait have shown that, for the most part, unconsolidated sediments form a thin layer, 1-2 m thick, which overlies cohesive Pleistocene clay and, close to reefs,

Environmental management Torres Strait

321

% Mud Content

0-10%

10<20% 20<40% 40<60%

i i i i 0 30km

60<80% >80%

Figure 27.5 Distribution of (a) mud and (b) carbonate in the surficial sediments of Torres Strait, based on analyses of 550 samples.

-0ms

Warrior Reefs

Missionary Pass

Figure 27.6 Examples of high-resolution boomer seismic profiles collected in 1991 during a cruise of the RV Sunbird (Harris et al 1991), using an ORE Boomer system operated at 350 Joules. The sections show examples of (a) tidal current eroded seabed in Missionary Pass with some bedded sediments near the Warrior Reefs; (b) fluvially incised surface draped by a 10-20 m-thick layer of Holocene sediments; (c) tidal current eroded channel located to the north of Saibai Island with flanking depositional areas, characterised by an incised surface draped by a 10-20 m-thick layer of Holocene sediments. See Figure 27.3 for locations of the seismic sections. Vertical scale in ms two-way travel time.

322

P. T. Harris

Figure 2 7 . 8 Surface water turbidity (mg/L) determined by filtration in (a) April 1990 and (b) September 1991. Circles indicate station locations. For April 1990, peak turbidity was recorded near the southern Warrior Reefs (28 mg/L) and at one station on the Fly Delta (35 mg/L). Note the turbidity in central Torres Strait, west of the Warrior Reefs, is consistently higher than that in the Great North East Channel. For September 1991, peak turbidity was recorded mainly inshore west of Saibai and north of Boigu Island, and also forming a plume extending from the Fly Delta and westward through Toro Pass.

cemented limestone (Harris 1994; Harris etal 1991, 1996). The clay and limestone are considered to have undergone subaerial exposure and were formed during the last glacial lowering of sea level 20 000 years ago (Harris et al 1992, 1996). In some areas,riverchannels have been incised into the older Pleistocene sediments, to be then partially or completely infilled with younger, Upper Pleistocene to Holocene sediments. These channels occur mainly within 20 km of the southern coastline of Papua New Guinea (Harris 1994). Inspection of seismic profiles collected in the constricted channels of northern Torres Strait and areas flanking the channels suggests a pattern of channel erosion and deposition of sediment in distal areas. For example, the seabed characterising Missionary Passage is rough and erosional in appearance, with outcrops of older sedimentary strata in many locations (Figure 27.6a). The northern margin of the Warrior Reefs appears to have a foundation of (unconsolidated?) sediments rather than reefal limestone judging from the seismic character of the strata (Figure 27.6a) and from one short piston core collected at Station 73 (Figures 27.3, and 27.7 on Plate 13). This core contained

an upper unit, 45 cm thick, of unconsolidated, olive-grey, calcareous muddy sand overlying a greyish-red montmorillonite clay, interpreted as a weathered Pleistocene soil (Harris et al 1992). Presumably, the clay (Figure 27.7 on Plate 13) is the upper part of a much thicker fluvio-estuarine/deltaic unit that is resolved in the seismic data (Figure 27.6a). Further west, the erosional surface is draped by Holocene (?) sediments which infill some incised channels (Harris 1994) (Figure 27.6b). These sediments were deposited probably during the post-glacial transgression (i.e. they are part of the transgressive systems tract) when the shelf was first flooded by the rising sea-level around 9000 years ago (Harris et al 1996). Subsequent erosion of the reef passes (such as Missionary Passage) also provided a local source which supplied sediment to depositional areas and produced the Highstand Systems Tract. In many cases, the relict fluvial channels remain unfilled or only partially filled with sediments (Harris 1994) (Figure 27.6b). Tidal currents accelerate in the narrow channel located between Saibai Island and the Papua New Guinea coastline, and the seabed here shows signs of extensive erosion, with flanking depositional areas (Figure 27.6c). The deposits lying above the erosional surface appear to be about 10 m thick, and shell hash from the base (248 cm) of a core collected at station 121 (Figure 27.3) gave an uncorrected Holocene radiocarbon age of 1540 ±340 a (Harris et al 1996). They are dark-grey to black, bioturbated, slightly shelly terrigenous muddy sands and sandy muds (Figure 27.7 on Plate 13). The clay fraction of these sediments is predominantly kaolinite and illite with minor chlorite and mica, which is similar to that presently deposited in the Fly Delta (Harris et al 1993) and which is distinct from the montmorillonite clays characterising the Pleistocene soils described above. Level 2: Sedimentary processes and deposits PHYSICAL PROCESSES Physical sedimentary processes were reviewed by Harris (1995a) in terms of bed-load and suspended-load sediment transport. The two processes are linked because an important source of turbidity is the tidal current reworking of calcareous gravelly sands that comprise mobile bedforms (Figure 27.3); reworking causes the soft carbonate grains to fracture and break down to form smaller silt-sized grains. These fine-grained products are transported in suspension away from the tidal scour zones (Harris et al 1995) i.e. areas of maximum tidal flow (Figure 27.2) and are deposited in the muddy, high-carbonate zones located in eastern and central Torres Strait (Figure 27.5a, b). Tidal currents combined with wind-generated waves give rise to a turbidity maximum in central Torres Strait (Harris & Baker 1991) which is detectable in air photographs and satellite images. Measurements of suspended-sediment concentrations in surface waters from April 1990 (Figure 27.8a) show that the highest levels occur in areas adjacent to constricted reef passes. Turbidity levels are generally higher in central Torres Strait than in the Great North East Channel (Figure 27.8a). The pattern for September 1991 (Figure 27.8b)

Environmental management Torres Strait

323

Spring Tide

•g i— ZJ

Days in February, 1989 o

| 60 8 . 2® CO

Days, September, 1991 Figure 2 7 . 9 Time-series plots showing (a) sea-level at Daru, current speed and surface water turbidity over a 24 hour period representing spring tides (from Station 21, Missionary Pass); (b) sea-level at Daru and surface water turbidity over a 24 hour period representing neap tides (from Station 21, Missionary Pass); (c) current speed and direction measured at 100 cm above the bed over a 5 day period (from Station 21, Missionary Pass); and (d) current speed and turbidity measured at 100 cm above the bed over an 8 day period (from Station 62, Saibai Island). Station 21 was in 18 m and Station 62 was in 10 m water depth. See Figure 27.3 for station locations.

324

P. T. Harris

Figure 2 7 . 1 0 Distribution of sedimentary facies in Torres Strait (modified from Harris 1995a to include the bedform distribution information shown in Figure 27.4).

Flocculation, semidiurnal and neap-spring variations in current speed and the associated lag time between deposition and erosion can all have an effect on surface water turbidity and suspended-load transport. Contemporaneous current and turbidity observations suggest that in Torres Strait the path followed by a suspended sediment particle is determined mainly by the direction of residual (tidally averaged) flow. For example, suspended-sediment concentration and current speed measurements from Missionary Passage (Figure 27.9a-c) show that turbidity rises and falls over 24 hours, but the pattern is different for spring and neap tides. During spring tides, the overall concentrations are higher (20-30 mg^L) than in neaps (<10 mg/L). For both neap and spring tides, there are peaks and troughs in the turbidity curve which coincide with the current speed curve (Figure 27.9a-c). A similar pattern is observed at station 62 north of Saibai Island (Figure 27.9d). Again, turbidity rises and falls with the tidal cycle and shows about a two-fold increase from neap to spring tides. Multiplying the instantaneous suspended-sediment concentration by the current-velocity vector and summing the vectors shows that the net suspended-sediment transport is towards the west for both the Missionary Passage and Saibai Island datasets and hence are flood-dominated. Therefore, the westward flowing, wind-driven residual current that dominates Torres Strait during the southeast trade-wind season appears to control the local advection of suspended sediments along the southern coastline of Papua New Guinea, between Daru and Saibai Island. GEOCHEMICAL AND BIOLOGICAL PROCESSES

% Terrigenous Mud

Figure 2 7 . 1 1 Scatter plot of copper concentration vs terrigenous mud content of surficial sediments (see also Baker 1990). Terrigenous mud refers to inorganic grains having a size <63pm.

shows a similar range of concentrations, with higher levels associated with the Fly Delta and southern coast of Papua New Guinea. Measurements of suspended-sediment concentration in the surface 1 m of water at 102 stations in Torres Strait in April 1990 (the northwest monsoon season) had a mean of 9.14 mg/L and ranged between 2.2 and 36.6 mg/L whereas measurements at 78 stations in September 1991 (southeast trade-wind season) had a mean of 5.4 mg/L and ranged between 0.90 and 35.2 mg/L (not including the Fly Delta stations). Although the two datasets are not directly comparable since different stations were occupied during the two surveys, the data probably represent the typical range of surface water turbidity occurring in Torres Strait.

The effects of geochemical and biological processes on sedimentation in Torres Strait have not received as much attention as physical processes. Keene and Harris (1995) showed that aragonite cements were formed in basal dune (bedform) deposits located in the southern Warrior Reefs. These cements have formed less than 500 years ago and are thought to be precipitated as water is pumped through the porous bedform deposits by the strong tidal currents (Keene & Harris 1995). Biological processes play an important role in the bioturbation of sediments in the lower energy (mud content >20%) sections of eastern Torres Strait. Bioturbation obliterates primary sedimentary structures, such as ripple and dune cross-bedding, and mixes surficial sediments 1 m or more below the seabed (Tudhope & Scoffin 1984). Bioerosion results in the breakdown of larger sand- and gravel-sized grains into smaller silt-sized grains; bioerosion is identified as an important source of the calcareous muds found in the depositional areas of eastern and central Torres Strait (Cole et al 1995). Recently, Ayukai and Wolanski (1997) reported some biological controls on the removal of heavy metals from the Fly River plume in the Great North East Channel area. Level 3: A facies model of Torres Strait The available information on bedform distribution and surficial sediments (based on 500 surface grab samples and 50 cores) was used by Harris (1995a) to define four sedimentary facies: (i) deltaic mud facies; (ii) high-energy transi-

Environmental management Torres Strait

Pteistoceno day fluvial-deltaic sediments

Holocene carbonate sediment drape

North

325

Relict fluvial channel partially infilled with Holocene sediment

t km

mmJ

Figure 27.12 Examples of high-resolution boomer seismic profiles collected in 1990 during a cruise of the HMAS Cook (Harris et al 1990) using an ORE Boomer system operated at 350 Joules. The sections show examples of (a) Pleistocene fluvial deltaic cut-and-fill sequences in the Great North East Channel, with Holocene carbonate muddy gravelly sand fill of up to 10 m locally [Core 17VC12 confirms the lithology and age of the units (Harris et al 1992)]; and (b) highly reflective surface interpreted as cemented limestone draped with Holocene carbonate sediments that thicken with increasing distance from Basilisk Passage. See Figure 27.2 for locations of the seismic sections. Vertical scale in ms two-way travel time.

