GSA Special Publication No.19: Tectonics & Metallogenesis of the New England Orogen

TECTONICS AND METALLOGENESIS OF THE NEW ENGLAND OROGEN

1997 THE GEOLOGICAL SOCIETY OF AUSTRALIA INCORPORATED This volume is published in the Special Publications series of The Geological Society of Australia REFERENCES General Reference: ASHLEY P. M. & FLOOD P. G. (Eds) 1997. Tectonics and Metallogenesis of the New England Orogen: Alan H. Voisey Memorial Volume. Geological Society of Australia, Special Publication 19, 303pp. Two forms of reference to specific papers are possible, as follows: MURRAY C. G. 1997. From geosyncline to fold belt: a personal perspective on the development of ideas regarding the tectonic evolution of the New England Orogen. Geological Society of Australia Special Publication 19, 1-28. MURRAY C. G. 1997. From geosyncline to fold belt: a personal perspective on the development of ideas regarding the tectonic evolution of the New England Orogen. In: Ashley P. M. & Flood P. G. (eds) Tectonics and metallogenesis of the New England Orogen: Alan H. Voisey Memorial Volume. Geological Society of Australia Special Publication 19, 1—28. First published in 1997 for The Geological Society of Australia by Conference Publications, PO Box 107, Springwood, NSW 2777, Australia. Typeset by Conference Publications, Springwood NSW; printed and bound by Southwood Press, Sydney. Registered in Australia for transmission by post as a book. Orders for Special Publication No. 19 shouldofbeAustralia, directed to: The Business Manager, Geological Society 706 Wynyard House, 301 George Street, Sydney, NSW 2000, AUSTRALIA. 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 written permission. Inquiries should be made to The Geological Society of Australia. © Geological Society of Australia Incorporated, 1997. ISSN 0072-1085. National Library ofAustralia Cataloguing-in-Publication data Tectonics and metallogenesis of the New England Orogen: Alan H. Voisey Memorial volume. Bibliography. Includes index. ISBN 1 876315 02 4. 1. Geology, Structural - New South Wales. 2. Geology, Structural - Queensland. 3. Metallogeny - New South Wales. 4. Metallogeny - Queensland. I. Ashley, P. M. II. Flood, P. G. (Peter Gerard), 1946- . III. Geological Society of Australia. (Series: Special publication (Geological Society of Australia); no. 19). 559.444

TECTONICS AND METALLOGENESIS of the NEW ENGLAND OROGEN ALANH. VOISEYMEMORIAL VOLUME

Edited by

P. M. Ashley and P. G. Flood

SPECIAL PUBLICATION NO. 19, GEOLOGICAL SOCIETY OF AUSTRALIA DECEMBER 1997

Professor Alan H. Voisey DSc (Sydney) 1911-1995 as remembered by most of his students.

TECTONICS A N D M E T A L L O G E N E S I S OF THE NEW E N G L A N D O R O G E N ALANH. VOISEYMEMORIAL VOLUME

Introduction THIS SPECIAL PUBLICATION is dedicated to the memory of Alan Heywood Voisey (19111995). A first-class honours graduate of the University of Sydney, Alan Voisey was appointed Lecturer-in-Charge of Geology and Geography at New England University College in 1939. Fifteen years later, in 1954, he was appointed Foundation Professor of Geology of the new University of New England where he continued until 1965. In 1966, he was appointed Foundation Professor of the School of Earth Sciences at Macquarie University and served that institution until his retirement from academia in 1971. Alan Voisey infected his pupils with his enthusiasm for geology. He was a character, an individual for whom students displayed great respect. Alan's research interests centred on understanding New England geology and he published more than 30 papers over the period 1939-1969. Most of these contained fundamental original results based on field mapping and these efforts formed the basis of first editions of 1:100 000 and 1:250 000 geological map sheets. His mapping of the KempseyTaree area remains unsurpassed to this day. His mapping of the Keepit and Manilla areas were masterpieces of field observations undertaken prior to the availability of aerial photographs or detailed base maps. His Presidential Address to the 1959 ANZAAS meeting discussed the opposing ideas as to whether the Australian craton had developed by successive accretion on to a continental nucleus, or by deposition on an existing continent of the same dimension. Following a sabbatical leave overseas, he returned to New England and applied Marshall Kay's 'geosynclinal concept' to this problem and concluded that the continent had advanced to the east by successive deformation. His contribution has greatly assisted the current studies of the tectonic development of the New England Orogen. In summary, Alan Voisey displayed diverse interests in geomorphology, stratigraphy, tectonics, metallogenesis and education. He will always be remembered by his students for the unselfconscious enthusiasm, displayed in all geological pursuits. For more than three decades he produced Earth Sciences graduates, many of whom have exerted a profound influence on the development and application of the discipline. Then, for almost another two decades (1970-1987), he applied his diverse knowledge to a variety of exploration and mining ventures as Chief Geologist, Exploration Manager, Director, and consultant. His memoirs Sixty Years on the Rocks were published in 1991 by the Earth Sciences History Group of the Geological Society of Australia (Harrington et al. 1991). Alan Voisey died on 14 April 1995, just three days after his 84th birthday. With his passing many geologists lost a friend and colleague. In this Special Publication, we have collected together a diverse group of twenty papers prepared by 47 authors who have undertaken recent research on the tectonics and metallogenesis of the New England Orogen, the most easterly, and youngest, of the collage of orogenic domains comprising the Tasman Orogenic Zone of eastern Australia. The paper by Murray opens the volume and gives a personal perspective on the development of ideas regarding tectonic evolution of the New England Orogen. It is followed by a contribution by Korsch, Johnstone and Wake-Dyster outlining the crustal structure of the New England Orogen based on the results of deep seismic refraction profiling. A group of papers on the v

northern portion of the New England Orogen follows with contributions from Holcombe, Stephens, Fielding, Gust, Little, Sliwa, Kassan, McPhie and Ewart (two papers), and Morand on tectonic evolution, a paper by Fielding, Stephens and Holcombe on Permian stratigraphy and palaeogeography in coastal central Queensland, a paper by Messenger, Golding and Taube on the volcanic setting of the Mt Morgan Au-Cu deposit and ore genesis implications, and two papers on the Gympie Province in southeast Queensland — one outlining general aspects of the geology and mineral deposits (Cranfield, Shorten, Scott and Barker) and the other the implications of geochemical and isotopic studies of Gympie Province basalts (Sivell and McCulloch). Several groups of papers constitute the results of research in the southern portion of the New England Orogen. Lennox and Flood outline the structure and age of the Texas megafold, Stratford and Aitchison report on the lithostratigraphy of the Gamilaroi terrane, and a related paper by Aitchison, Stratford and Buckman describes evidence for oblique-slip movement on the Peel-Manning Fault System. Three papers provide contributions on the dismembered ophiolitic rocks of the Weraerai terrane (Great Serpentinite Belt): Yang and Seccombe report on the geochemistry of mafic and ultramafic complexes, Ashley describes the nature of silica-carbonate alteration zones and related gold deposits, and Offler, O'Hanley and Lennox outline the tectonic significance of veins and associated structures. The New England Orogen contains abundant granitoids and related hydrothermal mineral deposits and three papers report on topics from the southern portion of the orogen. In their paper on the Clarence River Supersuite, Bryant, Cosca and Arculus provide new 40 Ar/ 3 9 Ar ages, and Kleeman, Plimer, Lu, Foster and Davidson date the timing of thermal and mineralisation events in the Mole Granite. The Dumboy-Gragin Granite, a recently recognised leucogranite, is described by Vickery, Ashley and Fanning. The application of vitrinite reflectance and stable isotopic studies at the Copeland Goldfield is reported by Gibson and Seccombe, and lastly, Lottermoser, Ashley, Muller and Whistler provide the results of an environmental geochemical study at the Halls Peak massive sulfide deposits. As editors of this Special Publication, we thank the authors and reviewers for their contributions and anticipate that the papers will be of value in guiding future research and assisting mineral exploration in the New England Orogen.

REFERENCE HARRINGTON H . J., Y E A T E S A. J, B R A N A G A N D . F. & M C N A L L Y G. H . (Eds) 1 9 9 1 . Sixty Years on the Rocks: The Memoirs of Professor Alan H. Voisey. Earth Sciences History Group, Geological Society of Australia, Sydney.

P. M. ASHLEY P. G. FLOOD Editors

Contents Introduction

C. G. Murray: From geosyncline to fold belt: a personal perspective on the development of ideas regarding the tectonic evolution of the New England Orogen

1

R. J. Korsch, D. W. Johnstone and K. D. Wake-Dyster: Crustal architecture of the New England Orogen based on deep seismic reflection profiling

29

R. J. Holcombe, C. J. Stephens, C. R. Fielding, D. Gust, T. A. Little, R. Sliwa, J. Kassan, J. McPhie and A. Ewart: Tectonic evolution of the northern New England Fold Belt: the Permian-Triassic Hunter-Bowen event

52

R. J. Holcombe, C. J. Stephens, C. R. Fielding, D. Gust, T. A. Little, R. Sliwa, J. McPhie and A. Ewart: Tectonic evolution of the northern New England Fold Belt: Carboniferous to Early Permian transition from active accretion to extension

66

C. R. Fielding, C. J. Stephens and R. J. Holcombe: Permian stratigraphy and palaeogeography of the eastern Bowen Basin, Gogango Overfolded Zone and Strathmuir Synclinorium in the Rockhampton-Mackay region, central Queensland

80

V. J. Morand: Tectonic significance of The Shacks Mylonite Zone and related shear zones, northern New England Fold Belt, Queensland

96

P. R. Messenger, S. D. Golding and A. Taube: Volcanic setting of the Mt Morgan Au-Cu deposit, central Queensland: implications for ore genesis

109

L. C. Cranfield, G. Shorten, M. Scott and R. M. Barker: Geology and mineralisation of the Gympie Province

128

W. J. Sivell and M. T. McCulloch: Geochemistry and Sm-Nd isotope systematics of Early Permian basalts from Gympie Province and fault basins in southeast Queensland: implications for mantle sources in a backarc setting at the Gondwana rim

148

P. G. Lennox and P. G. Flood: Age and structural characterisation of the Texas megafold, southern New England Orogen, eastern Australia

161

J. M. C. Stratford and J. C. Aitchison: Lithostratigraphy of the Gamilaroi terrane, upper Barnard region, northeastern New South Wales

178

J. C. Aitchison, J. M. C. Stratford and S. Buckman: Geology of the Upper Barnard region: evidence of Early Permian oblique-slip faulting along the Peel-Manning Fault System

188

K. Yang and P. K. Seccombe: Geochemistry of the mafic and ultramafic complexes of the northern Great Serpentinite Belt, New South Wales: implications for first-stage melting

197

P. M. Ashley: Silica-carbonate alteration zones and gold mineralisation in the Great Serpentinite Belt, New England Orogen, New South Wales

212

R. Offler, D. S. O'Hanley and P. Lennox: Tectonic significance of veins and associated structures at Woodsreef, New South Wales

226

40

39

C. J. Bryant, M. A. Cosca and R. J. Arculus: Ar/ Ar ages of Clarence River Supersuite intrusions from the northern portion of the New England Batholith, southern New England Orogen

Vll

242

J. D. Kleeman, I. R. Plimer, J. Lu, D. A. Foster and R. Davidson: Timing of thermal and mineralisation events associated with the Mole Granite, New South Wales

254

N. M. Vickery, P. M. Ashley and C. M. Fanning: Dumboy-Gragin Granite, northeastern New South Wales: age and compositional affinities

266

J. H. Gibson and P. K. Seccombe: Vitrinite reflectance and isotopic evidence for epithermal mineralisation at the Copeland goldfield, southern New England Fold Belt, New South Wales

272

B. G. Lottermoser P. M. Ashley, M. Muller and B. D. Whistler: Metal contamination at the abandoned Halls Peak massive sulfide deposits, New South Wales

290

Index

301

viii

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 1-28.

From geosyncline to fold belt: a personal perspective on the development of ideas regarding the tectonic evolution of the New England Orogen C. G. MURRAY Geological Survey of Queensland, GPO Box 194, Brisbane, Qld 4001, Australia. By late last century, the New England Orogen was known to have been the site of thick marine Devonian and Carboniferous deposition, but its presently accepted boundary was first delineated in 1930 as the western limit of Late Permian folding and granite emplacement The juxtaposition of two belts with contrasting stratigraphy and structure across the Peel and Yarrol Fault Systems has been the major impediment to development of an understanding of the geological history of the New England Orogen. For the first 70 years of this century, deformational history was used as an indicator of age, and the more strongly deformed and metamorphosed eastern sequence was almost universally regarded as older than the well-dated sedimentary rocks to the west. Eagerness to find fossils in the 'unfossiliferous' eastern sequence led to cases of misidentification and two palaeontological hoaxes; a surprisingly high proportion of real or imagined fossils from this sequence has subsequently been lost. It was only the application of plate-tectonic models interpreting the western belt as forearc basin strata, and the eastern belt as an accretionary-wedge assemblage, that explained how these structurally different sequences could be largely coeval. A Late rather than Early Palaeozoic age for the accretionary-wedge sequence, suggested by the occurrence of oolitic arenites, has been confirmed over recent years by the recovery of crinoid fragments, radiolarians and conodonts. Definite Early Palaeozoic rocks in the New England Orogen occur along the Peel Fault System in the New England region, and comprise dismembered ophiolitic complexes, volcaniclastic sediments, and limestones of Cambrian and Ordovician age. Also, Ordovician conodonts have recently been recovered from limestones adjacent to the Yarrol Fault System south of Gladstone. The original tectonic setting and mode of emplacement of this assemblage remains unclear. In these rocks, and also in the accretionary wedge and forearc basin sequences, a continuing problem has been the relative age of allochthonous limestone blocks and their host sediments. Large-scale fault displacements and major oroclinal bending have recently been proposed to account for the present-day patterns of rock distribution in the New England Orogen. These concepts are not new, but were first suggested or implied more than 70 years ago. Folding at the end of the Palaeozoic was documented late last century. This Hunter-Bowen Movement or Orogeny has been long recognised as the major fold-thrust event in the New England Orogen. The existence and significance of earlier orogenies in mid- to Late Carboniferous and Middle-Late Devonian time have been more controversial. Deformation styles typical of subduction zone processes have been described from the accretionarywedge assemblage. For the first half of this century, the main focus of debate on the tectonic evolution of eastern Australia was whether the Australian continent had originally been much smaller and had grown to its present size by accretion to the east, or whether it had been much larger and had subsequently collapsed. For the Tasman Geosyncline in general and the New England Orogen in particular, these opposing concepts implied deposition in an oceanic environment marginal to the continent, or in an intracontinental trough, respectively. An attempt to explain the development of the New England Orogen in terms of classical geosynclinal theory completely failed to account for the difference in stratigraphy and structure across the Peel and Yarrol Fault Systems. The dominance of platetectonic models over the last two decades is not just an indication of the general acceptance of this hypothesis, but, more importantly, is a result of the fact that these models explain the development of the New England Orogen much more satisfactorily than any fixist concept. When the method of multiple working hypotheses is applied to the New England Orogen, plate tectonics succeeds where other theories fail. Key words: continental accretion, geosynclines, New England Orogen, orogeny, plate tectonics, tectonics.

INTRODUCTION The N e w England Orogen is the easternmost tectonic element in the Tasman Orogenic Zone of eastern Australia. It extends for 1500 km from Newcastle in the south to Bowen in the north, and is bounded to the west by the Hunter-Mooki-Goondiwindi Fault System along which the N e w England Orogen has been thrust westward partly over the Sydney-Gunnedah-Bowen Basin.

For much of its length, the N e w England Orogen is divided into contrasting belts by a major fault zone marked by serpentinite lenses, the Peel-Yarrol Fault System (Figure 1). To the west, sequences are generally gently folded, forming coherent packages with wellconstrained ages determined from relatively abundant marine macrofaunas. In contrast, strata to the east are complexly deformed and strongly cleaved, with local development of melange structures, and the lack of

2

C. G. M U R R A Y

.CAMPWYN BLOCK

YARROL PROVINCE BOWEN lockhampton YARROL FAULT

MARYBOROUGH BASIN

GYMPIE >PROVINCE GREAT

^D-AGUILAR BLOCK

YARRAMAN' BLOCK GREAT MORETONFAUU

ARTESIAN

SILVERWOOD BLOCK,

BASIN

, Jprisbane

, CLARENCE /BEENLEIGH BLOCK VMORETONI U BASIN ^EMU CREEK BLOCK -COFFS HARBOUR BLOCK ^NAMBUCCA BLOCK ^HASTINGS BLOCK

-32° LACHLAN OROGEN

148° Figure

1

Structural

SYDNEY

tectonics was being invented by W. R. Morgan and D. P. McKenzie. As far as the New England Orogen was concerned, I had been thoroughly trained at university that the more deformed rocks must be the oldest, and that vertical tectonics and normal faulting were the dominant tectonic style. Having experienced at first hand the transition from geosynclines to plate tectonics, my main purpose in this paper is to demonstrate how effective prevailing dogma was in controlling the development of ideas on the tectonic evolution of the New England Orogen. I have also tried to acknowledge those who attempted, in many cases unsuccessfully, to introduce new concepts for the New England Orogen. The initial lack of acceptance of these concepts was not necessarily because they were incorrect, but because they were not consistent with prevailing dogma.

NEW ENGLAND PROVINCE Newcastle

units of eastern Australia

showing

subdivisions of the N e w England Orogen (shaded).

marker beds and sparsity of diagnostic fossils have made it impossible to recognise large-scale stratigraphy. The juxtaposition of these two contrasting belts across a structural boundary has been the major impediment to development of an understanding of the geological history of the New England Orogen. For most of this century, deformational history has been used as an indicator of age, and the more strongly deformed and metamorphosed eastern sequence was almost universally regarded as older than the well-dated sediments to the west (David & Browne 1950; Voisey 1959a, 1969; Hill 1960). It was only the application of plate-tectonic models, interpreting the western belt as forearc basin strata and the eastern belt as an accretionary-wedge assemblage, that led to the realisation that these structurally different sequences could be largely coeval (Scheibner 1973; Leitch 1974, 1975; Murray 1974, 1975). When I started mapping in the New England Orogen more than 30 years ago, the geosynclinal theory, as expressed most elegantly by Kay (1951), held sway. Sea-floor spreading, outlined by Hess (1962) in his 'essay on geopoetry' was still in its infancy, and plate

RECOGNITION AND DEFINITION OF THE NEW ENGLAND OROGEN The New England Orogen was first distinguished from other tectonic elements of the Tasman Orogenic Zone by the recognition of thick Devonian and Carboniferous marine strata last century. In time, this region came to be regarded as a separate depositional basin: 'During the whole of the Devonian Period the sea which covered [the Tamworth Province] appears to have been very definitely isolated from those seas which covered other parts of New South Wales' (Sussmilch 1923, p. 20). Palaeogeographic maps such as those of Jensen (1912) and Bryan (1926) showed the distribution of Devonian and Carboniferous deposition in the New England Orogen quite accurately. An interesting point noted by Bryan (1926) is that the earliest palaeogeographic maps of Queensland for the Carboniferous show a deep-sea area to the east of the shallow-marine deposits which were known at that time, but later maps show a land mass, named Tasmantis by Sussmilch (1920, p. 277). This change reflects the fact that early workers such as Jensen regarded 'the unfossiliferous Palaeozoic beds between Brisbane and Coffs Harbour, and also those north of Brisbane in the D'Aguilar Ranges, ... as a palaeozoic [sic] complex of deep sea deposits, ranging in age from, perhaps, Pre-Cambrian to Devonian or Carboniferous in age' (Jensen 1912, p. 164), whereas by the 1920s the prevailing opinion was that these unfossiliferous strata were Early Devonian and older, and had been uplifted by Carboniferous time. The New England Orogen was first clearly defined as a structural entity by Browne (1929) for New South Wales and Reid (1930) for Queensland, who delineated the New England Orogen (without naming it) as an area involved in Late Permian folding and granite emplacement (Figure 2). The Late Permian folding event, commencing at the end of the upper marine stage in the Sydney-Gunnedah-Bowen Basin, was named the Hunter-Bowen Movement by Carey and Browne (1938).'* They described what is now known as the * Notes 1 - 5 5 will be found at the end of the paper.

3

TECTONIC EVOLUTION, NEW ENGLAND OROGEN

rr—i i _ l

AREA CONSIDERED INVOLVED IN LATE PERMIAN FOLDING AND PLUTONIC INTRUSIONS (REID)

. . . . . . LIMITING UNE OF LATE PERMIAN STRONG FOLDING, INCLUDING ASSUMED CONTINUATION BENEATH MESOZOICS (REID) BOUNDARY OF FOLDED UPPER PALAEOZOIC STRATA IN NSW (BROWNE) |

I| Figure 2 The first map delineating the western structural boundary of the New England Orogen, from Reid (1930). The boundary of the New England Orogen in New South Wales is from Browne (1929).

Figure 3 Palaeogeographic map of eastern Australia for the Devonian (Andrews 1938). Andrews defined the New England Geosyncline as the narrow trough of thick sedimentation through western New England, but clearly showed this trough continuing north as far as Rockhampton in Queensland. which Devonian sediments of the Tamworth and Barraba Series (and also the Woolomin Series) were deposited in subparallel belts. Extension of this name northwards into Queensland is justified by the fact that Andrews showed the New England Geosyncline as the southern part of a belt of thick Devonian sedimentation which continues north to Rockhampton (Figure 3). This trough was bordered to the east by 'Ancestral New England' and its northern continuation along the coast of Queensland, including Tasmantis (Andrews 1938, pp. 161, 168). Early Carboniferous marine deposition, as depicted in Figure 6 of Andrews (1938, p. 175), was also concentrated along this belt. Voisey (1959b) was 2

New England Orogen as 'a wide belt of folded Upper Palaeozoic strata, bounded westward by thrust-faults and extending perhaps from Townsville to Maitland — a distance of some 900 miles — in a great curve concave to the west, which in late Palaeozoic time seems to have behaved as a tectonic unit under the influence of horizontally-acting forces directed from the Pacific Ocean' (Carey & Browne 1938, p. 608). The name New England was introduced by Andrews (1938, p. 161), who applied the term New England Geosyncline to a relatively narrow and sinking trough in

3

4

C. G. M U R R A Y

the first to use the name for the entire New England Orogen, but extended the northern limit of his New England Eugeosyncline as far as Princess Charlotte Bay on Cape York Peninsula. The term orogen is used in this paper rather than fold belt because the total development of the New England Orogen is considered, including both stratotectonic and structural aspects (see discussion by Scheibner 1993, p. 2; 1996, pp. 15, 20). Because of its division into southern and northern parts by Mesozoic cover of the Clarence-Moreton Basin, the combined name New England-Yarrol Orogen or Fold Belt has been used for the New England Orogen, for example on the tectonic map of Australia and New Guinea (Geological Society of Australia 1971), and it has even been separated into the Yarrol and New England Orogens (Roberts & Engel 1980; Harrington & Korsch 1985a). However, it is clearly a single tectonic unit and the simpler term New England Orogen is preferred as it has priority and is more widely used.

AGE AND SUBDIVISION OF THE ACCRETIONARY-WEDGE ASSEMBLAGE Early views on the age of the sequence of generally unfossiliferous rocks, now regarded as an accretionarywedge assemblage, reflected the primitive knowledge of regional geology last century. R. L. Jack included all these rocks in his Gympie beds, 4 a series of strata which attain an immense development in the south-eastern portion of [Queensland], and are probably represented in many portions of the eastern coast district to the North' (Jack & Etheridge, 1892, p. 72). Because the marine faunas in the type area at Gympie were then considered to be Carboniferous, the age assigned to the Gympie beds was Permo-Carboniferous (a combined term for the entire Carboniferous and Permian Periods). The presence of the Gympie Series in New England was recognised by David (1894, pp. 586-587) on the basis of Permian fossils at Vegetable Creek.4 Both the Gympie Group in the type area at Gympie and the sediments at Vegetable Creek are now interpreted as younger sequences unrelated to the accretionary-wedge assemblage. B. Dunstan, while investigating phosphate occurrences in the Gladstone-Yeppoon area, was the first to attempt to subdivide the accretionary-wedge rocks into discrete belts. Initially, he defined three parallel northwest-trending belts: Devonian limestone and slate in the west; serpentinite in the middle; and phosphate-bearing slates in the east (Dunstan 1904a, b). Subsequently, he recognised a belt of manganiferous slates and quartzites between the serpentinite and the phosphatic slates, and was able to trace the entire sequence of parallel belts as far south as Brisbane. The Devonian limestoneslate belt was considered to be the top of the series, and the phosphate-bearing slates, tentatively regarded as Ordovician by correlation with phosphatic sequences in Victoria, the base (Dunstan 1916; 1920, p. 250). Richards and Bryan (1928, p. 288) explained anomalous relationships at a number of localities where the

phosphatic slates appear to overlie the manganiferous jasper sequence as the result of 'heavy overfolding from the east'. This concept of a continuous sequence from the accretionary-wedge assemblage, across the Yarrol Fault System represented along much of its length by serpentinite lenses, and into fossiliferous Devonian strata of the forearc basin or its precursor, has been reiterated a number of times since Dunstan's original proposal. Whitehouse (1928, p. 441) considered 'that there is a perfectly conformable passage from the lowest beds of the [Yeppoon] area into the typical Devonian Series', and Dear et al. (1971, p. 10) quoted company mapping which suggested that the fossiliferous Devonian Calliope beds 'were conformable with the ?Lower Palaeozoic strata (Doonside Formation) that outcrop immediately to the east of them'. Another aspect of Dunstan's interpretation which persisted for almost the next 70 years was that the accretionarywedge assemblage was youngest in the west and became progressively older eastwards. In his detailed mapping of the Great Serpentine Belt of New England, Benson (1913) came to the same conclusion as Dunstan in central Queensland. He considered the Woolomin beds, east of the Peel Fault System, to be strictly conformable with the Tamworth beds to the west: 'In the region north-east of Manilla, where the Tamworth Beds are believed to occur east of the serpentine-line,5 there is no suggestion of any discordance between them and the enclosing Woolomin Beds' (Benson 1913, p. 496). As the limestones in the Tamworth beds were then considered to be Middle Devonian, and Benson interpreted the Woolomin beds as the lowest beds of the series, he suggested an Early Devonian age for the latter unit. Benson emphasised that the presence of radiolarian jaspers was the most notable feature of the Woolomin beds. The view that the Woolomin beds or Series was conformable with known Devonian strata west of the Peel Fault Zone persisted for at least 25 years, as indicated first by David (1932, p. 50) and then by Andrews (1938, p. 168) who stated that 'the Woolomin (Lower? Devonian), Tamworth (Middle Devonian) and Barraba (Upper Devonian) Series ... are conformable with each other'. Andrews (1938, p. 158) also continued to support the view that the rocks of the accretionary wedge become older to the east. The discovery by Richards and Bryan (1925) of radiolarian jaspers at the western edge of the Brisbane Schists near Fernvale, associated with a serpentinite belt, provided strong evidence for correlation with the Woolomin Series. They accepted an Early Devonian age for the Fernvale jaspers, and suggested that Dunstan's manganiferous slates must be Silurian if the easternmost phosphatic slates were Ordovician. Denmead (1928) carried out the first detailed mapping of the Brisbane Schists, and recognised four conformable subdivisions, the Greenstone, Bunya, Neranleigh and Fernvale Series. He accepted a Devonian age for the Fernvale Series by correlation with the Woolomin Series, and also emphasised the similarity of the sequence at Brisbane with the Gladstone-Curtis Island Series to the north. However, Denmead argued against

TECTONIC EVOLUTION, NEW ENGLAND OROGEN the correlation of the phosphatic sequences in Victoria and Queensland, and, following Bryan (1925, p. 47), suggested that no part of the Brisbane Schists could be older than Silurian because of the lack of evidence of any unconformity in the sequence, whereas intense folding had characterised the close of the Ordovician elsewhere in eastern Australia. The arguments presented by Bryan and Denmead were not accepted by David (1932, p. 40), who supported the correlation of the phosphatic sequences: 'A long intermittent belt of Upper Ordovician rocks, characterised by veins of turquoise and other hydrous phosphates, extends from the Ovens River, Victoria, through Bodalla and Murwillumbah in New South Wales, the Neranleigh and Bunya Series of Brisbane, and the cherts of Gladstone and Yeppoon, to Innisfail, south of Cairns, Queensland. The uniform character of this phosphatic zone suggests a more or less continuous development of Upper Ordovician rocks, along the eastern sea-board of Australia, for the distance of about 1,700 miles'. Assignment of a Late Ordovician age was no doubt strongly influenced by the discovery (by Denmead) of a supposed Diplograptus from south of Brisbane, as recorded by Whitehouse (1930) and David (1932, pp. 40, 44) (see subsequent discussion of reported fossil discoveries from the accretionary-wedge assemblage). The collection of Early Devonian corals from limestones of the Tamworth Series by Brown (1942) led her to deduce a pre-Devonian rather than Early Devonian age for the Woolomin Series, and for the next 30 years the accretionary-wedge assemblage was consigned to the Early Palaeozoic. Thus when David and Browne (1950) published their geology of Australia, the Woolomin, Fernvale and Neranleigh Series, as well as the correlative Coffs Harbour or Fitzroy Series, were listed as Silurian, the Bunya and Nambucca Series as Ordovician, and the Greenstone Series of the Brisbane Schists as Cambrian. By the time The Geology of Queensland was published (Hill & Denmead 1960), a major unconformity had been proposed within the Brisbane Metamorphics (Bryan & Jones 1954), and the Bunya Phyllite and Rocksberg Greenstones were both described in the Precambrian chapter.6 The Woolomin beds were still regarded as the youngest part of the accretionary-wedge assemblage, but their age could not be determined by comparison with rocks west of the Peel Fault System because 'no sequence which shows clearly the relationship between the Woolomin Group, regarded as being at the top, and the Devonian rocks has yet been found' (Voisey 1959a, p. 197). The first conclusive evidence that rocks of the accretionary-wedge sequence could be Late Palaeozoic rather than Early Palaeozoic in age was provided by Ball (1923), who collected the Early Carboniferous coral Lithostrotion from limestone lenses in manganesebearing jaspers and slates in the Texas area of southern Queensland. Ball (1923, p. 457) was in no doubt that the steeply dipping, complexly folded slates were also of Carboniferous age: 'Because of the development of a slightly schistose cleavage in the shales enclosing the manganese deposits, they are considered to belong to a

5

lower horizon than the limestones of Gore and Texas, but in the absence of fossil evidence I can recognise no justification for their removal from the carboniferous [sic]. Between Gore and Cobba-da-mana these schistose shales everywhere are steeply tilted, and multiple folding is suspected'. Yet despite the fact that limestone at Ashford across the border in New South Wales also proved to be Early Carboniferous (Raggatt 1941), the prevailing view was still that the host rocks were much older, and Ball's heretical interpretation was specifically targeted: 'The [Silurian] Fernvale Series has probable counterparts in the jaspers of Gore and Texas' (David & Browne 1950, p. 205). Voisey (1938, p. 463) noted the similarity of tuffs and olive-green mudstones west of Armidale containing plant and crinoid fragments to Early Carboniferous Burindi beds of the Hunter and Manning districts, but later correlated them with the Woolomin Series (Voisey 1942). His original interpretation was subsequently supported by the discovery of a Carboniferous lepidodendroid plant from the tuff and mudstone sequence (Crook 1958).7 But once again, the lithological and palaeontological evidence was overwhelmed by the established dogma that the 'unfossiliferous' rocks of the Tablelands Complex of New England were of Early Palaeozoic age.8 In the volume on the geology of New South Wales edited by Packham (1969a), the Ashford limestone and Crook's fossil locality were mentioned only briefly in the chapter on the Carboniferous system of the New England region (p. 258), and were not related in any way to the Woolomin beds (described in an earlier section on Lower Palaeozoic Systems), despite the fact that the Woolomin beds were considered to extend east of the New England Batholith into the Ashford and Armidale areas (Voisey 1969, p. 229). An Early Palaeozoic age for the Woolomin beds appeared to be confirmed by the discovery of limestone lenses containing Ordovician-Silurian coral faunas near Attunga (Chappell 1961). However, only the Silurian limestones are now considered to lie within the Woolomin beds (Hall 1978), and detailed observations by Fitzpatrick (1974; also see Korsch 1977, p. 343) indicated that even these could not definitely be considered to be part of the sequence. Hall (1978) described Early Silurian corals from the limestones in the Woolomin beds, but gave conflicting accounts of their relationships. He first reported that the coral faunas were from 'two limestone lenses interbedded with cherts and jaspers of the Woolomin beds', but later suggested that the coral age may not be applicable to the Woolomin beds 'because of a possible allochthonous origin for the limestones' (Hall 1978, p. 85). No further suggestions of a Late Palaeozoic age for the accretionary wedge were forthcoming until McKellar (1967) correlated particular lithologies in this assemblage with fossiliferous Early Carboniferous sedimentary rocks of the Yarrol Basin west of the Yarrol Fault System in the Monto district. 'Much of the cleaved undifferentiated Palaeozoic strata to the east of the Yarrol fault eventually may be proved to be Carboniferous in age ... Fracture cleaved, coarse oolitic

6

C. G. M U R R A Y

subgraywacke that must have its place in the [Lower Carboniferous] Caswell Creek Group has been identified from such strata two miles east of the Yarrol Fault near Kalpowar' (McKellar 1967, p. 25). The significance of this correlation for the palaeogeographic reconstruction of the Carboniferous Yarrol Basin was clearly recognised: 'Extension of the Upper Palaeozoic sedimentary basin to the east of the present structural boundary at the Yarrol Fault now seems to be beyond doubt' (McKellar 1967, p. 34). Over the last 30 years, evidence to support a Late Palaeozoic age for the bulk of the accretionary-wedge sequence has accumulated at an increasing rate. Limestones at Murgon in southeast Queensland, long considered to be interbedded in the Wondai Series that was correlated with the unfossiliferous accretionarywedge assemblage, yielded Late Carboniferous conodonts (Palmieri 1969).9 D. L. Strusz (Appendix 3 in Olgers & Flood 1970a) established that Early Carboniferous limestones were widely distributed in the Texas beds10, and confirmed the earlier interpretation of Ball (1923) that the host rocks were no older than Carboniferous. W. G. Whitaker and I collected granite boulders from conglomerates in the NeranleighFernvale beds11 south of Brisbane which were dated as mid-Carboniferous by Green (1973). Oolitic arenites collected by H. Schwarzbock in the NeranleighFernvale beds near Mt Nebo, northwest of Brisbane, and recognised as such by Whitaker (1972), yielded elliptical crinoid stems which proved a Late Palaeozoic age for these rocks (Fleming et al 1974). The occurrence of abundant silicified ooliths made an Early Carboniferous age most probable, because oolitic limestones are virtually confined to this age range in southeast Queensland (McKellar 1967; Kirkegaard et al 1970). The subsequent discovery of oolitic arenites, with or without crinoid fragments, along the entire strike length of the accretionary-wedge sequence (Whitaker et al 1974; Fleming et al 1975; E. C. Leitch pers. comm. 1976; Fergusson 1984; Fergusson and Flood 1984; Korsch 1984; P. G. Flood in Roberts 1987), parallel to fossiliferous oolitic limestones of the Tamworth Belt and Yarrol Basin to the west, provided conclusive evidence for an Early Carboniferous age for at least part of the sequence. Roberts (1985, 1987) found that one type of elliptical crinoid stem recovered from the Neranleigh-Fernvale beds and Wandilla Formation by Fleming et al (1974, 1975) was confined to the Early Carboniferous Schellwienella cf. burlingtonensis brachiopod zone in the Tamworth Belt.12 Cawood (1982a, 1983) and Korsch (1984) approached the problem of the age of the accretionarywedge assemblage from a different perspective by attempting to correlate sandstone compositions across the Peel Fault System. Whereas Cawood studied a relatively small area near Tamworth, Korsch compiled a large number of modal analyses from published and unpublished literature covering much of the New England region. He was able to suggest ages for a number of lithological associations in the accretionarywedge sequence by comparison with fossiliferous units

of the Tamworth Belt: Early to Middle Devonian for the Woolomin association; Late Devonian for the Sandon association; and Early Carboniferous to Early Permian for the Coffs Harbour association (Korsch 1984, p. 203). Although incorrect in detail, these correlations give a broad indication of the age range of the accretionary-wedge sequence in the Tablelands Complex. Over recent years, concerted attempts by J. C. Aitchison, H. Ishiga and B. L. Murchey to date jaspers and cherts of the accretionary wedge by extraction of radiolarians (and conodonts) has yielded a limited database (Aitchison 1988a, b, c, 1990; Aitchison & Flood 1990; Aitchison et al 1992a; Ishiga 1990; Ishiga et al 1988a, b; Blake & Murchey 1988). Although faunas as old as mid-Silurian have been reported from the Woolomin association and equivalents in Queensland, most dated samples range in age from Middle Devonian to Early Carboniferous13, suggesting that the major part of the accretionary wedge accumulated over a relatively short time-span. The advent of plate tectonics provided a model which was consistent with a Late Palaeozoic age for much of the accretionary-wedge sequence. One major result was the conclusion that the overall younging of this sequence was to the east and not to the west as had been accepted since the original subdivision by Dunstan (1904a, b). This was first explicitly stated on regional geological maps (Pogson 1972; Pogson & Hitchins 1973) that showed an eastward younging of the accretionary wedge in New England, from the SilurianDevonian or Ordovician-Silurian Woolomin beds in the west, through an unnamed Silurian-Devonian or Ordovician—Devonian unit, to the Carboniferous Sandon and Texas beds (including the Ashford limestone) in the east. The Woolomin beds and their equivalents in Queensland (Fernvale Series of Denmead 1928, and Doonside Formation of Kirkegaard et al 1970) were recognised as ocean-floor material forming the base of the accretionary pile, with the serpentinite belts along the Peel and Yarrol Fault Systems representing dismembered ophiolites (Oversby 1971; Scheibner 1972; Scheibner & Glen 1972; Fitzpatrick 1974; Murray 1974). It was not until a decade later that the classic structure of accretionary wedges, comprising imbricated thrust slices in which the younging direction is opposite to the overall younging determined from faunal and lithological evidence, was described by Cawood (1982b) and Fergusson (1982a) from the New England region. The age range of the accretionary-wedge assemblage remains one of the greatest problems in deciphering the geological history of the New England Orogen. Apart from the uncertainty of the lower age limit, which has already been discussed, the upper age limit is also known only in general terms. The youngest rocks of the accretionary wedge are presumably the easternmost exposures, which include anomalous quartz-rich sandstones. Limestone lenses in the Wondai Series were determined as Late Carboniferous (NamurianWestphalian) by Palmieri (1969), and similar limestones occur in the Good Night beds to the north. Questionable

TECTONIC EVOLUTION, NEW ENGLAND OROGEN Late Carboniferous bryozoans were identified from the Texas beds by Fleming (Appendix 6 in Olgers & Flood 1970a). Isotopic dates from the Hillgrove plutonic suite in New England (Collins et al. 1993) and equivalent intrusives in southeast Queensland (Little et al. 1993), and from metamorphosed accretionary-wedge rocks west of Coffs Harbour (Graham & Korsch 1985), in central New England (Watanabe et al. 1988), and in the D'Aguilar Block north of Brisbane (Little et al. 1993) indicate that accretion ceased before the end of the Carboniferous. From a personal point of view, it was only the platetectonic revolution which enabled me to reconcile conflicting aspects of the age relationships of the New England Orogen accretionary wedge. I was a member of the joint Geological Survey of Queensland - Bureau of Mineral Resources party which mapped the Rockhampton and Port Clinton 1:250 000 sheet areas in 1965 and 1966. During the course of this mapping, the accretionary-wedge assemblage was divided into three lithologically distinct units, named from west to east the Doonside, Wandilla and Shoalwater Formations, which together constitute the Curtis Island Group (Kirkegaard et al. 1970).14 Kirkegaard et al. (1970, p. 13) regarded the Curtis Island Group as 'lithologically distinct from the Devonian and younger rocks', mainly on the basis of the quartz-rich arenites of the Shoalwater Formation. They also emphasised differences in structural history and depositional environment, the Curtis Island Group being considered to have been deposited 'in very deep water' (Kirkegaard et al. 1970, p. II), 15 and suggested an ?Ordovician-Silurian age. However, they found calcareous and silicified ooliths and sparse crinoid columns in arenites in the middle part of the Wandilla Formation at two localities, similar to the rocks correlated with the Early Carboniferous Caswell Creek Group by McKellar (1967).16 The idea of a young Curtis Island Group appealed to me, but current tectonic models could not explain why two coeval units juxtaposed along the Yarrol Fault System should display such vastly different stratigraphic and structural styles. A few years later, the answer was provided when I took a course by C. A. Burk, an expert on continental margins (Burk & Drake 1974).17 From his regional mapping of the Alaskan Peninsula and Aleutian Arc (Burk 1965), he developed the concept that geosynclinal deposits at continental margins were progressively added to the continent by deformation and uplift. These rocks dipped towards the continent, but showed overall younging towards the ocean. Burk (1965, 1972) therefore recognised the essential features of subduction complexes or accretionary wedges, even though these terms had not yet been used. The age range of these deformed assemblages was the same as that of relatively undeformed shallow-water sedimentary rocks further inland. This model could be directly applied to central Queensland, and I became convinced firstly that the Curtis Island Group and its correlatives were of Late rather than Early Palaeozoic age, and secondly that oolitic arenites were a key to proving this idea.

7

REPORTED MACROFOSSILS FROM THE ACCRETIONARY-WEDGE ASSEMBLAGE IN QUEENSLAND The apparently unfossiliferous nature of the rocks of the accretionary-wedge assemblage of the New England Orogen posed a tremendous challenge to geologists. Their eagerness to find fossils in these enigmatic rocks led to cases of misidentification, and contributed to the success of two deliberate hoaxes. The first hoax was almost certainly perpetrated by James Smith,18 a collector employed by the Geological Survey of Queensland last century, and led to the following statement by R. L. Jack (Jack & Etheridge 1892, p. 91): 'The originally sedimentary nature of the Cawarral Serpentine [northeast of Rockhampton] however metamorphosed, is evidenced by the occurrence of a specimen, collected by Mr Smith, and now in the Geological Survey Collection, containing the impression of a small fossil, which my Colleague [Etheridge] says "may be either a perforate Spirifer or Athyris"'. The specimen is still in the Geological Survey of Queensland collection, but is merely an artefact made by carving a patterned impression with a finely serrated object. It was beautifully done, but it is astonishing that its non-organic origin was not recognised by Etheridge if he actually examined the specimen. In 1928, A. K. Denmead found a large number of carbonaceous fragments in rocks of the Brisbane Schists at Boyds Bay, Tweed Heads. He exhibited one of these fragments at a meeting of the Royal Society of Queensland in 1929. It had been identified by R. A. Keble, a Victorian graptolite expert, as belonging to the Diplograptidae, and assigned a Late Ordovician or Early Silurian age (Anonymous 1930; Whitehouse 1930; David 1932, pp. 40, 44). However, there was some doubt about the accuracy of this identification, as indicated by Andrews (1938, p. 146): 'The "Diplograptus" mentioned by David appears to have been mislaid, and moreover, there is a divergence of opinion as to its organic origin'.19 Bryan (1944, p. 54) confirmed that the specimen had been lost, but apparently still regarded it as genuine (Bryan & Jones 1944, pp. 15-16). In 1944, another infamous graptolite was found during a university geology excursion to Brookfield, a western suburb of Brisbane, in an area of outcrop of the Neranleigh Series of the Brisbane Schists. It was identified as the genus Diplograptus, possibly the subgenus Orthograptus, and indicated an Ordovician age (Anonymous 1945; Ball 1945). But this discovery proved to be another hoax, the graptolite being a Victorian specimen from one of the university college collections (Bryan 1948).20 The realisation that the Brookfield graptolite had been planted had the effect of raising yet more doubts about the specimen collected earlier by Denmead, and Bryan (1948, p. 204) concluded that 'there was now no authentic record of any graptolite having been collected in Queensland'. Even Denmead himself did not mention it when listing 'undoubted organic impressions suggestive of eurypterid fragments' which he also collected from

8

C. G. M U R R A Y

Tweed Heads in 1928 (Denmead 1960b, p. 137).21 L. C. Ball was one of the keenest searchers for fossils in the Brisbane Schists,22 and in 1946 exhibited polished blocks and a cast 'reminiscent of fragmentary Archaeocyathinae' from Brookfield, not far from the 'NBG slates' (Anonymous 1947). A group of us from the Geological Survey of Queensland checked out the locality and found that these specimens came from interstitial calcareous material in pillow basalts and were obviously not serious contenders as authentic fossils. The most substantial genuine macrofossil collected from the accretionary-wedge assemblage in Queensland was found by a schoolboy in coarse greywacke of the Neranleigh-Fernvale beds at Cedar Creek Falls, Mt Tamborine, where several crinoid columns had also been discovered (Denmead 1960b). This was a conulariid (.Paraconularia) which indicated a Late Palaeozoic age (Murray & Whitaker 1982, p. 87).23 EARLY PALAEOZOIC ROCKS IN THE NEW ENGLAND OROGEN Although evidence gathered over the last 30 years confirms a Late rather than Early Palaeozoic age for the accretionary-wedge assemblage of the New England Orogen, this same period has seen the recognition of definite Early Palaeozoic sequences along the Peel Fault System. The first report of confirmed pre-Silurian rocks was by Philip (1966) who listed Late Ordovician coral and conodont faunas from a small limestone outcrop, the Trelawney beds, on the western side of the Peel Fault System south of Tamworth. Limestones of the same age, in a similar structural position at the eastern margin of the Tamworth Belt north of Tamworth, were subsequently documented by Hall (1975)24 as the Uralba beds. The most substantial sequence of Early Palaeozoic rocks, comprising 1330 m of volcaniclastic sediments with Middle or early Late Cambrian limestone blocks at the base and Ordovician limestone clasts towards the top, was described by Cawood (1976) from the Woolomin area south of Tamworth.25 This sequence, which is bounded by the Peel Fault System to the east and by Devonian strata of the Tamworth Group to the west, was earlier mapped as part of the Tamworth Group by Crook (1961). It was used by Cawood (1976), Cawood and Leitch (1985) and Leitch and Cawood (1987) as evidence for an Early Palaeozoic arc-related history for the New England Orogen. However, because the faunas came 'from allochthonous limestone blocks commonly less than 1.5 m in diameter' (Cawood 1976, p. 318), and because of the existence of resedimented limestone blocks in olistostromal units in this area (Leitch & Cawood 1980), subsequent authors argued that the Cambro-Ordovician limestones are merely older clasts within the Devonian Tamworth Group, and discounted an Early Palaeozoic succession and history for the New England Orogen (see Korsch & Harrington 1981, p. 208; Murray 1988, p. 205; Aitchison & Flood

1995, pp. 158-159). The recent discovery by Stewart (1995) of abundant paraconodonts from thin-bedded chert within the sequence mapped by Cawood (1976) is consistent with a Middle or early Late Cambrian age, and proves that an Early Palaeozoic succession really does exist in the eastern part of the Tamworth Belt.26 However, there is still no evidence that any part of the accretionary-wedge assemblage is of comparable age. Further evidence of an Early Palaeozoic history for the New England Orogen is provided by U-Pb dates from zircons from ophiolitic complexes along the Peel Fault System, which give results close to 530 Ma, or Middle Cambrian (Aitchison et al 1992b; Aitchison & Ireland 1995). The close coincidence in age of the ophiolitic assemblage and the volcaniclastic sequence described by Cawood (1976) suggests a related origin, possibly in an intraoceanic arc setting (Glen & Scheibner 1993, p. 124). Fukui et al (1995, p. 1018) proposed that the ophiolitic rocks and the Middle Cambrian volcaniclastics were part of a convergent plate-boundary assemblage which also included midOrdovician high-pressure metamorphics. The timing and mechanism of emplacement of the Early Palaeozoic rocks to their present position are unknown. In spite of (or perhaps because of) the lack of data, a surprisingly large number of possibilities have been suggested: (i) they were firstly incorporated into Ordovician to mid-Devonian accretionary prisms associated with the Lachlan Orogen, and subsequently incorporated into the Late Devonian-Carboniferous accretionary prism of the New England Orogen (Glen & Scheibner 1993, p. 125); (ii) they were technically exhumed from a buried basement of Lachlan Fold Belt rocks along a west-dipping structure visible on a deep seismic profile (Glen & Scheibner 1993, p. 125; Korsch et al 1993, pp. 95-96); (iii) they were remnants of an allochthonous terrane formed in an oceanic setting and accreted to the continental margin during the main Devonian-Carboniferous episode of plate convergence (Fukui et al 1995, p. 1018); and (iv) the ophiolitic assemblage 'represents only portions of a Cambrian intra-oceanic arc rift sequence over which younger terranes... may have been thrust westwards as a series of thin-skinned nappes' during later accretionary processes (Aitchison & Ireland 1995, p. 20).27 High-pressure metamorphic rocks from several localities in the New England region have given Ordovician isotopic dates. Most if not all samples have come from blocks in serpentinite melanges. An age of 444 Ma was originally obtained from Port Macquarie (Scheibner 1985), and more recent dating by Fukui et al (1995) has produced consistent results ranging from 482 to 467 Ma for samples along the southern part of the Peel Fault System and at Port Macquarie. In Queensland, localised limestone clasts in the Early Devonian Rosenthal Creek Formation of the Silverwood Group south of Warwick contain Late Ordovician bryozoans (Wass & Dennis 1977). The Silverwood Group has been correlated with the Tamworth Group (Bryan 1925; Leitch 1975; Cawood & Leitch 1985), so this occurrence may be similar in some respects to the

TECTONIC EVOLUTION, NEW ENGLAND OROGEN Trelawney and Uralba beds along the Peel Fault System near Tamworth. Recently, Ordovician rocks have been discovered within the Calliope beds south of Gladstone, previously regarded as entirely of Devonian age. Mid-Ordovician conodonts have been recovered from a limestone bed in a thin sequence roughly conformable with the surrounding Devonian sediments. In addition, a conglomerate within the surrounding Devonian sequence contains limestone clasts of both Late Ordovician and Devonian age. The location of the Ordovician rocks, near the Yarrol Fault System, is similar to that of Cambrian and Ordovician strata in the Tamworth area of the southern New England Orogen (Murray et al. 1997).

MAJOR STRIKE-SLIP FAULT DISPLACEMENTS A common feature of convergent plate margins such as the New England Orogen is the occurrence of strike-slip faults in the forearc region due to oblique convergence (Fitch 1972). These faults typically are parallel or subparallel to regional structural trends, and are therefore difficult to recognise because they do not displace contrasting tectonic units unless total transcurrent movement on them is very large. It is astonishing, therefore, that the first suggestion of major fault displacement in the New England Orogen was made more than 70 years ago by Bryan (1925). Bryan (1925, pp. 57-60) noted a 'great break in trend' between all pre-Permian rock units in southern Queensland and northern New South Wales (Figure 4).28 He compared the trends and positions of four different groups of rocks that all indicated the same sense and similar amounts of displacement: the supposedly preDevonian Brisbane Schists and Coffs Harbour Schists; serpentinite belts; the Devonian Silverwood Series and Tamworth Series; and Carboniferous strata of the Rockhampton Series and the Tamworth Belt. Not surprisingly, in view of what was then known about strike-slip faulting, Bryan came up with the simplest explanation, which was dextral displacement across strike. He concluded that, 'while there may possibly be other ways of explaining all these harmonious data, the most reasonable hypothesis is that a lateral displacement of all series older than Permo-Carboniferous has taken place after those beds have been folded along N.N.W. axes, but before the deposition of the PermoCarboniferous Series. This displacement has taken place at right angles to the strike and axes of folding. The amount of such displacement was very large and was approximately 150 miles' (Bryan 1925, pp. 59-60). 29 A few years later, Bryan (1928) repeated his radical hypothesis: 'The Coffs Harbour schists in northern N.S.W. although out of alignment with the [Brisbane Schists], are regarded by most geologists as being part of the same series, the lack of alignment being due to a lateral break — a drag of the rocks for about 150 miles in a direction at right angles to the strike' (quoted by Voisey 1934, p. 344). Bryan's postulated 'great break in

9

trend' was not popular with geologists of the time, who were convinced by the quantitative arguments of physicists such as Sir Harold Jeffreys that large-scale movements of the Earth's crust were not possible, even over long periods of time (Jeffreys 1929, pp. 3043 05),30 and his imaginative hypothesis was quickly forgotten. This concept of major offset was reinvented by Rod (1966, p. 875) who proposed that the Birrie Geosuture (Darling River Lineament) cut across the New England Orogen in almost an east-west direction and displaced serpentinite belts in a right lateral sense by 160 km. Leitch (1975) was the next author to postulate an east-west displacement of units in the New England Orogen, again not aware of earlier proposals. He noted (p. 142) that 'it is difficult to link the inferred elements [of the New England Orogen], for those in the north appear displaced eastward relative to those farther south'. Leitch explained this displacement by movement on 'an east-northeast-striking transform-like fault', the Tenterfield Fault. It is interesting to note that the position of the Tenterfield Fault (Figure 5) corresponds almost exactly with that of Bryan's 'great break in trend' (Figure 4). As well as the east-northeast offset on the Tenterfield Fault, Leitch (1975, p. 142) also postulated that the north-trending Demon Fault had been the site of major strike-slip movement which brought the Coffs Harbour Block southwards from 'a position adjacent to the western volcanic chain in southern Queensland'. Earlier, Leitch (1974, p. 148) had specified this southward movement of the Coffs Harbour Block as 'several hundred kilometres in Permian time'. Scheibner and Glen (1972), Scheibner (1973) and Runnegar (1974) also suggested significant right lateral strike-slip faulting along the Demon Fault in preTriassic time, partly to explain the deformation of the Nambucca Slates. This general theme was also adopted by Murray and Whitaker (1982, p. 90), who postulated 'major southward movement of the Queensland portion of the New England Orogen along a north-northwest trending Late Carboniferous-Early Permian dextral transform fault similar to the present San Andreas Fault' (but which was not a precursor of the Demon Fault). The Peel Fault System is another major lineament on which substantial strike-slip faulting was proposed, but the sense of movement has been disputed. Scheibner and Glen (1972, p. 12) stated that up to 150 km of dextral strike-slip movement took place along it in the mid-Permian, whereas Cawood (1982c) required sinistral displacement at this time. Harrington and Korsch (1985a, p. 169) introduced the concept that the western boundary of the New England Orogen, the Mooki Fault System and its northern continuation along the eastern margin of the Bowen Basin in Queensland, was a major fault zone in the latest Carboniferous and earliest Permian, with up to 500 km of dextral strike-slip movement.31 Subsequent discussions of major strike-slip faulting were mainly concerned with the hypothesis of oroclinal bending of the New England Orogen.

10

C. G. M U R R A Y

Figure 4 Great break in trend affecting all pre-Permian tectonic units of the New England Orogen (Bryan 1925).

Figure 5 Interpretation of the New England Orogen by Leitch (1975), showing offset along the Tenterfield Fault which coincides with the break in trend of Bryan (1925).

OROCLINAL BENDING

'inclined to regard the Bracker Creek and GraysholmYuaraba outcrops as northern and southern limbs of a great fold' (Ball 1923, p. 457). 32 This was the first documentation of part of the structure which was later to be named the Texas Megafold or Orocline. The full extent of the Texas Megafold was mapped by Lucas (1960b, Figure 32), who showed rocks of his Thanes Creek Slate and Beacon Mudstone (both later included in the Texas beds by Olgers and Flood 1970a) curving through an angle of more than 90° from northwest trends in the east, through east-west trends in the north, to south-southwest trends in the southwest. Despite the uniformly steep dips of the folded strata, Lucas (1960b, p. 234) explained the structure by outwards thrusting in opposite directions from an indefinite northwest-trending axis. He expanded this concept to explain the apparent symmetry of the entire

The overall north-northwest structural grain of the New England Orogen was recognised late last century (David 1894), but regional discrepancies from this dominant trend soon became apparent. Ball (1923) observed anomalous trends in jasper, shale and schistose slate, regarded by him as Carboniferous and now included in the Texas beds, in the Warwick-Texas area of southern Queensland. In the northern part of the belt, the rocks are 'everywhere steeply tilted and with prevalent east-west strike, which is nearly at right angles to the dominant trend of eastern Australia' (Ball 1923, p. 457). Ball considered that these same rocks could be traced southwards past Bonshaw, and noted that their strike swung around first to the southwest and then to south-southwest. He was

TECTONIC EVOLUTION, NEW ENGLAND OROGEN

northern Tablelands Complex across this axis as far as the Peel Fault Zone in the west (produced by westward thrusting), and the Baryulgil Serpentinite in the east (a zone of eastward thrusting). The eastern thrust belt presumably included the Condamine Fault, which separated the Thanes Creek Slate (a correlative of the Woolomin beds) from the Devonian Silverwood Group to the east, and formed a mirror image of the Peel Fault System in the Tamworth area. The concept of the Texas Megafold was developed still further by Runnegar (1974) although, like both previous authors, he did not name the feature. Runnegar (1974, p. 11) suggested that the northern continuation of the Peel Fault System beneath platform cover of the Surat Basin, as indicated by satellite imagery and gravity trends, swung to the northeast, 'paralleling the strike of tightly folded Late Palaeozoic sediments of the Texas area'. Based on the well-documented change in strike of the exposed Texas beds, and the observation that the succession across the Condamine Thrust near Warwick was the reverse of that found across the Peel Fault System in the Tamworth area, Runnegar then postulated that the Peel and Condamine Faults were joined in an arcuate fashion (Figure 6) and were part of a single lobate thrust. He also suggested that the Baryulgil Serpentinite defined a southern continuation of the Condamine Thrust. To this stage, Runnegar's model is virtually identical to that of Lucas (1960b). However, Runnegar (1974, p. 16) believed that thrusting was solely in a westerly direction, and that the entire Tablelands Complex was a thrust sheet with serpentinite at the base which 'moved approximately 150 km to the west between the Late Carboniferous and the middle Permian'. In a specific reference to the large-scale folding of the Texas beds and the postulated arcuate join of the Peel and Condamine Faults, Runnegar (1974, p. 11) suggested that their similar curvature resulted from 'folding about a steeply plunging axis, and may be due to subsequent dextral movement on the Demon Fault', ranging from 100 to 150 km. The anomalous east-west trends of the Coffs Harbour or Fitzroy Series were first noted by Voisey (1934) and confirmed by Kenny (1937) who noted that 'a remarkable feature associated with the Fitzroy Series is the peculiar strike of the beds which is persistently within a few degrees of east and west'. These trends were documented in detail by Korsch (1973, 1975, 1981), who also observed a marked change in strike compared with rocks in coastal exposures further north. Korsch (1975, 1981) found that the main belt of Coffs Harbour beds south of Woolgoolga displayed arcuate trends ranging from northwest near the Demon Fault to east—west in coastal exposures, and was characterised by northward-younging indicators. North of Woolgoolga, strikes varied from north-south to northeast, with westward younging, and Korsch (1981, p. 20) proposed that 'the overall distribution of lithologic units suggest [sic] that the Coffs Harbour Block could be a large complex syncline'. The unusual nature of this 'syncline' was noted by Korsch in his PhD thesis: 'This macroscopic fold has an almost vertical axial plane striking in a WNW-ESE direction

11

and an almost vertical fold axis' (Korsch 1975, pp. 51— 52), and he suggested that the structure was possibly an orocline. This was the first time that the large-scale folds in the New England Orogen had been related to the orocline model of Carey (1955). The situation in the mid-1970s, therefore, was that large-scale east-west offset of tectonic elements of the New England Orogen in the region of the State border had been proposed in a number of publications, and had been explained either by strike-slip faulting in an eastnortheast direction (Bryan 1925; Rod 1966; Leitch 1975) or by thrusting (Lucas 1960b; Runnegar 1974). In addition, macroscopic folding through more than 90° of steeply dipping strata of the accretionary-wedge assemblage about vertical or near-vertical fold axes had been clearly documented in both the Texas beds of southern Queensland and the Coffs Harbour beds of northern New South Wales. The folding in the Texas beds had been ascribed to dextral strike-slip faulting on a precursor of the Demon Fault, and the suggestion had been made that the macroscopic fold in the Coffs Harbour beds could be an orocline. However, no direct connection had yet been made between the offset and the macroscopic folding. Over the next decade, tectonic models relating the offset to oroclinal bending were developed by several individuals in at least three groups, for much of the time operating independently, with increasing levels of sophistication. At the Geological Survey of Queensland, W. G. Whitaker and I became convinced that the offset first noted by Bryan (1925) and the large-scale folding of the accretionary-wedge sequence first postulated by Ball (1923) had a common cause. In 1979 I presented a seminar at the University of Queensland at which I compared the double orocline of the New England Orogen accretionary wedge with the Mendocino—Idaho double orocline of Carey (1958, 1976) (Figure 7), and related its formation to dextral transform faulting like that on the San Andreas Fault. The main site of strikeslip movement was along the Yarrol Fault System and its southern continuation through the Fernvale and Pine Mountain Serpentinites west of Brisbane and the Baryulgil Serpentinite, intersecting the coast just north of Coffs Harbour (Figure 7). Although the double orocline model explained many aspects of the distribution of rock units in the New England Orogen, I concluded that the model was highly improbable. The main problem was that the sequence in the Coffs Harbour Block was considered to become younger to the northeast (Leitch et al. 1971), whereas oroclinal bending would require an opposite direction. This problem was removed by subsequent structural and sedimentological studies by Fergusson (1982a, b, c), who showed that the sequence in the Coffs Harbour Block is a stack of fault slices typical of an accretionary wedge, with overall younging to the southwest despite the common occurrence of northeast youngings in individual fold limbs. It soon became apparent, however, that the Yarrol Fault System could not have been the main line of strike-slip movement, because the repetition of forearc basin and accretionary-wedge sequences across New 33

34

35

36

12

C. G. MURRAY

SURAT BRISBANE

BASIN

/ Inglewood Goondiwindi,vtr

'Vvi

V

LVN ^ > Drake, + Tenterfield +• ^fWarialdi ^Trough

X V

TABLELANDS + y COMPLEX,4 £ 4- y V V V( + + + J^Hcv V I + + v \ + + J v!K T\ / \\ / \\ / \\ / \ / lv av/ V<C\ 4-//A v V^AM +

+

AVvv

cherts

CflOSS/ MAGLEN / C o f f s Harbour

r\

Armidale NAMBUCCA SLATES

V V V ' V V \

Syntectonic granitoids (stressed)

Tertiary volcanics

Mesozoic sediments Serpentinite

it!

Post-tectonic Permian and Triassic volcanics

+ +

Post-tectonic granitoids (unstressed)

+

+

+

+

Thrust fault L

Probable thrust fault

Figure 6 Structural map of New England (Runnegar 1974), showing a curved join between the Peel Fault and the Condamine Thrust, subparallel to trends in the Texas beds to the south, and to regional gravity trends to the north. Note also the abrupt change in strike of the accretionary-wedge assemblage north of Coffs Harbour.

TECTONIC EVOLUTION, NEW ENGLAND OROGEN

13

Figure 7 Comparison of the inverted Mendocino-Idaho coupled oroclines and San Andreas Fault (Carey 1958, 1976) and the double orocline in the accretionary wedge of the New England Orogen, presented at a seminar at the University of Queensland in 1979.

England would not then be possible.37 Therefore, after numerous attempts at cutting and displacing maps of the New England Orogen, a new north-northwest-trending dextral strike-slip fault was deduced which transported the Emu Creek and Beenleigh Blocks (and much of the remainder of the Queensland portion of the New England Orogen as well) from the north (Murray & Whitaker 1982, p. 90). A slide presented at the first New England Orogen symposium in 1982 related this Palaeozoic equivalent of the San Andreas Fault to collision of a north—south-trending oceanic spreading ridge with the offshore trench east of the New England Orogen.38 In the same symposium volume, Flood and Fergusson (1982) published the first specific description of the double orocline, which they named the Texas-Coffs Harbour megafold. They acknowledged contributions to the development of this concept by Lucas (1960b),

Runnegar (1974) and Korsch (1975), but not by Ball (1923). The formation of the Z-shaped megafold was attributed to a major northwest-southeast dextral shear couple in Late Carboniferous-Early Permian time, similar to that suggested by Evans and Roberts (1980) to explain the origin of Permian basins in eastern Australia. Powell (1984a) regarded subsurface granites near Roma and the basal volcanic sequence of the Drummond Basin as a northern extension of the Late Devonian - Early Carboniferous magmatic arc which was assumed to have existed to the west of the Tamworth Belt. This northern extension was west of and parallel to the generally accepted position of the arc along the western side of the New England Orogen in Queensland. If these two arc segments were originally continuous, their present positions indicate 1000 km or more of dextral transcurrent movement on a north-

14 C. G. M U R R A Y northwest-trending fault system now buried beneath the Bowen Basin, which would have cut the coast near Coffs Harbour (Powell 1984a, figure 224). Powell suggested that this strike-slip faulting could have been responsible for the oroclinal bending described by Flood and Fergusson (1982). Flood and Fergusson (1984) presented a more detailed mechanism for development of the megafold or double orocline based on the present plate margin through western North America, where the San Andreas Fault is the site of major transcurrent movement. Collision of an oceanic spreading ridge with a trench produced a lengthening transform fault between a northward migrating ridge-trench-fault triple junction and a southward migrating fault—fault-trench triple junction. This concept was refined and elaborated by Murray et al (1987) who proposed that the main dextral strike-slip displacement took place on the GogangoBaryulgil Fault Zone which ended in the hinge of the Coffs Harbour Orocline. Korsch and Harrington (1987) accepted the TexasCoffs Harbour double orocline (and also a more southerly orocline, the Manning Orocline, with opposite vergence), but did not explain its formation beyond noting that the formation of oroclines followed development of a major transform fault along the Mooki Fault Zone and its northern continuation. Harrington and Korsch (1987, p. 798) suggested that the oroclines were formed by complex slip movement along the PeelYarrol Fault System at about the same time as the Mooki Fault Zone was active as a dextral transform fault, and that the Devonian-Carboniferous subduction zone acted as a detachment surface limiting the oroclinal bending at depth. 'During oblique subduction and strike-slip the oroclinal ribbon could have been torn away from the upper plate and partially carried along by the lower plate' (Harrington & Korsch 1987, p. 799). The age of the oroclinal bending is controversial and has been placed in the Late Carboniferous (Murray et al 1987), the Early Permian (Fergusson 1982b; Korsch & Harrington 1987; Harrington & Korsch 1987) and the Late Permian (Collins 1990; Collins et al 1993). 39

40

41

OROGENESIS Until relatively recently, attempts to identify orogenic events in the New England Orogen were based largely on four premises: (i) orogenies were short lived; (ii) each was characterised by particular directions of folding; (iii) they operated over very large areas synchronously; and (iv) if possible they should be correlated with classic orogenies in Europe and North America such as the Caledonian and Appalachian (Browne 1949). The dominant deformational event in the New England Orogen was undoubtedly the Hunter-Bowen Movement or Orogeny named by Carey and Browne (1938). Its effects were known late last century and were documented by David (1894, p. 593) who noted that 'in the Stroud district the whole of the Carboniferous rocks are highly folded on nearly meridional lines 42

inclining slightly to west of north. The folding here may have commenced in the Permo-Carboniferous, but for the most part it took place subsequent to the deposition of the Permo-Carboniferous rocks, as the latter have been involved in the folding'. Almost all subsequent authors have recognised the importance of the Hunter-Bowen Orogeny, a notable exception being Bryan (1925, p. 48), who stated categorically that 'there was no folding of any importance in Queensland at the close of the Palaeozoic era'. He was strongly influenced by the highly folded nature of rocks as young as Cretaceous in the Maryborough Basin, and the variable degree of deformation of Permian sequences. He concluded that the only Late Permian deformation was epeirogenic uplift, and that 'along the Queensland coast a folding movement at the close of the Mesozoic era folded all the formations from Permo-Carboniferous to the Burrum Series (of the Maryborough Basin) to much the same extent' (Bryan 1925, p. 62). Detailed studies by Osborne (1950) showed the complexities of the Hunter-Bowen Orogeny, which are still the subject of current debate (Roberts & Engel 1987; Collins 1991; Collins et al 1993). Recent evidence, including seismic data, suggests that the Hunter-Bowen Orogeny was more significant in the Sydney and Bowen Basins than originally envisaged, and became younger as it progressed westwards, affecting Triassic as well as Permian strata (Leitch 1969; Korsch et al 1990b; Fergusson 1991). The existence and significance of earlier orogenic events in the New England Orogen have been much more controversial. David (1894, p. 587) was the first to describe evidence for an older orogeny in the form of the pronounced angular unconformity between the gently inclined Permian Ashford Coal Measures and the underlying steeply dipping accretionary-wedge sequence (which David included in the Gympie Series at the time). Browne (1929) came to the conclusion that this orogeny was mid-Carboniferous in age, arguing that the rocks beneath the unconformity were no younger than Early Carboniferous, and that conglomerates at the top of the Early Carboniferous marine sequence of the Tamworth Belt were the most probable expression of this disturbance west of the Peel Fault System. He explained the comparative lack of folding in the Tamworth Belt by the fact that 'the compression was relieved to the west by a series of overthrusts along what is now the Great Serpentine Belt, which saved the strata to the west from being deformed' (Browne 1929, p. xxxii). This interpretation was accepted by David (1932) and Carey and Browne (1938), who considered that the mid-Carboniferous deformation was younger than the Kanimblan Orogeny, placed at that time at the end of the Devonian. Carey and Browne (1938, pp. 604-607) referred to the mid-Carboniferous event as the Drummond Movement, a name introduced by Reid (1930) for the terminal folding event in the Drummond Basin in central Queensland. By the time Browne (1949) and David and Browne (1950) were published, the age of the Kanimblan Orogeny had been revised to mid-Carboniferous, and the term Drummond Movement became superfluous. 43

44

TECTONIC EVOLUTION, NEW ENGLAND OROGEN Lucas (1960b) and Olgers and Flood (1970b) described several localities near the New South WalesQueensland border where Permian strata unconformably overlie Carboniferous rocks. Most of these sites were similar to that at Ashford, with Permian rocks overlying the strongly deformed accretionary wedge assemblage. However, Olgers and Flood (1970b) used a postulated unconformable relationship between fossiliferous Late Carboniferous and Early Permian strata in the Emu Creek Block at Paddys Flat to place the unconformity close to the Carboniferous-Permian boundary. Subsequent investigations (McCarthy et al 1974; Murray et al 1981) have questioned the existence of the unconformity at this locality in rocks typical of the forearc basin sequence. The lower age limit of the Permian-Carboniferous unconformity, therefore, is the youngest age of the accretionary-wedge assemblage. This may extend into the Late Carboniferous (see previous discussion), indicating that deformation, which was probably diachronous, continued later than the mid-Carboniferous Kanimblan Orogeny that folded the Drummond Basin in central Queensland. The apparent restriction of the Permian-Carboniferous unconformity to the accretionary-wedge sequence is clear evidence that the deformation was the result of processes in the active subduction zone. Cawood (1982b) and Fergusson (1982a, b) were the first to recognise the characteristic imbricate thrust structure of accretionary prisms, with individual thrust slices having opposite younging directions to the complete accretionary wedge. Kirkegaard et al (1970) introduced the concept of Middle to Late Devonian folding of the Yarrol Basin sequence in central Queensland based on two lines of evidence. The first was the pronounced cleavage development in most Late Silurian to Middle Devonian rocks compared with younger sequences. The second was an angular unconformity between Middle and Late Devonian strata near Mt Morgan, which represented an erosional break during which the Middle-Late Devonian Mt Morgan Tonalite was unroofed. The unconformity was not merely a basin-edge effect, because the magnitude of the erosional break increased eastwards together with the thickness of the post-Middle Devonian sequence, as revealed by the widespread occurrence of pebbles of Early or Middle Devonian limestone in Early Carboniferous strata, and probable unconformities between Early Devonian and Early Carboniferous and even Early Permian rocks (Kirkegaard et al. 1970). The significance of this Middle to Late Devonian deformational event was questioned by Morand (1993a), who found that cleavage development was not controlled by age but by structural position, being greatest close to the Yarrol Fault System and in stratigraphically lower units that were subsequently uplifted by thrusting. Cleavage and fold axes have similar orientations in Devonian, Carboniferous and Early Permian rocks, and Morand suggested that the structures were consistent with a single folding event, the mid- to Late Permian Hunter-Bowen Orogeny.

15

TECTONIC MODELS

Most authors during the first half of this century described the tectonic evolution of eastern Australia simply in terms of geosynclinal deposition and subsequent deformation and uplift. Very rarely was any causal mechanism proposed. From David (1911, pp. 38-39) and Jensen (1912, pp. 172-174) onwards, it was recognised that the compressional movements which folded the rocks of the New England Orogen, particularly the Hunter-Bowen Orogeny, were directed from the east towards the present continent (Carey & Browne 1938, p. 608). The main focus of debate was whether the Australian continent had originally been much smaller, and had grown to its present size by accretion to the east, or whether it had been much larger, and had subsequently collapsed leaving behind continental fragments such as New Zealand, New Caledonia and Fiji (Bryan 1944; Fisher 1974). For the Tasman Geosyncline in general and the New England Orogen in particular, these opposing concepts implied deposition in an oceanic environment marginal to the continent, or in an intracontinental trough, respectively. The hypothesis of eastwards continental growth was stated most explicitly by Jensen (1912), who concluded from a series of palaeogeographic maps that 'throughout the Palaeozoic periods the ancestral Australian Continent was continually adding to itself on the eastern side' (Jensen 1912, p. 167). The locus of thickest marine sedimentation moved progressively to the east as older deposits were uplifted to form dry land to the west. By Carboniferous time, the continent had extended so far to the east that 'it is not unlikely that portions of Australia were connected with New Zealand, Fiji, and New Caledonia' (Jensen 1912, p. 157). Contrary to the implication by Bryan (1944, p. 54), Jensen did not believe that Australia originally consisted of a relatively small Precambrian craton in the west which grew eastwards by accretion in the Palaeozoic, as 'rocks of [Precambrian] age must underlie all later formations' (Jensen 1912, p. 151). As pointed out by Fisher (1974, p. 8), Jensen 'was always a firm believer in the drifting of continents', and he certainly embraced this concept with enthusiasm at an early stage. He believed that 'Western Australia ... formed part of a Pre-Cambrian landmass which extended in a north-west direction to Arabia and Abyssinia and south-wards to Antarctica. This great landmass was probably fairly persistent throughout Palaeozoic times, since Africa, India and Australia had a similar flora up to the end of the Palaeozoic, namely the Lepidodendron flora of the Carboniferous and the Glossopteris flora of the Permo-Carboniferous' (p. 152), and that 'during the Permo-Carboniferous cataclysms commenced the break up of the great Gondwana continent and the separation of Australia from India and South Africa was commenced' (p. 158). Jensen also applied the term 'Atlantic type' to the coasts of Western Australia, India, Arabia, and East Africa, which represent fragments of a supercontinent broken up by faulting similar to that 'in progress at the present time in the Rift Valley of Africa' (Jensen 1912, p. 171). 45

16

C. G. M U R R A Y

The concept of eastwards growth came to be generally accepted, as indicated by David (1932, p. 49): 'as regards the evolution of the Australian continent, the point has frequently been stressed that after the disappearance of the Proterozoic Nullagine seas, deposition was mostly restricted to what is now the eastern portion of the continent, and that there was a steady gain of land eastwards at the expense of the sea'. According to Bryan (1944, p. 53), the opposing idea that the Australian continent became progressively smaller with time was first proposed by Schuchert (1916). The missing eastern portion, lost by fracturing and subsidence, was comparable in size to the present continent, being 'a land about 1,800 miles east and west and 2,200 miles north and south'. These two hypotheses required very different interpretations of the Tasman Geosyncline. 46 The first envisaged the Tasman Geosyncline (including the New England Orogen), through most of its history at least, as a marginal sea filled by erosion of the adjacent continent, and moving gradually eastward with time. For the second theory, the Tasman Geosyncline was an intracontinental seaway with a relatively fixed position, envisaged by Schuchert (1916) and Bryan (1944) as a large basin similar to the Mediterranean. Andrews (1938, p. 182) was a proponent of the continental growth (or welding) hypothesis, describing 'the steps whereby great arcuate belts of Silurian, Devonian, Carboniferous, and Permian rocks [in New England and eastern Queensland] were attached successively, as dry land, to the growing continent of Australia'. However, his palaeogeographic maps for the Devonian (Figure 3) and Early Carboniferous (his figure 6) show the New England or Tasman Geosyncline as 'an extraordinarily narrow, ribbon-like feature, confined within the present continental limits' (Bryan 1944, p. 54). It was bordered on the east by the landmass of Tasmantis, the greater proportion of which subsequently sank beneath the sea. For the New England Orogen, therefore, Andrews' ideas were little different from those of Schuchert for the Tasman Geosyncline.47 Voisey (1959b) favoured eastwards growth of the Australian continent by accretion of marginal volcanic geosynclines and island arcs. This 'easterly advance was made not by the successive additions of deposits on continental shelves, but by the accumulation of sediments in wide depressions or geosynclines with the development of ridges or volcanic islands within them. Subsequent leaping forward of the continental margin to the island arcs took place and the intervening region, which had formerly separated them, was consolidated through the development and deformation of secondary geosynclines' (Voisey 1959b, pp. 188-189). He attempted to explain the development of the New England Orogen in terms of the geosynclinal types defined by Kay (1951), but found severe problems in applying these to eastern Australia: 'It is ... current practice to call sedimentary sequences without volcanics, miogeosynclinal; and volcanic ones, eugeosynclinal. Actually it is a most difficult matter to draw a boundary in eastern Australia between the two types on this basis, as volcanic products are wide-spread and are associated

with a variety of sedimentary types. In principle, I prefer to think of the eugeosyncline as lying seawards of a volcanic island arc, possibly on a simatic floor and the miogeosyncline occupying a position between the arc and the craton, possibly, but not necessarily, on a sialic floor' (Voisey 1959b, p. 189). Voisey (1959a, b) believed that rocks of the New England Orogen regarded as Early Palaeozoic (now largely included in the accretionary-wedge assemblage) were deposited in a marginal eugeosyncline, the Brisbane Eugeosynclinal Belt, which was the site of essentially continuous sedimentation from Precambrian time. These Early Palaeozoic rocks were overlain by 'a conformable sequence of sediments from Lower Devonian to the end of Permian times totalling several miles in thickness' (Voisey 1959b, p. 194), indicating the continued existence of the geosyncline, the New England Eugeosyncline, into Late Palaeozoic time (Figure 8). Voisey (1959a) maintained the interpretation of a Devonian and Carboniferous landmass in central New England ('Ancestral New England' of Andrews 1938), which he envisaged as an island arc similar to the Japanese islands. Because of his revised definition of miogeosyncline and eugeosyncline, Voisey (1959b) was forced to include the entire New England Orogen in his New England Eugeosyncline, and showed two east-dipping serpentinite belts (Peel Fault System in the west and Baryulgil Serpentinite in the east) in the deformed and uplifted geosynclinal deposits (Figure 8). Thus this model completely failed to explain the major problem of the New England Orogen, the difference in stratigraphy and structure across the Peel and Yarrol Fault Systems. Possibly for this reason, no further attempts were made to interpret the tectonic history of the New England Orogen in terms of classical geosynclinal theory. Plate-tectonic models were first proposed for eastern Australia (including the New England Orogen) in the early 1970s. The most influential of these models, as far as I was concerned, was the detailed review by Marsden (1972) for the Devonian. His interpretation was hampered by the fact that the rocks of the accretionarywedge assemblage were still considered to be of Early Palaeozoic age. He regarded these as 'older trench and deep ocean deposits' (Marsden 1972, p. 155). Marsden considered that Late Silurian to Middle Devonian marine volcanic and sedimentary rocks of the Yarrol Basin and the Tamworth Trough, including widespread coralline limestones, formed in a volcanic island-arc environment related to a west-dipping subduction zone. Possible deformation between the Middle and Late Devonian may have marked a new phase of plate movement and westward subduction, producing a continental margin arc of Andean type which was almost superimposed on the earlier arc and which continued into the Carboniferous. The interpretation of Marsden (1972), with the significant difference that the accretionary-wedge assemblage was coeval, became the basis for the Devonian-Carboniferous interval of subsequent platetectonic models, including those of Geological Survey of Queensland geologists (Murray 1975, 1986, 1987b;

TECTONIC EVOLUTION, NEW ENGLAND OROGEN

17

Figure 8 Model for the evolution of the New England Orogen (Voisey 1959b) applying the geosynclinal classification of Kay (1951).

Day et al 1978; Murray et al 1987). Inherent in this interpretation was the implication that Late Silurian to Middle Devonian rocks, grouped together as the Calliope Island Arc, and regarded as forming in an oceanic environment, could be an exotic terrane.48 This was first explicitly stated by Veevers et al (1982) and Powell (1984b). The Middle to Late Devonian deformation postulated for central Queensland by Kirkegaard et al. (1970) could then reflect the collisional event during accretion. A similar model has recently been proposed for the southern New England Orogen, where the Devonian Gamilaroi terrane (Tamworth Group) has been interpreted as a rifted oceanic island arc which was accreted to the Australian continent in Late Devonian time (Flood & Aitchison 1992; Aitchison & Flood 1995).49

An alternative view that the Late Silurian to Middle Devonian rocks of the Calliope and Craigilee Blocks were related to a continental margin arc, rather than to an island arc, was proposed by Henderson (1980). He suggested that an active, cordilleran-type margin was located along the entire present coastal zone of Queensland at that time, linking the Hodgkinson-Broken River Orogen in the north with the New England Orogen.50 Morand (1993b) argued that the Calliope Volcanic Assemblage either formed on thin continental crust, or was a mature island arc located near a continental source of detritus, based on the presence of silicic volcanics and the local occurrence of quartzose sediments.51 Scheibner (1973) outlined a plate-tectonic model for New South Wales which related the development of both the Lachlan Fold Belt and the New England

18

C. G. M U R R A Y

Orogen to the episodic formation of west-dipping subduction zones that migrated eastwards from Late Silurian to Permian time. During this interval, the New England Orogen was the leading edge of the Australian continental plate. Progressively younger volcaniclastic flysch deposits, locally incorporating scraped-off oceanfloor sediments and volcanics, accumulated at the continental margin, represented by the Late Silurian Middle Devonian Woolomin beds in the west, through the Late Devonian - Early Carboniferous TexasSandon beds, to the Early Permian Nambucca beds in the east.52 The plate-tectonic model proposed by Leitch (1975) was the first specifically for the New England Orogen. This model suffered from the same deficiency as that of Marsden (1972), in that Leitch still accepted an Early Palaeozoic (Ordovician-Silurian) age for the Woolomin beds and equivalent rocks in Queensland.53 A volcanic chain along the entire western edge of the New England Orogen was formed by westward subduction commencing in the Early Devonian. Thick sequences of volcaniclastic sediments were laid down in a forearc basin (Tamworth Trough and Yarrol Basin). To explain the offset of tectonic elements between the New South Wales and Queensland parts of the fold belt, Leitch proposed that the arc-forearc basin system was broken in the Late Devonian by the east-northeast-trending Tenterfield Fault. North of the fault, the arc split, and the eastern part migrated eastward, forming an intra-arc basin behind it in which the Late DevonianCarboniferous Texas beds were deposited (Figure 5). This basin was closed, and the Texas beds deformed, by reversal of movement on the Tenterfield Fault in latest Carboniferous time.54 These early plate-tectonic models led to the development of what came to be known as the "consensus" model, which envisaged the New England Orogen as a convergent plate margin related to a west-dipping subduction zone through Devonian and Carboniferous time. The only radically different hypothesis was presented by Harrington and Korsch (1985a), who proposed that the Devonian-Carboniferous arc along the western edge of the New England Orogen formed above an eastdipping subduction zone, and that 'the Devonian sediments of the Tamworth Belt were deposited as a clastic wedge in a back-arc basin which passed eastwards to a deeper ocean basin in which the Woolomin Association was deposited' (Harrington & Korsch 1985a, p. 168).55 Flood (1988) quite rightly criticised the consensus model as being too simplistic, and pointed out several specific problems. However, while the terrane concept has provided the means for more objective scientific analysis of the tectonic evolution of the New England Orogen, and has enabled new understanding of the preLate Devonian history in particular, the consensus model still remains unchallenged (rightly or wrongly) as an explanation for the basic framework of the New England Orogen. A feature of the consensus model was the recognition that the Devonian-Carboniferous convergent plate margin was profoundly modified in the Early Permian,

which was characterised by extensional basin formation, probably in a backarc setting. Although sequences such as that in the Gympie Province may represent a volcanic arc to the east, no associated forearc basin or accretionary wedge assemblages of Early Permian or younger age have yet been recognised.

CONCLUSIONS The history of ideas on the development of the New England Orogen reflects very precisely the maturing of tectonic concepts over the last century, and also the continual gathering of new geological data. However, it was only in 1959 that the first attempt was made to analyse the geological evolution of the New England Orogen in terms of the most elegant expression of the prevailing tectonic model of the time, the geosynclinal theory as stated by Marshall Kay. The dominance of plate-tectonic models over the last 25 years is not just an indication of the general acceptance of this hypothesis, but, more importantly, is a result of the fact that these models explain the development of the New England Orogen much more satisfactorily than any fixist concept. When the method of multiple working hypotheses is applied to the New England Orogen, plate tectonics is a clear winner. The real breakthrough in understanding the New England Orogen was the realisation that plate tectonics was consistent with, and in fact demanded, the juxtaposition of strongly deformed rocks of a subduction complex or accretionary wedge against relatively undeformed coeval forearc basin strata at convergent plate margins. Up until this time, well-bedded and richly fossiliferous Devonian and Carboniferous strata west of the Peel-Yarrol Fault System were considered to be younger than the more strongly deformed and metamorphosed rocks to the east, despite the fact that the eastern sequence contained limestones with fossils as young as Carboniferous. Uniquely in the world, the presence along the entire strike length of this accretionary-wedge sequence of oolitic arenite horizons which could only have been derived from Early Carboniferous oolitic limestone banks in the forearc basin to the west, was crucial in establishing a Late rather than Early Palaeozoic age, as subsequently confirmed by radiolarian studies. Some recently proposed features of current tectonic models of the New England Orogen, such as large-scale strike-slip faulting and oroclinal bending, were first recognised more than 70 years ago. The lack of acceptance of these concepts at that time was not because they were incorrect, but because they were not consistent with prevailing dogma.

ACKNOWLEDGMENTS This paper is based on the work of many geologists over a period of a century and a half. In such a short review, it has only been possible to mention or refer to some of these. I have endeavoured to highlight contributions

TECTONIC EVOLUTION, NEW ENGLAND OROGEN which were at odds with the consensus thinking of the day, but which in many cases were reinvented and accepted at a later time when tectonic models had changed. In this, I have probably shown bias towards Queensland workers, because of my familiarity with geology, events and literature north of the State border. Bill Whitaker commented on an early version of this manuscript, and it was also improved as a result of reviews by Erwin Scheibner and particularly Peter Cawood. The figures were drafted by Lesley Blight, Graphical Services Unit, Department of Mines and Energy, and publication was approved by the Director, Geological Survey of Queensland.

NOTES 1 The choice of name reflected the geographical extent of the Late Permian orogeny. It did not imply that the deformational event affected the Bowen Basin, because this name was not yet in use, the structural element being called the Great Syncline by Reid (1930). 2 Throughout this review, I have largely followed original authors in their usage of terms such as Series and beds. 3 Andrews (1938, p. 174) confused the nomenclature by stating that sedimentation of Silurian to Early Carboniferous age 'appears to have proceeded conformably within the Tasman Geosyncline, a trough extending from the Hunter River area in the south to Cape York Peninsula in the north'. This is certainly not the generally accepted usage of this term as introduced by Schuchert (1916). 4 The problems faced by geologists last century in making sense of the complex geology of the New England Orogen are exemplified by David's interpretation of relationships between the Gympie Series and Lepidodendron australe beds of the Tamworth Belt to the west. The relative age of these sequences was still not resolved by the palaeontologists of the day, and David stated that '[the Gympie Series in the Vegetable Creek district] are far more strongly folded than the Lepidodendron australe beds which dip off them and may be unconformable to them' (David 1894, p. 587). 5 The outcrops east of the Peel Fault System correlated with the Tamworth beds by Benson (1913) are now included in the olistrostromal Wisemans Arm Formation (Leitch & Cawood 1980; Cawood 1982a). 6 In his article describing the Rocksberg Greenstones, however, Mathews (1960, p. 81) considered this unit to be of Early Palaeozoic age. 7 When Crook published his note, this specimen had already been destroyed by a fire at the University of New England. 8 This dogma that deformed 'unfossiliferous' rocks of the New England Orogen must be of Early Palaeozoic age also operated very effectively in Queensland. In 1930, an Early Permian fauna including Martiniopsis was found in the Mary River

19

Valley at Cambroon (Denmead 1960a; Hill 1961), in rocks traditionally included in the Amamoor Series of Dunstan (1916) which were correlated with the Curtis Island Series and Brisbane Schists. Although the extent of the Early Permian sequence was clearly defined in an unpublished report by Geological Survey of Queensland geologist J. E. Ridgway in 1936, it was not differentiated on geological maps of the Moreton district published in 1955, and the Gympie 1:250 000 sheet area published in 1975. 9 Because no Late Carboniferous conodonts had ever been found in eastern Australia before, because these rocks were accepted as Early Palaeozoic (probably Silurian), because the conodonts were poorly preserved, and because he was an expert on foraminifers, not conodonts, Palmieri (1967) initially interpreted the conodont fauna as Late Silurian-Early Devonian. As soon as the Late Carboniferous age was established, the limestones and their host rocks were separated from the Wondai Series (renamed the Maronghi Creek beds), and mapped as an undifferentiated Carboniferous to Permian unit (Murphy et al. 1976). Murray (1987a, 1990) cited new evidence that the limestone-bearing sequence was interbedded with the Wondai Series, and suggested that these rocks, together with the similar Good Night beds to the north, could be the youngest part of the accretionary-wedge assemblage. 10 When I discussed the age of the limestones in the Texas beds with her several years ago, Dorothy Hill was adamant that the identification of the Devonian coral Alveolites from 'Cooinoo' (see Lucas 1960a, p. 140) was correct, despite the fact that Strusz listed Early Carboniferous corals from this locality. 11 The combination of the Neranleigh and Fernvale Series of Denmead (1928) as the NeranleighFernvale Group by Bryan & Jones (1950) was unfortunate, as the Fernvale Series in particular was a distinctive unit easily correlated with the Woolomin beds. 12 Although initially very sceptical of the value of ooliths and elliptical crinoids in determining stratigraphic age, Roberts displayed admirable scientific discipline in searching his comprehensive collections of Tamworth Belt fossils (successfully) for elliptical crinoid stems. 13 The restricted age range attributed to the radiolarian faunas may be partly due to the lack of a defined radiolarian biostratigraphy for the pre-Middle Devonian. 14 The Doonside and Shoalwater Formations did not meet the formal requirements of the Australian Code of Stratigraphic Nomenclature at this time, because both western and eastern boundaries of the Curtis Island Group were not stratigraphic contacts. However, strict adherence to the code, and use of the more informal term beds for the sequence, would have made the lithological variation much less apparent. 15 Most previous authors had accepted a deep-marine

20

C. G. M U R R A Y

environment for the accretionary-wedge assemblage. David (1894, p. 594) proposed that 'the red jasperoid shales of the Nundle and Bingara districts with the associated serpentines may represent altered abysmal [sic] deposits'. 16 J. E. Thompson, who was Bureau of Mineral Resources supervisor of the joint mapping party, suggested during the regional mapping program that the Curtis Island Group represented deep-water equivalents of the Devonian and Carboniferous marine shelf sequence of the Yarrol Basin to the west. 17 After relative isolation during long field seasons in 1965—1967, it was a tremendous culture shock to go to Princeton University in 1968, where Harry Hess and Fred Vine were reminiscing about sea-floor spreading and magnetic stripes, and Jason Morgan, having completed his contribution to the platetectonic model, was considering mantle plumes as driving forces. 18 There seems little doubt that it was James Smith himself who perpetrated the hoax. Despite the endorsement of R. L. Jack (Jack & Etheridge 1892, p. xiii), 'Smith was not a particularly careful collector, and a number of his specimens in the Survey collections are labelled with localities which are obviously incorrect' (Whitehouse 1928, p. 441). Smith also may have been something of a practical joker, as revealed by his description of the geology of the Rockhampton region: 'In my opinion, the formation of Central Queensland was a deep bed of quartz. There was a superabundance of silica in our system in this part of the world for which there was not room in the three-fold mixture of the granite, just as we find the mountain of Assynt at the present day. This bed was torn by earthquakes, cracked by heat, tossed and pounded by tumultuous waves into rounded and oval seashore pebbles of all sizes, from that of a boy's marble to a large water-melon. These were sorted and sifted by the strong currents of the first impetuous tides, spread out into a stratum, and cemented together by an oxide, and now form the most obdurate, stubborn stuff man ever put a tool on. It extends from Keppel Bay to Morinish, and from the Fitzroy River at Lilymere to Clermont, a distance of 300 by 60 miles. It was subsequently torn and rent into a thousand gaping fissures, canted on its edge in parallel slices, with a general north and south trend, and now forms the great conglomerate dykes of Central Queensland' (Lukin 1888, pp. 45-46). 19 Once the Late Palaeozoic age of the accretionarywedge assemblage had been firmly established, the identification of this possible graptolite by an acknowledged expert in R. A. Keble became something of a problem. However, I later discovered that even experts are capable of error. In 1930 Keble submitted a paper to the Royal Society of Queensland on graptolites in samples from a spoil dump at Mt Isa, which he identified to species level and regarded as Silurian in age. 'General disagreement with the determination of the specimen as graptolitic

was expressed by several members' (Anonymous 1931), which is hardly surprising given that the shales at Mt Isa are Mesoproterozoic. Additional specimens were collected which showed 'the pseudograptolitic character of the hydrated and oxidised marcasite crystals on the joint planes and bedding planes of the country rock associated with the Rio Grande silver lead lode at Mt. Isa. A specimen of a bore core from depth showed the original pyritic character of the long, slender, and occasionally plate-like crystals very well, and serves to explain more clearly the true nature of the graptolite-like markings' (Anonymous, 1933, p. xi). 20 When I went on a university excursion to this locality, O. A. Jones, who was present on the 1944 excursion, referred to the rocks at this locality as the 'NBG slates' — N o Bloody Graptolites! 21 Denmead once told me that he shipped these 'eurypterid' fragments to Brisbane in boxes labelled 'Gympie gold specimens'. Not surprisingly, they never arrived, and became just another example of lost fossil specimens from the 'unfossiliferous' accretionary-wedge sequence. The statement that they were 'suggestive of eurypterid fragments' seems incredible. 22 In 1934, Ball found marine Permian fossils near the Brisbane River at Northbrook, in rocks which appeared 'to belong to the Fernvale series of the Brisbane schists' (Ball 1934). These rocks were later defined as a separate unit, the Northbrook beds. 23 The specimen was subsequently discarded by his mother when tidying the boy's room — only a plaster cast remains. 24 The Late Ordovician limestones were interpreted by Scheibner and Glen (1972, p. 10) as 'sedimentary cappings of guyots and seamounts' scraped off descending sea floor in an oceanic trench east of the present position of the Peel Fault System. 25 The presence of Ordovician rocks older than the Trelawney beds had been recorded in this area by Packham (1969b, p. 231) on the basis of graptolite faunas. 26 At about the same time, J. C. Aitchison recognised this possibility because poorly preserved radiolarians extracted from the Early Palaeozoic sequence were 'quite different to those found in the overlying Tamworth Group' (Aitchison & Ireland 1995, p. 19). 27 None of these models seems to provide a reasonable explanation for the observed distribution of Early Palaeozoic rocks. If the ophiolitic complexes and the Cambro-Ordovician volcaniclastic sediments are related, then any model must be compatible with the unconformity exposed in Copes Creek between the latter sequence and the overlying Early Devonian Tamworth Group (Cawood, 1976, p. 318). This would create problems for any model involving tectonic emplacement of the Early Palaeozoic assemblage, or tectonic transport of younger strata over them. The interpretation of the Tamworth Group itself as an exotic terrane (Aitchison & Flood,

TECTONIC EVOLUTION, NEW ENGLAND OROGEN

1995) is difficult to reconcile with simple models of accretion of the Early Palaeozoic rocks to the Australian continent in view of the relative tectonic positions and ages of the two sequences. Apart from the problem listed by Fukui et al (1995, p. 1019) regarding evidence for Ordovician subduction in the Lachlan Fold Belt, the possibility of preserving the almost continuous linear Great Serpentine Belt after recycling through two accretionary processes seems extremely remote. 28 Bryan suggested that the fact that the displacement had not been recognised previously was due to its location near the State boundary, because 'any discordance between the New South Wales and the Queensland formations would naturally be considered merely a reflection of differences of opinion between the geologists of the two States. Another explanation is that the break is to some extent obscured by more recent deposits' (Bryan 1925, p. 57). 29 One discrepancy in Bryan's analysis (Figure 4) was the occurrence of Carboniferous strata recognised by Ball (1923) in the Texas area, which are now regarded as part of the accretionary-wedge assemblage. 30 Geologists have always seemed at a loss to oppose quantitative arguments put forward by physicists, such as those of Jeffreys (1929, pp. 304-305) based on strength and rigidity of the Earth and the viscosity of crustal and lithospheric layers. An earlier example was the calculation of the age of the Earth from its cooling history by Lord Kelvin. 31 Strangely, this is the only proposal for major fault displacement within or adjacent to the New England Orogen which has been seriously challenged. Cherry (1989) disputed the proposed strike-slip movement on the basis of a provenance linkage between a quartzite clast in the Currabubula Formation of the New England Orogen containing Cyrtospirifer and probable source rocks of the Lambie facies in the adjacent Lachlan Fold Belt. The alternative source proposed by Korsch et al (1990a, p. 42) from the Drummond Basin in Queensland, which would be compatible with large-scale fault offset, can be definitely ruled out because Late Devonian sediments of the Drummond Basin are volcaniclastic, not quartzose (Olgers 1972). 32 Ball inadvertently reversed the positions of these outcrops. In fact Bracker Creek lies southwest of the Graysholm-Yuaraba (currently spelt GreysholmYuraraba) district. 33 Unfortunately, these more speculative concepts were not included in the subsequent published version (Korsch 1981). Korsch (1975, figure 79) was more emphatic in identifying the large-scale curvature of the southern part of the Tamworth Belt (including the Hastings Block) as an orocline. 34 As it was contained in a PhD thesis to which access was restricted, this information was not generally available at this time. 35 Carey (1958, p. 192) used the term 'coupled oroclines' for two oroclines in the same orogen

21

which 'may have dextral or sinistral coupling, implying respectively right hand or left hand wrench movement'. 36 This interpretation of the Coffs Harbour Block was disputed by Morris (1991), who accepted that the overall younging direction was to the northeast, making oroclinal bending unnecessary as a means of explaining the distribution of rock units in New England. Abandonment of the orocline model, however, requires an alternative explanation for the observed large-scale folding and changes in trend near Texas and Coffs Harbour. 37 For the same reason, the models of Bryan (1925), Rod (1966) and Leitch (1975) invoking eastnortheast-trending displacement cannot be valid. 38 Thisfigurewas not prepared in time for inclusion in the published paper submitted prior to the symposium (Murray & Whitaker 1982). 39 Runnegar (1984) also proposed a ridge—trench collision to account for the displacement and oroclinal bending. 40 Cawood (1982c) also proposed mid-Permian oroclinal bending and sinistral strike-slip faulting (along the Peel Fault System and other faults) to explain the curvature of the southern Tamworth Belt, the structural pattern in the adjacent accretionary wedge, and the apparent offset of the Hastings Block relative to the remainder of the southern New England Orogen (also see Cawood & Leitch 1985). 41 Determination of the precise mechanism proposed by Harrington and Korsch (1987) is difficult because their two figures are inconsistent with each other and with the text. The main problem is that their figure 2 does not continue south of the Coffs Harbour Orocline, and therefore does not explain the formation of the Manning Orocline with opposite vergence. 42 Browne (1949, pp. 40-41) stated that it is of interest to note that quite a number of our tectonic epochs can be correlated with some degree of probability — or at least of plausibility — with those of Europe and North America. ... it is surely something more than coincidence that they show so many synchronisms with those of the Northern Hemisphere. It would really seem that our continent vibrated and our geosynclines were wrinkled more or less in unison with those of Europe and North America'. He correlated twelve Australian tectonic epochs with events in the Northern Hemisphere. 43 The importance of the Cretaceous compressional event was recently reiterated by Elliott (1993). 44 Despite the fact that David eventually agreed with the view that there was a mid-Carboniferous deformation of the New England Orogen (David 1932; David & Browne 1950), it is apparent that for many years he believed in a Carboniferous—Permian unconformity, possibly based partly on his observations at Ashford. Browne (1949, p. 39), reviewing Australian orogenic epochs, expressed the view that 'much trouble and controversy could have been avoided in the last 20 or 30 years if only there was an angular unconformity, as I was taught when a

22

C. G. M U R R A Y student, between Carboniferous and Permian. But unfortunately there isn't! Sir Edgeworth David's conception of such a break was based mainly on a traverse across a line of junction between marine Permian and freshwater Carboniferous near Lochinvar, New South Wales, but later and more detailed investigations in the same area at a point where the sequence is fuller have shown that there is a conformable passage. Elsewhere, it is true, there are many unconformities between Permian and Lower Carboniferous beds, and these, no doubt, helped to foster the illusion of a CarboniferousPermian break, but they are due to Kanimblan folding. In Queensland and New South Wales, wherever there is a section showing uppermost Carboniferous and lowest Permian beds, there is no sign of angular discordance'. 45 What makes these statements by Jensen more remarkable is that they were published in the same year as Wegener's first statement of his theory of continental drift (Wegener 1912a, b), and three years prior to the publication of the first edition of his book (Wegener 1915). 46 The New England Geosyncline was not differentiated within the Tasman Geosyncline at this time, although it was recognised that the belt of thickest Devonian and Carboniferous sedimentation was positioned along the eastern margin of the present continent, east of the main areas of Ordovician and Silurian sedimentation. Jensen's Carboniferous palaeogeographic map of Australia (Jensen 1912, figure 6) clearly shows the New England Orogen as a region of 'heavy sedimentation'. 47 Even Jensen (1912, p. 156) considered that the Devonian basins of eastern Australia were 'probably trough subsidence areas (senkungs-feldter) in the Silurian platform', a concept which seems remarkably similar to that of Andrews. Jensen's palaeogeographic map for the Devonian, however, shows no eastern landmass. 48 Marsden suggested that the Late Silurian to Middle Devonian volcanic arc could have formed on oceanic crust, but was located close to its present position (Marsden 1972, p. 155, figure 4). 49 Morand (1993a) rejected an exotic origin for the Calliope Volcanic Assemblage, and concluded (p. 269) that 'an unconformity between Middle Devonian and Late Devonian volcanic arc and forearc basin rocks in the Rockhampton region probably resulted from local changes in plate motion at the subduetion zone and was not a major collisional event. This implies that the Calliope Volcanic Assemblage is not an exotic terrane, and that the Devonian history of the northern New England Fold Belt was similar to that of the southern part'. I find this statement intriguing in light of the fact that an exotic model has been proposed for the Gamilaroi terrane, with accretion in the Late Devonian (Flood & Aitchison 1992; Aitchison & Flood 1995), and has not been questioned, despite the fact that the conformable relationship of the Tamworth Group with the overlying Late Devonian strata seems to

preclude any collisional deformation at all. 50 Recent interpretations of the Hodgkinson - Broken River Orogen favour an extensional basin model for Late Silurian - Middle Devonian time, rather than a convergent plate margin (Withnall et al. 1987; Withnall & Lang 1990, 1993). 51 Any interpretation of the palaeogeographic setting of the Calliope Volcanic Assemblage must explain the fact that the entire exposed sequence appears to have been deposited in a submarine environment, and that coralline limestones occur in the Middle Devonian Capella Creek Formation only a few kilometres east of the Mt Morgan Tonalite, regarded by Morand (1993b, p. 15) as a high-level arc pluton which is part of the assemblage. Geochemically, the Mt Morgan Tonalite is an island arc pluton with primary-mantle-melt patterns similar to those of rocks from New Britain (Allen et al. 1994). 52 Several aspects of Scheibner's model differ from current interpretations, including the existence of an emergent area, the New England Arch, east of the Peel Fault System as early as mid-Carboniferous time, and acceptance of the Nambucca beds as the youngest part of the accretionary-wedge assemblage. 53 The Curtis Island Group in Queensland was considered to have been deposited on 'the continental rise off an Atlantic-type continental margin', and the Woolomin beds 'perhaps on an abyssal plain off a consuming or recently sealed plate boundary' (Leitch 1975, p. 142). These rocks were later 'plastered against and beneath the inner trench wall' associated with Devonian subduction (Leitch 1975, p. 143). 54 Like Scheibner (1973), Leitch continued the subduction-accretion model into the Early Permian, and identified rocks of this age (presumably the Nambucca beds) as trench fill which was 'fed at an increasing rate into the trench, much of it by slumping' (Leitch 1975, p. 144). 55 The inconsistency noted by Flood (1988) between this interpretation and the 'consensus' model accepted by Harrington and Korsch (1985b) and Korsch and Harrington (1985) is due to the fact that the original manuscript for Harrington and Korsch (1985a) was submitted at the beginning of 1981. The result of the extended delay in publishing what was obviously a controversial paper was that by the time it appeared, the authors had already abandoned some of the basic concepts in it. REFERENCES AITCHISON J. C. 1988a. Late Paleozoic radiolarian ages from

the Gwydir terrane, New England orogen, eastern Australia. Geology 16, 793^-795.

AITCHISON J. C. 1988b. Radiolaria from the southern part of

the New England Orogen, eastern Australia. In: Kleeman J. D. ed. New England Orogen — Tectonics and Metallogenesis, pp. 49-60. Department of Geology and Geophysics, University of New England, Armidale.

TECTONIC EVOLUTION, NEW ENGLAND OROGEN

AITCHISON J. C. 1988c. Early Carboniferous (Tournaisian)

radiolaria from the Neranleigh-Fernvale beds, Lake Manchester, Queensland, Australia. Queensland Government Mining Journal 89, 240-241. AITCHISON J. C. 1990. Significance of DevonianCarboniferous radiolarians from accretionary terranes of the New England Orogen, eastern Australia. Marine Micropaleontology 15, 365—378. AITCHISON J. C. & FLOOD P. G. 1990. Early Carboniferous radiolarian ages constrain the timing of sedimentation within the Anaiwan terrane, New England orogen, eastern Australia. Neues Jahrbuch fur Geologie und Palaontologie Abhandlungen 180, 1—19. AITCHISON J. C. & FLOOD P. G. 1995. Gamilaroi terrane: a Devonian rifted intra-oceanic island-arc assemblage, NSW, Australia. In: Smellie J. L. ed. Volcanism Associated with Extension at Consuming Plate Margins, pp. 155-168. Geological Society of London Special Publication 81. AITCHISON J. C., FLOOD P. G. & SPILLER F. C. P. 1992a. Tectonic setting and paleoenvironment of terranes in the southern New England orogen, eastern Australia as constrained by radiolarian biostratigraphy. Palaeogeography, Palaeoclimatology, Palaeoecology 94, 31—54. AITCHISON J. C. & IRELAND T. R. 1995. Age profile of ophiolitic rocks across the Late Palaeozoic New England Orogen, New South Wales: implications for tectonic models. Australian Journal of Earth Sciences 42, 11—23.

AITCHISON J. C., IRELAND T. R., BLAKE M . C. JR. & FLOOD

P. G. 1992b. 530 Ma zircon age for ophiolite from the New England orogen: oldest rocks known from eastern Australia. Geology 20, 125-128.

ALLEN C. M., CHAPPELL B. W., WOODEN J. L. & WILLIAMS I.

S. 1994. Ages, compositions, and sources of the Urannah Batholith. In: Henderson R. A. & Davis B. K. eds. New developments in geology and metallogeny: northern Tasman Orogenic Zone, Extended Conference Abstracts, pp. 107—110. James Cook University, Economic Geology Research Unit Contributions 50. ANDREWS E. C. 1938. The structural history of Australia during the Palaeozoic. Journal and Proceedings of the Royal Society of New South Wales 71,118-187. ANONYMOUS 1930. Abstract of proceedings, 29th April 1929. Proceedings of the Royal Society of Queensland 41, vi—viii. ANONYMOUS 1931. Abstract of proceedings, 28th April 1930. Proceedings of the Royal Society of Queensland 42, ix—xii. ANONYMOUS 1933. Abstract of proceedings, 27th June 1932. Proceedings of the Royal Society of Queensland 44, x-xi. ANONYMOUS 1945. Abstract of proceedings, 25th September 1944. Proceedings of the Royal Society of Queensland 56, xi—xii. ANONYMOUS 1947. Abstract of proceedings, 29th April 1946. Proceedings of the Royal Society of Queensland 58, vii—viii. BALL L. C. 1923. Notes on manganese on the Darling Downs. Queensland Government Mining Journal 24, 457-459. BALL L. C. 1934. Geology of the Brisbane Valley area. Queensland Government Mining Journal 35, 37. BALL L. C. 1945. An important palaeontological discovery. Queensland Government Mining Journal 46, 23. BENSON W. N. 1913. The geology and petrology of the Great Serpentine Belt of New South Wales. Part I. Proceedings of the Linnean Society of New South Wales 38, 4 9 0 - 5 1 7 . BLAKE M . C. JR. & MURCHEY B. L. 1988. A California model for the New England fold belt. In: Kleeman J. D. ed. New

23

England Orogen — Tectonics and Metallogenesis, pp. 2031. Department of Geology and Geophysics, University of New England, Armidate. BROWN I. A. 1942. The Tamworth Series (Lower and Middle Devonian) near Attunga, N.S.W. Journal and Proceedings of the Royal Society of New South Wales 76, 165—176. BROWNE W. R. 1929. An outline of the history of igneous action in New South Wales till the close of the Palaeozoic Era. Proceedings of the Linnean Society of New South Wales 54, ix-xxxix. BROWNE W. R. 1949. Some thoughts on the division of the geological record in the Commonwealth of Australia. In: Carey S. W. ed. Report of the Twenty-Seventh Meeting of the Australian and New Zealand Association for the Advancement of Science, pp. 35-^46. Government Printer, Hobart. BRYAN W . H. 1925. Earth movements in Queensland. Proceedings of the Royal Society of Queensland 37, 3-82. BRYAN W. H. 1926. Earlier palaeogeography of Queensland. Proceedings of the Royal Society of Queensland 38,

79-102. BRYAN W . H. 1928. A glossary of Queensland stratigraphy.

Department of Geology, University of Queensland Special Publication.

BRYAN W. H. 1944. The relationship of the Australian

continent to the Pacific Ocean — now and in the past. Journal and Proceedings of the Royal Society of New South Wales 78, 4 2 - 6 2 . BRYAN W. H. 1948. Palaeontological and lithological notes. Proceedings of the Royal Society of Queensland 59, 203-204. BRYAN W. H. & JONES O. A. 1944. A revised glossary of Queensland stratigraphy. Papers of the Department of Geology, University of Queensland 2(11). BRYAN W . H. & JONES O. A. 1950. Contributions to the geology of Brisbane. No 1. Local applications of the standard stratigraphic nomenclature. Proceedings of the Royal Society of Queensland 61, 13—18. BRYAN W. H. & JONES O. A. 1954. Contributions to the geology of Brisbane. No 2. The structural history of the Brisbane Metamorphics. Proceedings of the Royal Society of Queensland 65, 25—50. BURK C. A. 1965. Geology of the Alaska Peninsula — island arc and continental margin. Geological Society of America Memoir 99. BURK C. A. 1972. Uplifted eugeosynclines and continental margins. Geological Society of America Memoir 132, 75-85. BURK C. A. & DRAKE C. A. (Editors) 1974. The Geology of Continental Margins. Springer-Verlag, New York. CAREY S. W. 1955. The orocline concept in geotectonics. Part 1. Papers and Proceedings of the Royal Society of Tasmania 89, 255-288. CAREY S. W. 1958. A tectonic approach to continental drift. In: Continental Drift — A Symposium, pp. 177—355. Department of Geology, University of Tasmania, Hobart. C A R E Y S. W. 1976. The Expanding Earth. Elsevier, Amsterdam. C A R E Y S. W. & B R O W N E W. R. 1938. Review of the Carboniferous stratigraphy, tectonics and palaeogeography of New South Wales and Queensland. Journal and Proceedings of the Royal Society of New South Wales 71, 591-614. CAWOOD P. A. 1976. Cambro-Ordovician strata northern New South Wales. Search 7, 317-318.

24

C. G. MURRAY

CAWOOD P. A. 1982a. Correlation of stratigraphic units across the Peel Fault System. In: Flood P. G. & Runnegar B. eds. New England Geology, pp. 53-61. Department of Geology, University of New England and the AHV Club, Armidale. CAWOOD P. A. 1982b. Structural relations in the subduction complex of the Paleozoic New England Fold Belt, eastern Australia. Journal of Geology 90, 381-392. CAWOOD P. A. 1982c. Tectonic reconstruction of the New England Fold Belt in the Early Permian: an example of development at an oblique-slip margin. In: Flood P. G. & Runnegar B. eds. New England Geology, pp. 25-34. Department of Geology, University of New England and the AHV Club, Armidale. CAWOOD P. A. 1983. Modal composition and detrital clinopyroxene geochemistry of lithic sandstones from the New England Fold Belt (east Australia): a Paleozoic forearc terrane. Geological Society of America Bulletin 94, 1199-1214.

CAWOOD P. A. & LEITCH E. C. 1985. Accretion and dispersal

tectonics of the southern New England Fold Belt, eastern Australia. In: Howell D. G. ed. Tectonic Terranes of the Circum-Pacific Region, pp. 481-492. Circum-Pacific Council for Energy and Mineral Resources, Earth Science Series 1.

Australasian Medical Publishing Company Limited, Sydney. DAVID T . W . E . & BROWNE W . R . 1 9 5 0 . The Geology

Commonwealth London.

of Australia.

of

the

Edward Arnold & Co.,

DAY R . W . , MURRAY C . G . & WHITAKER W . G . 1 9 7 8 . T h e

eastern part of the Tasman Orogenic Zone. Tectonophysics 48, 327-364. DEAR J. F., MCKELLAR R . G . & TUCKER R . M . 1 9 7 1 . G e o l o g y

of the Monto 1:250 000 Sheet area. Geological Survey of Queensland Report 46. DENMEAD A. K. 1928. A survey of the Brisbane Schists. Proceedings of the Royal Society of Queensland 39, 71-106. DENMEAD A. K. 1960a. Permian. Gympie. In: Hill D. & Denmead A. K. eds. The Geology of Queensland, pp. 2 2 5 228. Journal of the Geological Society of Australia 7. DENMEAD A. K. 1960b. Silurian. South-eastern Queensland — The Neranleigh-Fernvale Group. The Moreton district. In: Hill D. & Denmead A. K. eds. The Geology of Queensland, pp. 136-137. Journal of the Geological Society of Australia 7. DUNSTAN B. 1904a. Records No. 1. III. Phosphate-bearing rocks in the Rockhampton district. Geological Survey of Queensland Publication 190, 10-11.

CHAPPELL B. W. 1961. The stratigraphy and structural geology of the Manilla-Moore Creek district, N.S.W. Journal and Proceedings of the Royal Society of New South Wales 95, 63-75.

DUNSTAN B. 1904b. Records No. 1. VI. Notes on the geological history of Keppel Bay. Geological Survey of Queensland Publication 190, 27-28.

CHERRY D. 1989. Geological note. The palaeogeographic significance of a Late Devonian Lambie Group clast in the Late Carboniferous Currabubula Formation, New South Wales. Australian Journal of Earth Sciences 36, 139—140.

DUNSTAN B. 1916. Appendix B. Queensland geological formations. In: Harrap G. ed. A School Geography of Queensland, pp. 164-176. Department of Public Instruction, Brisbane.

COLLINS W. J. 1990. A reinterpretation of Permian tectonism in the Sydney Basin and southern New England Fold Belt. In: Twenty-fourth Newcastle Symposium on Advances in the Study of the Sydney Basin, pp. 8-16. Department of Geology, University of Newcastle, Newcastle. COLLINS W. J. 1991. A reassessment of the 'Hunter-Bowen Orogeny': tectonic implications for the southern New England Fold Belt. Australian Journal of Earth Sciences 38, 409-423.

DUNSTAN B. 1920. Industrial minerals. Article 8. Manganese. Geological Survey of Queensland Publication 267, 2 4 7 -

COLLINS W . J., OFFLER R . , FARRELL T . R . & LANDENBERGER

B. 1993. A revised late Palaeozoic-early Mesozoic tectonic history for the southern New England Fold Belt. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 69-84. Department of Geology and Geophysics, University of New England, Armidale. CROOK K. A. W. 1958. Note on the occurrence of a Carboniferous plant near Armidale, N.S.W. Australian Journal of Science 20, 216-218. CROOK K. A. W. 1961. Stratigraphy of the Tamworth Group (Lower and Middle Devonian), Tamworth-Nundle district, N.S.W. Journal and Proceedings of the Royal Society of New South Wales 94, 173-188. DAVID T. W. E. 1894. A sketch of our present knowledge of the geological history of Australia, Tasmania and New Zealand from Archaean time down to the commencement of the Permo-Carboniferous Period. Proceedings of the Linnean Society of New South Wales 8, 547-607. DAVID T. W. E. 1911. Notes on some of the chief tectonic lines of Australia. Journal and Proceedings of the Royal Society of New South Wales 45, 4-60. DAVID T. W. E. 1932. Explanatory Notes to accompany a new Geological Map of the Commonwealth of Australia.

288.

ELLIOTT L. G. 1993. Post-Carboniferous tectonic evolution of eastern Australia. The APEA Journal 33, 215-236. EVANS P. R & ROBERTS J. 1980. Evolution of central eastern Australia during the late Palaeozoic and early Mesozoic. Journal of the Geological Society of Australia 26, 3 2 5 340.

FERGUSSON C. L. 1982a. Structure of the late Palaeozoic Coffs Harbour Beds, northeastern New South Wales. Journal of the Geological Society of Australia 29, 25-40. FERGUSSON C. L. 1982b. Structure and tectono-stratigraphy of the central Coffs Harbour Block, eastern New England Fold Belt. PhD thesis, University of New England, Armidale (unpubl.). FERGUSSON C. L. 1982C. An ancient accretionary terrain in eastern New England — evidence from the Coffs Harbour Block. In: Flood P. G. & Runnegar B. eds. New England Geology, pp. 63-70. Department of Geology, University of New England and the AHV Club, Armidale. FERGUSSON C. L. 1984. Tectono-stratigraphy of a Palaeozoic subduction complex in the central Coffs Harbour Block of north-eastern New South Wales. Australian Journal of Earth Sciences 31, 217-236. FERGUSSON C. L. 1991. Thin-skinned thrusting in the northern New England Orogen, central Queensland, Australia. Tectonics 10, 797-806. FERGUSSON C. L. & FLOOD P. G. 1984. A late Palaeozoic

subduction complex in the Border Rivers area of southeast Queensland. Proceedings of the Royal Society of Queensland 95, 47-55.

TECTONIC EVOLUTION, NEW ENGLAND OROGEN FISHER N. H. 1974. Progress in studies of the Tasman Geosyncline. In: Denmead A. K., Tweedale G. W. & Wilson A. F. eds. The Tasman Geosyncline — A Symposium, pp. 3-17. Geological Society of Queensland, Queensland Division, Brisbane. FITCH T. J. 1972. Plate convergence, transcurrent faults, and internal deformation adjacent to Southeast Asia and the Western Pacific. Journal of Geophysical Research 77, 4432-4460.

25

Royal Society of New South Wales 108, 75-93. HALL R. L. 1978. A Silurian (Upper Llandovery) coral fauna from the Woolomin Beds near Attunga, New South Wales. Proceedings of the Linnean Society of New South Wales 102,85-108. HARRINGTON H. J. & KORSCH R. J. 1985a. Tectonic model for the Devonian to middle Permian of the New England Orogen. Australian Journal of Earth Sciences 32, 163— 179.

FITZPATRICK K. R. 1974. New England Fold Belt. WoolominTexas Block. Woolomin Beds and associated sediments. In: Markham N. L. & Basden H. eds. The Mineral Deposits of New South Wales, pp. 339-349. Geological Survey of New South Wales, Sydney.

HARRINGTON H. J. & KORSCH R. J. 1985b. D e f o r m a t i o n

FLEMING P. J. G . , DAY R. W . , MURRAY C . G . & WHITAKER

HARRINGTON H. J. & KORSCH R. J. 1987. Oroclinal bending in

W. G. 1974. Late Palaeozoic invertebrate macrofossils in the Neranleigh-Fernvale Beds. Queensland Government Mining Journal IS, 104-107.

the evolution of the New England-Yarrol Orogen and the Moreton Basin. In: Proceedings Pacific Rim Congress 87, pp. 797-800. Australasian Institute of Mining and Metallurgy, Melbourne. HENDERSON R. A. 1980. Structural outline and summary geological history for northeastern Australia. In: Henderson R. A. & Stephenson P. J. eds. The Geology and Geophysics of Northeastern Australia, pp. 1-26. Geological Society of Australia, Queensland Division, Brisbane. HESS H. H. 1962. History of ocean basins. In: Engel A. E. J., James H. L. & Leonard B. F. eds. Petrologic Studies: A Volume in Honor of A. F. Buddington, pp. 599-620. Geological Society of America. HILL D. 1960. Geological structure. In: Hill D. & Denmead A. K. eds. The Geology of Queensland, pp. 1-19. Journal of the Geological Society of Australia 7. HILL D. 1961. Geology of south-eastern Queensland. In: Introducing Queensland — A Handbook prepared for the Thirty-fifth Congress of the Australian and New Zealand Association for the Advancement of Science, pp. 125—135. Government Printer, Brisbane.

FLEMING P. J. G . , MURRAY C . G . & WHITAKER W . G . 1 9 7 5 .

Late Palaeozoic invertebrate fossils in the Wandilla Formation, and the deposition of the Curtis Island Group. Queensland Government Mining Journal 76, 416-422. FLOOD P. G. 1988. New England Orogen: Geosyncline, mobile belt and terranes. In: Kleeman J. D. ed. New England Orogen — Tectonics and Metallogenesis, pp. 1 6. Department of Geology and Geophysics, University of New England, Armidale. FLOOD P. G. & AITCHISON J. C. 1992. Late Devonian

accretion of the Gamilaroi terrane to Gondwana: provenance linkage provided by quartzite clasts in the overlap sequence. Australian Journal of Earth Sciences 39, 5 3 9 - 5 4 4 . FLOOD P. G . & FERGUSSON C . L. 1 9 8 2 . T e c t o n o - s t r a t i g r a p h i c

units and structure of the Texas-Coffs Harbour region. In: Flood P. G. & Runnegar B. eds. New England Geology, pp. 71-78. Department of Geology, University of New England and the AHV Club, Armidale. FLOOD P. G . & FERGUSSON C . L .

1984. T h e

associated with the accretion of the Gympie terrane in eastern Australia. In Leitch E. C. ed. Third Circum-Pacific Terrane Conference Extended Abstracts, pp. 104-108. Geological Society of Australia Abstracts 14.

geological

HILL D. & DENMEAD A. K. (Editors) 1960. The Geology of

development of the northern New England Province of the New England Fold Belt. In: Herbert H. K. & Rynn J. M. W. eds. 1984 Field Conference. Volcanics, Granites and Mineralisation of the Stanthorpe-Emmaville-Drake region, pp. 1-19. Geological Society of Australia, Queensland Division, Brisbane.

Queensland. Journal of the Geological Society of Australia 7. ISHIGA H. 1990. Radiolarians from the Gympie Province, eastern Australia. In: Proceedings Pacific Rim Congress 90, Volume III, pp. 187-189. Australasian Institute of Mining and Metallurgy, Melbourne.

FUKUI S., WATANABE T . , ITAYA T . & LEITCH E. C . 1 9 9 5 .

Middle Ordovician high PT metamorphic rocks in eastern Australia: evidence from K-Ar ages. Tectonics 14, 10141020. GEOLOGICAL SOCIETY O F AUSTRALIA 1 9 7 1 . Tectonic

Australia and New Guinea Society of Australia, Sydney.

map

of

1:5 000 000. Geological

GLEN R. A. & SCHEIBNER E. 1993. Lachlan rocks in the N e w

England Orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 123-126. Department of Geology and Geophysics, University of New England, Armidale. GRAHAM I. J. & KORSCH R. J. 1985. R b - S r geochronology of

coarse-grained greywackes and argillites from the Coffs Harbour Block, eastern Australia. Isotope Geoscience 4, 45-54.

GREEN D. C. 1973. Radiometric evidence of the Kanimblan Orogeny in south-eastern Queensland and the age of the Neranleigh-Fernvale Group. Journal of the Geological Society of Australia 20, 153-160. HALL R. L. 1975. Late Ordovician coral faunas from northeastern New South Wales. Journal and Proceedings of the

ISHIGA H . , LEITCH E. C . , WATANABE T . , NAKA T . & IWASAKI

M. 1988a. Radiolarian and conodont biostratigraphy of siliceous rocks from the New England Fold Belt. Australian Journal of Earth Sciences 35, 73-80. ISHIGA H., WATANABE T. & LEITCH E. C. 1988b. Microfossil

biostratigraphy and lithologic association of bedded chert in the Macdonald Block of the New England Fold Belt, eastern Australia. In Preliminary Report on the Geology of the New England Fold Belt, Australia (No. 1'), pp. 47-59. Co-operative Research Group of Japan and Australia. JACK R . L . & E T H E R I D G E R . J r .

1892. T h e g e o l o g y

and

palaeontology of Queensland and New Guinea. Geological Survey of Queensland Publication 92. JEFFREYS H. 1929. The Earth. Its Origin, History and Physical Constitution. Second Edition. Cambridge University Press, London. JENSEN H. I. 1912. The building of eastern Australia. Proceedings of the Royal Society of Queensland 23, 149198.

KAY M. 1951. North American geosynclines. Geological Society of America Memoir 48.

26

C. G. MURRAY

KENNY E. J. 1937. Geological reconnaissance of the North Coast region. Progress Report. Annual Report of the Department of Mines New South Wales for 1936, 92. KIRKEGAARDA. G . , SHAW R. D . & MURRAY C. G . 1970.

Geology of the Rockhampton and Port Clinton 1:250 000 Sheet areas. Geological Survey of Queensland Report 38. KORSCH R. J. 1973. Structural analysis of the Palaeozoic sediments in the Woolgoolga district, North Coast, New South Wales. Journal and Proceedings of the Royal Society of New South Wales 106, 98-103. KORSCH R. J. 1975. Structural analysis and geological evolution of the Rockvale-Coffs Harbour region, northern New South Wales. PhD thesis, University of New England, Armidale (unpubl.). KORSCH R. J. 1977. A framework for the Palaeozoic geology of the southern part of the New England Geosyncline. Journal of the Geological Society of Australia 25, 339-355. KORSCH R. J. 1981. Deformational history of the Coffs Harbour Block. Journal and Proceedings of the Royal Society of New South Wales 114, 17-22. KORSCH R. J. 1984. Sandstone compositions from the New England Orogen, eastern Australia: implications for tectonic setting. Journal of Sedimentary Petrology 54, 192-211. KORSCH R. J. & HARRINGTON H. J. 1981. S t r a t i g r a p h i c a n d

structural synthesis of the New England Orogen. Journal of the Geological Society of Australia 28, 205—226. KORSCH R. J. & HARRINGTON H. J. 1985. F r a g m e n t a t i o n a n d

deformation of terranes in the New England Orogen, eastern Australia. In: Leitch E. C. ed. Third CircumPacific Terrane Conference Extended Abstracts, pp. 131— 135. Geological Society of Australia Abstracts 14. KORSCH R. J. & HARRINGTON H. J. 1987. Oroclinal bending,

fragmentation and deformation of terranes in the New England Orogen, eastern Australia. In: Leitch E. C. & Scheibner E. eds. Terrane Accretion and Orogenic Belts, pp. 129-139. American Geophysical Union Geodynamics Series 19. KORSCH R. J., HARRINGTON H. J., MURRAY C. G . , FERGUSSON

C. L. & FLOOD P. G. 1990a. Tectonics of the New England Orogen. In: Finlayson D. M. ed. The EromangaBrisbane Geoscience Transect: a guide to basin development across Phanerozoic Australia in southern Queensland, pp. 35—52. Bureau of Mineral Resources Bulletin 232. KORSCH R . J., WAKE-DYSTER K. D . & JOHNSTONE D . W .

1990b. Deep seismic profiling across the Bowen Basin. In: Beeston J. W. ed. Bowen Basin Symposium 1990 Proceedings, pp. 10-14. Geological Society of Australia, Queensland Division, Brisbane. KORSCH R . J., WAKE-DYSTER K. D. & JOHNSTONE D. W .

1993. The Gunnedah Basin-New England Orogen deep seismic reflection profile: implications for New England tectonics. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 85—100. Department of Geology and Geophysics, University of New England, Armidale. LEITCH E. C. 1969. Igneous activity and diastrophism in the Permian of New South Wales. Geological Society of Australia Special Publication 2, 21—37. LEITCH E. C. 1974. The geological development of the southern part of the New England Fold Belt. Journal of the Geological Society of Australia 21, 133-156. LEITCH E. C. 1975. Plate tectonic interpretation of the

Paleozoic history of the New England Fold Belt. Geological Society of America Bulletin 86, 141-144. LEITCH E. C. & CAWOOD P. A. 1980. Olistoliths and debris

flow deposits at ancient consuming plate margins: an eastern Australian example. Sedimentary Geology 25, 5-22. LEITCH E. C. & CAWOOD P. A. 1987. Provenance determina-

tion of volcaniclastic rocks: the nature and tectonic significance of a Cambrian conglomerate from the New England Fold Belt, eastern Australia. Journal of Sedimentary Petrology 57, 630-638. LEITCH E. C., NEILSONM. J. & HOBSON E. ( C o m p i l e r s ) 1971.

Dorrigo-Coffs Harbour 1:250 000 Geological Series Sheet SH 56-10&11. Geological Survey of New South Wales, Sydney. LITTLE T . A . , HOLCOMBE R . J. & MCWILLIAMS M . 1 9 9 3 . 40

Ar/ 3 9 Ar thermochronology of epidote-blueschists and granitoids from the North D'Aguilar block, Queensland: implications for the timing of metamorphism and unroofing in the New England Orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 549-556. Department of Geology and Geophysics, University of New England, Armidale. LUCAS K. G. 1960a. Silurian. Border rivers. In: Hill D. & Denmead A. K. eds. The Geology of Queensland, pp. 139140. Journal of the Geological Society of Australia 7. LUCAS K. G. 1960b. Permian. The Texas area. In: Hill D. & Denmead A. K. eds. The Geology of Queensland, pp. 229235. Journal of the Geological Society of Australia 7. LUKIN G. L. 1888. Rockhampton gold fields. Annual Report of the Department of Mines, Queensland for 1887, 44-49. MARSDEN M. A. H. 1972. The Devonian history of northeastern Australia. Journal of the Geological Society of Australia 19, 125-162. MATHEWS R. T. 1960. Precambrian. The Brisbane Metamorphics. The Rocksberg Greenstones. In: Hill D. & Denmead A. K. eds. The Geology of Queensland, pp. 8 1 82. Journal of the Geological Society of Australia 7. MCCARTHY B., RUNNEGAR B. & KLEEMAN J. D . 1974. L a t e

Carboniferous-Early Permian sequence north of Drake, N.S.W. dates deposition and deformation of Beenleigh Block. Queensland Government Mining Journal 75, 324—329.

MCKELLAR R. G. 1967. The geology of the Cannindah Creek area, Monto district, Queensland. Geological Survey of Queensland Publication 331. MORAND V. J. 1993a. Structure and metamorphism of the Calliope Volcanic Assemblage: implications for Middle to Late Devonian orogeny in the northern New England Fold Belt. Australian Journal of Earth Sciences 40, 257—270. MORAND V. J. 1993b. Stratigraphy and tectonic setting of the Calliope Volcanic Assemblage, Rockhampton area, Queensland. Australian Journal of Earth Sciences 40, 15-30. MORRIS W. 1991. The Coffs Harbour Block — not an accretionary prism. In: Twenty-fifth Newcastle Symposium on Advances in the Study of the Sydney Basin, pp. 61-1 A. Department of Geology, University of Newcastle, NSW. MURPHY P. R . , SCHWARZBOCK H . , CRANFIELD WITHNALL I. W . & MURRAY C . G . 1976. Geology

L. C., of the

Gympie 1:250 000 Sheet area. Geological Survey of Queensland Report 96. MURRAY C. G. 1974. Alpine-type ultramafics in the northern part of the Tasman Geosyncline — possible remnants of Palaeozoic ocean floor. In: Denmead A. K., Tweedale

TECTONIC EVOLUTION, NEW ENGLAND OROGEN G. W. & Wilson A. F. eds. The Tasman Geosyncline — A Symposium, pp. 161-181. Geological Society of Queensland, Queensland Division, Brisbane. MURRAY C. G. 1975. A plate tectonic interpretation of mineralization in the New England Geosyncline in Queensland. Geological Survey of Queensland Record 1975/13. MURRAY C. G. 1986. Metallogeny and tectonic development of the Tasman Fold Belt System in Queensland. Ore Geology Reviews 1, 315-400. MURRAY C. G. 1987a. Review of the geology of the Gympie district, southeast Queensland, with particular reference to stratigraphic, structural and tectonic problems. In: Murray C. G. & Waterhouse J. B. eds. 1987 Field Conference, Gympie District, pp. 1-19. Geological Society of Australia, Queensland Division, Brisbane. MURRAY C. G. 1987b. Tectonic evolution and metallogenesis of the New England Fold Belt, eastern Australia. In: Proceedings Pacific Rim Congress 87, pp. 353-358. Australasian Institute of Mining and Metallurgy, Melbourne. MURRAY C. G. 1988. Tectonic evolution and metallogenesis of the New England Orogen. In: Kleeman J. D ed. New England Orogen — Tectonics and Metallogenesis, pp. 204-210. Department of Geology and Geophysics, University of New England, Armidale. MURRAY C. G. 1990. Comparison of suspect terranes of the Gympie Province with other units of the New England Fold Belt, eastern Australia. In: Proceedings Pacific Rim Congress 90, Volume II, pp. 247-255. Australasian Institute of Mining and Metallurgy, Melbourne. MURRAY C. G., FERGUSSON C. L., FLOOD P. G., WHITAKER W . G. & KORSCH R. J. 1987. Plate tectonic model for the

Carboniferous evolution of the New England Fold Belt. Australian Journal of Earth Sciences 34, 213-236.

MURRAY C. G . , MCCLUNG G. R., WHITAKER W . G . & DEGELING P. R. 1981. Geology of late Palaeozoic

sequences at Mount Barney, Queensland and Paddys Flat, New South Wales. Queensland Government Mining Journal 82, 203-213.

MURRAY C. G . & THE YARROL PROJECT TEAM 1997. T h e

Yarrol Project — increasing the prospectivity of the New England Orogen in the Rockhampton-Monto region, central coastal Queensland. In: Beeston J. W. compiler. Proceedings of the Queensland Development 1997 Conference, 13-14 November, pp. 3 9 - 5 6 . Department of Mines and Energy, Brisbane. MURRAY C. G. & WHITAKER W . G. 1982. A review of the stratigraphy, structure and regional tectonic setting of the Brisbane Metamorphics. In: Flood P. G. & Runnegar B. eds. New England Geology, pp. 79-94. Department of Geology, University of New England and the AHV Club, Armidale. OLGERS F. 1 9 7 2 . Geology of the Drummond Basin, Queensland. Bureau of Mineral Resources Bulletin 132. OLGERS F. & FLOOD P. G. 1970a. Palaeozoic geology of the Warwick and Goondiwindi 1:250 000 Sheet areas, Queensland and New South Wales. Bureau of Mineral Resources Record 1970/6. OLGERS F. & F L O O D P. G . 1970b. An angular Permian/ Carboniferous unconformity in southeastern Queensland and northeastern New South Wales. Journal of the Geological Society of Australia 17, 81—85. OSBORNE G. D. 1950. The structural evolution of the HunterManning-Myall province, New South Wales. Royal Society of New South Wales Monograph 1.

27

OVERSBY B. 1971. Palaeozoic plate tectonics in the southern

Tasman Geosyncline. Nature 234,45-48.

PACKHAM G. H. (editor) 1969a. The Geology of New South

Wales. Journal of the Geological Society of Australia 16.

PACKHAM G. H. 1969b. The New England region. Stratig-

raphy. Lower Palaeozoic Systems. Tamworth area. In: Packham G. H. ed. The Geology of New South Wales, pp. 230-231. Journal of the Geological Society of Australia 16.

PALMIERI V. 1967. Upper Silurian-Lower Devonian cono-

donts from limestones of the Wondai 'Series' (Murgon district, Queensland). Australian Journal of Science 30, 67-68. PALMIERI V. 1969. Upper Carboniferous conodonts from limestones near Murgon, south-east Queensland. Geological Survey of Queensland Publication 341. PHILIP G. M. 1966. The occurrence and palaeogeographic significance of Ordovician strata in northern New South Wales. Australian Journal of Science 29, 112-113. POGSON D. J. 1972. Geological Map of New South Wales, scale 1:1 000 000. Geological Survey of New South Wales, Sydney. POGSON D. J. & HITCHINS B. L. 1 9 7 3 . New England 1:500 000 Geological Sheet. Geological Survey of New South Wales, Sydney. P O W E L L C. MCA. 1984a. Late Devonian and Early Carboniferous: continental magmatic arc along the eastern edge of the Lachlan Fold Belt. In: Veevers J. J. ed. Phanerozoic Earth History of Australia, pp. 329-340. Clarendon Press, Oxford. POWELL C. MCA. 1984b. Silurian to mid-Devonian — a dextral transcurrent margin. In: Veevers J. J. ed. Phanerozoic Earth History of Australia, pp. 309-329. Clarendon Press, Oxford. RAGGATT H. G. 1941. Geological age of Ashford Caves limestone, N.S.W. Australian Journal of Science 3, 170— 171. REID J. H. 1 9 3 0 .

The Queensland Upper Palaeozoic succession. Geological Survey of Queensland Publication 278. RICHARDS H. C. & BRYAN W . H. 1925. Radiolarian jaspers in the Brisbane Schist Series. Proceedings of the Royal Society of Queensland 36, 131—136. RICHARDS H. C. & BRYAN W . H. 1928. Note on the Devonian rocks of central and southern Queensland. Australasian Association for the Advancement of Science Report 18, 286-290. ROBERTS J. 1985. Carboniferous faunas of the Tasman Belt,

eastern Australia. In: Leitch E. C. ed. Third CircumPacific Terrane Conference Extended Abstracts, pp. 196— 201. Geological Society of Australia Abstracts 14. ROBERTS J. 1987. Carboniferous faunas: their role in the recognition of tectonostratigraphic terranes in the Tasman Belt, eastern Australia. In: Leitch E. C. & Scheibner E. eds. Terrane Accretion and Orogenic Belts, pp. 9 3 - 1 0 2 . American Geophysical Union Geodynamics Series 19. ROBERTS J. & EN G E L B . A. 1980. Carboniferous palaeogeography of the Yarrol and New England Orogens, eastern Australia. Journal of the Geological Society of Australia 27, 167-186. ROBERTS J. & ENGEL B. A. 1987. Depositional and tectonic history of the southern New England Orogen. Australian Journal of Earth Sciences 34, 1-20. ROD E. 1966. Clues to ancient Australian geosutures. Eclogae Geologicae Helvetiae 59, 849-883.

28

C. G. MURRAY

RUNNEGAR B. 1974. The geological framework of New England. In: Murray C. G. & Runnegar B. eds. 1974 Field Conference New England area, pp. 9-19. Geological Society of Australia, Queensland Division, Brisbane. RUNNEGAR B. 1984. The Permian of Gondwanaland. Proceedings of the 27 th International Geological Congress

1, 305-339.

SCHEIBNER E. 1972. Actualistic models in tectonic mapping. International Geological Congress, 24th, Canada, Reports 3,405-422. SCHEIBNER E. 1973. A plate tectonic model of the Palaeozoic tectonic history of New South Wales. Journal of the Geological Society of Australia 20, 405-426. SCHEIBNER E. 1985. Suspect terranes in the Tasman Fold Belt System. In: Howell D. G. ed. Tectonic Terranes of the Circum-Pacific Region, pp. 493-514. Circum-Pacific Council for Energy and Mineral Resources, Earth Science Series 1. SCHEIBNER E. 1993. Structural framework of New South Wales. Geological Survey of New South Wales Quarterly Notes 93, 1-36. SCHEIBNER E. 1996. Geology of New South Wales — Synthesis. Volume 1 Structural Framework. Geological Survey of New South Wales Memoir Geology 13 (1). SCHEIBNER E. & GLEN R. A. 1972. The Peel Thrust and its

tectonic history (New South Wales). Geological Survey of New South Wales Quarterly Notes 8, 2—14. SCHUCHERT C. 1916. The problem of continental fracturing and diastrophism in Oceanica. American Journal of Science 192,91-105. STEWART I. 1995. Cambrian age for the Pipeclay Creek Formation, Tamworth Belt, northern New South Wales. Courier Forschungsinstitut Senckenberg 182, 565-566. SUSSMILCH C. A. 1920. Sequence and glaciation of the Carboniferous rocks of the Hunter River district, New South Wales. Journal and Proceedings of the Royal Society of New South Wales 53, 246-282. SUSSMILCH C. A. 1923. The history of vulcanism in New South Wales. Journal and Proceedings of the Royal Society of New South Wales 57, 12-53. VEEVERS J. J., JONES J. G . & POWELL C. MCA. 1982. T e c t o n i c

framework of Australia's sedimentary basins. The APE A Journal 22, 283-300.

VOISEY A. H. 1934. A preliminary account of the geology of the middle North Coast district of New South Wales. Proceedings of the Linnean Society of New South Wales 59, 333—347. VOISEY A. H. 1938. The geology of the Armidale district. Proceedings of the Linnean Society of New South Wales 63, 463-467. VOISEY A. H. 1942. The geology of the County of Sandon, N.S.W. Proceedings of the Linnean Society of New South Wales 67, 288-293. VOISEY A. H. 1959a. Tectonic evolution of north-eastern New

South Wales, Australia. Journal and Proceedings of the Royal Society of New South Wales 92, 191-203. VOISEY A. H. 1959b. Australian geosynclines. Australian Journal of Science 22, 188-198. VOISEY A. H. 1969. The New England region. Stratigraphy. Lower Palaeozoic Systems. In: Packham G. H. ed. The Geology of New South Wales, pp. 229-230. Journal of the Geological Society of Australia 16. WASS R. & DENNIS D. M. 1977. Early Palaeozoic faunas from the Warwick-Stanthorpe region, Queensland. Search 8, 207-208. WATANABE T., ITAYA T., TAKESHITA H. & LEITCH E. C . 1988.

K-Ar dating of metamorphic rocks from the southern Yarrowitch and the southern Macdonald blocks. In: Preliminary Report on the Geology of the New England Fold Belt, Australia (No. 1), pp. 141-143. Co-operative Research Group of Japan and Australia. WEGENER A. 1912a. Die entstehung der kontinente und ozeane. Petermanns Mitteilungen 58, 185-195, 253-256, 305-309. WEGENER A. 1912b. Die entstehung der kontinente. Geologische Rundschau 3, 276-292. WEGENER A. 1915. Die Entstehung der Kontinente und Ozeane. Vieweg & Sohn, Brunswick. WHITAKER W. G. 1972. A note on the occurrence of phenoclasts of Lower Palaeozoic oolitic arenite in the Upper Triassic Kholo Sub-group. Queensland Government Mining Journal 73, 149-152. WHITAKER W . G . , MURPHY P. R . & ROLLASON R . 1 9 7 4 .

Geology of the Mundubbera 1:250 000 Sheet area. Geological Survey of Queensland Report 84. WHITEHOUSE F. W. 1928. Central Queensland geology. Queensland Government Mining Journal 29, 441-442. WHITEHOUSE F. W. 1930. The geology of Queensland. In: Bryan W. H., Longman H. A. & Reid J. F. F. eds. Handbook for Queensland, pp. 23-39. Australasian Association for the Advancement of Science, Brisbane. WLTHNALL I. W . , BULTITUDE R. J., LANG S. C., DONCHAK P. J.

& HAMMOND R. L. 1987. Geology and tectonic history of the Palaeozoic Hodgkinson and Broken River Provinces, north Queensland. In: Pacific Rim Congress 1987 Proceedings, pp. 495-498. Australasian Institute of Mining and Metallurgy, Melbourne. WLTHNALL I. W. & LANG S. C. 1990. Tectonic history of the

Palaeozoic Broken River Province, north Queensland. In: Pacific Rim Congress 1990 Proceedings, Volume / / , pp. 315-323. Australasian Institute of Mining and Metallurgy, Melbourne. WlTHNALL I. W. & LANG S. C. 1993. Geological and tectonic

history. In: Withnall I. W. & Lang S. C. eds. Geology of the Broken River Province, north Queensland, pp. 261— 273. Queensland Geology 4. (Received 15 May 1996; accepted 1 July 1997)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 29—51.

Crustal architecture of the New England Orogen based on deep seismic reflection profiling R. J. KORSCH, D. W. JOHNSTONE AND K. D. WAKE-DYSTER Australian Geodynamics Cooperative Research Centre, Australian Geological Survey Organisation, GPO Box 378 Canberra, ACT2601, Australia. Petroleum & Marine Division, Australian Geological Survey Organisation, GPO Box 378 Canberra, ACT2601, Australia. 1

1

2

2

2

Deep seismic reflection profiling in eastern Australia has provided images of the present-day crustal architecture of the New England Orogen, its relationship with the Lachlan and Thomson Orogens and the geometry of the Bowen—Gunnedah-Sydney Basin, a successor basin system. Crustal architecture along the suture between the New England and the Lachlan-Thomson Orogens varies considerably. In the north, the Bowen Basin seismic lines are dominated by east-dipping structures, with the New England Orogen thrust over the Thomson Orogen. The lines in southernmost Queensland are dominated by gently west-dipping structures in the lower crust, but eastdipping thrusts in the upper crust, with the accretionary-wedge part of the New England Orogen thrust beneath the easternmost Lachlan Orogen, but with the forearc basin component (Tamworth Belt) of the New England Orogen being thrust over (backthrust) the Lachlan Orogen. A profile in the southern New England Orogen contains both east- and west-dipping structures and shows similar relationships between the New England and Lachlan Orogens as in the lines in southernmost Queensland. Thus, the New England Orogen, at least in its southern part, is a doubly vergent orogen that developed in the Late Palaeozoic to Triassic. Key words: crustal architecture, New England Orogen, seismic profiles, tectonics.

INTRODUCTION The Tasman Orogenic Zone in eastern Australia is subdivided into the New England Orogen east of the Bowen-Gunnedah-Sydney basin system and the Lachlan and Thomson Orogens west of the basin system. To the east of the Lachlan and Thomson Orogens, tectonic development of eastern Australia was dominated during the Devonian and Carboniferous by a convergent plate margin setting, with west-dipping subduction taking place (Murray et al. 1987; Korsch et al. 1990a; Fergusson et al. 1993). An accretionary wedge formed the main part of the New England Orogen (the Tablelands Complex) and the present-day Tamworth and Yarrol belts correspond to the forearc basin (Figure 1). To the north in the Yarrol Province, the magmatic arc is represented by the Connors and Auburn volcanic arcs. In the southern part of the orogen, the New England Province, the volcanic arc is now missing, either concealed beneath the Gunnedah Basin or Tamworth Belt, or removed by erosion or strike-slip faulting or by a combination of these. The main part of the New England Orogen is related to west-dipping subduction, and at least two units, the Gympie Province and Beenleigh Block (Figure 1), have been postulated to be exotic terranes (Korsch & Harrington 1987; Murray 1988). The relationship between the Lachlan and Thomson Orogens west of the basin system and the New England Orogen east of the basin system is enigmatic because

the contact between them is concealed and presumably somewhere beneath the basins. The Australian Geological Survey Organisation [formerly Bureau of Mineral Resources (BMR)] has conducted three deep seismic surveys with 20 s record lengths (approximately 60 km depth) in eastern Australia across parts of the basin system and underlying orogens (Figure 1). Here we provide geological interpretations of the seismic profiles, compare the crustal architectures seen on the profiles, and comment on the relationships between the New England and Lachlan and Thomson Orogens. As a note of caution, the seismic reflection method provides an image of the current structure of the crust; hence the best images are often of the latest tectonic events, which overprint and often obliterate or obscure older structures. In eastern Australia, the products of the late contractional phases tend to dominate in the recorded sections. SEISMIC DATA In 1984 and 1986 the BMR, in conjunction with the Geological Survey of Queensland, shot a 6-fold, common depth point, explosive source, 20 s record length, seismic reflection profile across southernmost Queensland (Wake-Dyster et al 1987) (lines BMR84.14, BMR84.16, BMR86.15, BMR86.17, BMR86.18, BMR86.19; Figure 1). Almost the entire traverse was recorded on sedimentary rocks of the Eromanga,

30

R. J. KORSCH ET AL.

Two further surveys were undertaken as part of the National Geoscience Mapping Accord (NGMA) project, Sedimentary Basins of Eastern Australia. In 1989, the BMR conducted a deep seismic reflection survey across the Bowen Basin and New England Orogen in the vicinity of Blackwater and Marlborough in Queensland (Korsch et al. 1990b). Reflection data (20 s TWT; 8fold CMP; explosive source) were recorded along three lines totalling 254 km in length (Figure 1). In 1991, the BMR acquired deep seismic reflection data (also 20 s TWT; 8-fold CMP; explosive source) at the latitude of Boggabri and Manilla in New South Wales in the form of a single, east-west-oriented, deep seismic reflection profile 253 km long across the Surat and Gunnedah Basins and Tamworth Belt and farther east across the Peel Fault in the New England Orogen (Korsch et al. 1993a) (Figure 2). Prior to these exclusively deep (20 s TWT) seismic reflection surveys, the BMR had been experimenting with the technique since 1957 as part of normal shallow (4-6 s TWT) seismic reflection surveys (Mathur 1983). Thus some deep seismic data, recorded to 16 s TWT, were recorded in the Bowen Basin in 1978. SOUTHERN NEW ENGLAND OROGEN: DEEP SEISMIC PROFILE BMR91.G01 Only a summary of the interpretation of the main crustal elements seen in seismic profile BMR91.G01 will be provided here. Further details, including elaboration on the geometry of the Early Permian — Middle Triassic Gunnedah Basin, can be found in Korsch et al (1993a, b), and of the Tamworth Belt in Glen et al (1993).

New England Orogen BMR deep seismic traverse (with line number) Concealed margin of PermoTriassic sediments Western boundary of Gympie Province

Figure 1 Location of the New England Orogen and BowenSydney basin system in eastern Australia and locations of the deep seismic reflection profile. Seismic lines 2, BMR89.B02; 3, BMR89.B03; 14, BMR84.14; 15, BMR86.15; 16, BMR84.16; 17, BMR86.17; 18, BMR86.18; 19, BMR86.19.

Surat and Clarence-Moreton Basins. In the area of interest for this paper, the traverse crossed, in the subsurface, the Thomson and Lachlan Orogens, and the Tamworth Belt, South D'Aguilar Block and Beenleigh Block of the New England Orogen. Only the easternmost few shotpoints were recorded on exposures of the Beenleigh Block.

Geometry of Gunnedah Basin In the vicinity of seismic profile BMR91.G01 (Figure 3), the succession in the Surat and Gunnedah Basins is generally less than 1000 m to the top of basement, which in the eastern two-thirds of the basin is normally taken as the top of a thick volcanic succession (Leitch 1993). West of the Rocky Glen Ridge, the interpretation of the seismic profile indicates the presence of only a very thin sedimentary succession above a high-velocity reflector, which we interpret to be the top of a granite or intensely deformed metasedimentary rocks. This highvelocity reflector occurs at a depth of just over 400 m at the western end of the profile, and shallows to the east towards the Rocky Glen Ridge. Below this reflector is a zone of non-reflective rocks extending to over 3 s TWT (about 10 km) which we interpret to be rocks of the Lachlan Orogen, based on correlation with the nearest Lachlan Orogen outcrops and borehole information (see also Leitch 1993; Tadros 1993). This zone can be traced eastwards until it meets a major west-dipping structure beneath the eastern part of the Tamworth Belt. Figure 2 (opp) Location of the deep seismic reflection profile BMR91.G01 across the Gunnedah Basin and western New England Orogen. The enlarged portion of the map (B) shows the simplified geology of the Tamworth Belt in the vicinity of the seismic profile (modified from Glen & Brown 1993).

DEEP SEISMIC PROFILING, NEW ENGLAND OROGEN

31

32

R.J. K O R S C H ET

AL.

w

BMR91.G01

1

Boggabri Ridge

2

Maules Creek Sub-Basin

3

Early Permian volcanics

30 km

Figure 3 Unmigrated deep seismic profile BMR91.G01 across the Gunnedah Basin and also the Tamworth Belt and Tablelands an average crustal velocity of 6000 ins-'.

East of the Rocky Glen Ridge is the West Gunnedah Sub-basin. This sub-basin was renamed the Mullaley Sub-basin by Tadros (1993) but as no reason was given for the change we use the term West Gunnedah Subbasin because it has priority. On the western side of this sub-basin, the sedimentary succession thins west onto the Rocky Glen Ridge, and there is an angular relationship with the volcanics below. Beneath the subbasin, is a thick, well-layered succession, which we interpret to be interlayered Early Permian volcanics and sedimentary rocks, the upper parts of which are intersected in many petroleum and coal exploration drillholes (Leitch 1993). This volcanic succession is relatively thin on the eastern side of the Rocky Glen Ridge, but appears to thicken to at least 3 km thick towards the east beneath the West Gunnedah Sub-basin. The Meandarra Gravity Ridge (see Murray et al. 1989) forms a gravity high across the Gunnedah Basin. Modelling by Krassay et al. (in press) of gravity data collected along the deep seismic profile indicates that the high can be explained by a pile of mafic volcanic rocks, up to 8 km thick, with a half-graben geometry centred on the West Gunnedah Sub-basin (Figure 3). To the east beneath the Maules Creek Sub-basin, the gravity modelling predicts a volcanic pile less than 3 km thick. The gravity model is supported by the rocks intersected in the Kelvin 1 well (about 18 km south of

the deep seismic profile) where, although the entire volcanic succession was not penetrated, over 1480 m of volcanics were intersected. Interpretation of the deep seismic section by Korsch et al. (1993a,b) suggested there is no evidence to indicate that the later depocentres of the Gunnedah Basin (West Gunnedah and Maules Creek Sub-basins) were controlled by extensional fault systems. If extensional faults operated, they must have been active at the time of the volcanism prior to commencement of sedimentation in the Early Permian.

Mooki Fault

The succession in the Maules Creek Sub-basin thickens from the eastern margin of the Boggabri Ridge towards the east, and at the surface the eastern limit of the Gunnedah Basin is defined by the Mooki Fault. In the vicinity of the seismic line, Ramsay and Stanley (1976) calculated a dip of 25° to the east for the fault, using the results of a ground-based magnetometer survey. The seismic data (Figure 4) confirm this geometry (Korsch et al. 1993a; Glen et al. 1993; Beckett et al 1995), with the Tamworth Belt being thrust over the eastern margin of the Gunnedah Basin for up to 10 km (see also Glen & Brown 1993; Woodward 1995).

DEEP SEISMIC PROFILING, N E W E N G L A N D O R O G E N

33

Complex of the New England Orogen (see Figure 2 for location). Vertical scale approximately equals horizontal scale assuming

Figure 4 Enlargement of part of unmigrated seismic line BMR91.G01 across the Mooki Thrust, Kelvin Thrust and Rocky Creek Syncline of the Tamworth Belt (see Figure 2 for location). Vertical scale approximately equals horizontal scale assuming an average upper crustal velocity of 4000 ms-'.

34 R.J. K O R S C H ET AL. Harrington and Korsch (1985) postulated that, in the Late Carboniferous, a precursor to the Mooki Fault existed at about its present site. This precursor was inferred to be part of a major strike-slip fault system extending nearly the entire length of the New England Orogen. Evidence for this fault cannot be seen in the seismic profile (Figure 4), but vertical faults are very difficult to image using the seismic reflection technique. If the fault exists, it must be below the pile of Early Permian Boggabri Volcanics. Tamworth Belt The quality of seismic data from across the Tamworth Belt is quite variable. Extremely high-quality data were obtained from the western part where the surface dips of the beds are relatively shallow, but the resolution of the data decreases towards the east where the surface dips are much steeper. An interpretation of the seismic data across the Tamworth Belt is presented in Glen et al. (1993). The western part of the Tamworth Belt displays a relatively simple structural geometry dominated at the surface by the Rocky Creek Syncline (Figure 4) and the Klori Anticline. At the Kelvin Fault Devonian rocks are thrust on top of Carboniferous ones and there appears to be a greater displacement than the Mooki Fault, which thrusts Late Carboniferous rocks on top of Early Permian rocks (Glen et al. 1993; Woodward 1995). About 40 km south of the seismic line the two thrusts converge, suggesting that the Mooki Fault is the frontal splay fault in front of the more fundamental Kelvin Fault. Based on a series of balanced cross-sections, Woodward (1995) suggested that the Tamworth Belt had been displaced westwards by 75-85 km. Woodward's model assumes that the Tamworth Belt continues at depth to the west of the Mooki Fault beneath the Gunnedah Basin and Early Permian volcanic pile. This is not supported by the deep seismic data where there is a nonreflective zone which we interpret as subsurface Lachlan Orogen. Instead, it appears that the entire Tamworth Belt has been obducted westwards onto the continental crust of the Lachlan Orogen. Peel Fault The Peel Fault is the present-day boundary between the Tamworth Belt (forearc basin) and the Tablelands Complex (accretionary wedge). Ten kilometres south of the seismic line White (1964) recorded a dip on the fault of 65° to the east. This was supported by modelling the results of a magnetometer survey by Ramsay and Stanley (1976). Because of the steep dip, it is unlikely that the Peel Fault would be imaged seismically. On the seismic profile (Figure 5), directly beneath the surface position of the Peel Fault, at a depth of about 1 km (0.4 s TWT), reflections dip moderately to the west, placing significant constraints on the geometry of the Peel Fault at depth. The west-dipping structure would truncate an east-dipping Peel Fault, or an east-dipping

Peel Fault would sole onto it. We thus suggest that the Peel Fault represents a very-high-level splay fault off a major west-dipping fault (Figure 3). We also interpret the west-dipping reflections as part of a system which extends to the Moho below the Tamworth Belt (Figure 3). The west-dipping structure projects to the surface east of the Peel Fault and corresponds to a melange zone of serpentinite and related rocks (Korsch et al. 1993a). These rocks have been brought from the mantle and lower crust to the surface and now occur as a 2 km-wide belt located about 6 km to the northeast of the Peel Fault along the line of the seismic profile. Glen and Brown (1995) referred to structures in the schistose serpentinite that are consistent with such an interpretation. The serpentinites emplaced along the Peel Fault were possibly derived at depth from the west-dipping melange zone. Tablelands Complex To the east of the Peel Fault, the Tablelands Complex (Devonian—Carboniferous accretionary wedge) is intruded by younger granitic plutons, including the early Permian Bundarra Plutonic Suite. In that part of the Tablelands Complex between the Peel Fault and the Bundarra Plutonic Suite, the seismic profile has a series of reflections dipping moderately to the west (Figure 5). One of these bands of reflections represents the melange zone of serpentinite and associated rocks mentioned above. The west-dipping reflections could represent thrust faults that developed during the Late Devonian to Carboniferous subduction-related accretion. Thrust faults within the accretionary wedge have been mapped in surface exposures (Cawood 1982; Blake & Murchey 1988). On the surface, however, the rocks dip very steeply and the axial surfaces of folds tend to be sub vertical. Hence the shallow reflections are more likely to represent structures within the accretionary wedge that developed later than the accretion phase. Alternatively, these reflections could represent latest Carboniferous — Early Permian extensional faults, such as the low-angle detachment faulting proposed by Blake and Murchey (1988) to the north in the Woodsreef area. A third alternative, favoured here, is that the reflections could represent east-directed thrusts that post-date the emplacement of the Bundarra Plutonic Suite. Data for the eastern end of seismic profile BMR91.G01 were acquired on outcrops of the Bundarra Plutonic Suite, which is a high-level pluton intruded into low-grade metasedimentary rocks of the accretionary wedge. From the surface to depths of 6-9 km (2-3 s TWT) the seismic section is predominantly non-reflective (Figure 6). We interpret this to be due to the homogeneous lithology of the body of the pluton. Strong reflections at 6-9 km indicate a major lithological change to the top of what we interpret to be the floor of the pluton. The pluton also extends for about 12 km to the west of its outcrop position beneath a shallow roof of Tablelands Complex (see Pogson & Hitchins 1973). This allows the shape of this western part of the pluton to be defined as pancake-like (Figure 6). Although the

DEEP SEISMIC PROFILING, NEW E N G L A N D OROGEN Tamworth Belt

> <

Tablelands Complex

35 E

Figure 5 Enlargement of part of unmigrated seismic line BMR91.G01 across the Peel Fault separating the Tamworth Belt from the Tablelands Complex (see Figure 2 for location). Vertical scale approximately equals horizontal scale assuming an average upper crustal velocity of 4000 ms- 1 .

pluton is predominantly non-reflective, there are a few internal reflections which dip gently to the west, indicating the possibility of faulting or internal deformation within the pluton since its emplacement. These reflections may be related to similar west-dipping structures within the Tablelands Complex immediately to the west of the pluton. One of these dipping reflections appears to extend from the Tablelands Complex into the Bundarra Plutonic Suite (Figure 6), suggesting that all

the west-dipping reflectors could be younger than the Early Permian emplacement age of the pluton.

Crustal architecture in southern New England Orogen At the western end of profile BMR91.G01, rocks of the Lachlan Orogen occupy a zone of low seismic

BMR91.G01 Tablelands Complex

Bundarra Plutonic Suite Position of exposed contact

5200 5217

0

26/Q/442

6

10 km

Figure 6 Enlargement of part of seismic line BMR91.G01 across the Bundarra Plutonic Suite within the Tablelands Complex (see Figure 2 for location). Vertical scale approximately equals horizontal scale assuming an average upper crustal velocity of 4000 ms-'.

36

R.J. KORSCH ET AL. Bowen Basin

Moho Lachlan Orogen

BMR84.14

Surat Basin

? missing

D - C volcanic arc

20 50 km

Figure 7 Line diagram of deep seismic reflection profile BMR84.14 across the Bowen Basin and New England Orogen

reflectivity to a depth of about 10 km and occur below a very thin sedimentary cover of the Gilgandra Trough (Figure 3). Below this there is a more highly reflective middle to lower crust. The non-reflective zone thickens eastwards, extending to a depth of about 24 km (~8 s TWT) beneath the Maules Creek Sub-basin and western part of the Tamworth Belt. This zone, and the reflective lower crust below it, appear to be truncated by a major west-dipping structure, which we interpret as a fault that extends through the entire crust. At upper crustal levels, the fault is represented by the west-dipping reflections that occur immediately below the Peel Fault (Figure 5). Within the seismic section, beneath the Tablelands Complex and Bundarra Plutonic Suite, we note other strong reflections that also dip to the west (Figure 3). Because the reflections cannot be traced to the surface, there are no geological constraints on their character, but we speculate above that they possibly represent west-dipping faults or shear zones. The reflective lower crust beneath the Gunnedah Basin could represent the lower part of the Lachlan Orogen or Proterozoic crust that was extensively thinned during Late Neoproterozoic or Early Palaeozoic crustal extension. Below the Gunnedah Basin, the Early Permian volcanics are interpreted to thicken from the Rocky Glen Ridge towards the east, and they extend in the subsurface at least as far as beneath the Mooki Fault. Carboniferous felsic volcanic centres in the western part of the Tamworth Belt, well to the south in the Hunter Valley, are the only surface remnants of the Late Devonian - Carboniferous magmatic arc. When active, the arc would have been located to the west of the forearc basin (Tamworth Belt), but is effectively missing. This arc could be located at depth on the seismic section, beneath the Early Permian volcanics, and its eastern margin could be truncated by the westdipping fault that brought the serpentinites and related rocks to the surface. Thus, the thickened non-reflective zone beneath the Maules Creek Sub-basin (eastern part of Lachlan Orogen on Figure 3) could in part be rocks of the Devonian-Carboniferous magmatic arc. On the

other hand, Murray et al. (1989) and Scheibner (1996) considered that the arc remains concealed beneath the overthrust Tamworth Belt, although our interpretation of line BMR91.G01 (Figure 3) provides less support for this explanation. Alternatively, the older arc could have been transported away from this area by strike-slip movements, as suggested by Harrington and Korsch (1985). The Moho has a reasonably constant depth of 30-35 km (-10-12 s TWT) across the entire profile, consistent with other deep seismic reflection profiles and refraction surveys recorded in eastern Australia. For example, in a seismic refraction study in the southern New England Orogen recorded to the east of reflection profile BMR91.G01, Finlayson and Collins (1993) found a distinct Moho at 34-35 km depth. There is no noticeable offset or jump in the position of the Moho across the west-dipping structure, probably because of re-equilibration of the Moho due to underplating associated with Tertiary basaltic volcanism. Our interpretation of the above geometry (Figure 3) suggests that the Mooki Fault dips gently to the east and that the Tamworth Belt has been thrust to the west over the Gunnedah Basin, which continues some distance to the east beneath the Tamworth Belt. Farther east, the deep seismic profile is dominated by a west-dipping crustal-scale structure which defines the eastern limit of the Lachlan Orogen in the subsurface and also the western limit of the New England accretionary wedge. We consider this feature, described above in the Peel Fault section, to be the suture between the two orogens. During a major contractional event in the Late Permian to Triassic, the New England Orogen was further thrust beneath the Lachlan Orogen. The Tamworth Belt became detached and thin-skinned and was overthrust (obducted) to the west on a system of major backthrusts. Hence, seismic line BMR91.G01 (across the Gunnedah Basin and into the southern New England Orogen) shows a doubly vergent orogen dominated by a major west-dipping structure which, although since reactivated, we interpret as the original suture between the Lachlan and New England Orogens.

DEEP SEISMIC PROFILING, NEW ENGLAND OROGEN Clarence-Moreton Basin

37

Beenleigh Block

Tamworth Belt

BMR86.17

Accretionary wedge

BMR84.16 26/Q/443B

(see Figure l for location). Vertical scale approximately equals horizontal scale assuming an average crustal velocity of 6000 ms* . 1

In southernmost Queensland, several seismic lines were acquired by the BMR in 1984 and 1986 as part of the Eromanga-Brisbane Geoscience Transect, and have been described previously (Korsch et al 1986, 1989, 1992; Wake-Dyster et al 1987; Finlayson et al 1990; papers in Finlayson 1990).These interpretations are updated here in the light of more recent work. Of particular interest are lines BMR84.14, BMR84.16 and BMR86.17 (Figure 1). These lines are in an unusual geological position in that, beneath a thin sedimentary cover of the Surat and Clarence-Moreton Basins, they traverse the northern part of the Texas orocline and run for a considerable distance essentially parallel to strike across the subsurface extension of the Tamworth Belt.

a network of industry seismic sections in the vicinity of line BMR84.14 (Totterdell et al 1995; Korsch & Totterdell 1995a), provides no evidence for the Leichhardt and Moonie Faults to be extensional and bounding faults of a half-graben, as proposed by Korsch et al (1992). However, modelling by Krassay et al (in press) of gravity data collected along profile BMR84.14 indicates that the Meandarra Gravity Ridge, a gravity high across the Taroom Trough, can be explained by a pile of mafic volcanic rocks, up to 8.5 km thick lying in a half-graben that is down to the west, that is, with the opposite polarity to that proposed by Korsch et al (1992). Support for this interpretation of the gravity anomaly is provided by cores intersected by petroleum exploration wells below the margins of the Taroom Trough interpreted here to be part of the synrift volcanics; they are exclusively basalt and andesite (Murray 1994).

Bowen and Surat Basins

Tamworth Belt

As displayed on seismic line BMR84.14, the geometry of the Surat Basin and the Taroom Trough (Figure 1) of the Bowen Basin is well preserved (Figure 7). Deformation is confined to narrow zones associated with the thrust faults near the eastern margin of the Bowen Basin (Figure 8). This seismic section cuts the eastern fault system of the Bowen Basin at a relay or accommodation zone between the Leichhardt Fault to the north and the Moonie Fault to the south. The Moonie Fault is a low-angle thrust fault with a frontal or short-cut thrust with limited displacement. This fault shows a typical fault-bend fold geometry with major movement occurring after deposition ceased in the Bowen Basin, but prior to commencement of deposition in the Surat Basin, that is, in Mid-Late Triassic time (Korsch & Totterdell 1995a, b). The succession in the anticline, including that above the upper flat of the fault, has been eroded so that the Surat succession sits directly on a basement of the Tamworth Belt east of the Moonie Fault. This geometry, coupled with detailed mapping of

The northern part of the accretionary wedge in New England (Texas Block) has been involved in oroclinal bending which has produced megafolds in the Texas and Coffs Harbour areas, and has led to a widening of the New England Orogen in this area (Korsch & Harrington 1987; Murray et al 1987). Interpretation of aeromagnetic and gravity anomalies of the New England Orogen, including the area covered by younger sedimentary rocks, by Wellman and Korsch (1988) and Wellman (1990) supported the oroclinal bending in the Texas and Coffs Harbour regions. The magnetic anomaly pattern also implies that rocks of the forearc basin succession (Tamworth and Yarrol belts) have been involved in the oroclinal bending, as previously inferred by Korsch and Harrington (1987). On lines BMR84.14, BMR84.16 and BMR86.17 (Figure 1) the area immediately beneath the Surat and Clarence-Moreton Basins between the Bowen Basin in the west and Esk Trough in the east is interpreted to be the Tamworth Belt or its equivalent (Figure 7). We

NEW ENGLAND OROGEN BENEATH MESOZOIC SEDIMENTARY COVER: 1984 AND 1986 DEEP SEISMIC SURVEYS

38

R.J. K O R S C H ET AL.

6250

Figure 8 Panel of deep seismic data from BMR84.14 across the easternmost Bowen Basin showing crust dominated by major thrust faults (see Figure 1 for location). Vertical scale approximately equals horizontal scale assuming an average crustal velocity of 6000 ms- . 1

consider the eastern limit of the Tamworth Belt to be the West Ipswich Fault (part of the Great Moreton Fault System), which thus represents the contact between the forearc basin to the west and accretionary wedge (South D'Aguilar Block) to the east (Figure 7). Thus the Great Moreton Fault System is the equivalent of the Peel Fault in the southern New England Orogen (Figure 5). This fault has been reactivated in the Triassic as a transtensional fault and now forms the eastern boundary fault of the Esk Trough (Korsch et al. 1989). Thus its original nature has been destroyed. Harrington and Korsch (1987) considered that the oroclinal bending was confined to the upper part of the crust, with movement taking place on a gently dipping detachment that allowed sliding at shallow to midcrustal levels. This is because the orocline cannot be expected to continue to an indefinite depth. A midcrustal detachment dipping gently westwards (Figure 7) was inferred to be this detachment surface by Korsch et al. (1990a). Esk Trough The Esk Trough is a Late Permian to Early Triassic basin. Based on seismic line BMR84.16 (Figures 9, 10),

Korsch et al. (1989) interpreted it to be an obliqueextensional basin, although an alternative interpretation by Elliott (1993) considered the sediments to represent foreland basin fill related to west-directed thrusting at its eastern margin. However, Elliott's interpretation does not explain the presence of abundant volcanic rocks within the Esk Trough succession. We suggest that the extensional event represents the initiation of extensional post-orogenic collapse of the orogenic mountain belt. Collapse commenced in the east behind the westward migrating thrust front, which by the Late Permian — Early Triassic had propagated well to the west, probably just to the east of the present eastern boundary of the Bowen Basin. Accretionary wedge The easternmost end of seismic line BMR84.16 was recorded across the Beenleigh Block (Figure 9), which is interpreted as part of an accretionary wedge (Murray et al. 1987). The seismic section (Figure 10) is characterised by an upper poorly reflective zone which is underlain by a strongly reflective zone referred to as a deep layered sequence by Korsch et al. (1986). Recent field work to the north of the deep seismic profile by

D E E P S E I S M I C P R O F I L I N G , NEW E N G L A N D O R O G E N

39

bCaloundra 20 km

• Gatton

geenleigh

Surfers Paradise

Coolangatta

Warwick Quaternary sediments

| + + + | Late Palaeozoic - Mesozoic granitoids

Tertiary volcanics

j

Tertiary sediments

n

Nambour Basin (Triassic - Jurassic) Clarence-Moreton Basin (Triassic Jurassic) Ipswich Basin (Triassic) Esk Trough (Triassic)

' | Yarraman Block (Palaeozoic)

v, -

South D'Aguilar and Northbrook Blocks (Palaeozoic)

1

p ^ C i ] North D'Aguilar Block (Palaeozoic) llllll

Beenleigh Terrane (Palaeozoic)

16/Q/234

Fault

F i g u r e 9 Simplified geological map of southeastern Queensland showing location of the eastern part of seismic line BMR84.16 across the Clarence-Moreton Basin to the Beenleigh Block. X Y is that part shown in Figure 10.

40

R.J. K O R S C H ET AL. BMR84.16

TMI(nl) Bouguer gravity

(jim.s') 2

1700 J

Ipswich Basin

Esk Trough / 1800

gft m - ^ n t a r a n r p - Mnrptnn Basin g

/

2000

\

2100

Beenleigh Block 2200

Elevation (m)

L

Rjifsgnent^ j & g'

: •: ' South D'Aguilar

Block

j a m worth Belt

Accretionary wedge (tower plate)

26/Q/445

20

Figure 10 Part of unmigrated seismic line BMR84.16 in the easternmost New England Orogen (see Figure 9 for location).