tional facies; (iii) high-energy carbonate sand facies; and (iv) muddy carbonate facies. The spatial distribution of these facies are summarised in Figure 27.10 and they are defined below. The deltaic mud facies (Figure 27.10) comprises laminated to massively bedded terrigenous, deltaic muds, containing generally <10% fine sand. The bedding is typically 2-5 cm in thickness, equivalent to the annual sedimentation rate as determined by 2 1 0 Pb dating, interpreted by Harris et al (1993) as annual deposits (varves). The varves are formed in response to the seasonal variation in surface wave energy; large waves produced by the southeasterly trade winds cause winnowing of sediments, leaving a sandy layer which is draped by mud during the lower energy monsoon season (Harris et al 1993; Baker et al 1995). The general trend is for mud content to exceed 80% within an isobath-parallel belt which extends generally between 20 and 60 m water depth (i.e. the pro-deltaic environment: Figure 27.5a). These sediments are prograding southeastwards at a rate of about 6 m/y into the Gulf of Papua (Harris et al 1993); the seaward edge of the prograding delta lies within 10 km of Bramble Cay, the most northern cay in the eastern patch-reef province. This facies is interpreted to extend as far westwards as Saibai Island, along the southern coast of Papua New Guinea, on the bases of carbonate content (Figure 27.5b), clay mineralogy and the dispersal processes described above. The high-energy, transitional sand facies (Figure 27.10) is 'transitional' in the sense that these sediments grade from terrigenous-deltaic to carbonate-rich, reefal sediments

which are characteristic of central and eastern Torres Strait. Sediments in the transitional high-energy facies range from muddy-gravelly sands to well-sorted sands, and contain 20-80% carbonate and <20% mud (Figure 27.5a, b). Coarse-grained sands and gravels locally form mobile dunes and erosional sand-ribbons. This facies occurs as an east-west-trending belt along the southern coastline of Papua New Guinea and also in association with the quartzose sands forming a tidal sandbank complex in western Endeavour Strait. Sediments in the high-energy carbonate facies (Figure 27.10) contain >80% carbonate and <20% mud. In locations where tidal currents attain a local maximum (Figure 27.2), the seabed is typically scoured clear of any unconsolidated sediment leaving a lag gravel or limestone pavement behind. With increasing distance from the scoured channels, sediment of decreasing grainsize is deposited. Beyond the scour zone, sand/gravel ribbons give way to dunes and, at greater distances from the reef passages, current strengths are diminished enough to allow fine silt- and clay-sized particles to settle out; this pattern is typical of tidally dominated continental shelves (Harris et al 1995). In areas of weaker tidal currents, the muddy carbonate facies is dominant (Figure 27.10). Sediments in this facies are comprised of over 80% carbonate derived from the same types of organisms described above, but sediments in this facies contain >20% mud. Up to 80% of this mud is detrital carbonate, derived from the breakdown of bioclastic sand- and gravel-sized particles in the high-energy facies, which are then exported to the muddy carbonate

326

P. T. Harris

fades (Harris & Baker 1991; Cole et al 1995). Although mud is accumulating in the localised patch of muddy carbonate sediment in the centre of Torres Strait (located between the Warrior Reefs and the Torres Strait Islands) seismic profiling and coring have shown only 1-2 m of Holocene sediment is deposited in this area.

Management issues TRANSPORT PATHS OF MINING WASTE IN TORRES STRAIT

The major source of terrigenous sediments to Torres Strait is the Fly River (Figure 27.2) (see also Harris 1988). Prior to European settlement, the Fly River had a total sediment load of approximately 85 Mt/y (Harris et al 1993). This was a tremendous volume of sediment, when one considers that all of the Queensland rivers which empty into the Great Barrier Reef lagoon supply only about 18 Mt/y (Harris et al 1993). With the opening of the Ok Tedi gold and copper mine in the early 1980s, the sediment load of the Fly River was greatly increased; by 1992 it was about 125 MT/y (Harris et al 1993, 1996). The environmental impact of this sediment and its associated heavy metals on the Torres Strait marine environment has been the centre of controversy and much debate (Dent 1985; Lawrence & Cansfield-Smith 1990). Schneider (1989) reported that some heavy-metal concentrations (specifically cadmium) in commercial prawns taken from Torres Strait were above natural background levels and comparable to those found in some polluted areas. This report led to the Federal Government funding the Torres Strait Baseline Study, the results of which have been published by the Great Barrier Reef Marine Park Authority (Dight & Gladstone 1993; Gladstone 1996; Evans-Illidge 1997). These later studies confirmed Schneider's (1989) observation of high cadmium levels in commercial prawns, but EvansIllidge (1997) concluded that this is a natural phenomenon, characteristic of carbonate depositional environments and unrelated to pollution from the Fly River. Regional studies of the concentration of copper in surfical sediments, aimed specifically at determining the impact of mining activities, have identified a trend of higher copper, lead and zinc levels in Fly Delta sediments compared with lower levels in sediments from the adjacent Torres Strait (Baker & Harris 1991). The metal concentrations are strongly correlated (R2 = 0.88; n = 110) with terrigenous mud content in surficial sediments (Figure 27.11); hence any information on the processes controlling the dispersal and deposition of terrigenous mud is also relevant to the dispersal and deposition of the associated heavy metals. However, as recently as 1994 sediment sampling and coring in the distributary channels of the Fly Delta had not detected copper concentrations significantly higher than background (E. K. Baker pers comm. 1997) and so it appears that (as of 1994) no detectable increase heavy-metal levels has effected the Fly Delta or northern Torres Strait. The increased levels of terrigenous mud supplied by the Fly River will probably be dispersed by deltaic mud sedimentation in the areas shown in Figure 27.10. Harris et al (1993) estimated that less than 2% of the pre-mining sediment load of the Fly was deposited in Torres Strait, based on the assumption that all of the terrigenous mud presently accumulating in Torres Strait is sourced from the Fly.

Outcrops of eroded Pleistocene fluvial/deltaic sediments in Missionary Passage and the channel north of Saibai Island (Figure 27.6a, c) are also potential sources of terrigenous mud, but these probably supply montmorillonite-rich clays that are distinct from the predominantly kaolinite and illite clays supplied by the Fly (and which are accumulating at core sites 62 and 121: Figure 27.7 on Plate 13). Given the complexity of the sediment-dispersal systems in the region, it is clear that monitoring of terrigenous sediment and heavy-metal distributions needs to be continued. Such monitoring should focus on the areas most likely to be affected by changes in terrigenous sedimentation patterns. These areas include the southern margin of the Deltaic Mud facies in Figure 27.10, extending from the northern part of the Great North East Channel and eastwards to Daru and Saibai Island. With limited resources, spreading the monitoring effort over a wider area and thus spanning additional sedimentary facies (Figure 27.10) is unlikely to improve our ability to detect environmental changes (Harris 1995b). IMPLICATIONS OF THE PROPOSED GAS PIPELINE ACROSS TORRES STRAIT

The proposed gas pipeline could cause local current acceleration and subsequent seabed erosion. After the initial disruption caused by laying of the pipeline, any part of the pipeline which protrudes above the level of surrounding seabed will cause currents to be accelerated locally, from a background value Ux to a greater value U 2 , given by: U2=[U1z1]/z2

(1)

where zl and z 2 are the water depths away from and directly above the pipeline, respectively. Thus a current in 7 m water depth with initial speed 0.5 m/s will be accelerated to 0.58 m/s if the pipe is 1 m high. This has a larger effect on sediment mobility than might be expected, because the relationship between grainsize and threshold current speed is non-linear (Miller et al 1977). Equation (1) also shows that, for a given pipe diameter, acceleration will increase with decreasing water depth. The effects of erosion will also be exacerbated in locations like Missionary Passage that are already undergoing natural erosional processes (Figure 27.6a). In such locations, the effects of dredging and laying of a pipeline may pose a threat to the surrounding reef areas. This is because the fine-grained Pleistocene clays are covered and protected from erosive currents by a thin, coarse layer of calcareous, reefal debris (Figure 27.7 on Plate 13). Removal of this coarse layer over the length of the pipeline could cause widespread erosion of the soft clay, which could result in elevated turbidities and increased terrigenous sedimentation on adjacent reef tracts for some time (i.e. until a protective gravel cover is deposited). An obvious alternative is not to lay the pipeline through the narrow section of Missionary Passage, where tidal currents reach a peak velocity, and locate it westwards along the Papua New Guinea coast before turning south. In this way the pipeline would bypass the high-current-energy area and instead would cross through the more tranquil depositional zone west of Missionary Passage (i.e. at the

Environmental management Torres Strait location of the seismic profile shown in Figure 27.6b) where seabed erosion is not presently occurring. According to NSR (1997), avoiding Missionary Passage altogether and routing the pipeline southwards along the Great North East Channel is the preferred option (Figure 27.2). Current speeds tend to be lower and the water depth is greater in the northern part of the Great North East Channel (Figure 27.2) and hence current acceleration is not as great and seabed erosion is less likely to be a problem. However, in Basilisk Passage, where the pipeline route crosses the Warrior Reef complex, spring tidal currents exceed 80 cm/s and available bathymetric data (NSR 1997) shows an erosional topography, with >40 m-deep incised channels, similar to Missionary Passage. Seismic profiles and sediment cores are not available for Basilisk Passage itself. However, along the pipeline route in the Great North East Channel to the north, Pleistocene clay is seen to outcrop, or to be thinly draped with Holocene carbonate sediments (Figure 27.12a). This area is also used by prawn trawlers (NSR 1997), so burial of the pipeline may be necessary (the Missionary Passage route has the advantage that it avoids the prawn trawler grounds). Closer to Basilisk Passage, burial of the gas pipeline will be complicated by the occurrence of what is interpreted as a cemented limestone pavement (Figure 27.12b). The exact character of the seabed in Basilisk Passage is unknown and collection of further seismic and sediment core data is needed to determine whether Pleistocene clay deposits occur here. Exposure of the clay in Basilisk Passage should be avoided because seabed erosion will result in elevated turbidities which the adjacent coral reefs may not be able to tolerate. Whichever route is eventually chosen, it is clear that the marine geology of Torres Strait, including the Upper Pleistocene-Holocene stratigraphy, will have to be carefully considered before the gas pipeline development can safely proceed.