Little et al. (1992) indicated that accretionary-wedge rocks of the North D'Aguilar Block can be divided into two, separated by a major extensional fault, the Mt Mia Fault, which is folded. Following the interpretation of Little et al. (1992), the poorly reflective uppermost unit in the seismic data (Figure 10) is equivalent to the upper plate rocks of the South D'Aguilar and Beenleigh Blocks, and the deep layered sequence is equivalent to the lower plate, the contact between the two units being a southern equivalent of the folded Mt Mia Fault plunging to the south. Crustal architecture

The aeromagnetic image of Australia (gradientenhanced residuals of total magnetic intensity, Tarlowski et al. 1993) indicates that rocks of the Lachlan Orogen can be traced beneath the sediment cover on the western side of the Bowen Basin north to about Roma (Figure 1). This implies that rocks beneath the western Taroom Trough and associated Early Permian half-graben volcanic fill (Figure 7) represent

the Lachlan Orogen. Rocks beneath the far western side of the Bowen Basin to the west and north of Roma, however, are interpreted to be the Thomson Orogen by Wellman (1990, 1995). This suggests that in the vicinity of line BMR84.14, the Thomson Orogen is a wedge that thins to the east and is thrust above the Lachlan Orogen (Figure 7). We suggest that the eastern limit of the Lachlan Orogen is now defined by the low-angle fault that bounds the western margin of the Early Permian half-graben volcanic fill (Figure 7). Thus the Tamworth Belt (forearc) occurs between the Leichhardt Fault and the eastern margin of the Esk Trough and extends into the middle crust as far as the gently west-dipping detachment. To the east of the Esk Trough, we infer the rocks to be the upper plate of the accretionary wedge. Below the gently dipping detachment, which is the equivalent of the Mt Mia Fault, we infer the rocks to be the lower plate of the accretionary wedge. This indicates that the lower-plate component of the accretionary wedge extends well to the west, beneath the oroclinally folded Tamworth Belt. Thus beneath the Bowen Basin, rocks of the New England Orogen have been thrust beneath those of the Lachlan Orogen.

DEEP SEISMIC PROFILING, NEW ENGLAND OROGEN

i l l l

Gogango Thrust Zone

Shoalwater Terrane

Connors Arc

Yarrol Belt

Marlborough Terrane

Wandilla Terrane

41

Fault

Figure 11 Simplified geological map of the central Bowen Basin and New England Orogen showing locations of the 1989 BMR deep seismic reflection survey lines.

The base of the crust is interpreted as the bottom of the reflective zone (Figure 7); farther west in the Eromanga Basin, the base of the reflective zone correlates with the Moho defined by seismic refraction profiling (Finlayson & Collins 1987). A remaining problem to be addressed in this area is the location of the Devonian-Carboniferous magmatic arc. In the northern New England Orogen, the magmatic arc associated with the Late Devonian - Carboniferous forearc basin (Yarrol Belt) and accretionary wedge (Wandilla and Shoalwater terranes) is exposed as the Auburn Arc. South of about 26°S, however, the arc is not exposed, either buried by younger sediment, overthrust, or removed by thrusting or strike-slip faulting. We suggest that the arc could occupy the triangularshaped area beneath the Leichhardt Fault (Figure 7). This would then place it in the subsurface along strike from the known magmatic arc (Auburn Arc) which, to the north, is covered by shallow Surat Basin sediments and then exposed farther to the north. An alternative is that the triangular-shaped area could be part of the Tamworth Belt and that the magmatic arc, which is not necessarily continuous, is not present in this area.

NORTHERN NEW ENGLAND OROGEN: 1989 DEEP SEISMIC SURVEY Bowen Basin: Deep seismic line BMR89.B01 Hammond (1988, 1990) proposed an extensional origin for the Bowen Basin, postulating compartments bounded by northwest-striking extensional faults and northeast-striking transfer faults. Seismic line BMR89. B01 was positioned to follow a corridor between two inferred transfer faults postulated by Hammond (1988) and to test his extensional tectonics model. In the west, the line starts to the east of the northern Denison Trough and in the east ends just beyond the Eocene Duaringa Basin and above Permian rocks of the Gogango Overfolded Zone and older rocks of the Connors Arc (Figure 11). (The Gogango Overfolded Zone is dominated by thrust faults which appear to have controlled the deformation during shortening. Thus to call the zone an overfolded one may be a misnomer, and the name Gogango Thrust Zone may be more appropriate.) The western part of the Bowen Basin overlies the Thomson Orogen, whereas the highly deformed sedimentary

42

R. J. K O R S C H ET

AL. Blackwater Zone

Bouguer gravity

40Q

Elevation

400

(m,

BMR89.B03

C o m e t Ridge

0

0

1700 B a s e ™ a s s i c 1900

0 Base Permian sediments

25 km 26/Q/446A J

Figure 12 Unmigrated deep seismic profile BMR.B01 across the Bowen Basin and Duaringa Basin (see Figure 11 for location).

rocks of the Gogango Thrust Zone in the east (Figure 11) is considered to be part of the New England Orogen by, for example, Fergusson (1991), but part of the original Bowen Basin depositional system by Fielding and Stephens (1994). The upper part of the seismic section is dominated by an imbricate thrust fan that soles into an east-dipping detachment (Figure 12). At the Jellinbah Thrust, the Late Permian coal measures have been thrust over the Early Triassic Rewan Group. The detachment appears to flatten and become more diffuse in the ductile zone in the middle to lower crust (at ~7 s TWT beneath the Duaringa Basin, Figure 12). Within the section dominated by the dipping detachment, the Blackwater, Jellinbah and Dawson structural zones in the Bowen Basin, recognised by Hobbs (1985), exhibit distinct structural styles both in their surface geology and seismic characteristics (see Korsch et al. 1992). In the Blackwater Zone, to the west of the Jellinbah Thrust, the uppermost part of the seismic section is dominated by subhorizontal reflections, interpreted as

mainly sedimentary layering (Figure 12). The PermoTriassic wedge appears to consist of a conformable sedimentary succession, but subhorizontal, beddingparallel thrust faults with postulated displacements of up to 5 km are known (Hobbs 1985). In the Jellinbah Zone (previously referred to as the Yarrabee Zone, but renamed here to conform with the convention of naming thrust sheets after their underlying thrusts), the seismic section is dominated by a series of listric thrust faults in an imbricate thrust fan that soles into a detachment fault which dips gently to the east (Korsch et al. 1992). Ramp anticlines have developed in the hangingwall above some thrusts. The boundary between the Late Permian coal measures and overlying Early Triassic sandstones shows minor displacement at several localities (Korsch et al. 1992 figures 10, 11). Late Permian sedimentary rocks have been thrust over Early Triassic ones, indicating that the deformation, at least in this area, is Middle Triassic or younger. The Dawson Fold Zone, to the east of the Yarrabee Thrust, consists of tightly folded Late Permian

43

DEEP SEISMIC PROFILING, NEW ENGLAND OROGEN Jellinbah Zone

Dawson Fold Zone

BMR89.B01

Duaringa Basin

NE

Permian sediments

Bouguer gravity (nm.s

2)

Elevation (m)

Moho

Vertical scale approximately equals horizontal scale assuming a crustal velocity of 6000 ms-

sedimentary rocks on the surface (Fergusson 1991). Because of the steep dips, seismic reflectivity is low, but the most obvious features on the seismic section are strong reflections that dip moderately to the east. We interpret these as thrust faults above the east-dipping detachment. The surface folds thus are interpreted to be ramp anticlines related to underlying thrusts. The base of the Dawson Fold Zone is the major detachment. Below the Blackwater Zone and the east-dipping detachment, strong, east-dipping reflections also occur at levels deeper than 1.5 s TWT (Figure 12). We interpret these reflections as major thrust faults in the basement. In the eastern part of line BMR89.B01, the Eocene Duaringa Basin has a strong asymmetric geometry (Figure 12; see also Korsch et al. 1990b figure 4) and consists of over 1200 m of terrestrial sediments containing considerable resources of oil shale. The basin is bounded on the western side by a very steep to subvertical synsedimentary fault (Figure 12), which we interpret to have a strong strike-slip motion. Thus we

consider the basin to be transtensional in origin and was superimposed on top of the older Bowen Basin system. The easternmost part of BMR89.B01, beyond the eastern feather edge of the Duaringa Basin, is the Gogango Thrust Zone, with the last few shotpoints being recorded on rocks of the Connors Arc (Figure 12). This zone is dominated by thrust faults and highly cleaved, tight to isoclinal folds in rocks of Early Permian or older age. Here the section is again dominated by relatively planar reflections that dip to the east, and which we interpret as thrust faults related to the main contractional episode. The seismic data suggest that the dominant structures in the Gogango Thrust Zone, at least on a large scale, are these thrust faults. The Early Permian rocks of the Gogango Thrust Zone are intruded by post-orogenic plutons of probable Late Permian to Early Triassic age; hence the age of the thrusting is constrained to being younger than the Early Permian but older than latest Permian. Farther to the west, in the Dawson and Jellinbah Zones, the thrusts are younger, because Late Permian rocks are thrust over

44

R.J. KORSCH ET AL "Transfer zone"

.<

>.

Bowen Basin BMR89.B03

BMR89.B01

"Transfer zone" Bouguer gravity (jxm.s 2 ) Elevation (m)

Base Permian sediments

^Moho

Figure 13 Unmigrated deep seismic profile BMR89.B03 oriented approximately north-south across the Bowen Basin to test for the presence or absence of transfer faults (see Figure 11 for location). Vertical scale approximately equals horizontal scale assuming an average crustal velocity of 6000 ms- 1 .

DEEP SEISMIC PROFILING, NEW ENGLAND OROGEN Early Triassic ones. This is consistent with observations in other thrust belts around the world where the thrusts generally propagate into the foreland sequentially. Deep crustal reflections are evident between 7 and 12 s TWT but are strongest between about 9 and 12 s TWT (Figure 12). Reflections defining the Moho appear to deepen from about 10.5 s TWT under the western part of the line to about 12.5 s TWT under the eastern part of the basin. There are very few reflection events from the upper mantle. Refraction data collected in 1973 indicate a Moho at about 36 km depth under the central part of the seismic line (Collins 1978). In summary, seismic line BMR89.B01, across the exposed part of the Bowen Basin, is dominated by listric thrust faults that root onto a major east-dipping detachment which appears to flatten and diffuse in the middle to lower crust. In the eastern part of this area, the deformation is so intense that indications of the original geometry of the basin before thrusting are all but obliterated. The seismic data show the dominant role of low-angle thrust faults in controlling the last major deformational event in the sedimentary succession. This seismic section provides an excellent example of the thin-skinned style of thrust tectonics. Also, the line contains east-dipping structures in the basement, interpreted as part of a major crustal-scale thrust system. If they are thrusts, then the New England Orogen has been thrust westwards over the basement (= Thomson Orogen). The thrust front that initiated in the New England Orogen to the east has propagated a long way into the basin, and this is not seen elsewhere in the basin system to the south. Western Bowen Basin: Deep seismic line BMR89.B03 Seismic line BMR89.B03 was recorded in a north-south direction to run approximately parallel to the main structural grain of the basin and was positioned to intersect two transfer zones inferred by Hammond (1988). In this area, the Bowen Basin sits on crust of the Thomson Orogen, which consists of a non-reflective upper crust (0.5-5 s TWT) and a strongly reflective middle to lower crust (5-11 s TWT) (Figure 13). No evidence is seen for major crustal-scale dipping structures. The reflection Moho is defined as the base of the strongly reflective zone at about 11 s. The data in the upper part of the seismic section deteriorates towards the south as the Tertiary basalt capping gets thicker above the Bowen Basin (Figure 13). Relatively continuous reflections in the section suggest that the northern transfer zone does not exist. The seismic data do not deny the possible existence of the inferred southern transfer zone, but its maximum depth must be limited to less than about 5 s TWT. At this depth, the inferred transfer zone, if it exists, must either detach onto a subhorizontal fault or diffuse into the brittleductile transition zone, and thus cannot have been a deep-seated structure that acted as a plumbing system for the rise of magma, as postulated by Hammond (1988).

45

Northern New England Orogen: Deep seismic line BMR89.B02 Seismic line BMR89.B02 is the eastern extension of line BMR89.B01 and was planned to cross the eastern margin of the Bowen Basin onto basement of the New England Orogen. Because it was not possible to find a suitable route to the east of the eastern end of line BMR89.B01, it was recorded further to the north along the Marlborough-Sarina road (Figure 12). The western end of the line was recorded on the Connors Volcanics (magmatic arc) and the eastern end on Marlborough Serpentinite; in between, the seismic line crossed the northern part of the Gogango Thrust Zone. This line is dominated by reflections that dip moderately to the east (Figure 14), which we interpret as crustal-scale thrusts forming part of an imbricate thrust system, essentially confirming the geometry figured by Fergusson (1991). Between these faults, the rocks have a fabric with variable dips. The eastern end of line BMR89.B02 crosses an area of serpentinite, with minor quartz-mica schist and granite (Figure 15). Here an increase in gravity coincides with the western limit of the serpentinite. It is difficult to detect the relationships of the serpentinite to the schist and the felsic pluton, but the seismic data suggest that the serpentinite is essentially flat-lying and less than 1 s TWT, that is, about 3 km thick, confirming previous gravity modelling and geological interpretations that the Marlborough Serpentinite was a thin flat sheet (Murray 1974). Thus the serpentinite and associated quartz-mica schist and granite are relatively thin skinned and appear to have been emplaced late in the thrusting, above a deformed basement dominated by more steeply dipping earlier thrusts in the New England Orogen. (Figures 14, 15).

DISCUSSION Figure 16 shows our interpretation of the crustal architecture of the three main deep seismic profiles that cross the suture between the Lachlan and New England Orogens. These three sections show major differences in crustal architecture along the length of the New England Orogen and its foreland basin system. The Bowen Basin lines are dominated by east-dipping structures, with the New England Orogen thrust over the Thomson Orogen. The lines in the Surat and ClarenceMoreton Basins in southernmost Queensland are dominated by gently west-dipping structures in the lower crust, but east-dipping thrusts in the upper crust. The accretionary-wedge part of the New England Orogen is thrust beneath the easternmost Lachlan Orogen, but with the forearc basin (Tamworth Belt) of the New England Orogen being obducted (backthrust) on top of the Lachlan Orogen. The line in the Gunnedah Basin and southern New England Orogen contains both east and west dipping structures and shows similar relationships between the New England and Lachlan Orogens as in the lines in southernmost Queensland.

46

R.J. K O R S C H E T

AL.

W Connors Arc

BMR89.B02 Marlborough Serpentinite

Gogango Thrust Zone

400 0 -400

800

400

o

Bouguer gravity ([i m.s "2) Elevation (m)

2000

\ B a s e of serpentinite

Moho

20 km

Figure 14 Unmigrated deep seismic profile BMR89.B02 across the elements of the northern New England Orogen including the Connors Arc and Marlborough Serpentinite (see Figure 11 for location). Vertical scale approximately equals horizontal scale assuming an average crustal velocity of 6000 ms- . 1

47

DEEP SEISMIC PROFILING, NEW ENGLAND OROGEN SW

BMR89.B02

0

NE

10 km

26/Q/449

Figure 15 Enlargement of part of unmigrated seismic profile BMR89.B02 across the Marlborough Serpentinite and Gogango Thrust Zone (see Figure 11 for location). Vertical scale approximately equals horizontal scale assuming an upper crustal velocity of 6000 ms-'.

There are also major changes in the upper crustal geometries along the length of the orogens and basin system, best reflected by the change in the nature of the fault systems (Korsch & Totterdell 1995c). That is, the upper crust is partitioned into zones of different deformational style which is mirrored by fundamental changes in crustal architecture along the length of the basins. Thus the crustal geometries exert some control over the nature of the structures that developed during later contractional events. Also, our interpretation predicts very different relationships between the LachlanThomson and New England Orogens from north to south. Nevertheless, the crustal architecture inherits a strong component from its time as an active convergent margin. The differences in crustal architecture seen in the BMR profiles indicate that between the seismic profiles there must be zones where there are fundamental changes in the architecture of the crust. Thus the crust is divided into compartments which have behaved differently through time. The compartments are defined by the inherited crustal architecture of the basement (palaeo-continental crust) which exerts a lingering influence on later crustal geometry and upper crustal components such as basin geometry and nature of the thrust systems. The crustal architectures interpreted in the southernmost Queensland and New South Wales deep seismic

data show that the New England Orogen developed as a doubly vergent orogen (see dynamic models of Koons 1990; Willett et al 1993; that are underpinned by the kinematic model developed by Wellman 1979). Farther to the north, the lines across the Bowen Basin are dominated by east-dipping structures, which are equivalent to only the western components of the southern lines. That is, in the north, as far east as the present-day coastline, we only see the western part of the doubly vergent orogen. The seismic data suggest that Cambrian ophiolitic rocks from the Peel Fault system (Aitchison & Ireland 1995) form either the floor of the Tamworth Belt or represent fragments of the easternmost margin of the Lachlan Orogen that have been pulled from beneath the obducted Tam worth Belt (Korsch et al. 1993; Glen & Brown 1995). The west-dipping structure, which at the surface is just to the east of the Peel Fault, appears to extend to at least the Moho, and hence fragments of the Lachlan Orogen could be brought to the surface from the lower crust and upper mantle. Hence, we infer that this structure defines the eastern limit of the Lachlan Orogen at depth; to the east of it is crust only of the New England Orogen. The discovery of Cambrian paraconodonts from the Pipeclay Creek Formation (Stewart 1995) indicates the presence of Cambrian sedimentary rocks beneath the main Devonian-Carboniferous pile in the Tamworth Belt, supporting the idea that at least

48

R. J. K O R S C H ET

SOUTHWEST co 0

AL.

DUARINGA BASIN BOWEN BASIN _ , . Jellinbah „ _ ,, Comet Blackwater Z o n e Dawson Fold Ridge | Zone j r J | Zone

Gogango

Thrust Zone

NORTHEAST I Serpentinite

V -

= 1

H 50 km

l

^

^ ' X ^ / ^ ^ T !

6

V X ^ ^

^

1r h N / N i u P P e r M a n t l e x N / w \ i v , WEST BMR84.14 (EAST)

SURAT BASIN

>, \ X >, BMR89.BQ1

" " 7 r ; v x 1w v \

BMR89.BQ2""

/ \ /r - v w ^ v \ / \ / { - v \ / \ / \ / \ / ^ BMR86.17

CLARENCE-MORETON BASIN

BMRM.

BEENLEIGH BLOCK (UPPER PLATE) EAST

Figure 16 Diagrams of the crustal architectures in the vicinity of the three main deep seismic surveys showing the relationships between the New England Orogen and the Lachlan and Thomson Orogens.

some of the ophiolites were the floor of the Tamworth Belt. The deep seismic profiles show that, for the units interpreted as Thomson Orogen, there is a non-reflective upper crust, below which is a strongly layered middle to lower crust. This crustal character can be seen in line BMR89.B03 (Figure 13) and in the western part of BMR84.14 (Figure 7). The same non-reflective upper crust is seen on line BMR60.H (Robertson 1965) on the eastern Anakie Inlier which forms part of the Thomson Orogen, and beneath the Denison Trough in line BMR78.08 (Mathur 1983). The non-reflective zone is interpreted to be highly deformed metasedimentary rocks and/or homogeneous granites. This crustal character appears to be typical of the Thomson Orogen elsewhere, and also has been imaged by deep seismic surveys recorded farther to the west (Mathur 1984; Finlayson et al. 1990; Leven et al. 1990). The same crustal character is observed on the western end of BMR89.G01 (Figure 3), where it is interpreted to be the Lachlan Orogen. It is also seen elsewhere in the Lachlan Orogen in other deep seismic lines such as in the Cobar Basin in New South Wales (Drummond et al. 1992), the Heathcote fault zone in Victoria (Gray et al. 1991) and beneath the Otway Basin in southern Victoria (Finlayson et al. 1996).

The reflective lower crust beneath the Gunnedah Basin could represent the lower part of the Lachlan Orogen or Proterozoic crust that was extensively thinned during Late Neoproterozoic or Early Palaeozoic crustal extension. This helps to explain the suggestion that some granites and volcanics in the forearc were derived from either a Proterozoic source or from a source that contained Proterozoic-derived material (see Korsch etal. 1993a).

CONCLUSIONS

(1) Deep seismic data in eastern Australia have provided an insight into present-day crustal architecture, with the New England Orogen being a doubly vergent orogen that developed in the Late Palaeozoic to Triassic. (2) In eastern Australia, the main deep seismic lines spaced approximately 500 km apart, although seemingly in the same structural position, show a very different crustal architecture along the suture between the New England and the Lachlan-Thomson Orogens. (3) Compartments defined by the pre-existing basement crustal architecture (palaeo-continental crust) exert a lingering influence on later crustal geometry and

DEEP SEISMIC PROFILING, NEW ENGLAND OROGEN

upper crustal components such as basin geometry and younger thrust systems. (4) The suture between the New England Orogen and the Lachlan and Thomson Orogens is complex, having a different geometry at different positions along the suture. (5) In the southern New England Orogen, the boundary between the Tamworth Belt (forearc) and Tablelands Complex (accretionary wedge) is a major west-dipping fault that extends to at least the Moho. The Peel Fault is a late stage, high level structure that roots onto this west-dipping fault. Thus, the deep seismic data in eastern Australia show present-day crustal architectures that are the responses of the crust to interplate and later Permian and Mesozoic intraplate deformational events superimposed on a crustal architecture that was established by a subduction-related convergent plate margin in the Late Palaeozoic. ACKNOWLEDGMENTS This paper was presented at the A. H. Voisey Memorial Symposium, Macquarie University, 8 December 1995. RJK is pleased to contribute to a volume dedicated to the memory of Professor Voisey, because he was responsible for triggering RJK's interest in geology. We thank other members of the AGSO Land Seismic Group and members of the various seismic field parties for assisting with the collection of the deep seismic reflection data. Some of the data were collected as a contribution to the NGMA Sedimentary Basins of Eastern Australia project, which is a cooperative project between the Geological Survey of Queensland, Geological Survey of New South Wales and AGSO; we thank other members of the NGMA project in the three organisations for many useful discussions. We also thank Graham Brown and Pat Pratt of AGSO for drafting the figures and R. A. Glen, B. R. Goleby, J. M. Totterdell and an anonymous reviewer for their helpful comments on the manuscript. The paper is published with the permission of the Executive Director, Australian Geological Survey Organisation and the Director of the Australian Geodynamics Cooperative Research Centre. REFERENCES

AITCHISON J. C. & I R E L A N D T. R. 1 9 9 5 . Age profile of

ophiolitic rocks across the Late Palaeozoic New England Orogen, New South Wales: implications for tectonic models. Australian Journal of Earth Sciences 42, 1 1 - 2 3 .

BECKETT J., GLEN R., KORSCH R. J. & BOREHAM C. J. 1995.

Hydrocarbons below the Mooki Thrust: a Cordilleran model for petroleum potential in the Maules Creek Subbasin, Gunnedah, New South Wales. In: Morton D. J., Alder J. D., Grierson I. J. & Thomas C. E. eds. 1995 NSW Petroleum Symposium Proceedings, 16 pp. (no pagination). Petroleum Exploration Society of Australia, New South Wales Branch.

49

BLAKE M . C. JR. & MURCHEY B. L. 1988. A California model

for the New England Fold Belt. Quarterly Notes of the Geological Survey of New South Wales 72, 1-9. CAWOOD P. A. 1982. Structural relations in the subduction complex of the Paleozoic New England Fold Belt, eastern Australia. Journal of Geology 90, 381-392. COLLINS C. D. N. 1978. Crustal structure of the central Bowen Basin, Queensland. BMR Journal of Australian Geology & Geophysics 3, 203-209.

DRUMMOND B. J., GOLEBY B. R., WAKE-DYSTER K. D., GLEN R. A. & PALMER D. 1992. New tectonic model for the

Cobar Basin, NSW points to new exploration targets in

Lachlan Fold Belt. BMR Research Newsletter 16, 1 6 - 1 7 . ELLIOTT L. G. 1993. Post-Carboniferous tectonic evolution of eastern Australia. The APEA Journal 33, 215-236.

FERGUSSON C. L. 1991. Thin-skinned thrusting in the northern

New England Orogen, central Queensland, Australia. Tectonics 10, 797-806.

FERGUSSON C. L. HENDERSON R. A., LEITCH E. C. & ISHIGA

H. 1993. Lithology and structure of the Wandilla terrane, Gladstone-Yeppoon district, central Queensland, and an overview of the Palaeozoic subduction complex of the New England Fold belt. Australian Journal of Earth Sciences 40, 403-414. FIELDING C. R.& STEPHENS C. J. 1994. New insights into the eastern margin of the Bowen Basin, Queensland. Geological Society of Australia Abstracts 37, 108. FINLAYSON D. M. (Editor) 1990. The Eromanga-Brisbane geoscience transect: a guide to basin development across Phanerozoic Australia in southern Queensland. Bureau of Mineral Resources Bulletin 232. FINLAYSON D. M . & COLLINS C . D . N. 1 9 8 7 . Crustal differences between the Nebine Ridge and the central Eromanga Basin from seismic data. Australian Journal of Earth Sciences 34, 2 5 1 - 2 5 9 . FINLAYSON D. M . & COLLINS C. D . N. 1 9 9 3 . Lithospheric velocity structures under the southern New England Orogen: evidence for underplating at the Tasman Sea margin. Australian Journal of Earth Sciences 4 0 , 1 4 1 - 1 5 3 . FINLAYSON D. M . , JOHNSTONE D. W . , OWEN A . J. & WAKEDYSTER K. D. 1996. Deep seismic images and tectonic

framework of early rifting in the Otway Basin, Australian southern margin. Tectonophysics 264, 137-152.

FINLAYSON D. M . , WAKE-DYSTER K . D., LEVEN J. H. ETAL.

1990. Seismic imaging of major tectonic features in the crust of Phanerozoic eastern Australia. Tectonophysics 173,211-230. GLEN R. A. & BROWN R. E 1993. A transect through a forearc basin: preliminary results from a transect across the Tamworth Belt at Manilla N.S.W. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 105-111. Department of Geology & Geophysics, University of New England, Armidale. GLENR. A. & BROWN R. E 1995. Fault-related mineralisation in the Teatree Creek Goldfield, Tamworth Belt, New England Orogen. Geological Survey of New South Wales Quarterly Notes 97, 1-13.

GLEN R. A., KORSCH R . J. & WAKE-DYSTER K. D. 1993. A

deep seismic cross section through the Tamworth Belt: preliminary interpretation of 4 seconds two-way time data. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 1 0 1 - 1 0 4 . Department of Geology & Geophysics, University of New England, Armidale.

50

R.J. KORSCH ET AL.

G R A Y D . R . , WILSON C . J. L. & B A R T O N

T . J.

1991.

Intracrustal detachments and implications for crustal evolution within the Lachlan Fold belt, southeastern Australia. Geology 19, 574-577. HAMMOND R. L. 1988. The geological structure of the Bowen Basin. In: Mallett C. W., Hammond R. L., Leach J. H. J., Enever J. R. & Mengel C. eds. Bowen Basin—stress, structure and mining conditions: Assessment of mine planning, pp. 10-43. CSIRO Division of Geomechanics, NERDDC Project No. 901, Final Report. HAMMOND R. L. 1990. Modelling the geological structure of the Bowen Basin, present and future. In: Beeston J. W. ed. Bowen Basin Symposium 1990 Proceedings, pp. 7-9, Geological Society of Australia, Queensland Division, Brisbane. HARRINGTON H. J. & KORSCH R. J. 1985. Tectonic model for

the Devonian to middle Permian of the New England Orogen. Australian Journal of Earth Sciences 32, 163— 179.

HARRINGTON H. J. & KORSCH R. J. 1987. Oroclinal bending in

the evolution of the New England-Yarrol Orogen and the Moreton Basin. In: Pacific Rim Congress 1987, pp. 797800. Australasian Institute of Mining and Metallurgy, Melbourne. HOBBS B. E. 1985. Interpretation and analysis of structure in the Bowen Basin. Geological Society of Australia Abstracts 17, 151. KOONS P. O. 1990. Two-sided orogen: collision and erosion from the sandbox to the Southern Alps, New Zealand. Geology 18, 679-682. KORSCH R. J. & HARRINGTON H. J. 1987. Oroclinal bending,

fragmentation and deformation of terranes in the New England Orogen, eastern Australia. American Geophysical Union Geodynamics Series 19, 129-139. KORSCH R . J., HARRINGTON H . J., MURRAY C . G . , FERGUSSON

C. L. & FLOOD P. G. 1990a. Tectonics of the New England Orogen. Bureau of Mineral Resources Bulletin 232,35-52. KORSCH R . J., LINDSAY J. F., O'BRIEN P. B., SEXTON M . F. &

WAKE-DYSTER K. 1986. Deep crustal seismic reflection profiling, New England Orogen, eastern Australia: telescoping of the crust and a hidden deep layered sedimentary sequence. Geology 14, 982—985.

KORSCH R . J., WAKE-DYSTER K . D . & JOHNSTONE D .

W.

1990b. Deep seismic profiling across the Bowen Basin. In: Beeston J. W. ed. Bowen Basin Symposium 1990 Proceedings, pp. 10-14. Geological Society of Australia, Queensland Division, Brisbane. KORSCH R . J., W A K E - D Y S T E R K . D . & JOHNSTONE D.

W.

1992. Seismic imaging of extensional and contractional structures in the Bowen and Surat basins, eastern Australia. Tectonophysics 215, 273-294. KORSCH R . J., WAKE-DYSTER K . D . & JOHNSTONE D .

W.

1993a. The Gunnedah Basin - New England Orogen deep seismic reflection profile: implications for New England tectonics. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 85-100. Department of Geology & Geophysics, University of New England, Armidale. KORSCH R . J., WAKE-DYSTER K . D . & JOHNSTONE D .

W.

1993b. Deep seismic reflection profiling in the Gunnedah Basin: regional tectonics and basin responses. In: Swarbrick G. J. & Morton D. J. eds. Proceedings NSW Petroleum Symposium, 30 pp. (no pagination). Petroleum Exploration Society of Australia, New South Wales Branch. KRASSAY A . A . , KORSCH R . J. & D R U M M O N D B . J.

The

Meandarra Gravity High: symmetry elements of the gravity anomaly and its relationship to the BowenGunnedah-Sydney basin system. In: Korsch R. J. ed. Evolution and petroleum potential of the Bowen, Gunnedah and Surat Basins, eastern Australia. Australian Geological Survey Organisation Bulletin (in press). LEITCH E. C. 1993. The floor of the Gunnedah Basin north of the Liverpool Range. In: Tadros N. Z. ed. The Gunnedah Basin, New South Wales, pp. 335-348. Geological Survey of New South Wales Memoir Geology 12. LEVEN J. H . , FINLAYSON D . M . & WAKE-DYSTER K . D . 1 9 9 0 .

Mid-crustal detachments controlling basin deformation: ramp synforms in southwestern Queensland. Tectonophysics

173, 231-246.

LITTLE T . A . , HOLCOMBE R . J., GIBSON G . M . , O F F L E R R . , GANS P. B. & MCWILLIAMS M . O . 1 9 9 2 . E x h u m a t i o n o f

late Palaeozoic blueschists in Queensland, Australia, by extensional faulting. Geology 20, 231-234.

KORSCH R . J., O ' B R I E N P. E., SEXTON M . J., WAKE-DYSTER

MATHUR S. P. 1983. Deep reflection experiments in northeastern Australia, 1976-1978. Geophysics 48, 1588— 1597.

K. D. & WELLS A. T. 1989. Development of Mesozoic transtensional basins in easternmost Australia. Australian Journal of Earth Sciences 36, 13—28

MATHUR S. P. 1984. Deep crustal reflection results from the central Eromanga Basin, Australia. Tectonophysics 100, 163-173.

KORSCH R. J. & TOTTERDELL J. M. 1995a. Permian and

MURRAY C. G. 1974. Alpine-type ultramafics in the northern part of the Tasman Geosyncline—possible remnants of Palaeozoic ocean floor. In: Denmead A. K., Tweedale G. W. & Wilson A. F. eds. The Tasman Geosyncline—a Symposium, pp. 161-183. Geological Society of Australia, Queensland Division, Brisbane. MURRAY C. G. 1988. Tectonostratigraphic terranes in southeast Queensland. In: Hamilton L. H. ed. Field Excursions Handbook for the 9th Australian Geological Convention, pp. 19-51. Geological Society of Australia, Queensland Division, Brisbane.

Mesozoic tectonic and structural events in the Bowen and Surat Basins and New England Orogen, Southwest Pacific Rim. In: Mauk J. L. & St George J. D. eds Proceedings of the 1995 PACRIM Congress, pp. 305-310. Australasian Institute of Mining and Metallurgy Publication 9/95. KORSCH R. J. & TOTTERDELL J. M. 1995b. Structural events and deformational styles in the Bowen Basin. In: Follington I. W., Beeston J. W. & Hamilton L. H. eds. Bowen Basin Symposium 1995... 150 years on ... Proceedings, pp. 27-35. Geological Society of Australia, Coal Geology Group, Brisbane. KORSCH R. J. & TOTTERDELL J. M. 1995c. Eastern margin of the Bowen-Gunnedah-Sydney Basin: geometry of the Burunga - Leichhardt - Moonie - Goondiwindi - Mooki Hunter fault system. Geological Society of Australia Abstracts 40, 85-86.

MURRAY C. G. 1994. Basement cores from the Roma Shelf, New England Fold Belt and Taroom Trough beneath the Surat and Moreton basins. Queensland Government Mining Journal 95, 17—34. MURRAY C . G . , FERGUSSON C . L., FLOOD P. G . , WHITAKER W .

G. & KORSCH R. J. 1987. Plate tectonic model for the

51

DEEP SEISMIC PROFILING, NEW ENGLAND OROGEN Carboniferous evolution of the New England Fold Belt. Australian Journal of Earth Sciences 34, 213-236. MURRAY C . G . , SCHEIBNER E. & W A L K E R R . N .

1989.

Regional geological interpretation of a digital coloured residual Bouguer gravity image of eastern Australia with a wavelength cut-off of 250 km. Australian Journal of Earth Sciences 36, 423-449. POGSON D . J. & HITCHINS B. L. 1 9 7 3 . New

England

1:500

000 geological map sheet. Geological Survey of New South Wales, Sydney. RAMSAY W . R . H . & S T A N L E Y J . M .

1976.

Magnetic

anomalies over the western margin of the New England foldbelt, northeast New South Wales. Geological Society of America Bulletin 87, 1421-1428. ROBERTSON C. S. 1965. Emerald-Duaringa seismic survey, Queensland 1960. Bureau of Mineral Resources Record 1965/2. SCHEIBNER E. 1996. Geology of New South Wales— Synthesis. Volume 1 Structural framework. Geological Survey of New South Wales Memoir Geology 13. SHAW

S. E. &

FLOOD R . H .

1981. T h e N e w

England

Batholith, eastern Australia: geochemical variations in time and space. Journal of Geophysical Research 86, 10530-10544.

STEWART I. 1995. Cambrian age for the Pipeclay Creek Formation, Tamworth Belt, northern New South Wales. Courier Forschungsinstitut Senckenberg 182, 565-566. TADROS N. Z. 1993. The Gunnedah Basin New South Wales. Geological Survey of New South Wales Memoir Geology 12. TARLOWSKI C . , SIMONIS F. & MILLIGAN P . 1 9 9 3 .

Magnetic

anomaly map of Australia, scale 1:5 000 000. Australian Geological Survey Organisation, Canberra.

Hamilton L. H. eds. Bowen Basin Symposium 1995... 150 years on ... Proceedings, pp. 247-256. Geological Society of Australia, Coal Geology Group, Brisbane. WAKE-DYSTER K . D . , SEXTON M . J., JOHNSTONE

W.,

Society 89, 423-430.

WELLMAN H. W. 1979. An uplift map for the South Island of New Zealand, and a model for uplift of the Southern Alps. In: Walcott R. I & Cresswell M. M. eds. The origin of the Southern Alps, pp. 13-20. Royal Society of New Zealand Bulletin 18. WELLMAN P. 1990. A tectonic interpretation of the gravity and magnetic anomalies in southern Queensland. Bureau of Mineral Resources Bulletin 232, 21-34. WELLMAN P. 1995. Tasman Orogenic System: a model for its subdivision and growth history based on gravity and magnetic anomalies. Economic Geology 90, 1430-1442. WELLMAN P. & KORSCH R. J. 1988. Oroclinal bending of the

New England Orogen — a refinement using magnetic and gravity anomalies. Geological Society of Australia Abstracts 21, 417. WHITE A. H. 1964. Geological map of New England 1:100000, Attunga sheet (No. 321), with marginal text. Department of Geology, University of New England, Armidale. WILLETTS., BEAUMONT C . & FULLSACK P. 1 9 9 3 . M e c h a n i c a l

model for the tectonics of doubly vergent compressional orogens. Geology 21, 371-374. WOODWARD N. B. 1995. Thrust systems in the Tamworth Zone, southern New England Orogen, New South Wales. Australian Journal of Earth Sciences 42, 107—117.

TOTTERDELL J. M . , BRAKEL A . J., WELLS A . T . & HOFFMANN

K. L., 1995. Basin phases and sequence stratigraphy of the Bowen Basin. In: Follington I. W., Beeston J. W. &

D.

WRIGHT C. & FINLAYSON D. M. 1987. A deep seismic

profile of 800 km length recorded in southern Queensland, Australia. Geophysical Journal of the Royal Astronomical

(Received 24 April 1996; accepted 28 June 1997)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 52-65.

Tectonic evolution of the northern New England Fold Belt: the Permian-Triassic Hunter-Bowen event R. J. HOLCOMBE,1 C. J. STEPHENS,1* C. R. FIELDING,1 D. GUST,2 T. A. LITTLE,U R. SLIWA,11 J. KASSAN,1# J. McPHIE3 AND A. EWART1 2

1 Department of Earth Sciences, University of Queensland, Qld 4072, Australia. School of Geology, Queensland University of Technology, GPO Box 2434, Brisbane, Qld 4001, Australia. 3 CODES, University of Tasmania, GPO Box 252-79, Hobart, Tas. 7001, Australia.

The New England Fold Belt in Queensland is a complex arrangement of terranes with boundaries dominated by structures that were active during the contractional Hunter-Bowen Orogeny (event). This event extended for around 35 million years from Late Permian (ca 265 Ma) to late Middle Triassic (ca 230 Ma) time. The present north-northwest-trending structural grain of the fold belt is largely due to this deformation, but most faults have been reactivated during post-Late Triassic faulting. The northern New England Fold Belt can be subdivided into: (i) a northern region (Connors Arch and lateral structures) within which deformation is characterised by open folds and variable, but generally minor, thrusting; (ii) a central region of thin-skinned, fold-thrust deformation with cross-orogen tear faults, and within which strain is strongly partitioned and there is variable cleavage development (Gogango Overfolded Zone and more eastern terranes); and (iii) a southern region of thick-skinned deformation within which basement appears to have been involved in deformation. Within the central region, the Marlborough Block is a composite out-of-sequence thrust nappe terrane of ophiolitic components, low- and high-grade metasedimentary rocks, and metagranitoid, juxtaposed, at deeper structural levels, by early thrusting and finally thrust at least 80 km west over the fold-thrust belt. Various elements of the fold belt, including the Yarrol/Calliope terrane and Gympie Block, may be allochthonous elements transported for distances of tens to hundreds of kilometres within the fold belt. Calc-alkaline magmatism in the northern New England Fold Belt during the Early and Middle Triassic may have been in response to the initiation, or onshore migration, of a magmatic arc, and its termination coincided with the last phase of contraction. The tectonic regime became extensional by the Late Triassic with widespread granite intrusion and development of silicic volcanic complexes and localised extensional sedimentary basins. Pulses of contractional deformation in the fold belt are recorded by the distribution and nature of sediment in the foreland Bowen Basin to the west. Thrusting is indicated by the presence of coarse clastic wedges shed into the basin for greater distances at successively higher stratigraphic levels, reflecting the advancement of the thrust front. The final contractional event appears to have been re-initiated at the eastern margin of the fold belt, rather than stepped westward from the previous thrust front, and to have been more intense than previous pulses. Bowen Basin sedimentation closed at about 232 Ma following accumulation of the most voluminous of the clastic wedges. Fluid flow associated with shear and dilational structures during this final contractional event is thought to be responsible for gold mineralisation in the fold belt, and for widespread diagenetic mineralisation in sediments of the Bowen Basin. Key words: crustal convergence, fold-thrust belt, New England Fold Belt, tectonics.

INTRODUCTION

The present geometry of folded terranes in the New England Fold Belt in coastal Queensland (Figure 1) is largely controlled by structures produced during the Permo-Triassic Hunter-Bowen event yet its nature and timing have been poorly defined. In a companion paper (Holcombe et al. 1997), we discussed the constraints on the transition from active accretion in midCarboniferous times to widespread extension through the Late Carboniferous and Early Permian. This transition is interpreted in terms of eastward retreat of the convergent slab, and migration of the volcanic arc offshore. This paper presents a synthesis of our current understanding of the Late Permian to Late Triassic tectonic evolution of this region (Holcombe et al. 1997

figure 2). We will argue that this ~35 million year period records the westward migration of a continental magmatic arc during a period of crustal contraction, and subsequent transition to an extensional (and ultimately intraplate) setting. All of the terranes of the northern New England Fold Belt are affected by Permo-Triassic thrusts and folds, *Present address: Central Norseman Gold Corporation, Norseman WA 6443, Australia. tPresent address: Department of Geology, Victoria University of Wellington, Wellington, New Zealand. tPresent address: Etheridge, Henley & Williams, 457 Upper Edward Street, Spring Hill, Qld 4000, Australia. #Present address: 34—36 Whistler Court, Greenbank, Qld 4124, Australia.

P E R M O - T R I A S SIC H U N T E R - B O W E N E V E N T , Q L D although the greatest intensity of deformation occurs within the Gogango Overfolded Zone (Figure 1) and in the old accretionary terranes east of the Yarrol Fault. Folds associated with thrusts in the Gogango Overfolded Zone are characterised by pervasive slaty cleavage in contrast to the much more localised cleavage development north and south of this area. It is in the Gogango Overfolded Zone that the magnitude and style of this contractional event can be most clearly characterised, but Late Permian thrusts have now been mapped in the North D'Aguilar Block in southern Queensland (Little 1992; Sliwa 1994; Donchak et al 1995) although with eastward vergence.

53

the broad contractional cycle, while Late Triassic silicic caldera-related volcanics and granite plutons overprint the fold-thrust belt structures. The data presented here are derived from a number of research projects in southern and central coastal Queensland, as well as drawing on several Bowen Basin studies. In particular, the deformation history is based on ongoing study areas in the Fitzroy region around Rockhampton and Marlborough in coastal central Queensland (Figures 1, 2), and in southern Queensland (Figures 1, 3). Our understanding of the magmatic history is based on data gathered throughout the northern New England Fold Belt during our studies, and from the Connors Arch region (Dear 1989, 1994; Allen et al. 1994; C. M. Allen pers. comm. 1996).

FOLD-THRUST DEFORMATION Marlborough-Fitzroy area

Figure 1 Location map of major tectonic entities and study areas in the northern New England Fold Belt.

Explanations for Late Permian-Triassic sedimentation patterns in the Bowen Basin have evolved toward thrust-loaded foreland-basin models (Flood 1983; Murray 1990; Baker et al 1993), replacing earlier extensional or wrench-related models (Evans & Roberts 1980; Korsch et al 1988). We now regard the Bowen Basin as preserving in its sedimentological signature a large component of the post-Carboniferous evolutionary history of the adjacent New England Fold Belt. In particular, thick wedges of coarse clastic material along the eastern margin of the basin indicate episodic influx from the active fold belt and reflect the pulsed nature of the contractional event. The Late Permian to Late Triassic represents a major period of magmatism in the New England Fold Belt, although magmatic compositions appear to change through time and the igneous rocks of this interval can be separated into relatively discrete suites (Gust et al. 1993). Early Triassic volcanic and plutonic rocks with calc-alkaline geochemical characteristics occur within

Two fault styles dominate the Fitzroy region: Late Permian-Triassic thrusts and associated tear faults; and Cretaceous-Tertiary normal faults (Figure 2). Significant post-Triassic strike-slip movement has also occurred on some of the faults in the region (e.g. Broad Sound Fault). Many of the Cretaceous-Tertiary normal faults exploit the pre-existing thrust-tear fault architecture of the region to such an extent that almost all faults, of any age, have a Cretaceous-Tertiary brittle overprint. Many of the thrusts in Figure 2 are continuous with a normal fault that forms the boundary to an adjacent Cretaceous or Tertiary basin. Although there are undoubtedly elements of crustal fracturing in the region that controlled basin development during Early Permian extension (Hammond 1987; Fielding et al. 1998), we have not positively identified any regional fault patterns within the northern New England Fold Belt that can be associated directly with this event. Contractional structures developed during the PermoTriassic thrust event (thrusts, folds, cleavage) are pervasively developed for a distance of at least 100 km inland from the coast and continue with varying intensity into the Bowen Basin. Regional deformation related to this event is strongly heterogeneous. The most deformed terranes in the eastern part of the belt include remnants of the Devonian-Carboniferous accretionary terranes in the Coastal Block and parts of the Calliope terrane (Mt Holly beds). Further west, the Gogango Overfolded Zone describes an arcuate shape in plan (Figure 2) that includes a - 2 0 km-wide, west-verging thrust-fold belt along the trend of the Connors-Auburn Arch (Figure 1). Much of the intervening Yarrol Basin units and Calliope terrane are significantly less deformed, at least at outcrop scale. Several styles of contractional deformation are preserved across the region. In the Gogango Overfolded Zone, the most common style of deformation is multiple imbricate thrusts of small throw (few tens to few hundreds of metres) associated with well-developed mesoscopic folding and cleavage indicating a distributed ductile response. Broad areas of steep to

54

R. J. H O L C O M B E

ETAL. 22°S 50

23°S

Tertiary/Cretaceous fault basin KV Late Cretaceous volcanics plutons Permo-Triassic Bowen Basin units Connors/Camboon Volcanics /Berserker Beds (BB) foliated S-type granitoid (Broome Head Metamorphics - BHM)

accretionary terranes (including serpentinite) Calliope/Yarrol terranes major fault (SFZ - Stanage Fault Zone MT - Marlborough Thrust RT- Rookwood Thrust)

Figure 2 Structural framework of the Fitzroy region showing units and features discussed in other thrust/tear faults the text. Dotted lines labelled i, ii and iii are the cross-section normal fault (BF - Broad Sound Fault) lines AB, CD, and EF respectively of Figure 4.

overturned dips suggest that thrust-propagation folds are common. Fergusson et al. (1990, 1993) described similar pervasive folding and cleavage in the accretionary rocks of the Coastal Block. The intervening terranes are characterised by fewer thrust faults that have larger throw, accompanied by fault bend folds but little cleavage. The largest of these fault-bend folds is the Craigilee Anticline which appears to be a breached faultpropagation fold (Figure 4). The main floor thrust to this anticline (Rookwood Thrust, Figure 2) is a major shallow crustal discontinuity. Although time-equivalent Permian units are incorporated in both footwall and hangingwall, the thrust marks the eastern boundary of the well-cleaved rocks that define the Gogango Overfolded Zone. Rocks of the Yarrol and Calliope terranes (Figure 1) occur only within the hangingwall of the thrust, whereas rocks of the Permo-Carboniferous Connors-Auburn Arch occur only within the footwall. Early Permian marine units (e.g. Rookwood Volcanics, Berserker beds: Holcombe et al. 1998) apparently only occur to the east of the Rookwood Thrust.

The Marlborough Block is an enigmatic, thin (<2 km), composite terrane that was transported westward along a very low angle, out-of-sequence thrust (the Marlborough Thrust, Figure 2) over the earlier thrust packages (Figure 4). The basal thrust is a brittle structure with little ductile deformation, even within metres of the contact. In contrast, the Marlborough Block internally is an amalgam of numerous faultbounded packages of greenschist to amphibolite facies rocks with kilometre-scale, ductile response in the rocks adjacent to most faults. Some of these faults and associated shear zones have clear thrust geometry and kinematics whereas others are more ambiguous in either dip or sense-of-shear. These latter structures tend to occur near foliated S-type granitoids, and may be either thrusts that were rotated during translation of the block, or remnants of extensional faults that developed during emplacement of the granitoids, as have been observed in the North D'Aguilar Block in southern Queensland (Holcombe et al. 1998). The internal thrusts of the Marlborough Block are interpreted to have developed during thrusting in a deeper part of the thrust belt and

PERMO-TRIAS SIC HUNTER-BOWEN EVENT, QLD Gympie

j

AGYMPIE \BLOCKf .\V\N\V \ \ k\\ X\w%r \ \1- 1I \ \yV\sx-\ v% v\\ V\ T l\v - ^^ -\ A\ t/N \ .Os\N \ \ s

55

other faults NPF - North Pine Fault system BF - Bracalba Fault system Selected Permian & Triassic plutons SCA - Station Creek Adamellite MG - Monsildale Granodiorite Permian & Triassic units NB - Northbrook Block CGG - Cressbrook Creek Grp thrust faults CT - Claddagh Thrust system Early - Middle Permian marine fault basins CB - Cambroon Beds CV - Cederton Volcanics MB - Marumba Beds Late Carboniferous granitoids CG - Claddagh Granodiorite Devonian -Carboniferous accretionary terranes BB - Booloumba Beds JP - Jimna Phyllite

Figure3 Structural framework of the North D'Aguilar Block in southern Queensland showing units and features discussed in the text.

Figure 4 Cross-sections across the Gogango Overfolded Zone in the Fitzroy region. The Rookwood Thrust is a major thrust that carries Yarrol Block basement over Connors Arch basement in the footwall. Early Permian marine basins occur predominantly to the east (in the hangingwall) of this thrust. The Marlborough Thrust (section AB) is an out-of-sequence thin thrust nappe that overrides the earlier thrusts. The large antiformal structure in CD and EF is the Craigilee Anticline which is a breached propagation fold from a thrust that splays off the Rookwood Thrust. The steep faults are Cretaceous and Tertiary normal faults. The locations of section lines are shown on Figure 2. Both vertical and horizontal scale-bars are in 1 km increments.

-28°S

1

CLARENCE MORETON BASIN

| mid to Upper Permian Lower Permian - Rookwood Voles/ Youlambie - Carmiia

Marlborough block Carboniferous ~~ serpentinite/ | Yarrol Block/Connors voles metagranite ] Siluro-devonian

7T

3

56 R. J. H O L C O M B E ETAL. the package translated to its present structurally highlevel position along the younger Marlborough Thrust. In this model the Marlborough Thrust must represent an upper flat on a system that ramps down to the east. We interpret Ar/ Ar cooling ages of 248.8±0.5 Ma and 242.9±0.4 Ma on biotite from foliated granodiorite in the Marlborough Block (Holcombe et al. 1998) to reflect exhumation and cooling following the deeper level, more ductile, thrusting event. Deformation is also strongly partitioned along the fold belt, with tear faults separating compartments with fold-thrust packages of differing geometry (e.g. in the area shown in Figure 2). The northern termination of the Gogango Overfolded Zone fold-thrust belt is a tear fault system (Holcombe et al. 1995) equivalent to the Stanage Fault Zone of Henderson et al. (1993). The system is a complex zone of linked faults that also separates the allochthonous and para-allochthonous fold-thrust belt and Marlborough Block to the south from a gently folded, autochthonous terrane to the north. Fault styles developed within the zone include pure thrusts, strike-slip and oblique-slip vertical faults, and oblique-slip thrusts. The Stanage Fault Zone thus appears to have been a major tear fault system to both the nappe emplacement of the Marlborough Block, and to the in-sequence thrust belt in the footwall. The orientation of the overall thrust convergence vector appears to be southwest-directed, as indicated by the consistent orientation of tear faults. Strong strain partitioning in the fold-thrust belt makes both the construction of balanced cross-sections and precise estimation of the overall crustal shortening difficult. Fergusson (1991) estimated 60% shortening within the Gogango Overfolded Zone, an estimate consistent with the intensity of cleavage, and an overall upper crustal shortening of 50-90 km across the Gogango Overfolded Zone and Bowen Basin Folded Zone (Figure 2). By matching the easternmost exposures of rocks that we equate with Connors/Camboon Volcanics (Holcombe et al. 1998; Figure 1) north and south of the Stanage Fault Zone, we estimate thrust-tear fault contraction of >30 km along the Stanage Fault Zone, consistent with a 30 km translation proposed by Leitch et al. (1994). By correlating deformed syntectonic granite terranes in the Marlborough Block (Holcombe et al. 1997) with the Broome Head Metamorphics on the coast (Morand 1993) we estimate -50 km translation on the out-ofsequence Marlborough Block nappe system, in broad agreement with early estimates by Murray (1974). 40

39

North D'Aguilar Block The effects of the Permo-Triassic contractional deformation are subdued in the southeast Queensland section of the New England Fold Belt, relative to the Fitzroy region. The earliest recognisable thrust structure is the west-dipping Claddagh Thrust (Little 1992) in the northern North D'Aguilar Block. The thrust juxtaposes amphibolite facies rocks over low-grade (anchizonal) rocks within the older accretionary complex. The thrust can be traced for over 50 km to the south where it splits

into a system of imbricates (Sliwa 1994) that are responsible for the >5 km thickness of the Jimna Phyllite (Figure 3). Other east-verging thrust imbricates have now been mapped on the eastern margin of the North D'Aguilar Block (Donchak et al. 1995). The age of movement on the Claddagh Thrust and its imbricates is poorly constrained between Early Permian and Early Triassic. Both the ca 305 Ma Claddagh Granite (Little et al. 1995) and the ?Early Permian Marumba beds (Figure 3) are allochthonous within this thrust system. The upper limit for thrusting is the age of the Early and Middle Triassic Toogoolawah Group of the Esk Trough which unconformably overlies these early thrusts. Although poorly dated at this stage, a group of ca 241 Ma K/Ar whole rock and amphibole ages (240.9±11 Ma, hornblende; 241.9±8 Ma, 236±7, whole rock: Irwin 1973; Kerr 1974) from the Neara Volcanics is currently used to constrain the age of this group. The Monsildale Granodiorite is a multi-phased body that intrudes subvertical Marumba beds a few kilometres from where it is unconformably overlain by very gently dipping Toogoolawah Group rocks (Bryden Formation). Although it is not known whether the pluton stitches the thrusts at depth, K/Ar hornblende ages of 247.5±3 Ma (Sliwa 1994) and 240.1±3Ma (Kwiecien 1996) on the older phase of the granodiorite provide a minimum age on the thrusting and tilting, assuming that the granodiorite itself has not been involved in the tilting. Ar/ Ar ages on white mica from the Mt Mee area in the southern North D'Aguilar Block more tightly constrain the lower limit of thrust-related contraction in this area (Holcombe & Little 1994). Whereas the polymetamorphic rocks and syntectonic granitoids in the northern North D'Aguilar Block were exhumed through the blocking temperature for argon diffusion in these minerals (~350°C) during regional extension at about 296 Ma (Little et al. 1995), the structurally deeper epidote-blueschist rocks at Mt Mee remained below this blocking temperature until exhumed rapidly (-0.3 km/my) at ca 260 Ma (261.6±0.6Ma to 257.9±0.8 Ma: Holcombe & Little 1994). At Mt Mee, a regional upright antiform is associated with thirdgeneration axial plane fabrics which were initiated under greenschist facies conditions associated with the growth of coarse albite porphyroblasts. (Holcombe & Little 1994). These metamorphic fabrics, and associated mesoscopic folding, become very much more intense adjacent to the North Pine Fault which is the western boundary of this epidote-blueschist facies terrane. Although this fault is now defined by its post-Late Triassic movement, a precursor fault associated with contractional deformation clearly must have been active prior to the rocks passing through the ~350°C isotherm at ca 260 Ma. The rapid exhumation at ca 260 Ma is distinctly younger in age than the regional extension that initiated Bowen Basin sedimentation and is marginally younger than the initiation of thrust-loading sedimentation in the basin. Thus we would interpret the event creating the conditions for rapid exhumation at Mt Mee as being initiation of the regional thrust-fold contraction and the North Pine Fault as a likely early thrust. 40

39

PERMO-TRIAS SIC HUNTER-BOWEN EVENT, QLD Other ambiguous, open, upright folds (up to several kilometres wavelength) overprint earlier metamorphic fabrics throughout the North and South D'Aguilar Blocks (Holcombe 1977), and are overprinted by contact metamorphism adjacent to the ca 230-220 Ma suite of granitoids. In the central part of the North D'Aguilar Block one such set of large flexures fold the imbricate thrusts of the Jimna Phyllite (Figure 3; Sliwa 1994) and are likely related to the youngest contractional event described below. Esk Trough The Esk Trough is a narrow, north-northwest-trending belt of Early Triassic rocks, flanked by fault-bounded slivers of Permian rocks. The belt overlies and includes several regional lineaments in the northern New England Fold Belt and also potentially overlies the Early Carboniferous cratonic margin, separating accretionary complex rocks to the east from forearc basin rocks to the west. The structure of the belt is generally regarded as a graben, or half-graben, with its present margins approximating original rift margins. Seismic profiling across the southern extension of the belt clearly defines an asymmetric structure with steeply faulted eastern margin but no basal structures suggestive of an extensional origin (Korsch et al., 1989). The predominant rock types within the Esk Trough are a terrestrial sequence of andesitic, mainly volcaniclastic, rocks (Neara Volcanics), a package of interbedded clastic sedimentary rocks (Bryden Formation) that at least locally underlie these volcanic rocks, and a sandstone-dominated alluvial and lacustrine sequence (Esk Formation) overlying the volcanic strata. The present margins of the Esk Trough are sharply delineated faults that are characterised by Late Triassic or younger movement such that the original geometry of any trough is uncertain. In the central area of the Esk Trough we have observed local unconformable contacts with underlying sequences on both the eastern and western margin of the belt. At the western margin subvertical Esk Formation units overlie a similarly steeply dipping, slightly metamorphosed, pillow basaltic unit of unknown age. At the eastern margin of the belt in this area the shallowly dipping Bryden Formation is in strong angular unconformity with underlying ?Early Permian rocks, and further north gently tilted units of the Neara Volcanics onlap the basement rocks of the North D'Aguilar Block. Thus, the present western margin reflects strong post-depositional fold and fault structures (east-verging thrust) but the current eastern margin appears to be broadly depositional and only moderately modified by the late strike-slip faulting. Remnants of the Neara Volcanics that overlie the basement rocks are mainly volcanic-dominated, whereas thick beds of coarse volcaniclastic conglomerate ('boulder beds') characterise deposits within the axis of the belt. These data suggest that the Esk Trough was at least a depocentre, if not a fault-bounded basin, during the Early and Middle Triassic. Remnants of Permian marine sedimentary and volcanic sequences are preserved adjacent to the margins of the Esk Trough and

57

may indicate an earlier history to trough development. Ongoing studies indicate that the dominant palaeoflow direction in the Esk Formation was southward and westward. There is, to date, no sedimentological evidence that the present fault margins confined either the coarse conglomeratic facies or the overlying Esk Formation. Structures and stratigraphic relationships associated with the Esk Trough constrain elements of the Triassic part of the Hunter-Bowen event. The Neara Volcanics lie unconformably over east-verging thrusts along the western margin of the North D'Aguilar Block, providing a minimum age on the initiation of thrusting in this area. The sequences within the Esk Trough are also folded into > 1 km-wavelength, strongly asymmetric, east-verging, low- to moderate-amplitude folds and are unconformably overlain by flat-lying 228.4±0.6Ma volcanics of a later extensional magmatic phase (see below). This folding, constrained within the interval 241-228 Ma, appears to be the last Triassic contractional deformation in the area, and is interpreted as part of the final deformation associated with the Hunter—Bowen event. Fold axial traces generally lie parallel to the axis of the Esk Trough belt except near the fault margins where locally axial traces trend west to west-northwest in a sense that is consistent with a component of dextral wrenching parallel to the trough. West-trending axial traces also occur in fault slices of adjacent Late Permian fault blocks (Northbrook beds; Cressbrook Creek Group) that lie outside the belt. Deformation intensity in this folding event is low and the folds lack axial plane cleavage. Unlike that in the Fitzroy region, Permian and Triassic thrusting in southern Queensland appears to be thick-skinned and involve basement rocks, rather than thin-skinned. Except for the metamorphic rocks of the Mt Mee area and within the Gympie Block, cleavage is rarely associated with folds related to the PermoTriassic event in southern Queensland. Thrust vergence in southern Queensland is eastward, in contrast to the consistent westward vergence in central Queensland.

THRUST LOADING SIGNATURES IN BOWEN BASIN SEQUENCES Development of the Bowen Basin began with extensional sub-basins in the Early Permian (Phase 1 of Fielding et al. 1995) that was followed by a period of thermal sag in the latest Early Permian to early Late Permian (Phase 2). A major change in the petrology and depositional environment of basinal sediments in the Late Permian (Baker et al. 1993) was manifested in the introduction of first-cycle, volcanic lithic detritus shed from the east. The basin developed a marked crosssectional asymmetry typical of loaded foreland basins (Busby & Ingersoll 1995). Sediments shed westward across the Gogango Overfolded Zone joined major south-flowing, axial drainage systems (Fielding et al. 1995, 1998). Depositional environments were initially marine, but rapidly became coastal plain to alluvial plain systems as the basin was oversupplied with coarse sediment. This change is interpreted as a response to the

58 R. J. H O L C O M B E ETAL. onset of thrust loading of the eastern Bowen Basin, and accompanied the resurgence of volcanism to the east (Phase 3, see below). Carboniferous calc-alkaline volcanics in the Connors Arch (Figure 2) are widely interpreted as describing the position of a magmatic arc along the eastern margin of the Bowen Basin (Day et al 1978). Fergusson (1991), in contrast, showed the Connors Arch as a structural high produced during Permo-Triassic thrusting. The Connors Arch west of Marlborough is an open, antiformal structure with shallow (<30°) limb dips, complicated locally (particularly on the eastern limb) by steep faults. A complete Permian succession is preserved on both limbs of the antiform and around the southern limit of the Arch (Malone et al 1969). Continental and shallow-marine Permian sediments are transgressive across Connors Volcanics basement with only local angular discordance. West-directed palaeoflow directions occur in Phase 2 and Phase 3 sequences both east and west of the arch, and no basin-margin facies transitions are evident close to the arch. While Carboniferous volcanics along the arch may have been exposed during formation of the Early Permian extensional sub-basins, these had no topographic expression by the mid-Permian transgression. We suggest, on the basis of the characteristics of Triassic sediments within the Bowen Basin, that a topographic high did not form until the latest Middle Triassic as the thrust front migrated westward across the basin. The sedimentological transition from thermal sag to thrust loading is exposed in the Moah Creek beds along the Fitzroy River, west of Rockhampton (Fielding et al. 1997b figure 4) where a monotonous sequence of thinly interbedded marine siltstone and fine sandstone passes abruptly upward into disorganised conglomerate and diamictite enclosed within thinly bedded strata similar to the thick interval exposed below. Slump folds encased in a mud-breccia matrix are interpreted as the products of mainly debris and turbidity flow processes. The lower part of the exposed sequence is typical of the Phase 2 sag sequences and the first evidence of instability is a 5 m-thick, foundered horizon consisting of detached and transported masses of sandstone occurring within the siltstone package about 50 m below the main transition to conglomerate-rich units. The section is interpreted as having accumulated in an unstable submarine environment ahead of an approaching subsurface thrust front with the foundered horizon representing the earliest indication of the approaching front. Similar packages of coarse-grained mass-flow deposits have been recognised at this stratigraphic position over distances of at least 350 km along the eastern Bowen Basin margin (Fielding et al. 1997a). Within the eastern Bowen Basin, a number of distinct wedges of coarse clastic sediments occur in the Late Permian to Early Triassic succession (Figure 5; Kassan 1994; Fielding et al. 1995). These wedges comprise conglomerates and breccias of metasedimentary, volcanic and intrusive, and intraformational lithologies. Clast composition and palaeocurrent data indicate derivation from a highland to the east, that was associated with active volcanism. Successive wedges

appear to have penetrated further west into the basin (Figure 5). We interpret them as arising from pulses of thrusting in a rising hinterland to the east. In the case of the initial pulses recorded in the Late Permian sequences (such as the Moah Creek beds described above) this active volcanic hinterland was to the east of the current coastline. Elliott (1993) and Korsch and Totterdell (1995) have documented two discrete periods of thrust deformation within the Triassic Bowen Basin fill from seismic data. Each thrusting episode followed a period of westward propagation of a coarse clastic wedge, the later (late Middle Triassic) event terminating sediment accumulation across the entire basin. EARLY AND MIDDLE TRIASSIC CALCALKALINE SUITES

About half of the exposed granitoid ages in the northern New England Fold Belt have K/Ar ages between 250230 Ma (Gust et al 1993), within the later part of the broad Hunter-Bowen thrust event. The granitoids are widely distributed throughout the fold belt south of Broad Sound and east of the Gogango Overfolded Zone and are predominantly intermediate in composition. The plutonic rocks are coeval with terrestrial volcanism that is overwhelmingly andesitic in composition. Carnian

230-

Ladinian

^Anisian ,

240

CO CO

o 'Scythian 25a

"TJ CD

5 Tatarian'

3

S" =3

260-

Kazanian Ufimian

270-

Kungarian-

5> ZJ

Flat Top Fm Barfield Fm Mt Ox Subgroup

Figure 5 Time-stratigraphic distribution of coarse clastic sediment wedges shed into the Bowen Basin from the east, showing the extent and episodic nature of coarse siliciclastic sediment derived from the approaching thrust front to the east. Bar marked Au deposits indicates age range for K/Ar ages from structurally focused gold mineralisation in the northern New England Fold Belt. Small bars indicate specific K/Ar ages from illite in the Bowen Basin (Faraj et al. 1996) interpreted as fluid-flux event(s). Three periods of thrust deformation are shown; the earliest is constrained mainly by Ar/ Ar dating in southeast Queensland, whereas the second and third are recognised from seismic records across the Bowen Basin (Korsch & Totterdell 1995).

40

39

P E R M O - T R I A S SIC H U N T E R - B O W E N E V E N T , QLD 59 Volcanic rocks of this age are poorly represented canic structures was accompanied by development of north of the Stanage Fault Zone at Broad Sound, small to moderate-sized north-northwest-elongate although Triassic radiometric ages are recorded from basins with coarse-grained, grossly fining-upward, intrusives and minor volcanic rocks in the Whitsunday alluvial fills, thick localised coal bodies and evidence of region (Ewart et al 1992; Parianos 1993; Allen et al; synchronous bimodal volcanism. Basins of this nature C. M. Allen pers. comm. 1996) and from mafic dykes include the Ipswich (Falkner 1986; Staines et al 1995) and minor granitoid intrusions within the Urannah and Tarong Basins (Pegrem 1995) and part of the Complex (Allen et al 1994; C. M. Allen pers. comm. Callide Coal Measures (Biggs et al 1995). The Tarong 1996). Andesitic volcanic and volcaniclastic sequences Basin has a half-graben geometry (Williams 1993) as do south of Broad Sound unconformably overlie basement a number of less well-described coeval basins in the subsurface (Wiltshire 1982). Although the crossof Devonian-Carboniferous to Late Permian age. sectional geometry of the other major basins is not well constrained, our work (CRF unpubl. data) suggests that LATE TRIASSIC EXTENSIONAL BASINS AND these basins are probably similar in geometry. Bimodal MAGMATIC SUITES volcanism is most evident within the Ipswich Basin where basalt flows and a significant felsic pyroclastic High-level granites and volcanics unit (Brisbane/Hector Tuffs) lie at, or near, the base of the sequence. All three of the basins studied thus far In contrast to the intermediate-dominated composition of Early and Middle Triassic granitoids of southern contain abundant thin felsic tuffaceous horizons within coal-dominated sequences. Queensland, the Late Triassic (ca 230-220 Ma) is Models proposed for the origin of these basins are characterised by intrusions of predominantly silicic granite composition associated with the development of diverse (Murphy et al 1976; Falkner 1986; Korsch et volcanic complexes of rhyolite and minor mafic lava al 1989; Pegrem 1995) but on the basis of the and ignimbrite (Stephens et al 1993). One large-scale asymmetric cross-sectional basin geometry and regional caldera associated with ignimbrite development has association with extensive coeval bimodal volcanic been identified within these units (Stephens 1992) and sequences, we regard the basins as characteristic of an the restricted distribution of many of these sequences extensional tectonic environment. suggests that other calderas may be present. Ignimbrites correlated with this event are unconformable on Early to YOUNGER BASINS AND STRUCTURES Middle Triassic andesite in the northern Esk Trough. In southern Queensland flat-lying andesite giving an Ar/ Ar cooling age of 228.4±0.6 Ma on amphibole The post-Triassic evolution of the region is marked by unconformably overlies folded Early to Middle Triassic development of extensive, Jurassic-Cretaceous basins rocks (Little et al 1993). Elsewhere, hornblende in the (e.g. Surat Basin, Clarence-Moreton Basin, Marysame sequence yielded a concordant Ar/ Ar total borough Basin) across the Late Triassic extensional fusion age of 232±4 Ma (C. G. Murray pers. comm. basins. One of these basins that impacts on the interp1992) and ca 225 Ma K/Ar ages have been derived retation of structures seen within the fold belt is the from similar andesitic and rhyolitic flows and dykes that Maryborough Basin which, unlike others of its age, is overlie basement rocks in the North D'Aguilar Block gently to moderately folded. This deformation, which (Holcombe et al 1997 table 1; Roberts 1992; Sliwa is constrained stratigraphically as mid-Cretaceous or younger, is the only recognised post-Triassic con1994). The Station Creek Adamellite, which intrudes rocks tractional episode in the region. It is most likely the of the North D'Aguilar Block, provides important con- event in which much of the late regional faulting within straints on the timing of events in the area as it also the northern New England Fold Belt formed. Effects of intrudes the thrust sheets within the North D'Aguilar this deformation include small-displacement reverse Block and may be comagmatic with the overlying vol- faults that cut the Mesozoic basin rocks throughout canics. We have recently obtained Ar/ Ar step southern Queensland, and presumed coeval faults of heating plateau ages of ca 236 Ma and 232.8±0.4 Ma similar geometry and kinematics within the fold belt. Major faults, commonly with -10 km sinistral strikeon biotite from this body, although the data have not yet slip, broadly define the structural grain of southern been completely interpreted. Queensland. Almost all of the older terranes are These Late Triassic volcanics and plutons are largely undeformed, although the Station Creek Adamellite on bounded by these younger faults which locally displace the northeastern flank of the North D'Aguilar Block is plutons of the ca 230-220 Ma magmatic suite. Exlocally faulted and ductilely sheared (Edgar 1992; J. amples of these faults exploiting the older fault Tang pers. comm. 1996), and there is local brittle- architecture occur in the North D'Aguilar Block, where ductile deformation in the North Arm Volcanics, to the Early Permian synmetamorphic deformational structures intensify toward the North Pine Fault. This fault, east of the North D'Aguilar Block. however, appears to have sinistrally offset several Late Triassic plutons and a Late Triassic volcanic formation by ~8 km. The North Pine Fault is continuous with the Extensional basins composite Perry Lineament to the north, where a Late The change in character of magmatism to silicic vol- Triassic {ca 221 Ma) volcanic complex shows a sinistral 40

39

40

39

40

39

60 R.J. H O L C O M B E E T A L . strike displacement of about 9 km (Stephens 1991). Similar sinistral strike-slip faults occur in the central Queensland part of the northern New England Fold Belt [e.g. the Broad Sound Fault with -20 km strike separation (Figure 2); offshore in the Whitsunday region (Ewart et al 1992)] but fault patterns in the northern region are dominated by steep normal faults that bound the numerous Cretaceous and Tertiary Basins (e.g. the Cretaceous Styx Basin and the Tertiary Duaringa Basins). DISCUSSION The major contractional period defined here as the Hunter-Bowen event lasted for about 35 million years, from ca 265 to ca 230 Ma. Stratigraphic evidence from the foreland basin fill suggests that this event was strongly pulsed and that successive thrust pulses penetrated further into the basin. The final contractional event appears to have re-initiated at the eastern margin of the fold belt, rather than step westward from the previous thrust front, and to have been more intense than previous pulses. The presence of a broadly synchronous magmatic event within the northern New England Fold Belt suggests that arc magmatism was superimposed on the actively rising mountain belt. In the Fitzroy region the commencement of thrust contraction is constrained as Late Permian (ca 265 Ma). The oldest sedimentary unit in this region that was derived from the approaching thrust front (and subsequently involved in the thrusting) lies within the Late Permian Moah Creek beds, Barfield Formation and the equivalent Boomer Formation. The maximum age on thrusting is thus constrained to Kazanian on biostratigraphic evidence (Fielding et al 1997b). Episodic deformation in the Fitzroy area is indicated by the out-ofsequence thin-skinned emplacement of the Marlborough thrust nappe. This thrust overrides earlier thrusts that involve latest Permian rocks (Dinner Creek Conglomerate), and the only other constraint on the timing of this thrust are 248.8±0.9 Ma and 242.9+0.4 Ma Ar/ Ar cooling ages on biotite in sheared and foliated metagranites that we surmise were related to a deeper seated, earlier, thrust environment. The presence and magnitude of the nappe indicates a significant renewal of a contractional deformation from the east after this time. The initiation of thrusting is less constrained in southern Queensland, but there is a clear indication of two phases of contractional deformation separated by an interval of calc-alkaline magmatism. On the western margin of the North D'Aguilar Block, thrusts that carry the Late Carboniferous Claddagh Granite and the ?Early Permian Marumba beds are unconformably overlain by the Early to Middle Triassic (ca 241 Ma) volcanic sequence of the Esk Trough. The youngest age for these thrusts is thus ca 241 Ma, but the oldest age is poorly constrained. Nonetheless, the white mica Ar/ Ar ages from the Mt Mee area indicates rapid exhumation of the metamorphic basement rocks at ca 260 Ma, an event that may relate to the commencement of HunterBowen contractional deformation in this area and an age 40

40

39

39

that is consistent with initiation of this event elsewhere. Termination of the thrust and folding events in southern Queensland is constrained to the interval 241— 228 Ma. The folded Early and Middle Triassic volcanic succession within the Esk Trough is unconformably overlain by ca 228 Ma flat-lying intermediate volcanics but related rocks in the region have a range of K/Ar ages from ca 235 to 225 Ma. (Holcombe et al 1997 table 1). The new ca 235 Ma Ar/ Ar ages we have obtained on the Station Creek Adamellite (see above), might provide an even tighter constraint on the age of the terminal Hunter-Bowen folding in this area. Analysis of radiometric dates (Gust et al 1993) and local detailed mapping (Stephens 1992) clearly distinguishes the presence of Early and Late Triassic volcanic rocks of contrasting composition and style. All of these Triassic volcanics were grouped during the initial 1:250 000-scale mapping of the northern New England Fold Belt, and only recently has the presence of two compositionally distinct events been reflected in the stratigraphic nomenclature (Cranfield 1994). Early and Middle Triassic magmatism has not been systematically studied at this time, but data on granitoids and volcanics (unpublished theses at University of Queensland and Queensland University of Technology) and limited isotopic data from volcanics in the Esk Trough (Ewart et al 1992) show not only the calc-alkaline character of this event but the overwhelmingly intermediate composition of the rocks. These data are consistent with a period of continental margin arc-related magmatism during the Early and Middle Triassic. Such an interpretation is supported by the observation that the Late Permian-Early Triassic sediments derived from the approaching thrust front to the east contain first-cycle volcanic detritus. Holcombe et al (1990) and Fielding et al (1997) emphasise that there is no evidence in the Early to mid-Permian rocks of the northern New England Fold Belt for the presence of an arc-related volcanic terrane. In tectonic terms, the Permo-Triassic magmatism thus requires the initiation of subduction below the region, or the migration of the arc onshore from a position somewhere to the to east, during the Hunter-Bowen contraction event. Stephens (1992) and Stephens et al (1993) interpreted Late Triassic silicic volcanism in terms of an extensional tectonic environment. Criteria cited included the discrete, caldera-forming nature of the volcanism, characteristic of continental extensional environments, the bimodal, silicic-dominated composition of the volcanics, and the regional silicic granite-dominated composition of coeval intrusives. These data suggest that the relatively rapid re-establishment of voluminous arc magmatism within the New England Fold Belt during the Permo-Triassic, clearly associated with a broader cycle of tectonic contraction, was replaced during the Late Triassic by an extensional environment and crustal melting of the recently arc-impregnated crust. The latest Triassic is further characterised by localised, discrete caldera development and emplacement of granite with mild A-type geochemical affinities (Stephens 1992; Gust et al 1993), supporting the concept that the region underwent a transition from 40

39

PERMO-TRIASSIC HUNTER-BOWEN EVENT, QLD

convergence to extension that continued into the latest Triassic. The position and nature of any arc that operated after the Middle Triassic, or of any Permo-Triassic subduction complex, is uncertain. However, a possible mechanism for the transition from presumed subduction to extension during this time may lie in one of our suggested interpretations of the Late CarboniferousEarly Permian evolution of the northern New England Fold Belt (Holcombe et al. 1997). We suggest one possible scenario is that the subducting slab again underwent roll-back during the late Middle Triassic driving extension and resulting in re-establishment of the volcanic arc some distance to the present east of the New England Fold Belt (and the present coastline). Voluminous first-cycle volcaniclastic debris and numerous tuffs within the Surat Basin (Exon 1976) suggest that volcanism sourced from an unidentified terrane continued through the Jurassic leading up to the major Early Cretaceous breakup-related magmatism (Ewart etal. 1992). Metallogenic aspects Two major styles of mineralisation are associated with the broad Hunter-Bowen event in the northern New England Fold Belt. Porphyry-style mineralisation is commonly developed in association with the PermoTriassic calc-alkaline intrusives of the fold belt (Horton 1978). Of somewhat more enigmatic origin is the occurrence of epithermal gold mineralisation, associated with quartz-rich alteration systems, and locally with carbonate veining, that consistently gives ca 245235 Ma K/Ar alteration ages. Examples include the major deposits of Cracow and Gympie, plus the smaller deposits at Manumbar in the North D'Aguilar Block, Mt Mackenzie in the southern Connors Arch, and perhaps at Mt Wickham in the northern Connors Arch. We also include mineralisation at Rannes, in the Gogango Overfolded Zone, in this association on the basis of alteration style and structural association. In some instances, such as at Gympie and Rannes, mineralisation is within thrust-related sheared or cleaved rocks. In others, such as Cracow, Mt Mackenzie and Mt Wickham, mineralisation is more typical high level, mesothermal to epithermal in nature. A line of significant gold deposits occurs along the eastern margin of the Bowen Basin, including Cracow, Rannes, Mt Mackenzie and Mt Wickham. Of these occurrences, only Rannes occurs within strongly thrustdeformed rocks. At Rannes, well-developed silicic alteration systems associated with gold and minor basemetal mineralisation occur in locally sheared Camboon Volcanics. Mineralisation appears to occur both on thrusts and in zones that cut across the thrust trend at a high angle. The other deposits comprise more classical alteration systems of similar high-T, low-P grade, but also overprint the Late Carboniferous-Early Permian volcanic succession. Further east at Gympie, strain is strongly partitioned in the volcaniclastic sandstones of the Rammutt

61

Formation. Where cleavage is developed, it is a strong pressure-solution fabric and is accompanied by a marked stretching lineation defined both by the shape of pressure-solved clasts and, more particularly, by mica beards and fringes developed on clasts. A characteristic feature of these rocks is the development of a network of fine extensional veins perpendicular to the stretching lineation, and infilled with fibrous quartz (and minor carbonate) that are parallel to this lineation. Similar veins sets occur perpendicular to the stretching lineation in the gold-bearing black slates and are known locally as the 'Gympie vein set'. These veins (and the associated mineralisation) are thus syntectonic with the cleavage-forming deformation, and alteration associated with these veins has given a K/Ar age of ca 235 Ma (Cuneen 1994). Gold mineralisation at Manumbar in the North D'Aguilar Block occurs in carbonate-quartz veins within rocks equated to the Early Triassic Neara Volcanics. Mining is currently occurring in a single, major vein, but numerous en echelon swarms of fibrous extensional veins occur in the field. There is no other obvious deformation apart from the brittle-ductile deformation associated with the vein swarm, and alteration associated with mineralisation has yielded a ca 235 Ma age (M. Garmanpers. comm. 1995). In all cases, the deposits are localised in volcanic or volcaniclastic rocks, and in areas that are characterised by late (i.e. post-Permian) Hunter-Bowen structures. The line of deposits along the eastern margin of the Bowen Basin occur along the western limb of the structural arch that defines the basin margin and which we regard as forming during the Middle Triassic. The timing of mineralisation clearly just precedes the late, major contractional pulse that closed the basin, and is broadly coeval with K/Ar ages on mineralogically pure cleat-filling illite in the Late Permian coal measures (Figure 5). We believe that this late pulse of the HunterBowen event not only was responsible for the development of the structural Connors-Auburn Arch, but also promulgated a major fluid flux within structures deforming the northern New England Fold Belt and through sediments and structures within the eastern Bowen Basin (Faraj et al. 1996). The meteoric composition of the mineralising fluids at Cracow (Golding et al. 1987) is interpreted to reflect the nature of fluids generated during this event.

Possible allochthoneity of the Yarrol/Calliope terranes The thrust geometries of the Gogango Overfolded Zone shown in the cross-sections of Figure 5 are regarded as a reasonable extrapolation of the available surface data. Strong strain partitioning and disruption by Cretaceous and Tertiary faults, however, makes confident interpretations of these sections to depth difficult. One source of variation in interpretation based on extrapolation of these sections, however, is that placed on the geometry of the Berserker Block. We have noted that basement to the footwall rocks of the Rookwood Thrust is

62

R. J. H O L C O M B E

ETAL.

consistently Connors Volcanics or equivalents, whereas basement to the hangingwall rocks is consistently rocks of the Yarrol/Calliope terranes. If the Berserker Block is stratigraphically equivalent to the Connors Volcanics (,sensu lato), as suggested by Holcombe et al (1997), then the simplest geometry that satisfies this structural and stratigraphic interpretation is that the Berserker Block (Figure 2) is a window through the Rookwood Thrust system. The implication of this interpretation is that the hangingwall, with its Siluro-Devonian basement, is very thin skinned and must have a displacement of several tens of kilometres. The ramifications of such a model are that: (i) both the Calliope and Yarrol terranes may be very thin (<2 km); (ii) since both the hangingwall and footwall of the Rookwood Thrust contain Bowen Basin sequences, the thrust displacement of the allochthonous terranes would not be expected to greatly exceed a few tens of kilometres; and (iii) the Tungamull and Parkhurst Faults are part of a single imbricate thrust system (albeit reactivated during Cretaceous-Tertiary normal faulting) carrying allochthonous terranes that include both the accretionary rocks and elements of the Yarrol terrane. While these interpretations are highly speculative, they do bear upon matters such as the position of the continental margin and accretionary complex during the Early Carboniferous and older convergent tectonic events. Gympie Block: how allochthonous is it? We see a paradox in the histories of different age packages of cleaved rocks of the North D'Aguilar Block. The cleaved Early Permian Cambroon beds on the eastern margin of the North D'Aguilar Block are in fault contact to the west with the ?Early Carboniferous Booloumba beds of the accretionary terrane (Figure 3) and structural relationships suggest considerable translation on this fault (tentatively correlated with an early phase of movement on the Bracalba Fault: Sliwa 1994). The folded chert-argillite sequences of the Booloumba beds contain a single upright cleavage characteristic of the anchizonal (upper plate) accretionary rocks that we interpret as being developed during the midCarboniferous accretion. There are, however, no overprinting cleavage fabrics in the older rocks that correspond to the cleavage present in the adjacent Early Permian rocks. That the cleavage-forming event in the Early Permian rocks was not sufficiently intense to be transmitted into the basement rocks is considered unlikely, given the polydeformational fabrics in the Permian rocks at some localities. More likely, the fault separating the two units has considerable displacement and was active after cleavage formation in the Early Permian rocks, but before the boundary was intruded by Late Triassic (ca 220 Ma) granite. The amount of any such displacement is unknown but must be sufficient to have juxtaposed rocks of entirely different deformational responses to the same event. Terranes east of the Cambroon beds that may have been included in such a displacement include remnants

of the accretionary rocks and the Early Permian—Early Triassic units of the Gympie Block. The concentration of post-Early Permian brittle and ductile deformation along the eastern margin of the North D'Aguilar Block, the probability of significant translation on the Bracalba(?) Fault, and the presence of Middle Triassic cleavage-forming deformation in the Gympie Block, suggests that the units of the Gympie Block also have been translated, to some degree, into its present location. Holcombe et al (1997) noted the similarity of the Early Permian sediments and volcanics in the Gympie Block with other extensional marine basin sediments in the most eastern parts of the northern New England Fold Belt. In contrast with the much less cleaved rocks in the adjoining blocks in southern Queensland, the style of deformation in the Gympie Block with its widespread cleavage, and variable cleavage intensity and orientations typical of thrust terranes, is similar to that in the rocks of the fold-thrust belt that we have studied in the Fitzroy area to the north. We suggest that the Gympie Block may be an element of a more northerly terrane of the northern New England Fold Belt that has been displaced south by dextral strike-slip motion during the later part of the Hunter-Bowen event. We would speculate that it initiated as one of the suite of Early Permian marine extensional basins that formed within the old accretionary terranes along the New England Fold Belt, thus accounting for its present location to the east of the southern accretionary exposures. The well-cleaved ?Early Triassic Kin Kin Phyllite is the youngest rock unit in the Gympie Block and thus any major strike-slip displacement must post-date that time and yet be completed by the end of Hunter-Bowen contraction at ca 230 Ma. Typical strike-slip fault displacement rates on major faults in California are within the range of 1-10 mm/y, increasing to 25—35 mm/y for the San Andreas plate margin system (Petersen & Wesnousky 1994). Hence a moderately fast movement rate of 10 mm/y would produce 100 km of displacement over 10 million years. We have noted that the ultimate Hunter-Bowen contractional event was more intense than previous pulses, and that it produced outof-sequence thrust nappe structures at the eastern margin of the fold belt. It is this event that would be the most likely driving force for any displacement of the Gympie Block, although movement rates would have to be 10-20 mm/y. The New England Fold Belt 'double orocline' and dextral wrenching A major factor in the consideration of possible displaced terranes in the northern New England Fold Belt has been the problem of explaining the major double oroclinal flexure in northern New South Wales and southern Queensland. Murray et al. (1987) developed a model for the formation of the oroclinal flexure invoking large scale dextral displacement of terranes in the eastern New England Fold Belt on a transform fault during the Late Carboniferous. This model, and

PERMO-TRIAS SIC HUNTER-BOWEN EVENT, QLD subsequent variations (Fergusson et al. 1993) postulate a large displacement (-500 km) dextral strike-slip fault in the northern New England Fold Belt that accommodates the oroclinal bending to the south. A major problem with this model has been the lack of documented dextral strike-slip structures of that age in the northern New England Fold Belt, although any such structure could well be masked by the later contractional deformation. We would suggest that the most likely deformation event with the required geometry to produce the dextral oroclinal flexure would be during the Hunter-Bowen event. In the northern New England Fold Belt, the original meridional structural grain that was imparted by the Early Carboniferous accretionary events was overprinted by a north-northwest-trending grain transverse to west-southwest-verging thrusts during the Hunter-Bowen event. A west-southwest contractional vector would provide an ideal structural environment for dextral slip on the pre-existing structural grain. ACKNOWLEDGMENTS This paper incorporates the results of several independent ARC-funded projects (A38830041; A39130279; A39331366; A39232338, A3931187) over the past ten years. We particularly acknowledge the financial and logistical support of Queensland Metals Corporation (and in particular Ian Howard-Smith, David Horton and Darcy Milburn) without which the unifying work in the Fitzroy region would never have evolved. We are indebted to the many students who have contributed as part of the various projects. In particular, aspects of the PhD work of Joe Tang, Terry Harbort, Lorraine Campbell and Basim Faraj, the MSc work of Bill Kwiecien and Jim Hanson, and the Honours work of Terry Harbort, Vanessa Muscio, Murray Patterson, Bernadette Williams, Eris O'Brien, Sean Joyce, Craig Roberts, Jason Moultrie and Steve Downey have added to our understanding of the region. We must also acknowledge the discussions, sometimes spirited, over the years with Cec Murray, Russell Korsch, Chris Fergusson, John Draper and others which have refined our ideas. In a review paper like this it is likely that we have missed proper acknowledgment of some sources. Constructive reviews by Vince Morand, Peter Cawood and Cec Murray have considerably strengthened the paper.

63

complex back-arc extensional to foreland basin: the Bo wen Basin, eastern Australia. Journal of Sedimentary Petrology 63, 881-893. BIGGS M . , BURGESS A . W . & PATRICK R . B . 1 9 9 5 . C a l l i d e

Basin. In: Ward C. R., Harrington H. J., Mallett C. W. & Beeston J. W. eds. Geology of Australian Coal Basins, pp. 471-^488. Geological Society of Australia Coal Geology Group Special Publication 1. BRYAN W. H. 1925. Earth movements in Queensland. Proceedings of the Royal Society of Queensland 37, 3-82. BUSBY C. & INGERSOLL R. (Editors) 1995. Tectonics

D A Y R . W . , MURRAY C . G . & W H I T A K E R W . G . 1 9 7 8 . T h e

eastern part of the Tasman Orogenic Zone. In: Scheibner E. ed. The Phanerozoic structure of Australia and variations in tectonic style, pp. 327-364. Tectonophysics 48. D O N C H A K P . J. T . , LITTLE T . A . , SLIWA R . & HOLCOMBE R . J.

1995. Geology of metamorphic units of the North D'Aguilar Block—Goomeri, Nannago and Nambour 1:100 000 sheet areas. Queensland Geological Record 1995/7. EDGAR S. A. 1992. Geology and geochemistry of the Woolooga area, southeast Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). ELLIOTT L. G. 1993. Post-Carboniferous tectonic evolution of eastern Australia. The APEA Journal 33, 215—236. EWART A . , SCHON R . W . & C H A P P E L L B . W .

1992.

ALLEN C . M . , CHAPPELL B . W . , WOODEN J. L . & WILLIAMS I.

S. 1994. Ages, compositions and sources of the Urannah batholith. In: Henderson R. A. & Davis B. K. eds. Extended conference abstracts : new developments in geology and metallogeny: Northern Tasman orogenic zone, pp. 107—108. James Cook University, Townsville, Economic Geology Research Unit Contribution 50. BAKER J. C . , FIELDING C . R . , DE CARITAT P. & WILKINSON M .

M. 1993. Permian evolution of sandstone composition in a

The

Cretaceous volcanic- plutonic province of the central Queensland (Australia) coast—a rift-related 'calc-alkaline' province. Transactions of the Royal Society of Edinburgh 83, 3 2 7 - 3 4 6 .

EXON N. F. 1976. Geology of the Surat Basin in Queensland. Bureau of Mineral Resources Bulletin 166. FALKNER A. J. 1986. Sedimentology of the Blackstone Formation, Ipswich Coal Measures, southeast Queensland. PhD thesis, University of Queensland, Brisbane (unpubl.). FARAJ B . S . , FIELDING C . R . & MACKINNON I. D . R . 1 9 9 6 .

Cleat mineralisation of the Upper Permian Baralaba/ Rangal Coal Measures, Bowen Basin, Australia. In: Gayer R. & Harris I. eds. Coal bed Methane and Coal Geology, pp. 151-164. Geological Society of London Special Publication 97. FERGUSSON C. L. 1991. Thin-skinned thrusting in the northern New England Orogen, central Queensland, Australia. Tectonics 10, 797-806. FERGUSSON C . L . , HENDERSON R . A . & LEITCH E . C .

REFERENCES

of

Sedimentary Basins. Blackwell, Oxford. CRANFIELD L. C. 1994. 1:250 000 Geology Series Explanatory Notes, Maryborough, Queensland: SG56-6. Queensland Department of Minerals and Energy, Brisbane. CUNEEN R. 1994. New Insights into the Gympie Goldfield. In: Queensland Exploration Potential 1994, Symposium Abstracts, p. 26. Queensland Department of Minerals and Energy, Brisbane.

1990.

Structural history and tectonics of the Palaeozoic Shoalwater and Wandilla terranes, northern New England Orogen, Queensland. Australian Journal of Earth Sciences 37, 387-400. FERGUSSON C . L . , HENDERSON R . A . & LEITCH E . C .

1994.

Tectonics of the New England Fold Belt in the Rockhampton Gladstone region, central Queensland. In: Holcombe R. J., Stephens C. J. & Fielding C. R. eds. 1994 Field Conference Guidebook: Capricorn Region, central coastal Queensland, pp. 1-16. Geological Society of Australia, Queensland Division, Brisbane.

64

R. J. HOLCOMBE

ETAL.

FERGUSSON C . L., HENDERSON R. A . , LEITCH E. C. & ISHIGA

H., 1993. Lithology and structure of the Wandilla terrane, Gladstone-Yeppoon district, central Queensland, and an overview of the Palaeozoic subduction complex of the New England Fold Belt. Australian Journal of Earth Sciences 40: 403-414. FIELDING C . R., STEPHENS C . J. & HOLCOMBE R. J. 1997a.

Submarine mass-wasting deposits as an indicator of the onset of foreland thrust loading: Late Permian Bowen Basin, Queensland, Australia. Terra Nova 9, 14-18. FIELDING C . R., STEPHENS C. J. & HOLCOMBE R. J. 1 9 9 7 b .

Permian stratigraphy and palaeogeography of the eastern Bowen Basin, Gogango Overfolded Zone and Strathmuir Synclinorium in the Rockhampton-Mackay region, central Queensland. Geological Society of Australia Special Publication 19, 80-95. FIELDING C . R . , STEPHENS C. J., KASSAN J. & HOLCOMBE R. J.

1995. Revised palaeogeographic maps for the Bowen Basin, Central Queensland. In: Follington I. L., Beeston J. W. & Hamilton L. H. eds. Bowen Basin Symposium 1995, Proceedings, pp. 7-15. Geological Society of Australia, Queensland Division, Brisbane. FLOOD P. G. 1983. Tectonic setting and development of the Bowen Basin. In: Waterhouse J. B. ed. 1983 Field Conference — Permian areas, Biloela, Moura, Cracow, pp. 7-21. Geological Society of Australia, Queensland Division, Brisbane. GOLDING S. D., WILSON A . F., SCOTT M . ETAL. 1987. Isotopic

evidence for the diverse origins of gold mineralisation in Queensland. In: Herbert H. K. ed. Gold in Queensland, pp. 65—83. Papers of the Department of Geology, University of Queensland 12. GUST D . A . , STEPHENS C . J. & GRENFELL A . T .

1993.

Granitoids of the northern NEO: their distribution in time and space and their tectonic implications. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, Eastern Australia, NEO '93 Conference Proceedings, pp. 565— 572. University of New England, Armidale. HAMMOND R. 1987. The Bowen Basin, Queensland, Australia: an upper crustal extension model for its early history. Bureau of Mineral Resources Record 1987/51, 131-142. HENDERSON R . A., FERGUSSON C. L., LEITCH E. C., MORAND V. J., REINHARDT J. J. & CARR P. F. 1993. T e c t o n i c s o f the

northern New England Fold Belt. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, Eastern Australia, NEO '93 Conference Proceedings, pp. 505— 515. Department of Geology and Geophysics, University of New England, Armidale. HOLCOMBE R. J. 1977. Structure and tectonic history of the Brisbane Metamorphics in the Brisbane area. Journal of the Geological Society of Australia 24,475-490. HOLCOMBE R. J. & LITTLE T . A . 1994. B l u e s c h i s t s o f t h e

North D'Aguilar Block: structural development of the Rocksberg Greenstone and associated units near Mt Mee, southeast Queensland. Australian Journal of Earth Sciences 41, 115-130. HOLCOMBE R. J., FIELDING C. R. & STEPHENS C. J. 1995. T e a r

fault termination of the fold-thrust belt in the Northern New England Orogen. Geological Society of Australia Abstracts 40, 72. HOLCOMBE R . J., STEPHENS C . J., FIELDING C. R. ETAL.

1997.

Tectonic Evolution of the Northern New England Fold Belt: Carboniferous to Early Permian transition from acccretion to extension. Geological Society of Australia Special Publication 19, 66-79.

HORTON D. J. 1978. Porphyry-type copper—molybdenum mineralization belts in Eastern Queensland. Economic Geology 73, 904-921.

IRWIN M. J. 1973. The igneous and sedimentary geology of the Kinbombi area, southeast Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). JONES P. J. (Compiler). 1996. AGSO Phanerozoic Timescale 1995. Oxford University Press, Melbourne. KASSAN J. 1993. Basin analysis of the Triassic section in the Bowen Basin, Queensland. PhD thesis, University of Queensland, Brisbane (unpubl.). KERR I. D. 1976. Aspects of the geology and geochemistry of the Neara Volcanics and the Kilkivan Mercury Deposits, southeast Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). KORSCH R . J., HARRINGTON H. J., WAKE-DYSTER K . D . , O'BRIEN P. E. & FINLAYSON D . M . 1 9 8 8 . S e d i m e n t a r y

basins peripheral to the New England Orogen: their contribution to understanding New England tectonics. In: Kleeman J. D. ed. New England Orogen, Tectonics and Metallogenesis, pp. 134-140. University of New England, Armidale. KORSCH R. J., O'BRIEN P. E., SEXTON M . J., WAKE-DYSTER K.

D. & WELLS A. T. 1989. Development of Mesozoic transtensional basins in easternmost Australia. Australian Journal of Earth Sciences 36, 13—28. KORSCH R. J. & TOTTERDELL J. M. 1995. Structural events

and deformational styles in the Bowen Basin. In: Follington I. L., Beeston J. W. & Hamilton L. H. eds. Bowen Basin Symposium 1995 Proceedings, pp. 27-35. Geological Society of Australia, Queensland Division, Brisbane. KWIECIEN W. 1996. Geology, geochemistry, and petrogenesis of the Monsildale Granodiorite. MSc thesis, Queensland University of Technology, Brisbane (unpubl.). LEITCH E. C. 1975. Plate tectonic interpretation of the Palaeozoic history of the New England Fold Belt. Geological Society of America Bulletin 86, 141—144. LEITCH E . C . , FERGUSSON C. L. & HENDERSON R . A . 1994.

Ophiolitic and metamorphic rocks in the Percy Isles and the Shoalwater Bay region, New England Fold Belt, central Queensland. Australian Journal of Earth Sciences 41,571-579. LITTLE T. A. 1992. Geology of a northern part of the North D'Aguilar block, southeast Queensland. Queensland Department of Resource Industries Record 1992/22. LITTLE T . A . , HOLCOMBE R. J. & SLIWA R. 1993. S t r u c t u r a l

evidence for extensional exhumation of blueschist-bearing serpentinite—matrix melange, New England Orogen, southeast Queensland, Australia. Tectonics 12, 536-549. LITTLE T . A., MCWILLIAMS M . O . & HOLCOMBE R. J. 1995. 40

Ar/ 39 Ar thermochronology of epidote-blueschists from the North D'Aguilar Block, Queensland, Australia: timing and kinematics of subduction complex unroofing. Geological Society of America Bulletin 107, 520-535 MORAND V. J. 1993. The Broome Head Metamorphics: high grade metamorphism in the northern New England Fold Orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, Eastern Australia, NEO '93 Conference Proceedings, pp. 591-598. Department of Geology and Geophysics, University of New England, Armidale. MURPHY P. R . , SCHWARZBOCK H . , CRANFIELD L . C . , WITHNALLI. W . & MURRAY C. G . 1976. G e o l o g y o f t h e

Gympie 1:250 000 Sheet area. Geological Queensland Report 96.

Survey

of

P E R M O - T R I A S SIC H U N T E R - B O W E N E V E N T , Q L D MURRAY C. G. 1990. Tectonic evolution and metallogenesis of the Bowen Basin. In: Bowen Basin Symposium 1990 Proceedings, pp. 201-212. Geological Society of Australia, Queensland Division, Brisbane. MURRAY C. G. 1974. Alpine-type ultramafics in the northern part of the Tasman Geosyncline—Possible remnants of Palaeozoic ocean floor. In: Denmead A. K., Tweedale G. W. & Wilson A. F. eds. The Tasman Geosyncline—A Symposium, pp. 161-181. Geological Society of Australia, Queensland Division. MURRAY C . G . , FERGUSSON C . L., FLOOD P. G . , WHITAKER W .

G. & KORSCH R. J. 1987. Plate tectonic model for the Carboniferous evolution of the New England Fold Belt. Australian Journal of Earth Sciences 34, 213-236. PARIANOS J. 1993. Carboniferous to Tertiary geology of the Airlie Block, northeast Queensland. MSc thesis, University of Queensland, Brisbane (unpubl.). PEGREM B. J. 1995. Tarong Basin. In: Ward C. R., Harrington H. J., Mallett C. W. & Beeston J. W. eds. Geology of Australian Coal Basins, pp. 465-470. Geological Society of Australia Coal Geology Group Special Publication 1. PETERSEN M. D. & WESNOUSKY S. G. 1994. Fault slip rates and earthquake histories for active faults in southern California. Bulletin of the Seismological Society of America 84, 1608-1649. ROBERTS C. A. 1992. The geology of the Fat Hen Creek area, Kilkivan, Southeast Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). SLIWA R. 1994. Regional structural geology of the central North D'Aguilar Block, southeastern Queensland. PhD thesis, University of Queensland, Brisbane (unpubl.). SLIWA R . , HOLCOMBE R . J., FIELDING C . R . , LITTLE T . A . ,

BRYAN S. E., FIFOOT A. 1993. Early Permian marine fault basins formed during exhumation of the New England Orogen subduction complex in southeastern Queensland. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, Eastern Australia, NEO '93 Conference

65

Proceedings, pp. 557-564. Department of Geology and Geophysics, University of New England, Armidale. STAINES H . R . E., FALKNER A . J. & THORNTON M . P. 1 9 9 5 .

Ipswich Coalfield. In: Ward C. R., Harrington H. J., Mallett C. W. & Beeston J. W. eds. Geology of Australian Coal Basins, pp. 455-464. Geological Society of Australia Coal Geology Group Special Publication 1. STEPHENS C. J. 1991. The Mungore Cauldron and Gayndah Centre Late Triassic large-scale silicic volcanism in the New England Fold Belt near Gayndah, southeast Queensland. PhD thesis, University of Queensland, Brisbane (unpubl.). STEPHENS C . J., SCHONR. W . S. & EWART A . 1 9 9 3 . M e s o z o i c

crustal extension in the northern NEO: Geochemical and isotopic evidence from large scale silicic magmatism. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, Eastern Australia, NEO '93 Conference Proceedings, pp. 637-642. University of New England. WATERHOUSE J. B . & SIVELL W . J. 1 9 8 7 .

Trans-Tasman

relationships between the Permian-Triassic rocks of Gympie, southeast Queensland, and those of New Caledonia and New Zealand. In: Murray C. G & Waterhouse J. B. eds. 1987 Field Conference Guide, Gympie District, pp. 48-59. Geological Society of Australia, Queensland Division. WILLIAMS B. M. 1993. The Geology of the Meandu Coal Deposit and the Central Tarong Basin, southeast Queensland. BSc. (Hons) thesis, University of Queensland, Brisbane (unpubl.). WILTSHIRE M. J. 1982. Late Triassic and Early Jurassic sedimentation in the Great Artesian Basin. In: Moore P. S. & Mount T. J. eds. Eromanga Basin Symposium, Summary Papers, pp. 59-67. Geological Society of Australia and Petroleum Exploration Society of Australia, Adelaide.

(Received 24 April 1996; accepted 9 January 1997)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 66-79.

Tectonic evolution of the northern New England Fold Belt: Carboniferous to Early Permian transition from active accretion to extension R. J. HOLCOMBE,1 C. J. STEPHENS,1* C. R. FIELDING,1 D. GUST,2 T. A. LITTLE,1* R. SLIWA,lf J. McPHIE3 AND A. EWART1 2

1 Department of Earth Sciences, University of Queensland, Qld 4072, Australia. School of Geology, Queensland University of Technology, GPO Box 2434, Brisbane, Qld 4001, Australia. 3 CODES, University of Tasmania, GPO Box 252-79, Hobart, Tas. 7001, Australia.

The northern New England Fold Belt in central and southern Queensland comprises a collage of terranes the complexity and variety of which dominantly reflect tectonic events in the ~ 100 million year interval from midCarboniferous through the Middle Triassic. We present a synthesis of our current, but evolving, understanding of the Carboniferous to Late Permian tectonic evolution of this region which we argue was controlled by a transition from active continental-arc/accretionary processes to widespread crustal extension. Contrary to earlier interpretations, we find little evidence for the existence of an Early Carboniferous magmatic 'arc' (Connors-Auburn Volcanic Arc) that forms a companion to the more easterly interpreted forearc and accretionary complex elements. These elements, together, have been interpreted widely in terms of segments of a complete subduction complex. Dating of the magmatic elements of assumed Early Carboniferous age shows them to be overwhelmingly Late Carboniferous to Early Permian in age. The Connors-Auburn Volcanic Arc and the Early Permian Camboon Volcanic Arc are thus part of a single, broad magmatic cycle that we propose should be called the Connors-Camboon Province. Significant intraformational unconformities within the volcanics of this province are interpreted as evidence for ongoing extension. The present data allow two alternative end-member interpretations, or perhaps more realistically mixtures of these end-members. These are: (i) the existence of Late Carboniferous subduction and associated magmatism that is transitional in nature to Early Permian extension; and (ii) a more controversial interpretation of a passive margin within which all magmatism of Late Carboniferous to Early Permian age records a continental-scale extension and thermal event. Evidence for this thermal event is recorded much further east of the traditionally recognised magmatic province in the Connors-Camboon Province as extension and crustal melting in the earlier forearc assemblages. Irrespective of these interpretations, the calcalkaline-dominated phase of volcanism is transitional into the late Early Permian and formation of the Bowen Basin, a basin that originally extended eastward of the present structural margin to a shallow-marine coastline. The extensional phase of basin formation is characterised by clastic sedimentation into developing grabens and halfgrabens and associated mafic-dominated volcanism. Elements of basin formation are recognised throughout the northern New England Fold Belt and the distribution of 'Bowen Basin' rocks differs significantly from traditional models. Key words: continental accretion, crustal extension, New England Fold Belt, tectonics.

INTRODUCTION The N e w England Fold Belt of eastern Australia (Figure 1) comprises a n u m b e r of variably deformed terranes, ranging in age f r o m Early Palaeozoic to Late Triassic. E a c h terrane is g e n e r a l l y interpreted to reflect a different tectonic element or record a different orogenic episode, and generally has been regarded as forming within a broadly convergent continental margin setting (Day et al. 1978, 1983; Henderson 1980). This paper, and its companion (Holcombe et al. 1997), present a synthesis of our current, but evolving, understanding of the post-Devonian t i m e - s p a c e framework and provide constraints on tectonic m o d e l s for the northern N e w England Fold Belt in central Queensland. The m a i n tectonic elements of the northern N e w England Fold Belt had been recognised by the mid-

1970s. Day et al. (1978), in a m a j o r summary of the regional geological history, presented a model involving the periodic development of a magmatic arc along what is now the Connors-Auburn Arch. This was interpreted in terms of a convergent continental margin, with most of the northern N e w England Fold Belt elements in the arc-forearc region. The tectonic history was generally thought of in terms of three arc-producing events: (i) the Siluro-Devonian Calliope Arc, developed as an island arc then subsequently accreted; (ii) the D e v o n i a n * Present address: Central Norseman Gold Corporation, Norseman WA 6443, Australia. {Present address: Department of Geology, Victoria University of Wellington, Wellington, New Zealand. tPresent address: Etheridge, Henley & Williams, 457 Upper Edward Street, Spring Hill, Qld 4000, Australia.

PERMO-CARBONIFEROUS EVOLUTION, QLD Carboniferous Connors-Auburn Volcanic Arc, developed along an Andean-style margin; and (iii) the Early Permian Camboon Volcanic Arc, developed essentially on top of the Connors-Auburn Volcanic Arc. Of these separate orogenic' events, however, only the Late Devonian-Carboniferous arc was associated with most of the tectonic elements expected at a continental margin subduction orogen (Day et al 1978; Murray et al 1987). These elements comprise the Wandilla Slope and Basin accretionary complex, the Yarrol forearc basin, and the coeval, possibly backarc extensional Drummond Basin (Johnson & Henderson 1991; Henderson et al 1993). The forearc-setting interpretations have been supported by more recent work (Fergusson et al 1990; Leitch et al 1994), and Leitch et al (1993) suggested that the final timing of active accretion persisted into the latest Carboniferous. Several authors have argued for possible allochthonous elements within the fold belt. Both the Palaeozoic North D'Aguilar and the Permo-Triassic Gympie Blocks, in particular, are characterised by units that are dissimilar in both composition and metamorphic grade to adjacent rocks thought to be of the same age. It has also been suggested that the North D'Aguilar Block, with its dominantly mafic metavolcanic units and large serpentinite bodies, was an accreted arc or backarc (Waterhouse & Sivell 1987a). Various origins have been suggested for the Gympie Block ranging from an accreted arc (Sivell & Waterhouse 1987) to an exotic displaced terrane derived perhaps from New Zealand (Harrington 1974, 1983; Waterhouse & Sivell 1987b). In the northern part of the fold belt, the origin of the serpentinites, and greenschist and amphibolite facies schists of the Marlborough Block remained somewhat enigmatic. The ultramafic components have generally been regarded as ophiolitic in origin (Murray 1974; Leitch et al 1994). The schists were regarded as distinct from the surrounding coastal accretionary terranes (Henderson et al 1993).

67

4

Figure 1 Location of major terranes of the northern New England Fold Belt.

The last decade has seen a resurgence of new research in the northern New England Fold Belt in Queensland. Independent groups working on projects as diverse as southeast Queensland structure and tectonics, Bowen Basin sedimentological studies, and Permian volcanic petrology began to converge by the early 1990s toward common conclusions that, in some respects, differ considerably from the earlier tectonic models. In particular, we now recognise two departures from previous models: (i) we argue that there are limited data to support the existence of an Early Carboniferous volcanic arc that developed coevally

Bowen Connors- Gogango Yarrol Basin Auburn Overfolded Block Block Zone granites, caldera complexes ^ extensional basins

coastal blocks East

Ma West

Figure 2 Time—space plot of major tectonostratigraphic terranes and events in the northern New England Fold Belt. The horizontal 'space' scale is relative to the present position of (from right to left) the accretionary terranes, The Yarrol forearc basin, the Connors-Auburn 'arc' and the eastern Bowen Basin margin.

thermal sag extensional sub-basins Camboon P 0 Connors-cAuburn—^ | Urannah 300 310

Crustal events

Slab retreat <d Slab £ Thrust-folding O c CQ CD advance calc-alkaline "V CD CD 2 magmatism CO NEFB Thermal sag entirely within backarcProtracted foreland (including Gympie Block?) crustal extension Extension

^Ton^iianne marin^ ^^ forearc? (Neerkol Fm)

syntectonic S-type granitoids

?arc to west? "forearc" core complexes

320

i _UJV p-

330 340

Torsdale Arc?

forearc basin (Yarrol Basin)

350 360 370 380

arc/forearc? (Calliope Assemblage)

. accretionary I complex 1 (Wandilla, | Shoalwater q 7 terranes)

i

Slab retreat?

accretion/ (arc??)

Subduction/ accretion?

J.-si?

J

gpo

68

R.J. H O L C O M B E E T A L . Beenleigh and South D'Aguilar Blocks (Figures 1, 3) comprise low grade, internally discontinuous (faultbounded) units of ribbon chert, argillite, chert-argillite melange, greywacke and metabasaltic greenstone (including pillow basalt). The rocks (NeranleighFernvale beds) are moderately to strongly deformed, generally containing a single generation, steeply westdipping slaty cleavage accompanying a transposition layering. Remnants of similar rocks occur throughout the North D'Aguilar Block (Wide Bay Creek Broken Formation, Booloumba beds, Amamoor beds: Table 1). Metamorphic grades are low, prehnite-pumpellyite or prehnite-actinolite facies assemblages occur in the basaltic rocks; anchizonal conditions, with T<250°C and moderate pressure, are inferred from illite crystallinity and b 0 data in the metasediments (Sliwa 1994). These rocks, containing Late Devonian to Early Carboniferous macroinvertebrates and radiolarians (Fleming et al 1974; Aitchison 1988; Ishiga 1990), are interpreted as representing upper levels of the accretionary prism, and are typical of other accretionary assemblages in the Woolomin terrane in the southern New England Fold Belt (Cawood & Leitch 1985) and the Wandilla/Shoalwater terranes in the northern New England Fold Belt (Fleming et al. 1975; Fergusson etal 1993).

with, and to the west of, the subduction complex elements; and (ii) we also recognise a major crustal thermal event in the Late Carboniferous-earliest Permian, that arguably records a transition from a convergent to an extensional, possibly backarc, setting. Figure 2 is a time-space plot of the northern New England Fold Belt that summarises our current state of knowledge of the critical terranes on which we base much of our tectonic analysis of the post-Devonian history. Elements of this framework have appeared in numerous conference proceedings and recent unpublished theses. The stratigraphic time-scale used for correlation is based on Jones (1996) and future refinements of this will change some critical event boundaries, particularly in the Permian.

MID-CARBONIFEROUS ACCRETIONARY ELEMENTS Southeast Queensland The oldest rocks in the southern Queensland section of the New England Fold Belt are strongly deformed, Devonian?-Carboniferous accretionary rocks exposed within several discrete 'basement blocks'. The

younger faults NPF - North Pine Fault system

Gympie GYMPIE VBLOCK/

Permian & Triassic units/plutons MG - Monsildale Granodiorite

Thrust faults CT - Claddagh Thrust system Early - Middle Permian marine fault basins CV - Cederton Volcanics CB - Cambroon Beds MB - Marumba Beds Late Carboniferous syntectonic granitoids CG - Claddagh Granodiorite GG - Gallangowan Granodiorite Accretionary terranes

Upper Plate simply deformed, low grade /multiply deformed, low grade WB - Wide Bay Creek Broken Fm BB - Booloumba Beds AB - Amamoor Beds YB - Yarraman Block (accretionary?) BP - Bunya Phyllite

Mt Mia _ detachment fault Lower Plate

V l T CLARENCE MORETON 28°S BASIN 1^3°

:: SOUTH;:: D'AGUILAR Sj BLOCKS

polydeformed schists /serpentinite MM - Mt Mia Serpentinite -matrix Melange RG - Rocksberg Greenstone

0

10

20

30km

Figure 3 Structural f r a m e work of the North D'Aguilar Block in southern Queensland showing units and features discussed in the text.

PERMO-CARBONIFEROUS EVOLUTION, QLD

69

Table 1 Compilation of previously unpublished K-Ar isotopic data used in these studies, or data previously published only informally in conference abstracts. Nominal map unit

Location

Rock type

granite Gayndah granite Gayndah granite Gayndah rhyolite Gayndah granite Gayndah granite Gayndah granite Gayndah granite Gayndah granite Gayndah gabbro Kilkivan granite Esk Trough North D'Aguilar Block andesite Undifferentiated volcanics North D'Aguilar Block andesite Undifferentiated volcanics North D'Aguilar Block andesite basalt Gayndah Wetheron Basalt Connors Arch ignimbrite Connors Volcanics basalt Cracow Unnamed volcanics

Mt Walsh Granite Mt Walsh Granite Mt Walsh Granite Aranbanga Volcanics Mungore Granite Mungore Granite Mungore Granite Mungore Granite Mungore Granite Unnamed gabbro Unnamed granite Undifferentiated volcanics

Age 170.6 + 2.2 172.2 + 2 175.3 ± 3 224.2 + 3.2 215.2 ± 2 215.8 + 2 228.4 ± 2.5 228.8 ± 2.4 230 ± 3 231.1 + 3 255.4 ± 5

Longitude E Latitude S

Sample No.*#

References

-25.65847 -25.65847 -25.65847 -25.5252 -25.74698 -25.74698 -25.83645 -25.83645 -25.78297 -26.1455 -26.64118

B2203/2 (b) B2203/2 (b) B2203/2 (b) S987/4 (b) G3058/1 (b) G3058/1 (b) G3187/2 (b) G3187/2 (b) B2314/2 (b) SR2 (p) RG50 (a)

Stephens 1991 Stephens 1991 Stephens 1991 Stephens 1991 Stephens 1991 Stephens 1991 Stephens 1991 Stephens 1991 Stephens 1991 This paper Grayson 1995

152.07042 152.07042 152.07042 151.65251 151.94112 151.94112 151.83457 151.83457 152.0545 152.28968 152.28865

224.6 ± 4

152.41534

-26.42506

RS350 (wr)* Sliwa 1994

224.7 ± 3

152.27883

-26.12106

CR003 (wr)* Roberts 1992

234.8 + 3 221.2 + 5.8 228.6 + 3 230.2+ 10

152.41534 151.77403 149.52694 150.28317

-26.42506 -25.51272 -22.82883 -25.16689

RS361 (wr)* Sliwa 1994 Stephens 1991 S060/4 (p) ER046 (wr)* O'Brien 1994 C95-627 (p) J. A. Jones unpubl. data

152.43032

-26.64181

RS316 (wr)* Sliwa 1994

152.43447

-26.61474

RS345 (a)*

Sliwa 1994

149.85108

-23.02845

CSUQ72 (p)

O'Connell 1995

Camboon Volcanics

North D'Aguilar andesite 243.3 + 3 boulder Block North D'Aguilar granodiorite 311.4 ± 4.3 boulder Block Gogango Overfolded Zone 258.1 ± 5 dolerite Gogango Overfolded Zone 252.9 + 4 dolerite Gogango Overfolded Zone 241.6 ± 3 gabbro granodiorite 227 ± 3 Esk Trough granodiorite 229.2 + 3 Esk Trough 240.1 + 3 gabbro Esk Trough granodiorite 247.5 ± 3 Esk Trough Gogango Overfolded Zone 200.8 ± 6 gabbro 277.9 + 4 andesite Cracow

Camboon Volcanics Camboon Volcanics

Cracow Cracow

260.2 + 3 282.8 ± 5

144.26311 150.28041

-25.26535 -25.37982

Camboon Volcanics Nob Creek Granite

296.3 ± 4 andesite Cracow Marlborough Block 268.4 ± 3 granite North D'Aguilar Block granodiorite 309.3 + 4 North D'Aguilar Block granodiorite 289.3 ± 4 North D'Aguilar Block granodiorite 310.6 + 4 North D'Aguilar Block quartz-mica schist 282.0 ± 4 312 + 4 ignimbrite Cracow

150.26311

Marumba beds Marumba beds Rookwood Volcanics Youlambie Conglomerate Eulogie Park Gabbro Monsildale Granite Monsildale Granite Monsildale Granite Monsildale Granite Westwood Gabbro

Gallangowan Granite Yabba Creek Granite Yabba Creek Granite Booloumba beds Torsdale beds

andesite basalt

150.06285

-23.35226

CSUQ23 (p) This paper

150.40566 152.39678 152.38729 152.38506 152.40099

-23.86149 -26.7148 -26.702 -26.68723 -26.67871

CSUQ82 (p) This paper Sliwa 1994 RS363 (a)* QUT245 (a)* Kwiecien 1996 QUT294 (a)* Kwiecien 1996 Sliwa 1994 RS431 (a)*

150.72021 144.26311

-23.61514 -25.26535

-25.26535

CSUQ81 (p) This paper J. A. Jones C57(p) etal 1996 Jones et al. 1996 C63 (p) J. A. Jones C8 (p) unpubl. data Jones et al. 1996 C63 (p)

149.94453

-22.71334

VNM (b)*

Muscio 1994

152.45696

-26.53086

RS339 (b)*

Sliwa 1994

152.48695

-26.55986

RS374 (a)*

Sliwa 1994

152.48695

-26.55986

RS374 (a)*

Sliwa 1994

152.40385 150.3162

-26.70761 -25.15306

Casley 1993 3D1 (m)* C94-546 (a) Jones etal. 1996

* A correction has been applied to all data derived from the University of Queensland laboratory between 1991 and 1995 and arises from a rounding error in data calculation; the corrected data are indicated by *. # The dated mineral is shown in brackets after the sample number: m, muscovite; b, biotite; p, plagioclase; a, amphibole; h, hornblende; wr, whole rock.

70

R. J. HOLCOMBE ET AL.

The North D'Aguilar Block (Figure 1) is a composite terrane consisting of deeply subducted (18-20 km), polymetamorphic, but originally epidote-blueschist facies, ophiolitic assemblages (Rocksberg Greenstone, Mt Mia serpentinite-matrix melange: Table 1) juxtaposed against the low-grade accretionary rocks by a complex system of low-angle detachment faults, steep normal faults, and younger thrust faults (Little et al. 1992, 1993; Sliwa 1994). The Rocksberg Greenstone outcrops as a semi-coherent, mafic metavolcaniclastic unit in the southern part of the North D'Aguilar Block, and as detached blocks and 'knockers' within the serpentinite-matrix melange in the northern part of the block. Although strongly overprinted by later deformation, the initial fabric in the Rocksberg Greenstone is steeply dipping, and is interpreted to be an accretionrelated fabric similar to that in the lower grade basement blocks. Strain is highly partitioned in this earliest deformation and the highly schistose, map-scale zones within which epidote-blueschist assemblages are preferentially developed are thought to represent deep fluid-focusing shear zones within the subducting pile (Holcombe & Little 1994). These rocks are interpreted as underplated ophiolitic components of the accretionary wedge, that were detached and subducted to depths > 18 km before being exhumed in the footwall of an extensional detachment in the Late Carboniferous (see below). Marlborough Block The Marlborough Block near the northern end of the New England Fold Belt is a thin-skinned nappe sheet (Holcombe et al 1995; 1997) carrying composite terranes that are very similar to those in the North D'Aguilar Block in southeast Queensland. Within the block, serpentinised ultramafic rocks, including harzburgite, cumulate, and gabbro are juxtaposed by steep (rotated?) thrusts and shear zones against chertargillite, amphibolite, schist, and locally marble of greenschist to amphibolite facies metamorphic grade. We now recognise that the schist units include deformed equivalents of foliated S-type granitoids that occur elsewhere within the block (see below), and we believe that much of the chert-argillite component resembles the Devonian-Carboniferous accretionary rocks exposed further to the east (but see Henderson et al 1993). Unlike the North D'Aguilar Block, serpentinite-matrix melange is rare and most of the serpentinite bodies in the Marlborough Block are serpentinised massive ultramafic rocks. Sheared serpentinite is most commonly associated with Cretaceous or Tertiary faults. Age constraints on accretion The age of the first deformational fabric in the accretionary rocks, and hence on the timing of active accretion, is poorly constrained at mid-Carboniferous. An absolute upper limit is imposed by the tightly constrained ca 305 Ma overprinting metamorphic and

magmatic event (below). Aitchison (1988) described Early Carboniferous (Tournaisian) radiolarians from ribbon cherts of the Neranleigh-Fernvale beds. Little et al. (1995) interpret a complex Ar/ Ar whole-rock spectrum taken from a slate in the higher level accretionary rocks as indicating a minimum age on the subduction-related fabric of ca 315 Ma but with most of the high K/Ca steps >330 Ma. 40

39

LATE CARBONIFEROUS EXTENSIONAL STRUCTURES AND S-TYPE GRANITOIDS Southeast Queensland In the far northern part of the North D'Aguilar Block (Figures 1, 3), tectonostratigraphic relationships between the polymetamorphic, higher grade, ophiolitic rocks and the low-grade, single-fabric, accretionary rocks are well-defined (Little 1992; Donchak et al. 1995). The higher grade rocks are overlain structurally by the thick Mt Mia serpentinite-matrix melange sheet, which is separated in turn from the overlying low-grade rocks by a gently arched detachment fault (Little et al 1993). Shallowly dipping, second generation, deformation fabrics (subparallel to the detachment) dominate most outcrops of the higher grade rocks and these fabrics are accompanied by a regional greenschist facies overprint of the earlier epidote-blueschist facies mineral assemblages. Similar fabrics, with similar detachmentparallel orientation, occur within fault-bounded phyllitic units exposed along the detachment. These appear to represent hangingwall accretionary rocks (e.g. Anderson Creek Phyllite and probably Bunya Phyllite) that were derived from intermediate depth levels of the complex, heated and recrystallised by contact with the footwall, and then accreted to the lower plate. The Greenbank Layered Sequence, with its shallowly dipping reflectors seismically imaged at depths equivalent to > 1.5 s twoway time below the Beenleigh Block (Korsch et al 1989), probably represents these same fabrics and detachment system. The pressure gap across the detachment fault between the epidote-blueschist facies rocks (>0.6 GPa) and the anchizonal accretionary sediments is -0.3 GPa, representing a depth difference of - 9 km (Little et al 1995). The geometry is that of a core complex where a crustal-scale extensional detachment has exhumed a lower plate of deep-seated, ductilely deformed rocks against an upper plate of high-level, more brittly deformed rocks (Little et al 1993). Similar extensional core complexes occur in other convergent margin settings (Lee & Lister 1992; Wijbrans et al 1993). There is some question as to whether the arched nature of the detachment system in the northern North D'Aguilar Block is a function of the original core complex geometry as suggested by Little et al (1993) or has been overprinted by later flexuring. Recent mapping in the area indicates multiple thrust imbrication of some units (Donchak et al 1995), suggesting that this later contraction (Holcombe et al 1997) may control the overall geometry. Broad folding of shallowly dipping

PERMO-CARBONIFEROUS EVOLUTION, QLD

thrust fabrics also occurs in the central part of the block (Sliwa 1994). A suite of Late Carboniferous, mildly S-type, granodiorite and diorite plutons are exposed within a thrust sheet (Claddagh Thrust) in the northern North D'Aguilar Block (Little et al 1993). The plutons intrude the lower plate, polymetamorphic, rocks and include the Claddagh and Gallangowan granodiorites and several smaller unnamed plutons (Sliwa 1994). Granodiorite intrusion was synkinematic with the deformation forming the detachment-parallel fabrics in the higher grade rocks (Little et al 1993, 1994, 1995). The granodiorite bodies are variably foliated, and xenoliths of more deformed (mylonitic) phases are common. The regional greenschist facies thermal metamorphic overprint increases to amphibolite facies around the igneous bodies, and the foliation within the granodiorite is parallel to the detachment-parallel fabrics in the aureole rocks. Because of the demonstrable structural relationship between magmatism, metamorphism and deformation, the timing of the deformation fabrics accompanying exhumation can be reliably dated. Ar/ Ar dating of hornblende and white mica from the Claddagh granodiorite and its aureole indicate syntectonic crystallisation at ca 306 Ma (306.5±0.6; 306.9±1.2 Ma ages on hornblende), followed by continued exhumation and cooling until all rocks had passed through the ~350°C white mica blocking temperature by ca 296 Ma (Little et al 1995). K/Ar ages on these foliated granodiorites are regarded as less precise and yield slightly older ages of ca 310 Ma Sliwa (1994; Table 1). The precise dating of a significant crustal extension, and crustal melting, event at ca 305 Ma is a major constraint on our tectonic models and poses a significant problem when related to the coeval peak of apparent arc magmatism occurring to the west. 40

39

Other foliated S-type granitoid terranes Foliated S-type granitoids similar to the CladdaghGallangowan bodies described above occur sporadically throughout the New England Fold Belt. They are generally intrusive into the earlier accretionary rocks and are locally associated with metamorphic culminations in the fold belt. In the southern New England Fold Belt these include the Hillgrove igneous suite which is considered to have an emplacement age of ca 300 Ma but for which locally other origins have been proposed [thrust model (Dirks et al 1993); diapiric models (Shaw & Flood 1981)]. A widespread, poorly exposed, area of garnet(cordierite)-bearing gneissic granitoids occurs about 90 km west of the North D'Aguilar Block bodies. These metagranitoids, that we have termed the Chahpingah Complex (Figure 1), are problematic in terms of their regional affinities and have been undescribed previously except in unpublished Honours theses (Arthur 1969; Virisheff 1974). They have a gently folded, shallowly dipping, locally shear-dominated, foliation. Preliminary interpretation of Ar/ Ar step-heating spectra of 40

39

71

biotite and muscovite from the metagranitoids indicates distinct groupings of cooling ages between 250 and 225 Ma as these rocks passed through the ~300°C isotherm. We surmise that they were emplaced syntectonically, probably before ca 250 Ma and remained at depth until emplaced to their present structural position during the Permo-Triassic thrust event and then were subsequently exhumed. We have mapped similar foliated garnet-bearing granitoid terranes within the Marlborough Block thrust nappe in the northern New England Fold Belt. Locally we correlate the gneissic metagranite and their amphibolite facies schistose aureole with the Broome Head Metamorphics (Morand 1993) to the east. Isotopic ages derived from these rocks are ambiguous. Leitch et al (1993) reported a Rb/Sr age of ca 255 Ma from the Broome Head Metamorphics and Muscio (1994) derived a biotite K/Ar date of 268 ±3 Ma from one of the least-deformed foliated granites in the Marlborough Block. Preliminary interpretation of Ar/ Ar dating of biotite from other granitoids in this block yield cooling ages of 248.8±0.5 Ma from an unfoliated granodiorite and 242.8±0.4 Ma from a strongly foliated phase (Table 1). We suspect that the intrusion age of these Stype granitoids is probably older than Late PermianEarly Triassic and that the younger ages reflect exhumation and cooling during the massive thrust event in this region. 40

39

CARBONIFEROUS TO EARLY PERMIAN CALC-ALKALINE MAGMATISM Calc-alkaline rocks of the Connors and Torsdale beds are considered to represent the Early to mid-Carboniferous arc and the Early Permian Camboon Volcanics are the basis for postulating an Early Permian arc (Day et al 1978). Field and geochronologic studies, directed towards the distinction between the two groups of calcalkaline rocks, indicate major problems with the previous arc models. The Camboon Volcanics and Torsdale beds (Jones 1994; Jones et al 1996) occur on the western side of the Auburn Arch. The Torsdale beds are dominantly silicic ignimbrites that are intruded by numerous small granite bodies. Recent K/Ar dating of hornblende from a sample collected from near the unconformity with the overlying Camboon Volcanics gave an age of 312±4Ma (Jones et al 1996). The unconformity is marked by coarse basal conglomerates composed predominanty of granitoid and felsic volcanic clasts. The Camboon Volcanics consists of a lower bimodal mafic lava/felsic ignimbrite unit and an upper terrestrial andesitic sequence from which a K/Ar age on plagioclase of 282.8±5 Ma was obtained at the top of the sequence (Jones et al 1996). The Camboon Volcanics are unconformably overlain by the late Artinskian {ca 280-275 Ma: Jones 1996) marine Buffel Formation of the Bowen Basin. Inliers of similar volcanics are exposed within the Gogango Overfolded Zone along the trend of the Connors-Auburn Arch. These were interpreted on

72

R. J. HOLCOMBE ETAL.

1:250 000 maps as Camboon Volcanics in the south and as components of the Rannes beds in the north (Malone et al 1969; Kirkegaard et al 1970; Dear et al 1971). Our mapping shows that in both areas, these volcanics are the structurally deepest rocks exposed at the base of imbricate thrust sheets or within antiformal cores (Holcombe et al 1995, 1997). Rocks mapped as Camboon Volcanics are variably deformed intermediate volcanic and volcaniclastic rocks. These rocks are predominantly subaerial, although intermediatecomposition pillow lavas are locally developed. In the north, rocks mapped as Connors Volcanics or Rannes beds consist predominantly of silicic volcanic and volcaniclastic rocks with relatively minor intermediate and mafic volcanic rocks. Within the Connors Arch, the Connors Volcanics, and the lower parts of units mapped as Lizzie Creek Volcanics and Carmila beds, consist of a stratigraphically complex sequence dominated by dacitic to felsic ignimbrite, rhyolite and volcaniclastic rocks, with lesser andesite and basalt. This sequence has an upper contact with a predominantly fine-grained, lacustrine sedimentary succession that typically is associated with basaltic volcanism, and which we equate on lithologic and structural grounds with the Early Permian extension phase of the Bowen Basin (see below, and Fielding et al 1997). In rocks mapped as Carmila beds north of the Marlborough Block (Kirkegaard 1970) we have observed a mixed sequence of fine-grained sedimentary, rhyolitic and coarse-grained dacitic volcaniclastic rocks containing Early Permian Fauna 1 marine fossils occurring within tens of metres of the contact with the overlying fine-grained succession. EARLY PERMIAN RIFT-SAG BASINS Bowen Basin elements in the Gogango Overfolded Zone Following on from the work of others, Fielding et al. (1990, 1995) and Baker et al. (1993) presented a threepart model for the Bowen Basin, based on stratigraphic correlations, interpreted sediment dispersal patterns and sandstone petrology. According to this model, the stratigraphic record of the Bowen Basin may be divided into three suites, reflecting (i) an Early Permian phase of extensional subsidence and magmatic activity, (ii) an early Late Permian period of mainly passive, thermal subsidence, and (iii) a late Late Permian to early Late Triassic phase of foreland loading and contractional deformation. The last phase is associated with the onset of the Hunter-Bowen contractional event (Holcombe et al. 1997) and is not treated here. Tectonic interpretation of deformed Permian rocks within the Gogango Overfolded Zone has been hampered by a stratigraphic nomenclature based in part on deformation style rather than stratigraphic criteria (e.g. Rannes beds: Holcombe et al. 1997) and inconsistent stratigraphic ordering and correlation of units particulaly between adjacent map sheets. A critical step in our understanding of the Permian framework of the northern

New England Fold Belt has been regional stratigraphic correlations based on recognition of distinct assemblages of sedimentary lithofacies and associated igneous rock types, sediment dispersal patterns and interpreted palaeoenvironments, that are characteristic of the three phase evolution of the Bowen Basin. The issues summarised below are developed in more detail by Fielding et al. (1997). Phase 1 rocks are a diverse assemblage of fine- and coarse-grained sedimentary rocks, interbedded with a mafic-dominated bimodal igneous suite of mafic lavas and sills, minor ignimbrite and associated tuffs. Palaeocurrent distributions and sediment compositions indicate mainly local sediment derivation from multiple sources. Sequences in the northern New England Fold Belt that we correlate with Phase 1 are the Youlambie Conglomerate and upper, fine-grained sedimentary parts of both the Carmila beds and Lizzie Creek Volcanics. Locally, the marine Rookwood Volcanics are laterally transitional to the upper part of the Youlambie Conglomerate. Except for the Rookwood Volcanics, depositional environments during much of this phase were non-marine, mainly alluvial and lacustrine. Sediment accumulation is interpreted to have occurred within active grabens and half-grabens, separated by uplifted basement blocks. A major change in depositional environment occurred towards the close of the extensional phase involving a diachronous marine transgression which flooded much of the basin and led to the widespread establishment of coastal to marine shelf conditions. Bioclastic limestones were sporadically (but extensively) developed on remnant basement highs. The top of this succession is a basin-wide hiatal surface that almost certainly relates to tectonic processes within or marginal to the basin. Phase 2 sediments form a continuous sheet across the basin, in contrast to underlying Phase 1 deposits, and sediment dispersal was westward (towards the Bowen Basin) on both sides of the 'Connors Arch'. Resulting, mainly fine-grained, clastic sediments are now represented by the undifferentiated Back Creek Group ('Pb' of the local map sheets) and parts of the Rannes beds, Moah Creek beds and Boomer Formation. Rookwood Volcanics The marine Rookwood Volcanics has been one of the most enigmatic units in the northern New England Fold Belt. The unit was ascribed a late Early Permian age on the basis of a single fossil locality (Briggs 1993). The stratigraphic context of the unit, nonetheless, is poorly constrained on published map sheets. The Rookwood Volcanics consist almost entirely of pillow lavas and high-level intrusives with minor associated volcaniclastic sediments. Sediment intercalations are rare within the Rookwood Volcanics, but locally form accumulations ranging from a few centimetres up to at least 300 m-thick. Slices of massive rhyodacite and pillow basalt occur together within several thrust imbricates at Comanche Station west of Rockhampton but their contact relationships are unclear.

PERMO-CARBONIFEROUS EVOLUTION, QLD Most contacts between the Rookwood Volcanics and other units are either faulted or obscured by poor outcrop, but the Rookwood Volcanics commonly structurally overlie the Early Permian Youlambie Conglomerate. In a complete section that we have mapped through the Rookwood Volcanics near the Fitzroy River, west of Rockhampton (and see O'Connell 1995) the Rookwood Volcanics both overlie and pass laterally into the Youlambie Conglomerate. The unit here is overlain by rocks we equate with Phase 2. Early Permian (Artinskian) fossils have been found in the uppermost Youlambie Conglomerate and near the top of the Rookwood Volcanics. K/Ar dating of two dolerites in the Rookwood Volcanics yield ages of 258±5 Ma and 253±4 Ma (O'Connell 1995; Table 1) but we interpret these to be ages that have been reset during the subsequent contractional event. Locally abundant trace fossils (Cruziana ichnofacies), macroinvertebrate fragments and the one foraminifer locality, indicate a marine origin for the sequence, with water no deeper than shelfal (-200 m). Small-scale, soft-sediment deformation is common, and attests to the likely unstable nature of the depositional basin. The abrupt lateral thinning of the Rookwood Volcanics and equivalence with the uppermost Youlambie Conglomerate at this locality is interpreted to define the hinge margin of a developing half-graben. In a regional sense, the Rookwood Volcanics are petrographically indistinguishable from other Phase 1 lavas. Malone et al. (1969) interpreted the textures and petrographic characteristics of the Rookwood Volcanics as due to spilitisation, and this interpretation has been raised consistently as evidence for a deep-marine eruptive environment. The identical secondary mineralogy and textures observed within all Phase 1 lavas, some of which were erupted subaerially, are clearly due to burial metamorphism during later basin evolution. Berserker beds The Berserker beds in the Rockhampton area are another marine volcanic unit within the northern New England Fold Belt for which regional correlations are poorly constrained. The unit is composed of intermediate and felsic volcanic rocks and volcaniclastic sedimentary rocks, apparently passing upward into finegrained marine, commonly bioturbated, sedimentary units containing abundant trace fossils (Cruziana ichnofacies) and Early Permian shallow-marine macrofossils (late Sakmarian to early Artinskian: Kirkegaard et al. 1970; Sainty 1992; Briggs 1993). The unit is host to the Mt Chalmers volcanic-hosted massive sulfide-style deposit and, as with the Rookwood Volcanics, there is a contradiction between the shallow-marine conditions inferred from the fossil assemblage, and the deepmarine (>1000 m) conditions commonly inferred for exhalative sulfides (Hunns 1994). The internal structure is one of gently undulating dips, but the unit is entirely exposed between two northwesttrending thrust faults (Wilmott et al. 1986). The southera end of this belt passes offshore, and the northern end

73

is covered by the Marlborough Thrust nappe, where several isolated klippen of serpentinite overlying the Berserker beds attest to the subhorizontal nature of the floor thrust. The trace fossil assemblage and character of the sedimentary units in the upper part of the stratigraphy are similar to those in sedimentary members in the nearby Rookwood Volcanics, and the fossil age ranges may overlap. The geochemical character of the two sequences of volcanics is markedly dissimilar. An area of felsic volcanic, volcaniclastic and sedimentary rocks that we correlate with Connors Volcanics/Camboon Andesite (see above and Fielding et al. 1997) is exposed to the north of the Marlborough Block, at Charon Point. We suggest this area and the Berserker beds may be contiguous, and have been overridden by the Marlborough Block nappe. If this is the case, then the Berserker beds are most likely the regional equivalents of the Carboniferous-Early Permian volcanics, and the later Early Permian sediments equivalent to Phase 1 sedimentary strata further west. Southeast Queensland basins Several small, fault-bounded basins and basin remnants, characterised by indurated, labile sedimentary rocks, overlie the exhumed accretionary rocks of the North D'Aguilar Block. The principal unit is the Marumba beds. Rock types include lithic sandstone, clast and matrix-supported conglomerate, megabreccia (clasts up to 10 m-diameter), massive argillite (locally calcareous) and minor basalt (Sliwa 1994). Internal disruption of bedding, consistent with synsedimentary deformation in a local fault basin, is common. Clasts are dominantly intermediate but also include chert-argillite rocks similar to those in the low-grade accretionary rocks and mainly unfoliated granitoid. No blue amphibole-bearing rocks or serpentinite, rock types characteristic of the deepest structural levels of the adjacent metamorphic basement, have been found as clasts, although abundant foliated S-type granodiorite clasts similar to the adjacent Gallangowan Granodiorite have been found at one locality. The sequence is moderately to steeply dipping and relatively uncleaved, and exposed within the same thrust sheet that contains Late Carboniferous foliated granitoid. Similar Early Permian basins, containing abundant mud-matrix conglomerate and breccia and interpreted as extensional in origin, occur extensively within the New England Fold Belt [Barnard Basin (Leitch 1988); Reids Dome beds (Draper & Beeston 1985); Fitzroy region (Fielding et al. 1997)]. We propose that the Marumba beds may represent an early extensional manifestation of the Esk Trough. We correlate andesitic rocks of the Cedarton Volcanics with the Marumba beds on the basis of the similar degree of metamorphism and presence of Early Permian macrofossils (Hill et al. 1972; Murphy et al. 1976; Murray et al. 1979). The Marumba beds may also be equivalent to the seismically-imaged 'deep riftfill' interpreted by Korsch et al. (1989) to underlie the Esk Trough.

74 R . J . H O L C O M B E ETAL. The Cambroon beds (formerly included as part of the Amamoor beds), are a deformed sequence of sedimentary rocks within the North D'Aguilar Block containing mud-matrix conglomerate and Early Permian macrofossils (Murray et al. 1979) and radiolarians (Ishiga 1990). In contrast to the other Early Permian units in southeast Queensland, the Cambroon beds are strongly cleaved and locally polydeformed. Their juxtaposition against accretionary-complex rocks (Bouloomba beds) that preserve an older, less complex, deformational history is an important element in assessing the tectonics of the nearby Gympie Block (see below). Gympie Block The Early Permian units of the Gympie Block (Highbury Volcanics, Rammutt Formation) are dominated by basaltic and andesitic volcanics and volcaniclastics, an association quite distinct in character from the mud-matrix conglomerate-bearing units of the Marumba and Cambroon beds. Both sequences, nonetheless, contain abundant volcanic detritus of basic to intermediate composition and there is evidence in the nearby Cressbrook Creek Block for similar age marine fossil-bearing assemblages to the overlying midPermian South Curra Limestone. There are no recognised equivalents to the Permian and Triassic sequences that overlie the limestone. Permian sequences in the Gympie Block are not unlike those of similar age elsewhere in the northern New England Fold Belt. The lower volcanic and sedimentary units are similar to Phase 1 extensional lithologies although, like the Berserker beds, may contain older volcanic units. The northeasterly trending normal growth faults in the Rammutt Formation, which are important in localising gold mineralisation in the black slate facies at Gympie, are also consistent with formation in an extensional basin (Cuneen 1994). The overlying South Curra Limestone is similar in age and lithology to the late Phase 1 and Phase 2 marine sequences in the Gogango Overfolded Zone to the north, and the Late Permian transition into the overlying clastic sequences may include the Phase 2-Phase 3 thrust loading transition. The structural style of the Gympie Block rocks has contributed as much as anything else to misconceptions and speculation about the origin of the sequence. The (apparently) most deformed rocks are those of the strongly cleaved Triassic Kin Kin Phyllite at the top of both the structural and stratigraphic sequence. However, the metamorphic grade (very low greenschist facies) and degree of cleavage in the Kin Kin Phyllite is similar in all units. Both the intensity and orientation of cleavage vary considerably, a feature consistent with development in an easterly dipping shear regime with local strain partitioning. Strain is distributed more uniformly in the Kin Kin Phyllite, the weakest unit in the sequence. The large kink folds that are common in this unit represent very low strain magnitudes and may have formed as late as the Cretaceous deformation in the nearby Maryborough Basin.

The lithologies and deformation history of the rocks of the Gympie Block are very similar to those of the Bowen Basin terranes within the northernmost part of the northern New England Fold Belt. We interpret the Gympie Block as a fragment of the Bowen Basin terranes from the northern part of the fold belt that has been displaced southward into its present position during the latter part of the Triassic thrust contraction (Holcombe et al. 1997). DISCUSSION Connors-Auburn-Urannah-Camboon calc-alkaline suite: not an Early Carboniferous magmatic arc A major constraint of previous tectonic interpretations of the New England Fold Belt has been the accepted existence of a complete convergent-margin architecture in the northern New England Fold Belt in the Early to mid-Carboniferous, with a well-defined accretionary complex, forearc basin (Yarrol Basin), and magmatic arc system (Connors—Auburn Volcanic Arc). The Camboon Volcanics are widely considered to represent a superimposed, Early Permian subduction-related episode, although any associated forearc basin or accretionary wedge elements have not been recognised. Our field studies (in both the Marlborough and Cracow areas), in combination with reinterpreted and new age data (Gust et al. 1993; C. M. Allen pers. comm. 1996) suggest that both of these volcanic episodes are part of a broad, Late Carboniferous through Early Permian magmatic event. Furthermore, we now question whether any arc-signature volcanic rocks can be recognised that relate to the interpreted Early Carboniferous forearc elements. The concept of the Connors-Auburn Volcanic Arc was introduced by Day et al. (1978) who suggested the existence of an Andean-style volcanic arc along the site of the present Connors-Auburn Arch. The Yarrol Province was interpreted as an unstable forearc basin, with a complementary accretionary complex (Wandilla Terrane) originally attributed by Day et al. (1978) to be slope and abyssal plain deposits. The age of the elements of the subduction complex was best known from the Yarrol Province shelfal sequence, where oolitic limestone was widely developed. A unifying theme in associating the shelf and slope deposits was the presence of detrital oolites in the slope sequences. The age and existence of the volcanic arc was poorly constrained, based on: (i) the presence of volcanics intruded by Late Carboniferous granites and thus broadly interpreted as Devonian/Carboniferous in age (Malone et al. 1966; Jensen et al. 1966), within the Connors Arch; and (ii) a single 343 Ma K/Ar determination from an outlier of felsic volcanics in the southern Auburn Arch (Whitaker et al. 1974) correlated with the Torsdale beds of the Auburn Arch. Volcanism was interpreted to have ceased during the Late Carboniferous, coincident with the intrusion of voluminous granite batholiths, and sedimentation limited to relatively quartz-rich sediments on the shelf

PERMO-CARBONIFEROUS EVOLUTION, QLD

75

and slope. The principal metamorphism in the accretionary complex rocks was attributed a Late Carboniferous age associated with an orogeny that varied widely in age and intensity. During the Early Permian, the Camboon Volcanic Arc was interpreted to have developed over the prior Connors-Auburn Volcanic Arc, although there was no identified corresponding accretionary complex. Dear (1994) described three sequences (his cycles 1 3) of predominantly felsic volcanic and volcaniclastic rocks in the southern Connors Arch, the lowest two of which are intruded by 316-287 Ma granites and are thus at least early Late Carboniferous in age. Dear correlated his Cycle 2 rocks with the Torsdale beds on the basis of the intrusive ages, and his Cycle 3, which unconformably overlies a 316 Ma granite, with the Camboon Volcanics. The 312±4Ma whole-rock date from the Torsdale beds (Jones et al. 1996; Table 1) and unpublished palaeomagnetic data (CSIRO report to Queensland Metals Corporation) suggest that the oldest sequences mapped by Dear are no older than midCarboniferous in age. Recent SHRIMP dating of zircons from granites of the Urannah Suite at the northern end of the Connors Arch (Allen 1994; C. M. Allen pers. comm. 1996) suggests that the main phase of activity is about 305 Ma, and that batholith development ceased by about 286 Ma. Thus, although we recognise that there are significant internal unconformities within the volcanics, we consider this entire suite, comprising Torsdale beds, Connors Volcanics, Auburn Complex, Urannah Complex, and Camboon Volcanics, to be part of a broad magmatic system. We propose that these elements be collectively called the Connors-Camboon Province, to differentiate it from the structural entity that is the Connors-Auburn Arch. This episode spanned about 40 million years, peaking at 305 Ma and ranging from the early Late Carboniferous, at ca 320 Ma, and continuing into the Early Permian until ca 280 Ma. The single 343 Ma K-Ar age from the outlier of Torsdale beds is thus the only evidence in the northern New England Fold Belt for an Early Carboniferous magmatic episode that may be coeval with development of the recognised accretionary complex and forearc basin.

granites overprint the older accretionary terranes and must represent a major crustal melting event associated with a mid-crustal deformation. In the North D'Aguilar Block the metamorphic rocks and syntectonic granitoids are exposed in an extensional core complex. Metamorphic culminations containing similar syntectonic granites occur sporadically along the whole length of the orogen, from the Tia and Wongwibinda Complexes of the Late Carboniferous Hillgrove suite in the southern New England Fold Belt in New South Wales, to the Broome Head Metamorphics, and their correlatives described above in the Marlborough Block, at the northern end. We would suggest a common extensional origin for all, although this viewpoint is not widely accepted for the southern New England Fold Belt, where a thrust-related origin has been postulated (Dirks et al. 1993). We have tightly constrained the timing of syntectonic intrusion in the North D'Aguilar Block at ca 305 Ma, and a recent zircon age from the Tia Complex has given a preliminary age of ca 300 Ma (Dirks et al. 1993). The crystallisation ages of the other complexes is currently unknown as most age dates have been derived by K-Ar dating and have been disrupted by more recent tectonic events. We believe that the widespread occurrence of ages at 305-300 Ma indicates a non-diachronous extensional event rather than with oblique ridge subduction and translation of a triple junction southwards along the orogen as has been proposed elsewhere (Murray et al 1987; Fergusson et al. 1993). The age range of the Connors-Camboon Province, ca 320-280 Ma and peaking at ca 305 Ma, is essentially identical to the range of ages recorded from the Bulgonunna Volcanics and associated granites/volcanic complexes to the west of the Bowen Basin and throughout north Queensland (Webb & McDougall 1968; Branch 1972; Oversby et al. 1980; Black et al. 1981; C. M. Allen pers. comm. 1996). C. M. Allen pers. comm. 1996 has correlated the Urannah and Bulgonunna Suites on the basis of similar geochemistry. The development of large-scale calderas, batholith emplacement and ring complexes that characterises these events is characteristic of continental extensional environments and record a significant episode of crustal melting.

Late Carboniferous magmatic belt

The paradox of Late Carboniferous magmatism

Coeval, but contrasting, magmatic suites dominate the Late Carboniferous at ca 305 Ma. The interpretation of Day et al. (1978) that the rocks of the ConnorsCamboon Province represent a continental-margin arc is widely accepted. The igneous rocks of the Province are broadly calc-alkaline in character (Jones 1994; Dear 1994) and Allen (1994) argued that the batholith that underlies these volcanics was similar to other interpreted continental arc batholiths. These data are consistent with the notion that a north-south-trending magmatic arc of Late Carboniferous to Early Permian age developed during this time. At the same time that calc-alkaline magmatic activity peaks in the ConnorsCamboon Province a suite of syntectonic S-type

Several significant factors bear upon any tectonic interpretation for the Connors-Camboon Province: (i) the present linear geometry of the province is largely an artefact of later structuring and basin formation (Holcombe et al. 1997); (ii) active extension and crustal melting occurred in the potential position of any forearc; (iii) widespread continental crustal melting, probably accompanying extension, occurred to the west and north; and (iv) there is no clearly recognised forearc complex rocks that may have developed during this time. The first three points have been discussed above; the fourth point is addressed below. The Neerkol Formation and equivalents, in the Yarrol Province, comprise Late Carboniferous shallow-marine

76 R . J . H O L C O M B E ETAL. sediments that are broadly coeval with both the Connors-Camboon suite and the forearc S-type magmatic terranes. Leitch et al (1994) speculated that the Neerkol Formation could be a Late Carboniferous forearc basin, paired with the Shoalwater accretionary terrane, on the basis that: (i) units correlated with the Neerkol Formation on the coast east of Marlborough have a strong volcaniclastic component; and (ii) the Shoalwater terrane is possibly younger than the other coastal accretionary terranes based on its structural position. Although these sedimentary rocks now occupy a position that could have represented a Late Carboniferous forearc basin, they generally contain only a limited component of volcaniclastic detritus, inconsistent with the presence of a nearby active volcanic province. They are dominated by a shallow-water marine mud facies with widespread, and characteristic, development of bryozoans implying quiet conditions; thus they are not typical of extensional basins that might be expected to develop within a forearc environment. The coastal rocks that Leitch et al (1994) described are highly atypical of the Neerkol Formation regionally and contain no fossils (elsewhere a characteristic of the unit). Interpretation of the Neerkol Formation and equivalents is problematic and may be a key to future tectonic analyses. We recognise two interpretations of these observations dependent on whether the Connors-Camboon Province is interpreted as a continental arc, or whether it is part of a broader province of crustal melting associated with continental extension. (1) If the calc-alkaline rocks represent a magmatic arc, then tectonic scenarios must account for the coeval forearc extension and strong regional thermal perturbation. The implication is that an underlying fertile slab must persist below the arc and yet at the same time, magmatic production related to the downgoing slab must be overtaken by a regional thermal perturbation. In earlier papers (Little et al 1993; Stephens et al 1994) we have suggested that synchronous slab roll-back to the east, with extension driven by the trench-suction force, may have initiated just prior to ca 305 Ma and, during retreat, hot mantle has been introduced beneath this region and induced mantle and crustal melting. The fact that arc magmatism does not migrate eastward with the roll-back requires that melt production from the slab continues in situ and we interpreted this in terms of either a bent slab geometry (as for example in the present Tonga-Kermadec slab) or a detached slab remnant. In either case, the slab must ultimately become detached and sink into the mantle because arc magmatism continues in place and wanes slowly into the Early Permian (Camboon Volcanics time). Note that since we have linked the Early Permian Camboon Volcanics with the Late Carboniferous magmatism, we no longer require a separate discrete Permian arc episode ('Camboon Arc') as is commonly envisaged. (2) If the Connors-Camboon Province is part of a broad extensional event then the isotopic compositions and calc-alkaline signature of the magmatism (C. M. Allen pers. comm. 1996) must be interpreted in terms of

remelting of relatively young crust that is calc-alkaline in composition. This model has been argued for Late Triassic and Early Cretaceous volcanism in the New England Fold Belt (Stephens 1991, Ewart et al 1992; Stephens et al 1993) but requires the existence of earlier calc-alkaline magmatism beneath the region. Contrary to the earlier interpretation of Allen (1994), the Urannah Suite is most likely related to extension at a time when subduction has slowed or stopped (C. M. Allen pers. comm. 1996). Given the regional extensional signature of surrounding Permo-Carboniferous magmatic terrains, and the continuation of extension through the Early Permian (Fielding et al 1997), we are gradually being drawn to the interpretation that the Connorsr-Camboon Province represents an extensional event that is not necessarily related in any respect to active subduction. We regard the position and nature of the plate boundary at this time as uncertain, although the foregoing discussion of the Berserker beds suggests that increasingly marine conditions may have existed progressively to the east. The key to both tectonic and age interpretations of the Connors-Camboon Province lies in further, detailed research into the Connors volcanics. The nature of the driving force for such a broad regional extension is also uncertain. Simple slab retreat, as suggested above, would seem to be insufficient as a mechanism for such a major crustal and thermal perturbation and is unlikely to have occurred synchronously over such a large length of subduction margin. This event, and the ensuing Early Permian extensional period, may mark a major tectonic event occurring along a passive margin. Certainly coeval events of the same magnitude, such as termination of the Alice Springs Orogeny in central Australia, must be considered when determining the fundamental tectonic dynamics. Connors Arch and 'Grantleigh' Trough The Permian sedimentary units exposed to the east of the Connors Arch, within the Gogango Overfolded Zone were previously interpreted as having been deposited in a deep marine basin called the Grantleigh Trough that had opened within the forearc (Malone et al 1966; Kirkegaard et al 1970; Day et al 1983). The Connors Arch has been widely regarded as representing the eastern margin of the depositional Bowen Basin (see Fielding et al 1997 for discussion). Fergusson (1991) put forward a structural model suggesting that the Connors Arch is a structural high developed above an antiformal duplex stack in the basement rocks, and produced during the PermoTriassic thrust event. The depositional Bowen Basin was shown continuous, but deepening, into the Grantleigh Trough. We agree broadly with this view and interpret the 'Connors Arch' northwest of Rockhampton as the exposed core of a simple antiformal flexure that is the westernmost fold in a series of low-amplitude regional folds extending to the east (Apis Creek Syncline/Strathmuir Synclinorium).

PERMO-CARBONIFEROUS EVOLUTION, QLD Fielding et al (1994, 1997) argued on sedimentological and structural grounds that there is no need to invoke a separate, or deep-water, sedimentary basin in order to explain the observed regional stratigraphic variations, and indeed no evidence to support such an interpretation. A critical aspect of interpretations invoking the existence of the Grantleigh Trough has been the nature and age of the R o o k w o o d Volcanics, addressed in this paper, and the internally, structurally c o m p l e x R a n n e s beds. T h e t h i c k n e s s of s e d i m e n t originally estimated by Kirkegaard et al (1970) for the R a n n e s beds (3000 m ) is grossly exaggerated by s t r u c t u r a l t h i c k e n i n g ( H o l c o m b e et al 1997). In addition, the depositional characteristics of the lessd e f o r m e d R a n n e s b e d s are identical to t h o s e of sediments in the adjacent B o w e n Basin west of the G o g a n g o Overfolded Zone. The sedimentary rocks of the Rannes beds are clearly not deep-water turbidites as stated by a variety of authors (Kirkegaard et al. 1970; D a y et al 1983; Fergusson 1991), but are t e c h n i c a l l y deformed shallow-marine sediments. They are widely, and i n a p p r o p r i a t e l y , d e s c r i b e d in the literature as 'flysch'.

ACKNOWLEDGMENTS This paper incorporates the results of several independent A R C - f u n d e d projects ( A 3 8 8 3 0 0 4 1 ; A 3 9 1 3 0 2 7 9 ; A 3 9 3 3 1 3 6 6 ; A 3 9 2 3 2 3 3 8 ) over the past ten years. W e are indebted to the m a n y students who have contributed as part of the various projects. In particular, aspects of the P h D w o r k of Terry Harbort, A m a n d a Jones, and Paul M e s s e n g e r and of the H o n o u r s w o r k of Terry Harbort, Mac Denton, Vanessa Muscio, Murray Patterson, Shane O ' C o n n e l l , and Eris O ' B r i e n have added to our understanding of the region. Discussions over the years with Cec Murray, Russell Korsch, Chris Fergusson, John Draper, and others have refined our ideas. In a review paper like this it is likely that w e have m i s s e d p r o p e r a c k n o w l e d g m e n t of s o m e sources. C o n s t r u c t i v e r e v i e w s b y V i n c e M o r a n d and Peter Cawood have considerably strengthened the paper.

REFERENCES

77

BAKER J. C., FIELDING C. R . , DE CARITATP. & WILKINSON M .

M. 1993. Permian evolution of sandstone composition in a complex back-arc extensional to foreland basin: the Bowen Basin, eastern Australia. Journal of Sedimentary Petrology 63, 881-893. BRANCH C. D. 1966. Volcanic cauldrons, ring complexes, and associated granites of the Georgetown Inlier. Queensland. Bureau of Mineral Resources Bulletin 76. BRIGGS D. J. C. 1993. Time control in the Permian of the New England Orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 283—292. University of New England, Armidale. CASLEY Z. W. D. 1993. Geology and environmental geochemistry of the Agricola Gold Mine, Southeast Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). CAWOOD P. A. & LEITCH E. C. 1985. Accretion and dispersal

tectonics of the southern New England Fold Belt, eastern Australia. In: Howell D. G. ed. Tectonostratigraphic terranes of the Circum-Pacific Region, pp. 481—492. Circum-Pacific Council for Energy and Mineral Resources, Earth Science Series 1. CUNEEN R. 1994. New Insights into the Gympie Goldfield. In: Queensland Exploration Potential 1994, Symposium Abstracts, p. 26. Queensland Department of Minerals and Energy, Brisbane. DAY R. W . , MURRAY C . G . & WHITAKER W . G . 1 9 7 8 . T h e

eastern part of the Tasman Orogenic Zone. In: Scheibner E. ed. The Phanerozoic structure of Australia and variations in tectonic style, pp. 3 2 7 - 3 6 4 . Tectonophysics 48. DAY R . W . , WHITAKER W . G . , MURRAY C . G . , WILSON I. H .

& GRIMES K. G. 1983. Queensland Geology. Geological Survey of Queensland Publication 383. DEAR J. F. 1994. Major cycles of volcanism in the southern Connors Arch. In: Holcombe R. J. et al eds. 1994 Field Conference Guidebook: Capricorn Region, central coastal Queensland, pp. 31-45. Geological Society of Australia, Queensland Division. DEAR J. F., M C K E L L A R R . G . & TUCKER R . M . 1971. G e o l o g y

of the Monto 1:250 000 sheet area. Geological Survey of Queensland Report 46. DIRKS P. H. G . M . , OFFLER R. & COLLINS W . J. 1993. T i m i n g

of emplacement and deformation of the Tia Granodiorite, southern New England Fold Belt, NSW: implications for the metamorphic history. Australian Journal of Earth Sciences 40, 103-108. DONCHAKP. J. T . , LITTLE T . A . , SLIWA R . & HOLCOMBE R . J.

AITCHISON J. C. 1988. Early Carboniferous (Tournaisian) Radiolaria from the Neranleigh-Fernvale beds, Lake Manchester, Queensland, Australia. Queensland Government Mining Journal 89, 240-241. ALLEN C . M . , CHAPPEL B. W . , WOODEN J. L. & WILLIAMS

I. S. 1994. Ages, compositions and sources of the Urannah batholith. In: Henderson R. A. & Davis B. K. eds. Extended conference abstracts : new developments in geology and metallogeny: Northern Tasman orogenic zone, pp. 107-108. James Cook University, Townsville, Economic Geology Research Unit Contribution 50. ARTHUR G. C. 1969. Granitic and basaltic intrusions and regional metamorphic rocks from the Chapinga district southeast Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.).

1995. Geology of metamorphic units of the North D'Aguilar Block-Goomeri, Nannago and Nambour 1:100 000 sheet areas. Queensland Geological Record 1995/7. DRAPER J. J. & BEESTON J. W. 1985. Depositional aspects of

the Reids Dome Beds, Denison Trough. Government Mining Journal 86, 200-210. EWART A . , SCHON R . W . & CHAPPELL B . W .

Queensland 1992.

The

Cretaceous volcanic- plutonic province of the central Queensland (Australia) coast—a rift-related 'calc-alkaline' province. Transactions of the Royal Society of Edinburgh 83, 3 2 7 - 3 4 6 .

FERGUSSON C. L. 1991. Thin-skinned thrusting in the northern New England Orogen, central Queensland, Australia. Tectonics 10, 797-806.

78

R. J. H O L C O M B E E T A L .

FERGUSSON C . L . , HENDERSON R . A . & LEITCH E . C . 1 9 9 0 .

HOLCOMBE R . J. & LITTLE T . A . 1 9 9 4 . B l u e s c h i s t s o f t h e

Subduction complex melange, Palaeozoic New England Orogen, central Queensland, Australia. Journal of Structural Geology 12, 591-599.

North D'Aguilar Block: structural development of the Rocksberg Greenstone and associated units near Mt Mee, southeast Queensland. Australian Journal of Earth Sciences 41, 115-130.

FERGUSSON C . L., HENDERSON R . A . , LEITCH E. C . & ISHIGA

H., 1993. Lithology and structure of the Wandilla terrane, Gladstone-Yeppoon district, central Queensland, and an overview of the Palaeozoic subduction complex of the New England Fold Belt. Australian Journal of Earth Sciences 40, 403-414. FIELDING C . R . , FALKNER A . J., KASSAN J. & DRAPER J. J.

1990. Permian and Triassic depositional systems in the Bowen Basin. In: Bowen Basin Symposium 1990, Proceedings, pp. 21-25. Geological Society of Australia, Queensland Division. FIELDING C . R . , STEPHENS C . J., & HOLCOMBE R . J. 1 9 9 4 . A

critical evaluation of the Grantleigh Trough, east-central Queensland. In: Holcombe R. J. et al. eds. 1994 Field Conference Guidebook: Capricorn Region, central coastal Queensland, pp. 17—30. Geological Society of Australia, Queensland Division. FIELDING C . R . , STEPHENS C . J. & HOLCOMBE R . J.

1997.

Permian stratigraphy and palaeogeography of the eastern Bowen Basin, Gogango Overfolded Zone and Strathmuir Synclinorium in the Rockhampton—Mackay region, central Queensland. Geological Society of Australia Special Publication

19, 80-95.

FIELDING C . R . , STEPHENS C . J., KASSAN J. & HOLCOMBE R . J.

1995. Revised palaeogeographic maps for the Bowen Basin, Central Queensland. In: Follington I. L., Beeston J. W. & Hamilton L. H. eds. Bowen Basin Symposium 1995, Proceedings, pp. 7—15. Geological Society of Australia, Queensland Division. FLEMING P. J. G . , MURRAY C . G . & WHITAKER W . G . 1 9 7 5 .

Late Palaeozoic invertebrate fossils in the Wandilla Formation and the deposition of the Curtis Island Group. Queensland Government Mining Journal 76, 416-422.

GRAYSON R. 1995. The Avoca Creek Granodiorite, Southeast Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). HARRINGTON H. J. 1974. The Tasman Geosyncline in Australia. In: Denmead A. K., Tweedale G. W. & Wilson A. K. eds. The Tasman Geosyncline—-A symposium, pp. 383-407. Geological Society of Australia, Queensland Division. HARRINGTON H. J. 1983. Correlation of the Permian and Triassic Gympie terrane of Queensland with the Brook Street and Maitai terranes of New Zealand. In: Permian Geology of Queensland, pp. 431-436. Geological Society of Australia, Queensland Division. HENDERSON R . A . , FERGUSSON C . L., LEITCH E. C . , MORAND V . J., REINHARDT J. J. & CARR P. F. 1993. T e c t o n i c s o f t h e

northern New England Fold Belt. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 505—515. Department of Geology and Geophysics, University of New England, Armidale. HILL D., PLAYFORD G., & WOODS J. T. 1972. Permian

Fossils

of Queensland. Queensland Palaeontographical Society Publication. HOLCOMBE R . J., FIELDING C . R . , & STEPHENS C . J. 1 9 9 5 .

Tear fault termination of the fold-thrust belt in the Northern New England Orogen. Geological Society of Australia Abstracts 40, 72.

HOLCOMBE R . J., STEPHENS C . J., FIELDING C . R . ETAL.

1997.

Tectonic evolution of the northern New England Fold Belt: the Permian-Triassic Hunter-Bowen event. Geological Society of Australia Special Publication 19, 52-65. HUNNS S. R. 1994. Geology of the Mount Chalmers volcanichosted massive sulphide, and implications for its formation. In: Holcombe R. J. et al. eds. 1994 Field Conference Guidebook: Capricorn Region, central coastal Queensland, pp. 80-92. Geological Society of Australia, Queensland Division. ISHIGA H. 1990. Radiolarians from the Gympie province. In: Proceedings Pacific Rim Congress 90 III, pp. 187-189. Australasian Institute of Mining and Metallurgy, Melbourne. JENSEN A . R . , GREGORY

C . M . & FORBES V . R .

1966.

Geology of the Mackay 1:250 000 Sheet area Queensland. Bureau of Mineral Resources Report 104. JOHNSON

S.

E. &

HENDERSON

R. A.

1991.

Tectonic

development of the Drummond Basin, eastern Australia; backarc extension and inversion in a late Palaeozoic active margin setting. Basin Research 3 , 197-213.

JONES J. A. 1994. Character and setting of the Camboon Volcanics Central Queensland: implications for the Early Permian tectonics of the New England Fold Belt. Geological Society of Australia Abstracts 37, 204. JONES J. A . , STEPHENS C . J. & EWART A . 1 9 9 6 . C o n s t r a i n t s o n

the Early Permian and Late Carboniferous of the northern New England Fold Belt from the Camboon Volcanics and the Torsdale Beds. Geological Society of Australia Abstracts 41, 222. JONES P. J. (Compiler). 1996. AGSO Phanerozoic 1995. Oxford University Press, Melbourne.

Timescale

KIRKEGAARD A . G . , SHAW R . D . & MURRAY C . G .

1970.

Geology of the Rockhampton and Port Clinton 1:250 000 sheet areas, Queensland. Geological Survey of Queensland Report 38. KORSCH R . J., O ' B R I E N P. E., SEXTON M . J., WAKE-DYSTER

K. D. & WELLS A. T. 1989. Development of Mesozoic transtensional basins in easternmost Australia. Australian Journal of Earth Sciences 36, 13—28. KWIECIEN W. 1996. Geology, geochemistry, and petrogenesis of the Monsildale Granodiorite. MSc thesis, Queensland University of Technology, Brisbane (unpubl.). LEE J. & LISTER G. S. 1992. Late Miocene ductile extension and detachment faulting, Mykonos, Greece. Geology 20, 121-124. LEITCH E. C. 1988. The Barnard Basin and the early Permian development of the southern part of the New England Fold Belt. In: Kleeman J. D. ed. New England Orogen— Tectonics and Metallogenesis, pp. 61-67. Department of Geology and Geophysics, University of New England, Armidale. LEITCH E . C . , FERGUSSON

C . L . , HENDERSON R .

A.

&

MORAND V. J. 1994. Late Palaeozoic arc flank and fore-arc basin sequence of the New England Fold Belt in the Stanage Bay region, central Queensland. Australian Journal of Earth Sciences 41, 301-310.

PERMO-CARBONIFEROUS EVOLUTION, QLD LEITCH E. C., MORAND V . J., FERGUSSON C. L., HENDERSON

R. A. & CARR P. F. 1993. Accretion and post-accretion metamorphism in subduction complex terranes of the New England fold belt, eastern Australia. Journal of Metamorphic Geology 11, 309-318. LITTLE T. A. 1992. Geology of a northern part of the North D'Aguilar block, southeast Queensland. Queensland Department of Resource Industries Record 1992/22. LITTLE T . A . , HOLCOMBE R. J., GIBSON G . M . , OFFLER R . , G A N S P . B. & MCWILLIAMS M . O. 1992. E x h u m a t i o n o f

late Paleozoic blueschists in Queensland, Australia, by extensional faulting. Geology 20, 231-234. LITTLE T . A., HOLCOMBE R. J. & SLIWA R . 1993. S t r u c t u r a l

evidence for extensional exhumation of blueschist-bearing serpentinite-matrix melange, New England Orogen, southeast Queensland, Australia. Tectonics 12, 536-549. LITTLE T . A., MCWILLIAMS M . O . & HOLCOMBE R. J. 1995. 40

Ar/ 39 Ar thermochronology of epidote-blueschists from the North D'Aguilar Block, Queensland, Australia: timing and kinematics of subduction complex unroofing. Geological Society of America Bulletin 107, 520-535

MALONE E. J., OLGERS F. & KIRKEGAARD A . G . 1969. T h e

geology of the Duaringa and Saint Lawrence 1:250 000 sheet areas, Queensland. Bureau of Mineral Resources Report 121. MORAND V. J. 1993. The Broome Head Metamorphics: high grade metamorphism in the northern New England Fold Orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 591-598. Department of Geology and Geophysics, University of New England, Armidale. MURPHY P. R . , SCHWARZBOCK H . , CRANFIELD L . C . , WITHNALL I. W . & MURRAY C. G . 1976. G e o l o g y o f t h e

Gympie 1:250 000 Sheet area. Geological Queensland Report 96.

Survey

of

MURRAY C. G. 1974. Alpine-type ultramafics in the northern part of the Tasman Geosyncline—Possible remnants of Palaeozoic ocean floor. In: Denmead A. K., Tweedale G. W. & Wilson A. F. eds. The Tasman Geosyncline—a Symposium, pp. 161-181. Geological Society of Australia, Queensland Division. MURRAY C . G., FERGUSSON C. L., FLOOD P. G., WHITAKER W .

G. & KORSCH R. J. 1987. Plate tectonic model for the Carboniferous evolution of the New England Fold Belt. Australian Journal of Earth Sciences 34, 213-236. MURRAY C. G . , MCCLUNG G . R . & WHITAKER W . G . 1979.

Early Permian fossils from the Amamoor beds and the age of the Booloumba beds—a new stratigraphic unit of the North D'Aguilar Block. Queensland Government Mining Journal 80,71-78. MUSCIO V. 1994. Structural Analysis of the Nob Creek region, Marlborough Block, central Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). O'BRIEN E. 1994. Volcanic character of the 'Langdale Inlier' southern Connors Arch, Marlborough district, central Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). O'CONNELL S. A. 1995. Geology and structure of the Aeroview area, Rockhampton region, central Queensland: implications for mineralisation of the lower Permian Rookwood Volcanics. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.).