CONCLUSIONS Environmental management of marine systems requires a pyramid of geological knowledge at three levels; (i) basic descriptive data, comprising an information system; (ii) an understanding of the physical, chemical and biological processes controlling sedimentation; and (iii) a facies model which can be used to predict the character and order of succession of sedimentary environments in a spatial and stratigraphic context. In Torres Strait, the pyramid of knowledge becomes increasingly less reliable moving from the first to the third level. Level 1 (information system) is available, although there are significant gaps in the dataset, particularly in northern Torres Strait and along the southern Papua New Guinea coastline. Level 2 (knowledge of sedimentary processes) is only based on few observations; in particular the processes controlling the supply of terrigenous sediments to northern Torres Strait are only poorly understood. Hence, the third level (facies model) must be viewed as a 'working modeP which will require refinements as new data become available. Threats to the Torres Strait marine environment from mining waste derived from the Fly River mainly involve the

327

northern section, particularly along the southern Papua New Guinea coastline between Daru and Saibai Island. Here, the deltaic mud facies (Figure 27.10) extends from the Fly Delta into Torres Strait as a continuous belt; this mud is transported towards the west in low concentrations (<50 mg/L) of suspended sediment as measured in Missionary Passage and in the channel north of Saibai island. No increase in copper levels above background has yet been detected in the Fly Delta; nevertheless, monitoring of this northern Torres Strait area is warranted given the limitations of our present understanding of this complex sedimentary system. Emplacement of the proposed Torres Strait pipeline may cause local erosion and elevated turbidity along the pipeline route. The deeply incised section of Missionary Passage (and possibly Basilisk Passage) should be avoided because unconsolidated Pleistocene clays may outcrop near the surface in this area; erosion of the clay is likely to occur under the strong currents and the potential for environmental impact on the nearby coral reefs is high. An alternative is to change the route to pass through a lower energy depositional area west of Missionary Passage. Further seismic profiling and coring work is needed for the Basilisk Passage route, to determine whether or not erosion of the clay is a possibility in this area.

ACKNOWLEDGMENTS The author is grateful for the financial support of the Australian Defence Science and Technology Organisation and of the Australian Research Council (Ref. No. A39131170) which provided for the collection of the data. Thanks to Howard Stagg and Neville Exon for their reviews of an earlier version of this paper, and to Vic McGraw (South Pacific Pipelines Inc.) for providing information on the proposed gas pipeline. Published with the permission of the Executive Director, Australian Geological Survey Organisation.

REFERENCES AMIN M. 1978. A statistical analysis of storm surges in Torres Strait. Australian Journal of Marine and Freshwater Research 29, 479-496. AYUKAI T. & WOLANSKI E. 1997. Importance of biologically mediated removal of fine sediments from the Fly River plume, Papua New Guinea. Estuarine, Coastal and Shelf Science 44, 629-639. BAKER E. K. 1990. Origin, distribution and toxicological implications of suspended particulate matter in Torres Shelf seawater. MAppSc thesis, University of Technology, Sydney (unpubl.). BAKER E. K. & HARRIS P. T. 1991. Copper, lead and zinc in the sediments of the Fly River Delta and Torres Strait. Marine Pollution Bulletin 22, 614-618. BAKER E . K . , HARRIS P. T., SHORT S . A. & KEENE J. B . 1 9 9 5 . P a t t e r n s o f

sedimentation in the Fly River Delta. In: Flemming B. W. & Bartholoma A. eds. Tidal Signatures in Modern and Ancient Sediments, pp. 193-211. International Association of Sedimentologists Special Publication 24. BODE L. & MASON L. B. 1995. Tidal modelling in Torres Strait and the Gulf of Papua. In: Bellwood O., H. Choat H. & Saxena N. eds. Recent Advances in Marine Science and Technology '94 Townsville, pp. 55-66. James Cook University, Townsville.

COLE A. R . , HARRIS P. T. & KEENE J. B . 1 9 9 5 . F o r a m i n i f e r s a s f a c i e s

indicators in tropical, subtidal environments: Torres Strait-Fly River Delta, southern Papua New Guinea. In: Flemming B. W. &

328

P. T. Harris Bartholoma A. eds. Tidal Signatures in Modern and Ancient Sediments, pp. 212-224. International Association of Sedimentologists Special Publication 24.

DE PABLO C . L . , DE AGAR P. M . , BARTUREN R . , NICOLAS J . P. & PINEDA F.

D. 1994. Design of an information system for environmental planning and management ( S I P A ) . Journal of Environmental Management 40, 231-243. DENT J . 1 9 8 5 . Possible impact of Ok Tedi mining wastes on the waters of Torres Strait. In: Haines A. K., Williams G. C. & Coats D. eds. Torres Strait Fisheries Seminar, Port Moresby, Papua New Guinea, February 1985, pp. 2 9 3 - 2 9 8 . Australian Government Printing Service, Canberra. DIGHT I. J . & GLADSTONE W. 1993. Torres Strait baseline study: pilot study and final report. Great Barrier Reef Marine Park Authority Research Publication 29. EVANS-ILLIDGE E . 1997. Heavy metals in commercial prawns and crayfish species in the Torres Strait. Great Barrier Reef Marine Park Authority Report Series 5b. GLADSTONE W. 1996. Trace metals in sediments, indicator organisms and traditional seafoods of the Torres Strait. Great Barrier Reef Marine Park Authority Report Series 5a. HARRIS P. T. 1 9 8 8 . Sediments, bedforms and bedload transport pathways on the continental shelf adjacent to Torres Strait, Australia-Papua New Guinea. Continental Shelf Research 8 , 9 7 9 - 1 0 0 3 . HARRIS P. T. 1989. Sandwave movement under tidal and wind-driven currents in a shallow marine environment: Adolphus Channel, northeastern Australia. Continental Shelf Research 9, 981-1002. HARRIS P. T. 1991. Reversal of subtidal dune asymmetries caused by seasonally reversing wind-driven currents in Torres Strait, northeastern Australia. Continental Shelf Research 11, 655-662. HARRIS P. T. 1 9 9 4 . Incised valleys and backstepping deltaic deposits in a foreland-basin setting, Torres Strait and Gulf of Papua, Australia. In: Dalrymple R. W., Boyd R. & Zaitlin B. eds. Incised Valley Systems: Origin and Sedimentary Sequences, pp. 9 7 - 1 0 8 . Society of Economic Paleontologists and Mineralogists Special Publication 51. HARRIS P. T. 1995a. Muddy waters: the physical sedimentology of Torres Strait. In: Bellwood O., H. Choat H. & Saxena N. eds. Recent Advances in Marine Science and Technology '94 Townsville, pp. 149-160. James Cook University, Townsville. HARRIS P. T. 1995b Interpretation of marine sediment samples: temporal and stratigraphic issues. In: Ocean Rescue 2000 Workshop on Marine Data Standards, CSIRO Marine Laboratories, Hobart, 5-6 December, 1995 (http://www.environment.gov.au/marine/ manuals_reports/standards/geoscience/geoscience.html). HARRIS P. T. & BAKER E. K. 1991. The nature of sediments forming the Torres Strait turbidity maximum. Australian Journal of Earth Sciences 38, 65-78. HARRIS P. T., BAKER E. K. & COLE A. R. 1990. Sandwave movement, currents and sedimentation in Torres Strait: results obtained during a cruise of HMAS Cook in April 1990. University of Sydney, Ocean Sciences Institute Report 43. HARRIS P. T., BAKER E. K. & COLE A. R. 1992. Late Quaternary sedimentation at the Fly River-Great Barrier Reef junction (northeastern Australia). In: Proceedings of the 7th International Coral Reef Symposium, Guam, pp. 1159-1168. University of Guam, Guam. HARRIS P. T., BAKER E. K., COLE A. R. & KEENE J. B . 1991. Final Report: sandwave movement, currents and sedimentation in Torres Strait Ocean. Sciences Institute, University of Sydney, Report 47. HARRIS P. T., BAKER E. K., COLE A. R. & SHORT S . A. 1993. Preliminary

study of sedimentation in the tidally dominated Fly River Delta, Gulf of Papua. Continental Shelf Research 13, 441-472. HARRIS P. T., PATTIARATCHI C . B . , COLLINS M . B . & DALRYMPLE R .

HARRIS

W.,

1995. What is a bedload parting? In: Flemming B. W. & Bartholoma A. eds. Tidal Signatures in Modem and Ancient Sediments, pp. 1-18. International Association of Sedimentologists Special Publication 24. P. T . ,

PATTIARATCHI C .

B.,

KEENE J .

B.,

DALRYMPLE

GARDNER J . , BAKER E . K . , COLE A . R . , MITCHELL D . , G I B B S P.

1996. Late Quaternary deltaic and carbonate sedimentation in the Gulf of Papua foreland basin: response to sea-level change. Journal of Sedimentary Research 66, 801-819. LAWRENCE D. & CANSFIELD-SMITH T. (Editors) 1990. Sustainable Development for Traditional Inhabitants of the Torres Strait Region. Proceedings of the Torres Strait Baseline Study Conference, November 1990. Great Barrier Reef Marine Park Authority, Townsville, Workshop Series 16 KEENE J . B. & HARRIS P. T. 1 9 9 5 . Submarine cementation in tide generated bioclastic sand dunes: epicontinental seaway, Torres Strait, northeast Australia. In: Flemming B. W. & Bartholoma A. eds. Tidal Signatures in Modern and Ancient Sediments, pp. 2 2 5 - 2 3 6 . International Association of Sedimentologists Special Publication 24. MAXWELL W . G. H . 1968. Atlas of the Great Barrier Reef Elsevier, Amsterdam. MACNEILL J. W. 1971. Environmental Management. Information Canada, Ottawa. MILLER M . C . , MCCAVE I. N . & KOMAR P. D . 1 9 7 7 . Threshold of sediment motion under unidirectional currents. Sedimentology 24, SCHROEDER W . W .

507-527.