ROBERTS C. A. 1992. The geology of the Fat Hen Creek area, Kilkivan, Southeast Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). SAINTY R. A. 1992. Shallow-water stratigraphy at the Mount Chalmers volcanic-hosted massive sulfide deposit, Queensland, Australia. Economic Geology 87, 812-824. SIVELL W. J. & WATERHOUSE J. B. 1987. Volcanic rocks of

the Highbury Volcanics and Rammutt Formation— remnant portion of an Early Permian intra-oceanic arc. In: Murray C. G. & Waterhouse J. B. eds. 1987 Field Conference—Gympie District, pp. 34-43. Geological Society of Australia, Queensland Division. SLIWA R. 1994. Regional structural geology of the central North D'Aguilar Block, southeastern Queensland. PhD thesis, University of Queensland, Brisbane (unpubl.). STEPHENS C. J. 1991. The Mungore Cauldron and Gayndah Centre: Late Triassic large-scale silicic volcanism in the New England Fold Belt near Gayndah, southeast Queensland. PhD thesis, University of Queensland, Brisbane (unpubl.). STEPHENS S. J., EWART A., FIELDING C. R., HOLCOMBE R. J. &

SCHON R. W. 1994. Location and timing of convergent margin volcanism during the Late Palaeozoic-Mesozoic evolution of Eastern Australia. Geological Society of Australia Abstracts 37, 417. VIRISCHEFF A. S. 1974. Petrology of the metamorphic and granitic rocks of Glenrock near Kumbia, south east Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). WARD P. L. 1995. Subduction cycles under western North America during the Mesozoic and Cainozoic eras. In: Miller D. M. & Busby C. eds. Jurassic Magmatism and Tectonics of the North American Cordillera, pp. 1-46. Geological Society of America Special Paper 299. WATERHOUSE J. B. & SIVELL W. J. 1987a. Trans-Tasman

relationships between the Permian—Triassic rocks of Gympie, southeast Queensland, and those of New Caledonia and New Zealand. In: Murray C. G & Waterhouse J. B. eds. 1987 Field Conference Guide, Gympie District, pp. 48-59. Geological Society of Australia, Queensland Division. WATERHOUSE J. B. & SIVELL W. J. 1987b. Permian evidence for trans-Tasmanian relationships between east Australia, New Caledonia and New Zealand. Tectonophysics 142, 227-240. WEBB A . W . & MCDOUGALL I. 1968. T h e g e o c h r o n o l o g y o f

the igneous rocks of eastern Queensland. Journal of the Geological Society of Australia 15, 313—346. WHITAKER W . G., MURPHY P. R . & ROLLASON R . G . 1974.

Geology of the Mundubbera 1:250000 sheet area. Geological Survey of Queensland Report 84. WIJBRANS J. R . , VAN W E E S J. D . , STEPHENSON R . A .

volcanism in

northeastern

&

CLOETINGH S. A. P. L. 1993. Pressure-temperature-time

evolution of the high-pressure metamorphic complex of Sifnos, Greece. Geology 93, 443-446. WILLMOTT W . F., O'FLYNN M . L. & TREZISE D. L. 1986.

Rockhampton Region, Queensland. 1:100 000 Geological Map Commentary. Geological Survey of Queensland.

OVERSBY B. S., BLACK L. P. & SHERATON J. W . 1980. L a t e

Palaeozoic continental

79

Queensland. In: Henderson R. A. & Stephenson P. J. eds. The Geology and Geophysics of northeastern Australia, pp. 247-268. Geological Society of Australia, Queensland Division.

(Received 24 April 1996; accepted 9 January 1997)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 80-95.

Permian stratigraphy and palaeogeography of the eastern Bowen Basin, Gogango Overfolded Zone and Strathmuir Synclinorium in the Rockhamptonh-Mackay region, central Queensland C. R. FIELDING, C. J. STEPHENS AND R. J. HOLCOMBE Department of Earth Sciences, University of Queensland, Qld 4072, Australia.

Variably deformed Permian sedimentary rocks are extensively preserved in the northernmost New England Fold Belt of coastal central Queensland. Publications arising from joint Bureau of Mineral Resources — Geological Survey of Queensland mapping of the area suggested that the Bowen Basin originally covered much of what is now referred to as the northernmost New England Fold Belt, with Permian sediment accumulating in a series of discrete but interconnected north to north-northwest-elongate depocentres. Subsequent work, however, developed a model in which Permian sediments of the Gogango Overfolded Zone (in the New England Fold Belt) accumulated in a deep-marine basin (the Grantleigh Trough) that was separated from the Bowen Basin to the west by a volcanically active, physiographic barrier along the line of the Connors—Auburn Arch. According to this model, the Connors-Auburn Arch was the site of a continental volcanic arc (the Camboon Volcanic Arc) which was active during Early Permian, basin-forming times. Stratigraphic and sedimentological data presented herein suggest that the Connors-Auburn Arch did not form a basin-margin physiographic feature during accumulation of the Bowen Basin succession, and that Permian strata now incorporated in the New England Fold Belt were continuous with the Bowen Basin. Within correlative formations, facies assemblages are identical on either side of the Connors-Auburn Arch, and no evidence of basin-margin facies is found adjacent to that basement feature. Palaeocurrent data collected from a variety of stratigraphic levels and geographic locations indicate sediment dispersal consistently westward from a source at least as far east as the present coastline, across the ConnorsAuburn Arch and into the Bowen Basin. Therefore, the deep-marine Grantleigh Trough does not exist and it is proposed that use of this term be discontinued. Key words: Bowen Basin, New England Fold Belt, Permian, stratigraphy.

INTRODUCTION Historically, the interpretation of variably deformed Permian and Triassic rocks preserved in the New England Fold Belt of coastal central Queensland (Figure 1) has been controversial, and several critical issues remain unresolved by previous work. The area, which lies to the east of the structural Bowen Basin (Folded Zone and Nebo Synclinorium), includes northnorthwest-elongated outcrop belts of (i) inferred Devonian to Lower Permian volcanic and intrusive rocks (Connors and Auburn Arch: Malone et al. 1966; Dear 1994); (ii) Permian sedimentary rocks (Gogango Overfolded Zone: Malone et al. 1969; Dickins & Malone 1973); (iii) Devonian to Permian sedimentary and volcanic rocks (Yarrol and Calliope Blocks: Kirkegaard et al. 1970); (iv) ophiolitic and metamorphic rocks of uncertain age (Marlborough Block: Kirkegaard et al. 1970; Murray 1974); and (v) Devonian to Carboniferous metasedimentary rocks (Wandilla or Coastal Block: Fergusson et al. 1993). Most of these domains are bounded by faults that separate rocks of disparate age and character developed during the Permo-Triassic contractional, Hunter-Bowen event (Holcombe et al. 1997b).

Contrasting interpretations of the Connors Arch basement ridge and adjacent Gogango Overfolded Zone, and their relationship to the Bowen Basin to the west (Figure 1), have been published. In contrast to earlier work by Malone, Dickins and others, Day et al. (1978) proposed that the Connors Arch was the site of a continental volcanic arc that formed the eastern margin of the Bowen Basin during Early Permian times and that sedimentary rocks of the Gogango Overfolded Zone formed in a separate, deep-water basin, the 'Grantleigh Trough' (Kirkegaard et al. 1970). As will be discussed in more detail below, this view remains popular despite significant contradictory evidence. In this paper, we report new stratigraphic and sedimentological data which complement the structural analysis summarised in two companion papers by Holcombe et al. (1997a, b). These data indicate that the Connors Arch was not a physiographic barrier during accumulation of the Bowen Basin succession, that the rocks of the Gogango Overfolded Zone accumulated within a contiguous Bowen Basin, and that the entire area achieved its present configuration following crustal shortening in the latest Permian to early Late Triassic (the Hunter-Bowen Event sensu Holcombe et al. 1997b). We suggest that there is no

PERMIAN, NORTHERN NEW ENGL AND FOLD BELT

evidence to support the notion of a Permian volcanic arc along the Connors Arch, active during formation of the Bowen Basin, nor of the existence of a deep marine trough in the position of the so-called Grantleigh Trough. We propose that use of the term 'Grantleigh Trough' be discontinued. DEFINITION OF THE EASTERN BOWEN BASIN MARGIN Early investigations in the Bowen Basin (e.g. in the Collinsville area: Reid 1929, 1930) identified that, at least locally, the eastern margin of the basin is a structural margin coincident with the western edge of an uplifted igneous complex (the Connors Arch of Malone et al 1966; Connors-Auburn Arch of Day et al 1978). The geological map of Queensland (Geological Survey of Queensland 1953) applied the earlier terminology of Lower Bowen Volcanics used by Jack and Etheridge (1892) to all of the volcano-sedimentary sequences exposed at the base of the Permian succession within

Figure 1 Major structural elements in the eastern Bowen Basin and adjacent New England Fold Belt of coastal central Queensland. The 'Grantleigh Trough' of Kirkegaard et al. (1970) covers the area shown as Gogango Overfolded Zone.

81 the basin and extending across the Connors Arch, although most of the Permian rocks within the New England Fold Belt, or Eungella-Cracow Mobile Belt as it was then known, were identified as undifferentiated Upper Palaeozoic. The early history of the Bowen Basin was summarised by Malone (1964) and Dickins and Malone (1973), drawing on the results of 1:250 000-scale mapping (Malone et al 1964, 1966, 1969; Jensen et al 1966). They showed a series of palaeogeographic maps, based principally on a stratigraphic analysis, in which the basin developed in terms of an early Early Permian series of volcanic provinces and volcano-sedimentary sub-basins (the Reids Dome beds in the southwest, the Lizzie Creek Volcanics and Carmila beds in the north, and the Camboon Andesite in the southeast) overlain by marine sedimentary rocks (Back Creek Group) that were deposited across the southern Connors Arch and farther east than the present coastline. This interpretation recognised that the original area of sedimentation extended outside the present structure that is the Bowen Basin.

82

C. R. FIELDING ETAL.

Dickins and Malone (1973), nonetheless, expressed difficulty in resolving the stratigraphic association of formations that overlay the Connors-Auburn Arch due to the effects of deformation in the Gogango Overfolded Zone, and gave a confused assessment of all elements in the southeastern part of the basin. It appears that they regarded the Rannes beds as a composite stratigraphic package that could contain rocks ranging in age from Siluro-Devonian through to pre-Late Permian, broadly characterising them as 'Lower Permian and older'. They described the Rannes beds as partly equivalent to, but predominantly younger than, the Camboon Volcanics, a relationship supported by Kirkegaard et al. (1970). Similarities in lithology to sediments of the Back Creek Group and Boomer Formation were noted and it appears that they regarded local in-faulting as a possible cause for the inclusion of later Permian strata in the Rannes beds. Uncertainty as to the stratigraphic significance of the Rannes beds also led Dickins and Malone (1973) to difficulties in the interpretation of sequences preserved east of the Bowen Basin in the New England Fold Belt. Correlation of the Youlambie Conglomerate with the Carmila beds or lower Back Creek Group was suggested on fossil evidence, although they were unable to determine whether the Youlambie Conglomerate formed in an eastern basin that was continuous with, or separated from, the Carmila beds or Back Creek Group. In their summary, they did not develop a model that allowed for deposition outside the preserved areas of exposure of all of the stratigraphic units. Kirkegaard et al. (1970) regarded the Rannes beds as a discrete formation deposited within the Grantleigh Trough, a downwarp that included the terrestrial Camboon Volcanics in the west, and marine rocks of the Rannes beds and Rookwood Volcanics. In a general sense, Dickins and Malone (1973) followed this interpretation with their early sub-basin model. They suggested that this region may have been deformed prior to the deposition of Late Permian marine sediments (upper Back Creek Group), whilst Kirkegaard et al. (1970) suggested uplift possibly accompanied by deformation prior to accumulation of the Moah Creek beds, which they equated 'at least in part' with the Boomer Formation. Day et al. (1978), in developing a model for the tectonic evolution of the Tasman Fold Belt, showed an initial downwarp developing over the Bowen Basin during the Late Carboniferous to Early Permian (the Combarngo Volcanics and associated continental sediments: Day et al. 1983) followed by Early to midPermian development of the Bowen Basin 'to the west of the Camboon Volcanic Arc'. The Camboon Volcanic Arc was interpreted to have developed along the site of the present Connors-Auburn Arch. They specified the volcanics along the western side of the Connors Arch as forming the basal sequence of the Bowen Basin, and the Grantleigh Trough as a deep trough east of the arc. The model of Day et al. (1978) marks a significant change from the earlier history outlined by Dickins and Malone (1973) in that: (i) volcanism during the early history of the basin is sourced from along the Connors-

Auburn Arch, rather than from within the troughs as envisaged by Dickins and Malone, implying the presence of a significant constructive edifice that should reasonably have supplied abundant primary and reworked volcanic detritus to the basin; the concept of the arc was also developed as a feature related to westwarddirected subduction, a concept not embodied in any of the previous literature; (ii) volcanics cropping out along the western flanks of the Connors and Auburn Arches were described as the 'basal sequence of the Bowen Basin' whilst volcanic and volcaniclastic rocks east of the Connors Arch were not related to the basin, leaving the resultant impression that the eastern margin of the basin was coincident with the western margin of the Connors and Auburn Arches; and (iii) the concept of the Grantleigh Trough as a separate elongate basin formed east of, and coevally with, the arc was further developed. The impression as to the limited extent of the basin was reinforced by the absence of any further discussion of basin development. Day et al. (1983) later stated that the main marine depositional trough of the basin was along the presently defined northeastern structural margin, associating the sediment accumulation with local formation names (Gebbie and Blenheim Subgroups). They also emphasised the distinction of the Grantleigh Trough as a deep-water trough containing 'flysch-type sediments' and spilitic pillow basalts 'with gross features typical of ocean floor basalts'. Subsequent interpretations of the geological evolution of the eastern Bowen Basin and northern New England Fold Belt (Henderson 1980; Harrington & Korsch 1985; Murray et al. 1987; Hammond 1987; Fergusson & Leitch 1993) have not significantly challenged the palaeogeographic implications of the Day et al. model for the Permian. Fergusson (1991) developed the first well-constrained model for the crustal architecture beneath the southern Connors Arch as a structural culmination developed during the latest Permian to Middle Triassic contractional deformation that closed the basin (the Hunter— Bowen Orogeny). Our research suggests, in addition, that the north-northwest trend of the Connors Arch north of Rockhampton may be a consequence of westsouthwest-directed thrusting which, as well as producing the intense deformation within the Gogango Overfolded Zone, may have imposed a regional structural fabric that does not reflect the original orientation of the basin margin. Sedimentological data presented here indicate that, at least until the latest Permian, the present eastern limit of exposed basin rocks should not be interpreted in terms of a basin margin (Fielding et al. 1994; in general agreement with Malone 1964; Dickins and Malone et al. 1973). Stratigraphic units can be correlated over at least 350 km south to north from the southeast Bowen Basin into the Gogango Overfolded Zone, and from west to east across and around the basement inliers of the Connors and Auburn Arches. Facies within stratigraphically equivalent formations on either side of the Connors Arch, particularly in the middle Permian marine units, are identical and do not indicate a basin-

PERMIAN, NORTHERN NEW ENGLAND FOLD BELT margin location. Palaeocurrent data gathered from these units indicate similar (westward) sediment dispersal directions on both sides of the arch. On this basis we support the earlier conclusion of Reid, Malone, Dickins and others that during the Permian period, the Bowen Basin extended for some distance to the east of the present position of the Connors Arch. Given the intensity and nature of deformation associated with closure of the basin during the Hunter-Bowen Orogeny, the original margin may be only locally preserved within onshore Queensland and its position remains speculative. PERMIAN STRATIGRAPHY

The major Permian stratigraphic units encountered in the Strathmuir Synclinorium and northern Gogango Overfolded Zone can be correlated with lithologically similar units of equivalent age in the eastern Bowen Basin (Figure 2). North-south correlations along the present eastern edge of the Bowen Basin and into the Gogango Overfolded Zone are illustrated in Figure 3.

w Bowen Basin Folded Zone Rewan Fm.

83

Detailed logs of measured sections through parts of these formations are shown in Figure 4 as representative illustrations of facies. In an earlier paper (Fielding et al. 1995), the Bowen Basin stratigraphy was divided into three gross packages, interpreted in terms of Early Permian crustal extension, mid Permian passive thermal subsidence, and Late Permian to Middle Triassic, foreland thrust loadinduced subsidence. Each phase of basin-forming activity has left a distinctive record in terms of gross cross-sectional stratal geometry, sediment composition (petrofacies), lithofacies assemblage and interpreted sediment dispersal patterns. The Permian stratigraphy of the northernmost New England Fold Belt can also be reconciled to this model. The characteristics of the various formations are discussed in more detail below, in ascending stratigraphic order. Throughout this paper, correlations used are consistent with those of Draper et al. (1990), with reference to additional biostratigraphic data of Dickins and Malone (1973) and Briggs (1993). The absolute time-frame is that of Jones (1996).

Connors- Strathmuir Synclinorium Gogango Carmila Syncline Overfolded/ Crai 9 ilee Auburn Block Arch Zone

Ma 250

Dinner Creek Conglomerate

Baralaba Coal Measures

Tr Tatarian

260

Kazanian Ufimian 270Kungurian 280 Artinskian Carmila Beds/ Lizzie Creek Volcanics (sedimentary . , package) , \ , volcanic / I package^ \X v v v/v

290

x

v

v

volcan[c package H

v

v

Neerkol j=

x N

m

Sakmarian Asselian

300

c

Figure 2 East-west time-space diagram illustrating the time equivalence of Permian stratigraphic units from the eastern Bowen Basin across the Strathmuir Synclinorium, Gogango Overfolded Zone and Craigilee Block. Time-scale is that of Jones (1996). Notation adjacent to time-scale is: 1, Early Permian extension; 2, mid-Permian passive thermal subsidence; 3, Late Permian to Middle Triassic foreland thrust loading. Diagonal ruling indicates major gaps in the stratigraphic record. All formations are constrained by biostratigraphic data except where indicated by '?'. Stratigraphic names are conventional, but interpretations of their ranges are ours (see text for details).

84

C. R. FIELDING ETAL. BOOMER RANGE

0 N

BANANA/ MOURA

FITZROY RIVER

CRACOW

DCC

500

MCB

1000

? ...

COCKATOO-1 S BCM

GY

FT ;

BAR

1500

BF

MCB BAR

2000

BAR

YC

25Q0J metres

VOLCANIC BASEMENT _

1

w c/>

o en

1

Basement Over much of the exposed eastern margin of the structural Bowen Basin, late Early Permian sedimentary strata unconformably or disconformably overlie volcanic, volcaniclastic and minor sedimentary rocks of mid-Carboniferous to Early Permian age, referred to the Camboon, Connors and Lizzie Creek Volcanics (Dickins & Malone 1973; Dear 1994) and considered by these authors as components of the Connors-Auburn Arch. Similar relationships have been noted on the eastern side of the Connors Arch, where Permian sediments of the Carmila beds and Calen Coal Measures overlie Connors and Lizzie Creek Volcanics. We have found volcanic rocks of similar compositional variety and mode of origin in the Lizzie Creek and Connors Volcanics on both sides of the Connors-Auburn Arch, in the Camboon Volcanics on the western side and in the lower part of the Carmila beds on the eastern side. Accordingly, given the limited available geochronological data (see Dear 1994; Holcombe et al. 1998a, for more details), we regard these volcanic successions as broadly coeval. In each case, the top of the volcanic succession is marked by an abrupt contact with overlying, mainly fine-grained, Early Permian sedimentary rocks (Figure 4a). Locally, the Permian sedimentary section contains volcanic and intrusive rocks of predominantly basaltic composition, with minor felsic ignimbrites (Figure 4b). Such relationships have been

coal

Figure 3 Interpreted northsouth stratigraphic cross-section along the eastern Bowen Basin illustrating continuity of the major middle and Late Permian formations over at least 350 km from the northern Gogango Overfolded Zone in the north to the southeast Bowen Basin in the south (see Figure 1 for location of sections). Dominant sediment dispersal direction in these formations was east to west (Figure 5). YC, Youlambie Conglomerate; BU/OX, Buffel /Oxtrack Formations; BAR, Barfield Formation; MCB, Moah Creek beds; BF, Boomer Formation; FT, Flat Top Formation; GY, Gyranda Formation; BCM, Baralaba Coal Measures; DCC, Dinner Creek Conglomerate. Dotted lines indicate correlations. The packages of interbedded massflow conglomerates noted in the text are shown by shading.

observed throughout the northern New England Fold Belt in the study area, although in some eastern parts (e.g. east of the Rookwood Thrust; Figure 2), sedimentary rocks of the Yarrol Block form basement to Permian strata. The Connors Volcanics and equivalents are a composite succession of lavas, intrusives and volcaniclastics of widely varying composition. In the southern Connors Arch, Dear (1994) recognised four discrete and correlatable volcanic units separated by extensive conglomerate bodies. Although these units cannot as yet be correlated regionally, a similar, composite architecture has been noted in the Camboon Volcanics at Cracow (Jones 1994) and the Connors Volcanics near Nebo (J. McPhie & C. R. Fielding unpubl. data). The context of these volcanic rocks is as yet uncertain (see discussion in Holcombe et al. 1997a), and there is no clear indication as to whether a significant constructive edifice existed during the eruption of these volcanics (along the present Connors Arch). The consistent eastward palaeocurrent direction evident from sedimentary structures in volcanic clast conglomerates within and immediately overlying the volcanics on both sides of the Connors Arch may, however, indicate the existence of a northsouth extensive, high-profile volcanic terrain during Late Carboniferous to earliest Permian time (Figure 5a). A marine environment to the east of the volcanic system is indicated locally by the occurrence of Early Permian marine invertebrate fossils (Fauna 1 of Dickins et al.

PERMIAN, NORTHERN NEW ENGLAND FOLD BELT 1964; see also Briggs 1993) within the basal conglomerates (e.g. at Charon Point peninsula, BMR locality DU179: Figure 4a). The context of other similar fossil occurrences within the Lizzie Creek Volcanics of the Connors Arch (e.g. BMR localities SL59, SL60: Malone et al 1969) is as yet unclear. The contact between volcanic basement and Permian sedimentary strata may be diachronous. Radiometric dates (summarised in Webb & McDougall 1968; Holcombe et al 1997a; C. M. Allen, I. S. Williams & C. J. Stephens pers. comm. 1996) indicate that magmatism in the Connors-Auburn Arch extended into the Permian, whereas sedimentary rocks equated with the overlying Bowen Basin succession are known to have formed as far back as the basal Permian (e.g. SHRIMP date of 293 Ma from zircons in felsic ignimbrite within sediments of Carmila beds at Dumbleton Rocks near Mackay; Figure 4b; C. M. Allen, I. S. Williams & C. J. Stephens pers. comm. 1996). From these relationships, it is suggested that the extensional event that formed the Bowen Basin may have begun in the earliest Permian or even Late Carboniferous, and may have been continuous through the magmatic episode that produced the Connors-Camboon Volcanics into basin formation (Holcombe et al 1997a).

Early Permian extensional package Thick (up to 3000 m: Kirkegaard et al 1970), laterally variable successions of mainly non-marine, Early Permian clastic sedimentary rocks overlie volcanic or sedimentary basement over large parts of the eastern Bowen Basin and adjacent New England Fold Belt. West and south of Rockhampton, these strata form the Youlambie Conglomerate and locally the uppermost parts of the Camboon Volcanics, whereas to the north they form the upper parts of the Carmila beds and Lizzie Creek Volcanics, and lower part of the Calen Coal Measures. Complete sections through this thick succession (including enclosing units) have been logged at Scrub Creek west of Rockhampton (8951—GR 990158 to 957127; Figure 6) and Hazelwood Creek southwest of Eungella (8555—GR 537499 to 476492; C. R Fielding, C. J. Stephens, J. McPhie, unpubl. data). In both cases, the largely non-marine Permian sedimentary strata overlie volcanic and sedimentary basement of uncertain age. Outcrop patterns in the Scrub Creek area demonstrate interfingering between the Youlambie Conglomerate and Rookwood Volcanics (which contain Artinskian foraminifers: O'Connell 1995), and marine invertebrate macrofossils referable to the Artinskian E. warwicki or lower E. preovalis biozones of Briggs (1993; pers. comm. 1995) have been found at a horizon in the Youlambie Conglomerate below the Rookwood Volcanics (Figure 6). In general, and in both the sections noted above, the Early Permian rocks are overlain by sandstones and siltstones containing middle Permian marine faunas. Conglomerates dominated by subangular to wellrounded clasts of volcanic and intrusive rocks with a

85

volcanic lithic matrix are commonly, but not ubiquitously, developed at the base of the succession. At a few localities near the present coast, the conglomerates are interbedded with bioclastic (and siliciclastic debris-rich) limestones which contain Sakmarian fossils (noted above; Figure 4a). Elsewhere (e.g. Hazelwood Creek to Nebo), the basal part of the unit is dominated by basaltic sandstones rich in fresh, first-cycle volcanic detritus. Where present, the basal, coarse clastic rocks typically pass upwards into thick successions of interbedded claystones and siltstones, thin-bedded siltstone - fine sandstone, discrete sandstone bodies, minor coal and locally thick conglomerates with well-rounded clasts of granite and volcanic lithologies (Figure 4b). In some areas (e.g. along the western edge of the northern Connors Arch between Nebo and Blenheim station), the succession contains numerous intraformational breccias with clasts of sandstone and siltstone up to several metres in diameter. Direct evidence of contemporaneous basaltic and felsic volcanic activity is locally preserved in the form of basalt lavas and high-level, pillowed intrusives, plus minor dacitic to rhyolitic ignimbrites and air-fall tuffs, while indirect evidence abounds in the abundance of first-cycle volcanic clasts and the typically Figure 4 (next two pages) Selected logged sections through parts of Permian formations in the eastern Bowen Basin and adjacent New England Fold Belt discussed in the text, (a) Carmila beds near Charon's Point within the northern Gogango Overfolded Zone (8852—GR 936984), showing volcanic rocks of interpreted Connors Volcanics overlain by coarse basal conglomerates and interbedded marine bioclastic limestones, in turn fining upward into interbedded siltstonesandstone strata typical of the upper Carmila beds, (b) Carmila beds at Dumbleton Rocks, near Mackay (8755—GR 156606), showing facies typical of the Early Permian extensional package. Note the intraformational unconformity at 150-152 m and the presence of two silicic ignimbrites, zircons from the lower of which has yielded a SHRIMP age of 293 Ma (C. M. Allen, I. S. Williams & C. J. Stephens pers. comm. 1996). (c) Undifferentiated Back Creek Group overlying Connors Volcanics near Yatton Creek (8752—GR 406795) in the Nebo Synclinorium of the Bowen Basin, showing marine strata characteristic of the thermal subsidence package. (d) Undifferentiated Back Creek Group overlying Connors Volcanics in the headwaters of Leura Creek in the southern Strathmuir Synclinorium (8851—GR 732393), showing basal volcanic lithic sandstones containing marine fossils in their upper part reminiscent of the Fairyland Formation, overlain by basement-clast-rich bioclastic limestones containing Fauna II fossils, passing in turn upward into bioturbated siltstone— sandstone strata typical of the Barfield Formation and equivalents, (e) Undifferentiated Back Creek Group at Apis Creek in the Strathmuir Synclinorium (8852—GR 595677), showing typical middle Permian shallow-marine facies. (f) Boomer Formation in the lower Leura Creek in the Nebo Synclinorium (8851—GR 630346), comprising similar strata to (e). (g) Boomer Formation exposed in a road cutting in the Broadsound Range (northern Gogango Overfolded Zone: 8852—GR 754656), of similar origin to (e) and (f). (h) Uppermost Moah Creek beds and overlying Dinner Creek Conglomerate in a road cutting in the Native Cat Range in the Gogango Overfolded Zone (8951—GR 068974), showing abrupt transition from marine strata of the Moah Creek beds into non-marine facies of the Dinner Creek Conglomerate.

86

C. R. FIELDING ETAL.

( A ) CHARON POINT PENINSULA "LOST LIMESTONE

H

. SI

CLAY SILT

(B)

-

DUMBLETON

ROCKS

BLUFF'

HIGH ENERGY CDASTAL AND SHALLOW MARINE

CLAY SILT

SAND RAVEL

(C) OLD SARINA ROAD NORTH OF YATTON CREEK

SAND SWVEL

CLAY SILT

SAND BWVEL

( D ) BOOMER RANGE, TRIBUTARY OF LEURA CREEK

TOP OF M EXPOSED SECTION 50

OFFSHORE MARINE

(E)

APIS

CREEK TOP OF EXPOSED M SECTION

OFFSHORE MARINE

OFFSHORE MARINE

y

y '

Planolltes

ZD

V

OFFSHORE

Planolltes etc.

~ Jl _

m

t/ OFFSHORE MARINE

OFFSHORE MARINE 8 ol

Z

U

Skolithos

SHOREFACE

, EXPOSURE 0» CONTINUES

CLAY SILT

SAN) CRAVEL

SECTION REPEATED 0 BY FOLDING

CLAY SILT

SAND WAVEL

1-

CLAY

SILT SAND GRAVCL

|SHOREFACE

PERMIAN, NORTHERN NEW ENGL AND FOLD BELT (G)

BROADSOUND RANGE

TOP or M EXPOSED SECTION 50

(F)

LEURA CREEK

SHOREFACE Planoll-tes

i f / / *

• = r ~ j f

OFFSHORE MARINE SHDREFACE

OFFSHORE MARINE

^

Ptanolrtes Telchlchnus Zoophycos

SECTION REPEATED 0 - |

CL HI ' JILT ' SANJ OtAVEL

BY FOLDING

ROAD CUTTINGS, NATIVE CAT RANGE

SHOREFACE ,

CUT SLT

SAND ORAVQ.

KEY TO SYMBDLS USED IN STRATIGRAPHIC SECTIONS BED CONTACTS m

.

SEDIMENTARY STRUCTURES MEAN

GRADA TIONAL I

1

PALAEOCURRENT

DIRECTION

GRADATIONAL

(= WEST)

EROSION

SURFACE

SHARP/EROSIVE OR IRREGULAR

FLAT

SHARP/PLANAR

CROSS-BEDDING

LAMINATION

LOW-ANGLE (MAINLY

LAMINATION

TROUGH)

HUMMOCKY

CROSS-

STRATIFICATION UNDULATORY CURRENT

LITHOLOGICAL QUALIFIERS

CURRENT

COMBINED-FLOW COMBINED LENTICULAR WAVY

1

BALLS

CONVOLUTE

•A.

WATER ESCAPE

JJ

TUFF

<s>

CARBONATE

NODULE

CARBONATE

CEMENT

u/c

UNCONFORMITY

BASE OF

CASTING

LOAD

JL

CLAST

COHERENT

LOAD

IGN1MBRITE

(X)

CLAY SILT SAND GRAVEL SECTION

RIPPLES

FELSIC

¥

—

0 -I

LAMINATION

WAVE XT

O 9

BEDDING

BEDDING

LINSEN

LAVA

V Palaeophycus

0

FLOW

MAFIC

FELSIC

OUTSIZE

CLAST

EXTRAFORMA TIONAL

CLAST

INTRAFORMATIONAL

CLAST

IMBRICATION

RIPPLE

CROSS-LAMINATION

COAL/CARBONACEOUS SHALE

AAAA

LENTICULAR

BEDDING

CLAST-SUPPORTED CONGLOMERATE INTERBEDDING OF DIFFERENT GRAIN SIZES

AAAAA

BEDDING

RIPPLE

CROSS-LAMINATION

STRUCTURES

BEDDING

SYNAERESIS

CRACKS

RECUMBENT

SOFT-SEDIMENT

FOLD

STRUCTURE)

(LOOP

COALY ^

BEDDING

DISTURBED

TRACES

MARINE

INVERTEBRATE

FOSSILS BURROWS

(GENERA

/ r

PLANT

FOLIAGE

//

PLANT

STEMS

(FRAGMENTED) TRACKS/TRAILS

SPECIFIED)

Figure 5 Summary of palaeocurrent data collected from Permian strata in the eastern Bowen Basin and adjacent New England Fold Belt, (a) Connors Volcanics (basement) and overlying coarse, basal conglomerates, (b) Carmila Beds (upper sedimentary package), Youlambie Conglomerate and Lizzie Creek Volcanics (upper sedimentary package), and equivalents— products of Early Permian extension (Phase 1).

Figure 5 (cont.) (c) Oxtrack/Barfield Formations and equivalents—products of mid-Permian, passive thermal subsidence (Phase 2). (d) Gyranda Formation, Dinner Creek Conglomerate, Baralaba Coal Measures and equivalents—products of Late Permian thrust loading (Phase 3).

oo

90

C. R. F I E L D I N G ET AL.

first-cycle volcanic provenance of sandstones. A feature of these rocks is the preservation of intraformational unconformities (Figure 4b). These facies are interpreted as the products of lacustrine and fluvial environments in rapidly subsiding, technically and volcanically active basins, and are similar in most respects to the stratigraphically equivalent Reid's Dome beds in the Denison Trough to the west (Draper & Beeston 1985). Basement clast conglomerates are interpreted as alluvial fan and plain deposits shed from elevated terrains composed of older volcanic and intrusive rocks. Local occurrences of fossiliferous, bioclastic limestones indicate occasional marine transgressions to at least the easternmost part of the study area. Intraformational breccias, which are particularly characteristic along the western side of the northern Connors Arch, are interpreted as subaqueous slide and mass-flow deposits down a technically steepened, subaqueous surface near an active fault or faults. Palaeocurrent data from these rocks typically indicate complex patterns of sediment dispersal, in some cases centripetal into areas of known outcrop (Figure 5b). Dickins and Malone (1973) suggested that these sediments accumulated in discrete sub-basins, a view we support. We further suggest that these sub-basins were formed by limited continental extension in the Early Permian, as part of a suite of such basins that includes the Denison, Arbroath and Bogong Troughs further to the west and southwest. Remnants of two such (possibly half-graben) sub-basins can be defined along the western and eastern sides of the Connors Arch, possibly originally separated by a basement ridge that supplied some sediment (Figure 2) and further basin complexes occupied parts of the Gogango Overfolded Zone and the area around Biloela in the southeast (see palaeogeographic maps of Fielding et al. 1995). This succession of predominantly fine-grained sedimentary rocks forms the earliest, clearly distinguishable record of Bowen Basin sediment accumulation. In the Gogango Overfolded Zone west of Rockhampton, the upper part of the extensional-phase succession contains a thick pile of basaltic intrusive and subaqueous eruptive rocks with minor sediments (the Rookwood Volcanics: Kirkegaard et al. 1970; O'Connell 1995). Our mapping has shown that this unit overlies, but also passes eastward into, the Youlambie Conglomerate (Figure 6), and is overlain by a finegrained marine sedimentary sequence. The whole package is carried on the Rookwood Thrust which surfaces to the west (Holcombe et al. 1997b figure 2). Facies relationships (O'Connell 1995) suggest sediment and lava accumulation on an uneven, technically active submarine surface of uncertain depth. Marine fossil-bearing strata are scattered throughout the Early Permian extensional package, but are particularly abundant in the uppermost part. These clastic and carbonate sedimentary rocks, biostratigraphically dated as late Artinskian (Briggs 1993) (Figure 2), include the Fairyland and Buffel Formations of the Cracow area, the Tiverton Formation further north and

unnamed equivalents. Parfrey (1986) interpreted a fauna collected from volcanogenic sandstones, mapped as part of the Camboon Volcanics near Biloela, as correlative with the Buffel and Tiverton Formations. Lateral facies and thickness changes are characteristic of these formations, and Draper (1988) has demonstrated their confinement within extensional sub-basins in the Cracow area. Locally, well-sorted bioclastic limestones of this age occur on both the western (e.g. Yatton Limestone) and eastern (e.g. in the headwaters of Leura Creek: Figure 4d) sides of the Connors Arch (Malone et al. 1969), and of the Auburn Arch farther south (Dear et al. 1971). Mid-Permian thermal subsidence package Overlying the Lower Permian units described above, and locally directly overlying basement (Figure 4c), are laterally extensive, sheet-like accumulations of marine strata (Figure 2). In the southeast, these rocks are referred to the Oxtrack and Barfield Formations (Dear et al. 1971), which have been mapped both to the west and east of the Auburn Arch between Cracow and Biloela. Farther north, identical facies of equivalent age have been mapped as Gebbie Subgroup and undifferentiated Back Creek Group on the western side of the Connors Arch, and as undifferentiated Back Creek Group and Moah Creek beds to the east (Malone et al. 1969; Kirkegaard et al. 1970). Elsewhere across the Gogango Overfolded Zone, more strongly deformed but otherwise identical strata have been mapped as Rannes beds and Boomer Formation (Figure 2; Malone et al. 1969; Kirkegaard et al. 1970; Holcombe et al. 1997b). The base of this marine succession is a hiatal surface across much of the Bowen Basin (Fielding et al. 1995). Overlying this surface in some areas, notably on the flanks of basement blocks, are thin bioclastic limestone or calcareous sandstone units. In the south, the carbonates are referred to the Oxtrack Formation (Kungurian to Ufimian), and have again been recognised both to the west and east of the Auburn Arch. These limestones contain a distinctly different fauna to that of the underlying Buffel Formation and equivalents (Fauna IV: Dickins & Malone 1973; Echinolosia n. sp. B/C biozones of Briggs 1993). The carbonates are overlain, or in places the basement is directly overlain, by thick, monotonous successions of fossiliferous, thin-bedded sandstone-siltstone strata (Figure 3) containing Ufimian to Kazanian faunas (Fauna IV, Echinolosia n. sp. D to Pseudostrophalosia n. sp. Zones). Interbedding is rhythmic (Figure 4c-g), with varying proportions of the two principal lithologies. Sandstone beds vary from ~0.1 m to 1.0 m in thickness overall, but individual exposures typically show relative consistency of bed thickness. Sedimentary structures are mainly ripple cross-lamination with some cross-bedding, flat lamination and soft-sediment structures, and rare flute and other sole marks. Palaeocurrent data collected from such structures consistently indicate westward sediment dispersal (Figure 5c). A variety of marine body and trace fossils has been recorded from

PERMIAN, NORTHERN NEW E N G L A N D FOLD BELT

Figure 6 Geological map of the Aeroview—Scrub Creek area in the central Gogango Overfolded Zone, showing the stratigraphic equivalence of the Rookwood Volcanics with the uppermost Youlambie Conglomerate. Note the two marine fossil localities which also constrain the unit biostratigraphically.

\ \

strike of bed with dip thrust fault fault geological boundary trend line

dashed where approximate

(2) microfossil locality (J) macrofossil locality <XD fossil wood locality

these rocks, along with scattered wood debris. Locally (e.g. in road cuttings near the crest of the Gogango Range: 8950—GR 994736), hummocky cross-stratification was noted but this structure is uncommon. The carbonates are interpreted as coastal and shallowmarine deposits, formed during the initial stages of passive, slow subsidence (Draper 1988). The overlying clastics are interpreted as deposits of mainly offshore marine shelf or ramp environments, where coarse sediment was introduced by offshore-directed (westward), currents, possibly as turbulent underflows. Hummocky cross-stratified units may reflect intermittent shallower water depths in some areas. Notwithstanding the variation in structural style, the middle Permian marine strata are palaeontologically similar (Dickins & Malone 1973) and are here regarded as part of the same succession (Figure 2). Graphic logs from sections exposed in units mapped as 'undifferentiated Back Creek Group', Boomer Formation and Moah Creek beds on both western and eastern sides of the Connors Arch (Figure 4) indicate no significant differences in lithofacies assemblage. Furthermore, the eastern part of the Duaringa 1:250000 geological map shows along-strike contacts between rocks mapped as Rannes beds and Boomer Formation. The Rannes beds are distinguished only by their more intense structural deformation, and are therefore included in this association. None of the facies recognised in this suite of rocks is considered indicative of basin marginal situations: there are virtually no basement clast conglomerates, and no facies transitions that might indicate shoaling onto an emergent Connors-Auburn Arch. Furthermore, palaeocurrent data collected from widely distributed localities (Figure 5c) indicate consistent

Permian

•

150'E

Boomer Formation/ Moah Creek beds Rookwood Volcanics

rhyolitic pumice breccia pillow basalt/dolerite basalt & sedimentary units sedimentary units

Youlambie Conglomerate

91

Carboniferous

Neerkol Formation Rockhampton Group

Siluro-Devonian

111

Undifferentiated

westward sediment dispersal on both sides of the arch, and are interpreted to reflect sediment derivation from a source terrain to the east of the present outcrop belt (and by inference, east of the present coastline). The succession accumulated during a period of regional, passive subsidence, interpreted to reflect thermal relaxation of the earlier extensional terrain. Punctuating this monotonous succession of thinbedded strata are at least two discrete intervals dominated by chaotic breccias, diamictites and conglomerates. These intervals have been recognised north-south for at least 350 km (Figure 3), and occur both on the west and east sides of the Connors-Auburn Arch. The coarse-grained rocks contain both rounded basement clasts and blocks and slabs of intraformational sandstone and siltstone, typically suspended within a poorly sorted, fine-grained matrix which is also rich in detrital wood. They are interbedded with, and enclosed by, interbedded sandstone-siltstone strata identical to those noted above and which contain marine trace and body fossils. Fielding et al. (in press) interpret these rocks as the products of tectonic destabilisation of a submarine surface, and suggest that they may indicate the onset of foreland thrust-induced subsidence in the Bowen Basin. The middle Permian marine succession may therefore be said to contain a record of the transition from passive, thermal subsidence to thrust load-induced, foreland basin-style subsidence. Late Permian foreland basin package The middle Permian marine strata coarsen upward into coarse-grained clastic facies that are devoid of marine

92

C. R. FIELDING ETAL.

biota. In the south, this occurs in the lower part of the Gyranda Formation (Banana Member), which is only recognised on the western side of the Auburn Arch. Further north, the same transition occurs at the boundary between the Moah Creek beds and Dinner Creek Conglomerate (Figure 4h), but has to date been recognised only east of the Connors Arch in the Gogango Overfolded Zone. This most likely reflects the poor quality of exposure in the Bowen Basin proper and the masking effect of the Tertiary Duaringa Basin, rather than its explicit non-occurrence. In its most easterly exposure, the Dinner Creek Conglomerate lies unconformably on Carboniferous sedimentary rocks of the Yarrol Block (Kirkegaard et al. 1970). The lower Gyranda Formation (Banana Member) coarsens up from body and trace fossil-bearing, tuffaceous siltstones into sandstone-dominated strata containing a variety of current-generated structures. The upper half of the unit is predominantly interbedded tuffaceous sandstones and conglomerates which contain well-rounded clasts of mainly volcanic lithologies. Palaeocurrent data indicate sediment transport to the west and south (Figure 5d; Miller 1992). The Moah Creek beds — Dinner Creek Conglomerate transition is similar (Figure 4h), and coarsens upward into wellrounded clast conglomerates dominated by metasedimentary quartzite clasts. Tuffaceous siltstones and fine sandstones that are interbedded with conglomerate bodies are rich in plant fossil debris, including wellpreserved Glossopteris leaves. Palaeocurrent data indicate sediment transport towards the west-southwest (Figure 5d). The coarsening-upward sequences described above are interpreted to record the final infilling of the marine Bowen Basin. The thickness of these coarseningupward intervals gives an approximation of water depth of 20-100 m during this sedimentation. The abrupt transition between offshore marine silts and non-marine coarse clastics, illustrated in Figure 4h, on the other hand is suggestive of a relative base-level drop causing incision of the depositional surface. The notion of a drop in relative base-level preceding accumulation of the conglomerates is also supported by the direct unconformity between the Dinner Creek Conglomerate and Carboniferous basement at Stanwell. The conglomerates are interpreted as the products of mainly high-energy fluvial processes, which delivered sediment into the basin from a source area to the east. Contemporaneous volcanic activity is indicated by the overwhelmingly volcanic lithic and tuffaceous composition of silts and sands. The source and context of this volcanic activity are at present uncertain. In the south, the Gyranda Formation is overlain by the uppermost Permian Baralaba Coal Measures and equivalents, which are extensive across the entire Bowen Basin (Fielding et al. 1993). A conglomeratedominated succession close to the present structural Bowen Basin margin (from Theodore southward) passes westward towards Moura and upward stratigraphically into more sandstone-dominated and coal-rich facies. Palaeocurrent data indicate westward sediment dispersal (Figure 5d). These patterns can be traced northward as

far as Baralaba, but are then obscured by Tertiary and Quaternary strata of the Duaringa Basin. No latest Permian coal measures are known from within or east of the Gogango Overfolded Zone—this may be due to non-deposition, subsequent erosion, or may indicate that the Dinner Creek Conglomerate is in part equivalent to the Baralaba Coal Measures. The latest Permian coal measures are interpreted as the products of high-energy alluvial environments close to the basin margin, which passed downdip (westward) into deposits of large, meandering rivers and extensive floodbasin and wetland environments (see Fielding et al. 1993, 1996 for more details). As such, the coarse alluvial facies may indicate the early stages of westward migration of the eastern basin margin during the latest Permian.

DISCUSSION The similarity in character of the Permian succession across the Bowen Basin and Gogango Overfolded Zone suggests that it accumulated in a single sedimentary basin. Stratigraphic thicknesses, sequence architecture, lithofacies assemblages and palaeocurrent distributions are remarkably similar. On this basis, we propose that the Permian succession of the northern New England Fold Belt described herein formed within a single depositional entity of which the structurally defined Bowen Basin is only a part. The succession began to accumulate during the earliest Permian, at least locally, in response to a longlived episode of modest crustal extension. This extension pulled apart a basement terrain which, in the study area, was a complex of volcanic, intrusive, sedimentary and metasedimentary rocks of Devonian to earliest Permian age. A series of discrete, probably fault-bounded sub-basins formed, some of which were initially sites of basaltic (and occasionally felsic) magmatic activity. A basin-and-range topography caused complex patterns of sediment dispersal into the early sub-basins. Some coarse sediment was supplied from emergent basement highs (including, we suggest, a ridge or ridges along parts of the present ConnorsAuburn Arch; Figure 2), while much of the material was probably derived from contemporaneous silicic and/or basaltic volcanism. During the latter stages of this extensional phase, particularly in the east, marine incursions became increasingly frequent. Individual sub-basins, now parts of the allochthonous Gogango Overfolded Zone, became the sites of voluminous, subaqueous basaltic magmatic activity (represented by the Rookwood Volcanics; see Fielding et al. 1995 for a palaeogeographic map). Lithostratigraphic and palaeontological data (O'Connell 1995) argue strongly for a shallow-marine environment of formation for the Rookwood Volcanics, in contrast to the deep marine, 'flysch trough' envisaged by Day et al. (1983). Importantly, no evidence exists for a volcanic 'arc' active within the region during this part of the Permian period. This period of extension was followed by a hiatus

PERMIAN, N O R T H E R N N E W E N G L A N D FOLD BELT 93 (Figure 2), which may have been related to upwarp of are no geochronological data and field relationships are the thinned, extended crust as has been described from equivocal. Based on cross-cutting relationships in the other extensional basins (Busby & Ingersoll 1995). Bowen Basin, Korsch et al (1992, 1995) interpreted the Fielding et al (1995) suggested that this hiatus might basin-wide thrusting event of the Hunter-Bowen correlate with a 260 Ma contractional event recognised Orogeny as Late Triassic in age, which is in accord with in southeast Queensland, but recent revision of the our data from the Capricorn Region (Holcombe et al absolute time-scale suggests that this hiatus occurred ca 1997b). We agree with the structural interpretation of 275-272 Ma (Jones 1996; Figure 2). The largely infilled Fergusson (1991 et seq.) that the Connors Arch became extensional topography subsided slowly and evenly an elevated basement-high during this thrusting episode. across the entire area, giving rise to a basin-wide transgression and the accumulation of successions of relatively fine-grained, shallow-marine strata. Facies CONCLUSIONS and palaeocurrent data demonstrate that, after the initial stages of the transgression when basement highs may Permian sedimentary successions are preserved within a have remained as relative highs (e.g. giving rise to the fold-thrust belt of Late Triassic age in the northernmost carbonate deposits of the Oxtrack Formation, etc.), none New England Fold Belt of coastal central Queensland. of the original basement terrains that separated exten- The nature of these strata, in terms of lithostratigraphic sional sub-basins remained emergent. We confirm the architecture, lithofacies assemblages and palaeocurrent conclusion of some earlier workers (summarised by distributions suggests that they formed as part of the Dickins & Malone 1973) that marine sediments must Bowen Basin. This interpretation is in contrast with have covered the entire area now forming the Connors some previous interpretations which regard the Bowen and Auburn Arch during the middle part of the Permian, Basin as separated from a deep-marine basin to the east and suggest that these sediments were sourced from an (the Grantleigh Trough) by volcanism along a basement emergent terrain some distance to the east of the present ridge (the Camboon Volcanic Arc along the Connors Arch). We find no evidence to support the notion of an limit of exposure. The foregoing argues that there is no basis for the upstanding Permian volcanic arc along the Connorsnotion of a deep-marine, 'flysch trough' in what is now Auburn Arch during sediment accumulation in the the Gogango Overfolded Zone. Palaeocurrent data indi- Bowen Basin, and no evidence that Permian rocks of cate sediment dispersal across the entire area towards the Gogango Overfolded Zone were separate from the the main part of the Bowen Basin, rather than focusing basin except during Early Permian extension when the into a discrete basin (Figure 5). No facies that might area was divided into discrete sub-basins. In accord with Malone, Dickins and coworkers, we be indicative of a deep, technically active basin are preserved, and the thickness estimates provided by pre- suggest that the original Bowen Basin extended at least vious workers are exaggerated by structural repetition as far east as the present coastline during the Permian, (e.g. the Rannes beds; Holcombe et al 1997a, b). We and that source areas for the preserved Permian strata therefore propose that use of the term 'Grantleigh subsequently have been rifted offshore during Cretaceous continental breakup (Falvey & Mutter 1981; Trough' be discontinued. The period of passive thermal subsidence {ca 272- Ewart et al 1992). We suspect that the basin margin 267 Ma; Figure 2) was terminated by the onset of thrust advanced progressively westward in the Early and load-induced subsidence, as recorded by the dis- Middle Triassic in response to a mountain building organised submarine conglomerates in the Barfield episode associated with the Hunter—Bowen Orogeny. Formation and Moah Creek beds. The succeeding strati- We support the view put forward by Fergusson (1991) graphic units are interpreted to record gradual, probably that the Connors Arch is an allochthonous basement pulsed encroachment of thrust fronts from the east block emplaced during Hunter—Bowen thrusting in the (Holcombe et al 1997b), leading to occlusion of the Late Triassic, when much, if not all, of the coastal marine basin and establishment of alluvial plain con- central Queensland area was transported westward as ditions across the entire region. As for the thermal sag- thrust sheets during contraction (Holcombe et al phase sediments, facies and palaeocurrent patterns 1997b). The data presented here, and new dating by us and indicate that sediments accumulated continuously, and other workers, requires significant revision of current without interruption, across present basement ridges. concepts of Bowen Basin history and definition of the Sediment accumulation in alluvial and lacustrine environments continued in the Bowen Basin proper Bowen Basin. The three-phase evolution for the basin until latest Middle Triassic times (Fielding et al 1995). development, and our stratigraphic analysis, clearly Thin-skinned thrust deformation of the entire basin defines two distinct tectonic events that gave rise to occurred at around 235—230 Ma, terminating sediment sediment accumulation in the basin, an early extensional accumulation. Little record of the Triassic is preserved environment and a later contractional environment. in the northernmost New England Fold Belt. One of few These events need not have formed a related tectonic possible Triassic units is the Native Cat Andesite, which cycle. In contrast with common usage and the ideas crops out in the area around Stanwell west of Rock- presented in many publications, the early basin was a hampton. Fergusson et al (1994) used the apparently much more extensive structure that subsequently fiat-lying aspect of this unit as evidence that it post- developed its present structural eastern boundary in dates the main Triassic thrusting event, although there response to the Hunter-Bowen contraction. The original

94

C. R. FIELDING ETAL.

geometry of the eastern margin was significantly modified, in position and probably in orientation, during this event. In addition, the definition of the base of the sedimentary sequence that filled the basin is unclear. Our interpretations suggest that dated sequences with sedimentological characteristics of half-graben development, and hence characteristics of extension basin formation, began forming while relatively productive magmatism was still occurring, as evidenced by Early Permian dates for volcanism in the Connors-Camboon Province and intrusions of the Urannah Suite. The magmatic rocks also show evidence of an extensional tectonic setting (Holcombe et al., 1998a). The character and distribution of the early sedimentary sequences are not well-constrained and require considerable new field work. However, we recognise a change, typically across a disconformity or unconformity, upward to a dominantly fine-grained, still predominantly grabenconfined, sedimentary sequence associated explicitly with mafic-dominated volcanism, in which there is little or limited input from a silicic basement or volcanic terrain. This sequence is overlain conformably by finegrained sediments of the thermal sag phase which define the first continuous sheet-like sediments in the basin. Whether the base of the basin is regarded as the base of the earliest extensional fill, the base of the transition to fine-grained rocks associated with mafic volcanism, or the arrival of the first sheet-like deposits is likely to engender considerable debate depending on the preference of the researcher. We do not explicitly recommend a convention here as considerable new work is yet required. It is, nonetheless, critical that the concept of the original basin be divorced from the present structural feature in any discussions of basin evolution, such that the great body of information preserved in rocks residing outside the 'Bowen Basin' not be excluded. ACKNOWLEDGMENTS This research was supported by an ARC Large Grant to A. Ewart and J. McPhie (A39232338), and by a research grant to the authors from Queensland Metals Corporation. Both sources of funding are gratefully acknowledged. David Briggs and Vince Palmieri provided useful advice on biostratigraphic issues, and reviews by C. L. Fergusson, R. A. Henderson and C. G. Murray improved the manuscript. REFERENCES BRIGGS D. J. C. 1993. Time control in the Permian of the New

England Orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, Eastern Australia, pp. 283—291. NEO '93 Conference Proceedings, University of New England, Armidale. BUSBY C. & INGERSOLL R. (Editors) 1995. Tectonics of Sedimentary Basins. Blackwell, Oxford.

DAY R. W., MURRAY C. G. & WHITAKER W. G. 1978. The

eastern part of the Tasman Orogenic Zone. In: Scheibner E.

ed. The Phanerozoic structure of Australia and variations in tectonic style, pp. 327—364. Tectonophysics 48.

DAYR. W., WHITAKER W. G., MURRAY C. G., WILSON I. H. & GRIMES K. G. 1983. Queensland Geology. Geological

Survey of Queensland Publication 383.

DEAR J. F. 1994. Major cycles of volcanism in the southern

Connors Arch. In: Holcombe R. J., Stephens C. J. & Fielding C. R. eds. 1994 Field Conference Guidebook: Capricorn Region, central coastal Queensland, pp. 31—45. Geological Society of Australia, Queensland Division, Brisbane. DEAR J. F., MCKELLARR. G. & TUCKER R. M. 1971. Geology of the Monto 1:250 000 Sheet area. Geological Survey of Queensland Report 46. DICKINS J. M. & MALONE E. J. 1973. Geology of the Bowen Basin, Queensland. Bureau of Mineral Resources Bulletin 130. DICKINS J. M., MALONE E. J. & JENSEN A. R. 1964.

Subdivision and correlation of the Middle Bowen Beds. Bureau of Mineral Resources Report 70.

DRAPER J. J. 1988. Permian limestone in the southeastern

Bowen Basin, Queensland: an example of temperate carbonate deposition. Sedimentary Geology 60, 155-162. DRAPER J. J. & BEESTON J. W. 1985. Depositional aspects of the Reid's Dome Beds, Denison Trough. Queensland Government Mining Journal 86, 200-210.

DRAPER J. J., PALMIERI V., PRICE P. L., BRIGGS D. J. C. & PARFREY S. M. 1990. A biostratigraphic framework for the

Bowen Basin. In: Bowen Basin 1990 Symposium Proceedings, pp. 26-35. Geological Society of Australia, Queensland Division, Brisbane.

EWART A., SCHON R. W. & CHAPPELL B. W . 1992. T h e

Cretaceous volcanic—plutonic province of the central Queensland (Australia) coast: a rift-related 'calc-alkaline' province. Transactions of the Royal Society of Edinburgh

83, 327-346. FALVEY D. A. & MUTTER J. C. 1981. Regional plate tectonics

and evolution of Australia's passive continental margin. Bureau of Mineral Resources Bulletin 6, 1—29.

FERGUSSON C. L. 1991. Thin-skinned thrusting in the northern

New England Orogen, central Queensland, Australia. Tectonics 10, 797-806.

FERGUSSON C. L. & LEITCH E. C. 1993. Late Carboniferous to

Early Triassic tectonics of the New England Fold Belt, eastern Australia. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 53-59. NEO '93 Conference Proceedings, University of New England, Armidale.

FERGUSSON C. L., HENDERSON R. A., LEITCH E. C. & ISHIGA

H. 1993. Lithology and structure of the Wandilla terrane, Gladstone-Yeppoon district, central Queensland, and an overview of the Palaeozoic subduction complex of the New England Fold Belt. Australian Journal of Earth Sciences 40, 403-414.

FERGUSSON C. L., HENDERSON R. A. & LEITCH E. C. 1994.

Tectonics of the New England Fold Belt in the Rockhampton Gladstone region, central Queensland. In: Holcombe R. J., Stephens C. J. & Fielding C. R. eds. 1994 Field Conference, Capricorn Region, Central Coastal Queensland, pp. 1-16. Geological Society of Australia, Queensland Division, Brisbane. FIELDING C. R., FALKNER A. J. & SCOTT S. G. 1993. Fluvial response to foreland basin overfilling: the Late Permian Rangal Coal Measures in the Bowen Basin, Queensland, Australia. Sedimentary Geology 85, 475-497.

PERMIAN, NORTHERN NEW ENGLAND FOLD BELT

95

JONES P. J. (Compiler) 1996. AGSO Phanerozoic 1995. Oxford University Press, Melbourne.

Timescale

FIELDING C. R., HOLCOMBE R . J. & STEPHENS C. J. 1 9 9 4 . A

critical evaluation of the Grantleigh Trough, east-central Queensland. In: Holcombe R. J., Stephens C. J. & Fielding C. R. eds. 1994 Field Conference, Capricorn Region, Central Coastal Queensland, pp. 17-30. Geological Society of Australia, Queensland Division, Brisbane. FIELDING C . R., STEPHENS C . J., KASSAN J. & HOLCOMBE R. J.

1995. Revised palaeogeographic maps for the Bowen Basin, central Queensland. In: Follington I. L., Beeston J. W. & Hamilton L. H. eds. Bowen Basin 1995 Proceedings, pp. 7—15. Geological Society of Australia, Coal Geology Group, Brisbane. FIELDING C . R . , KASSAN J. & DRAPER J. J. 1996. G e o l o g y o f

the Bowen and Surat Basins, eastern Queensland. Australasian Sedimentologists Group Field Guide Series 8. FIELDING C . R . , STEPHENS C . J. & HOLCOMBE R . J. 1997.

Submarine mass-wasting deposits as an indicator of the onset of foreland thrust loading: Late Permian Bowen Basin, Queensland, Australia. Terra Nova 9, 14—18. GEOLOGICAL SURVEY O F QUEENSLAND 1953. Geological

map

of Queensland, 40 miles to 1 inch. Geological Survey of Queensland, Brisbane. HAMMOND R. 1987. The Bowen Basin, Queensland, Australia: an upper crustal extension model for its early history. Bureau of Mineral Resources Record 1987/51, 131-142. HARRINGTON H. J. & KORSCH R. J. 1985. Tectonic model for

the Devonian to middle Permian of the New England Orogen. Australian Journal of Earth Sciences 32, 163— 179. HENDERSON R. A. 1980. Structural outline and summary geological history for northeastern Australia. In: Henderson R. A. & Stephenson P. J. eds. Geology and Geophysics of Northeastern Australia, pp. 1—26. Geological Society of Australia, Queensland Division, Brisbane. HOLCOMBER. J., STEPHENS C. J., FIELDING C. R. ETAL. 1997a. Tectonic evolution of the Northern New England Fold Belt: Carboniferous to Early Permian transition from accretion to extension. Geological Society of Australia Special Publication 19, 66-79. HOLCOMBE R. J., STEPHENS C. J., FIELDING C. R. ETAL. 1997b. Tectonic evolution of the Northern New England Fold Belt: the Permian—Triassic Hunter-Bowen event. Geological Society of Australia Special Publication 19, 52-65. JACK

R.

L.

&

ETHERIDGE R . JR.

1892.

The

geology

and palaeontology of Queensland and New Guinea. Publications of the Geological Survey of Queensland 92. JENSEN A . R . , GREGORY

C . M . & FORBES V . R .

1966.

Geology of the Mackay 1:250 000 Sheet area, Queensland. Bureau of Mineral Resources Report 104. JONES J. A. 1994. Character and setting of the Camboon Volcanics, central Queensland: implications for the Early Permian tectonics of the New England Fold Belt. Geological Society of Australia Abstracts 37, 204. JONES J. A . , STEPHENS C . J. & EWART A . 1996. C o n s t r a i n t s o n

the Early Permian and Late Carboniferous of the northern New England Fold Belt from the Camboon Volcanics and the Torsdale Beds. Geological Society of Australia Abstracts 41, 222.

KIRKEGAARDA. G . , SHAW R . D . & MURRAY C. G . 1 9 7 0 .

Geology of the Rockhampton and Port Clinton 1:250 000 sheet areas. Geological Survey of Queensland Report 38. KORSCH R. J. & TOTTERDELL J. M. 1995. Structural events

and deformational styles in the Bowen Basin. In: Follington I. L., Beeston J. W. & Hamilton L. H. eds. Bowen Basin Symposium 1995 Proceedings, pp. 27—35. Geological Society of Australia Coal Geology Group, Brisbane. KORSCH R . J., WAKE-DYSTERK. D . & JOHNSTONE D. W . 1992

Seismic imaging of Late Palaeozoic - Early Mesozoic extensional and contractional structures in the Bowen and Surat Basins, eastern Australia. Tectonophysics 215, 273-294.

MALONE E. J. 1964. Depositional evolution of the Bowen Basin. Journal of the Geological Society of Australia 11, 263-282. MALONE E. J., CORBETT D . W . P. & JENSEN A . R . 1 9 6 4 .

Geology of the Mount Coolon 1:250 000 Sheet area. Bureau of Mineral Resources Report 64. MALONE E. J., JENSEN A . R . , GREGORYC. M . & FORBES V . R .

1966. Geology of the southern half of the Bowen 1:250 000 Sheet area, Queensland. Bureau of Mineral Resources Report 100. MALONE E. J., OLGERS F. & KIRKEGAARD A . G . 1 9 6 9 . T h e

geology of the Duaringa and Saint Lawrence 1:250 000 sheet areas, Queensland. Bureau of Mineral Resources Report 121. MILLER D. E. 1992. Geology of the southern Moura Mine and Malakoff Range, southeastern Bowen Basin, central Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). MURRAY C. G. 1974. Alpine-type ultramafics in the northern part of the Tasman Geosyncline — possible remnants of Palaeozoic ocean floor. In: Denmead A. K., Tweedale G. W. & Wilson A. F. eds. The Tasman Geosyncline — a Symposium, pp. 161-181. Geological Society of Australia, Queensland Division, Brisbane. MURRAY C . G . , FERGUSSON C. L., FLOOD P. G . , WHITAKER W .

G. & KORSCH R. J. 1987. Plate tectonic model for the Carboniferous evolution of the New England Fold Belt. Australian Journal of Earth Sciences 34, 213-236. O'CONNELL S. A. 1995. Geology and structure of the Aeroview area, Rockhampton region, central Queensland: implications for mineralisation of the Lower Permian Rookwood Volcanics. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). REID J. H. 1929. Geology of the Bowen River coalfield. Geological Survey of Queensland Publication 276. REID J. H. 1930. The Queensland Upper Palaeozoic succession. Geological Survey of Queensland Publication 278. WEBB A . W . & MCDOUGALL I. 1968. T h e g e o c h r o n o l o g y o f

the igneous rocks of eastern Queensland. Journal of the Geological Society of Australia 15, 313—346.

(Received 24 April 1996; accepted 12 February 1997)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 96-108.

Tectonic significance of The Shacks Mylonite Zone and related shear zones, northern New England Fold Belt, Queensland V. J. MORAND*

Department of Geology, University ofBallarat, PO Box 663, Ballarat, Vic. 3353, Australia.

The Shacks Mylonite Zone is part of the Stanage Fault Zone, a north-northeast-trending fault in the central Queensland coastal region. Between Shoalwater Bay and Broad Sound, the Stanage Fault Zone outcrops as a 3 km-wide mylonite zone—The Shacks Mylonite Zone—which juxtaposes the high-grade metamorphosed Shoalwater terrane (the Broome Head Metamorphics) against low-grade Yarrol terrane rocks. The Shacks Mylonite Zone strikes northeast and the dominant foliation S m dips moderately southeast. Metamorphic grade decreases across the shear zone from amphibolite facies in the hangingwall to greenschist facies in the footwall. A stretching lineation L m plunges shallowly to the east and northeast, and shear-sense indicators show oblique dextral thrust movement, with transport to the west-southwest. Fault movement occurred during the HunterBowen Orogeny in the Late Permian. Displacement along the Stanage Fault Zone could be up to 100 km, based on the separation of elements of the Marlborough terrane along the fault. Farther south, this displacement was manifest as major thrusting along the Gogango Overfolded Zone and the leading edge of the Marlborough terrane thrust sheet. A possible explanation for these structures is that a rigid indentor (now hidden under the sediments of the continental shelf) impinged on the fold belt in this area during collisional orogenesis, causing major southwest thrusting and tearing the fabric of the fold belt along the Stanage Fault Zone. Key words: mylonite, New England Fold Belt, Stanage Fault Zone, structural terranes, The Shacks Mylonite Zone, thrust faults.

INTRODUCTION The New England Fold Belt in central eastern Queensland is dominated by northwesterly trending structures between Gladstone and Mackay (Figure 1). In this region there are three main tectonic units running parallel to the present coastline, the Yarrol terrane (Late Palaeozoic arc and forearc basin), the Wandilla terrane (mainly Early Carboniferous accretionary prism) and the Shoalwater terrane (undated accretionary prism) (Fergusson et al. 1990; Henderson et al. 1993). The Marlborough terrane is a strongly deformed ophiolitic sequence located on the boundary of the Yarrol and Wandilla terranes, at the northern end of the Yarrol Fault (Figure 1). In the Shoalwater Bay area a major fault system, the Stanage Fault Zone (Henderson et al. 1993), disrupts this pattern. Here the structural trends are north-northeast, and the Wandilla, Shoalwater and Marlborough terranes are terminated by the fault zone (Figure 1). Much of this fault zone is either under water or obscured by granite, but it crops out on the Stanage Peninsula, between Broad Sound and Shoalwater Bay, as a 3 km-wide zone of mylonitic rocks known as The Shacks Mylonite Zone (Morand 1993a), named after the locality The Shacks on the coast of Shoalwater Bay. The main fold and cleavage trends adjacent to the Stanage Fault Zone curve around the Marlborough terrane to become the Gogango Overfolded Zone (Figure 1), a major fold-thrust belt (Fergusson 1991).

Hence The Shacks Mylonite Zone is part of a major crustal-scale shear zone, and deciphering its structural history is essential to understanding the tectonic development of the northern New England Fold Belt. REGIONAL SETTING The complex geology of the Shoalwater Bay area has been documented by Leitch et al. (1994a, b). Limestone-bearing volcanic and sedimentary sequences of the Yarrol terrane are found on the west side of the Stanage Peninsula and on the Duke Islands to the northeast (Figure 2). Rocks of the Shoalwater terrane occur on the east side of the Stanage Peninsula and to the east of Shoalwater Bay, with an isolated occurrence in the Percy Isles (northern part of Figure 2). They range from low-grade quartz arenite, mudstone and chert (Shoalwater Formation) to upper amphibolite facies migmatitic gneiss and amphibolite, known as the Broome Head Metamorphics (Morand 1993a). A faultbounded belt of low-grade Shoalwater Formation is interleaved with Yarrol terrane strata to the west of The Shacks Mylonite Zone (Leitch et al. 1994b; Figure 3). South Percy Island (Figure 2) contains a fragment of the Marlborough terrane adjacent to the Stanage Fault Zone. The rocks here are mainly pillow basalt and serpentinite (Leitch et al. 1994b). * Present address: Geological Survey of Victoria, PO Box 2145, MDC Fitzroy, Vic. 3065, Australia.

THE SHACKS M Y L O N I T E ZONE

b rcy V,/

150°

S r ISLES /

0 5 10 15 20 25 1 l l i i i km

/

97

Fault

N i

DUKE • ISLANDS,^

-22°

A

K

/Or v Stanage

/CO

Townshend

THE SHACKS

Figure 1 Distribution of terranes in the northern New England Fold Belt. Outcrop areas of the Connors and Auburn Arcs are also shown. The Shoalwater terrane is postulated to continue offshore northwest of the Percy Isles. The 200 m bathymetric contour marks the edge of the continental shelf.

r T 1 Cretaceous volcanic and 1 * I sedimentary strata

f T I Shoalwater terrane (low L _ l to medium grade) Broome Head Metamorphics

^T+J Late Permian to Triassic • t -H intrusions

Marlborough

* | ?Permian acid volcanics Yarrol

terrane

m

terrane

High strain zone

| W | Wandilla terrane

The structure in the area of Figure 2 is dominated by a sequence of east-dipping thrust faults with westward transport, exposing older and higher grade strata to the east (Leitch et al. 1994a). Strain also intensifies eastward from Long Island across the Stanage Peninsula, culminating in the Broome Head Metamorphics which exhibit four generations of folds (Morand 1993a). The pattern of north- to northeast-striking thrust faults has been complicated by later faults such as the Emily Peak Fault (Figure 3). Two major shear zones bound the Broome Head Metamorphics: The Shacks Mylonite Zone in the northwest and the Sabina Point Shear Zone in the southeast. They are partly obscured by large posttectonic intrusions of granite and diorite (Figures 2, 3) but The Shacks Mylonite Zone is well exposed along the coast at The Shacks, and the Sabina Point Shear Zone has reasonable exposures in road cuttings in the Pine Mountain area, although Sabina Point itself is composed of laterite.

Figure 2 Basement geology of the Broad Sound-Shoalwater Bay region, based partly on interpretation of BMR aeromagnetics of the Port Clinton 1:250 000 sheet.

THE SHACKS MYLONITE ZONE General description The Shacks Mylonite Zone is cut by at least one later fault, and is also folded such that its strike changes from northeast to north. It consists of two distinct parts, separated by the Emily Peak Fault (Figure 3), which comprise completely different rock types, although the structures are similar throughout the mylonite zone. East of the Emily Peak Fault, The Shacks Mylonite Zone strikes northeast, is about 3 km width in outcrop (2 km true thickness) and separates the Broome Head Metamorphics (sillimanite—garnet gneiss, quartzite,

98

V. J. M O R A N D ,AAAAA/<(AAAAAAAAAAAA A A A A A A! A AAAAAAAAAAA j - A A A A A AJA AAAAAAAAAAA iA A A A A A AA / AAAAAAAAAAA - rrrVo , a a ^aaaaaaaaaaa f'5640/ AAAAAAAAAAA 7fi-J MAAAAAAAAAAAAA.VAA /

Bedding

X S2

Sm

Lm

Fault / —

Road

Figure 3 Geological map of The Shacks Mylonite Zone and adjacent rocks.

amphibolite and migmatite) from the Late Silurian to Early Devonian Mt Holly Formation (Figure 3). The latter is greenschist facies grade basic to acid volcanic and volcaniclastic rocks interbedded with marble and graphitic slate, all with a well developed cleavage, S i, striking north-northeast (Morand 1993b, c). Rock types in this eastern part of the mylonite zone appear to be schists in hand specimen, but microscopic examination reveals that they are a mixture of strongly foliated schist and protomylonite. Classic mylonitic fabrics with rounded porphyroclasts are not present. The southeast side of the mylonite zone is composed of mica schist and amphibolite derived from the Broome Head Metamorphics, and the northwest side consists of strongly cleaved low-grade marble and schistose volcaniclastics

derived from the Mt Holly Formation (Figure 4). Metamorphic assemblages on the northwest side of the mylonite zone are calcite-quartz—muscovite ± chlorite in marble and chlorite-muscovite-quartz-albite ± biotite in mylonitised volcaniclastics, indicative of the greenschist facies. On the southeast side of the shear zone metamorphic assemblages in metasedimentary units are quartz-muscovite-biotite-plagioclase ± garnet ± microcline, and within mylonitised amphibolite the assemblage hornblende-plagioclase-quartz-ilmenite is typical, both assemblages indicating amphibolite facies grade. Retrograde assemblages of clinozoisitemuscovite-sphene ± albite ± actinolite in mafic rocks post-date the mylonitisation. Thus metamorphic grade increases from greenschist facies on the northwest side

THE SHACKS M Y L O N I T E ZONE

99

2 km

-J

\

\ s - | in MHF, Pzs; S 2 in Pzm

A

MHF \

Mylonitised

Mylonitised

MHF

\

\Sm \Scren, S4 in Pzm 1 \ Pzm

Pzm

1

\ Mylonitised 7

Pzs

\

Pzs

Foliated granite

Figure 4 Cross-sections A—B and C - D (see Figure 3 for location). MHF, Mt Holly Formation; Pzm, Broome Head metamorphics; Pzs, low-grade Shoalwater Formation. SHOALWATER <(37

17

X

41

BAY

\

Figure 5 Geological map of the coast near The Shacks, showing transition from gneiss of the Broome Head Metamorphics to The Shacks Mylonite Zone. Cross-sections A—A' and B - C - D are also shown.

to amphibolite facies on the southeast side, where the mylonite grades into the Broome Head Metamorphics. The transition into the Broome Head Metamorphics is best seen along the coast near The Shacks (Figure 5). West of the Emily Peak Fault, The Shacks Mylonite Zone has an outcrop width of 2 to 3 km and separates quartzite and quartz arenite of the Shoalwater Formation from a granite containing quartz, plagioclase (oligoclase—andesine), perthitic microcline, brown biotite, olive green to brown hornblende and accessory allanite, zircon and opaque minerals. The age and affinities of this granite are unknown, but it must pre-date the posttectonic I-type granitoids of the area which cut the mylonite zone, and is not apparently related to the small S-type granite bodies derived by partial melting of the Broome Head Metamorphics (Morand 1993a). The metamorphic grade in the Shoalwater Formation is greenschist facies or lower (quartz-muscovite ± chlorite). The northwestern half of the mylonite zone is composed of fine-grained mylonitic quartz-rich schist

derived from the Shoalwater Formation, with greenschist facies assemblages of quartz—muscovite-chloritecarbonaceous opaques ± pyrite and locally biotite, or quartz-muscovite-chlorite—hematite. This gives way rather abruptly to mylonitic granite forming the southeastern half of the mylonite zone (Figure 3). Metamorphic minerals in the mylonitic granite are quartz, albite, biotite (green to brown), epidote, sphene, white mica, with some samples containing chlorite and/or actinolite, also indicating greenschist facies. No detectable increase in grade occurs across the mylonite zone west of the Emily Peak Fault.

Structure and movement sense One main mylonitic foliation, S m , is present in The Shacks Mylonite Zone, striking northeast and dipping moderately southeast, with more variable orientations in the west (Figures 3, 6). S m is a fine lamination

100

V.J. M O R A N D

Figure 6 Lower hemisphere, equal-area projections of poles to mylonitic layering (Sm), crenulation cleavage (S c r e n ) and axial planes of mesoscopic folds, and orientations of the stretching lineation (L m ), crenulation lineation (L c r e n ) and mesoscopic fold axes in The Shacks Mylonite Zone.

defined by muscovite and biotite in metasedimentary mylonites, although lensoidal muscovite fish (Lister & Snoke 1984) at an oblique angle to S m are present in some mylonites derived from gneiss. A well-developed mineral lineation, L m , is present on S111 surfaces (Figure 7a) pitching at a moderate angle to the northeast and east (Figures 3, 6). Some mylonites have an S-C fabric, with the intersection of S and C perpendicular to L m , thus supporting the interpretation of this lineation as the extension direction. Shear sense indicators such as S-C fabrics, quartz fabrics oblique to S m and mica fish (Figure 8a), are not common, and are generally seen microscopically. Data from oriented thin-sections indicate that The Shacks Mylonite Zone is an oblique thrust with a large component of dextral strike-slip. The transport direction, parallel to the stretching lineation L m , was to the westsouthwest in the eastern part of the mylonite zone, and generally to the west-northwest in the western part (Figure 6). On the coast near The Shacks (Figure 5) the transition from Broome Head Metamorphics, which have undergone six phases of deformation (Morand 1993a), to mylonite involves: (i) tightening of F 4 folds; (ii) a change from abundant complex mesoscopic folds (Morand 1993a) to one prominent foliation (Sm) with a consistent orientation; (iii) development of a prominent stretching lineation (Lm); (iv) a well-developed foliation in granite layers (in contrast to the massive to weakly foliated granite within the Broome Head Metamorphics); and (v) boudinage of granite layers along S m (Figure 7b). At The Shacks, S m is locally folded into isoclinal reclined folds plunging parallel to L m . Amphibolites from the mylonite zone do not have typical mylonite fabrics but are coarse grained with a well-developed schistosity, with hornblende prisms aligned within S m to form the mineral-stretching linea-

tion (Figure 8c). This contrasts with amphibolite in the Broome Head Metamorphics (Figure 8b) which has only a very weak fabric. The transition from mylonite to Mt Holly Formation is covered by sand dunes on the coast, but inland outcrops suggest that the one intense slaty cleavage in the Mt Holly Formation grades into S m , becoming less steep as the mylonite zone is entered (Figure 4), and the steeply plunging lineation L (Morand 1993c) grades into L m with a decrease in pitch. On the coast (Figure 3) a crenulation cleavage S 2 affects the Mt Holly Formation as The Shacks Mylonite Zone is approached (Morand 1993c). West of the Emily Peak Fault, the Shoalwater Formation has one moderately developed foliation S b and in sandstones the detrital quartz grains are clearly evident. These rocks become more intensely foliated and recrystallised as the mylonite zone is approached, with detrital quartz grains becoming elongated and developing strong undulose extinction and recrystallised grain boundaries, indicative of a dislocation flow mechanism. Within the mylonite zone, quartz is recrystallised into a fine-grained mosaic of equant grains, while opaque grains and aligned muscovite define S m laminations. Most specimens have only one strongly developed lamination S m , containing a stretching lineation L m , but some show microscopic isoclinal folding of S m . The recovery of quartz to produce granoblastic textures suggests a slow decline in temperature after mylonitisation, or a static thermal event post-dating fault movement. Figure 8d, e shows the transition from strongly cleaved Shoalwater Formation to intensely deformed quartz-graphite mylonite. Textures in foliated granite west of the Emily Peak Fault grade from mylonitic in the centre of The Shacks Mylonite Zone to protomylonitic at its southeast edge. Much of this granite is covered by coastal swamp and

THE SHACKS M Y L O N I T E ZONE

101

creri Figure 7 (a) Outcrop of mylonitic amphibolite near The Shacks, showing moderately pitching L m (parallel to the pen) on S m surfaces, (b) Boudinaged granite layer (g) in outcrop of mylonitic metasedimentary gneiss dominated by the foliation S m : 600 m southeast of The Shacks, (c) Mylonitised Mt Holly Formation showing mylonitic lamination S m folded into tight folds with S c r e n as axial plane: 2 km southwest of The Shacks, (d) Mylonitic quartzite from The Shacks Mylonite Zone, showing stretching lineation L m cut at a high angle by crenulation lineation L cren . Specimen is 15 cm wide.

all outcrops appear to be within the mylonite zone with foliation intensity decreasing to the southeast. In mylonitic samples all original igneous minerals are deformed, with recrystallised quartz and biotite anastomosing around stronger feldspar and hornblende remnants (Figure 8f). Protomylonitic granite has less intense foliation, but all igneous minerals show some solid-state deformation features, including undulose extinction, fracturing, kinking and recrystallisation, with fine-grained biotite and quartz defining S m and filling fractures in feldspars. Metamorphic biotite is green in contrast to the brown colour of igneous biotite. Late north-trending open to close, rounded to kinklike folds with steep axial planes deform S m and are correlated with F 6 folds developed during the last deformation event in the Broome Head Metamorphics (Morand 1993a) as they have the same orientation and the same open style, and some have monoclinal style with east-side-down asymmetry (z-shaped looking north) which is characteristic of F 6 (Morand 1993a). In some outcrops the axial plane is marked by a crenulation cleavage S cren (Figure 6), best developed where the folds are locally close to tight (Figure 7c). These folds are generally small (wavelength < 3 cm, amplitude < 1 cm) and they form a prominent crenulation lineation L cren on S m , generally at a high angle to

L m (Figure 7d). These may be responsible for bending the western part of the mylonite zone into a north-south strike. The spread of S m poles east of the Emily Peak Fault into a partial great circle girdle with its pole coincident with L cren (Figure 6) indicates that Sm has been deformed by these late folds on a scale larger than a typical outcrop. Low-plunging open folds, with no axial plane fabric but variable axial plane orientations, also deform S m (Figure 6), but they are sporadic and have not been correlated with any folds outside the mylonite zone.

SABINA POINT SHEAR ZONE The Sabina Point Shear Zone is a zone of highly schistose rocks at the southern boundary of the Broome Head Metamorphics (Figure 2) which, in the area adjacent to the shear zone, are biotite schist and pegmatite complexly folded on a small scale, in contrast to the shear-zone rocks which are dominated by one planar foliation. To the south of the shear zone the dominant rock type is low-grade quartz arenite of the Shoalwater Formation, where a weak north- to northwest-trending cleavage is present and clastic grains are evident in thin-section. Massive granite and diorite

102

V. J. M O R A N D

Figure 8 (a) Mylonitised metasedimentary gneiss from The Shacks Mylonite Zone at The Shacks, with muscovite fish in the centre of the photo indicating dextral shear. Most of the quartz and feldspar has undergone static recrystallisation after shearing to produce a granoblastic texture, while aligned micas define S m . Crossed polars, scale bar = 2 mm. Specimen 31830. (b) Typical texture of amphibolite from the Broome Head Metamorphics, with very weak preferred orientation of minerals. The minerals are mainly hornblende and plagioclase. Plane-polarised light, scale bar = 2 mm. Specimen 31824. (c) Amphibolite from The Shacks Mylonite Zone 800 m southeast of The Shacks, with strongly aligned hornblende defining S m . Minerals are hornblende and plagioclase. Plane-polarised light, scale bar = 2 mm. Specimen 31827. (d) Low-grade, Shoalwater Formation quartzite, 600 m west of The Shacks Mylonite Zone, west of the Emily Peak Fault. Clastic grains are evident but strongly elongated to form a welldeveloped cleavage which is cut by later stylolites. Plane-polarised light, scale bar = 2 mm. Specimen 31841. (e) Mylonitic quartzite from The Shacks Mylonite Zone west of the Emily Peak Fault, consisting of quartz and carbonaceous opaques with some pyrite and muscovite. An intense schistosity/layering S m is folded isoclinally and cut by later microfaults and quartz veins. All quartz has been statically recrystallised. Plane-polarised light, scale bar = 2 mm. Specimen 31813. (f) Mylonitic granite from The Shacks Mylonite Zone west of the Emily Peak Fault. A large plagioclase grain at the bottom of the photo is fractured and has undulose extinction, and the mylonitic foliation S m anastomoses around it. Relics of quartz grains are recrystallised and elongated along S m (running N—S in this photo), which is defined by trails of fine-grained quartz, biotite, white mica, epidote, albite and sphene. Crossed polars, scale bar = 2 mm. Specimen 31837. All thin-sections are lodged at the Department of Geology, James Cook University, Townsville.

THE SHACKS MYLONITE ZONE

103

Figure 9 Basement geology of the Sabina Point Shear Zone, a high-strain zone of schist separating schist and gneiss of the Broome Head Metamorphics on the north from low-grade quartz arenite of the Shoalwater Formation to the south. Much of the map area is covered by Cenozoic laterite and sand.

of unknown but probable Late Permian to Triassic age intrude and terminate the shear zone, and a small undated latite intrusion forms Pine Mountain (Kirkegaard et al. 1970) (Figure 9). These intrusive rocks have caused minor contact metamorphism of the rocks in the shear zone. The Sabina Point Shear Zone consists of quartz-mica schist with the metamorphic assemblage of quartzbiotite-muscovite-opaques ± plagioclase, with accessory tourmaline and zircon. One schistosity/metamorphic lamination, S s , and a lineation L, are well developed in all outcrops. S s in some outcrops is observed to be the axial plane to small-scale isoclinal folds of either quartz veins or an earlier schistosity/layering, presumably the dominant S 2 / S 3 schistosity in the Broome Head Metamorphics (Morand 1993a). This is more evident in thin-section, where remnants of an earlier schistosity are commonly seen, and uncommon rootless folds have S s as axial plane schistosity. These folds are parallel to L, which is a rodding of quartz-rich

layers and an alignment of mica and elongated quartz grains on mica-rich layers. L is interpreted as a stretching lineation, as the foliation is more intense in sections parallel to L than in sections perpendicular to it. A second lineation L min observed in some outcrops is an alignment of biotite crystals. It is locally parallel to L, but usually pitches about 10-20° away from L. L min may be a relic of the L 3 lineation that is prominent in the Broome Head Metamorphics (Morand 1993a). S s varies in strike between southeast and northeast and dips mainly at a moderate angle to the south, while L pitches at a low angle to both east and west (Figures 9, 10). Most rocks do not show mylonitic fabrics, but appear to be schists both in outcrop and thin-section, although recrystallisation in the vicinity of post-tectonic intrusions may have destroyed such features. Nevertheless, these schists form a high-strain zone separating rock units of markedly different metamorphic grade, and a few samples show probable shear bands with a sinistral sense of shear. The generally low plunge of the

Figure 10 Lower hemisphere equal-area projections of lineations (L and L j ) , a mesoscopic fold axis and poles to structural surfaces (S s and S c r e n ) of the Sabina Point Shear Zone. m

n

104

V. J. M O R A N D

EMILY PEAK FAULT

Zone along the Emily Peak Fault is dextral in sense, this could have resulted from pure dip-slip or pure strikeslip displacement, or some combination of both. If the movement was purely dip-slip, then the shear sense would have been east-side-up. The amount of vertical displacement needed to produce the 6.5 km strike separation is 6 km—a seemingly large amount which should result in a marked difference in metamorphic grade either side of the fault. Although the hangingwall side of The Shacks Mylonite Zone is amphibolite facies east of the Emily Peak Fault and greenschist facies west of this fault, consistent with east-side-up movement, no such grade change is evident in the north part of Figure 3, where rocks on both sides of the Emily Peak Fault are greenschist facies. Thus pure dip-slip movement is not favoured for the Emily Peak Fault and no outcrops of the fault plane have been found which might help determine the type of movement it has undergone. Pure dextral strike-slip displacement along the fault is not likely either, because none of the rock units on the west side can be matched with units on the east side, suggesting that units on one side of the fault have been uplifted and eroded away. Therefore a dominantly dextral strike-slip displacement with a component of east side up movement is favoured for the youngest movement on the Emily Peak Fault.

The Shacks Mylonite Zone is offset with a dextral strike separation of about 6.5 km by the north-striking Emily Peak Fault (Figures 2, 3), one of the younger faults in the region. To the north this fault separates Carboniferous strata and low-grade Shoalwater Formation on the west from the Mt Holly Formation on the east, but in the area of Figure 3 it is mostly covered by coastal swamp. At Campbell Peak, in the centre of Figure 3, a body of tourmaline-bearing pegmatite with accessory garnet trends parallel to the fault. Much of this pegmatite contains a steep, north-striking foliation which appears to have formed by movement along the Emily Peak Fault. The steep dip of the foliation implies a steep dip on the Emily Peak Fault. Foliation in the pegmatite is mylonitic, and in a few outcrops S-C fabrics are developed, with C dipping steeply east and S dipping moderately east, indicating east-side-down or normal movement. Evidence for this movement sense is also seen on the axial planes of F 6 folds in the Broome Head Metamorphics (Morand 1993a) and suggests that the same deformation was responsible for the F 6 folds and movement on the Emily Peak Fault and also the development of S cren in The Shacks Mylonite Zone and the Sabina Point Shear Zone, as discussed above. It should be noted that east-side-down, dip-slip movement is not compatible with the outcrop pattern of The Shacks Mylonite Zone, for this would result in a sinistral strike separation along the Emily Peak Fault, rather than the observed dextral apparent offset. Thus it is inferred that the map pattern results from a separate episode of movement along the fault, presumably postdating that correlated with F6. Although the apparent offset of The Shacks Mylonite

Movement on The Shacks Mylonite Zone occurred while the Broome Head Metamorphics were still hot, as evidenced by amphibolite facies assemblages defining the mylonite foliation and lineation. The tightening of D4 structures and the gradation from complex structures of the metamorphics into the simple mylonite zone structure (Figure 5) indicates that shearing took place during or after D4, but before D6 which deformed the mylonite zone fabrics. Metamorphic grade decreases across the mylonite zone into the greenschist facies Mt Holly Formation, so it appears that the Broome Head Metamorphics were brought up from depth during or soon after peak metamorphism and thrust over the lower grade rocks of the footwall. Relative timing of movement on the Sabina Point Shear Zone is less well constrained, but it also must post-date much of the complex folding of the Broome Head Metamorphics, as well as pre-dating the east-west shortening of D6. The Broome Head Metamorphics were last at amphibolite facies temperatures during the HunterBowen Orogeny in the Late Permian (Leitch et al. 1993), suggesting that The Shacks Mylonite Zone and the Sabina Point Shear Zone were active during this orogeny. None of the massive, post-tectonic granite and diorite intrusions shown on Figures 3 and 9 have been dated, but they are thought to be Late Permian to Early Triassic (Kirkegaard et al. 1970), the age of most plutons in this part of the New England Fold Belt. This places a probable younger age limit of Early Triassic on shear-zone movement.

stretching lineation indicates a dominantly strike-slip movement sense, but some outcrops have a dip-slip sense, possibly suggesting more than one episode of movement along the shear zone. S s is folded into open folds trending approximately east, thus causing variations in its dip (Figure 9) and a scattering of poles along an ill-defined north-south girdle (Figure 10). A few such folds were observed in outcrop, but they must also be larger than outcrop scale, causing dip variations over hundreds of metres (Figure 9). A later set of structures has folded S s into open, upright, north-northwest-trending folds a few kilometres in wavelength. These structures also occur as open, locally kink-like, mesoscopic folds with a wavelength of about 1 cm, and a crenulation cleavage Scren as axial plane fabric. S cren has a generally steep dip and north strike (Figure 10). A weak girdle of Ss poles has a P axis plunging gently south (Figure 10), indicating the orientation of the late folds. These latest folds have the same style and orientation as the latest folds affecting The Shacks Mylonite Zone and are also correlated with F6 in the Broome Head Metamorphics (Morand 1993a).

TIMING OF FAULT MOVEMENT

THE SHACKS M Y L O N I T E ZONE TECTONIC IMPLICATIONS The Shacks Mylonite Zone is a major oblique dextral thrust fault that has transported amphibolite facies rocks buried at 15 to 20 km depth (Morand 1993a) over greenschist facies rocks. It is the most important of a series of northeast-trending thrusts in the Stanage B a y Percy Isles region (Leitch et al. 1994a, b). This crustalscale shear zone has brought the Shoalwater terrane into contact with the Yarrol terrane, these two terranes being elsewhere separated by the Wandilla terrane (Figure 1).

Figure 11 Palinspastic reconstruction of the northern New England Fold Belt during the Permian, prior to the HunterBowen Orogeny, showing main palaeogeographic units. The present positions of Mackay (M), Rockhampton (R) and Gladstone (G) are shown for comparison with Figure 1. SFZ, Stanage Fault Zone; BHM, Broome Head Metamorphics; MT, Marlborough terrane.

The Sabina Point Shear Zone appears to be a sinistral strike-slip shear zone, although most strike-slip faults have a steep dip. Its current moderate dip to the south may not be the original dip, as it has been folded by both east-trending and north-trending folds. If it was a sinistral strike-slip fault, and was active concurrently with The Shacks Mylonite Zone, then the Broome Head Metamorphics have moved as a thrust sheet toward the west-southwest, with the two bounding shear zones acting as complex transfer faults with opposite senses of strike-slip movement. Thrusting in this part of the New England Fold Belt during the Hunter—Bowen Orogeny has resulted in major disruption of terranes, with the Marlborough terrane being thrust in a southwest

105

direction (Murray 1974) over the Yarrol terrane (Figure 1) toward the Gogango Overfolded Zone. The last-named resulted from thrusting in the same direction (Fergusson 1991) and curves around to become continuous with the Stanage Fault Zone (Figure 1). This continuity has been modified by several kilometres of late sinistral displacement along the Yarrol Fault (Morand 1993c). The eastward outstepping of Yarrol terrane rocks north of the Stanage Fault Zone (Figure 1) was explained by Leitch et al. (1994b) as a promontory on the initial coastline of the active continental margin during the Late Palaeozoic, with the Stanage Fault Zone initiated along the southeast side of this promontory. An alternative explanation is offered here, whereby the offset of palaeogeographic features across the Stanage Fault Zone was caused by major displacement along the fault. The dextral strike-slip component on The Shacks Mylonite Zone is assumed to be typical of the Stanage Fault Zone as a whole and this information can be used to estimate the amount of displacement along the Stanage Fault Zone. On Figure 2 the Stanage Fault Zone is shown running east of the Percy Isles. The Marlborough terrane rocks on South Percy Island can thus be linked up with the rest of the terrane near Marlborough if dextral movement along the Stanage Fault Zone is reversed. This gives a displacement along the fault zone of about 100 km. Figure 11 is a palinspastic reconstruction of the northern New England Fold Belt in the mid-Permian, prior to the Hunter—Bowen Orogeny. Shortening is 60% in the Folded Zone of the Bowen Basin and 70% in the Gogango Overfolded Zone (Fergusson 1991); crosssections in Morand (1993c) indicate 15% shortening in most of the Yarrol terrane and 40% in the eastern part of this terrane. Yarrol terrane rocks north of the Stanage Fault Zone are generally weakly deformed with dips less than 30° (Fergusson et al. 1994), and shortening for these rocks is estimated at 15%. Shortening in the Connors-Auburn Arc is estimated at 10%, while the accretionary prism (Wandilla and Shoalwater terranes) is estimated to have been shortened at least 50% during the Hunter-Bowen Orogeny, on the basis of strongly developed cleavage in both terranes and multiple folding in the Shoalwater terrane (Fergusson et al. 1990). The western limit of major Hunter—Bowen contraction is the Folded Zone of the Bowen Basin (Fergusson 1991) in the north, the Gogango Overfolded Zone in the centre, and the western edge of the Auburn magmatic arc in the south. Using these shortening estimates, strata have been unfolded perpendicular to the trend of regional folds and cleavage to generate Figure 11. Large amounts of shortening due to thrusting and cleavage development south of the Stanage Fault Zone compared with generally weakly deformed rocks north of the fault zone (Campwyn Volcanics), are compatible with large displacement along this fault, with the Marlborough terrane having been thrust over the Yarrol forearc basin, which in turn was thrust over the Connors-Auburn magmatic arc along the Gogango Overfolded Zone. The Broome Head Metamorphics are

106

V. J. M O R A N D

shown as originating well within the outboard Shoalwater terrane. The reconstruction in Figure 11 shows a weak promontory or bend in the arc and forearc region southeast of the Stanage Fault Zone, but this does not seem significant enough to initiate a major fault, as required by Leitch et al. (1994b). Furthermore, the bend occurs to the southeast of the Stanage Fault,

Figure 12 Distribution of plutons in the Queensland part of the New England Fold Belt (after Gust et al. 1993 and Geological Survey of Queensland 1975). Major structures shown are Stanage Fault Zone (SFZ), Gogango Overfolded Zone (GOZ), Yarrol Fault (YF) and North Pine Fault (NPF) which forms the western boundary of the Gympie Province.

whereas Leitch et al. (1994b) require the promontory to be northwest of the fault. North of the Stanage Fault Zone, the eastern boundary of the Yarrol terrane is not exposed, and hence its position on Figure 11 is speculative. It is shown as running parallel to the arc from Marlborough terrane rocks on South Percy Island, as this terrane occurs on the eastern boundary of the Yarrol terrane south of the Stanage Fault Zone. Figure 11 shows the Wandilla terrane terminating near the Stanage Fault Zone. Granite plutons obscure the area where this terrane would be expected to crop out north of Marlborough (Figure 2), but it is likely that the Broome Head Metamorphics and other units of the Shoalwater terrane have been thrust over the Wandilla terrane here. The Wandilla terrane may have been faulted out in this area by the Gogango—Baryulgil Transform (Murray et al. 1987), which is postulated to have run through this area in the Late Carboniferous, or alternatively it may have continued northwards, outboard of the Yarrol terrane. Recent work on the Bowen Basin (Fielding et al. 1994) indicates that Bowen Basin strata overlapped the Connors—Auburn Arc and much of the Yarrol forearc basin during back-arc extension in the Early Permian, but these strata have been omitted from Figure 11 in order to show the underlying basement features. In Figure 11 the continental magmatic arc is shown as extending westward under the Bowen Basin, based on cross-sections in Fergusson (1991), and is about 100 km wide. Seaward of the arc is the forearc basin (100— 150 km wide) and the accretionary prism (perhaps 200 km wide). These values are typical of present-day convergent margins such as Sumatra, where the a r c trench gap is up to 300 km (Hamilton 1988). It can be seen from Figure 11 that the Stanage Fault Zone is a major tear fault or complex lateral ramp, running at a high angle to the general trend of the fold belt, with major strike-slip displacement. As a result of this generally southwestward movement, the rocks of the region were deformed into several large, complex thrust sheets, all with southwestward transport: the Broome Head Metamorphics, bounded laterally by The Shacks Mylonite Zone and the Sabina Point Shear Zone; the Marlborough terrane thrust sheet (Murray 1974, Fergusson 1991); and the Gogango Overfolded Zone carrying part of the Yarrol terrane as a thrust sheet (Figure 1). Figure 1 also shows possible Shoalwater terrane continuing northwest from the Percy Isles, where this terrane is exposed on Northeast Island (Leitch et al 1994b), outboard of the Yarrol terrane. No Palaeozoic rocks are exposed northwest of the Percy Isles, but the continental shelf is over 200 km wide here, leaving room for more New England Fold Belt rocks to occur. Not only does the Stanage Fault Zone mark the northern limit of intense deformation in the outcropping northern New England Fold Belt, but it also appears to mark the northern limit of Late Permian to Triassic (i.e. Hunter-Bowen) plutons in the fold belt (Figure 12). Gust et al. (1993) used this pattern of pluton ages to suggest that the Connors Arc is not part of the New

England Fold Belt, which they claimed stops at the northern limit of Hunter-Bowen plutons. Figure 11 shows that the Connors and Auburn Arcs formed a continuous magmatic arc prior to the Hunter-Bowen Orogeny, and the middle part of this arc was overridden by thrust sheets along the Gogango Overfolded Zone. Thus the New England Fold Belt continues well to the north of Mackay, as does its associated foreland basin, the Bowen Basin (Fielding et al 1994). To explain the pluton age distribution of Figure 12, two possibilities are entertained. The first is that Hunter-Bowen plutons simply continue northwestward from the Stanage Fault Zone, parallel to the coast but offshore (no Palaeozoic rocks are exposed on the islands in this region). The second is that HunterBowen plutons do in fact stop at the Stanage Fault Zone. In this case the plutons must be directly related to crustal thickening due to folding and thrusting, or else they are subduction-related (Gust et al 1993; Kleeman 1988; Bryant & Arculus 1993) and the Stanage Fault Zone marks the northern limit of subduction during the Hunter-Bowen Orogeny, or at least an outstepping of the subduction zone along a transform fault. In the absence of detailed aeromagnetic data in this offshore region, it is difficult to say which alternative is most likely. The cause of the Hunter-Bowen Orogeny is not clear, but large-scale folding and thrusting of terranes along an active continental margin suggest the likelihood of collision with a crustal block brought into the subduction zone from the ocean to the east. Mediumpressure series metamorphism in the Broome Head Metamorphics is also consistent with collisional orogenesis (Morand 1993a). As the Hunter-Bowen plutons are post-tectonic, and are thought to have formed by subduction processes (Gust et al 1993), it appears that subduction resumed soon after the collision. Thus the colliding body must have been relatively small in width, although well over 1000 km long, and embedded within a largely oceanic plate—perhaps an island arc or oceanic plateau. The Stanage Fault Zone may have formed in response to a rigid indentor within the colliding mass causing maximum deformation, overthrusting and exhumation (Gogango Overfolded Zone, Marlborough thrust sheet, Broome Head Metamorphics) where it impinged on the continental margin. The northern limit of the indentor coincided with the Stanage Fault Zone, which may have been a reactivated basement fault. The identity of the colliding body is unknown—it presumably lies under the continental shelf or offshore plateaus—but one possibility is one or more of the constituent terranes of the Gympie Province. The collision of this body was sufficient to tear the fabric of the northern New England Fold Belt and cause intense deformation, metamorphism and uplift. CONCLUSIONS The Shacks Mylonite Zone is part of the Stanage Fault Zone, which runs in a generally north-northeast

THE SHACKS M Y L O N I T E ZONE 107 direction, at a high angle to the regional trends of the northern New England Fold Belt. It crops out as a 3 kmwide zone of strongly sheared and mylonitic rocks showing oblique dextral thrust movement in which the Broome Head Metamorphics have been thrust over the low-grade Mt Holly Formation. A related shear zone, the Sabina Point Shear Zone, bounds the Broome Head Metamorphics to the south and appears to have moved sinistrally. Elements of the Marlborough terrane have been displaced by about 100 km along the Stanage Fault Zone, and this movement is linked to major thrusting along the Gogango Overfolded Zone. Fault movement occurred during the Hunter-Bowen Orogeny in the Late Permian to Early Triassic. The Stanage Fault Zone is interpreted as a major tear fault formed by the impingement of a rigid indentor of the colliding body responsible for the Hunter-Bowen Orogeny in the northern New England Fold Belt. ACKNOWLEDGMENTS This work was supported by ARGS Grant A38615703 and James Cook University Special Research Grant 2790. Discussions with Chris Fergusson, Evan Leitch and Bob Henderson have contributed to ideas on the New England Fold Belt, but I take full blame for any weird concepts in this paper. Chris Fergusson and Robin Offler provided useful reviews. REFERENCES BRYANT C. J. & ARCULUS R. J. 1993. Geochemistry of the

Clarence River Suite granitoids. In: Flood P. G. and Aitchison J. C. eds. New England Orogen, Eastern Australia, pp. 3 4 9 - 3 5 2 . Department of Geology and Geophysics, University of New England, Armidale. FERGUSSON C. L. 1991. Thin-skinned thrusting in the northern New England Orogen, central Queensland, Australia. Tectonics 10, 797-806. FERGUSSON C. L., HENDERSON R. A. & LEITCH E. C. 1990.

Structural history and tectonics of the Palaeozoic Shoalwater and Wandilla terranes, northern New England Orogen, Queensland. Australian Journal of Earth Sciences 37, 387-400.

FERGUSSON C. L., HENDERSON R. A. & WRIGHT J. V. 1994.

Facies in a Devonian—Carboniferous volcanic fore-arc succession, Campwyn Volcanics, Mackay district, central Queensland. Australian Journal of Earth Sciences 41, 287-300.

FIELDING C. R., HOLCOMBE R. J. & STEPHENS C. J. 1994. A

critical evaluation of the Grantleigh Trough, east-central Queensland. In: Holcombe R. J., Stephens C. J. and Fielding C. R. eds. Capricorn region central coastal Queensland, 1994 Field Conference, pp. 1 7 - 3 0 . Geological Society of Australia, Queensland Division, Brisbane. GEOLOGICAL SURVEY OF QUEENSLAND 1 9 7 5 . Queensland Geology, Scale 1:2 500 000. Department of Mines, Brisbane.

108

V. J. MORAND

GUST D. A.,

STEPHENS C . J. & G R E N F E L L A . T .

1993.

Granitoids of the northern NEO: their distribution in time and space and their tectonic implications. In: Flood P. G. and Aitchison J. C. eds. New England Orogen, Eastern Australia, pp. 565—572. Department of Geology and Geophysics, University of New England, Armidale. HAMILTON W. B. 1988. Plate tectonics and island arcs. Geological Society of America Bulletin 100, 1503-1527. HENDERSON R . A . , FERGUSSON C . L., LEITCH E. C . , MORAND V . J., REINHARDT J. J. & CARR P. F. 1993. T e c t o n i c s o f t h e

northern New England Fold Belt. In: Flood P. G. and Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 505—515. Department of Geology and Geophysics, University of New England, Armidale. KLEEMAN J. D. 1988. Constraints from granite intrusives and felsic extrusives on the tectonics of the southern New England Orogen in the Late Palaeozoic-Early Triassic. In: Kleeman J. D. ed. New England Orogen Tectonics and Metallogenesis, pp. 129-133. Department of Geology and Geophysics, University of New England, Armidale. KIRKEGAARDA. G . , SHAW R . D . & MURRAY C . G . 1 9 7 0 .

Geology of the Rockhampton and Port Clinton 1:250 000 Sheet areas. Geological Survey of Queensland Report 38. LEITCH E. C . , MORAND V . J., FERGUSSON C . L., HENDERSON

R. A. & CARR P. F. 1993. Accretion and post-accretion metamorphism in subduction complex terranes of the New England fold belt, eastern Australia. Journal of Metamorphic Geology 11, 309-318. LEITCH E. C . , FERGUSSON

C . L . , HENDERSON R .

A.

&

MORAND V. J. 1994a. The Late Palaeozoic arc flank and fore-arc basin sequence of the New England Fold Belt in the Stanage Bay region, central Queensland. Australian Journal of Earth Sciences 41,301—310.

LEITCH E . C . , FERGUSSON

C . L . , HENDERSON R .

A.

&

MORAND V. J. 1994b. Ophiolitic and metamorphic rocks in the Percy Isles and Shoalwater Bay region, New England Fold Belt, central Queensland. Australian Journal of Earth Sciences 41, 571-579. LISTER G. S. & SNOKE A. W. 1984. S - C mylonites. Journal

of

Structural Geology 6, 617-638. MORAND V. J. 1993a. The Broome Head Metamorphics: high grade metamorphism in the northern New England Orogen. In: Flood P. G. and Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 591-598. Department of Geology and Geophysics, University of New England, Armidale. MORAND V. J. 1993b. Stratigraphy and tectonic setting of the Calliope Volcanic Assemblage, Rockhampton area, Queensland. Australian Journal of Earth Sciences 40, 15-30. MORAND V. J. 1993c. Structure and metamorphism of the Calliope Volcanic Assemblage: implications for Middle to Late Devonian orogeny in the northern New England Fold Belt. Australian Journal of Earth Sciences 40, 257-270. MURRAY C. G. 1974. Alpine-type ultramafics in the northern part of the Tasman Geosyncline: possible remnants of Palaeozoic ocean floor. In: Denmead A. K., Tweedale G. W. and Wilson A. F. eds. The Tasman Geosyncline: A symposium, pp. 161-181. Geological Society of Australia, Queensland Division, Brisbane. MURRAY C . G . , FERGUSSON C . L., FLOOD P. G . , WHITAKER W .

G. & KORSCH R. J. 1987. Plate tectonic model for the Carboniferous evolution of the New England Fold Belt. Australian Journal of Earth Sciences 34, 213—236. (Received 10 April 1996; accepted 17 August 1996)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 109-127.

Volcanic setting of the Mt Morgan Au-Cu deposit, central Queensland: implications for ore genesis P. R. MESSENGER,1* S. D. GOLDING1 AND A. TAUBE2 1

Department of Earth Sciences, University of Queensland, Qld 4072, Australia. Mount Morgan Joint Venture, PO Box 72, Mt Morgan, Qld 4714, Australia.

2

Low-K rhyolitic volcanics of arc derivation comprise the lower part of the Middle Devonian Capella Creek Group which forms the Mt Morgan Mine Corridor and is exposed extensively along the Dee Range in central Queensland. New research is presented which provides a reinterpretation of these rocks and their significance to the volcanic setting of the host sequence to the Mt Morgan Au-Cu deposit. The predominant lithofacies association at Mt Morgan comprises very thick (-10 to >110m), massive to crudely graded, crystal-rich, pumiceous volcaniclastic mass-flow deposits. These represent subaqueously emplaced, syneruptive deposits from contemporary pyroclastic eruptions, and/or resedimented deposits of non-welded primary pyroclastic detritus. Subordinate lithologies include detrital limestone, jasper, chert and tuffaceous mudstone, siltstone and sandstone. The fossil assemblage of fine-grained sedimentary rocks indicates a marine depositional environment. Numerous small (<1 km diameter), dome-like quartz-plagioclase porphyry intrusions cut this sequence. The porphyry intrusions are mineralogically and chemically similar to the volcaniclastic mass-flow deposits. Porphyry contacts with the host rocks are highly irregular and exhibit textures ranging from delicate interpenetration of jasper, siltstone and porphyry (ragged peperite) to quench fragmented porphyry within tuffaceous sandstone-breccia (blocky peperite). In both peperite types, discolouration of adjacent sedimentary rocks reflects thermal induration during porphyry emplacement. Peperitic contacts indicate that porphyritic magma intruded wet, unconsolidated sediments and is therefore broadly coeval with the pyroclastic eruptions that sourced the enclosing volcaniclastic mass-flow deposits. A distinctive porphyry type occurs in intimate association with predominantly fine-grained chemical and clastic sedimentary rocks in the 200 m thick Banded Mine Sequence at Mt Morgan. Clasts of porphyry up to 2 m long are enclosed by jasper, chert and tuffaceous mudstone or siltstone. Many porphyry clasts are flattened with shapes resembling fiamme. Importantly, the adjacent host rocks lack evidence of thermal induration but commonly show the effects of soft sediment deformation. The morphology, groundmass textures and lithofacies association of these porphyry clasts is most consistent with their interpretation as variably, diagenetically compacted, vesiculated and non-vesiculated lava blocks. It is postulated that these blocks were sourced from one or more submarine lava domes and deposited proximally from suspension into a sequence of fine-grained, chemical and clastic sediments. Mineralogical and chemical similarities between the rhyolitic lithofacies, together with their close spatial and temporal relationships suggest that they are genetically related. A volcanic facies model is presented which provides an interpretation of the rhyolitic host sequence to the Mt Morgan Au-Cu deposit in terms of a proximal, intrabasinal, effusive-explosive facies association related to a sub volcanic trondhjemitic magma chamber which eventually intruded its own eruptive pile. The spatial and temporal association between low-K rhyolitic volcanism and exhalite formation at Mt Morgan provides compelling evidence for a genetic link between rhyolitic volcanism and hydrothermal activity, thereby providing a new perspective for the evaluation of genetic models of Au-Cu mineralisation at Mt Morgan. Key words: gold, copper, mineralisation, mass-flow deposits, peperite, rhyolite, trondhjemite, volcanism.

INTRODUCTION

The Mt Morgan Mine, situated in central Queensland (Figure 1), produced 250 t of Au and 360 000 t of Cu between 1882 and 1989 (Golding et al. 1993) and remains one of the most significant mines in Australia in terms of total gold production. Mineralisation was hosted by a belt of porphyritic and crystal-bearing (quartz + plagioclase) volcanic rocks of low-K rhyolite composition with subordinate exhalites and limestones of Eifelian age (Fordham & Taube 1994). The nature of the rhyolitic lithologies at Mt Morgan has been a matter of debate with genetic interpretations ranging from intrusive (Reid 1947; Cornelius 1969) to pyroclastic

origins (Hawkins & Witcher 1961; Frets & Balde 1975; Taube 1986; Arnold & Sillitoe 1989). Sixty percent of the ore at Mt Morgan was mined at or about the level of the Banded Mine Sequence which comprises a mixture of porphyry, jasper, chert, and tuffaceous siltstone and mudstone. Complex textures within this sequence have proven to be particularly problematic. Cornelius (1969) interpreted the sequence as a succession of tuffs and lavas, whereas Frets (1970) suggested that synsedimentary gel retexturing, as proposed by Elliston (1969), could account for most of * Present address: Great Central Mines Limited, 46-50 Kings Park Road, West Perth, WA 6005, Australia.

110

P. R. M E S S E N G E R £ T y 4 Z , .

New England Fold Belt

X.20

Study area

j V-r, 1000 km

egBSB

Burnett Highway UND 17 77/8

kilometres 0

5

Volcano-plutonic stratigraphy

Devoniar

+

Mt Morgan Tonalite

+ [

upper Capella Creek beds Quartz-plagioclase porphyry lower Capella Creek beds

Limestone Banded Mine/Banded Mineralised Sequence

+

wmmm M M— I mim

|

Specific lithologies

Permian granitoids

X

f

000 mE

"^•V-

X

f

240

-

Alteration zone/ base metal prospect

V20

Bedding orientation Fault Anticline/syncline Major road

N

A

liilllil 31 km to Mt Alma

Figure 1 Locality map and general geology of the Mt Morgan — Dee Range area showing the distribution of major rock units. Locality marked A indicates the position of peperite in Figure 7. Coordinates refer to Australian Metric Grid Zone 56.

the textural ambiguity of these rocks. The nature of the host lithologies at Mt Morgan is central to the more general debate regarding the origin of the Au-Cu deposit itself, for which the two main theories are: (i) a Devonian volcanogenic massive sulfide pipe deposit (Paltridge 1967; Frets 1974; Gibbons 1974; Frets & Balde 1975; Lawrence 1977; Golding & Wilson 1981; Fedikow & Govett 1985; Taube 1986; Large, 1992; Golding et al. 1993); and (ii) a structurally controlled Devonian replacement body related to the Mt Morgan Tonalite (Newman & Campbell-Brown 1911; Fraser 1914; Staff Mt Morgan Limited 1965; Arnold & Sillitoe 1989). Although the host lithologies have undergone texture-

destructive alteration within a subvertical alteration pipe, lithologies away from this pipe retain relic primary textures. The aim of this paper is to provide detailed lithological descriptions of the host sequence to the Mt Morgan Au-Cu deposit drawn from outcrop mapping, drillcore logging and thin-section studies, and to evaluate its palaeo-volcanological setting. Summary geochemical data are presented to aid interpretation of lithofacies relationships; however, a more detailed geochemical discussion will be presented elsewhere. The observations and interpretations presented here provide constraints on the depositional environment of the host succession at Mt Morgan and a basis for the evaluation of ore genesis theories.

GEOLOGICAL SETTING

MT MORGAN Au-Cu DEPOSIT, QLD

Fault blocks comprising Early to Middle Devonian volcano-sedimentary sequences are the oldest recognised elements of the northern New England Fold Belt in central Queensland. These blocks consist of rhyolitic and lesser mafic to intermediate volcanic successions interbedded with subordinate coralline limestone and jasper/chert horizons. In the Mt Morgan area, the sequence was intruded by the polyphase Mt Morgan Tonalite, which together with the volcanosedimentary successions were termed the Calliope Volcanic Assemblage by Morand (1993). Regional geology of the area was described by Kirkegaard et al. (1970) and Dear et al. (1971). The Capella Creek Group occupies the westernmost

111

fault block of the Calliope Volcanic Assemblage and is exposed in a northwest-trending belt extending from near Mt Usher in the north to south of Mt Alma (Figure 1). This group comprises a lowermost rhyolitic division, correlated with the host sequence of the Mt Morgan Au-Cu deposit by Fordham and Taube (1994), and an upper dacitic division. Fordham and Taube (1994) concluded that limestones in the lower division and at the base of the upper division of the Capella Creek Group are both middle Eifelian (384-382 Ma). Fossil assemblages in limestone and jasper lenses throughout the Capella Creek Group indicate a marine depositional environment. No evidence exists to accurately constrain water depth during deposition of the rhyolitic division, although the lack of structures in

250

500

^ ^

Triassic dyke

[" + |

Mt Morgan Tonalite dacitic lithologies basalt/dacite rhyolite quartz-plagioclase porphyry crystal-rich pumiceous breccia-sandstone

• |

Banded Mine Sequence Lower Mine Siltstone Alteration zone and Au-Cu orebody (exhausted) Faults: accurate approximate inferred

^ ^

A

Peperite locality discussed in text 38-01 Diamond drillhole P collar and trace

v\/ *\

Figure 2 Geology of the Mt Morgan Mine Corridor.

112

P. R. MESSENGER

ETAL.

Table 1 Stratigraphy of the Mt Morgan Mine Corridor Volcanics.

Unit (thickness)

Description

Interpretation

Upper Mine Sequence (750-900 m)

Crystal-rich (quartz + plagioclase), pumiceous breccia-sandstonesiltstone (lithofacies association 1); peperitic quartz + plagioclase porphyries (lithofacies association 2); and aphanitic rhyolite; minor carbonate lenses or blocks (lithofacies association 5).

Crystal and pumice-rich volcaniclastic mass-flow deposits and synsedimentary porphyritic intrusions, aphanitic rhyolite lavas or high-level sills.

Banded Mine Sequence (120-200 m)

Thinly bedded jasper, chert, tuffaceous mudstone-siltstone-sandstone, carbonate lenses, porphyry clasts, (lithofacies associations 3, 5 & 6); peperitic quartz + plagioclase porphyries (lithofacies association 2).

Interbedded exhalites, reworked volcaniclastic turbidites, water-settled ash-fall and pelagic deposits, locally derived vesiculated and non-vesiculated porphyritic lava blocks and synsedimentary porphyritic intrusions.

Middle Mine Sequence ( - 7 0 m)

Crystal-rich (quartz + plagioclase), pumiceous breccia-sandstone-siltstone (lithofacies association 1).

Crystal and pumice-rich volcaniclastic mass-flow deposits.

Lower Mine Siltstone

Massive to diffusely laminated rhyolitic mudstone-siltstone, minor crystal-rich sandstone, banded chert, oxide-sulfide facies and thin carbonate horizons [lithofacies association 3 (mudstonesiltstone dominant interval), & lithofacies associations 4, 5 & 6].

Reworked volcaniclastic turbidites, water-settled ash-fall and pelagic deposits, detrital limestone lenses, exhalites and replacement massive sulfide mineralisation.

Crystal-rich (quartz + plagioclase) pumiceous breccia-sandstone-siltstone (lithofacies association 1); synsedimentary quartz + plagioclase porphyries (lithofacies association 2).

Crystal and pumice-rich volcaniclastic mass-flow deposits and synsedimentary porphyritic intrusions.

(-80 m)

Lower Mine Sequence (>200 m)

fine-grained sedimentary beds indicative of reworking by bottom or surface currents suggests deposition below storm wave-base. By contrast, the more widespread occurrence of coralline limestone and cross-bedded calcareous sandstone in the upper, dacitic division, indicate shallow-water sedimentation following the cessation of rhyolitic volcanism. The Mine Corridor Volcanics (Taube 1986), which are the immediate host to the Mt Morgan Au-Cu deposit, form a north-northwest-trending rhyolitic belt, approximately 6 km long and up to 2 km wide, enclosed by the Mt Morgan Tonalite (Figure 2). Informal subdivisions of the Mine Corridor Volcanics are based on the recognition of mostly fine-grained chemical and/or clastic sedimentary units between thickly bedded, coarser grained volcaniclastic deposits (Table 1). Dacitic lithologies in faulted contact with this stratigraphic sequence immediately south of the Mt Morgan Mine (Figure 2) were termed the Graben Sequence by Taube (1986). These lithologies are excluded from the general stratigraphic succession here, as their primary relationship to the other units is unclear. The Mine Corridor Volcanics and regional correlatives were intruded by synsedimentary, quartz-

plagioclase porphyries and later dykes and irregularshaped volcanic intrusions of basalt to rhyolite, but predominantly dacite composition. All rock types underwent variable alkali metasomatism and propylitic alteration during or soon after emplacement. The Capella Creek Group was intruded subsequently by the multiphase Mt Morgan Tonalite (381 ± 9 Ma 2a: Golding et al. 1994). Field relationships (Messenger & Golding 1986), as well as isotope and trace-element geochemistry (Messenger et al. 1996) indicate a oogenetic relationship between the Capella Creek Group and the Mt Morgan Tonalite. Regional structure along the Dee Range east of Mt Morgan is dominated by the northwest-trending Gracemere Anticline that is characterised by gentle open folds (Figure 1). The structure of the Mt Morgan Mine Corridor is difficult to resolve because of the massive nature of much of the sequence and the presence of steep northwest- and northeast-trending faults. Throughout much of the Mine Corridor bedding appears to be mostly flat lying. By contrast, the immediate mine area forms a faulted, steep-sided halfdome, centred on the ore deposit. This domal structure is cut by arcuate, inward-dipping normal faults and

MT MORGAN Au-Cu DEPOSIT, QLD

major, steeply dipping, northwest- and northeasttrending faults (Figure 2). The arcuate faults are believed to be the oldest known faults in the Mine Corridor and were reactivated with normal dip slip prior to emplacement of a set of northwest-trending Triassic dykes (Figure 2). The northwest- and northeast-trending faults also show evidence of reactivation, with both dip and strike-slip reversal in the case of the Slide Fault (Figure 2) following emplacement of the Triassic dykes. The western half of the dome is obscured by the Mt Morgan Tonalite. No pervasive cleavage is present in the Capella Creek Group and the whole sequence is metamorphosed to a lower greenschist facies assemblage. LITHOFACIES ASSOCIATIONS

The major lithofacies associations of the Mt Morgan Mine Corridor and regional correlatives are described in this section. Table 1 illustrates the relationships between these lithofacies associations and the stratigraphy at Mt Morgan. Lithofacies association 1: crystal-rich, pumiceous volcaniclastic breccia-sandstone DISTRIBUTION

Crystal-rich, pumiceous volcaniclastic breccia and sandstone largely comprise the Lower Mine Sequence, Middle Mine Sequence and Upper Mine Sequence (Table 1, Figure 2). This association dominates the lower Capella Creek Group and at Mt Morgan it is > 900 m thick. It is extensively exposed along the Dee Range from near Mt Usher in the north to south of Mt Hopeful (Figure 1). DESCRIPTION AND PETROGRAPHY

This lithofacies association comprises thick, massive to crudely graded beds, some exceeding 110 m in

p p 21

113

PD 11

100 m -

MF1

200 m.

ra

shear

200 m

Basaltic andesite Dacite Porphyry-jasper-mudstone Crystal-rich pumiceous breccia Volcanic siltstone-sandstone

Lithics

Figure 3 Graphic-logs of crystal-rich, pumiceous mass-flow units MF 1 and MF 2) from the Middle Mine Sequence at Mt Morgan. Note the sharp basal contacts, massive interior, gradational upper contacts and lithic clast concentration zones. Drillhole collar locations are shown on Figure 2; depths are in metres below the collar.

thickness. Beds are tabular with sharp basal contacts. They comprise a massive lower division of crystal-rich pumiceous breccia with or without a relatively thin (<2—3 m) gradational upper division composed of massive or diffusely bedded, tuffaceous sandstone and siltstone (Figure 3). The lower division is poorly sorted and comprises 15-40 vol.% crystals of altered plagioclase (andesine to oligoclase where fresh) and quartz which is extensively embayed and averages

Figure 4 Photomicrographs of pumiceous breccia adjacent to quartz-plagioclase porphyry intrusion in DDH 38-13. Porphyry intrusion lies between 384.6 m and 391.0 m below the collar, (a) At 30 cm below contact, non-welded, blocky tube pumice clasts (arrows) are arranged in random orientations, (b) At 20 cm below the contact, eutaxitic texture is well developed within a zone of secondary welding aligned parallel to the contact. Scale bar is 1 mm long. Plane-polarised light.

114

P. R. M E S S E N G E R

2 mm in diameter. Crystals are set in a microcrystalline matrix composed of quartz + albite + chlorite + opaques ± epidote ± calcite. In least-altered zones, the lower division can be shown to consist predominantly of tube pumice clasts in which tube vesicles are defined by parallel laminae now composed of quartz + plagioclase and non-birefringent phyllosilicates (Figure 4a). These clasts are blocky and rarely exceed 6 cm in length. They are porphyritic with 10-15 vol.% quartz + plagioclase phenocrysts in subequal amounts. Vesicles are noncompacted and the tube pumice clasts have random orientations, suggesting that they are not welded. Although sparsely preserved, tube-pumice clasts are an important original component of this lithofacies assocation. Lithic clasts are common in the basal zone of most beds and can range up to several metres in diameter. Lithic clasts are of volcanic and non-volcanic rock types including quartz-plagioclase porphyry, dacite, limestone, chert and trondhjemite. Immediately adjacent to some quartz-plagioclase porphyry and other intrusions, pumice breccia exhibits a narrow zone of highly attenuated fiamme aligned parallel to the intrusive contact (Figure 4b). This texture was recognised in two drillholes (DDH 38-13, 391 m and DDH 63-06, 106 m) and in both cases is restricted to a zone extending <20 cm from the intrusive contact.

Lithofacies association 2: quartz-plagioclase porphyry DISTRIBUTION

Irregularly shaped quartz-plagioclase porphyry bodies are common along the Dee Range (Figure 1) and in the Mine Corridor (Figure 2). The exposed dimensions of these bodies vary from <50 m to >1 km in longest dimension. At Mt Morgan, the largest exposed porphyry intrusion is ~500 m in diameter (Figure 2). Porphyry bodies are confined to the lower Capella Creek Group with the exception of a group of east-west dykes that cut the upper Capella Creek Group in the north of the study area (Figure 1). Numerous small outcrops of porphyry are exposed in the benches and roads adjacent to the opencut at Mt Morgan. The true extent of porphyry within the now flooded opencut is difficult to determine given the degree of alteration associated with mineralisation; however, it is likely that the quartzporphyry alteration pipe underlying the Au-Cu ore deposit is an intensely altered dome-like porphyry intrusion. Evidence for this interpretation includes the transgressive nature of altered, quartz ± plagioclasebearing rock (>730 m thick) that cuts the lower part of the stratigraphy here and the domal structure of the immediately overlying strata.

INTERPRETATION DESCRIPTION AND PETROGRAPHY

The great thickness and compositional uniformity of this lithofacies association together with bedding characteristics and internal organisation are features most consistent with deposition from high particle concentration, water-supported, pumice-rich volcaniclastic mass flows (Lowe 1982; McPhie et al. 1993). The fossil assemblage of interbedded sedimentary units indicates submarine deposition. The composition and great thickness of single beds, the abundance of blocky pyroclasts, quartz and plagioclase crystals suggest syneruptive deposition sourced from major, contemporaneous, rhyolitic, pyroclastic eruptions and/or resedimentation of non-welded, primary pyroclastic deposits. The variable, generally high, crystal content of this lithofacies association compared with the phenocryst content of both contained pumice clasts and compositionally similar, synsedimentary porphyritic intrusions (Lithofacies association 2) suggest that crystal enrichment and partial loss of vitric ash occurred prior to final deposition (Fiske & Matsuda 1964; Walker 1972; Cas 1983; Cas & Wright 1991). Transport distance from source is poorly constrained but the thickness of the association and individual beds within it indicate vent-proximal deposition which could have occurred within a couple of kilometres or up to a few tens of kilometres from source. The restriction of highly attenuated fiamme aligned parallel to intrusive contacts to a narrow selvedge immediately adjacent to some intrusions suggests that this foliation is a result of secondary welding compaction of pumice clasts caused by heat induced from the adjacent intrusion (McPhie & Hunns 1995).

Quartz-plagioclase porphyry is massive and composed of evenly distributed quartz and plagioclase (albitised) phenocrysts in a non-vesicular, microcrystalline groundmass. Sparsely preserved, isolated spherulites suggest that, in places, the groundmass was originally partly glassy. Euhedral, partially resorbed quartz and plagioclase phenocrysts account for 5—20 vol.% and average 1-2 mm in diameter. Plagioclase is the predominant phenocryst phase. The Baree Porphyry is a textural variant that contains an excess of quartz over plagioclase phenocrysts. Both phenocryst phases of the Baree Porphyry are typically larger than other porphyry bodies with quartz phenocrysts averaging 2-A mm in diameter and plagioclase phenocrysts averaging 1—3 mm in length. Quartz-plagioclase porphyry bodies display mineralogical and chemical similarities (Table 2) to the crystal-rich volcaniclastic mass-flow deposits of lithofacies association 1 and both are interpreted to be genetically related.

CONTACT RELATIONSHIPS

At Mt Morgan, contacts between porphyry and sedimentary lithologies are well exposed in several outcrops (Figure 2, localities A, B, D, E). Many examples of small (< 1 m) lobes and amoeboid porphyry bodies enclosed by fine-grained jasper are exposed on the northern benches of the open cut (Figure 2, locality F). A small outcrop immediately southeast of the Mt Morgan opencut (Figure 2, locality A) shows a

MT MORGAN A u - C u DEPOSIT, QLD

115

Table 2 Average whole-rock geochemistry of major lithofacies types at Mt Morgan.

LAI Pumice breccia «=8

LA 2 Porphyry

Si0 2 Ti0 2 A1203 Fe 2 0 3 MnO MgO CaO Na 2 0 K20 P2O5 LOI Total

76.1 ± 1.3 0.36 ±0.1 12.2 + 0.6 3.0 ±0.3 0.08 ± 0.0 1.8 ±0.5 1.4 ±0.7 4.2 ± 1.4 0.9 ± 0.7 0.05 ± 0.0 1.5 ±0.3 99.8 ± 1.2

Rb Ba Sr Zr Zn La Ce Y Cu Ni Co Cr V Nb U Th Pb Nd Sc

9 ±5.9 114± 111.7 104 ±38.8 135 ± 12 74 ±43.1 6 ±3.3 13 ±5.4 40 ±6.1 7 ± 12.1 1 ±0.4 66 ± 18.5 4+1.8 28 ± 14.3 2 ±0.3 0 ± 0.4 2±1 4±1.4 13 ± 3.7 12 ±2.5

Zr/Ti Zr/Y

0.06 3.38

LA 3 Tuffaceous mudstone n = 10

LA7 Trondhjemite

LA 7 Tonalite

LA 7 Gabbro

n= 4

LA 3 Porphyry clasts* n=2

n-1

n=4

n=3

78.6 ± 0.6 0.2 ± 0 11.45 ±0.2 2.52 ±0.6 0.04 ± 0 0.63 ±0.1 0.35 ±0.1 6.13 ±0.2 0.09 ± 0 0.01 ± 0 0.66 ±0.1 100.15 ±0.3

66.4 ± 6.8 0.3 ±0.1 18.9 ±4.4 4.9 ± 0.9 0.1 ± 0 2.8 ±0.4 0.9 ±0.7 1.2 ±0.8 4.5 ± 1 0.00 ± 0 3.6 ±0.5 99.5 + 1

77.13 ± 2 0.28 ±0.1 11.58 ±0.8 3.14 ±0.9 0.08 ± 0 1.28 ±0.2 0.95 ± 0.3 3.88 ± 1.1 1.63 ±0.8 0.05 ± 0 1.28 ±0.2 100.49 ±0.5

72.75 ± 3.3 0.51 ±0.1 12.88 ±1.1 4.41 ±0.8 0.08 ± 0 1.17 ±0.4 2.87 ± 1.2 4.46 ± 0.7 0.78 ± 0.3 0.08 ± 0 1.56 ±0.6 99.77 ± 0.6

64.46 ± 0.7 0.67 ±0.1 15.60 ±0.7 6.22 ± 0.6 0.12 ± 0 2.53 ±0.5 5.31 ±0.3 4.03 ± 0.2 0.93 ± 0.2 0.13 ± 0 1.44 ±0.3 100.41 ±0.1

50.38 ±1.2 0.51 ±0.2 18.37 ± 1.2 9.58 ±2.4 0.24 ±0.1 6.7 ± 0.6 11.8 ± 1 2.2 ± 0.2 0.17 ± 0.1 0.05 ± 0 1.03 ±0.8 100.86 ± 0

46 ± 1.1 319 ±167.3 44 ±7.1 300 ± 63.4 65 ±0.8 11 ± 1.3 27 ± 10.8 72 ± 16.2 0±0 4 ±3.6 19 ±5.6 2 ±1.9 20 ± 13.6 4 ±0.7 1 ±0.1 4 ±0.5 8±1 24 ±5.1 24 ± 3.3

15 ±8.9 177 ±95.1 79 ±37.1 155 ±24.9 91 ±46.7 6 ±4.2 23 ± 12.7 44 ± 8.8 3 ±4.6 5 ±8.3 60 ±23.1 3±2 11 ± 6.9 2 db 1.1 1 ±0.7 2 ±1.4 10 ± 7 18 ±7.4 13 ±4.7

11 ± 4.7 182 ±70.8 128 ±43.3 135 ± 4 2 47 ±24.1 7 ±5.7 20 ±8.5 35 ± 5 15 ±22.4 2 ± 1.2 67 ± 2 8 8 ±5.9 59 ±28.4 2 ±0.6 1 ±0.6 2 ±0.6 2 ±0.5 14 ±2.4 15 ±4.2

15 ±2.4 245 ± 39.9 300 ± 78.6 105 ±55.7 74 ± 25.6 2 ± 1.7 13 ±2.7 26 ±7.3 31 ±21.5 7 ±3.7 63 ±22.3 21 ±12.1 119 ± 11.7 1 ± 1.2 0 ± 0.4 0 ± 0.5 5 ±3.9 14 ±2.9 20 ±4.8

2 ±1.6 42 ±21.8 240 ± 38 17 ±8.7 144 ± 114. 5 ±5.4 1 ± 1.8 11 ±2.5 49 ± 27.9 37 ± 16.7 56 ±2.1 154 ±62.6 245 ± 72 0±0 0±0 0 ± 0.5 2 ±0.6 4±3 33 ± 6.8

0.17 4.17

0.09 3.52

0.04 3.86

0.03 4.04

1 ±0.1 19 ±6.9 30 ± 7.4 171 ±22.8 60 ±20.1 6 ±0.8 15 ±4.3 59 ± 8 9 ± 16.1 0±0 69 ± 5 1 ± 1.4 1 ±0.5 3 ±0.2 1 ±0.6 2 ±0.8 2 ±0.8 16 ±2.5 11 ± 1.9 0.18 2.9

0.01 1.55

All analyses carried out by X-ray fluorescence at the University of Queensland and are reported ± l a . Major elements are in wt% recalculated on a water-free basis; trace elements in ppm. n, number of samples analysed. Full data tables are available from the senior author upon request. •Note that the degree of alteration of these porphyry clasts has resulted in substantial loss of Si0 2 and Na 2 0, gain of K 2 0 and apparent gain of all other major elements as well as Zr.

porphyry sill immediately underlying a well-stratified rhyolitic siltstone. Although exhibiting gross conformity, the upper sill margin is highly irregular and numerous, small porphyry lobes (<10 cm wide) penetrate the overlying siltstone. Rounded blebs of bleached and indurated siltstone are common within these lobes. Porphyry—siltstone contacts are sharp and highly irregular (Figure 5a, b). The immediately enclosing siltstone is discoloured from purple or bluegreen to black with complete disruption of bedding fabric extending 1-3 cm from the contact (Figure 5a, b), whereas at greater distances, bedding is undisturbed. Porphyry—jasper contacts are typically marked by an alteration of jasper from red to black (Figure 5c).

Another example of a complex porphyry—siltstone contact is exposed at Upper Nine Mile Creek on the Dee Range (Figure 1, Locality A). Here, blocky, curviplanar porphyry clasts showing jigsaw-fit textures are enclosed within purple siltstone (Figure 6). The massive nature of most other intruded sedimentary units precludes the determination of bedding orientations. Nevertheless, the irregular outcrop pattern of most porphyry bodies suggests that they are grossly discordant to bedding. Highly irregular contacts between porphyry and volcaniclastic deposits of lithofacies association 1 are exposed along the Lookout Road (Figure 2, locality D). Here, curviplanar fractures, filled with tuffaceous sandstone, extend into the

116

P. R. M E S S E N G E R E T A L .

Figure 5 Examples of rhyolite porphyry — siltstone and rhyolite porphyry — jasper ragged peperite from Mt Morgan, (a) Polished slab showing rhyolite porphyry (P) contact with well-laminated tuffaceous siltstone (S) at locality A (Figure 2). Note the lack of quench fragmentation of porphyry, the delicate interpenetration of porphyry lobes and the complete disruption and discolouration of bedding immediately adjacent to the porphyry, (b) Rough slab showing porphyry (P) -siltstone (S) contact (arrows) at the same locality, (c) Polished slab showing the contact between porphyry (P), mudstone (M) and jasper (J) from Locality C (Figure 2). Note the delicate nature of porphyry lobe (L) with bulbous head and thinned neck zone. The contact is marked by alteration from blue-green to black mudstone and red to black jasper. Sedimentary layering is disrupted around the porphyry contact. Scale bar is 1 cm long.

porphyry. In addition, bulbous porphyry lobes extend into the tuffaceous sandstone (Figure 7a), isolated porphyry clasts are enclosed by tuffaceous sandstone (Figure 7a) and rounded blebs of tuffaceous sandstone are enclosed by porphyry (Figure 7b). Elsewhere (Figure 2, Locality E), blocky, porphyry clasts exhibit curviplanar fragmentation surfaces and jigsaw-fit texture (Figure 7c). In all cases, tuffaceous sandstone immediately adjacent to the porphyry is massive and altered to a blue-black colour. INTERPRETATION AND DISCUSSION OF PORPHYRY CONTACTS

Quartz—plagioclase porphyry bodies are interpreted to represent high-level, synvolcanic porphyritic intrusions. The interpenetration and mingling of porphyry and sedimentary units at their exposed contacts is interpreted to result from the intrusion of magma into wet, unconsolidated sediment with attendant dewatering and thermal induration of host sediment. Similar mixtures of volcanic and sedimentary components are well documented in the literature (Kokelaar 1982; Busby-Spera & White, 1987; Kano 1991; Hanson 1991; Rawlings 1993; McPhie 1993; Boulter 1993; Hanson & Wilson 1993; Brooks 1995) and are commonly referred to as peperite after Scrope (1827).

At Mt Morgan, peperites show a range of textures defined by the degree and style of porphyry fragmentation. Where porphyry fragmentation is minor and porphyry-sedimentary rock contacts are lensoidal or flame-like and highly interpenetrative, the resulting facies is here termed ragged peperite (Figure 5). Ragged peperite resembles the fluidal peperite described by Busby-Spera and White (1987) in that spalled margins, blocky fragments and jigsaw-puzzle clasts are absent. In contrast to fluidal peperite, ragged peperite lacks the

MT MORGAN Au-Cu DEPOSIT, QLD ^

7a

Cf8><rV c- r

( r

^ /

* ,

*

Lithofacies association 3: jasper-mudstoneporphyry DISTRIBUTION

The most distinctive lithofacies association at Mt Morgan comprises interbeds of ribbon jasper, tuffaceous mudstone, siltstone and sandstone, and porphyry. The major occurrence of this association is within the Banded Mine Sequence (Table 1) which is up to 200 m thick and is known to extend in the subsurface from Figure 7 (a) Outcrop sketch of porphyry—tuffaceous sandstone (lithofacies association 1) contact at Locality D (Figure 2). Note the bulbous protrusion of porphyry into sandstone (arrow) and detached porphyry blocks and blebs within sandstone, (b) Outcrop adjacent to above locality showing porphyry (white) with numerous globular, massive black sandstone inclusions (arrow), (c) Blocky peperite from breccia zone adjacent to massive rhyolite porphyry intrusion locality E (Figure 1). Note the curviplanar fractures and jigsaw-puzzle texture of porphyry clasts (arrows). Hammer in (a) is 30 cm long; pencil in (b, c) is 14 cm long.

x> *

v

*

+ % ML

*

*

extensive development of magma globules. This textural difference is likely to be due to the higher viscosity of rhyolite magma compared to basaltic magma resulting in the suppression of magma - wet sediment mixing. Where porphyry fragmentation is well developed, a breccia comprising blocky, curviplanar porphyry clasts and thermally indurated sedimentary rock is referred to as blocky peperite after Busby-Spera and White (1987). Blocky peperite is more commonly developed at Mt Morgan where the host sedimentary unit is coarse grained and relatively poorly sorted, whereas ragged peperite is restricted to the contact zone between porphyry and fine-grained, well-sorted chemical or clastic sedimentary rocks.

*

/ * I* *

tO

Figure 6 Sample of blocky peperite collected from Upper Nine Mile Creek, Locality A (Figure 1) showing jigsaw-fit texture (arrow) and sediment-filled fractures in porphyry. Purple siltstone (dark) shows complete disruption of bedding fabric. Scale bar is 5 cm long.

J

+SS

+&

T

C-

117

V

+

^

v^

^

*

V

*

+ + *

+ A

*

*

*

+

^

*

118

P. R. M E S S E N G E R £ T y l Z .

DDH 63-03 in the north to DDH 38-13 in the south (Figure 2). Outcrop is restricted to the northern and eastern benches of the opencut and to the area immediately southeast of the opencut. Along the Dee Range, this lithofacies association was recognised in numerous, isolated outcrops extending from a tributary of Crocodile Creek in the north to south of Mt Hopeful and is locally referred to as the 'Banded Mineralised Sequence' (Figure 1). The Lower Mine Siltstone at Mt Morgan is composed of tuffaceous mudstones of this lithofacies association; however, jasper and porphyry are rare or absent. DESCRIPTION

Figure 8 illustrates some of the variations within this

lithofacies association both vertically and laterally. At Mt Morgan, intervals of the Banded Mine Sequence dominated by outsize porphyry clasts interbedded with jasper, chert and tuffaceous mudstone and siltstone alternate with intervals dominated by diffusely to welllaminated tuffaceous mudstone and siltstone with minor jasper lenses and small porphyry clasts (Figure 8). In the Banded Mine Sequence, porphyry clast-dominant intervals range up to 30 m thick, whereas tuffaceous mudstone-siltstone-dominant intervals are typically < 1 0 m thick. The porphyry clast-dominant intervals contain up to 50 vol.% porphyry clasts that range from < 1 cm to ~2 m in length, are lensoid and have ragged terminations (Figure 9). Ribbon jasper lenses rarely exceed 5 cm in thickness at Mt Morgan, are highly contorted and pinch and swell on the scale of tens of

Figure 8 Stratigraphic columns s h o w i n g the Banded Mine Sequence at Mt Morgan and the correlative Banded Mineralised Sequence along the Dee Range. The Mt Usher section was measured in a tributary of Crocodile Creek, locality B (Figure 1), UND 17 and 7 7 / 8 sections are from drillcore at Upper Nine Mile Creek (Figure 1), and the 38-01 section is from drillcore at Mt Morgan (Figure 2). Metres refer to distance above the base of the Banded Mine Sequence or Banded Mineralised Sequence.

MTMORGAN Au-Cu DEPOSIT, QLD

119

Porphyry clasts are evenly porphyritic, containing 5— 12 vol.% quartz and 10-12 vol.% plagioclase (oligoclase to andesine) phenocrysts set in either a siliceous or sericite ± chlorite altered groundmass (Figure 10a, b). Both phenocryst phases are euhedral and partly resorbed. Equant quartz phenocrysts range up to 5 mm in diameter and plagioclase phenocrysts are <3 mm long. Siliceous porphyry clasts contain abundant isolated spherulites suggesting that these clasts were originally glassy. The groundmass of the phyllosilicate-altered porphyry clasts is either massive, composed of fine-grained sericite, or exhibits vesicular textures. Vesicles can be partly flattened (Figure 10c) and are filled with microcrystalline quartz + albite ± iron oxides, whereas vesicle walls are strongly altered to sericite. In some cases, phyllosilicate-altered porphyry clasts contain a rim, <3 cm wide, of siliceous porphyry.

INTERPRETATION

Figure 9 Outcrop at locality C (Figure 2) showing a typical porphyry clast-dominant interval of the j a s p e r - m u d s t o n e porphyry association in the Banded Mine Sequence. Ragged, attenuated porphyry clasts (P) are interbedded with jasper and fine-grained mudstone-siltstone (S). Hammer is 30 cm long.

centimetres to metres. Porphyry-dominant intervals are rare or absent along the Dee Range, and jasper ribbons are typically thicker here, up to 1 m thick. An outcrop of well-stratified tuffaceous siltstone exposed at Mt Morgan (Figure 2, locality A), exhibits some typical features of the tuffaceous mudstone— siltstone-dominant intervals. The siltstone is generally well-sorted but contains < 1 vol.% scattered, subrounded to angular, quartz—plagioclase porphyry clasts (< 3 cm in diameter). Alternating light and dark (1—5 mm) laminations are defined by the relative abundance of fine hematite in the siltstone, whereas elsewhere, bedding laminations are more diffuse. Liquefaction structures, slump folds and bedding-parallel shears, all indicative of soft-sediment deformation, are common and reflect disruption due either to emplacement of an underlying porphyry sill or slope instability. Bedding in the siltstone underlying porphyry clasts is typically disrupted, whereas overlying beds drape the porphyry clasts. There is no discolouration indicative of thermal induration of siltstone adjacent to these clasts.

PETROGRAPHY

Diffuse to well-laminated, grey-green, tuffaceous mudstone-siltstone comprises <1 vol.% angular, unstrained quartz and plagioclase crystal fragments (<450 |um in diameter) set in a fine, granular mosaic of anhedral quartz + albite with disseminated iron oxides and minor chlorite. Jasper ribbons comprise globular hematite and cryptocrystalline quartz in the approximate proportions of 1:4, and typically contain <1 vol.% siliceous sponge spicules and radiolarian fragments.

The homogeneous composition and presence of broken, volcanic quartz and plagioclase crystal fragments in the tuffaceous mudstone-siltstone suggest that these are volcanic deposits probably derived from explosive pyroclastic eruptions. Good sorting of the finer components, parallel continuous bedding at outcrop scale and local mantling of porphyry clasts are consistent with suspension sedimentation. The compositional similarity (Table 2) between the tuffaceous mudstonesiltstone and both quartz-plagioclase porphyry (lithofacies association 2) and crystal-rich pumiceous breccias (lithofacies association 1) suggests that they are cogenetic. It is therefore likely that the tuffaceous mudstone-siltstone was derived either from pyroclastic fallout that subsequently settled through the water column, or from dilute, turbulent suspensions trailing pumiceous mass flows. The coarser porphyry clasts could represent buoyant pumice clasts that floated for some distance before becoming waterlogged and sinking to the site of deposition, or could represent proximally derived, porphyry clasts. The jasper beds of this association were interpreted to be chemical exhalite by Staines (1953), and this is consistent with their composition, texture and bedding characteristics. The microfaunal assemblage of the jasper beds restrict their deposition to a marine environment. The combination of porphyritic textures and sedimentary bedding in this lithofacies association has proved difficult to interpret. Cornelius (1969) considered the association to represent a sequence of striped lavas and tuffs. Frets (1970) proposed a sedimentary origin for the porphyry, suggesting that gel retexturing (Elliston 1969) could produce pseudoporphyritic textures. He conceded, however, that textures typical of gel crystallisation such as ovoid feldspar and quartz crystals, syneresis cracks and conjoined globular crystals were absent. McPhie (1995) interpreted this facies association in terms of mudstone and sandstone containing outsize, compacted clasts of quartz—feldspar-phyric pumice. The euhedral shape and even distribution of quartz and plagioclase

120

P. R. M E S S E N G E R E T A L .

Figure 10 Photomicrographs of the porphyry clasts from the jasper-mudstone-porphyry (lithofacies association 3) in the Banded Mine Sequence at Mt Morgan, (a) Euhedral, partly resorbed and embayed quartz phenocrysts (q) within a sericite-altered, foliated and vesiculated groundmass. Sample collected from the Mt Morgan opencut at AMG 231912E, 7383424N. (b) As above showing partly resorbed, euhedral, glomerophyric plagioclase phenocrysts. (c) Close-up of groundmass showing vesicle walls (W) and partly flattened vesicles (V). Vesicles are filled with fine-grained quartz + albite + chlorite + iron oxides, vesicle walls are altered to sericite. (d) Lenticular porphyry clasts (P) interbedded with jasper (J). Note the thread-like end terminations of porphyry clasts. Sample collected from the Mt Morgan opencut at AMG 232100E, 7383130N. Scale bar is 1 mm long. Plane-polarised light.

crystals in the porphyry clasts of this association strongly suggests that they are igneous phenocrysts (Figure 10a, b). The 'flamme-like' shapes of many porphyry clasts were first noted by Cornelius (1968). These shapes could be a result of shearing and tearing apart of lowviscosity magma during eruption, or of compaction of originally highly vesiculated clasts, an interpretation supported by the partial preservation of vesicular texture (Figure 10c). Compaction of pumice clasts to produce fiamme occurs in subaerial ignimbrites where heat retention and load result in the plastic deformation of juvenile pumice and glass shards (Smith 1960). Fiamme have also been documented in welded pyroclastic fall deposits (Sparks & Wright 1979; Mahood 1984) and in pumiceous deposits adjacent to lava flows or dykes (Christiansen & Lipman 1966; Schmincke 1967; McPhie & Hunns 1995). Fiamme can also form during diagenesis and burial compaction of pumiceous deposits (Fiske 1969; Branney & Sparks 1990). In this case, alteration of original glass to a phyllosilicate or clay assemblage precedes or accompanies burial compaction.

Fiamme can form in this way regardless of the thickness of the pumiceous deposit. Bedding in sedimentary units enclosing the poiphyry clasts is commonly contorted due to differential compaction and soft-sediment deformation but is well preserved and there is no sign of discolouration or thermal induration. This suggests that the porphyry was in a cold state at the time of emplacement and compaction cannot be attributed to welding. These contact relationships, together with porphyry groundmass textures and the degree and style of alteration, distinguish the porphyry clasts of this lithofacies association from ragged peperite (lithofacies association 2). The extensive phyllosilicate alteration, highly attenuated shape and thread-like end terminations of porphyry clasts (Figure lOd) are typical of fiamme formed by diagenetic compaction of pumice (Branney & Sparks 1990) and this explanation is most consistent with their lithofacies association. The presence of irregularly shaped siliceous porphyry clasts containing isolated spherulites indicates that some porphyry clasts were originally composed of dense,

MTMORGAN Au-CuDEPOSIT, QLD

non-vesiculated glass. The occurrence of large clasts of originally vesiculated and non-vesiculated lava within predominantly fine-grained sedimentary rocks suggests proximity to nearby pumiceous lavas or domes. These porphyry clasts could represent vesiculated lava blocks, with or without a quenched glassy rind, that became detached and floated prior to being waterlogged and sinking into the sediments. A similar process was interpreted for the giant pumice block beds within the caldera lake sediments of Sierra La Primavera, Mexico (Clough et al. 1981) and Lake Taupo, New Zealand (Wilson & Walker 1985). Alternatively, McPhie (1995) suggested that porphyry clasts of this association originated as buoyant pumice blocks that entered suspension in the water column from large-volume pumiceous mass flows (lithofacies association 1). Although all rhyolitic lithologies in the lower Capella Creek Group are geochemically similar, subtle variations in the Zr/Ti ratio exist. These data suggest a direct correlation between the porphyry clasts of this lithofacies association and the porphyry intrusions of lithofacies association 2, despite significant major element redistribution (Table 2). It is therefore concluded that the porphyry clasts of this association were derived from porphyry lava domes associated with lithofacies association 2. Lithofacies association 4: mudstone-chert-banded oxide ± sulfide facies

121

INTERPRETATION

The fine-grain size and thin bedding of this lithofacies association suggest a low sedimentation rate. Chert and iron oxide-rich sedimentary units are interpreted to represent suspension-settled, chemical exhalative deposits derived from a nearby, actively venting hydrothermal system. Interbedded mudstones represent ash-rich deposits associated with rhyolitic volcanism. Massive and stringer sulfides are most likely to represent sub-sea-floor replacement mineralisation probably following burial of sea-floor hydrothermal vents by thick, overlying volcaniclastic mass flows of the Middle Mine Sequence. Lithofacies association 5: carbonate facies DISTRIBUTION AND DESCRIPTION

Fossil fragment-rich, variably altered and recrystallised carbonate lenses up to 10 m thick occur throughout the lower Capella Creek Group at Mt Morgan and along the Dee Range. Individual horizons can be traced over several hundred metres at Mt Morgan and typically pinch and swell along strike. Where fossil fragments can be recognised they include parts of crinoids, rugose corals and brachiopods.

DISTRIBUTION AND DESCRIPTION

INTERPRETATION

Well laminated to very thinly parallel-bedded mudstone, chert, magnetite ± sulfide-rich layers (or banded ironformation) and massive sulfide lenses occur at several intervals, a few metres thick, overlying the Lower Mine Siltstone. This lithofacies association was intersected by numerous drillholes some 600 m southeast of the opencut at Mt Morgan. In places, this association is dominated by thin stratabound horizons of high-grade sphalerite. The footwall to this lithofacies association is strongly altered to an assemblage of sericite + chlorite to several hundred metres depth, and contains crosscutting massive and stringer sulfides. By contrast, hangingwall sericite-chlorite-pyrite alteration is weaker and narrow stringer sulfide veins are restricted to <30 m from the top of this association. The major sulfide and oxide ore minerals of this association include pyrite, magnetite, sphalerite, chalcopyrite and pyrrhotite, with minor galena, hematite, leucoxene and tetrahedrite-tennantite. The stratabound mineralisation consists of pyrite-rich bands with anhedral pyrite aggregates, subhedral to euhedral pyrite grains and lesser magnetite and quartz, as well as siliceous pyrite-poor bands. Textures indicative of magnetite replacement by pyrite are typical, although several examples of magnetite replacing pyrite were observed. Sphalerite is most abundant in the stratabound and hangingwall zones, whereas chalcopyrite is most abundant in the footwall to this association (Golding 1995).

The fossil assemblage of this lithofacies indicates derivation from a shallow-marine source. The broken nature of fossil fragments suggests a degree of reworking of these components, which are likely to have been derived from carbonate debris temporarily stored and reworked up-slope prior to resedimentation. Lithofacies association 6: crystal-4ithic volcanic breccia-sandstone DISTRIBUTION AND DESCRIPTION

Purple, blue or green, thin to medium bedded (< 10 cm1 m), moderately well-sorted, crystal (quartz + plagioclase) and/or lithic-rich breccias and tuffaceous sandstones are a common, minor component of the Banded Mine Sequence and Lower Mine Siltstone at Mt Morgan and along the Dee Range. Beds are normally graded with sharp basal contacts and diffusely laminated purple or green siltstone tops. Lithic clasts in basal breccias are angular and generally <2 cm in size. These include quartz-plagioclase porphyry, aphyric dacite and jasperous siltstone. This lithofacies association is most common in drillholes along the Dee Range (e.g. DDH 77/8, Figure 8).

122

P. R. MESSENGER

ETAL.

INTERPRETATION

The bedding characteristics of this lithofacies association are similar to the rhyolitic volcaniclastic mass-flow deposits of lithofacies association 1 suggesting sedimentation by turbidity currents. The small-scale of bedding compared to lithofacies association 1, moderate sorting and medium to fine grainsize suggest that these deposits are reworked and resedimented volcaniclastic deposits. The similar composition to lithofacies association 1 suggests that these deposits represent reworking of crystal- or lithic-rich rhyolitic volcaniclastic deposits temporarily stored upslope. Lithofacies association 7: trondhjemite-tonalitequartz gabbro DISTRIBUTION

A contiguous group of elongate stocks of various compositions intrude the Capella Creek Group in the Mt Morgan district and are collectively known as the Mt Morgan Tonalite. This intrusive complex is elongate north-northwest, has maximum exposed dimensions of 28 x 6 km and completely encloses the Mt Morgan Mine Corridor (Figure 1). DESCRIPTION AND PETROGRAPHY

The Mt Morgan Tonalite displays intrusive contacts with the Capella Creek Group along its eastern margin and at Mt Morgan, Its western margin is obscured by the non-conformably overlying Frasnian Dee Volcanics (Kirkegaard et al 1970; Messenger & Golding 1996), and its northwest margin is obscured by the Jurassic Razorback beds. Radiometric age determinations on tonalite samples (381 ± 9 Ma: Golding et al 1994) together with field constraints suggest that emplacement occurred during the Givetian (381-377 Ma) within a few million years of volcanism. The petrology and petrogenesis of the Mt Morgan Tonalite will be discussed more fully elsewhere, but field relationships and petrographic aspects of the constituent plutons are briefly outlined here. Trondhjemite is the dominant lithofacies of the Mt Morgan Tonalite, whereas tonalite is common to the west of the Mine Corridor, and subordinate mafic plutons occur throughout. Trondhjemite is a sodic granite composed of quartz and plagioclase (An _4 ), with <10% relict hornblende, minor Fe-Ti oxides, zircon and apatite. Trondhjemites are mineralogically and chemically similar to the low-K rhyolites of the lower Capella Creek Group. Grainsize is typically <2 mm and there is a textural spectrum that ranges from hypidiomorphic granular trondhjemite through granophyric, porphyritic trondhjemite to quartzplagioclase porphyry (lithofacies association 2). In general, trondhjemites are moderately to strongly metasomatised with plagioclase altered to albite and hornblende altered to chlorite ± actinolite ± biotite. 38

5

Alteration of the trondhjemites is therefore similar to the quartz-plagioclase porphyries (lithofacies association 2). An irregular, subhorizontal intrusive contact was reported between trondhjemite and the Mine Corridor Volcanics approximately 2 km north of the Mt Morgan opencut, and in places here, both the volcanics and trondhjemite are heavily brecciated, hydrothermally altered and weakly mineralised (Sillitoe 1987). Tonalite is a medium-grained, hypidiomorphic granular rock with quartz, plagioclase (An 3 5_ 5 5 ), green hornblende, Fe-Ti oxides and minor zircon and apatite. Tonalite is typically less altered than trondhjemite and exhibits patchy hydrothermal alteration to an assemblage of chlorite + albite + actinolite ± sericite ± calcite ± biotite ± sphene. Tonalite plutons contain ubiquitous, fine- to medium-grained, rounded to angular, mafic inclusions that range from < 1 cm to approximately 15 cm in diameter. Megaxenoliths of basaltic andesite and dacite up to 10 m in length are exposed in tonalite on the northern side of the Mt Morgan opencut. Cornelius (1968) provided photographic evidence for a sharp intrusive contact between tonalite (his quartz diorite) and mineralised rhyolite in the northwest part of the opencut and numerous examples of relatively fresh tonalite cross-cutting mineralisation at Mt Morgan were intersected by drillholes. At the Baree Prospect, -2.5 km north of the Mt Morgan opencut, relatively fresh tonalite intrudes strongly altered and brecciated rhyolitic volcanics and similarly altered and brecciated trondhjemite (Sillitoe 1987). Several small, quartz gabbro plutons occur within tonalite and trondhjemite and appear to be broadly coeval with tonalite emplacement. Quartz gabbro is medium grained and composed of plagioclase (An o_ ), diopsidic augite, hypersthene, Fe-Ti oxides and minor interstitial quartz. These intrusions typically lack extensive alteration with generally fresh pyroxenes and plagioclase. 6

75

INTERPRETATION

The medium grainsize, complete lack of foliation of all plutons and association with chemically similar effusive rocks indicate emplacement into a low-stress, subvolcanic environment. The elongate nature of the complex and individual stocks suggests a strong structural control on emplacement. The chemical and mineralogical similarity between trondhjemite, porphyritic trondhjemite and quartzplagioclase porphyry (Table 2), and the textural spectrum displayed by these rocks suggests a genetic relationship between all three and indicates that the present level of exposure preserves the transition zone between volcanic and plutonic environments. The similar style and degree of alteration of the trondhjemites and rhyolitic volcanic lithofacies suggest that the trondhjemites where emplaced into the volcanic pile during a period of large-scale hydrothermal alteration. The lack of pervasive alteration of the tonalitic and gabbroic plutons suggests that these were emplaced following cessation of large-scale

123 MT MORGAN A u - C u DEPOSIT, QLD hydrothermal alteration. This together with the sharp volcaniclastic deposits of the Bunga beds. So, whereas intrusive contact between tonalite and both altered the style of volcanism at Mt Morgan is likely to have trondhjemite and volcanics at Mt Morgan indicates that been similar to the 1953-1957 eruption of Tuluman trondhjemite was emplaced prior to both tonalitic and Volcano and that interpreted for the Bunga beds, the mafic plutons. Furthermore, the lack of alteration shown scale of eruptions leading to deposition of the lower by tonalite stocks and the cross-cutting relationships Capella Creek Group was significantly greater. This is between these stocks and altered and mineralised rocks consistent with the proximity of a subvolcanic trondat Mt Morgan indicates that tonalite was emplaced after hjemite magma chamber (represented by lithofacies association 7) at Mt Morgan. This magma chamber was mineralisation. available to supply a large volume of magma to the dome-top vent/s. The proximal relationship with a subvolcanic magma chamber could also explain the DISCUSSION large-scale hydrothermal cell that resulted in wideVolcanic lithofacies of the lower Capella Creek Group spread alteration, exhalative activity and ore formation at Mt Morgan. at Mt Morgan and along the Dee Range record a period It is concluded that the rhyolitic volcanic lithofacies of subaerial or shallow-marine, rhyolitic volcanism. Volcanism resulted in deposition of thick, subaqueously associations at Mt Morgan and their regional correlaemplaced volcaniclastic mass-flow deposits (lithofacies tives represent a cogenetic, intrabasinal, proximal, exassociation 1), together with intrusive (lithofacies asso- plosive-effusive volcanic facies association genetically ciation 2) and probable submarine, effusive, rhyolitic related to the trondhjemite plutons of the Mt Morgan activity (porphyry clasts of lithofacies association 3). Tonalite. Tonalitic and gabbroic stocks were emplaced The spatial relationship of these associations with chem- relatively later in the Devonian history of the belt. The ically similar, genetically related stocks (lithofacies evolution of local dome-related volcanism and exhalite association 7) reflects the proximal nature of volcanism formation at Mt Morgan is summarised in Figure 11, above a sub volcanic trondhjemitic magma chamber, and the interpreted volcanic architecture of the lower stocks from which eventually rose to intrude the Capella Creek Group is presented in Figure 12. volcanic pile. In terms of volcanic processes, the major lithofacies associations of the lower Capella Creek Group can be IMPLICATIONS FOR ORE GENESIS compared with the observed 1953-1957 eruption of Tuluman Volcano, Papua New Guinea (Reynolds & The Mt Morgan Au-Cu deposit was an irregular body Best 1976; Reynolds et al 1980). In addition, a facies of massive and disseminated pyrite-chalcopyrite model for subaqueous, rhyolitic, dome-top tuff cones, mineralisation that transgressed stratigraphic units but developed by Cas et al (1990) for the Devonian Bunga was cross-cut by a relatively unaltered tonalite stock. beds in New South Wales and based partly on the 1953- The abundance of relatively insoluble metals such as 1957 eruption of Tuluman Volcano, can also aid in the Cu, Bi and Te, and the magmatic component of the oreinterpretation of lithofacies relationships at Mt Morgan. fluid, indicated by S and Pb isotope studies (Eadington The first stage of the Cas et al (1990) four-stage model et al 1974; Golding et al 1993), have been used by involves emplacement of a rhyolitic dome beneath the some workers to argue in favour of a tonalite-related sea floor and this is represented at Mt Morgan by origin for the Mt Morgan Au-Cu deposit. These numerous synsedimentary, quartz-plagioclase porphyry features, however, are also consistent with a proximal, intrusions (lithofacies association 2). The second stage volcanogenic origin involving a mixture of magmatic of the model involves breaching of the sea floor by the and seawater-derived fluids. Furthermore, a volcanorising dome with the release of pumiceous lava blocks, genic origin is consistent with radiogenic and stablepartial dome collapse and resedimentation of rhyolite isotope data that indicate involvement of a hybrid, and updomed sediment down the flanks of the dome. seawater-diluted magmatic ore fluid (Golding et al During the submarine phase of the 1953-1957 eruption 1994), and the broad synvolcanic timing of mineralisaof Tuluman Volcano, large blocks of highly vesiculated tion established by field, petrographic and geochemical lava were observed to float to the sea surface before criteria (Messenger 1996). Having formed broadly coeval with the period of sinking after a few minutes (Reynolds & Best 1976; Reynolds et al 1980). These blocks were interpreted to rhyolitic volcanism described in this paper, the Mt have been released from subaqueous lava domes. This Morgan Au-Cu deposit is by definition an end-member stage of development is a valid analogy for the origin of of the general class of volcanogenic massive sulfide ore the porphyry clasts of the jasper-mudstone-porphyry deposits in terms of its sub-sea floor replacement pipe (lithofacies association 3) at Mt Morgan. Stages 3 and 4 morphology (Large 1992). It is distinguished from the of the Cas et al (1990) dome-top tuff-cone model in- classical Kuroko-style volcanogenic massive sulfide volve shallow, subaqueous to subaerial explosive erup- deposits as well as many eastern Australian deposits by tion and dome collapse that produced lithofacies similar its form and metal content, being Au- and Cu-rich and to the volcaniclastic deposits of lithofacies associations Pb-Zn-poor. These differences are the result of differing volcanic and tectonic settings and crustal 1 and 6 at Mt Morgan. Individual beds that comprise lithofacies association thicknesses and in particular, reflect the low-K or tholeiitic nature of volcanism at Mt Morgan compared 1, however, are significantly thicker than the

124

P. R . M E S S E N G E R £ T

200 m to 2 km

I

Pumice rafts

I Quartz-plagioclase porphyry |

Blocky peperite

Banded Mine Sequence

Exhalite

|§S| Lower Mine Siltstone

Black smoker chimney and alteration channel

Crystal-rich pumiceous breccia-sandstone (lithofacies association 1)

Hydrothermal fluid circulation path

lithofacies identified at Mt Morgan

Fringing coral reefs Vesiculated lava blocks Stratiform & transgressive exhalite/massive sulphide Re-sedimented dome-related breccia Rhyolitic, pyroclast and crystal-rich volcaniclastics

sea level

W

Quartz-plagioclase porphyry Early-Middle Devonian volcanic basement

with the calc-alkaline settings typical of eastern Australian and Japanese volcanogenic massive sulfide districts. In terms of its metal content, morphology, host lithologies, volcanic setting and chemistry of both

Figure 11 Subaqueous lava dome model for development of synsedimentary intrusions (lithofacies association 2) and porphyry clasts of lithofacies association 3, based on the Tulluman Volcano analogue discussed in the text.

Fault blocks

Hydrothermal fluid circulation path

Figure 12 Schematic volcanic facies model for the lower Capella Creek Group prior to deposition of the Upper Mine Sequence illustrating the relationship between rhyolite porphyry intrusion/extrusion and voluminous, shallowmarine to subaerial, explosive volcanism. The lithofacies identified at Mt Morgan are indicated

volcanics and subvolcanic granitoids, the Mt Morgan Au-Cu deposit is most similar to the Archaean Home deposit in Canada. The empirical observation that AuCu-rich volcanogenic massive sulfide deposits at both

MT MORGAN A u - C u D E P O S I T , QLD Mt Morgan and Home are proximally related to low-K rhyolitic volcanics and associated subvolcanic trondhjemites suggests that this lithofacies association is a prime exploration target for this important class of ore deposit.

CONCLUSIONS (1) The dominant lithofacies association of the lower Capella Creek Group comprises crystal-rich, pumiceous volcaniclastic breccia and tuffaceous sandstone deposits (lithofacies association 1). This association represents submarine emplaced, high particle concentration, crystal- and pumice-rich volcaniclastic mass-flow deposits, sourced from contemporaneous, rhyolitic, shallow-marine or subaerial pyroclastic eruptions and/or resedimentation of non-welded, primary pyroclastic debris. (2) Massive, coherent quartz-plagioclase porphyry bodies (lithofacies association 2) represent synsedimentary, dome-like intrusions emplaced throughout the stratigraphy of the lower Capella Creek Group prior to dewatering and lithification. Both blocky peperite and ragged peperite occur in the contact zone between porphyry intrusions and sedimentary lithologies, although the former are more common where host sedimentary units are coarse grained and poorly sorted. (3) The Banded Mine Sequence at Mt Morgan comprises a mixture of jasper, tuffaceous mudstone and siltstone, together with large, compacted, porphyritic pumice clasts plus subordinate, non-compacted, nonvesiculated, porphyritic lava blocks (lithofacies association 3). Vesiculated and non-vesiculated porphyritic lava clasts are interpreted to have been derived predominantly from proximal, effusive lava dome/s genetically associated with the deposits of lithofacies association 1. (4) The major volcanic lithofacies of the lower Capella Creek Group at Mt Morgan and along the Dee Range form a proximal, intrabasinal, co-genetic, rhyolitic, explosive-effusive facies association related to a subvolcanic intrusive complex dominated by trondhjemite. (5) Rhyolitic volcanism was accompanied by widespread hydrothermal activity leading to the formation of chert, jasper and banded iron-formation and to sub-sea-floor massive sulfide mineralisation at Mt Morgan (lithofacies associations 3 and 4). (6) The Mt Morgan Au-Cu deposit represents an end-member of the class of volcanogenic massive sulfide deposits and many of its characteristics result from its proximal relationship to low-K rhyolite/ trondhjemite magmatism. ACKNOWLEDGMENTS The authors would like to acknowledge J. McPhie for her work in the area and many fruitful discussions. Both J. McPhie and R. Cas are thanked for their helpful

125

reviews. We also thank S. Bryan and C. Stephens who kindly read an earlier draft of the paper and made useful comments regarding its structure and content. Perilya Mines NL and CRAE are thanked for granting permission to publish graphic logs and maps from Mt Morgan and the Dee Range respectively. This work was partly funded by ARC Grant A39130274 to S. D. Golding and A. H. White.

REFERENCES ARNOLD G. O. & SILLITOE R. H. 1989. Mt Morgan g o l d -

copper deposit, Queensland, Australia: evidence for an intrusion-related replacement origin. Economic Geology 84, 1805-1816. BOULTER C. A. 1993. High level peperitic sills at Rio Tinto, Spain: implications for stratigraphy and mineralisation. Transactions Institute of Mining and Metallurgy 102, B30-B38.

BRANNEY M. J. & SPARKS R. S. J. 1990. Fiamme formed by

diagenesis and burial-compaction in soils and subaqueous sediments. Journal of the Geological Society of London 147,919-922. BROOKS E. R. 1995. Paleozoic fluidisation, folding and peperite formation, northern Sierra Nevada, California. Canadian Journal of Earth Sciences 32, 314—324. BUSBY-SPERA C . J. & W H I T E J. D . L . 1 9 8 7 . V a r i a t i o n

in

peperite textures associated with differing host-sediment properties. Bulletin ofVolcanology 49, 765-775. CAS R. A. F. 1983. Submarine 'crystal tuffs': their origin using a Lower Devonian example from southeastern Australia. Geological Magazine 120,471-486. CAS R. A. F. & WRIGHT J. V. 1991. Subaqueous pyroclastic flows and ignimbrites: an assessment. Bulletin of Volcanology 53, 357-380. CAS R . A . F., ALLEN R . L . , BULL S. W . , CLIFFORD B . A . &

WRIGHT J. V. 1990. Subaqueous rhyolitic dome-top tuff cones: a model based on the Devonian Bunga Beds and a modern analogue. Bulletin of Volcanology 52, 159-174. CHRISTIANSEN R . L . & LIPMAN P. W . 1 9 6 6 . E m p l a c e m e n t a n d

thermal history of a rhyolite lava flow near Forty Mile Canyon, southern Nevada. Geological Society of America Bulletin 77, 671-684. CLOUGH B . J., WRIGHT J. V . & WALKER G . P. L. 1 9 8 1 . A n

unusual bed of giant pumice in Mexico. Nature 289, 4 9 50. CORNELIUS K. D. 1968. The ore deposit and general geology of the Mt Morgan area. PhD thesis, University of Queensland, Brisbane (unpubl.). CORNELIUS K. D. 1969. The Mt Morgan mine, Queensland a massive gold-copper pyritic replacement deposit. Economic Geology 64, 885-902. DEAR J. F., MCKELLAR R . G . & TUCKER R . M . 1 9 7 1 . G e o l o g y

of the Monto 1:250 000 sheet area, Queensland. Queensland Geological Survey Report 46. EADINGTON P. J., SMITH J. W . & WILKINS R . W . T .

1974.

Fluid inclusion and sulphur isotope research, Mount Morgan, Queensland. In: Southern and Central Queensland Conference, Papers, pp. 4 4 1 ^ 4 4 . Australasian Institute of Mining and Metallurgy, Brisbane. ELLISTON J. 1969. The genesis of some epigenetic type ore deposits. 23rd International Geological Congress 14 (2), 129-139.

126

P. R. M E S S E N G E R S T A L .

FEDIKOW M . A . F. & GOVETT G . J. S. 1 9 8 5 . G e o c h e m i c a l

alteration halos around the Mount Morgan gold-copper deposit, Queensland, Australia. Journal of Geochemical Exploration 24, 247—272. FISKE R. S. 1969. Recognition and significance of pumice in marine pyroclastic rocks. Geological Society of America Bulletin 80, 1-8. FISKE R. S. & MATSUDA T. 1964. Submarine equivalents of

ash flows in the Tokiwa Formation, Japan. American Journal of Science 262, 76-106. FORDHAM

B. G.

& TAUBE

A.

1994.

Application

of

stratigraphic correlation to exploration in equivalents of the Mine Corridor Sequence at Mt Morgan. In: Queensland Department of Minerals and Energy Symposium. Queensland Exploration Potential 1994. Handbook, p.33. Queensland Department of Minerals and Energy, Brisbane. FRASER C. 1914. Geological report on Mount Morgan mine. Report, Mount Morgan Gold Mining Company Limited (unpubl.). FRETS D. C. 1970. Mine geology and exploration future— Mount Morgan Area. Report, Geopeko Limited east Queensland Group, Mount Morgan (unpubl.). FRETS D. C. 1974. Rock relationships and mineralisation at Mt Morgan. In Australasian Institute of Mining and Metallurgy Southern and Central Queensland Conference Papers, pp. 425-440. Australasian Institute of Mining and Metallurgy, Brisbane. FRETS D. C. & BALDE R. 1975. Mount Morgan copper-gold deposit. Australasian Institute of Mining and Metallurgy Monograph

5, 778-785.

KANO K. 1991. Miocene pillowed sills in the Shimane Peninsula, SW Japan. Journal of Volcanology and Geothermal Research 48, 359-366. KIRKEGAARD A . G . , SHAW R . D . & MURRAY C . G .

1970.