NASA 1988. Earth System Science: a Closer View. National Aeronautics and Space Administration, Washington, DC. NSR 1997. Papua New Guinea-Queensland Gas Project, Progress Report on Interim Alignment for Special Area No. 2, Torres Strait. NSR Environmental Consultants Pty Ltd. report to Southern Pacific Pipeline Company Pty Ltd, July 1997. RENNIE H . , BIERWIRTH P. & SCOTT G. 1997. Torres Strait satellite image of the sea floor. 1:300 000 scale. Australian Geological Survey Organisation, Canberra. SCHNEIDER P. 1989. Metal source and distribution in Torres Strait and related impacts to the Tiger Prawn Penaeus esculentus. MAppSc thesis, University of Technology, Sydney (unpubl.). TUDHOPE A. W. & SCOFFIN T. P. 1984. The effects of Callianassa bioturbation on the preservation of carbonate grains in Davies Reef lagoon, Great Barrier Reef, Australia. Journal of Sedimentary Petrology 54, 1091-1096. WALKER R. G. & JAMES N. P. (Editors) 1992. Fades Models: Response to Sea Level Change. Geological Association of Canada, St Johns, Newfoundland. WILLIAMS G. C. & STAPLES D. J. 1990. Australia's fisheries research in the Torres Strait protected zone. In: Lawrence D. & CansfieldSmith T. eds. Sustainable Development for Traditional Inhabitants of the Torres Strait Region. Proceedings of the Torres Strait Baseline Study Conference, November 1990, pp. 229-237. Great Barrier Reef Marine Park Authority, Townsville, Workshop Series 16. WOLANSKI E., RIDD P. & INOUE M. 1988. Currents through Torres Strait. Journal of Physical Oceanography 18, 1535-1545. Received 4 August 1997; accepted 6 November 2000

Geological Society of Australia Special Publication 21, 329-337

CHAPTER 28—Australia's marine jurisdiction: baseline information for seabed management and monitoring climate change I

H. M. J. STAGG, N. F. EXON, P. J. HILL, D. T. HEGGIE AND P. E. O'BRIEN

Australian Geological Survey Organisation, GPO Box 378, Canberra ACT 2601 Australia. In ratifying the United Nations Convention on the Law of the Sea (UNCLOS) in 1994, Australia took responsibility for a marine jurisdiction that is one and a half to two times the size of onshore Australia, depending on definitions. Within this jurisdiction, resource-based industries (principally petroleum and fisheries) are annually worth approximately $10 billion to the economy, a figure projected to grow to $35-40 billion within the next 20-25 years, with commensurate pressures on the marine environment. While ownership of offshore resources provided under UNCLOS bestows certain rights (including the right to exploit resources), it also carries a range of obligations to preserve and protect the marine environment. Although the state of health of this environment around Australia has been described as generally good, particularly in comparison to onshore, the diversity and complexity of issues and the gaps in knowledge make it very difficult to simply and scientifically assess this state. In particular, there is a key need for the collection and interpretation of fundamental baseline information that will allow future environmental changes to be accurately monitored, both onshore and offshore. The crucial time for the acquisition of these data is now, before the marine environment is substantially changed. From the geoscience perspective, the most important baseline information is the detailed depth and form of the sea floor and the nature of shallow substrate in high-use areas. Such information underpins the activities of virtually all the government, industry and academic sectors working offshore. The current state of such knowledge is extremely variable, with modern 3D swath-mapping data being available over only about 5% of the Australian Marine Jurisdiction, mainly in deep water. The main technical requirement is for the acquisition of a national capacity for seabed swath-mapping, accompanied by systematic seabed sampling to ground-truth the data. While the Australian Hydrographic Office has acquired swath-mapping capability for the morphologic shelf, deep-water work by Australia can currently only be carried out using foreign research vessels as the opportunity arises, with the consequence that national priorities for survey areas take second place to institutional research priorities. Determination of priorities for a national seabed-mapping program must take into account the range of interests of the individual sectors for which such a program is relevant. With these interests in mind, the southeast quadrant of the continent is the highest priority area overall. Data in this region will provide valuable new information on the tectonic fabric and breakup history, fisheries habitats and human impact on the environment offshore from Australia's principal population centres. Improving agricultural planning through the understanding of climatic change is of great importance in drought-prone Australia. The study of sediment cores from lakes or the ocean can extend our knowledge of climatic variability back far beyond the 150 years of climate records, with the level of detail increasing toward the present. Growth phases in cores from living large colonial corals provide an exceedingly accurate record of climatic changes over a few hundred years. Study of fossil corals may eventually extend this record back a few thousand years. Such studies, and those of sediment cores going back millions of years, can improve the climate models by increasing the time span of the databases on which they are based, back much further than the modeller's primary historical data. The information bank of offshore cores needs to be greatly increased, both geographically and by taking longer cores. This can most logically be done as part of the proposed seabed-mapping program. Longer cores can be taken by using new technology available from inside Australia and overseas. Giant piston cores (40 m long) can be taken in soft sediments from large research vessels. A portable, remotely operated drill should soon be available to take long cores in sedimentary and igneous rocks (100 m). The Ocean Drilling Program can drill and core soft sediments, and sedimentary and igneous rocks, for up to 2000 m below the seabed. KEY WORDS: Law of the Sea, management, marine environment, marine resources, seabed mapping,

INTRODUCTION On 5 October 1994, Australia ratified the 1982 United Nations Convention on the Law of the Sea (UNCLOS) and

thus became an original party to the new convention when it entered into force on 16 November 1994. The implementation of the new UNCLOS regime was a key event for Australia in that it provided new rules for the definition of

330

H. M. J. Stagg et al.

Table 28.1 Approximate areas under Australian Marine Jurisdiction (after Symonds & Willcox 1989). Area (x 10 6 km2) Australian landmass Continental shelf (geomorphological) Australian Exclusive Economic Zone (AEEZ) Australia + Lord Howe overlap Australia + island territories Australia + island territories + Australian Antarctic Territory (AAT) Extended Continental Shelf beyond AEEZ South to AAT (60°S) South to EEZ around AAT Total Australian Marine Jurisdiction Australia + territories, south to AAT Australia + territories, south to EEZ around AAT Australia + territories + EEZ around AAT

7.8 2.4 6.7 8.6 11.1 3.3 3.7 11.9 12.3 14.8

Table 28.2 Approximate values and employment levels of principal marine-based industries and sectors. Sector Oil and gas Commercial fishing and aquiculture Tourism and recreation Shipping and shipbuilding Pharmaceuticals and biotechnology Total (1994)

Value of output ($ billion)

Employment

8 1.6 15 3.7 virtually untapped -28

6 200 19 000 188 000 13 000 minimal - 2 2 6 000

These figures are averaged from several sources, including Veron (1995), Australian Maritime Industries and Sciences Council (1997) and Australian Centre for Maritime Studies (1997).

Australia's vast marine zones, as well as setting out Australia's rights and obligations for managing the environment and resources of those zones. The challenges presented by these rights and obligations will be one of the major issues confronting Australia as it enters the twentyfirst century. Marine geoscience is probably the most widely relevant scientific discipline involved in addressing these challenges, as almost every sector of the marine economy needs marine geoscience information either directly or indirectly. The intention of this chapter is to outline the key issues facing some of these sectors and the role that marine geoscience has to play. A number of marine zones are legally defined under UNCLOS. Of these areas, two are of particular relevance (Figure 28.1 on Plate 14). (1) The Australian Exclusive Economic Zone (AEEZ) extends 2 0 0 nautical miles from the territorial-sea baselines. Within the 11.1 million km 2 of this zone [including the EEZ around the Australian Antarctic Territory (AAT): Table 28.1] Australia has sovereign rights to explore and exploit, and conserve and manage the natural resources of the water column, the seabed and the subsoil. (2) The Legal Continental Shelf* encompasses the AEEZ and those areas which are natural prolongations of the continent extending beyond the AEEZ, according to a detailed set of definitions. Within that part of the Legal Continental Shelf extending beyond the AEEZ, Australia has sovereign rights over the natural resources of the seabed (including organisms attached to the seabed) and

the subsoil, but not over the water column. Preliminary estimates (Table 28.1) of the area of the Legal Continental Shelf beyond the AEEZ (including around the AAT) are about 4.5 x 10 6 km 2 , or about 60% of the area of onshore Australia. Table 28.1 sets out the approximate areas of the different zones that will make up what we will refer to in this chapter as the 'Australian Marine Jurisdiction'. It is immediately obvious that, depending on whether the marine zones of the Australian Antarctic Territory are included, the offshore area is from one-and-a-half to two times the size of onshore Australia. That is, about two-thirds of Australia lies beneath the sea. Australia's immediate priority with respect to UNCLOS is to support its claim to the Legal Continental Shelf beyond the EEZ. This requires the acquisition, processing and interpretation of new data and the preparation of a supporting case for the UN Commission on the Limits of the Continental Shelf by late 2 0 0 4 . Acquisition of the basic data by the Australian Geological Survey Organisation (AGSO) was largely completed by mid-1998, although there are some areas (Macquarie Ridge and east of Norfolk Island) where further information is required. It is expected that the documentation for the supporting case will largely be complete by 2 0 0 1 . * The 'Legal Continental Shelf', defined by a complex series of rules or formulae, is quite distinct from the geomorphic continental shelf as understood by a marine scientist. The Legal Continental Shelf includes the geomorphic continental shelf, the continental slope, marginal plateaus and sometimes the deep ocean.

Australia's Marine Jurisdiction

The value of marine-based industries to the Australian economy has been estimated in a number of reports. It was estimated at $28 billion in 1994 (Table 28.2), with marinebased industries employing in excess of 220 000 people. With real annual growth rates of about 8%, the 1997 value of these industries was estimated to be about $35 billion. This value has been predicted to grow to $50-80 billion by 2020 (Veron 1995). The impact on the marine environment can be expected to be commensurate with this growth. The intention of the remainder of this chapter is to outline: (i) some of the environmental issues, both offshore and onshore, that can be addressed through marine geoscience; and (ii) the marine geoscience program that is required to provide the baseline information to allow seabed management and assist in the monitoring of climate change.

GEOSCIENCE AND THE HEALTH OF THE MARINE ENVIRONMENT In 1995, the Commonwealth Minister for the Environment released the State of the Marine Environment Report (SOMER) (Zann 1995), a comprehensive description of Australia's marine environment, its uses and values, the issues and threats affecting it, and its management. SOMER noted that, because of the vast size of the Australian Marine Jurisdiction, the diversity and complexity of issues facing it, and the major gaps in scientific knowledge, it is not possible to simply and precisely assess the state of the marine environment. However, on the basis of the existing limited information and in comparison with neighbouring countries and equivalent developed countries in the northern hemisphere, the condition or 'health' of Australia's marine environment could be rated as 'generally good', but with many caveats or qualifiers. In general, it was believed that the condition of specific areas varied from almost pristine in remote, undeveloped areas, to locally poor off many highly developed urban, industrial and intensively farmed areas, particularly in the southeast and southwest of the continent. Inshore environments are in noticeably worse condition that environments further offshore, due to the increased dilution of pollutants with distance from source. SOMER identified several major issues affecting Australia's marine environment which have special relevance to geoscience. These included: (i) declining sediment and water quality as a result of inappropriate catchment land-use (rural, urban and industrial) practices; and (ii) loss of marine and coastal habitat. Marine environmental quality is intimately linked with the sustainable use of renewable and non-renewable coastal (estuaries, ports and harbours) and marine (continental shelf) resources, particularly seabed resources, and the concept of submerged land-use. Marine environmental quality has implications for wild fisheries and aquiculture. Petroleum exploration and development, and mineral exploration and extraction can have adverse effects. Declining sediment and water quality can result from the introduction of nutrients and sediments, oil pollution (including ship spills and formation-water discharge from production platforms), heavy metals, agrichemicals (pesti-

331

cides), introduced marine pests and litter. Eutrophication (the excessive growth of algae, often caused by anthropogenic nutrient loadings), and the accumulation of heavy metals, agricultural chemicals and petroleum hydrocarbons (particularly in the sediments) interfere with the sustainable use of coastal and marine resources, notably aquiculture industries in estuaries and the maintenance of wild fisheries. Another problem could arise from the extraction of sea-floor minerals (including construction materials) from the nearshore shelf, particularly in areas near potential urban markets and shipping lanes. Declining water and sediment quality is an important issue, Australia-wide, but particularly in areas of urban growth and rural activity, such as the southeast and southwest of Australia, and on the Great Barrier Reef. Formationwater discharge from offshore petroleum production platforms, while not currently significant in Bass Strait in terms of measurable environmental impact, still warrants ongoing monitoring. Likewise on the North West Shelf, where exploration and development in proximity to important environmental resources and marine parks is increasing, there is a need for such monitoring. The coastal zone is the recipient of inputs from catchments (rural activities) and urban areas, and processes occurring in the estuaries, coastal lakes and embayments control the delivery of pollutants from the land to the sea and Australia's EEZ. The loss of marine and coastal habitat is an important issue affecting sustainable use of wild fisheries and the development of aquiculture. Estuaries and coastal lakes, particularly in the southeast and southwest of Australia, have degraded sediment and water quality, and introduced marine pests have significantly degraded marine and coastal habitats, thus impacting on fisheries and aquiculture industries. The Australia-wide decline in temperate seagrass communities and the loss of mangrove and saltmarsh habitats, have reduced seafood resources, because these areas are important breeding grounds and nurseries for wild fish and aquiculture species.

CLIMATE CHANGE: BASELINE INFORMATION FROM MARINE GEOSCIENCE One of the major contributions geoscience can make to society is by helping agricultural planning, through the understanding of long-term climatic change. The study of sediment cores from lakes or the deep ocean can extend our knowledge of climatic variability back many millions of years. The younger the core, the better the accuracy and definition. The study of growth phases in cores from living large colonial corals (especially Pontes), provides an exceedingly accurate record of climatic changes over the last few hundred years. Fossil corals may eventually extend this record back a few thousand years. These studies extend the databases of climatic modellers back much further than the historical records they are more used to dealing with, thus improving their models. Australia has one of the most variable climatic regimes of any continent, causing great problems for agriculture. The history of early settlement and establishment of agriculture has seen cycles in which settlement advances during high rainfall episodes, and retreats during drought,

332

H. M. J. Stagg et al.

Figure 28.4 Schematic of Simrad EMI 2D hull-mounted multibeam swath-mapping system showing the acoustic beam configuration (after Hill et al. 1995).

with accompanying land degradation. Even today, farm economics are strongly controlled by rainfall variability. Rainfall is predominantly controlled by sea surface temperatures in the surrounding ocean and its variation— warmer surface water produces moisture laden air which brings rain, whereas colder surface water results in less rain. It can be shown that a strong correlation exists between the value of Australia's grain harvest and sea surface temperatures to Australia's north. Better prediction of rainfall is the driving force behind climatic modelling in this country. The best-known phenomenon affecting Australian rainfall is the El Nino-Southern Oscillation (ENSO), driven by fluctuations in sea surface temperature. In ENSO times, the warm water normally present north of Australia moves eastward across the Pacific, producing heavy rain in the central Pacific and South America and drought in Australia. Such ENSO events usually last for one year, but a recent event persisted from 1991 to 1995. The cause of these events is still poorly understood, and prediction still only allows about 6 months warning of an ENSO event. Longer records of ENSO events would help improve computer models that could extend the early warning period of impending drought and predict its duration. The impact of longer term climate fluctuations on Australian agriculture, such as those resulting from global warming, are even more poorly understood. Research in the North Atlantic suggests that major short-term changes in climate (5-8°C in a decade) may be caused by reorganisation of global oceanic circulation. This research is based on an extensive library of marine cores from the Atlantic: such a library does not exist in the Australian region. Geological study of the planktonic fossils, chemistry and inorganic grains in marine cores can provide a longterm record of sea surface temperature variations and hence of climate fluctuations, that will help us assess the risk of severe drought, and the probability of multi-year

drought. Weather records extend back only 150 years, whereas the marine record, particularly as recorded in corals and high-resolution sediment cores, can extend this record much further back. The marine record can also be used to predict the frequency of droughts in a warmer world, by examining records from past warm periods (e.g. Holocene warm phase 5000-8000 years ago; last interglacial about 130 000-180 000 years ago). These data can be used to assist in planning more sustainable agriculture and in agricultural risk analysis. In summary, although meteorological research provides evidence of changes over several hundred years and provides climatic models, climate research also requires ancient sea-surface temperature information from marine sediment cores. These cores can be obtained from coral reefs and the deep ocean, and their analysis can improve the models. Deep-sea cores require a ship with capacity to deploy and retrieve gravity and piston corers and the capability of acquiring swath-mapping images and high-resolution reflection seismic data to target the right sediments. Corals can be sampled with light-weight drill rigs. Analysis of such geological records in cores requires laboratories equipped to log a large number of parameters, particularly isotope geochemistry on carbonate grains.

KEY MARINE SECTOR ISSUES In early 1997, AGSO convened a marine geoscience planning workshop with the objective of determining national priorities for a seabed-mapping program in the Australian Marine Jurisdiction that incorporated the needs of most of the principal sectors working offshore (Stagg et al 1997). The principal marine sectors and their particular interests included: (i) government policy—including marine jurisdiction, multiple and sequential land use; (ii) petroleum exploration, development and field abandonment; (iii) minerals exploration and development—particularly including construction materials and diamonds; (iv) defence research— particularly in anti-submarine and mine warfare; (v) hydrography—safety of navigation; (vi) engineering— seabed and sub-seabed structures, including petroleum production platforms and pipelines and communications cables; (vii) fisheries and mariculture; (viii) introduced pests and diseases; (ix) environment—including the record of climate change and impact on agriculture; (x) pharmaceuticals and biotechnology; (xi) tourism; and (xii) academia and research—universities and government instrumentalities. Most of the sectors have their main interests directed at the water column, sea floor, or shallow substrate (Figure 28.2 on Plate 15). Only the petroleum exploration, government policy sectors and research sectors stand out as having a level of focus directed significantly below the seabed. Two major, common priorities were identified at the planning workshop. (1) The paramount need for detailed information on the depth and form of the sea floor throughout the marine zones, focused on high-use areas. The associated issue of which organisation should be responsible for the collation, storage and dissemination of bathymetric data and information is also fundamental.

Australia's Marine Jurisdiction

333

140° E

Figure 28.5 Areas around Australia which have been swath-mapped prior to 2000, and priority areas for future swath-mapping.

(2) Information on the nature of the sea floor and shallow substrate (sediment type, physical properties etc.) is required at all water depths. While sediment information is required by a wide range of organisations, usually for a single purpose, there is little use of samples for purposes beyond those for which they were acquired. It is evident that there is scope for increased efficiencies in marine science through more widespread and rational use of sediment samples. It is on these two requirements that the remainder of this chapter will be focused.

STATE OF EXISTING SEABED KNOWLEDGE While bathymetry and seabed-morphology data are undoubtedly the most fundamental information that is required to understand, manage and exploit the resources the Australian Marine Jurisdiction, the detail in which this information is available is extremely variable. Even on the morphologic shelf, large tracts of the seabed are relatively sparsely surveyed, particularly with modern techniques. In past decades, national responsibility for shelf mapping has primarily been carried by the former Division of National Mapping and the Australian Hydrographic Office of the Royal Australian Navy (which

absorbed Division of National Mapping hydrographic functions in 1986). The primary requirement for such mapping has been the safety of navigation, and secondarily the support of the Australian defence forces. Consequently, hydrographic mapping has been concentrated in areas that are hazardous to shipping, particularly along the major shipping routes, port approaches and 'choke' points where shipping routes are constricted by geography and bathymetry, as in Torres Strait. Large parts of the continental shelf are still only sparsely covered by modern bathymetric data, most of which are not available in digital form through a national data centre. Beyond the areas covered in detail for navigational purposes, there are localised areas where detailed information has been acquired, generally for commercial exploration and engineering infrastructure development. Thus some quarters of the petroleum exploration industry have acquired highly detailed bathymetric data in areas of existing or prospective petroleum developments, while similar detailed information along narrow corridors has been acquired to support cable or pipeline laying. However, these datasets are usually acquired for proprietary purposes and their release into the public arena is not guaranteed. Beyond the morphologic shelf, bathymetric profiles are not only irregularly spaced, but also are generally very

334

H. M. J. Stagg et al.

sparse. These data have been acquired by a wide variety of organisations—government, academic institutions, industry—but rarely on a systematic basis. Such data are most detailed in areas of active petroleum exploration. Elsewhere, particularly around Australia's island territories, bathymetric data may only be available along profiles separated by tens of kilometres. The bathymetric maps that can be produced from such sparse datasets often appear deceptively good. Often the inadequacy of the data is not apparent until the area is resurveyed by high-resolution multibeam 'swath-mapping' systems that can produce detailed three-dimensional images of seabed morphology. Figure 28.3 (on Plate 16) illustrates the orders-of-magnitude improvement that can be achieved with this technology. Figure 28.3a is the best bathymetric map of the South Tasman Rise that could be produced from digital data in the early 1990s, based on generally randomly spaced bathymetric profiles acquired by a wide range of agencies. By contrast, Figure 28.3b illustrates the degree of detail that can be provided in approximately two weeks of dedicated surveying using a multibeam swathmapping system (Figure 28.4). Clearly, while the previous map is broadly correct bathymetrically, it is incapable of delivering sufficient detail, for example for fisheries management or reliable regional tectonic interpretations. Seabed mapping in broad swaths can be carried out either by towed sidescan sonar systems, or by hullmounted multibeam sonar systems (see review in Exon & Hill 1999). These two swath-mapping systems are capable of mapping in great detail in shallow water, or in less detail in deeper water. Although initially sidescan sonar produced only backscatter imagery, and multibeam systems produced only bathymetric contours, many sidescan sonar systems now produce contours, and many multibeam systems produce imagery. With sidescan sonar, a sound pulse is emitted in a narrow beam at right angles to the ship's track. Echo arrival times and amplitudes are recorded to produce a line segment, and a swath map is built up from successive pulses as the ship moves. A slant-range correction is usually applied to give a reasonably accurate location. The sonar image shows variations in sea bed hardness, roughness and slope. Some sidescan systems now have a bathymetric capability, with both sidescan images and bathymetry recorded simultaneously. Bathymetry across the swath is determined by measuring phase differences of acoustic arrivals on a pair of transducer arrays on each side of a towfish. The phase difference gives the direction of the backscattered return, which combined with the arrival time, yields both depth and position. In very simplistic terms, multibeam sonar systems can be thought of as a large number of conventional singlebeam echo sounders with their beams set in a fan acrosstrack. Multibeam systems have transmit and receive transducer arrays mounted at right angles on the hull, with the transmit array(s) aligned along-ship. The transmit beam is narrow along-ship and wide athwartships. Individual receivers (162 in the case of the Simrad EMI 2D shown in Figure 28.4), which have a narrow athwartships focus, span the transmit sector, so that each received beam has a small footprint (sampling area) on the sea floor. The beams are electronically stabilised to counter roll and