Geology of the Rockhampton and Port Clinton 1:250 000 sheet areas, Queensland. Queensland Geological Survey Report 38. KOKELAAR B. P. 1982. Fluidisation of wet sediment during emplacement and cooling of various igneous rocks. Journal of the Geological Society of London 139, 21—33. LARGE R. L. 1992. Australian volcanic-hosted massive sulphide deposits: features, styles, and genetic models. Economic Geology 87, 471-510. LAWRENCE L. J. 1977. The syngenetic-epigenetic transition, an Australian example. In: Klemm D. D. & Schneider H. J. eds. Time and strata-bound ore deposits, pp. 46-54. Springer-Verlag, Berlin. LOWE D. R. 1982. Sediment gravity flows: II. Depositional models with special reference to the deposits of highdensity turbidity currents. Journal of Sedimentary Petrology 52, 279-297. MCPHIE J. 1993. The Tennant Creek Porphyry revisited: a synsedimentary sill with peperite margins, Early Proterozoic, Northern Territory. Australian Journal of Earth Sciences 40, 545-558. MCPHIE J. 1995. Report: Lithofacies and stratigraphy of volcanic and sedimentary rocks in PD 5, PD 11, PD 20 (Mount Morgan mine slag heap area), Mount Morgan. Report, Perilya Mines NL, Mt Morgan (unpubl.). MCPHIE J., DOYLE M. & ALLEN R. 1993. Volcanic

Textures:

GOLDING S. D. 1995. Report to Mount Morgan Joint Venture on Mineragraphy of core samples. Report, Perylia Mines NL (unpubl.).

A Guide to the Interpretation of Textures in Volcanic Rocks. Centre for Ore Deposit and Exploration Studies, University of Tasmania, Hobart.

GOLDING S. D. & WILSON A. F. 1981. An oxygen and carbon

MCPHIE J. & HUNNS S. R. 1995. Secondary welding of

isotope study of some gold deposits of eastern Australia. Proceedings of the Australasian Institute of Mining and Metallurgy 278, 13-21. GOLDING S. D., HUSTON D. L., DEAN J. A. ETAL. 1993. Mount Morgan gold-copper deposit: the 1992 perspective. In: Centenary Conference, pp. 95—111. Australasian Institute of Mining and Metallurgy, Melbourne.

submarine, pumice-lithic breccia at Mt Chalmers, Queensland, Australia. Bulletin of Volcanology 57, 170-178. MAHOOD G. A. 1984. Pyroclastic rocks and calderas associated with strongly peralkaline volcanism. Journal of Geophysical Research 89, 8540-8552. MESSENGER P. R. 1996. Relationships between Devonian magmatism and Au-Cu mineralisation at Mt Morgan, central Queensland. PhD thesis, University of Queensland, Brisbane (unpubl.).

GOLDING S. D . , MESSENGER P. R . , DEAN J. A . , PERKINS C . ,

HUSTON D. A. & WHITE A. H. 1994. Mt Morgan g o l d -

copper deposit: geochemical constraints on the source of volatiles and lead and the age of mineralisation. In: Henderson R. A. & Davis B. K. eds. Extended conference abstracts — New developments in the geology and metallogeny: Northern Tasman Orogenic Zone, pp. 89-95. Economic Geology Research Unit, James Cook University, Townsville, Contribution 50. GIBBONS G. S. 1974. Mineralogical studies at Mount Morgan Queensland. In: Australasian Institute of Mining and Metallurgy, Southern and Central Queensland Conference, Papers 1974. pp. 445-462. Australasian Institute of Mining and Metallurgy, Brisbane.

MESSENGER P. R . & G O L D I N G S. D . 1 9 9 6 .

Relationships

between Devonian volcanic and plutonic units in the Mt Morgan area, central Queensland. Australian Journal of Earth Sciences 43, 97-101. MESSENGER P. R . , COLLERSON K . D . & GOLDING S. D . 1996.

Isotope and trace element geochemistry of a Middle Devonian tonalite-trondhjemite-sodic rhyolite centre: Mt Morgan, Queensland, Australia. EOS 77 (22), W149. MORAND V. J. 1993. Stratigraphy and tectonic setting of the Calliope Volcanic Assemblage, Rockhampton area, Queensland. Australian Journal of Earth Sciences 40, 15-30.

HANSON R. E. 1991. Quenching and hyaloclastic disruption of andesitic to rhyolitic intrusions in a submarine island arc sequence, northern Sierra Nevada, California. Geological Society of America Bulletin 103, 804-816.

NEWMAN J. M . & CAMPBELL-BROWN G . F. 1911. N o t e s o n t h e

HANSON R. E. & WILSON T. J. 1993. Large-scale rhyolite

PALTRIDGE I. M. 1967. Breccia pipe mineralisation at Mount Morgan. Economic Geology 62, 861-863. RAWLINGS D. J. 1993. Mafic peperite from the Gold Creek Volcanics in the Middle Proterozoic McArthur Basin, Northern Territory. Australian Journal of Earth Sciences 40, 109-114.

peperites (Jurassic, southern Chile). Journal gy and Geothermal Research 54, 247-264.

ofVolcanolo-

HAWKINS B. & WITCHER I. 1961. Geological environment of

the Mt Morgan orebody. Report, Enterprise Exploration Company Pty. Ltd (unpubl.).

geology of Mt Morgan, Queensland. Transactions of the Australian Institute of Mining and Metallurgy 40, 439-470.

MT MORGAN A u - C u DEPOSIT, QLD REID J. H. 1947. The Mount Morgan orebody: a new interpretation of its origin. Report, Mount Morgan Limited (unpubl.). REYNOLDS M. A. & BEST J. G. 1976. Summary of the 1953—

1957 eruption of Tuluman Volcano, Papua New Guinea. In: Johnson R. W. ed. Volcanism in Australasia, pp. 287— 296. Elsevier, Amsterdam. REYNOLDS M . A . , BEST J. G . & JOHNSON R . W . 1 9 8 0 . 1 9 5 3 -

57 Eruption of Tuluman Volcano: rhyolitic volcanic activity in the northern Bismark Sea. Geological Survey of Papua New Guinea Memoir 7 SCHMINKE H. U. 1967. Fused tuff and peperites in southern central Washington. Geological Society of America Bulletin 78, 319-330.

SCROPE G. P. 1827. Memoir on the Geology of Central France; Including the Volcanic Formations of Auvergne, the Velay and the Vivarais. Longman, Rees, Orme, Brown and Green, London. SILLITOE R. H. 1987. A re-appraisal of the Mt Morgan goldcopper deposit and environs, Queensland, Australia: implications for exploration. Report, Goldfields Exploration Pty Limited (unpubl.). SMITH R. L. 1960. Zones and zonal variation in welded ash flows. US Geological Survey Professional Paper

127

354F, 149-159. SPARKS R. S. J. & WRIGHT J. V. 1979. Welded air-fall tuffs. In: Chapin C. E. & Elston W. E. eds. Ash flow tuffs, pp. 150-166. Geological Society of America Special Paper 180. STAFF, MOUNT MORGAN LIMITED 1 9 6 5 . T h e M o u n t M o r g a n

mine. 8th Commonwealth Mining and Metallurgical Congress Publications 1, 364—369. STAINES H. R. E. 1953. Mount Morgan copper and gold mine. In: Edwards A. B. ed. Geology of Australian ore deposits, pp. 732-750. 5th Empire Mining and Metallurgical Congress, Melbourne. TAUBE A. 1986. The Mt Morgan gold-copper mine and environment, Queensland: a volcanogenic massive sulphide deposit associated with penecontemporaneous faulting. Economic Geology 81, 1322-1340. WALKER G. P. L. 1972. Crystal concentration in ignimbrites. Contributions to Mineralogy and Petrology 36, 135-146. WILSON C . J. N . & W A L K E R G . P. L . 1 9 8 5 . T h e

Taupo

eruption, New Zealand. I. General aspects. Philosophical Transactions of the Royal Society of London 314, 199-228.

(Received 24 April 1996; accepted 1 July 1997)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 128—147.

Geology and mineralisation of the Gympie Province L. C. CRANFIELD, G. SHORTEN, M. SCOTT AND R. M. BARKER Department of Mines and Energy, GPO Box 194, Brisbane, Qld 4001, Australia.

This paper outlines the probable depositional environments of the Gympie Province and its extent in space and time. Relationships between units of the Gympie Province and older rocks of the Yarrol Province to the west are complex. The mode of emplacement of the Gympie Province to its present position and its structural complexity are reviewed in terms of the present model for evolution of the Bowen Basin. Evolution of the Gympie Province includes development of the Gympie arc in either the latest Carboniferous or earliest Permian to the east of the Australian continent, possibly at about the same time as the Camboon Volcanics were deposited in the Bowen Basin. The Gympie arc that produced the Highbury Volcanics and Mant Basalt was followed in the Early Permian by a period of extension. This extension produced the Rammutt and Kolbar Formations in the Gympie Province and the Cambroon beds and Cedarton Volcanics in extensional basins within the Late Devonian to Carboniferous rocks of the accretionary wedge (D'Aguilar Block, Good Night beds). A mid-Permian sag phase deposited the South Curra/Gigoomgan Limestones and the Tamaree/ Teebar Formations. A regression in the Early Triassic deposited continental deposits in the Brooweena and Keefton Formations followed by a return to marine conditions in the upper part of the Brooweena Formation and Kin Kin beds. Collision of the Gympie Province with the continent in the Middle to Late Triassic is probably coincident with the foreland loading phase that the Hunter-Bowen Orogeny produced in the Bowen Basin. Deformation of the Gympie Province in this event produced three discrete structural zones in the Gympie Province: (i) an eastern zone of shallowly dipping cleavage with east-over-west sense of thrusting; (ii) a central weakly deformed zone; and (iii) a poorly exposed western zone of steep to overturned bedding and zones of west-dipping thrusts. The formation of steep west-dipping thrusts and overturned beds in the western zone is considered to be due to reverse movement on extensional faults produced in the Early Permian extensional phase that deposited the Rammutt and Kolbar Formations in the Gympie Province and backarc extensional basins over the Carboniferous subduction complex. Mineralisation in the Gympie Province is mainly epigenetic and is dominantly of Middle Triassic and Late Triassic age. The mineralisation style that developed during the Early to Middle Triassic includes the Gympie Goldfield. This mineralisation is described and compared with Late Triassic mineralisation. Styles of mineralisation that accompanied each of these mineralising events are compared and contrasted with those elsewhere in the Gympie Province. Key words: geology, Gympie Province, mineralisation, stratigraphy, tectonics.

INTRODUCTION

The Gympie Province has been discussed over the past twenty years, however, its extent both in space and time has been poorly understood or defined by most authors. The origin of the Gympie Province has been conjectural since Harrington (1974) noted it as unique in the geology of eastern Australia because of the presence of marine Triassic rocks. Day et al. (1978) discussed the tectonostratigraphic nature of the Gympie Province (which then included the Amamoor beds, Good Night beds, Marumba beds, Rocksberg Greenstone and Kurwongbah beds of the North D'Aguilar Block) and identified it as not fitting the palaeogeographical pattern of the New England Orogen. Linked to this concept, the Gympie region was proposed as an exotic terrane by Harrington (1983, 1987), Harrington and Korsch (1985), Cawood (1984), Waterhouse and Sivell (1987a, b), and Cranfield (1990). Murray (1987a, 1988) and Murray et al. (1989) suggested that the Gympie Province is composed of at least three discrete suspect terranes one of which is composite. The metallogenesis of the Gympie Province as part of the New England

Orogen was discussed in Murray (1986, 1987b). Wholerock geochemistry studies include samples from Permian Highbury Volcanics (Sivell & Waterhouse 1987a, b, 1988; Sivell 1990; Cranfield & Scott 1993) and undivided Gympie Group Volcanics (Cranfield & Murray 1989). Sivell (1990) and Sivell et al. (1990) also carried out studies on the Amamoor beds and Rocksberg Greenstone which they considered to be part of Gympie Province. Nd isotope studies were carried out on the Highbury Volcanics, Cedarton Volcanics and 'Amamoor' beds by Sivell and McCulloch (1993,1996). Palaeontological investigations of brachiopod macrofossils by Waterhouse and Balfe (1987) linked fossils of the Gympie Group with those from the Bowen and Sydney Basins. Waterhouse and Sivell (1987a, b) equated Permian to Triassic rocks of the Gympie area with sequences in New Caledonia and New Zealand. Ishiga (1990) described radiolarian occurrences from the Gympie and Nambour 1:100 000 Sheet areas. These descriptions showed that areas previously mapped as Gympie Group were Early Carboniferous in age, and were reassigned to the Amamoor beds. Early Permian radiolarians were identified from rocks previously

mapped as Amamoor beds near Kenilworth. These rocks are defined as the Cambroon beds (Sliwa et al 1993; Donchak et al 1995), which are separated from the Cedarton Volcanics to the south by a thin cover of Tertiary basalt. The Cambroon beds and Cedarton Volcanics are considered by Donchak et al (1995) to be the same unit due to similarities in rock types (especially the andesitic volcanic component), deformation style and marine macrofossils (as outlined in Murphy et al 1976). The evolution of the North D'Aguilar Block and Early Permian fault basins that took place by extension and exhumation of the New England subduction complex was discussed by Little (1993), Holcombe et al (1993) and Sliwa et al (1993). PRESENT IDEAS ON THE TECTONOSTRATIGRAPHY OF THE GYMPIE PROVINCE The Amamoor beds (North D'Aguilar Block) and Good Night beds (Good Night Block), formerly considered as part of the Gympie Province by Cranfield (1990), are now regarded as part of New England Orogen proper. They represent deformed and metamorphosed remnants of an accretionary complex formed during Late Devonian to Middle Carboniferous Andean-style subduction at the eastern margin of the Australian continent. They show evidence of a two-stage deformational and metamorphic history, characterised by early subduction-related features overprinted by extensionrelated features. Extension was centred on a major crustal detachment that juxtaposes lower plate higher grade rocks and upper plate lower grade rocks (Little 1993). The structural architecture of the western margin of the North D'Aguilar Block is further modified by Late Permian imbricate thrusts (Donchak et al 1995). The deformed accretionary basement rocks enclose Early Permian marine fault-bound basins containing rift-fill sediments and volcanics—Cedarton Volcanics and Cambroon beds. The Gympie Province (Figure 1) comprises arcrelated mafic to felsic volcanics, volcaniclastics, and marine and non-marine sediments of Early Permian to Early Triassic age within a southern area centred on Gympie and a northern area centred on Dallarnil. Contacts in all areas between the Gympie Province and the Amamoor beds appear tectonic. Melange incorporating rock types of Gympie Group and Amamoor beds (disrupted chert, fine arenite, rounded clasts of chert and rare marble) in a matrix of sheared metapelite extends 10 km north along the Mary River from the Gympie aerodrome (Cranfield & Scott 1993). Faults and shear zones separate the northern part of the Gympie Province from the New England Orogen's Amamoor and Good Night beds. The Gympie Province is probably an allochthonous terrane accreted to the craton by the Middle Triassic. Mesozoic and Cenozoic rocks of the Maryborough and Pomona Basins unconformably overlie, or are faulted against the Gympie Province. A time-space plot (Figure 2) shows the Gympie Province in relation to other parts of New England Orogen. Units within the

GYMPIE PROVINCE, QLD 129 Gympie Province and evolution of the Gympie Province are compared in Figure 3 with stages of evolution of the Bowen Basin described in Baker et al (1993). UNITS OF THE GYMPIE PROVINCE Gympie Group Type sections of the Gympie Group—[Highbury Volcanics (base), Rammutt Formation, South Curra Formation and Tamaree Formation (top)]—were defined in the vicinity of Gympie township by Runnegar and Ferguson (1969). Ellis (1968) mapped Permian rocks in the Maryborough 1:250 000 Sheet area in the eastern part of the 'Biggenden beds'. Fieldwork by Cranfield (1989) redefined the eastern 'Biggenden beds' as Gympie Group and designated type sections for Brown's (1964) Mant Basalt (base), Kolbar Formation, Gigoomgan Limestone and Teebar Formation (top). Near Dallarnil the Gympie Group is poorly exposed, structurally deformed and is either undivided or mapped as Gympie Group Volcanics or limestone (Cranfield 1994). The basal unconformity of the Gympie Group with older rocks is not exposed in either the Gympie or Maryborough areas. Mant Basalt Near Gigoomgan the Mant Basalt is a mainly tuffaceous basalt sequence with pillow lavas overlain by pillow lavas and amygdaloidal basalts and minor pyroclastic breccias. Amygdaloidal basalts contain altered albitic plagioclase, chlorite, and fine-grained quartz and calcite inclusions, in a groundmass of fine magnetite and hematite. Amygdales contain calcite, chlorite and microcrystalline quartz. Rare tuffaceous units are present towards the top of the unit. The groundmass of the breccia is fine grained and greenish due to chlorite and actinolite alteration. South of Gigoomgan the Mant Basalt comprises basalt breccia, amygdaloidal basalt and porphyritic basalt. Clinopyroxene crystals are present in some specimens and amygdales are dominantly filled with calcite. Here the top unit of the Mant Basalt is a reddish-brown altered amygdaloidal basalt. Undivided (Gympie Group) volcanics North of Gigoomgan, undivided Gympie Group volcanics occupy a similar stratigraphic level to the Highbury Volcanics and the Mant Basalt. The unit was described by Cranfield (1994) as comprising mainly altered calcite-veined basalt. The basalts are dominantly amygdaloidal and have prominent pyroxene phenocrysts. In the area between Biggenden and Wallaville the unit is generally poorly exposed and strongly sheared particularly on its western margin. Strongly elongated, sheared amygdales are a feature of this unit (Cranfield 1994).

130

L. C. CRANFIELD ETAL.

Highbury Volcanics

The Gympie area was remapped in 1990 to 1992 as part of the Department of Mines and Energy's GEOMAP 2005 program by Cranfield and Scott (1993). They showed that the Highbury Volcanics range from basalt to andesite and contain pillow lavas and volcanic breccias. Typical Highbury Volcanics basalt north of Gympie contains andesine or labradorite and clinopyroxene phenocrysts in a groundmass of andesine, clinopyroxene and magnetite. Amygdales are filled with chlorite, calcite or a mixture of actinolite, prehnite, zeolite, epidote and calcite. Commonly the volcanics are altered and locally brecciated. Altered basalt to andesite with phenocrystic clinopyroxene occurs in several bores in the Gympie area. The top of the Highbury Volcanics in the Gympie area was defined by Cranfield and Scott

<22 LUO

m A-B

Kolbar Formation The Kolbar Formation forms a narrow linear belt along the hinge zone of the Gigoomgan Anticline in the north of the Gympie Province. Exposures of Kolbar Formation are present along the western slopes of the Urah Range (Cranfield 1994). The base of the Kolbar

- FAULTS SERPENTINITE

II

(1993) as the top pyroxene-phyric flow or pyroclastic below either a basalt-derived conglomerate or dacitic crystal tuff. The boundary between the Highbury Volcanics and the Amamoor beds is faulted in all exposures. Shear zones occur along the whole of this boundary. In these zones, phacoids of Highbury Volcanics, other units of the Gympie Group, and Amamoor beds are included in a tectonised, scaly clay matrix.

VA

^ SHEARZONE

(Bmoweena Formation, Kin Kin beds, Keefton Formation)

<3>

FOSSIL LOCALITY

ET

EARLY TRIASSIC

PERMIAN

LP

LATE PERMIAN

EP

EARLYPERMIAN

EARLY TRIASSIC

(Gympie Group, Cambroon beds, Cedarton Volcanics, Marvmba beds)

MIDDLE CARBONIFEROUS LATE DEVONIAN TO EARLY CARBONIFEROUS UNDIVIDED PALAEOZOIC

CROSS SECTIION (SEE FIGURE 5)

P

PERMIAN

MC

MID CARBONIFEROUS

EC

EARLY CARBONIFEROUS

Figure 1 Geological map showing the extent of the Gympie Province.

GYMPIE PROVINCE, QLD Formation overlying the Mant Basalt is locally a thin shell bed containing mainly Atomodesma fragments and prisms. Several thin shelly beds are also present in the lower part of the unit, which is a sequence of interbedded indurated buff siltstones and clastsupported arenites. Arenite from the Kolbar Formation contains mainly sedimentary rock fragments with zoned plagioclase and quartz. The contact to the overlying Gigoomgan Limestone is poorly exposed.

131

conglomerate with minor carbonaceous 'breaks'. Roach (1990) defined five facies in the Rammutt Formation near the West of Scotland shaft, Monkland— lower basalt, middle clastic, upper clastic, upper andesite, and upper conglomerate. Dunstan's basal 'Third Slate Group' was not recognised by Roach in the Monkland area. Roach (1990) identified unconformities at the base of the middle clastic, upper andesite, and upper conglomerate. The relationship between the units defined by Roach and those of Dunstan is shown in Table 1.

Rammutt Formation Table 1 Relationship of Rammutt Formation units.

The Rammutt Formation hosts most of the hard-rock gold mineralisation in the Gympie Province. The Rammutt Formation comprises a large variety of different rock types and has a strongly enhanced beddingplane fissility. Thin-sections from the unit show that it has an anastomosing cleavage (at a low angle to bedding) and stylolitic dissolution of grains. Dunstan (1911) subdivided the Rammutt Formation into six informal units (from bottom up): (i) 'Third Slate Group', a basal shale-tuff - rare conglomerate unit; (ii) 'Second Volcanic Group', volcanic flows, tuff, conglomerate; (iii) 'Second Slate Group', shalesandstone-conglomerate; (iv) 'First Volcanic Group', altered volcanic flows-tuff-sandstone-conglomerate; (v) 'First Slate Belt', shale—sandstone-conglomerate; and (vi) 'Conglomerate Group', a top unit of

Dunstan (1911)

Roach(1990)

Conglomerate Group First Volcanic Group First Slate Belt Second Volcanic Group

upper Conglomerate Facies upper andesite upper clastic facies middle clastic facies

The base of the Rammutt Formation (GR 4585 70095) (all grid references are to the Gympie 1:100 000 Sheet 9445) is a matrix-supported conglomerate which contains dark-grey clasts of altered basalt in a clay matrix (Cranfield & Scott 1993). At Chatsworth (GR 4615 71083), Dunstan's 'Third Slate Group' consists of

EARLY PERMIAN (FAULT BASINS)

EARLY PERMIAN (FAULT BASINS)

GYMPIE PROVINCE

UjS

§

111

§8

§

£

§

IS i s 5oq

A /WW

a r— w

V

IfL

03 =

p^n— ^ - -

"

ft

<ww

T ft Plutonic rocks

WEDGE

§5S

I

Si

ii

111 AGE

i i

I

3§

fl TP ^ R - s — . — ± — /WW /WW /V^K/V/

FAULT BASIN

/WW

ACCRETIONARY WEDGE

-

§5

I

iii ™

FAULT

/V^A/ ACCRETIONARY WEDGE

(Ma) -200 o -220 C CO O

<

C -240 D 1— FORELAND z LOADING -260 < THERMAL SAG dc -280 L QU _ CO EXTENSION -300 =5 O VOLCANIC DU C GYMPIE ARC L -320 LL. z o C O -340 ^< DC -360

ts

z < z -380 o > LU -400 Q

Figure 2 Time-space plot showing the Gympie Province in relation to other parts of the New England Orogen (adapted from Murray 1990).

132

L. C. CRANFIELD BOWEN BASIN STAGES

AGE

ETAL. GYMPIE

MARYBOROUGH

Kin Kin beds

EARLY TRIASSIC

F — — — —

- o - o - B R 0 0 W E E N A FORMATION- -

- OOOOOOOOOOOOO

F-

(oooooooooooooo • ° KEEFTON FORMATION ° 0 o ou oU oLi o o o o o o o o o o o o O— ° - - 0 , - 0 O - - 2 / - ° - — — - / - — — /UNDIVIDED' •GYMPIE _ GROUP-

EARLY PERMIAN

- UNDIVIDED _"GYMPIE_ GROUP LIMESTONE",

SOUTHCURRA. LIMESTONE

GIGOOMGAN LIMESTONE

7

IZ

~ ' t

:b i

/

' - \ v '

iz:

Z 7/ i ^ ,

. r i ^ ./

x 7I RAM ' ;u MUTT ^

KOLBAR FORMATION ^

K..

v

V

v

-y

FFORMATION O R M A T I O N !;

^ J. —

—

—

< A A AU A - A < U A < L- A < U AUK l>

UNDIVIDED . A 1 U A -7 LAI LAI CAMBOON >• " G Y M P I E ^ n ^ o ( v ^ . , n v . , N V . , fv VOLCANICS ^ GROUP - . ^ M A H T BASALT VOLCANICS"

? LATE CARB.

TAMAREE FORMATION

/ , / 7" z ; z

7

LATE PERMIAN

— / — z

/ ^ ^ , -TEEBAR FORMATION —A

U A AU -7A ^ U Z.AA-7 ^ ULAIA *^ U A L Af*. 1-7 Z L- A -7 LAI \ < : : HIGHBURYyOLCANICS N• U A

N N

Marine o oc _j2

a.

Continental

F

Limestone & turbidites Dacitic-Andesitic deponts WJ

Basic

F

U-u-

volcanics

thick-bedded, pencil-jointed, olive to buff mudstone and thin-bedded fine-grained sandstone. The thickness of this part of the sequence probably does not exceed 150 m. The overlying 'Second Volcanic Group' comprises interbedded tuff, shale, sandstone and minor conglomerate and is thicker bedded than the basal unit. The sandstone, conglomerate, and tuffaceous interbeds of the 'Second Volcanic Group' are not obviously strongly cleaved and contain prominent dissolution seams visible in thin-sections of the sandstone. Mudstone beds are ubiquitously strongly cleaved in the Bruce Highway section of the unit from GR 4572 71103 to 4582 71092. Dunstan's 'Second Slate Group' is not readily differentiable as a unit and has been included in descriptions of the 'Second Volcanic Group'. Dunstan's 'First Slate Group' contains the 'first break shale' that hosts the richest part of the gold-bearing 'Gympie veins'. In the Monkland area this part of the formation contains mud- and debris-flows, lahars, lithic tuffs, conglomerate and minor porphyritic andesite flows, shale (locally carbonaceous) and arenite. Roach (1990) inferred an unconformity at the base of this facies and considered it to represent uplift and rapid erosion probably due to the renewal of volcanic activity that continued into the overlying upper andesite facies. North of Monkland, the upper clastic facies consists of

Fault Unconformity Inferred unconformity Plant fossils Marine fossils

Figure 3 Evolution of the Gympie Province compared with stages in the evolution of the Bowen Basin.

interbedded sandstone, siltstone, shale and thin beds of conglomerate. Dunstan's 'First Volcanic Group' contains dacitic crystal-poor (quartz, plagioclase) tuff, ignimbrite, sparsely porphyritic dacite flows, lapilli tuff and airfall tuff. Facies changes along strike make this unit equate in stratigraphic position with Roach's 'upper andesite facies'. In the Monkland area, Roach's (1990) 'upper andesite facies' with thick flows of strongly chloritised, porphyritic andesite to basalt, and tuff with zoned clinopyroxene phenocrysts appears similar to the Highbury Volcanics. However, in contrast to the Highbury Volcanics, alkali feldspar is the dominant feldspar phenocryst. The facies occurs south of the Mary River along the Gympie-Glastonbury Road and south to the junction of the Mary River and Mcintosh Creek but is absent north of Monkland, north of the Mary River. Dunstan's 'Upper Conglomerate Group' is younger and disconformable or unconformable on the rest of the Rammutt Formation. It comprises three thick bands of conglomerate split by two purplish to black sheared mudstone called the 'Upper and Lower Plumbago' by Dunstan (1911). North to south variation in clast composition indicates local derivation, for example clasts in the Chatsworth area are mainly dacite with rare deeply weathered, altered granite. South of the

GYMPIE PROVINCE, QLD Inglewood Fault clasts are porphyritic andesite, reworked volcaniclastic sedimentary rock, and ?porphyritic andesitic tuff. Granite is conspicuously absent. Conglomerate in the Chatsworth area varies from pebble to boulder size, is clast-supported, and has a well-developed imbrication parallel to bedding and a relatively intense cleavage at a low angle to layering shown by stylolitic dissolution seams. The Rammutt Formation is disconformable to unconformable on the underlying Highbury Volcanics. The boundary with the overlying South Curra Limestone is conformable.

Gigoomgan Limestone Brown (1964) showed the Gigoomgan Limestone comprised three main subdivisions: (i) a basal coarse bioclastic grainstone containing tabulate corals, molluscs and brachiopod shells; (ii) a middle sequence containing limestone lenses dominated by prisms of Atomodesma and containing thin interbeds of fine calcareous arenite and siltstone; and (iii) a top sequence dominated by bioclastic grainstone with thin interbeds of coarse arenite and siltstone. The unit is generally poorly exposed and is represented by isolated boulder outcrop

133

Tamaree Formation. Clasts are skeletal crystalline calcite, brachiopod shells, bryozoans, crinoids and Atomodesma prisms. Two biomicrite facies were noted at Curra Quarry. The basal biomicrite directly overlies the basal Atomodesma calcarenite and comprises numerous thin beds of fossiliferous micritic limestone and interbedded siltstone. Atomodesma prisms are the dominant fossils near the base of this facies. Shearing is synchronous with development of chlorite and muscovite. The upper micrite facies at the Curra Quarry section may be a repetition of the basal micrite facies as it contains the same features as the basal micrite facies. The intervening facies is a siliceous biosparite, which appears equivalent to the top siliceous biosparite. Consequently the South Curra Limestone is probably represented by at most three separate units. Intraformational thrust faults and recumbent folds are a feature from Curra Quarry to due east of Chatsworth. The east-dipping Laing Slide of Dunstan (1911) is apparently a thrust fault with east-over-west displacement that juxtaposes South Curra Limestone against Highbury Volcanics. The South Curra Limestone is conformably overlain by the Tamaree Formation.

Teebar Formation South Curra Limestone Brodie's (1991) work on the South Curra Limestone from Curra and Tamaree quarries and boreholes in the Monkland area delineated two distinct facies at Tamaree Quarry, five at Curra Quarry, and two at Monkland. The greater number of facies at Curra Quarry may be due to a repetition of the sequence at this locality. The basal Atomodesma calcarenite facies at all localities contains graded feldspathic calcarenite and thin bioturbated siltstone. The facies becomes finer grained and more fossiliferous towards the top. Numerous Atomodesma prisms are the dominant fossils. Clasts comprise subrounded volcanics, quartz, plagioclase and chert. Skeletal clasts of non-ferroan calcite are abraded and partially replaced by fine silica. Alteration features are replacement of siliceouscarbonate cement by fine iron-rich chlorite. Secondary chlorite-sericite-carbonate replacement of lithic and siliceous clasts and sericite follow a bedding planeparallel cleavage. In the middle of the basal calcarenite, siltstones are essentially non-fossiliferous, and contain well-sorted, siliceous and lithic fragments, and fine chlorite. Stylolites and sericite alteration form an anastomosing cleavage. Thin siltstone beds are disrupted by stages of calcite veining and subsequent shearing. Brodie (1991) showed that limestone in the southern part of Gympie township and south of Gympie is more arenaceous than the dominantly micritic and sparry limestones at Curra and Tamaree Quarries. Overlying the basal calcarenite facies in drillholes in the West of Scotland Mine area, is siliceous biosparite with minor interbedded biomicrite (Brodie 1991). This unit becomes more siliceous at the contact with the

In the Gigoomgan area, the Teebar Formation consists of rhythmically bedded conglomerate, arenite, siltstone and mudstone. Variations in the Teebar Formation are in the bedding thickness of the units and an increase in intensity of cleavage to the east. Conglomerate and arenite of the Teebar Formation are essentially similar in composition. They contain more than 50% basaltic volcaniclastic rock fragments, quartz and feldspar grains (10-15%) and lesser amounts of bioclastic fragments in a clay matrix and a clear calcite cement and are apparently derived mainly from the Mant Basalt (Cranfield 1989,1994).

Tamaree Formation The Tamaree Formation is dominated by rhythmically bedded arenite, mudstone, siltstone and shale. The unit is characterised by a well-developed penetrative cleavage at a low angle to bedding. Near Monkland, Roach (1990) described three facies of upward-fining shale, siltstone, mudstone and arenite. The lowest facies contains volcanolithic arenite, siltstone and minor shale. Graded bedding is common within the arenites, and some reverse grading is locally present. Roach indicated that this facies was the thickest bedded part of the Tamaree Formation. Thick-bedded siltstone to mudstone with thin arenite and thick-bedded mudstone and thin-bedded siltstone occur at GR 4697 71001. The middle facies is dominated by upward-fining siltstone with minor arenite and mudstone. The arenites are typically medium grained, feldspathic and volcanolithic in composition. The upper facies contains strongly bioturbated mudstone, shale, siltstone, minor feldspathic

134

L. C. CRANFIELD ETAL.

arenite and rare lithic arenite. This facies contains thin (3 cm) coarse-grained, quartz-bearing volcanolithic arenite that scours underlying siltstone. Even, very thinly interbedded arenite and mudstone (2-3 cm in arenite, 1-2 cm in mudstone) occur at the base, middle and top of the formation. Thick-bedded feldspathic arenite (to 40 cm) occurs closer to the top of the unit. Thick-bedded granule to pebble conglomerate with pink quartzite pebbles in a fine clayey and ferruginous matrix and feldspathic arenite occur at the top of the unit. The source of the Tamaree Formation is acid to intermediate volcanics of the Rammutt Formation. Clasts in the unit are partly sericitised dacitic volcanics?, sericitised plagioclase, quartz, minor K-feldspar and rare partly chloritised volcanics. The percentage of quartz apparently increases upward through the unit. A prominent cleavage in the unit is represented in thin-section as black, anastomosing, stylolitic dissolution seams. Quartz clasts are strongly recrystallised, forming sub-grain boundaries. The top of the Tamaree Formation in the area immediately north of Six Mile Creek is predominantly thick-bedded and grades upwards into conglomerate of the Keefton Formation. Keefton Formation The Keefton Formation forms a continuous belt from the Belli Park area in the south to Coles Creek. North of Coles Creek, the unit is discontinuously exposed to south of Gympie in the Noosa road area. Outcrop in general is poor and the Keefton Formation often occupies low-lying country. The Keefton Formation is a strongly cleaved unit of coarse to pebbly lithic arenite with interbedded conglomerate and dark-grey shale. Red, locally pyritic, shales are interbedded with conglomerate in the type area along the North Coast Railway. There are no red shale beds in the unit north of Six Mile Creek. The conglomerates are poorly to moderately well sorted, and the arenites are moderately well sorted. Thin-sections show that the clasts consist of angular quartz (sharp extinction), plagioclase, moderately rounded composite quartz-rich metasediment, chert, rare mica, pyroxene and acid volcanics. The percentage of different clasts is often difficult to determine, particularly in the arenites, because clasts of very fine-grained recrystallised volcanics and chert are not readily distinguishable. Clasts indicate a source in the Rammutt Formation and rock types similar to the Amamoor beds. Original grain shapes of component clasts are commonly modified by stylolitic pressure solution. In coarser grained rocks this pressure solution mimics foliation in the interbedded pelites. The relationship between the Keefton Formation and the Gympie Group is unclear because of poor exposure. Near the West of Scotland and Mt Wonga mines, and north of Six Mile Creek the boundary between the Tamaree and Keefton Formations appears conformable. At these localities, the top of the Tamaree Formation contains thick-bedded arenite and conglomerate that

passes upwards into conglomerate and arenite of the Keefton Formation. Both units have similar bedding orientation and are lithologically very similar. At Mt Tuchekoi conglomerate of the Keefton Formation is faulted against the Amamoor beds. Kin Kin beds The Kin Kin beds comprise strongly cleaved and deformed mainly fine-grained metasediments. Topographically higher areas are faulted or hornfelsed by various intrusive bodies. A major foliation S] is characteristic of the unit. This foliation is locally overprinted and folded by a penetrative foliation S and by kinks of different orientations (Cranfield & Scott 1993). The type area designated by Murphy et al. (1976) was supplemented with reference sections from Gympie North railway station, Coles Creek Road and Neerdie Road by Cranfield and Scott (1993). At Gympie North Railway Station, olive green-brown and black phyllitic metasediments are intercalated with lenses and thin beds of quartz—volcaniclastic metaarenites and rare beds of calcareous meta-arenites containing unidentifiable crinoid and shell fragments. Bedding is not clearly evident and the most prominent feature is a strong foliation overprinted by multiple kinking events. Quartz veining is associated with late kinking and crenulation. This locality best matches rock types of the type area and is included because of the extent of excellent uninterrupted exposure. In thinsection foliation transects at low angles or is parallel to bedding. The foliation is highlighted by micas and goethite that anastomose around deformed, resistant angular quartz (15—20%), cryptocrystalline— recrystallised volcanics, chert? (25—30%) and plagioclase (5-7%) clasts. Clast shape is modified by pressure solution, corrosion, and recrystallisation to form sub-grain boundaries. Other components of the meta-arenites and metapelites are minor hornblende and tourmaline. Locally, secondary crenulations overprint the dominant foliation. The ubiquitous mica varies in abundance and type. Where the dominant foliation is defined by chlorite or opaques along stylolites, relict muscovite locally defines bedding. In other sections relict green biotite phenocrysts are rotated or partially rotated parallel to the dominant foliation. Locally, biotite is degraded and partially or totally replaced by chlorite-muscovite intergrowths. Conglomerate at GR 4748 70994 comprises clasts of coarse, dark, wellrounded concretionary lithics and red angular metapelite in grey or red muddy matrix. The metapelite clasts are very similar to metapelite of the Amamoor beds. Ridley (1962) and Wilson (1968) described andalusitemuscovite hornfels and sillimanite-cordierite hornfels contact-metamorphic assemblages from the aureole of the Goomboorian Intrusive Complex. Biotite-staurolite hornfels occurs along the southeastern margin of the Woondum Granite. The Coles Creek Road section south of Cooran is typical of the Kin Kin beds south of the Woondum Granite. Rock types are laminated dark-grey metapelite 2

and pale-coloured fine-grained meta-arenite. Foliation is well developed but not as pronounced as at Gympie North Railway Station. Bedding is clearly evident over a distance of 1 km along Coles Creek road and numerous changes in vergence between bedding and foliation indicate a west-directed sense of tectonic transport for this southern domain (Cranfield & Scott 1993). Bedding to cleavage relationships at this locality also highlight ductile shear zones. These narrow (5-7 cm wide) zones contain preserved tight fold hinges. Adjacent beds have abrupt changes in orientation and chaotic distortion, which suggest incomplete lithification during deformation (Scott & Cranfield 1993). At outcrop scale, Coles Creek and Gympie North Railway Station areas can be readily differentiated. Thin-sections from both areas are texturally and mineralogically similar with clasts of quartz, cryptocrystalline recrystallised aggregates (volcanics? + some chert?), plagioclase and mica (green biotite, chloritemuscovite intergrowths, chlorite, muscovite)). In both areas the foliation is defined by alignment of fine micas, reoriented mica xenocrysts and stylolitic? concentrations of insoluble opaques. Differentiation of bedding from foliation and diffraction of foliation is more apparent across coarser grained beds at Coles Creek. Neerdie Road exposures are medium to thick, regularly interbedded, light-buff, meta-arenites and metapelites. Thicker and coarser grained beds in this area are atypical of most of the Kin Kin beds and they have well-preserved sedimentary structures (mainly ripple cross-laminations and burrows). A single dominant foliation is evident at Neerdie. Near Mt Goomboorian, a meta-arenite bed is transposed parallel to the main foliation, and this foliation is overprinted by a later (second generation) foliation. The arenite beds are coarser grained than elsewhere in the Kin Kin beds, but they contain essentially the same clasts. Northwest-trending magnetic lineaments from geophysical imagery over the Kin Kin beds are a series of dioritic dykes (R. Cunneen, pers. comm. 1995). The apparent increase in deformation in the Kin Kin beds compared with rocks of the Gympie Group led previous workers to suggest an Early Palaeozoic age for much of the Kin Kin beds, with the exception of a western belt (which included an Early Triassic marine fossil locality). Approximately 2.2 km east of Gympie Railway Station, Ball mapped a north-northwesttrending 'Gympie Great Fault', separating shales to the west that he identified as Gympie Group from phyllite of probable Early Palaeozoic age to the east. Rollason (in Day et al 1974) also believed that a major northnorthwest-trending thrust fault separated an older eastern sequence from less-deformed fossiliferous Early Triassic rocks at Woondum. Harrington (1983) accepted the presence of the Gympie Great Fault and suggested that the Kin Kin beds east of the fault are correlatives of the Late Permian Tamaree Formation. To check whether the Kin Kin beds contain stratigraphic units of significantly different age and metamorphic grade, combined petrographic and illite crystallinity studies were undertaken by Cranfield and Scott (1993) in an

GYMPIE PROVINCE, QLD

135

east-west section across the Kin Kin beds. The evidence, although inconclusive, showed that there was no systematic variation in rock type or metamorphic grade from east to west across the Kin Kin beds, except for thick sandy units in the Neerdie area. This study also appeared not to support the presence of the Gympie Great Fault and the Kin Kin beds are regarded as an entirely Triassic unit. The boundary between the Kin Kin beds and the underlying Keefton Formation is faulted north of Six Mile Creek and sheared at Mt Wonga mine. Brooweena Formation The Brooweena Formation contains both marine and continental deposits. Marine deposits, which form the lower part of the unit, are mainly fine-grained silty to shaly clastic sediments with very rare limestone lenses. The continental deposits are fluvial clastic sediments containing quartz, feldspar and chert as the main clasts apparently derived from a source area dominated by metamorphic rocks and granite. The Brooweena Formation contains rare (mainly andesite) volcanic flows, but no obvious pyroclastics. The unit most commonly dips moderately to the east with local dip reversals (steep westerly) at fault zones. There is apparent conformity between the marine and continental parts of the unit. The boundary between the Brooweena Formation and the Gympie Group is poorly exposed and commonly sheared. It is disconformable or unconformable over the Gympie Group to the south of Biggenden and is locally faulted in other areas. PROPOSED EVOLUTION OF THE GYMPIE PROVINCE Geochemistry of basaltic volcanics Geochemical and Nd-isotope studies (Sivell & Waterhouse 1988; Sivell & McCulloch 1996) indicate that the Highbury Volcanics were island-arc tholeiites derived from subduction in the Early Permian. Ndisotopes from Sivell and McCulloch's (1996) 'Amamoor beds' and Cedarton Volcanics show typical chemistry of MORB and backarc basalts. Recent mapping by Donchak et al. (1995) shows that Sivell and McCulloch's 'Amamoor beds' would be now included in the Permian Cambroon beds of Sliwa (1994) and that the Cambroon beds may be in part equivalent to the Cedarton Volcanics. The variable nature of geochemistry of undivided Gympie Group volcanics in the Gayndah region noted by Cranfield and Murray (1989) show that these rocks have mixed geochemical signatures indicative of both island-arc tholeiites and backarc basalts. This may be also the result of juxtaposition of basalt from Early Permian extensional basins with arc-related basalts as a result of Early to Middle Triassic convergence during the HunterBowen Orogeny.

136

L. C. CRANFIELD

ETAL.

Fossil evidence Cranfield (1990) summarised the main fossil evidence from the Gympie Province. Critical features are as follows. (1) The Early Permian Rammutt and Kolbar Formations in the Gympie Province are about the same age (Early Permian) as nearby Early Permian Cambroon beds and Cedarton Volcanics. (2) Cambroon beds, Cedarton Volcanics and undivided Early Permian rocks west of Gigoomgan are dominated by Eurydesma spp. indicating a shoreline facies. This contrasts with index fossils from assemblages in the Rammutt Formation that indicate of neritic environment depths (Palmieri 1987). (3) The unconformity at the base of the 'Conglomerate Group' of the Rammutt Formation indicates that the 'Conglomerate Group' is mid-Permian in age. The age of the South Curra Limestone from work by Waterhouse and Balfe (1987) is mid-Permian. Palmieri (1987) showed that the South Curra Limestone and Gigoomgan Limestones range from mid- to Late Permian (Ufimian) (South Curra Limestone) to Kazanian (Gigoomgan Limestone). Fossil evidence (Palmieri 1988) from lenses of undivided Gympie Group limestone in the limb areas of fold axes west of Gigoomgan shows that some lenses are the same age as the Gigoomgan Limestone [i.e. Late Permian (Kazanian)]. Based on sparse collections the Tamaree Formation was considered by Waterhouse and Balfe (1987) as probable mid- to Late Permian in age. This age was updated to Kazanian by Palmieri (1988). The gradational contact between conglomerates of the Tamaree and Keefton Formations north of Six Mile Creek suggest that the Tamaree Formation may range into the Early Triassic. The age of the Keefton Formation was determined by plant fossils by Dear (1962). The Brooweena Formation contains marine deposits with macrofaunas of probable Early Triassic age (Fleming 1976). The continental deposits contain macrofloras that are similar to those of the Toogoolawah Group to the west (early Middle Triassic). Attempts to obtain miospores from the Brooweena Formation during drilling programs in the Maryborough Sheet area proved fruitless (Cranfield 1994). The only known fossils from the Kin Kin beds are ammonites (marine) of Early Triassic age collected from the old Woondum Railway Station site (Runnegar & Ferguson 1969). The age of the Highbury Volcanics is unknown and it is unconformably overlain by the Rammutt Formation. The evolution of the Gympie Province probably commenced in the latest Carboniferous or earliest Permian when an oceanic island arc developed east of the Australian continental margin. Figure 4a suggests that the basal Camboon Volcanics of the Bowen Basin may have developed in a backarc environment at possibly much the same time as the early evolution of the Gympie arc. Development of the Bowen Basin in eastern Queensland has been discussed as part of a three-stage process by Baker et al. (1993) and Fielding et al.

(1994). The first stage of the Bowen Basin outlined in Fielding et al. (1994) was a phase of extensional subsidence and magmatic activity. In the Gympie Province this phase is probably represented by fine- and coarse-grained sedimentary rocks and bimodal andesitic to felsic lavas of the Rammutt Formation. At about the same time as the Rammutt Formation was being deposited in the Gympie Province, the Cambroon beds and Cedarton Volcanics originated in a shallow-marine extensional basin developed on rocks of the North D'Aguilar Block (Figure 4b). In the Gympie Province, the Rammutt Formation has abundant examples of low-energy deposition. Horizontal bedding, and low-angle ripple crosslamination are prominent features of this formation. Conglomerate at the top of the formation has a welldeveloped imbrication and may be an alluvial-fan deposit. Measurements of rare palaeocurrent data from the Rammutt Formation indicate a source area from the west or east (Cranfield & Scott 1993). Variations in clast composition (Cranfield & Scott 1993) indicate multiple local source areas for the unit. The unconformity separating stage 1 from stage 2 in the Gympie Province probably occurs at the base of the 'conglomerate facies' of the Rammutt Formation. The main sag phase (stage 2) in the Bowen Basin (Figure 4c) is represented in the Gympie Province by the South Curra Limestone and Tamaree Formation. Sedimentary structures in the South Curra Limestone are dominated by planar bedding and minor rip-up clasts. Bioturbation is common in the basal calcarenite facies of the limestone and disrupts bedding. Palaeocurrent data from the Tamaree Formation (Cranfield & Scott 1993) indicates a source from the west. Roach (1990) recorded wavy, graded, small-scale crossbedding, ripple laminations, bioturbation, flame structures and parallel, flat laminations in the Tamaree Formation. He indicated that the unit had more features of tempestites than turbidites and regarded the lack of hummocky cross-stratification as indicating that the Tamaree Formation was deposited distally to the shoreline. Stage 3 in the Gympie Province is coincident with the onset of foreland loading that accompanied the start of the Hunter Bowen in the Bowen Basin. This stage formed structural zones in the Gympie Province.

Structural zones The onset of the Hunter-Bowen orogeny in the Gympie Province (Figure 4d) is reflected in the development of shallow east-dipping thrust faults and intense cleavage development in the east (Eastern zone), shallow dips in the central area (Central zone) and steep dips, overturned strata and thrusting along the western margin of the province (Western zone) (Figure 5). Deformation in the Eastern Zone is best recorded in the Kin Kin beds, but occurs in other units as an anastomosing stylolitic dissolution (Rammutt and Tamaree Formations, South Curra Limestone) and as recumbent folds (South Curra Limestone near Curra

GYMPIE PROVINCE, QLD

137

(b) Early Permian - extension Future Esk Trough

Future Brooweena Keefton Fms ? Kin Kin beds Sea Level

Sea Level

Accretionaiy Wodga

Gigoomgan Umestone, Tamaree & Teebar Formations

(c) Late Permian - sag phase

Figure 4 Stages in the evolution of the Gympie Province, (a) Late Carboniferous to Early Permian, (b) Early Permian, extension phase, (c) Late Permian, sag phase, (d) Early to Middle Triassic, Hunter-Bowen Orogeny.

(d) Early - Middle Triassic - Hunter Bowen Orogeny

Quarry). Of the Triassic units in the Gympie Province there is an increase in deformation in the south; the Keefton Formation and the Kin Kin beds both have a strongly developed cleavage whereas cleavage in the

Brooweena Formation is only weakly developed. The difference between the intensity of deformation of the Kin Kin beds - Keefton Formation and the Brooweena Formation may be due to faulting along a possible

138

L. C. CRANFIELD ET AL.

VlJU J J*. /

Mcilange j |

Amamoor

ml

beds

11

S

S H E j

V?

Western Zone

Central Zone

1L\

Eastern Zone

SO/S1 SURFACE IN GYMPIE PROVINCE

Figure 5 Structural zones of the Gympie Province.

transform fault between the two parts of the Gympie Province. The shallow-dipping Central zone is typified by the Bruce Highway section north of Rammutt Road and the axial area of the Gioomgan Anticline. In this zone the units are shallowly dipping and only weakly cleaved. Thin-sections show that the anastomosing stylolitic cleavage is still present in this zone. The Hunter-Bowen Orogeny is accompanied by porphyry-style copper and gold mineralisation and skarn mineralisation in the Gympie Province. The Western zone is a major zone of shearing and melange development along the western margin due to collision of the Gympie Province with the New England Orogen in the Middle to Late Triassic apparently as a result of the Hunter-Bowen Orogeny. The shear fabric in the southern part of this zone dips steeply to the south-west (Cranfield & Scott 1993). This orientation is probably due to reactivation of Early Permian extensional faults (formed during deposition of the Rammutt Formation and Cambroon beds/Cedarton Volcanics) as thrusts during the foreland loading phase. Similar thrust reactivation structures of early formed extensional features have been reported from the Bowen Basin by Korsch (1996). In addition reorientation of the boundary between the Gympie Province and New England Orogen was affected by Cretaceous extensional faulting that formed the Coral Sea. The first cover sequences that post-date the HunterBowen Orogeny are a series of extension-related highlevel granite plutons (Cranfield 1994) and associated felsic volcanics (Aranbanga Volcanic Group and North Arm Volcanics). The granite plutons and volcanics are associated locally with a major mineralisation epoch in the Gympie Province. MINERALISATION IN THE GYMPIE PROVINCE Mining history The history of mining in the Gympie Province is dominated by the gold and silver in the Gympie

§

£

Goldfield. Other significant areas of mining have been in the Biggenden area, Cordalba area, Ban Ban area and the Glenbar-Marodian area. The mining history in the Gympie Goldfield is summarised by Barker et al. (1993). The major producing mines were the Scottish Gympie (19.2 t Au bullion); Great Eastern No. 2 South (10.5 t Au bullion); Glanmire and Monkland South (6.3 t Au bullion); and the Phoenix No. 1 North (6.1 t Au bullion). In the northern part of the Gympie Province gold was first worked at the Mt Biggenden Mine in the Biggenden area in 1888. This mine was reopened as a magnetite mine in 1966. Since 1980, Devex Limited, alone or with various joint-venture partners, has explored the Gympie area, focusing on the Inglewood structure at the southern end. The West of Scotland shaft has been reconditioned, extensive underground exploration undertaken and a panel of ore blocked out. Production from 1995 to date is of the order of 120 000 t with an average grade of 8 g/t Au. Mineral exploration Major contributors to literature on the economic geology of the Gympie Goldfield were Aplin (1868), Rands (1889, 1891, 1894, 1899, 1901) who described and mapped the Gympie Goldfield in detail, and Dunstan (1910, 1911) who published detailed topographic and geological maps of the goldfield. Dunstan made an extensive collection of mine plans and sections preparatory to producing a series of geological sections of the goldfield. R. L. Jack (from 1886 to 1905), L. C. Ball (from 1901 to 1947), C. C. Morton (from 1920 to 1950), A. K. Denmead (from 1932 to 1987) and other geologists of the Geological Survey of Queensland inspected and reported on mining and mining proposals in the goldfield throughout the late 1800s to mid-1900s, most reports appearing in the Queensland Government Mining Journal and summarised in Barker et al (1993). Recent activity on the goldfield, centred on the West of Scotland shaft, has been summarised by Kitch and Murphy (1990) and Cunneen (1996). Siemon (1974)

GYMPIE PROVINCE, QLD

139

described the Neardie antimony deposits, and Krosch (1975) outlined limestone deposits of the Gympie Group at Gympie. Scott et al (1990) summarised mineralisation trends and exploration in the Gympie Province. A summary of exploration in the Gympie Province in the Gympie 1:250000 Sheet area prior to 1976 was incorporated in the description of the economic geology of the Gympie 1:250000 Sheet area given by Murphy et al (1976). Scott (1989) summarised company exploration for the Gympie, Goomeri, Nambour and Nanango 1:100000 Sheet areas for the period 1976 to 1988, and Barker et al (1993) summarised exploration to 1992. In the northern Gympie Province exploration prior to 1968 was described by Ellis (1968), from 1968 to 1979 by McLeod (1979) and from 1979 to 1991 by Cranfield and Garrad (1991). Mineralising epochs in the Gympie Province Mineralisation in the Gympie Province is dominated by gold; sulfide mineralisation is subsidiary and is associated mainly as accessory minerals. Mineralising epochs are genetically related to the intrusion and extrusion of Early to Middle Triassic and Late Triassic plutonic and volcanic rocks. The earliest mineralisation is apparently coincident or immediately post-dates the main foreland-loading deformation associated with the Hunter-Bowen Orogeny. This mineralisation is represented by thrust-related veining and porphyry mineralisation associated with the Early to Middle Triassic Calgoa Diorite and smaller intrusions of dioritic composition. Mineral deposits are porphyry-style copper-silver—lead—gold and gold deposits of the Calgoa and Coalstoun areas. These occur within the Calgoa Diorite close to the contact between the North D'Aguilar Block and the Gympie Province. Skarn mineralisation and vein deposits are associated with this event, but there was little wall-rock alteration. In contrast the Late Triassic mineralisation is associated with metasomatism in the case of skarn deposits and development of more extensive wall-rock alteration (argillic, propylitic and sericitic) in vein and intrusive contact-brecciated deposits (Cranfield & Garrad 1991). Mineral commodities include precious metals, copper, lead, zinc, antimony, manganese, cobalt, molybdenum, mercury, barite and gemstones (Figure 6). EARLY TO MIDDLE TRIASSIC MINERALISATION

Precious-metal-bearing quartz veins (some associated with granitic intrusive rocks) are hosted by Permian volcaniclastic and sedimentary rocks and Early Triassic sedimentary rocks. Gold also occurs in alluvial deposits derived from the gold lodes. Base-metal-bearing quartz veins are hosted by Triassic intrusive rocks. Antimony mineralisation occurs in Permian limestone and Early Triassic sedimentary rocks. Most of the major vein mineralisation is apparently associated with porphyry intrusion that accompanied or immediately post-dated the Hunter-Bowen Orogeny

Figure 6 Mineral occurrences in the Gympie Province.

Gold mineralisation in this epoch occurs in five main styles of deposits: (i) quartz (fissure) veins in volcaniclastic, sedimentary and metasedimentary rocks that have no apparent relation to intrusive rocks (e.g. Gympie Goldfield, Stanton Harcourt); (ii) quartz vein mineralisation hosted by the Brooweena Formation at the contact with small diorite intrusions at Glenbar, Dawn, Moonlight and Mt Scougall mines; (iii) disseminated stockworks and breccia deposits at or near intrusive contacts; (iv) skarn deposits hosted in mid- to Late Permian limestone at or near the contact with the Early to Middle Triassic granitoid intrusion of the Calgoa Diorite at Mt Allen, Mt Suthers and Munna mines; and (v) Holocene alluvial and eluvial deposits.

140

L. C. CRANFIELD ETAL.

Table 2 Styles of mineralisation in the Gympie Province. Mine/ prospect

Mineralisation, alteration and production

FISSURE VEIN DEPOSITS

Gympie Goldfield Mineralisation probably Early-Middle Triassic in age, associated with low temperature hydrothermal fluids from diorite to dolerite intrusions. Gympie veins: parallel tension gash veins N of Inglewood Fault. Gympie veins are higher grade than Inglewood lodes and highest at 'carbonaceous shale breaks'. Inglewood lodes are lower grade but less dependent on lithology. They form veins and stockworks on NE margin of the Inglewood structure. Reticulated fine gold production from Gympie Goldfield including alluvial gold (fineness value = 820) is 105 258 kg. Current silver estimates of Gympie bullion are 7.5-10%, indicating a silver yield between 9000 and 13 000 kg. Queensland Department of Mines (1953) estimated alluvial production 1867-1868 at about 2600 kg of gold bullion. Lees (1899) estimated 9128 kg of gold bullion was obtained from 1867 to 1872. Denmead (1933) believed that this was 'probably nearly all' alluvial gold. Kin Kin area Au—Py—Pbs—CuFeS in N-NW-trending, shallow SW to moderately E-dipping quartz reefs mainly in N Kin Kin beds. Dykes are common and locally control gold grades. Propylitic alteration is common in gold-bearing veins. Quartz vein orientation is similar to Gympie veins. Early to Middle Triassic source? Stanton Harcourt Au: fissure veins associated with faults /shear zones; Goldfield lodes trend NE, E and NW, worked 1886-1901, 1920 and sporadically to present. 2.9 kg Au. Early to Middle Triassic source? Dallarnil Silver Ag-Pb: fissure veins along NW-trending fault. Early to Mine Middle Triassic mineralising event? 2

Young Australian Mine Commonwealth (Mt Havilah)

Au: quartz-sericite-pyrite veins and breccia at contact with aplite and quartz-feldspar porphyry dykes. Early to Middle Triassic mineralising event? Cu—Fe-Ag—Au-Bi: shallow dipping fissure veins at contact with Late Triassic Degilbo Granite

Glenbar Goldfield Au-Pb-Ag-Zn-Cu-Fe-Sb: vertical NW-trending stibnite lode at Glenbar; lodes along NW-trending shear zones elsewhere, mineralised quartz-calcite veins of variable orientation at XYZ Mine; K-silicate alteration, silicification; production from this part of Glenbar field -10 kg Au and 58 t antimony. Late Triassic source? INTRUSIVE CONTACT DEPOSITS

Mt Scougall

Defiance (Bonnie Jan)

Au-Cu-Fe-Zn: fracture fill in coarse acid tuff at contact with granodiorite; worked 1925, 1952, 1960-1962; produced 2701 of ore for 622 g Au. Late Triassic volcanics. Au-Cu-As: brecciation, followed by tourmalinisation silicification, deposition of silica and sulphur minerals, chloritisation and carbonate vein deposition; current mining leases. Intruded by Late Triassic Broomfield Granite

Host rocks

References

Gympie Group; dominantly Rammutt Formation, Highbury Volcanics

Barker etal. 1993; Denmead 1933; Queensland Dept Mines 1953; Murphy et al. 1976; Lees 1899; Roach 1990

Kin Kin beds, highly altered basic to intermediate dykes.

Ball 1933; Barker etal 1993; Cribb 1935a, b, 1936; Dunstan 1911

Undivided Gympie Group volcanics

Ball 1901: Cranfield & Garrad 1991; McLeod 1979; Morton 1920b, 1930b, 1934a; Rands 1886c Cranfield & Garrad 1991; Cribb 1938

Undivided Gympie Group volcanics Undivided Gympie Group

Ball 1903; Cranfield & Garrad 1991; Ellis 1968

Undivided Gympie Group Volcanics Brooweena & Kolbar Formations

Ball 1902a, 1904, 1915, 1936; Cranfield & Garrad 1991; Dunstan 1907 Ann. Rept Dept. Mines 1917, 1920-1922, 1931, 1933; Ball 1936; Cranfield & Garrad 1991; Cribb 1945; Dunstan 1916a; Morton 1920a, 1933; Wallis 1990

Ooramera Volcanics

Ann. Rept. Dept Mines 1926, 1960, 1962; Cranfield & Garrad 1991;Doherty 1986; Stapelton 1961 Ball 1947; Brooks 1962a, b; Cranfield &Garrad 1991; de Havelland 1987; Ellis 1968; Huber 1979; Krosch 1978, 1979, 1980; McLeod 1979; Still well 1951; Woodcock & Dunkin 1952

Brooweena Formation

GYMPIE PROVINCE, QLD

141

Table 2 continued. Mine/prospect

Mineralisation, alteration and production

Host rocks

References

DISSEMINATED-STOCKWORKS-MILLED BRECCIA DEPOSITS

Mt Ideal

Au-Ag: fractured, central argillic zone flanked by siliceous zones NW & S; chlorite zone weakly developed; some sericitic alteration; worked intermittently 1895-1933. Late Triassic Broomfield Granite

Brooweena Formation

Ball 1947; Brooks 1962a; Cranfield &Garrad 1991; de Havelland 1987; Huber 1979; Krosch 1979, 1980; McLeod 1979; Morton 1935; Stillwell 1951; Woodcock & Dunkin 1952

Bullant

Ag-FeAsS-Au: breccia fill and veins, argillic, sericitic alteration. Late Triassic mineralisation

Brooweena Formation

de Havelland 1987; Denmead 1933; McLeod 1979

Wild Irishman

Au: breccia-fill deposit with veins; worked 1883, 1933-1936; 1361 ore for 2.38 kg Au. Later Triassic mineralisation

Undivided Gympie Group

Cranfield & Garrad 1991; Cribb 1936; de Havelland 1987; Ellis 1968; McLeod 1979

Mt Biggenden

Au-Cu-Bi-magnetite: metasomatic skarn at the contact; >186 kg Au, >4 t magnetite; contact with Late Triassic Degilbo Granite

Undivided Gympie Group volcanics

Ball 1902a, 1904, 1936; Bottomer 1970; Clarke 1963, 1969; Cranfield & Garrad 1991; Forster 1985; Hudson 1966; Siemon 1971; Siemon etal 1977

SKARN-TYPE DEPOSITS

Ban Ban Zinc Lode

Zn-Pb-Cu-Ag: metasomatic hydrothermal skarn; chlorite alteration at the contact with a probable Late Triassic granodiorite

Undivided Gympie Group volcanics

Ashley 1980; Ashley & Plimer 1988; Ball 1910; Brooks et al. 1974; Dunstan 1916a; Daviesl970, 1975; Green 1975; Horton 1978, 1982; Loutheanl992; McLeod 1979; Siemon etal. 1977; Stanton 1987

Mt Suthers

Cu-Ag-Au: skarn, minor magnetite-rich body; worked prior to 1917. At contact with Early-Middle Triassic Calgoa Diorite

Gigoomgan Limestone

Cribb 1945; Hill 1985; McLeod 1979; Wilcox 1917; Ziebell 1972

Mt Allen

Cu-Ag-Au-W: calc-silicate skarn with garnet and epidote at contact with Early -Middle Triassic Calgoa Diorite

Gigoomgan Limestone

Ball 1902b; Dunstan 1913; McLeod 1979; Morton 1926; Rands 1890a; SaintSmith 1917

Munna

Au-Cu-W: skarns containing wollastonite, garnet and scheelite at contact with Early-Middle Calgoa Diorite

Gigoomgan Limestone

Ball 1902b, 1915; Cranfield & Garrad 1991; Dunstan 1913; Ellis 1968; Rollason 1970

PLACER DEPOSITS

In 1867-1868 about 2600 kg of alluvial gold was mined in the central part. Alluvial platinum has been found near the N end of the Gympie Goldfield. Devex Limited (1987) estimated 150x106 m3 of Mary River alluvium in AP 3991M and 4255M. Gympie Gold Dredging No. 1 NL defined a measured resource of 9.5x106 m3 grading 0.263 g/m 3 gold. From 1981 to 1991 a ).7—1.8 m thick wash layer in North Creek N of Gympie yielded 0.3 to >1 g/m 3 gold.

Dunstan 1911, 1913; Barker et al. 1993; Devex Limited 1987

Stanton Harcourt

Eluvial and alluvial deposits near fissure veins

Ball 1901

Booyal

Mainly alluvial gold at the contact of Tawah Granite with limestone

Gympie Goldfield

142

L. C. C R A N F I E L D

ETAL.

The majority of the gold produced in the Gympie Province was from quartz reefs cutting rocks of the Rammutt Formation of the Gympie Group in the Gympie Goldfield. Two main lode types were recognised by Kitch and Murphy (1990) — 'Gympie vein' quartz lodes and 'Inglewood' lodes.

the classification between fineness and broad deposit type developed by Morrison et al. (1991), the Gympie deposits (fineness of 820) is more closely allied with the 'plutonic' type (average fineness = 825) than with the 'slate belt' type (average fineness = 940). Roach (1990) reported the silver content of the bullion to average 7.5— 10%. Murphy et al. (1976) noted that the Gympie mineralising event(s) probably post-dated folding of the Gympie Group and suggested a Middle Triassic age. If mineralisation is linked with the emplacement of the Woondum Granite, then a Late Triassic age applies. Melsom and Rowlands (1988) and Kitch (1989), on the basis of overprinting evidence, considered Gympie vein lodes pre-dated Inglewood lodes. Kitch and Murphy (1990), however, reported the age relationship between the two styles to be uncertain. Recent work by Gympie Eldorado Gold Mines (Cunneen 1996) has indicated that Gympie veins and the Inglewood lode developed contemporaneously and that the 'Inglewood Fault vein system' was the feeder and controlling structure in the southern part of the field. Roach (1990) considered that the gold mineralisation appeared to post-date the deposition of the sulphide minerals.

Gympie veins Gympie veins were the main source of gold in the Gympie Goldfield. About 70 reefs occur in a 3 km wide by 10 km long northwest-trending zone. Gympie veins, described by Kitch (1989) as a 'tension gash vein system' typically strike north-northwest to north and dip steeply south-southwest to west and follow the most prominent joint direction. Higher grade Gympie veins occurs at intersection with beds of carbonaceous shale or siltstone in the Early Permian Rammutt Formation (Rands 1889, 1891, 1894, 1901; Morton 1930). The host rocks and Gympie veins have a similar strike and, as the Rammutt Formation in this area dips consistently to the east at about 20°, the payable ore zones form gently pitching shoots. Patches of high-grade ore were attributed to intersection of the reefs with graphitic bedding-plane faults or were apparently related to later cross-cutting faults (Denmead 1960). The grade control may be related to reduction due to carbonaceous composition and the characteristic deformation of the host rocks (Kitch & Murphy 1990). The veins commonly occupy fault zones and locally shared these zones of weakness with altered andesite dykes (Queensland Department of Mines 1953). Wilson (1987) suggested a granitic source for the mineralising fluid based on the presence of tellurides and sulfides. Oxygen-isotope data for quartz are consistent with either a magmatic or metamorphic source (Golding et al. 1987). Platinum is associated with gold, pyrite and galena at the Lady Mary North (Warren Hastings P.C.) mine and in nearby alluvial deposits (Dunstan 1913). Wilson (1987) postulated that the platinum might have been scavenged from ultramafic rocks lining hydrothermal fluid pathways at depth beneath the goldfield. The fineness of gold bullion from Gympie was reported to average about 820 (Fisher 1945). Based on

Inglewood lodes The northwest-trending Inglewood structure is a major, steeply dipping fault zone that contains a series of dykes as well as the Inglewood reef. The Inglewood structure formed along the northeastern margin of a broad, northwest-trending shear zone (Kitch 1989). The structure comprises a fault zone intruded by a complex series of fractionated dykes up to 30 m wide; it dips steeply to the northeast at the surface but, on its southeastern end it rolls over at depth and dips steeply to the southwest. Other possible analogues to the Inglewood lodes are: the Sovereign and Dawn 'Crosscourses' (faults). The Sovereign Crosscourse (Dunstan 1911) contains an andesite dyke along part of its length. The Dawn Crosscourse is a fault that hosts a quartz-veined andesite dyke. Morton (1934) showed that gold grades in the Dawn reef close to the Dawn Fault (a southwest-dipping fault) were dissimilar to Gympie veins and that grades were unrelated to the host volcanic stratigraphy. The Inglewood lode comprises vein quartz, minor calcite and brecciated volcanic rocks of the Rammutt Formation or dyke rock (Kitch 1989). The quartz is generally massive, with some laminated zones. The mineralisation is characterised by free gold, finer grained than the Gympie veins and pyrite, chalcopyrite and galena. Gold mineralisation in the Inglewood lode is broadly controlled by the host stratigraphy — the lode carries gold mainly where it cuts the upper (shalebearing) portion of the Rammutt Formation. No correlation of grade with individual shale beds has been found as elsewhere in the Gympie veins. Higher grade mineralisation occurs in steeply pitching shoots (Kitch 1989), possibly due to a subtle structural control. Recent radiometric age dating of dykes in the Inglewood structure indicates that mineralisation is Early to Middle Triassic in age (R. Cunneen pers. comm. 1995).

LATE TRIASSIC MINERALISATION

Late Triassic mineralisation in the Gympie Province is associated with metasomatic skarn-related deposits hosted by: (i) Early Permian limestones that occur with undivided Gympie Group volcanics at Mt Biggenden and adjacent areas (Mt Havilah, Mt Hastings, Mt Hudave): and (ii) hydrothermal alteration of Brooweena Formation — hosted gold and epithermal alteration of wall rock related to the Broomfield Granite. Table 2 outlines the mineralisation styles in the Gympie Province.

Southern Gympie Province GOLD DEPOSITS

GYMPIE PROVINCE, QLD

BASE METALS

Accessory sulfides in gold-bearing quartz veins are galena, sphalerite, chalcopyrite and tetrahedrite. Sphalerite is associated with quartz vein-hosted antimony mineralisation at Neerdie. Malachite coats joints in the South Curra Limestone at Curra Limestone Quarry, north of Gympie. Antimony mineralisation occurs at Neerdie, Chatsworth and the Blue Bell mine. In the Neerdie area, antimony ore has been mined at Simpsons, McMahons and Neardie mines. Neardie, one of the larger historical antimony mines in Queensland (Wallis 1993) contains primary stibnite ore in a series of north-trending, steeply dipping shear zones in the Kin Kin beds. Moderately north-plunging, high-grade ore shoots within the shear zones are structurally controlled and contain arsenopyrite, sphalerite and magnetite, and traces of gold and silver. At Neardie mine there is a mercury content of <4 ppm (Siemon 1974). Near Chatsworth, stibnite, minor gold and silver occur in narrow calcite-quartz veins in the South Curra Limestone. Production has been recorded from the Good Luck and Gympie Antimony mines. Molybdenite occurs in narrow quartz veins and disseminations in a Tertiary comendite plug and adjacent silicified Kin Kin beds and is disseminated along the margins of an aplite dyke that cuts the Woondum Granite (Cribb 1943). Molybdenite also occurs in the Great Eastern No. 2 South mine at Gympie associated with gold mineralisation in quartz veins (Dunstan 1913). Northern Gympie Province Mining in the Biggenden and Calgoa area occurred at a similar time to the commencement of mining at Mt Perry in 1860. Mineralisation in this area is hydrothermal and metasomatic and related to Early to Middle and Late Triassic intrusives (Calgoa Diorite and Degilbo Granite of Cranfield 1994). Associations are Cu-Au in the Early to Middle Triassic and Au-Cu-Bi at Mt Biggenden. Styles are either fissure veins (such as at Stanton Harcourt and Glenbar), disseminated stockwork breccias (Defiance and Mt Ideal) in the Brooweena Formation or skarn-related from contacts against limestones of the Gympie Group or placer deposits. Northwest-trending fissure-vein type lodes are a feature of mineralisation at Stanton Harcourt and in the Glenbar Goldfield. At Stanton Harcourt these veins apparently formed by permeation of mineralising solutions into fault and shear zones at the contact with the Tawah Granite. In the Glenbar Goldfield the veins are present at the contacts with small unnamed intrusions. Metal associations include zones of molybdenum, copper, copper-gold and epithermal gold. Common mineral association include copper with secondary carbonate enrichment zones and pyrite, arsenopyrite and copper sulfides. Some of these ore bodies contain gold while others (e.g. Mt Biggenden) contain molybdenite and gold.

CONCLUSIONS

143

The Gympie Province has been shown not to fit the evolution of New England Orogen and has phases of sedimentological development and faunal assemblages that closely link to the evolutionary sequence of the Bowen Basin. The early extensional phase of the Bowen Basin is represented in the Gympie Province by the Rammutt and Kolbar Formations. The change from the extensional to sag phase in the Bowen Basin is represented in the Gympie Province by the unconformity at the base of the 'conglomerate group' at the top of the Rammutt Formation. The sag phase of the Bowen Basin is represented in the Gympie Province by the deposition of the South Curra and Gigoomgan Limestones and the Tamaree and Teebar Formations. The onset of foreland loading in the Gympie Province is represented by continental-derived conglomerates in the Keefton and Brooweena Formations. The present close spatial relationship of arc-derived basalts of the Gympie Province with other backarc volcanics (Cedarton Volcanics and Cambroon beds) and the higher degree of deformation of the backarc volcanics suggests that they were deformed through reactivation of early extensional faults as thrusts during the Hunter-Bowen Orogeny. Mineralisation in the Gympie Province is coincident with or immediately post-dates the main Hunter-Bowen Orogeny and is of Early to Middle Triassic age and formed veins, porphyry mineralisation, and skarns in the Marodian area. Mineralisation in the Gympie Goldfield also probably occurred at this time. Late Triassic mineralisation is hydrothermal or epithermal in nature and is associated with high-level granite and felsic volcanics. More extensive wall-rock alteration accompanies this later mineralising event. ACKNOWLEDGMENTS This paper is published with the permission of the Director-General of the Department of Mines and Energy, Queensland. REFERENCES ASHLEY P. M. 1980. Geology of the Ban Ban zinc deposit, a

sulfide-bearing skarn, southeast Queensland, Australia. Economic Geology 75, 15-29.

ASHLEY P. M. & PLIMER I. R. 1988. T h e B a n B a n a n d

Willi Willi stratabound skarn deposits, New England Orogen: new data and comments on their formation. In: Kleeman J. D. ed. New England orogen—Tectonics and Metallogenesis, pp. 2 4 0 - 2 4 8 . Department of Geology and Geophysics, University of New England, Armidale. APLIN C. D'O. H. 1868. Report from the Government Geologist, south Queensland, on the geological and mining features of the Gympie Gold Field. Votes and Proceedings of the Legislative Assembly of Queensland 1868. BAKER J. C., FIELDING C. R., CARITAT P. D. E. & WILKINSON

M. M. 1993. Permian evolution of sandstone composition in a complex back-arc extensional to foreland basin: the

144

L. C. CRANFIELD

ETAL.

Bowen Basin eastern Australia. Journal of Sedimentary Petrology 63, 881—893. BALL L. C. 1 9 0 1 . On the Stanton Harcourt diggings and the Mount Shamrock Mine. Geological Survey of Queensland Publication 168 (also in Queensland Government Mining Journal 2, 83-88). BALL L. C. 1902a. On the Mount Biggenden gold and bismuth mine and the Paradise Goldfield. Geological Survey of Queensland Publication 173 (also in Queensland Government Mining Journal 3, 298—302). BALL L. C. 1902b. On Yorkey's Goldfield and the Marodian Gold and Copper Field. Geological Survey of Queensland Publication 179 (also in Queensland Government Mining Journal 3, 563-567. BALL L. C. 1903. Young Australian Gold Mine and Commonwealth Bismuth Mine, Degilbo. Geological Survey of Queensland Publication 1 8 5 (also in Queensland Government Mining Journal 4, 584-585, 641). BALL L . C. 1 9 0 4 . Certain iron ore, manganese ore and limestone deposits in the Central and Southern districts. Part 3. Geological Survey of Queensland Publication 194 (also in Queensland Government Mining Journal 5, 605-609). BALL L. C. 1910. Ban Ban Zinc. In: Progress Report for the year 1910, pp. 189-190. Annual Report of the Department of Mines Queensland 1910. BALL L. C. 1915. Copper deposits of the Wide Bay and Burnett Districts. Queensland Government Mining Journal 16,618. BALL L. C. 1933. Advance notes on a visit (extending from 1st to 7th February) to Gympie. Geological Survey of Queensland, Commodity File, Gympie—Gold 1, 15 (unpubl.). BALL L . C. 1936. Queensland mineral production. Various minerals indexed to localities on Queensland map. Queensland Government Mining Journal 37, Bismuth 4-5, Antimony 46-47, Manganese 92-93, Arsenic 120-121, Wolfram 156-157, Molybdenum 232-233, Mercury 272273, Barytes 376-377. L. C. 1947. New Ideal Claim Cordalba. Geological Survey of Queensland File 4-4-55 (unpubl).

BALL

BARKER R . M . , BURROWS P. E . , SCOTT M . , CRANFIELD L. C . & GENN D. 1993. Mineral occurrences—Gympie and

Laguna Bay 1:100 000 Sheet areas. Queensland Geological Record 1993 /14. BOTTOMER L. R. 1970. Aspects of the geology of the Biggenden magnetite deposits with particular reference to those in the Mount Hastings area. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). G. 1991. Geology of the South Curra Limestone, Gympie area, Southeast Queensland. BSc (Hons) thesis, University of Southern Queensland, Toowoomba (unpubl.).

BRODIE

1962a. Defiance Gold deposit Woco Creek Cordalba. Memo to Chief Government Geologist. Geological Survey of Queensland File 4-4-55. (unpubl.).

BROOKS J. H .

1962b Application for Assistance Defiance Mines Pty Ltd 62/1514. Memo to Chief Government Geologist 2/1 1/62. Geological Survey of Queensland File 4-4-55 (unpubl.). BROOKS J. H . , SYVRET J. N. & SAWERS J. D . 1 9 7 4 . Mineral resources of the Kilkivan district. Geological Survey of Queensland Report 60. BROOKS J. H .

D. 1964. The stratigraphy and faunas of the Permian marine sediments of the Gigoomgan area, Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). CAWOOD P. A. 1984. The development of the SW Pacific margin of Gondwana; correlations between the Rangitata and New England orogens. Tectonics 3, 539-553. CLARKE D. E. 1963. Geology of the Mount Biggenden Gold and Bismuth mine and environs. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). CLARKE D. E. 1969. The geology of the Mount Biggenden Gold and Bismuth mine and environs. Geological Survey of Queensland Report 32. CRANFIELD L. C . 1989. New Palaeozoic stratigraphic units in the Maryborough 1:250 000 Sheet area, southeast Queensland. Queensland Government Mining Journal 90, 115-120. CRANFIELD L. C . 1990. The Gympie Group, and other Permian strata of the Gympie composite terrane. In: Pacific Rim Congress 90 Proceedings, III, pp. 99-108. Australasian Institute of Mining and Metallurgy, Melbourne. CRANFIELD L. C . 1994. Maryborough, Queensland, Sheet SG56-6. Geological Survey of Queensland, 1:250 000 Geological Series Explanatory Notes. CRANFIELD L. C . & GARRAD P. D. 1991. Mines and prospects in the Maryborough 1:250 000 Sheet area. Department of Resource Industries Record 1991/3. CRANFIELD L. C . & MURRAY C . G. 1989. Geochemistry of basaltic rocks from the Gayndah region in the Maryborough 1:250 000 Sheet area. Queensland Government Mining Journal 90, 469-480. CRANFIELD L. C . & SCOTT M. 1993. Geology of the Gympie 1:100 000 Special area (9445, 9545). Queensland Geological Record 1993 / 20. CRIBB H. G. S. 1935a. Re New Veteran, Gympie. Geological Survey of Queensland Commodity File, Gympie—Gold 9, 2332. CRIBB H. G. S. 1935b. Gympie Gold N.L., G.M.L.1660. Geological Survey of Queensland Commodity File, Gympie—Gold 6, 1381. CRIBB H. G. S. 1936. Memo Chief Government Geologist re Wild Irishman Mine. Geological Survey of Queensland File 4-4-29 (unpubl.). C R I B B H. G. S. 1938. Dallarnil Silver Mine. Geological Survey of Queensland File 4-5-15 (unpubl.). BROWN C.

H. G. S. 1943. Re molybdenite on the Cooran Tableland. Geological Survey of Queensland Commodity File, Molybdenite—Cooran Tableland 4-44-5. C R I B B H. G. S. 1945. Glenbar goldfield and vicinity. Queensland Government Mining Journal 46, 109-112. CUNNEEN R.1996. Recent developments in the geology of the Gympie Goldfields. Geological Society of Australia Abstracts 43, 163-166. DAVIES E. R. 1970. Final report for Authority to Prospect 398M near Ban Ban Springs. Report to Queensland Department of Minerals and Energy CR 3619 (unpubl.). DAVIES E. R. 1975. Ban Ban Zinc Lode. In: Knight C. L. ed. Economic Geology of Australia and Papua New Guinea, 1. Metals, pp. 797-798. Australasian Institute of Mining Metallurgy, Melbourne. CRIBB

D A Y R . W . , CRANFIELD L . C . & SCHWARZBOCK H .

1974.

Stratigraphy of Mesozoic basins in southeastern Queensland and northeastern New South Wales. In: Denmead A. K., Tweedale G. W. & Wilson A. F. eds.

GYMPIE PROVINCE, QLD

The Tasman Geosyncline—a Symposium, pp. 319-362. Geological Society of Australia, Queensland Division, Brisbane.

DAY R. W . , MURRAY C. G. & WHITAKER W . G. 1978. T h e

eastern part of the Tasman Orogenic Zone. Tectonophysics 48, 327-364.

DEAR J. F. 1962. Report on plant fossils from near Carter's

Ridge School. Geological Survey of Queensland Report (unpubl.). Volume 3. Queensland — Central and Southern Districts. Hesperian Press, Perth. DENMEAD A. K. 1933. Application for assistance. Geological Survey of Queensland Commodity File, Gympie—Gold 3, 593. DENMEAD A. K. 1960. Permian, Gympie. In: Hill D. & Denmead A. K. eds. The Geology of Queensland, pp. 225228. Journal of the Geological Society of Australia 7. DEVEX LIMITED 1987. Annual Report. DOHERTY G. 1986. Geology of the Glenbar area. 3rd year student project thesis, Queensland Institute of Technology, Brisbane (unpubl.). D E HAVELLAND D. 1987. Gold and Ghosts.

DONCHAK P. J. T., LITTLE T. A., SLIWA R. & HOLCOMBE R. J.

1995. Geology of metamorphic units of the North D'Aguilar Block — Goomeri, Nanango and Nambour 1:100 000 Sheet areas. Geological Survey of Queensland Record 1995 /7. DUNSTAN B. 1907: Some mines in the Burnett district. Geological Survey of Queensland Publication 207. DUNSTAN B. 1910: Topographical map of Gympie and environs. Geological Survey of Queensland Publication 221a. DUNSTAN B. 1911: Geological map of Gympie and environs. Geological Survey of Queensland Publication 221b. DUNSTAN B. 1913: Queensland mineral index and guide. Gympie. Geological Survey of Queensland Commodity File, Gympie—Gold 2, 459. DUNSTAN B. 1916a: Mount Melanie gold mine, Glenbar, west ofTiaro. Queensland Government Mining Journal 17, 124. DUNSTAN B. 1916b: Ban Ban Lode, near Biggenden. Queensland Government Mining Journal 17, 125. ELLIS P. L. 1968. Geology of the Maryborough 1:250 000 Sheet area. Geological Survey of Queensland Report 26. FIELDING C. R., HOLCOMBE R. J. & STEPHENS C. J. 1994. A

critical evaluation of the Grantleigh Trough, east-central Queensland. In: Field Conference '94 Capricorn Region, pp. 17-30. Geological Society of Australia, Queensland Division, Brisbane. FISHER N. H. 1945. The fineness of gold with special reference to the Morobe Goldfield, New Guinea. Economic Geology 40, 4 4 9 - 4 9 5 , 5 3 7 - 5 6 3 . FLEMING P. J. G. 1976. Records of Mesozoic fossils in the Maryborough 1:250 000 Sheet area. Geological Survey of Queensland Report (unpubl.). FORSTER R. M. 1985. Geology of the Mount Biggenden Mine area southeast Queensland. 3rd year student project thesis, Queensland Institute of Technology, Brisbane (unpubl.). GOLDING S. D., WILSON A. F., SCOTT M . ETAL. 1987. Isotopic evidence for the diverse origins of gold mineralisation in Queensland. University of Queensland Department of Geology Papers 12 (1), 6 5 - 8 3 . GREEN P. M. 1975. The geology of the Ban Ban Zinc Lode. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.).

145

The Tasman Geosyncline in Australia. In: Denmead A. K. Tweedale G. W. & Wilson A. F. eds. The Tasman Geosyncline—a Symposium, pp. 383. Geological Society of Australia, Queensland Division, Brisbane. HARRINGTON H. J. 1983. Correlation of the Permian and Triassic Gympie Terrane of Queensland with the Brooks Street and Maitai terranes of New Zealand. In: Permian Geology of Queensland, pp. 431-^36. Geological Society of Australia, Queensland Division, Brisbane. HARRINGTON H. J. 1987. Geological units common to eastern Australia and New Zealand. In: Proceedings Pacific Rim Congress 87, pp. 801-804. Australasian Institute of Mining and Metallurgy, Melbourne. HARRINGTON H. J. & KORSCH R . J. 1985. Deformation associated with the Gympie terrane in eastern Australia. In: Third Pacific Terrane conference Extended Abstracts, pp. 104—108. Geological Society of Australia Abstracts 14. HILL J. 1985. The geology of the Copper Creek — Mount Suthers area, Marodian southeast Queensland. Project thesis, Queensland Institute of Technology, Brisbane, (unpubl.). HARRINGTON H. J. 1974.

HOLCOMBE R. J., LITTLE T. A., SLIWA R. & FIELDING C. R.

1993. Tectonic elements and evolution of SE Queensland during the Late Palaeozoic to Early Mesozoic. In: Flood P. G. & Aitchison J. A. eds. New England Orogen, eastern Australia. NEO '93 Conference, pp. 517-528. Department of Geology and Mineralogy, University of New England, Armidale. HORTON D. J. 1978. Porphyry-type copper-molybdenum mineralisation belts in eastern Queensland, Australia. Economic Geology 73, 904—921 (Queensland Government Mining Journal 79, 474-^89). HORTON D. J. 1982. Porphyry type copper and molybdenum mineralisation in eastern Queensland. Geological Survey of Queensland Publication 378. HUBER R. D. 1979. Bonnie Jan Mine Cordalba, induced polarisation and resistivity survey. Geological Survey of Queensland Record 1979/9. HUDSON S. B. 1966. Dry concentration of magnetite from Biggenden, Queensland. CSIRO Investigation Report 662. ISHIGA H. 1990. Radiolarians from the Gympie Province, eastern Australia. In: Proceedings Pacific Rim Congress 90, pp. 187-190. Australasian Institute of Mining and Metallurgy, Melbourne. JACK R. L. 1886. Gympie Gold Field. Annual Progress Report for the Geological Survey of Queensland 1885, 58— 59. JACK R. L. 1888. The mineral wealth of Queensland. Geological Survey of Queensland Publication 48. JACK R. L. 1905. Scottish Gympie gold mines. Queensland Government Mining Journal 6, 2 1 - 2 4 . KITCH R. B. 1989. B H P Gold Mines/Devex Gympie Joint Venture project and the geology of the goldfield. In: North Queensland Gold '89 Conference, pp. 8 7 - 9 3 . Australasian Institute of Mining and Metallurgy, Melbourne. KITCH R. B. & MURPHY R. W. 1990. Gympie Gold Field. In: Hughes F. E. ed. Geology of the Mineral Deposits of Australia and Papua New Guinea, p. 1515. Australasian Institute of Mining and Metallurgy, Melbourne. KORSCH R. J. 1996. Mesozoic deformational events in eastern Australia and their impact on onshore sedimentary basins. Geological Society of Australia Abstracts 43, 308—318. KROSCH N. J. 1978. Inspection of Bonnie Jan gold copper mine, Cordalba ML 146. Memorandum to Chief

146

L. C. CRANFIELD

ETAL.