pitch of the ship. Detection of the bottom in the weak signal from the outer beams is achieved by comparing the phase of the returns at the two rows of transducers in the receive array. Seabed backscatter images can also be produced from the beam-formed signals of a multibeam system, and these can be directly superimposed on bathymetric contour maps. For the research community, both towed and shipmounted systems have great value, with the type of system matching the task. One practical advantage of a towed sidescan system is that it can readily be deployed on any suitable research vessel, as it has no requirement for hullmounted transducers. However, a hull-mounted multibeam system requires no lowering or recovering of a towed fish, has better navigational accuracy, and can be used more readily with other towed equipment. Figure 28.5 illustrates those parts of the Australian Marine Jurisdiction for which high-quality swath-mapping data were publicly available prior to 2000 and the priority areas for future work. The existing data are generally only available off the morphologic shelf and the total area covered amounts to no more than about 5% of the Australian Marine Jurisdiction. These surveys include: (i) combined GLORIA (ultra long-range sidescan sonar) and Seabeam (multibeam mapping) surveys by the HMAS Cook off northeast Australia (Johnson et al 1992), New South Wales (Jenkins & Lawrence 1990), canyon systems south of the Murray River mouth (von der Borch & Hughes Clarke 1993) and part of the southern coast of Western Australia; (ii) a survey of part of the Macquarie Ridge in 1994 using the HMR-1 towed sidescan swath-mapping system with bathymetric capability, aboard the RV Rig Seismic (Coffin et al 1994); (iii) a survey of the sea floor west and south of Tasmania in 1994 using a hull-mounted Simrad EMI 2D multibeam mapping system aboard the French RV UAtalante (Hill et al. 1995); (iv) a survey of the sea floor of the eastern margin of Tasmania and the Bass Canyon using the multibeam Seabeam system aboard the Scripps Institution of Oceanography's RV Melville (Exon et al 1999); and (v) a survey of Sydney Harbour in late 1997 using a high-resolution Atlas Fansweep 20 system aboard the RV Oceanographer. A notable feature of all of these multibeam surveys is that, with the exception of the now-obsolete Seabeam system on HMAS Cook and the recent survey of Sydney Harbour, they were carried out with foreign-flag vessels of opportunity, as there is currently no Australian swath-mapping capability. On the morphologic shelf, this situation will begin to be rectified from 1998 when the new RAN hydrographic vessels, Melville (not related to the Scripps Melville, referred to above) and Leeuwin, come into service. Both of these vessels, together with their total of six survey motor launches, have hull-mounted swath-mapping capabilities (Atlas Fansweep 30) of hydrographic standard. While both vessels will be tasked primarily with routine hydrographic surveying, concentrating in northern Australian waters, it is expected that they may also undertake some non-hydrographic national-priority seabed-mapping work. The second major component of a seabed-mapping program is the ground-truthing of the sonar images through sampling of the shallow substrate. A large amount of seabed sampling has been carried out in the past 30 years

Australia's Marine Jurisdiction

by a range of organisations, including: (i) AGSO, through its wide range of research projects, with samples being acquired by multiple techniques, including gravity, piston and vibrocoring, grab sampling and dredging; (ii) the Australian Hydrographic Office, which acquires grab samples as a routine component of its hydrographic surveying; (iii) universities, through a range of academic research programs; and (iv) industry (petroleum exploration and engineering), which has acquired a range of shallow sample data, primarily for engineering site-survey purposes. Seabed samples have also been acquired as an incidental byproduct of operations in non-geoscience sectors— for instance rocks have been dredged by fishing trawlers; however, these samples are acquired on a non-systematic basis and have rarely been archived. While a large amount of sediment data has been acquired, it is only in the past few years that serious attempts have been made to assemble the information in a form that could be accessible to a national seabed-mapping program. Development of the AUSEABED database by the Ocean Sciences Institute (Jenkins & Searle 1997; Jenkins 1997) and AGSO's MARS database are attempts to collate the available information into a form that is accessible by modern technology. However, there is currently no national sample repository (or network of repositories), which means that it is difficult in many cases to access the original samples. It should also be noted that many of the samples have been collected from only the top metre or less of the sedimentary section. Sampling below 10 m requires more advanced coring or drilling techniques than those available from standard research vessels. Cores up to 40 m long can be taken in soft deep-water sediments by giant piston corers, such as those on the French RV Marion Dufresne, which took such cores on the South Tasman Rise and elsewhere in 1997. The Ocean Drilling Program (ODP) (Ocean Drilling Program 1996) is capable of taking continuous cores in soft sediments, sedimentary rocks and igneous rocks in almost any water depth greater than 100 m, down to 2000 m below the seabed, although few holes have yet gone deeper than 1500 m. A major campaign off Australia, in 1988-89, saw many thousands of metres of core taken on the Exmouth Plateau, in the Argo and Gascoyne Abyssal Plains, and in the Great Barrier Reef. This campaign greatly elucidated the geological history of the areas studied. ODP is now back in the Australasian region and drilling legs are programmed from 1998 to 2000 for the Woodlark Basin in Papua New Guinea, the Great Australian Bight and the Antarctic Discordance on the oceanic spreading ridge south of Australia. The scientific value of such coring is tremendous. A consortium based at the University of Sydney has built a Portable Remotely Operated Drill (PROD) designed to recover up to 100 m of core from research or industry vessels in water depths of up to 2000 m. Such a drill would fill the gap between conventional gravity, piston and vibrocoring, and ODP or industrial drilling. A complete seabed survey of all deep-water areas of Australian jurisdiction beyond the morphological continental shelf (9.5 million km 2 , excluding the Australian Antarctic Territory), using a single broad-swath system like the Simrad EMI 2D, has been estimated by Exon and Hill

335

(1999) to take about 12 years at a rate of 180 days at sea per year. The 12 years would include transits and groundtruthing to characterise the rocks and sediments producing characteristic acoustic patterns. The program of swathmapping and sampling beyond the morphologic shelf would include acquisition, processing, interpreting, archiving, retrieving and production in various formats to customers as a public good.

NATIONAL PRIORITIES The foregoing discussion clearly indicates that there is an identifiable national priority to carry out mapping of the seabed and shallow substrate throughout the Australian Marine Jurisdiction. Such a program will provide baseline information that will underpin virtually all of the sectors currently working offshore. It is also clear that the existing Australian capabilities for providing this information to aid management of the Australian Marine Jurisdiction are inadequate if the information is to be available in a timely manner. In late 1999, Australia still had no national capacity to be able to swathmap the seabed in sufficient three-dimensional detail. Modern, multibeam mapping systems have the ability to map a swath of seabed that is up to seven times the water depth (the effective deep-water swath width is about 20 km). Such systems can map up to - 4 0 0 km2/hour at water depths of 3000 m, and 15 km2/hour at average shelf depths of 100 m. The key national need for seabed mapping information has been recognised in Australia's Oceans Policy (Environment Australia 1998) and in the development of the Marine Science and Technology Plan (Marine Science and Technology Plan Working Group 1998, 1999). Program 1, Objective 2 of the Marine Science and Technology Plan is defined as: 'To map the form and nature of the seabed of Australia's Marine Jurisdiction'. Within this objective, several priorities and strategies were identified as relevant to a program of seabed mapping around Australia. The most relevant of these were: (i) 'implementation of a national swath mapping program and seabed data digitisation program, on a priority basis determined by the regional marine planning process'; and (ii) 'develop and implement a systematic national program to swath-map the seabed, and to sample the seabed, substrate and benthic biota of the Australian Marine Jurisdiction...'. During the development of the Marine Science and Technology Plan, a number of priority areas were identified for seabed mapping (Marine Science and Technology Plan Working Group 1998). To some extent, those priorities were superseded by the Oceans Policy (Environment Australia 1998) and the release of the final Marine Science and Technology Plan. Figure 28.5 shows the areas which are now considered to be the highest priority for seabed mapping. The priority 1 areas are: (1) Southeast Australia. Complete surveying of the continental slope between eastern South Australia and southern New South Wales to provide input to the Southeast Australia Regional Marine Plan (Environment Australia 1998). This work will provide information of value to hydrocarbon exploration, commercial and recreational fisheries,

336

H. M. J. Stagg

etol.

biodiversity and water and sediment quality adjacent to Australia's most densely populated areas. (2) Three Kings Rise/Norfolk Basin. The extent of the Legal Continental Shelf beyond the EEZ around Norfolk Island is uncertain and bathymetric data are required to delineate Australia's claim. (3) Macquarie Ridge. Detailed bathymetric data are required to determine the extent of the Legal Continental Shelf south of Macquarie Island along the Macquarie Ridge. In addition, as detailed bathymetry data are required to assist in the definition of the Macquarie Marine Protected Area, there is a requirement to fill in the gaps that were left after the 1994 swath-mapping survey These areas are all scheduled to be surveyed in late 1999-early 2000, using the French RV L'Atalante. The areas shown in Figure 28.5 as being priority 2 are not currently scheduled for surveying. These areas, and the aims of the surveying, include: (1) Southwest Australia. Delineate the tectonic fabric that was developed during the early separation of Australia, Greater India and Antarctica; provide information on fisheries habitats. (2) Norfolk Ridge. Elucidate the tectonic fabric of the Lord Howe Rise-Norfolk Ridge region, a major area of medium- to long-term hydrocarbon potential; potential fish habitats and biodiversity in this region are almost entirely unknown. (3) Mellish Rise/Kenn Plateau. Provide geological framework information of direct relevance to any assessment of the hydrocarbon potential of the northern Lord Howe Rise and northeast Australia. (4) Exmouth/Scott Plateau margins. Provide new framework information that will lead to an improved understanding of the region adjacent to Australia's premier hydrocarbon province.

CONCLUSIONS Mapping of the shape and character of the seabed is a key to properly managing the Australian Marine Jurisdiction and preserving its environment. Currently, this can best be done using swath-mapping technology to obtain accurate contour maps and acoustic images. These digital data can be combined ultimately to allow production of a map of any area at any scale at the touch of a button. Interpretation of the swath maps can be greatly aided by samples obtained through dredging and coring, particularly in areas of outcrop. Once this is done geological maps can be combined with the bathymetric maps to provide basic information for research, resource exploitation, the building of structures, pipelines and cables, and for management. Additional sampling, in the form of long sediment cores and cores through large colonial corals, could be used for studies of climatic change, leading to better climatic modelling and hence to assist in better forecasting for Australia's drought-prone agricultural industries. The Oceans Policy and the Marine Science and Technology Plan both recognise the need for a coordinated program of swath-mapping, ground-truthing and other coring, both on the 2.4 million km 2 continental shelf, where the RAN Hydrographic Service would generally provide the

swath-mapping facilities, and in deep water, where other facilities would be needed. A multipurpose program of seabed mapping and sampling using one vessel with a broad-swath multibeam system for 180 days a year, could map the 10 million km 2 of deep-water Australian Marine Jurisdiction in 12 years. The acquisition of a suitable vessel would bring Australia's blue-water research fleet to the equivalent of three vessels, a small number for a country with Australia's vast offshore area. Most of the initial priorities are in the southeast of the continent, where the bulk of the population lives. The results of this work will be of critical importance to national and state research and management agencies.

ACKNOWLEDGEMENTS W e wish to thank J. B. Colwell and J. B. Willcox for their constructive reviews of this manuscript. This paper is published with the permission of the Chief Executive Officer, Australian Geological Survey Organisation.

REFERENCES AUSTRALIAN CENTRE FOR MARITIME STUDIES 1 9 9 7 .

the facts. Australian Maritime Digest 46.

Australia's oceans—

AUSTRALIAN MARITIME INDUSTRIES AND SCIENCES COUNCIL 1 9 9 7 . Marine

industry strategy development. Department of Industry, Science and Tourism, Canberra. M. E , KARNER G . D . & FALVEY D . A . 1 9 9 4 . Research cruise yields new details of Macquarie Ridge Complex. EOS 75,

COFFIN

561-564.

ENVIRONMENT AUSTRALIA 1 9 9 8 . Australia's Oceans Policy. Environment

Australia, Canberra. EXON N. E & HILL P. J. 1999. Seabed mapping using multibeam swath-mapping systems: an essential technology for mapping Australia's margins. AGSO Journal of Australian Geology & Geophysics 17, 1-16. EXON N. F., HILL P. J., KEENE J. B. & SMITH S . M. 1999. The Sojourn 7 research cruise off eastern Tasmania and in the Gippsland Basin, using R.V. Melville equipped with Seabeam 2000 swath-mapping system. Australian Geological Survey Organisation Record 1999/7. HILL P. J., EXON N. F. & ROYER J. Y. 1995. Swath-mapping the Australian continental margin: results from offshore Tasmania. Exploration Geophysics 26, 403-411. JENKINS C. J . 1997. Building a national scale offshore soils database from both word-based and numeric datasets. Sea Technology 38(12), 2 5 - 2 8 . JENKINS C . J . & LAWRENCE M. W . 1 9 9 0 . Report on the RAN-Marconi GLORIA survey of the EAXA: continental margin of southeastern Australia. University of Sydney, Ocean Sciences Institute Technical Report 22, 1-17 (unpubl.). JENKINS C . & SEARLE B . 1 9 9 7 . A database/GIS of Australian offshore soils. In: 3rd Australian Hydrographic Symposium Papers, p. 164. Hydrographic Society Special Publication 38. JOHNSON D . P., HUGHES CLARKE J . E . & VON DER BORCH C . C .

1992.

MARINE

SCIENCE

AND

TECHNOLOGY

PLAN

WORKING

GROUP

1998.

MARINE

SCIENCE

AND

TECHNOLOGY

PLAN

WORKING

GROUP

1999.

Depositional systems of trough-filling seaward of the Great Barrier Reef. 1st Australian Marine Geoscience Meeting Abstracts Volume (unpaginated). Australian National University, Canberra.

Australia's Marine Science and Technology Plan draft for consultation. Department of Industry, Science & Tourism, Canberra.

Australia's Marine Science and Technology Plan. Department of Industry, Science & Resources, Canberra. OCEAN DRILLING PROGRAM 1 9 9 6 . Understanding our dynamic Earth through ocean drilling: Ocean Drilling Program long range plan. Ocean Drilling Program, Washington, DC. STAGG H . M . J . , EXON N . F., COLWELL J . B . , SYMONDS P. A . , HEGGIE D . ,

Australia's Marine Jurisdiction HILL P. J. & WILLCOX J. B. 1997. Report of the planning workshop on a marine geoscience mapping program in the Australian ocean territory. Australian Geological Survey Organisation Record 1997/18. SYMONDS P. A., MURPHY B., RAMSAY D . , LOCKWOOD K. & BORISSOVA I.

1998. The outer limits of Australia's resource jurisdiction off Western Australia. In: Purcell P. G. & Purcell R. R. eds. The Sedimentary Basins of Western Australia 2, pp. 3-18. Proceedings of Petroleum Exploration Society of Australia Symposium Perth. SYMONDS P. A. & WILLCOX J. B. 1989. Australia's petroleum potential in areans beyond an Exclusive Zone. BMR Journal of Australian Geology & Geophysics 11,11-36.

337

VERON J. 1995. Australia's Ocean Age: Science and Technology for Managing our Ocean Territory. Department of the Prime Minister & Cabinet Office of the Chief Scientist, Canberra. VON DER BORCH C. C. & HUGHES CLARKE J. E. 1 9 9 3 . S l o p e m o r p h o l o g y

adjacent to the cool-water carbonate shelf of South Australia: GLORIA and Seabeam imaging. Australian Journal of Earth

Sciences 4 0 , 5 7 - 6 4 .

ZANN L. P. (Compiler) 1995. Our sea, our future. Major findings of the State of the Marine Environment Report for Australia. Department of the Environment, Sport and Territories Canberra. Received 20 May 1998; accepted 13 September 2000

Geological Society of Australia Special Publication 21, 339-340

Indexes

AUTHOR INDEX Belperio A. P. 201 Birch G. E 243 Bourman R. R 201 Boyd W. E. 257, 267 Brodie R. S- 153 Bush R. T. 257 Chappell J. 303 Chartres C. J. ix Chenhall B. E. 227, 235 Chiaradia M. 235 Chowdhury R. N. 67 Clark M. W. 257 Clarke J. D. A. 49 Cotter M. M. 267 Crabb P. 145 Depers A. M. 227, 235 Exon N. R 329 Fitzpatrick R. W. 27 Freeman M. J. 37 Ghobadi M. H. 67 Goss K. 145 Gostin V. A. xi Granger K. 61 Green D. 95 Gulson B. L. 235 Habermehl ML A. 127 Harris P. T. 317 Harvey N. 201 Hazelton P. A. 27 Heggie D. T. 329 Hill R J. 329 Hodgson L. 161 Hooper W. C. 95 Jacobson G. 171 Jones B. G. 7, 227, 235 Larcombe R 281, 295 Lovering J. R 145 Major R. 111 Matthai C. 243 McGrath R. 267 McNally G. H.13 McPhail D. C. 95 Nichol S. L. 215 O'Brien R E. 329 Pathirana S. 267 Semeniuk C. A. 181 Semeniuk V 181 Slade P. 27 Smith J. V. 257 Specht A. 267 Stagg H. M. J. 329

Sullivan L. A. 257 Taylor G. R 77 Taylor G. 13 Taylor S.E. 243 Williams M. A. J. 3 WischusenJ. 171 Woolfe K. J. 295

SUBJECT INDEX Aboriginal communities 171 acid drainage 77, 95 acid sulfate soils 215, 303 anthropogenic factor 181, 227, 235 artesian basins 127 atmospheric carbon dioxide 3 Avon River 37 Cairns 61 Captains Flat 77 climate change 3, 201, 329 coal mining 161 coastal management 201 coastal sediments 201, 215, 267 coasts 303 community relations 145 conservation 181 construction materials 13 contamination 77, 95 coral reefs 281 corrosive soils 27 cosmic rays 111 cultural heritage 267 cyclones 295 dredging 281 dunes 181 duricrust 13 earthquakes 61 Eastern Highlands 13 engineering geology 67 environmental change 3 environmental geology 49, 127, 181, 243 environmental geoscience 153, 161 environmental impact 161 environmental management 37, 77, 215, 227, 257, 267, 281, 303, 317 environmental science 243 estuaries 215, 243, 303 estuarine sediments 215 expansive soils 27 fire management 267 floods 61, 215

Fly River 317 Foreward ix gas pipeline 317 geochemistry 95 geohazards 61 geoheritage 181 geological history 267 geological processes 257 geology 161 geomorphology 37, 49, 161, 303 Gippsland Basin 161 Great Artesian Basin 127 Great Barrier Reef 281, 295 greenhouse effect 201 groundwater 13, 37, 95, 145, 153, 161, 171 groundwater assessment 127 groundwater management 127 groundwater mining 267 gypsum 37 heavy metals 243 heavy minerals 145 Holocene 201, 215, 281, 295, 303 hydrogeology 95, 127 hydrology 127 in situ leaching 111 Introduction xi Kambalda 49 King River 95, Lake Eyre 3 Lake Illawarra 227, 235 lakes 3 land subsidence 161 landslides 61, 67 Law of the Sea 329 lead isotopes 235 management 329 marine environment 329 marine geology 303 marine resources 329 marine sediments 49 mine tailings 95 mine wastes 317 mining 49, 145 Mt Lyell 77 Mt Morgan 77 Murray Basin 153 Murray-Darling Basin 145, 153 natural hazards 67 natural-resources management 145 neotectonics 201 New South Wales 67, 215, 227, 243, 257, 267

Geological Society of Australia Special Publication 21, 339-340 northern Australia 303 Northern Territory 171 nuclear radiation 111 numerical models 153 Olympic Dam 111 optical-fibre cables 27 oxidation 77 palaeodrainage 37 palaeogeography 3 pedology 27 petroleum industry 161 photosynthesis 3 Pine Creek 77 pollution 227, 235, 243, 257 potassium-40 111 problem soils 13 public risk 61 quarries 257 Quaternary 3, 181, 201 Queensland 257, 267 radioactive wastes 111 radioactivity 111 radium 111

radon gas 111 regolith 13, 49 rehabilitation 77, 145 remediation 67, 243, 257 risk analysis (hazards) 61 risk management 61 Rum Jungle 77 saline soils 27 salinisation 153 salinity 153, 161 sea-level 201, 215, 281, 303 sea-level rise 201 seabed management 329 seabed mapping 329 sediment dispersal 295 sedimentary facies 317 sedimentation 227, 281 sedimentology 243, 317 sediments 95 slumping 67 smelters 235 soil maps 27 South Australia 201

sulfides 77, 95, 257 Swan Coastal Plain 181 swath mapping 329 Sydney Harbour 243 Tasmania 95 tidal currents 317 Torres Strait 317 tourism 161 trace metals 227 turbidity 317 uranium mines 111 vegetation management 267 Victoria 161 water chemistry 127 water management 145, 257 water pipes 27 water quality 257 water supply 171 Western Australia 37, 49, 181 wetlands 181, 303 wind-blown dust 3 Woodlawn 77 Yilgarn Craton 37

Plate 1

(Clockwise from top left) Figure 8.1 Acid drainage in seepage pond below tailings dam, CSA Mine, Cobar, New South Wales. Figure 8.2 Yellow-brown secondary mineral precipitates from acid seepage, Brukunga Mine, South Australia. Figure 8.3 Green filamentous algae in pH = 2.8 seepage below tailings dam wall, Brukunga Mine, South Australia. Figure 8.4 Efflorescence of secondary minerals on surface of pyritic tailings, Thalanga, Queensland. Figure 8.5 Tree mortality due to acid environment in tailings dam, Elura Mine, New South Wales

341

342

Plate 2 Figure 8.6 Cation solubility vs pH. o -2

-4

log (M total) . 6

-8 CATIONS -10

-12

Figure 8.7 Rehabilitation of north dump at Captains Flat, New South Wales.

Figure 8.8 Acid seepage from tailings after rehabilitation, Captains Flat, New South Wales.

Figure 8.9 Waste-rock dump (60-70 Mt) which has been compacted, covered with waste rock and soil, and revegetated with grasses, shrubs and trees, Woodlawn, New South Wales.

Plate 3 (Clockwise from top right) Figure 8 . 1 0 Openpit at Mt Morgan, Queensland, containing low-quality water overlying twice-treated tailings. Figure 8.11 Collapsed slag with overlying sulfidic waste rock, Mt Morgan, Queensland, which is presently a heritage site. Figure 8 . 1 5 Water ponded on sulfidic uranium tailings at Quirke Mine, Elliot Lake, Ontario, Canada. Figure 8 . 1 6 Poor rehabilitation technology has led to reoccurrence of acid drainage, death of vegetation and erosion, Captains Flat, New South Wales. Figure 8 . 1 7 Hardpan/cement formed on surface of tailings by sulfide oxidation, Elura, New South Wales.

343

Plate 4

344

D e p t h (m) o f b a s e o f the R o l l i n g D o w n s G r o u p relative t o m e a n s e a level ( M S L )

0-300

r:\>i

200-0 -700 -200 -1200--700 -1700--1200 -2200--1700 Pre Cadna-owie Fm/Hooray Sst (Jurassic and Triassic units) —

Location of cross section A - B in Figure 4

— —

Cadna-owie Formation and Hooray Sandstone

——

Westbourne Formation Adori Sandstone Birkhead Formation Hutton Sandstone dematis Sandstone

—

» Afice Springs

Northern Territory

Roxby Downs

South Australia

• Broken Hi Part Augusta

New South Wales I

Figure 1 1 . 3

•Dubbo

E x t e n t of t h e a q u i f e r s a n d structure c o n t o u r s o n t h e b a s e of t h e Rolling D o w n s G r o u p - t o p o f C a d n a - o w i e F o r m a t i o n -

H o o r a y S a n d s t o n e a n d e q u i v a l e n t s . A - B is l o c a t i o n o f c r o s s - s e c t i o n in Figure 1 1 . 4 ( o n P l a t e 5 ) .

CTQ

n Vertical Exaggeration X87.5

Great Dividing Range

c era

&00

ft)

Sea Leve

Cooper Creek

Barcoo River Cooper Creet

Sea Level 500

i E

fll OT

era c

C a -iozoIc (Quaternary

and ~er ary un ts)

| Ja | Ador Sandstone

| JKh | -too 91 Sandstone

Mian Mudstone

Coorlklana Sandstone

Habermeh MA &lau, J.E, 1<»97 - Hydrogeology of the Great Aleslar Basin, Australia (Map a* scale :2 500 000

Jlp | Prec olce Sandstone

Coreen* Member

Oodnadatta ^ormatlor

Australian Geologlca Survey Organisation Canberrj

Jmb

Pi kheac Formatlor

Doncaster Mpmber

Wlnton Formation

Poolowpnna cormatlon

Mackunda c o matlo^

C Bmati8 Sandstone

Algebucklna Sandstone

Toolebuc c ormatior

Moolayember formation

Westbourne Formatlo-

Wa lumbi a Formation

Jlh

McKlniav Member JKmr

Murta Format on

| JKn | Namu' Sandstone WtSm Evergreen cormat!o<-

Hutton Sandstone

| Kco | Cadna-o* *> cormation

Bulldog Shnlft

R«f«r to =|gure

for locatior of cross-sectior A-B

Great Artesian Basin Cross-Section A-B

346

Plate 6

Hydrochemistry Total dissolved solids (mg/L) 0-500 500-1000 1000-1500 1500 - 3000 3000 - 7000 >7000 Chemical composition percentage of total equivalents per million K Na

" H

Mg Ca|•

5, SO4 CO3 HCO3

Alice Springs

Northern Territory

Figure 1 1 . 1 1 Groundwater quality (total dissolved solids) and major ion hydrochemistry (chemical composition of groundwater) of selected flowing artesian waterbores in the Great Artesian Basin (after Habermehl 1980; Habermehl & Lau 1997).

Plate 7

1 . Rivers of the MDB

2. FUture Salinities in Basin Rivers

QUEENSLAND

SOUTH AUSTRALIA

Figure 12.1 Map of the Murray-Darling Basin showing major rivers and water storages (Crabb 1997). Figure 12.2 Future salinities in Murray-Darling Basin rivers (Murray-Darling Basin Ministerial Council 1999).

347

348

Plate 8

Cuspate forelands north of Perth

Geomorphic Units Of&hore reefs/rtdges jijljj Offshore Islands Swan CoasTo) Plain ~>j Dandaragan Plateau Daring Plateau

Selected Physical Features Cuspate forelands Location of smaller \ cuspate forelands Wetlands ^r^Rlver and estuary Fautr

Hnes/Valfeys Dunes and ridaes including star-dunes, Hrvearmego-dunes Figure 16.2

Figure 16.2 Line diagram and coloured satellite image centred on Perth and the Swan River, showing natural features of interest on the Swan Coastal Plain. To the east is the fractured Precambrian rock terrain of the Yilgarn Craton that, with associated laterite and sand, comprises the Darling Plateau. Features marked on the map are the large double cuspate foreland of the Becher Point area, the small cuspate forelands north of Perth, offshore ridges and reefs, chains of linear wetlands and of round wetlands, the more conspicuous various other large wetlands, the various linear and star-dune forms and star-dune networks. The extent of urbanisation in the region is evident on the image but is not shown in the line drawing.

Plate 10

Port Jackson

Sydney CBD Figure 21.4 Copper concentrations in fine (<62.5 pm) surficial sediments in Parramatta River/Port Jackson estuary. Heavy-metal concentrations in pg

350

Cu in sediment fine fraction

>300 150 100 <70

Middle Harbour Lane Cove River

Platell

351

120m . 200m

TASMAN SEA

Kilometers

TASMAN SEA 0

25 50 Kilometers

TASMAN SEA 0

Kilometers

Figure 21.11 The distribution of Cu (jag g" ) in total surficial sediments on the continental margin of central New South Wales is controlled by the distribution of fine sediment (<62.5 pm) on the middle shelf. However, the distribution of Cu in the fine fraction of the sediment clearly shows the effect of the major population centres on the marine environment. These distributions demonstrate the advantages of size-normalisation in detecting impacted areas and contaminant source, as well as determining dispersion patterns. 1

352

Plate 12

^'•"fei:.

Figure 24.9 (a) View of the reef flat of Paluma Shoals at low water on the lowest spring tides. Note the turbid water stirred up at the person's feet, (b) Two coral microatolls (Goniastrea sp.) each approximately 1 m in diameter. The central portions of the microatolls contain significant deposits of grey mud, colonised by green algae in the example in the background.

Plate 13

SB91 73PC6

353

SB91 62PC5

r- Ocm Holocene poorly sorted calcareous gravelly muddy sand

- 10

-75cm

-80

K M ffifi \% - 30 85 - 40 Bioturbated transition layer - 50

- 90 Pleistocene very well sorted montmorillonite clay

Figure 27.7 Photographs of cores 73PC6 (Station 73 in Figure 27.3) and 62PC5 (Station 62 in Figure 27.3) from northern Torres Strait. The upper part of core 73PC6 is comprised of calcareous gravelly muddy sand (interpreted as Holocene age) with abundant gravel-sized mollusc fragments. This overlies a cohesive pale bluish-grey to reddish-brown clay, which exhibits dewatering structures (mudcracks) indicating subaerial exposure; it is interpreted as a Pleistocene-aged deposit. The two beds are separated by a transitional layer thai comprised of a bioturbated mixture of carbonate sand and clay. Core 62PC5 is a light olive-grey, mediun sandy clay with minor carbonate (mainly benthic foraminifers) and characterises the terrigenous facies' of Figure 27.10) that are found north or Saibai Island and along the southern Pap

354

Plate 14

(a) I uj CO

Sovereign rights for exploring and exploiting non-living resources o fsea-bed and subsoil,plus sedentary species.

Common T M heritage of man- : kind

Payment for exploitation beyond200nM

Sovereign rights for exploring,exploiting, conserving and managing living and non-living resources of the water,sea-bed and subsoil Give access to surplus allowable catch

LOWEU [SLOPE i RISE J OCEAN 1

(b)

~1 Australian Exclusive I bconomtc Zone (AEEZ)

J Preliminary area of Continental I Shelf extending beyond AEEZ

noAflte

Maximum passible *v*»nried Continental Shelf beyond AEEZ adjacent to AAT (based on 380 nmile cutoff) Boundary to be neaotiated/ratified

Figure 28.1 (a) Marine jurisdiction zones contained in the 1982 United Nations Convention on the Law of the Sea. (b) Sketch map showing the main marine jurisdiction zones around Australia and its territories (after Symonds et al. 1998).

Figure 28.2 The four linked regimes of the marine environment and their relevance to the key marine sectors.

Jurissdiction Fisheries

Pests

Climate & Agriculture

Environment

Engineering

Petroleum Exploration

Multiple & Sequential Land Use

Atmosphere

Shallow substrate

Plate 15 355

356

Plate 16

Figure 2 8 . 3 Bathymetry of the northern sector of the South Tasman Rise, (a) Contours derived from all available digital d. early 1990s, (b) Colour-shaded relief image derived from approximately two weeks of multibeam swath-mapping. Note the losolution along the left-hand border of the image where swath-mapped data are not available. The areas shown in the two'ma, identical. Tasmania is at the top of each map.

0 Ma

QUATERNARYTERTIARY

High Sea Level

JURASSIC

TRIASSIC 111

Low Sea Level

PERMIAN 295 Ma CARBONIFEROUS

DEVONIAN

434 ORDOVICIAN

ZMi:-'- '

WM,

llSfll®! I

, *K £ « V *j

ailra^^WFal

^ ^ -*