Government Geologist. Geological Survey of Queensland File 4-4-55 (unpubl.). KROSCH N. J. 1979. Inspection of Bonnie Jan Gold-Copper Mine Cordalba ML 146 (Bundaberg). Geological Survey of Queensland File 4-4-55 (unpubl.). KROSCH N. J. 1980. Department drilling program. Bonnie Jan Gold and copper mine ML 146. Department of Mines and Energy Queensland Report (unpubl.). LEES W. 1899. Gympie Goldfield. In: The Goldfields of Queensland. Outridge Printing Company Ltd, Brisbane. LITTLE T. A. 1993. Geology of a northern part of the North D'Aguilar Block, southeast Queensland. Department of Minerals and Energy Record 1993/3. LOUTHEAN R. (Editor) 1992. Register of Australian Mining 1992/93. Resource Information Unit Ltd, Perth. MCLEOD R. L. 1979. Mineral exploration in the Maryborough 1:250 000 Sheet area southeastern Queensland. Queensland Government Mining Journal 80, 572-581. MELSOM R. & ROWLANDS N. 1988. G y m p i e Goldfield,

Gympie, Queensland. In: Bloom M. S. & Parrington P. J. Bicentennial Gold 88, Core Shed Guide Book, pp. 16-17. Monash University, Melbourne. MORRISON G . W . , ROSE W . J. & JAIRETH S. 1991. G e o l o g i c a l

and geochemical controls on the silver content (fineness) of gold in gold-silver deposits. Ore Geology Reviews 6, 333-364. MORTON C. C. 1920a. The XYZ Mine near Glenbar. Geological Survey of Queensland File 4-4-45 (unpubl.). MORTON C. C. 1920b. Stanton Harcourt Goldfield. Memorandum to Chief Government Geologist. Geological Survey of Queensland File 4-4-45 (unpubl.). MORTON C. C. 1926. Copper prospecting at Mount Allen. Geological Survey of Queensland File 4-3-7 (unpubl.). MORTON C. C. 1930a. Gold resources of Queensland. In: The Gold Resources of the World, an Inquiry made upon the Initiation of the Organising Committee of the 15th International Geological Congress, South Africa, 1929, pp. 279-293. Bureau of the Congress, Pretoria. MORTON C. C. 1930b. Re Cornfield Creek, Stanton Harcourt Goldfield. Memorandum to Chief Government Geologist. Geological Survey of Queensland File 4-4-184 (unpubl.). MORTON C. C. 1933. Glenbar Goldfield. Geological Survey of Queensland File 4-4-78 (unpubl.). MORTON C. C. 1934. Re Stanton Harcourt Goldfield. Memorandum to Chief Government Geologist. Geological Survey of Queensland File 4-4-184 (unpubl.). MORTON C. C. 1935. Mount Ideal Cordalba. Government Mining Journal 36, 41.

Queensland

MURPHY P. R . , SCHWARZBOCK H . , CRANFIELD L . C . , WITHNALLI. W . & MURRAY C . G . 1976. G e o l o g y o f t h e

Gympie 1:250 000 Sheet area. Geological Survey of Queensland Report 96. MURRAY C. G. 1986. Metallogeny and tectonic development of the Tasman Fold Belt System in Queensland. Ore Geology Reviews 1,315-^00. MURRAY C. G., 1987a. Review of the geology of the Gympie district, southeast Queensland, with particular reference to stratigraphic, structural and tectonic problems. In: Murray C. G. & Waterhouse J. B. eds. 1987 Field Conference Gympie District, pp. 1-19. Geological Society of Australia, Queensland Division, Brisbane. MURRAY C. G. 1987b. Tectonic evolution and metallogenesis of the New England Fold Belt, Eastern Australia. In: Proceedings Pacific Rim Congress 87, pp. 353-358.

Australasian Institute of Mining and Metallurgy, Melbourne. MURRAY C. G. 1988. Tectonostratigraphic terranes of southeast Queensland, In: Hamilton L. ed. Field Excursions Handbook for the Ninth Australian Convention, pp. 19-51. Geological Society of Australia, Queensland Division, Brisbane. MURRAY C. G. 1990. Comparison of Suspect Terranes of the Gympie Province with other units of the New England Fold Belt, eastern Australia. In: Pacific Rim Congress 90, Proceedings Volume //, pp. 247-255. Australasian Institute of Mining and Metallurgy, Melbourne. MURRAY C . G . , HOLCOMBE R . & ISHIGA H . 1 9 8 9 . S u s p e c t

terranes of the Gympie Province, New England Fold Belt. Geological Society of Australia Abstracts 24, 99-100. PALMIERI V. 1987. Report on micropalaeontological analysis of limestone samples from the Permian of the Gympie Basin. Geological Survey of Queensland file report (unpubl.). PALMIERI V. 1988. Re age of limestone samples in the Degilbo and Red Hill area near Biggenden. Geological Survey of Queensland file report (unpubl.). QUEENSLAND DEPARTMENT O F M I N E S

1953.

Gympie

Goldfield. In: Edwards A. B. ed. Geology of Australian Ore Deposits, pp. 813—816. Australasian Institute of Mining and Metallurgy, Melbourne. RANDS W. H. 1886. Stanton Harcourt Gold Field. Geological Survey of Queensland File 4-4-184 (unpubl.). RANDS W. H. 1889. On the Gympie Goldfield. Geological Survey of Queensland Publication 52. RANDS W. H. 1890a. On the Tiaro District Coal Measures, Neardie Antimony Mine and Teebar and Calgoa Copper Lodes. Geological Survey of Queensland Publication 59. RANDS W. H. 1890b. On Mount Biggenden Bismuth Mine, Gebangle and the Mount Shamrock Mine. Geological Survey of Queensland Publication 60. RANDS W. H. 1891. Gympie Gold Field (second report). Geological Survey of Queensland Publication 76. RANDS W. H. 1894. Gympie Gold Field (third report). Geological Survey of Queensland Publication 100. RANDS W. H. 1899. Geological map of the Gympie Goldfield. Geological Survey of Queensland Publication 139. RANDS W. H. 1901. Fourth report on the Gympie Gold Field, having special reference to the Inglewood Dyke and the eastern leases. Geological Survey of Queensland Publication 166 (also in Queensland Government Mining Journal

2, 3 5 6 - 3 5 8 , 5 2 5 - 5 2 9 , 5 9 4 - 5 9 8 .

RIDLEY W. H. 1962. Geological studies in the Maryborough Basin, southeast Queensland. Proceedings of the Royal Society of Queensland 72, 83-89. ROACH A. 1990. Geology of the South Gympie goldfield. BSc (Hons) thesis, Queensland University of Technology, Brisbane (unpubl.). ROLLASON R. G. 1970. The geology of the Marodian area, near Gympie, south-east Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). RUNNEGAR B. N . & FERGUSON J. A . 1969. S t r a t i g r a p h y o f t h e

Permian and Lower Triassic marine sediments of the Gympie district, Queensland. University of Queensland Department of Geology Papers 7 (9). SAINT-SMITH E. C. 1917. Re Mt Allen. Geological Survey of Queensland File 4-3-6 (unpubl.). SCOTT M. 1989. Exploration summary—Gympie/Goomeri/ Nambour/Nanango 1:100 000 Sheet areas (1976-1988).

GYMPIE PROVINCE, QLD Queensland Government Mining Journal 90, 521-532. SCOTT M . , CRANFIELD L. C . , ROBERTSON A . D. C . & MURRAY

C. G. 1990. Base and precious metal mineralisation in the Gympie composite terrane. In: Proceedings of Pacific Rim Congress 90, Volume III, pp. 217-227. Australasian Institute of Mining and Metallurgy, Melbourne. SIEMON J. E. 1971. Diamond drilling program Mount Biggenden magnetite and bismuth mine. Geological Survey of Queensland Report 67. SIEMON J. E. 1974. Neardie antimony deposits, Gympie. Geological Survey of Queensland Report 82. SIEMON J. E . , GREEN P. M . & HORTON D . J. 1 9 7 7 . S o m e

Mines and mineral deposits of the Gayndah Biggenden area. In: Day R. W. ed. 1977 Field Conference Lady Elliot Island—Fraser Island—Gayndah-Biggenden, pp. 79-86. Geological Society of Australia, Queensland Division, Brisbane. SIVELL W. J. 1990. Late Palaeozoic metavolcanic suites from Gympie area: implications for terranes. In: Proceedings of Pacific Rim Congress 90, Volume //, pp. 277-282. Australasian Institute of Mining and Metallurgy, Melbourne. SIVELL W . J. & MCCULLOUGH M . T . 1 9 9 3 . N d - i s o t o p e o f a

Permian Island arc — back-arc basin system in Gympie Province. In: Flood P. G. & Aitchison J. A. New England Orogen, eastern Australia. NEO '93 Conference, pp. 581— 586. Department of Geology and Geophysics, University of New England, Armidale. SIVELL W . J. & MCCULLOUGH M . T .

1996.

Nd-isotopic

evolution of a nascent Permian back-arc basin — island arc system in Gympie Province. Geological Society of Australia Abstracts 41, 403. SIVELL W . J., STOCKSIEK C . M . & WATERHOUSE J. B. 1 9 9 0 .

Metavolcanic suites from suspect Late Palaeozoic terranes of the Gympie Province, southeast Queensland. Geological Society of Australia Abstracts 25, 197. SIVELL W. J. & WATERHOUSE J. B. 1987. Volcanic rocks of

the Highbury Volcanics and Rammutt Formation— remnant portion of an Early Permian intra-oceanic arc. In: Murray C. G. & Waterhouse J. B. eds. 1987 Field Conference Gympie District. Geological Society of Australia, Queensland Division, Brisbane, pp. 34-43. SIVELL W . J. & WATERHOUSE J. B . 1 9 8 8 . P e d o g e n e s i s o f

Gympie Group volcanics: evidence for remnants of an Early Permian volcanic arc in eastern Australia. Lithos 21, 81-95.

SLIWA R. 1994: Regional structural geology of the central North D'Aguilar Block, southeastern Queensland. PhD thesis, University of Queensland, Brisbane (unpubl.). SLIWA R . , HOLCOMBE R . J., FIELDING C . R . , LITTLE T . A . ,

BRYAN S. & FIFOOT A. 1993. Early marine fault basins formed during exhumation of the New England Orogen subduction complex in southeastern Queensland. In: Flood P. G. & Aitchison J. A. eds. New England Orogen, eastern

147

Australia. NEO '93 Conference, pp. 557-564. Department of Geology and Mineralogy, University of New England, Armidale. STANTON R. L. 1987. Constitutional features, and some exploration implications of three zinc-bearing stratiform skarns of eastern Australia. Transactions of the Institution of Mining and Metallurgy (Section B: Applied Earth Science) 96, B37-B57. STAPELTON J. 1961. Memo to Chief Government Geologist 7 / 2 / 6 1 . Geological Survey of Queensland file 4-4-77 (unpubl.). STILL WELL F. L. 1951. Auriferous ore from Mount Ideal Mine, Cordalba, Queensland. CS1RO Mineragraphic Investigation 480. WALLIS D. S. 1990. Antimony in Queensland. In: Proceeding of the Pacific Rim Congress 90, Volume II, pp. 240-248. Australasian Institute of Mining and Metallurgy, Melbourne. WATERHOUSE J. B. & BALFE P. 1987. Stratigraphic and faunal

subdivisions of the Permian rocks at Gympie In: Murray C. G. & Waterhouse J. B. eds. 1987 Field Conference Gympie District, pp. 20-33. Geological Society of Australia, Queensland Division, Brisbane. WATERHOUSE J. B. & SIVELL W. J. 1987a. Trans-Tasman

relationships between the Permian—Triassic rocks of Gympie, southeast Queensland, and those of New Caledonia and New Zealand. In: Murray C. G. & Waterhouse J. B. eds. 1987 Field Conference Gympie District, pp. 48-59. Geological Society of Australia, Queensland Division, Brisbane (also Tectonophysics 142, 227-240. WATERHOUSE J. B. & SIVELL W. J. 1987b. Permian evidence for trans-Tasman relationships between east Australia, New Caledonia and New Zealand. Tectonophysics 142, 227-240. WOODCOCK J. T . & DUNKIN H . H . 1 9 5 2 . C o n c e n t r a t i o n s o f

auriferous copper ore from the new Ideal Mine Cordalba, Queensland. CSIRO Ore dressing Investigation 429. WILSON A. F. 1987. Some highlights of the gold deposits of Queensland—past and present, and some future prospects. In: Herbert H. K. ed. Proceedings of the One-day Symposium—Gold in Queensland, pp. 1—16. University of Queensland, Department of Geology Papers 12 (1). WILSON I. H. 1968. Geology of the southern Goomborian pluton. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). WILCOX B. 1917. Maryborough district. Copper. Annual Report of the Department of Mines for 1917, 111. ZIEBELL R. O. 1972. The geology of the Mount Suthers area, south-east of Biggenden, south-east Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). (Received 15 May 1996; accepted 29 July 1997)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 148-160.

Geochemistry and Sm-Nd isotope systematics of Early Permian basalts from Gympie Province and fault basins in southeast Queensland: implications for mantle sources in a backarc setting at the Gondwana rim W. J. SIVELL 12 AND M. T. McCULLOCH2 1 2

Department of Geology and Geophysics, University of New England, Armidale NSW 2351, Australia. Research School of Earth Sciences, Australian National University, Canberra ACT 0200, Australia.

By the Early Permian, numerous extensional basins containing marine mass-flow deposits and mafic to silicic volcanics had developed in the New England Orogen. Distinct Sm-Nd isotopic trends recognised for Early Permian basalts along an - 4 0 km transect through the Gympie Province and Early Permian fault basins in southeast Queensland shed light on the origin and relation to regional crustal evolution and convergent margin tectonics of one of these extensional basins at the palaeo-Pacific Gondwana rim. Nd-isotope Trend 1 is shown by Cambroon basalts that possess high initial 143Nd/ 144Nd ratios which decrease only slightly (£ N( j (270 Ma) = +10.3 to +9) with decreasing l47 Sm/ 144Nd. These basalts were derived predominantly from a mantle source with s N d values (+10 to +12) similar to those expected for Permian MORB-type mantle. By contrast, the Highbury and Cedarton Volcanics and some Cambroon basalts show marked decreases of s N d (+10 to +5) in sympathy with 147 Sm/ 144Nd (Trends 2A and 2B). Samples with the lowest eN(j are the Highbury Volcanics from the Gympie Group immediately to the east of the Cedarton-Cambroon sequence. The Highbury basalts are island-arc tholeiites. Their relatively low initial 143 Nd/ 144 Nd ratios, together with high La/Nb, Zr/Nb, Ba/Zr and Sr/Nd typical of arc basalts, are due to the addition of components from subducted oceanic crust to the mantle wedge overlying a west-dipping Early Permian Benioff Zone seaward of the palaeo-Pacific Gondwana rim. However Ti/Y, Ti / V, Zr/Y and Nb/Zr ratios, which are relatively unaffected by slab-derived contributions, are lower for Highbury Volcanics than for Cambroon and Cedarton basalts. Prior to subduction-related metasomatism of the mantle wedge, the mantle source for the Highbury island-arc tholeiites was even more depleted than the source for the Cedarton and Cambroon lavas which probably formed in a backarc basin behind the Gympie (Highbury) island arc. Mixing between isotopically enriched (Highbury-type) subduction-modified mantle and MORB-type mantle in the evolving backarc basin gave rise to isotopic Trend 2. The small isotopic shift that accompanies decreasing l47 Sm/ 1 4 4 Nd, A l 2 0 3 / T i 0 2 and MgO/Zr, and increasing (Ce/Yb) N in Trend 1 indicates the presence of an additional isotopically enriched component (similar to enriched (E-) MORB or ocean island basalt source mantle) in the mantle wedge. This component was preferentially sampled at low degrees of partial melting. Prior to its removal due to continued melting, mixing between this isotopically enriched end-member and subductionmodified (Highbury-type) mantle during nascent backarc basin volcanism gave rise to the isotopic and geochemical features of the Cedarton basalts which are transitional between those of island-arc tholeiites and enriched (E-) MORB (Trend 2B). Trend 2A Cambroon basalts reflect mixing between (diminishing) subductionrelated components and the remaining depleted (N-) MORB source mantle beneath the more mature rift. Key words: basalt, geochemistry, Gympie Province, neodymium isotopes, Permian.

INTRODUCTION New ocean crust is generated as a result of lithospheric extension, upwelling and partial melting of mantle beneath backarc basins, concomitant with volcanism in related arc systems (Saunders & Tarney 1979; Hawkins & Melchoir 1985). The nature of chemically and isotopically distinct mantle sources contributing to magmatism in backarc regimes and the extent of interactions between these various mantle end-members, including relatively enriched and depleted components, remain largely unresolved (Volpe et al. 1988; Iyeda & Yuasa 1989). Possible contributions from subducted lithosphere in these settings (especially in nascent backarc basins) must also be assessed and may vary with time as basins evolve. Where slab influences are

pronounced they may partly mask or obscure more subtle geochemical or isotopic signatures due to compositional variations inherently present in mantle domains newly tapped by backarc volcanism. In contrast to long-lived spreading centres at major mid-oceanic ridges, basaltic volcanism in backarc environments may reveal the type of material that is tapped during the earliest stages of melting of (shallow) asthenospheric mantle domains. The purpose of this paper is to present combined trace element and Sm—Nd isotopic data for basalts from an immature Permian backarc basin-island-arc complex at the palaeo-Pacific Gondwana rim in the Cedarton-Cambroon sequence and Gympie Province, southeast Queensland, and to use this data to evaluate magmatic processes and particularly the geochemical and isotopic evolution of mantle

GEOCHEMISTRY, PERMIAN BASALTS, QLD source regions in this setting during the Late Palaeozoic. The wide range of chemical and isotopic variations among discrete basalt suites in the Early Permian fault basins and the Gympie Province, their relatively restricted geographic distribution, and good correlations between their trace element and isotopic ratios, offer an ideal opportunity to constrain the nature of their source mantle. This includes the causes of relative depletions or enrichments in the Permian backarc mantle and the nature of possible interactions among the inferred mantle sources. Some implications of the observed compositional trends of these Early Permian basalts (and hence of their sources) for the tectonic evolution of the palaeo-Pacific Gondwana margin are outlined. G E O L O G I C A L SETTING

In southeast Queensland, Early Permian sediments and volcanics of the Cambroon beds, Cedarton Volcanics and Gympie Group comprise suspect (in some cases composite) terranes of probable island-arc-marginalbasin origin, emplaced at the palaeo-Pacific margin of Gondwana (Figure 1). In the Gympie Province, the shallow-marine association of the Gympie Group and overlying Kin Kin beds constitutes a discrete exotic terrane which accreted to the Gondwana rim in Triassic time (Harrington 1983). Predominant tuff-breccias within the mafic volcanic succession of the Early Permian Highbury Volcanics of the Gympie Group represent an early submarine stage of volcanism in an island-arc environment (Sivell & Waterhouse 1988). The overlying (Triassic) Kin Kin beds represent a possible forearc assemblage. The Gympie island arc may have been partly coeval with the Camboon Volcanic Arc, which developed as a major Cordilleran-style magmatic arc with associated north-south-trending forearc basin and subduction complexes (tripartite stratotectonic framework of the Yarrol Orogen) along the Gondwana margin from the Late Carboniferous to the Early Permian. Volcanics from the Camboon Volcanic Arc have given isotopic ages as young as 280 and 270 Ma (Webb & McDougall 1968). Thus, in the Early Permian, there may have existed two partly contemporaneous arcs: (i) the continental margin Camboon Volcanic Arc; and (ii) the Gympie island arc. To the west, the Amamoor beds include shale, pebbly mudstone, greywacke and impure limestone, as well as basaltic flows, banded chert and jasper (Murray 1987). These rocks are more complexly deformed than the Gympie Group, with both Sj slaty cleavage and S 2 crenulation developed. The Amamoor beds are mapped as a composite terrane comprising technically juxtaposed Carboniferous and Permian sequences. Abundant Early Permian bivalves occur in local coarse volcaniclastics (Murray et al. 1979) now mapped as the Cambroon beds. The recent discovery of Early Carboniferous radiolarians from chert lenses in both the Amamoor beds and the southern part of the Gympie Group (Murray et al. 1989) indicates that portions of these units are older and unrelated. The boundaries of

149

the Carboniferous part of the Amamoor beds have been delineated by Cranfield and Scott (1993). The rocks investigated in this study are from portions of the volcanic successions that are Early Permian in age. Type 1 basaltic volcanics of the Cambroon beds form discontinuous lenses of related flows several kilometres in extent, locally in contact with rare small intrusive bodies (dykes?) of massive, coarser grained (Type 2) ophitic basalt. Some flows grade into volcanic breccia, with intimate mixing between red mud and brecciated lava indicative of slumping on steep slopes of submarine volcanoes. This is in marked contrast to complexes of intrusives, sheeted dykes and pillow lavas formed by long-term, near steady-state volcanism at major ocean ridges. Similar sequences are inferred to have accumulated in backarc basins where there is a general absence of well-defined symmetric spreading ridges [e.g. young sea floor in the Lau Basin (Lawver et al. 1976)]. To the south of the Cambroon beds, and separated from them by only a thin cover of Tertiary basalt, are the Cedarton Volcanics that include mainly basaltic tuffs and tuff-breccias. A Permian marine fauna similar

153°E

Kilkivan

10 i—

-26°30'S

km

20 _i—

30

Gympie Block m Beds

edarton Volcanic

E a r l y T r i a s s i c [\\\\'\ X J Kin Kin B e d s Granitoids 1+ + + | V o l c a n i c s

Queensland

and

Gympie Group, I Cambroon Beds and Cedarton Volcanics Amamoor Beds Serpentinite

fit

_ . . . I W 1 I.Connors-Auburn Volcanic Arc |7TH 2.Yarrol Forearc Basin c^rq 3.Wandilla Slope and Basin (Subduction Complex)

Rocksberg

I Gc or er er ne sl at ot ni vee s a n d

Figure 1 Simplified geological map showing pre-Middle Triassic terranes of Gympie Province, southeast Queensland (modified after Murray 1988). Inset shows relation of Gympie rocks to Late Palaeozoic (Gondwana) continental margin stratotectonic units (Yarrol Orogen).

150 W. J. SIVELL AND M.T. McCULLOCH to that in the Cambroon beds has been recognised in this study include the least recrystallised basalts. Sivell and Waterhouse (1988) indicated that intro(Murray et al 1979; Waterhouse & Balfe 1987). duction of Na 0 accompanied the albitisation of plagioclase in the Highbury basalts, but that abundances PETROGRAPHY OF THE BASALT SUITES of ferromagnesian elements were little affected by lowThe porphyritic basalts which comprise the tuff-breccias grade metamorphism with these basalts maintaining and lavas that dominate the Highbury Volcanics of the their subalkaline character as demonstrated for spilitic Gympie Group contain prominent phenocrysts of suites elsewhere (Cann 1969). For the analysed samples, clinopyroxene (up to 8 mm in diameter), together with good correlations among a number of elements (e.g. Ti, subordinate olivine and plagioclase, in a groundmass of V, Zr, Nb, Y, Ba, Al and Mg) as well as high corpyroxene, plagioclase, magnetite and altered brown and relations between inter-element ratios, show that green glass. Total phenocryst content is typically in the metamorphic transformations did not appreciably range 20-30 modal percent. Basalts from the Cedarton affect the abundances of these elements. Particularly Volcanics are more highly porphyritic than the significant are correlations of trace element ratios with Highbury rocks (total phenocryst content may exceed isotopic ratios, with individual basalt suites displaying 40 modal percent), and plagioclase (with subordinate systematic geochemical and isotopic trends expected for clinopyroxene and olivine) dominates the phenocryst igneous rocks and which reflect crystal fractionation assemblages. Clinopyroxene crystals are generally 2— trends and tectonic setting. Relatively constant trace 4 mm in diameter. Relict pyroclastic textures persist in element ratios for samples from individual suites accord the form of altered glassy shards. The glassy or well with basaltic series related by crystal fractionation and are consistent with observed phenocryst aphanitic groundmass may be rich in Fe-Ti oxides. There are two petrographically distinct types of assemblages. In this study, significant differences in Cambroon beds basalts, which correspond to chemically inter-element ratios between different basalt suites are different groups discussed in the following section. considered to result from contrasting sources for their Amygdaloidal, variolitic Cambroon Type 1 basalts primary magmas. This is particularly the case for ratios may be aphyric or include sparse (<5 modal percent) involving elements such as Zr, Nb, Cr, Ti, Y and V phenocrysts of plagioclase (up to 2 mm in length) which are typically regarded as immobile during lowand lesser olivine (including a high proportion of grade metamorphism (Winchester & Floyd 1977). For skeletal microphenocrysts). Groundmass minerals the analysed basalts, intra-suite variation in the degree include plagioclase, clinopyroxene and Fe-Ti oxides, of REE fractionation correlates with chemical indices of with textures ranging from intersertal to intergranular. the extent of partial melting (e.g. Zr/Mg and Al/Ti). Cambroon Type 2 basalts include (i) medium- to Although light-REE (LREE)-enrichment has been coarse-grained rocks in which clinopyroxene ophitically described in association with strong alteration of basalts encloses plagioclase; and (ii) porphyritic basalts with (Frey et al 1974) such processes have not affected these ~15 modal percent phenocrysts of predominant rocks which show little or no relative fractionation plagioclase (up to 8 mm in length) and olivine among heavy-REE and lack significant Eu anomalies (<3 mm). The phenocrysts occur in a hypocrystalline which can result from hydrothermal alteration in oceangroundmass of plagioclase and columnar, skeletal floor basalts. Basalts from the Highbury Volcanics are island-arc clinopyroxene. The majority of these basalts are partly recrystallised. Plagioclase is albitised, olivine is tholeiites with elevated FeO /MgO ratios and A1 0 pseudomorphed by chlorite, and altered glass contains contents that increase to relatively high levels (>18 chlorite, saussurite and Fe-Ti oxides. Clinopyroxene wt%) with decrease in MgO, as often observed in arc is the only unaltered primary mineral. Abundant magmas (Crawford et al 1987; Perfit et al 1980). The amygdales are infilled by pumpellyite, epidote, chlorite, Highbury basalts have high Ba and low high-field strength element (HFSE: Ti, Zr and Nb) contents. albite, quartz and calcite. High La/Nb, Sr/Nd, Ba/Ti and Ba/Nb ratios, together with low Ti/V, Ti/Y, Zr/Y and Nb/Zr (Figures 2, 3) GEOCHEMISTRY reflect metasomatism of depleted mantle by large-ion Major and trace-element contents of representative lithophile element (LILE)-enriched subduction-related samples from basalt suites are presented in Table 1 and fluids derived from subducting oceanic crust (Perfit et rare-earth element (REE) abundances for some samples al 1980; Saunders & Tarney 1979). Slightly LREEare listed in Table 2. Studies of the mobilities of enriched REE distributions [(Sm/Nd) = 0.79-0.83] elements suggest that abundances of some major and (Figure 7) for these rocks [which have Mg up to 70 and trace elements (especially alkalis) are susceptible to high Al 0 /Ti0 ratios (16-19)], together with the lack alteration during low-grade metamorphism (Cann 1969; of major Eu anomalies, most likely reflect the extent of Grapes 1976). Basalts comprising the Highbury REE fractionation in their mantle source region (Sun et Volcanics, Cedarton Volcanics and Cambroon beds al 1979) and moderate to high degrees of partial have been metamorphosed chiefly to prehnite- melting, at least for the most primitive basalts. pumpellyite facies assemblages. Any assessment of The Cambroon basalts include low A1 0 (Type 1) magmatic processes that took place prior to meta- and high A1 0 (Type 2) chemical groups, which morphism of the basalt suites is dependent on the degree correspond to petrographic groups discussed above. to which metamorphism was isochemical. Samples used Type 1 samples vary from LREE-depleted to slightly 2

t

2

CH

#

2

3

2

2

2

3

3

3

Table 1 Major and trace element analyses of representative basalts from the Gympie Province.

Analysis No.* Sample No.

1 CV22

2 CV30

3 CVIXB

4 CV37A

5 AM 16

6 AM 14

7 AM 10

8 AM16B

9 AM9B

10 AM42

11 HV4

12 HV9

13 HV39

Si0 2 Ti0 2 A1 2 0 3 Fe 2 0 3 MnO MgO CaO Na 2 0 K20 P205 LOI I 100 Mg / (Mg + Fe2+)

47.60 0.99 19.05 10.48 0.16 4.65 10.30 2.63 0.82 0.13 3.21 100.02 49.7

46.51 0.92 17.59 11.30 0.19 6.74 10.79 1.38 0.11 0.13 4.00 99.66 57.1

48.63 1.08 18.57 12.78 0.11 3.46 7.85 2.93 1.40 0.23 2.60 99.64 37.9

47.45 0.75 16.08 10.26 0.16 8.90 9.95 1.92 0.28 0.11 3.91 99.77 66.2

50.82 1.31 14.64 9.77 0.16 6.13 10.06 3.59 0.40 0.12 2.82 99.83 58.5

46.80 0.99 16.31 8.28 0.09 7.39 9.79 3.80 0.20 0.10 6.12 99.86 66.7

46.61 1.69 14.72 10.12 0.15 7.58 9.16 2.10 2.30 0.26 5.40 100.09 62.8

52.33 1.13 13.46 11.08 0.16 4.76 9.53 3.24 0.17 0.11 3.40 99.37 49.2

46.93 0.84 18.67 8.26 0.15 6.73 10.17 2.98 0.58 0.07 4.41 99.80 64.7

47.23 0.67 16.50 7.71 0.12 8.04 9.97 1.74 2.15 0.06 5.62 99.80 58.4

53.26 0.98 17.07 10.34 0.20 4.67 6.70 5.83 0.75 0.32 2.81 99.30 50.5

53.50 0.82 15.48 10.70 0.16 5.56 8.32 5.46 0.66 0.36 2.13 100.16 53.8

54.12 0.74 13.81 8.83 0.20 9.31 7.98 4.08 1.05 0.15 2.90 99.57 70.2

Trace elements Cr Ni V Rb Sr Ba Zr Nb Y Th

197 62 192 13.2 261 320 58 9.1 21 2.4

286 107 304 7.1 340 156 66 9.7 22 3.0

48 32 397 29.8 268 287 59.8 8.3 40 4.2

554 181 273 8.4 349 164 52 7.5 18 2.0

313 72 291 7.3 69 94 81 8.7 33 1.8

581 216 185 2.8 80 13 65 8.7 20 2.3

195 56 289 23.7 182 102 121 24.0 26 2.8

281 102 253 6.0 192 63 94 7.9 31 3.3

244 70 175 6.4 177 147 42 6.3 19 1.3

394 117 173 41.0 51 117 32 8.2 18 1.7

118 57 288 20.2 271 372 71 2.4 31

188 83 352 15.2 126 245 64 1.7 34

527 164 203 37.5 107 302 60 3.0 18

—

—

—

O m

o

o ffi M 5 M GO H *

* hd ffl

6

* 1-4, Cedarton basalts; 5—8, Cambroon Type 1 basalts; 9, 10, Cambroon Type 2 basalts; 11-13, Highbury basalts. Major oxides for 1-10 were analysed by X-ray fluorescence in the Department of Geology, University of Sydney. All other XRF determinations using the Siemens SRS-spectrometer at the University of Adelaide, following the method of Norrish and Hutton (1969). International standards and Abbey's (1980) recommended values were used for calibration during trace element determinations.

> w

>

GO >

r H (Z>

O r

o

152

W. J. SIVELL AND M.T. McCULLOCH

Table 2 Rare earth element abundances for Gympie Province basalts. Analysis No.* Sample No.

1 AM9B

2 AM10

3 AM16

4 CV22

5 HV4

6 HV39

La Ce Nd Sm Eu Gd Tb Dy Ho Er Yb Lu (Ce/Yb)N

0.75 2.43 4.00 1.60 0.67

8.20 18.50 11.50 3.41 1.15

3.13 9.06 8.07 3.00 1.15

6.64 15.30 10.10 2.80 0.94

-

-

-

-

9.44 20.49 15.52 4.10 1.29 4.72

5.01 13.33 10.33 2.96 0.91 3.31

—

-

-

-

0.63

0.85

—

—

-

-

-

-

-

-

2.92 2.60

-

1.90 0.26 2.06

2.77 0.40 0.84

-

4.62

0.74

1.06

2.05 0.31 2.31

1.63 0.27 0.38

0.54

0.74

0.62

0.45

2.02

3.48

2.13 2.04 1.67

* 1, Cambroon Type 1 basalt; 2, 3, Cambroon Type 2 basalts; 4, Cedarton basalt; 5, 6, Highbury basalts. REE abundances determined by: 1-4, neutron activation analysis by Becquerel Laboratories, Lucas Heights Nuclear Facility; 5-6, isotope dilution mass spectrometry using the method described in Sun and Nesbitt (1978).

LREE-enriched [(Sm/Nd) = 1.09-0.82] (Figure 7) and have low Ba/Nb and Ba/Ti ratios, as well as high Ti/V and Nd/Sr (-0.08) (Figures 2, 3, 7). These ratios are similar to those of mid-ocean ridge basalts (MORB) (both E-type and N-type: Sun & McDonough 1989) and some backarc basin basalts, especially those from the Mariana Trough interpreted as due to mixing between MORB-like and arc-like melts (Volpe et al 1987). Al 0 /Ti0 ratios show a wide range from 8-16 (Figure 6). Type 2 Cambroon samples are LREEdepleted [(Sm/Nd) = 1.2-1.3] with significantly higher Ba/Sm and Ba/Ti ratios and slightly higher Ba/Nb ratios than Type 1 basalts. These ratios are intermediate between those of MORB and island-arc tholeiites (Sun 1980; Sun & McDonough 1989; Chow et al 1980). Type 2 basalts plot (along with the Highbury and Cedarton Volcanics) in the island-arc tholeiite fields on Ti/Cr vs Ni and Cr vs Y (Pearce 1980) plots (Figure 4) in contrast to Type 1 basalts which plot as MORB. They also have slightly lower Ti/V than Type 1 rocks and are transitional to arc basalts. Al 0 /Ti0 ratios are high (-25). The Cedarton basalts have the highest A1 0 contents (16-18.5 wt%), with Al, Fe, Ti and V increasing as MgO decreases, as for the Highbury basalts. The Cedarton basalts also have low Ti0 , Zr, Ti/V, Ti/Y and Zr/Y, and relatively high abundances of LILE (K, Rb and Ba). Ba/Ti and Ba/Nb range from low values characteristic of E-MORB or ocean island basalts to high values typical of arc basalts (Figure 2). These ratios are intermediate between the systematically higher and lower values of the Highbury island-arc CH

2

3

2

CH

2

3

2

2

2

3

tholeiites and MORB-like Cambroon basalt suites respectively. Nd/Sr ratios (0.04-0.06) are transitional between those of E-MORB (or ocean island basalts) and island-arc tholeiites (Sun & McDonough 1989) and are distinctly lower than in Cambroon rocks. Arc-like features of the Cedarton Volcanics are also evident on various trace-element discriminant plots (Figure 4) where these rocks lie in fields of island-arc tholeiites, distinct from MORB-like Cambroon Type 1 basalts. All Cedarton Volcanics show slight LREE-enrichment [(Sm/Nd) = 0.81-0.87] (Figure 7) similar to the Highbury basalts. Al 0 /Ti0 ratios are also like those of the Highbury basalts. CH

2

3

2

Sm-Nd ISOTOPES Representative Sm-Nd isotopic compositions of the basalts are shown in Table 3. A large variation in initial Nd/ Nd ratios is observed (Figure 5) (s^d (270 Ma) = 10.3 to 4.1). Overall high e values clearly indicate a predominantly depleted (MORB-type) mantle source. However, large variation in initial Nd/ Nd ratios and sympathetic variation in ratios of incompatible elements of the basalts precludes any proposed origin in which their chemistry is controlled only by differences in partial melting of a homogeneous mantle source. This isotopic diversity requires manifold heterogeneity in the Early Permian mantle beneath Gympie Province. High e^d values of +4 to +10 for these rocks (Table 3) indicates their derivation from mantle with long-term depletion of Nd relative to Sm, resulting in higher Nd isotopic ratios relative to bulk 143

144

N d

143

l44

GEOCHEMISTRY, PERMIAN BASALTS, QLD

153

(a) 10

10

N-MORB

N-MORB MTB

t' MTB •l o

\ d r

C\J

E-MORB

LBB-1N U

o v. A

^ \A

OJ

°ib n

A^ -I ,

<M

V

5 6

\

MIAB

0.08

0.06

0.04 Ba/Ti

0.02 (b) 10

.

V

8

20

10

Cambroon Type 1 Cambroon Type 2 Highbury IAT Cedarton 40

30

Ti/V

/ E-MORB

*r^ i

CO

A

oibD

(c) AA 10 jfc^N-MORB MTB

1000

MIAB.

40

120

80 Ba/Sm • • • a

• 500 ; E Q_ Q. o

Cambroon Type 1 Cambroon Type 2 Highbury I AT Cedarton

100

L E-MORB

50 Y (ppm.)

o c\j 400 CO e

MIAB j~~ •

• A • a

Figure 3 e Nd vs Ti/V for Early Permian basalts. Symbols as in Figure 2. LBB-1N and LBB-1E indicate ranges for Lau Basin basalts with N- and E-MORB affinities (Volpe et al. 1988).

N-MORB

• U® MTB \ ° CM

OIB LBB-1E A)

MIAB

CO

•4!

n

/

I ^

(b)

AI

OIB

IAT 40

80 Ba/Nb

160

120

Figure 2 e N d vs Ba/Ti, Ba/Sm and Ba/Nb for Gympie Province basalts. Fields are indicated for average LREEdepleted and enriched mid-ocean ridge basalts (N-and EMORB respectively); ocean island basalt (OIB); Mariana Trough basalts (MTB) and Mariana Island Arc basalt (MIAB).

earth values. However low Sm/Nd elemental ratios present in some of the samples indicates that a relatively recent episode of LREE-enrichment also affected the compositions of some basalts. The 147Sm/.144Nd enrichment factor ySm/Nd -

(147Sm/l44Nd) sampl e _(147Sm/144Nd) chondrit e.

where ( 147 Sm/ 144 Nd) chondri te = 0.1967,

-1

I/A &

^

!•//•/

J

MORB ALK

E

g: 200

Cambroon Type 1 Cambroon Type 2 Highbury IAT Cedarton JL

12 8 Ti (ppm.) *10 3

_L 16

Figure 4 (a) Cr vs Y (Pearce 1980) and (b) Ti vs V (Shervais 1982) for Early Permian basalts. Symbols as in Figure 2. Fields are IAT, island-arc tholeiites; MORB, mid-ocean ridge basalts; ALK, alkalic ocean island basalts.

indicates the degree of enrichment or depletion of Sm compared to Nd, relative to the chondritic ratio. Highbury Volcanics have / S m / N d ranging from -0.215 to -0.178, similar to values for the Cedarton Volcanics

154

W. J. SIVELL AND M.T. McCULLOCH

Table 3 Representative Sm-Nd isotope data.

Sm (ppm)

Nd (ppm)

Nd/ Nd 144

£ d( )*

Cambroon Type 1 basalts AM 10 AM 14 AM 16 AM16B AM43

3.57 2.10 4.01 3.47 3.78

13.51 6.41 11.67 10.78 10.77

0.1596 0.2210 0.2077 0.1946 0.2121

0.513047 ± 7 0.513210 ± 9 0.513116 ± 8 0.513048 ± 7 0.513179 ± 9

9.00 9.17 8.72 7.84 9.80

Cambroon Type 2 basalts AM9B AM42

1.99 1.69

5.09 4.13

0.2364 0.2471

0.513249 + 9 0.513258 ± 9

10.33 10.14

Cedarton basalts CV22 CV31 CV30 CV1X1B

3.04 3.51 3.93 3.38

11.18 12.76 14.52 12.86

0.1646 0.1665 0.1638 0.1587

0.512913 ± 8 0.512974+ 10 0.512947 ± 8 0.512917 ± 8

6.23 7.36 6.92 6.51

Highbury basalts GY9 GY4

5.23 3.78

20.47 14.16

0.1544 0.1616

0.512842 ± 8 0.512862 + 9

5.20 4.14

147

Sm/ Nd 144

143

N

T

* Nd CO calculated for emplacement age ca 270 Ma. The procedures used for the isotope analyses of Sm and Nd are described elsewhere (McCulloch & Chappell 1982). The N d / N d ratio of La Jolla standard measured at the Australian National University during the course of this work is 0.511872 + 5. Total procedural blanks for Nd are <300 pg. The s ^ values are defined as: ySm/Nd - '(143Nd/^Nd)init x 10 ( Nd/ Nd) JHUR £

,43

144

4

143

where: ( Nd/ Nd)^ l43

= ( Nd/ Nd)°

144

,43

144

- ( W Nd)°

144

, 4

(e Sm _ i)

1 4 4

x

T

Present day reference values (Jacobsen & Wasserburg 1980) for the chondritic unfractionated reservoir (CHUR) (De Paolo & Wasserburg 1976) are ( Nd/ Nd)? u = 0.512650 and ( Sm/ Nd)2 uR = 0.1967; ?i m = 6.54 x 10" y ; ( Nd/ Nd) hit is the measured ratio in the rock, corrected for decay since the time of crystallisation T. ,43

144

H

,47

R

(-0.193 to -0.139), while for the Cambroon basalts, jfSm/Nd f _o.l88 to +0.123 (Type 1) and from +0.256 to +0.202 (Type 2). The latter samples are like LREE-depleted MORB (positive /Sm/Nd positive s ) , while most of the other basalts show LREE-enrichment (negative / , positive s ^ ) similar to island-arc basalts and ocean island basalts. Initial Nd-isotopic ratios of the Gympie rocks range from values similar to N-MORB and the highest values shown by E-MORB ( e ^ ~ +10) to near the lower limit of ratios for intra-oceanic island-arc tholeiites (sNd 6) (Figure 5) (McCulloch & Gamble 1991). This isotopic diversity among closely related rocks implies close proximity of both a long-term depleted (MORB-like) source and a source(s) with relative enrichment (lower v a r i e s

r o m

?

Nd

S m / N d

<+

,44

H

S

12

1

143

144

'arc-like' or 'ocean island basalts-like' integrated Nd/ Nd). Nd isotope compositions of the Early Permian basalts display a bimodal distribution which is related to geographic location. Two distinct isotopic trends are evident. Trend 1 is shown by Cambroon basalts that possess high Nd/ Nd ratios which increase only slightly ( e = +9 to +10.3) with increasing Sm/ Nd. These basalts were derived from a long-term LREE-depleted mantle source with z^d values (+10 to +12) similar to that expected for Permian N-MORB type mantle. Trend 2 is exhibited by Highbury and Cedarton Volcanics and some Cambroon basalts and reveals a much more pronounced decrease in e ^ (+10 to +4) sympathetic with Sm/ Nd. The isotopically most enriched 143

144

143

144

147

Nd

147

144

144

GEOCHEMISTRY, PERMIAN BASALTS, QLD samples at the low-e d end of Trend 2 are the Highbury island-arc tholeiites. The range in Nd isotopic compositions is gradational for the Cambroon, Cedarton and Highbury rocks. The high precision of the data allows, in detail, two sub-trends to be identified within overall Trend 2. Trend 2A comprises the main linear distribution of the data extending from the isotopically most depleted Cambroon Type 2 samples towards the low-s i Highbury basalts, but limited to the upper half of this trajectory. Cedarton basalts comprise curvilinear Trend 2B which at its low-e i end lies along a line joining the Highbury island-arc tholeiites with the lowest Sm/Nd Cambroon Trend 1 samples, but significantly, at its high-SNd end defines a path back towards the Cambroon basalts of the main Trend 2A. The fit of the data to linear-curvilinear trends suggests that the isotopic variation can be explained by various 'two component' mixing scenarios. Correlations between Nd isotope ratios and parent daughter ratio Sm/Nd are interpreted as due to recent mixing rather than closed system evolution of a common homogeneous source. Fundamental differences in trace-element chemistry between the volcanic suites correlate well with the observed isotopic variations. This enables constraints to be placed upon the nature of isotopically distinct endmember mixing components. Sivell and Waterhouse (1988) concluded that the Highbury island-arc tholeiites formed during an early submarine stage of an intraoceanic island arc. The Cedarton and Cambroon basalts possess even more pronounced MORB-like geochemical affinities and show no evidence to indicate any involvement of continental crust in their genesis (such as decreasing Ti/Yb or increasing K/P). However, good inverse correlations are observed between Nd isotope compositions and chemical ratios (in particular Ba/Ti, Nd/Sr and Ba/Nb: Figure 2) which are not controlled by fractionation, but reflect the chemistry of their source regions. These are consistent with the involvement of subduction-related components in the genesis of the low-e^ rocks. LILE/HFSE and LILE/REE ratios [typically high in arc rocks (Perfit et al 1980; Saunders & Tarney 1979) as well as in the Highbury and Cedarton Volcanics] reflect the addition N

N(

Nc

DISCUSSION Backarc evolution Previous geochemical studies (Gill 1976; Hawkins 1976; Hawkins & Melchoir 1985) have noted broad similarities between tholeiitic basalts from backarc basins and those generated at mid-ocean ridges. Relative to normal MORB, some backarc basin basalts (e.g. Mariana Trough basalts and samples from the marginal parts of the Lau Basin near the Lau Ridge and Tonga arc) have higher abundances of LILE (K, Rb, Ba, Sr and LREE) while abundances of HFSE (Ti, Zr, Nb)

0.51282

o

(Depleted mantle)

0.51276 ~

CO

h"O 0.51270 Z

Figure 5 Sm/ Nd vs N d / N d for Early Permian basalts. Symbols as in Figure 2. Error bar shows maximum total procedural analytical uncertainty for Nd-isotope determinations. 147

143

,44

144

t5 0.51264 z 3 0.51258 0.15

155

of slab-derived components [hydrousfluidsrich in LILE and LREE—including isotopically enriched (i.e. unradiogenic) Nd] to the overlying supra-subduction zone mantle wedge. This does not mean that contributions from subduction-related fluids are solely responsible for the observed isotopic variations, because good positive correlations are also apparent between and trace element ratios such as Ti/V, Ti/Y and Zr/Y (Figure 3). These ratios are largely unaffected by slab-derived contributions and reflect the degree of depletion in incompatible elements of the mantle source prior to interaction with slab components (Woodhead et al. 1993). Covariations of isotopic values with LILE/HFSE and LILE/REE ratios as well as ratios involving incompatible elements (e.g. Ti, V, Zr and Y) that are not readily incorporated in hydrous fluids, is best explained by mixing between actual depleted and (relatively) enriched mantle material, or mixing of melts derived from these at or close to their source. Addition of LILE-enriched slab-derived fluids alone to the suprasubduction zone mantle would increase Ba/Ti and cause concomitant lowering of Sm/Nd and e w but would not appreciably affect ratios like Ti/V. Concurrent changes in all these ratios implies mixing between mantle sources that already possessed different degrees of depletion prior to subduction. One mantle end-member may, however, have incorporated a slabderived component and represent subduction-modified mantle.

Cambroon Type 1 Cambroon Type 2 Cedarton Highbury

j(Subd'n. modified mantle)

0.20 147

Sm/144Nd

zQ. hO -si o

0.25

156

W. J. SIVELL AND M.T. McCULLOCH

are lower and transitional to those of island-arc basalts (Volpe et al. 1987). Other backarc basin basalts, such as basalts from small seamounts in the central Lau Basin (Volpe et al. 1988) and some East Scotia Sea samples (Saunders & Tarney 1979), have trace element abundances and ratios resembling incompatible element and LREE-enriched basalts (E-MORB and ocean island basalts). The influence of subduction-related processes leading to high LILE abundances may be greatest during the early stages of backarc basin development, diminishing as the basin widens and spreading centres move away from the associated (proximal) island arc system (Iyeda & Yuasa 1989). The Highbury Volcanics are clearly island-arc tholeiites, representing an immature submarine stage of island arc development (Sivell & Waterhouse 1988). By contrast, the Cedarton Volcanics and Cambroon basalts show chemical features that typify backarc basin basalts — including Nd isotopic compositions and interelement ratios (e.g. Sr/Nd, Ti/V and Ba/Nb) intermediate between MORB and island-arc tholeiites. Nb/Zr ratios (0.15-0.25) in these rocks are like those of E-MORB or ocean island basalts and are much higher than those of the Highbury basalts. None of the Cambroon or Cedarton basalts exhibit the selective Rb, Ba and K enrichment coupled with pronounced Zr and Nb depletion and lower absolute REE abundances which are typical geochemical signatures of island-arc tholeiites and transitional backarc basin basalts (Saunders & Tarney 1979). In this respect they more closely resemble basalts from the Lau Basin (Volpe et al. 1988) and East Scotia Sea (Saunders & Tarney 1979) that have geochemical affinities with E-MORB. Relatively high Ba/Ti and Ba/Sm in the Cedarton basalts, however, do most likely reflect introduction into their mantle source regions of (soluble) LILE and water-rich fluids derived from dehydration of subducted lithosphere. Higher Ti/V, Ti/Y and Zr/Y ratios for the Cambroon (and to a lesser extent the Cedarton) basalts compared with the Highbury island-arc tholeiites indicate that the Cambroon-Cedarton backarc basin basalts were derived from a mantle source that was less depleted than the source of the Highbury island-arc tholeiites (prior to subduction-related metasomatism of the arc source). Woodhead et al. (1993) have observed that backarc basin basalts from western Pacific island-arc-backarc basin systems are typically derived from less depleted mantle sources than those in the associated intra-oceanic arc. Subduction-related modification of the sub-arc mantle wedge, causing elevated LILE/HFSE and LREE/HFSE ratios and lowering of s^d values, may mask the initial extent of mantle depletion in arc sources, as demonstrated by the Highbury island-arc tholeiites. The discontinuous lensoid occurrences of the Cambroon basalts suggests that sea-floor spreading in the backarc basin was diffuse and irregular, in contrast to near steady-state volcanism at major ocean ridges. Similar styles of volcanism have been recognised in other backarc settings (e.g. Lau Basin: Lawver et al. 1976) which also lack well-developed spreading ridges. Whether or not due to sea-floor

spreading, and consequent widening of the more mature backarc basin, the source of the Cambroon basalts, with their more MORB-like geochemical affinities (especially in the Type 1 basalts), was more isolated from the influence of components derived from the subducting slab.

Heterogeneous mantle melting Geochemical and isotopic features of Cambroon basalts which follow the high-e^d isotopic trend (Trend 1) provide further information about melting processes and the nature of mantle sources in the Early Permian backarc environment at the Gondwana rim. For these basalts, rapidly decreasing Sm/ Nd is accompanied by only a slight decrease in s^d- However, major and trace-element compositions are highly variable even among Cambroon basalts with similar MgO contents (~8 wt%) which were collected in close proximity from 147

°20

60

25

(b) \

100

140

Cambroon Type 1 basalts Cambroon Type 2 basalts Cedarton basalts

A A\ A

MORB^ f x

^ 15

Zr (ppm.) • A A

20

144

#

_C\J <

•jJ

10

50. 60

1.0 Ti0 2 (wt. %)

1.4

1.8

Figure 6 (a) MgO vs Zr for Early Permian basalts. Large arrow shows trend for increasing percentage of mantle melting. Solid arrows are liquid lines of descent for fractionating magmas. REE/ CH boxes show sketch chondritenormalised REE profiles at different degrees of melting. Dashed balloon shows field of East Scotia Sea backarc basin basalts (Saunders & Tarney 1979). (b) Al 0 /Ti0 vs Ti0 . Field of primitive MORB (Sun et al. 1979) spanning a range of partial melting (see text) is indicated. Symbols as in Figure 2. 2

2

3

2

GEOCHEMISTRY, PERMIAN BASALTS, QLD a single lens of related flows and small intrusive bodies. Chemically distinct basalts occur in intercalated flows on an outcrop scale. Decreasing 147 Sm/ 144 Nd is accompanied by a large decrease in Al 2 0 3 /Ti0 2 (16 to 8) (Figure 6b). Hawkins and Melchoir (1985) suggested that Mariana Trough basalts with low Ti and high Al/Ti may correspond to extensive (25-30%) partial melts of a model pyrolite mantle source. In high-degree partial melts, Al/Ti reflects source composition, while at lower percentages of melting Ti behaves as an incompatible element, becoming more enriched in the melt relative to Al. The lack of negative Eu anomalies in REE profiles for the Cambroon basalts indicates that decreasing Al/Ti is unlikely to be due to major plagioclase feldspar fractionation from parental magmas. Rather, Cambroon (Trend 1) basalts with the highest Ti and lowest Al/Ti ratios most likely correspond to less extensive partial melts. On the MgO vs Zr plot (Figure 6), the most primitive Cambroon basalts define a subhorizontal trend of markedly decreasing Zr content (120-30 ppm) at essentially constant MgO (-8 wt%), concomitant with increasing Al 2 0 3 /Ti0 2 (increasing extent of melting). Similar trends have been observed for other backarc basin basalts suites (e.g. East Scotia Sea: Saunders & Tarney 1979). More fractionated basalts follow linear trends at a high angle to the proposed melting trend, and represent crystal fractionation paths for discrete magma batches generated at varying degrees of melting. REE distributions for the Cambroon Trend 1 basalts (Figure 7) also imply differential partial melting. Modelling of the REE fractionation produced in liquids which are variable-degree fractional melts of mantle mineral assemblages indicate that the total variation in REE fractionation shown by the Cambroon basalts [(Sm/Nd)CH = 0 . 8 - 1.3] can be achieved over a 10 percent range of melting [e.g. 15-25%, depending on the degree of initial depletion of the source (Loubet et al. 1975; Wood 1979)]. For partial melting of a source that is homogeneous with respect to Nd isotopic composition the 143Nd/ 144 Nd ratio in the melt would be unchanged by differential partial melting, with Sm/Nd decreasing at progressively smaller degrees of melting. But partial melting of a source that was isotopically heterogeneous on the scale of melting, or mixing between isotopically distinct sources/melts, could result both in variable Sm/Nd and Nd isotopic ratios. For the Cambroon Trend 1 basalts, the small but systematic isotopic shift ( 8 N d = +10.3 to +9.0) which accompanies varying extent of mantle melting is most likely due to either: (i) preferential incorporation of an isotopically enriched low-temperature melting component (at lower degrees of melting) during backarc volcanism; or (ii) mixing between depleted and enriched components in the source region, possibly involving influx of more fertile mantle into the supra-subduction zone mantle wedge. Geochemical features of the Cambroon basalts (e.g. relatively low Zr/Nb) indicate that this enriched component constitutes a mixing end-member additional to the 1 o w - 8 i mantle (depleted mantle wedge metasomatised by slab-derived fluids) that exerts the main control on the overall isotopic and chemical N c

LU LU DC

1

I

I I

LaCe

I Nd

157

—•—AM9B —o—AM10 —•—AM16 - - x - - Modelling I I l I I I l—I—I—l—I SmEuGdTb DyHo Er

Yb

Figure 7 Chondrite-normalised REE plots for (a) Highbury and Cedarton basalts, and (a) Cambroon basalts. Dashed profiles define the variation of REE fractionation produced in model melts over a 10% range of variable-degree fractional melting of mantle mineral assemblages, with residual assemblage adjusted for melting mode (Kelemen et al 1990).

variation of the studied basalt suites, and which predominates in the Highbury island-arc tholeiites (Zr/Nb>20). Enriched mantle components have previously been inferred to be present beneath backarc basins (Volpe et al 1987, 1988; Iyeda & Yuasa 1989), and the less depleted nature of backarc basin basalts relative to proximal island-arc tholeiites has been noted (Woodhead et al 1993). Excluding the effects of subduction-related metasomatism of the mantle wedge, the incorporation of even small amounts of other enriched components in backarc basin magmas will mask the intrinsically depleted character of suprasubduction zone melts. An important question concerns the scale and timing of development of the inferred isotopic heterogeneities in the wedge. Batiza (1980) found samples from volcanic seamounts near the East Pacific Rise show large differences in (La/Sm) N even in a single volcano, implying small-scale heterogeneities in their source. The (Trend 1) Cambroon samples (with a range of isotopic compositions) were likewise collected from a series of closely related flows and minor intrusives, so their observed range of isotopic compositions may imply that the enriched material which they sampled was distributed on a fine scale within the overall depleted mantle source region.

158

W. J. SIVELL AND M.T. McCULLOCH

However, it is likely that early formed, isotopically enriched upper mantle heterogeneities would have been eliminated prior to second-stage melting in the backarc regime. The enriched component more likely reflects upwelling of more fertile mantle from depth into the wedge region at the onset of backarc basin development (Iyeda & Yuasa 1989). On a range of geochemical plots shown here (Figures 2, 3, 5) good correlations are also evident between isotopic and chemical compositions of both Trend 2A and Trend 2B samples. The curvilinear isotopic trend of the Cedarton basalts (Trend 2B) in Figure 5 is particularly significant and may shed light on interactions among the three mantle end-members noted above, that is: (i) depleted N-MORB source mantle; (ii) subduction-modified isotopically enriched mantle (predominant in the Highbury island-arc tholeiites); and (iii) non-subduction-related enriched mantle similar to the source of E-MORB or ocean island basalts. The lowest e^d Cedarton basalts (those with the greatest slab-derived component) follow a trajectory from close to the field of the Highbury data initially toward the low Sm/Nd end of the Cambroon Trend 1 (non-slab-related enriched mantle source). The generation of these basalts probably involved mixing between sub-arc and E-MORB or ocean island basalts-like enriched mantle components. By contrast, the highest £Nd Cedarton basalts begin to plot away from this trend, defining a curve towards the Trend 2A Cambroon sample field. This trend is expected to result from the gradual depletion of the volumetrically minor isotopically enriched component in the backarc mantle with continued melt extraction in the evolving rift. Analogous trends are evident on chemical plots involving trace-element ratios which characterise the various proposed mantle end-members. They reflect initial mixing between E-MORB components (prior to their elimination) and island-arc tholeiite components, possibly during nascent backarc basin development (Iyeda & Yuasa 1989). Subsequently, mixing between N-MORB and island-arc tholeiite end-members became more predominant as the rift evolved. In the mature rift, the subduction component diminished with time, basin widening removing backarc eruptive centres further from the influence of the slab. Continental growth at the Gondwana rim Geochemical and isotopic data presented in this paper imply distinctive magmatic evolutions for Early Permian mafic volcanic suites in southeast Queensland. These require substantially differing tectonic settings for the origin of their host terranes, including intraoceanic island arc and nascent to more mature backarc basin environments. Here we examine briefly the relation of the Gympie Province and Early Permian fault basins to subduction and consequent crustal growth at the Gondwana rim and suggest some possible terrane interactions. Time-transgressive lateral accretion models for continental growth, which involve addition to continental lithosphere of terranes formed at active plate margins

have been applied to numerous circum-Pacific regions (Seely et al 1974; Marriner & Millward 1984). Plate convergence (with or without the added complexity of transform faulting) can cause terranes of diverse origin (e.g. magmatic arc, subduction complex or island arc) to migrate as coherent structural units to ultimately assemble at the continental edge forming composite suspect terranes or blocks. Some of the resulting juxtaposed terranes may represent exotic blocks that collided with the continent subsequent to their transport from a distal source, for example remnants of intraoceanic island arcs rafted by sea-floor spreading to the continental margin. The Highbury Volcanics were generated in an intraoceanic region remote from any continental influence. The shallow marine association of the Gympie Group (Highbury Volcanics and overlying sediments) constitutes a discrete exotic terrane requiring tectonic translation towards the palaeo-Pacific Gondwana rim, ultimately accreting to the Gondwana (continental) margin in Triassic time (Harrington 1983). To the west of the Gympie Group, the Cambroon beds are more complexly deformed than the Highbury Volcanics and their associated (forearc) sediments (Murray 1988) and both the Cambroon beds and Cedarton Volcanics are chemically distinct from the Highbury basalts. Neither the Cambroon beds nor the Cedarton Volcanics can be regarded as facies equivalents of the Gympie Group. Geochemical and Nd isotope data reported here indicate that the Cambroon beds and Cedarton Volcanics formed in a backarc setting, presumably to the rear of the Gympie (Highbury) island arc and the data are compatible with the expected sequential involvement of mantle end-members during progressive evolution of the backarc regime. The probable Early Permian fauna recognised in some strata from the Cambroon beds and Cedarton Volcanics indicates that these units at least partly overlap the Gympie Group in age (Murray et al 1979). [There are also some older (Carboniferous) elements present, at least within the Amamoor beds (Murray et al 1989), of probable sea-floor origin (slivers of oceanic basement upon which the backarc sequence accumulated?)]. Although clearly distinct from the Highbury islandarc tholeiites, the Cedarton Volcanics also possess a number of close petrographic, chemical and isotopic affinities with the arc volcanics. The curvilinear isotopic trend of the Cedarton basalts (Trend 2B), with its implied mixing between depleted mantle and rapidly diminishing subduction-related and enriched mantle end-members, suggests that they may represent early formed volcanics in a (nascent) backarc basin. By contrast, distinct MORB-like affinities of the Cambroon basalts suggest that these rocks were emplaced during more mature stages of marginal basin evolution. Tectonic translation of the Gympie arc towards the Gondwana rim, due to either subduction or strikeslip faulting, may have telescoped the facies of the marginal basin to the rear of the island arc. This could also explain the juxtaposition of the older (Early Carboniferous) sea-floor fragments of the Amamoor beds (not discussed in this paper) with younger Early

GEOCHEMISTRY, PERMIAN BASALTS, QLD Permian sequences. Further isotopic studies on both Permian and Carboniferous volcanic rocks from (eastern and western) Gympie Province and other Early Permian extensional basins within the New England Orogen should yield a better understanding of the nature of the collage of diverse terranes assembled at this part of the Gondwana rim, and how the growth of the continent is linked to the compositional evolution of the Late Palaeozoic mantle. ACKNOWLEDGMENTS WJS is grateful for assistance with mass spectrometry procedures by G. E. Mortimer, L. Kinsley and P. Oswald-Sealy at the Research School of Earth Sciences, Australian National University. Constructive reviews were provided by D. A. Gust and L. C. Cranfield. REFERENCES ABBEY S. 1980. Studies in 'standard samples' for use in the

general analysis of silicate rocks and minerals: Part 6. 1979 edition of 'usable' values. Geological Survey of Canada Paper 80-14. ALLEGRE C . J. 1987. Isotope geodynamics. Earth and Planetary Science Letters 86, 175-203. BATIZA R. 1980. The origin and petrology of young oceanic central volcanoes: are most tholeiitic rather than alkalic? Geology 8, 447-452. CANN J. R. 1969. Spilites from the Carlsberg Ridge, Indian Ocean. Journal of Petrology 10, 1-19. CHOW T. J., STERN R. J. & DIXON T. H. 1980. Absolute and relative abundances of K, Rb, Sr and Ba in circum-Pacific island arc magmas, with special reference to the Marianas. Chemical Geology 28, 111-121. CRANFIELD L. C . & SCOTT M. 1993. Geology of the Gympie 1:100 000 Special area (9445, 9545). Queensland Geological Record 1993 / 20. CRAWFORD A . J., FALLOON T . J. & EGGINS S. 1 9 8 7 . T h e

origin of island arc high alumina basalts. Contributions to Mineralogy and Petrology 97, 417-30. J. & WASSERBURG G. J. 1976. Nd isotopic variations and petrogenetic models. Geophysical Research Letters 3, 249-252.

D E PAOLO D.

FREY F. A., BRYAN W . B. & THOMPSON G . 1974. Atlantic

Ocean floor: geochemistry and petrology of basalts from Leg 2 and 3 of the Deep Sea Drilling Project. Journal of Geophysical Research 79, 5507-5527. GALER S. J. G . , GOLDSTEIN S. L. & O ' N I O N S R. K. 1 9 8 9 .

Limits on chemical and convective isolation in the earth's interior. Chemical Geology 75, 257-290. GILL J. B. 1976. Composition and age of Lau Basin and Ridge volcanic rocks: implications for evolution of an interarc basin and remnant arc. Geological Society of America Bulletin 87, 1 3 8 4 - 1 3 9 5 . GRAPES R. H. 1976. Low grade alteration of basic rocks in Hokkaido, Japan. Journal of the Faculty of Science Hokkaido University Series 4,17, 373-385. HARRINGTON H. J. 1983. Correlation of the Permian and Triassic Gympie terrane of Queensland with the Brook Street and Maitai terranes of New Zealand. In: Permian Geology of Queensland, pp. 431-436. Geological Society of Australia, Queensland Division, Brisbane.

159

HAWKINS J.

W. JR. 1976. Petrology and geochemistry of basaltic rocks of the Lau basin. Earth and Planetary Science Letters 28, 283-297.

HAWKINS J. W . & MELCHOIR J. T. 1985. Petrology of Mariana

Trough and Lau basin basalts. Journal of Geophysical Research 90, 1 1 4 3 1 - 1 1 4 6 8 . IYEDA Y . & YUASA M. 1989. Volcanism in nascent back-arc

basins behind the Shichito Ridge and adjacent areas in the Izu-Ogasawara arc, north-west Pacific: evidence for mixing between E-type MORB and island arc magmas at the initiation of back-arc rifting. Contributions to Mineralogy and Petrology 101, 377—393. JACOBSEN S. B. & WASSERBURG G . J. 1980. SM-Nd isotopic evolution of chondrites. Earth and Planetary Science Letters 50, 1 3 9 - 1 5 5 . KELEMENP. B., JOHNSON K . T . M . , KINZLER R. J. & IRVING

A. J. 1990. High-field-strength element depletions in arc basalts due to mantle-magma interaction. Nature 345, 521-524 LAWVER L . A . , HAWKINS J. W . & SCLATER J. G .

1976.

Magnetic anomalies and crustal dilation in the Lau Basin. Earth and Planetary Science Letters 3 3 , 27—35. LOUBET M . , SHIMIZU N . & ALLEGRE C . J. 1975. Rare earth elements in alpine peridotites. Contributions to Mineralogy and Petrology 53, 1-12. MARRINER G. F. & MILLWARD D. 1984. The petrology and geochemistry of Cretaceous to Recent volcanism in Colombia: the magmatic history of an accretionary plate margin. Journal of the Geological Society of London 141, 473-486. MCCULLOCH M . T. & CHAPPELL B. W. 1982. Nd isotopic characteristics of S- and 1-type granites. Earth and Planetary Science Letters 58, 51-64. MCCULLOCH M . T. & GAMBLE J. A., 1991. Geochemical and geodynamical constraints on subduction zone magmatism. Earth and Planetary Science Letters 102, 358-374. MURRAY C. G. 1987. Review of the geology of the Gympie district, south-east Queensland, with particular reference to stratigraphic, structural and tectonic problems. In: Murray C. G. & Waterhouse J. B. eds. 1987 Field Conference, Gympie District, pp. 1—19. Geological Society of Australia, Queensland Division, Brisbane. MURRAY C . G. 1988. Tectonostratigraphic terranes of southeast Queensland. In: Hamilton L. H. ed. Field excursions handbook for the Ninth Australian Geological Convention, pp. 1 9 - 5 1 . Geological Society of Australia, Queensland Division, Brisbane. MURRAY C. G., HOLCOMBE R. J. & ISHIGA H. 1989. Suspect terranes of the Gympie Province, New England Fold Belt. Geological Society ofAustralia Abstracts 24, 9 9 - 1 0 0 . MURRAY C . G . , MCLUNG G . R . & WHITAKER W . G . 1 9 7 9 .

Early Permian fossils from the Amamoor beds and the age of the Booloumba beds—a new stratigraphic unit of the North D'Aguilar Block. Queensland Government Mining Journal 80,71-78. HUTTON J. T . 1 9 6 9 . An accurate X-ray spectrograph^ method for the analysis of a wide range of geologic samples. Geochimica et Cosmochimica Acta 33,

NORRISH K . &

431-451. PEARCE J. A.

1980. Geochemical evidence for the genesis and eruptive setting of lavas from Tethyan ophiolites. In: Panayitou A. ed. Ophiolites: Proceedings, International Ophiolite Symposium, Cyprus, pp. 261—262. Ministry of Agriculture and Natural Resources, Geological Survey Department, Nicosia, Cyprus.

160

W. J. SIVELL AND M.T. McCULLOCH

PERFIT M . R., GUST D. A., BENCE A. E., ARCULUS R. J. &

TAYLOR S. R. 1980. Chemical characteristics of island arc basalts: implications of mantle sources. In: Le Maitre R. W. & Cundari A. eds. Chemical characterization of tectonic provinces, pp. 227-256. Chemical Geology 30. RINGWOOD A. E. 1966. The chemical composition and origin of the earth. In: Hurley P. M. ed. Advances in Earth Science, pp. 287-356. Massachusetts Institute of Technology Press. ROBSON D. & CANN J. R. 1982. A geochemical model of mid-

ocean ridge magma chambers. Earth and Planetary Science Letters 60, 93-104.

SUN S. S. & NESBITT R. W. 1978. Petrogenesis of Archaean ultrabasic and basic volcanics: evidence from rare earth elements. Contributions to Mineralogy and Petrology 65, 301-325. SUN S. S., NESBITT R . W . & SHARASKIN A. Y . A. 1979.

Geochemical characteristics of mid-ocean ridge basalts. Earth and Planetary Science Letters 44, 119-138. VOLPE A. M . , MACDOUGALL J. D. & HAWKINS J. W . 1987.

Mariana Trough basalts (MTB): trace element and Sr-Nd isotopic evidence for mixing between MORB-like and arc-like melts. Earth and Planetary Science Letters 82, 241-254.

SAUNDERS A. D. & TARNEY J. 1979. T h e g e o c h e m i s t r y o f

VOLPE A. M . , MACDOUGALL J. D. & HAWKINS J. W . 1988.

basalts from a back-arc spreading centre in the East Scotia Sea. Geochimica et Cosmochimica Acta 43, 555-572.

Lau Basin basalts (LBB): trace element and Sr-Nd isotopic evidence for heterogeneity in back-arc basin mantle. Earth and Planetary Science Letters 90, 174-186.

SEELY P. R., VAIL P. R. & WALTON G. G. 1974. Trench slope

model. In: Burke C. A. & Drake C. L. eds. The Geology of Continental Margins, pp. 853-872. Springer-Verlag, Berlin. SIVELL W . J. & WATERHOUSE J. B. 1988. P e t r o g e n e s i s of

Gympie Group volcanics: evidence for remnants of an Early Permian volcanic arc in eastern Australia. Lithos 21, 81-95. SIVELL W . J., STOCKSIEK C . M . & WATERHOUSE J. B 1990.

Metavolcanic suites from suspect late Palaeozoic terranes of the Gympie Province, south-east Queensland. Geological Society of Australia Abstracts 25, 197-198. SHERVAIS J. W. 1982. Ti-V plots and the petrogenesis of modern and ophiolitic lavas. Earth and Planetary Science Letters 59, 101-118.

SUN S. S. 1980. Lead isotopic study of young volcanic rocks from mid-ocean ridges, ocean islands and island arcs. Philosophical Transactions of the Royal Society of London A297, 409-445. SUN S. S. & MCDONOUGH W. F. 1989. Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: Saunders A. D. & Norry M. J. eds. Magmatism in the Ocean Basins, pp. 313-345. Geological Society of America Special Paper 42.

WATERHOUSE J. B. & BALFE P. E. 1987. S t r a t i g r a p h i c a n d

faunal subdivisions of the Permian rocks at Gympie. In: Murray C. G. & Waterhouse J. B. eds. 1987 Field Conference, Gympie District, pp. 20-33. Geological Society of Australia, Queensland Division, Brisbane. WEBB A . W . & MCDOUGALL I. 1968. T h e g e o c h r o n o l o g y of

the igneous rocks of eastern Queensland. Journal of the Geological Society of Australia 15, 313-346. WINCHESTER J. A . & FLOYD P. A. 1977. G e o c h e m i c a l dis-

crimination of different magma series and their differentiation products using immobile elements. Chemical Geology 20, 325-343. WOOD D. A. 1979. A variably veined suboceanic upper mantle: genetic significance for mid-ocean ridge basalts from geochemical evidence. Geology 7, 499-503. WOODHEAD J., EGGINS S. & GAMBLE J. 1993. High field

strength and transition element systematics in island arc and back-arc basin basalts: evidence for multi-stage melt extraction and an ultra-depleted mantle wedge. Earth and Planetary Science Letters 114, 491-504.

(Received 10 April 1996; accepted 22 October 1996)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 161-177.

Age and structural characterisation of the Texas megafold, southern New England Orogen, eastern Australia P. G. LENNOX 1 A N D P. G. FLOOD 2 1

2

Department of Applied Geology, University of New South Wales, NSW 2052, Australia. Department of Geology and Geophysics, University of New England, Armidale, NSW 2351, Australia. Analysis of the bedding and cleavage data around the Texas megafold demonstrates the structure is a steeply south-southeast-plunging, tight to isoclinal syncline. Two or possibly three post-bedding foliations have been observed both in the field and thin-section; the first is oriented subparallel to bedding and is the dominant foliation. The patchily developed, secondary foliation of variable morphology was probably formed during the early stages of bending of the originally meridionally oriented Texas beds. The third foliation is oriented east-west and occurs in both Texas beds and the Permian basins. The dominant foliation was folded during megafold formation, and is cross-cut by the secondary foliation, which is fanned around the megafold. The megafold developed over a period from pre-late Asselian until at least post-middle Artinskian on the basis of the Permian faunal assemblages, deformation of the Permian basins—Texas beds and independent palaeomagnetic studies. This is contrary to existing tectonic models, which propose Late Carboniferous to Early Permian megafold formation. Key words: New England Orogen, structural analysis, tectonic model, Texas megafold.

INTRODUCTION The New England Orogen consists of a forearc basin in the west separated by the Peel Fault from an accretionsubduction complex to the east, and granites (Leitch 1974). The doubling of the accretion-subduction complex through megafolding could account for the offset between the arc-forearc succession of the southern New England Orogen and the northern New England Orogen. The Texas megafold was first described by Lucas (1960), with further studies by Flood and Fergusson (1984), Fergusson and Flood (1984) and Korsch and Harrington (1987). The temporal and spatial character of structures in the Texas beds and Permian slope basins has important implications for the development of the megafold and hence the tectonic history of the southern New England Orogen. This paper examines the pattern of foliations around the Texas megafold, in the Texas beds and Permian slope basins, with the aim of characterising the folding event and providing age constraints on the formation of the megafold. Previous studies of the Texas area (Figure 1) have been restricted to localised Masters or Honours theses (Lucas 1960; van Noord 1993; Boxall 1972; Oxley 1972; Moore 1982; Forster 1991; Cohen 1991) or broad regional studies (Olgers & Flood 1974; Butler 1974; Fergusson & Flood 1984), and more recent dating of some of the cherts within the sequence using radiolarians (Aitchison & Flood 1990; Aitchison et al. 1992). Palaeomagnetic studies (Aubourg et al. 1994), deep seismic reflection profiling (Korsch et al. 1997) and lineament analysis (Vinayan et al. 1993) have supported a megafold model for this area (Figure 1). Lucas (1960) recognised the presence of a regional

synform, while Fergusson and Flood (1984) were the first to provide stratigraphic and structural evidence for the presence of a megafold. GEOLOGY AROUND THE MEGAFOLD The most comprehensive account of the lithologies exposed around the megafold is that of Olgers and Flood (1974), and summarised by Fergusson and Flood (1984). They described the Texas beds, the stratigraphic unit which is folded by the megafold, as a 'monotonous turbidite unit with minor cherts, greenstone, and limestone'. They recognised four tectonostratigraphic units informally referred to as Ctl, Ct2, Ct3 and Ct4, and proposed that the Texas beds formed a Carboniferous subduction complex. Regional mapping has shown that the Texas beds consist of many steeply dipping fault slices, which internally contain stratigraphic sequences several kilometres in thickness (Cohen 1991; Forster 1991). Wellman et al. (1994) provided an interpretation of the regional gravity and magnetic patterns in the southern New England Orogen consistent with megafolding. However, their interpretation of the lithological units is at variance with that proposed by Flood and Fergusson (1982), Fergusson and Flood (1984), Cross et al. (1987) and Murray et al (1987). Resolution of this divergence of ideas is beyond the scope of the current project. Unconformably overlying the Texas beds are several isolated Permian slope basin deposits of varying ages. The age progression from oldest to youngest is Alum Rock, Terrica, Silver Spur and ?Glenlyon, ?Pikedale and Glenmore. D. J. C. Briggs (pers. comm. 1996) has identified latest Asselian to Sakmarian faunas in Alum Rock sedimentary rocks, and similar but not precisely

162

b

P. G . L E N N O X A N D P. G. F L O O D

)

\ • '

ir

^ \\\\ ^

\m ^ ^

•28° 30'

\ \ \ \ \ \

\v > . v

N

\

2

v \

Great Artesian Basin

X

C Mperm ian

-++U \\ J : \ - + + + -hV 1 / • * K++++JV ( ' \ ++/ * J \ s:: I + +I

»

v

\ \ x

:i5j°30'\ : i

diagnostic fossils in the Terrica sedimentary rocks suggestive of a younger age than the Alum Rock rocks. He has also identified Sakmarian to middle Artinskian faunas in rocks from the Silver Spur area. Roberts et al. (1995) reported an age of about 294 Ma for volcanics from within the Alum Rock slope basin. Several of the Permian slope basins display folding and cleavage development, and some were deformed during the formation of the megafold. This is consistent with the suggestion of Korsch in Murray et al. (1987) that the megafold formed in the late Early Permian. STRUCTURAL ANALYSIS OF THE TEXAS MEGAFOLD Texas beds First we will examine the spatial and temporal relations of bedding, non-bedding foliations and limited beddingfoliation intersection lineation data around the megafold in suitable domains to show the overall pattern and what this may show us about the tectonic development of this region. BEDDING

The Texas megafold is a part of the Texas-Coffs Harbour orocline (Korsch & Harrington 1987), a fold couple of half-wavelength about 120 km and amplitude 135 km and as such the Texas megafold covers much of nine 1:100 000 map sheets. Six domains were characterised, three each on either the northeastern (2, 4 & 6) or northwestern (1, 3 & 5) limbs of the megafold (Figures 1, 2). The gross map pattern of bedding overall reflects a megafold, but there are a number of stations where the bedding pattern is complex. The contoured plots of poles to bedding for Domains

Figure 1 (a) Locality Map of the New England Orogen, eastern Australia, (b) The Texas and Coffs Harbour megafold defined by generalised structural and lithological trends with the six structural domains defined by lightly dashed grid lines; the numbers in the top right-hand corner of each of the structural domains are used to identify that domain. The heavy and light dashed lines denote the inferred outline of the megafold and gross bedding strike respectively.

2, 4 and 6 are similar reflecting the common bedding on the northeast limb of the megafold (Figure 3b, d, f). The small-circle scatter of poles to bedding for this northeast limb reflects in part, the inherent variability of bedding expected on this limb of the megafold, the inclusion of hinge zone bedding readings because of the choice of domain boundaries or the effects of refolding.The plot of poles to bedding for Domain 1 is consistent with the generally east-west strike of bedding measured in this area (Figure 3a). The Domain 3 plot of contoured poles to bedding which includes limited data from the adjoining Yetman sheet, shows broadly northwest—southeast, north-northeast- and east—weststriking bedding as this domain lies in part, over both the northwest limb and hinge zone of the megafold (Figures 2, 3c). The pole to the best-fit great-circle girdle is oriented 70°->017° consistent with the axial trace in this area determined from lineament analysis (Vinayan et al. 1993). The Domain 5 plot includes data from the Ashford sheet combined with limited data from the Yallaroi sheet to the west and the Inverell sheet to the south (Figure 1). This plot has two or three point populations representing bedding striking north-northwest—southsoutheast (Ashford sheet) and two other populations representing bedding oriented northeast-southwest (Inverell sheet) and north-south (Yallaroi and Inverell sheets) (Figure 3e). The great-circle best-fit for Domain 5 is consistent with a fold plunging 75°->153° consistent with the axial trace determined by lineament analysis in this area (Vinayan et al. 1993) and the observed macroscopic fold (Figure 2). The intrusion of the Bundarra pluton at 280±13 Ma (S. Shaw pers. comm. 1996) possibly into a strike-slip fault undergoing dextral translation (Aitchison et al. 1992) may explain the truncation of macroscopic folds south of Ashford and bending of bedding in this domain (Figure 2).

TEXAS MEGAFOLD, NSW The map of bedding readings around the Texas megafold together with the age assignment of tectonostratigraphic units of Fergusson and Flood (1984) indicate the megafold is a synclinal fold, although somewhat complicated by mesoscopic folding and faulting. The presence of younger rocks, occurring as Permian slope basins in its core and in partly

Figure 2 Map of representative bedding trends from the Texas beds and Permian basins showing the broad megafold shape, six structural domains and labelled Permian basins (P). The short dashed lines represent roads and the long dashed line, south of Ashford, represents the axial-surface trace of a macroscopic synform. The solid lines are geological boundaries.

163

fault-bounded slivers on the northeast limb parallel to the regional strike of the Texas beds is consistent with this structure being younger in age than the Permian basins it deforms. The megafold axis plunges steeply south-southeast with an axial surface near vertical, oriented northwest-southeast to north-south. This contrasts with the Coffs Harbour orocline

164

P. G. L E N N O X A N D P. G. F L O O D

Figure 3 Mostly contoured plots of the poles to bedding in the six domains which cover the Texas megafold. Domain 3 includes data from the adjoining Yetman 1: 100000 sheet and Domain 5 includes data from the adjoining Yallaroi and Inverell 1: 100000 sheets (see Figure 1). All stereographic projections in this paper are constructed using an equal-area Schmidt net (contours at multiples of 1 % per one percent area) with, in most cases, the best-fit great-circle girdle and its pole (*) calculated. This pole corresponds to the derived fold axis for the domain. The number in the bottom right hand corner of each plot is the total number of poles from which the stereographic projection was constructed. The derived fold axis for each plot is oriented: (a) 9°->274°; (b) 45°->320°; (c) 70°->017°; (d) 50°-»167°; (e) 75°-> 153°; (f) 17°->350°. Note the overall moderate to steeply, north to northwest or south to southeast plunge of derived folds for most domains, which are comparable to the calculated beddingfoliation lineations (Figure 13c) and mesoscopic fold data (Figure 13d,).

which Korsch (1975) suggests plunges steeply northwest. DOMINANT NON-BEDDING FOLIATION AND STRUCTURES

The morphology of the dominant foliation and structures in thin-sections of the Texas beds will be described and compared briefly with structures observed in other accretion-subduction complex rocks, so as to document similarities and differences. The dominant non-bedding foliation in this region varies from a penetrative slaty cleavage to a Type A or

B opaque seam cleavage using the criterion of Gray (1978). The slaty cleavage may be domainal and almost invariably this cleavage is at a low angle to bedding or parallel to bedding. In some cases the slaty cleavage is axial planar to transposed layers within the specimen. Some of the structures observed in thin-section are comparable with structures described by Barnes and Korsch (1990 figure 3) and Kiyokawa (1992 figures 4, 5) for the Cretaceous accretionary complexes in New Zealand and Japan respectively. The Texas beds contain examples of probable slump folds, transposition, early isoclinal folding and veining, and regional shearing consistent with formation of the Texas beds as part of

TEXAS MEGAFOLD,NSW an accretionary complex. Further research would be required to properly document the structures formed during subduction-accretion, but this is beyond the scope of this paper. EVIDENCE FOR CROSS-CUTTING RELATIONSHIP

Cohen (1991) documented exposures containing two non-bedding foliations varying in appearance from two

165

pencil cleavages in mudstones to a spaced crenulation cleavage crenulating a bedding-subparallel foliation (Figure 4a). The secondary foliation cross-cutting the dominant bedding-subparallel foliation outcrops as either kinks (Figure 4b) or C shears (Figure 4c). In rare exposures of Permian and Texas beds on either limb of the megafold, an east-west-oriented cleavage occurs but in no exposure was this foliation found to be crosscutting the secondary foliations. Hence, it is possible that this foliation in fact may be merely the secondary foliation in these limb areas. However Forster (1991) and reinterpretation of Oxley's (1975) map show all three cleavages within the Permian rocks of the Pikedale and Silver Spur basins respectively. There are no exposures in these basins in which these nonbedding foliations cross-cut one another. The map and plots of the foliations in Domains 2 and 4 (Figures 5, 6b, d) contain one similarly oriented foliation striking northwest—southeast and dipping steeply to the southwest or northeast and a minor population striking almost east—west. Domain 3 contains two main foliation populations (Figures 5, 6c); one striking northeast-southwest and another population striking northwest-southeast subparallel to the major population on the bedding plot (Figure 3c). The limited foliation data in Domain 5 are oriented either parallel or normal to the bedding in this area (Figure 6e). The two separate foliations identified in the Texas beds can be related to the tectonic development of this area: (i) the bedding-subparallel foliation, now defining the outline of the megafold; and (ii) a megafold formation-related foliation, now fanned around the structure (Figure 5). The foliation orientation pattern more clearly shows the form of the Texas megafold compared with the pattern of bedding (Figure 5 and cf. Figure 2). The megafold-related secondary foliation is oriented northwest-southeast on the northwest limb of the megafold (Domains 3, 5) and northeast-southwest on the northeast limb of the megafold (Domains 4, 6). BEDDING-FOLIATION INTERSECTION LINEATION

Figure 4 Field photograph and photomicrographs of second generation foliations in the Texas beds, (a) Crenulation cleavage in Texas beds on the northeast limb of megafold. Marking pen is 14.5 cm long. View is to the south S 0 / 1 57°->255°; S 2 65°->040° (GR555480, Allora sheet), (b) Photomicrograph of kinking of dominant foliation (S 0/1 ) within Texas beds from the Cunningham Highway near Yurababa (GR419641, Inglewood sheet). The longer side of the figure is 6 mm across, (c) Photomicrograph of sinistral C shear of dominant foliation from Texas beds from Arcot Road junction on Texas-Inglewood road (GR212158, Texas sheet). The longer side of the figure is 6 mm across.

Intersection lineations were not often observed in the field exposures of the Texas beds. Laboratory calculation of the bedding-cleavage intersection lineation in the Texas beds for the different domains provides useful information on the megafold (Figure 7). Domains 1-4 all contain a common population of steeply, mainly northwest-plunging, intersection lineation (Figure 8 a d). The northwest-plunge of the dominant lineation is comparable to the derived fold axis from cleavage plots (Figure 12c). Some lineations are aligned subparallel (?Lj) and others subnormal (L2 or L 3 ) to the form lines defined by the dominant foliation (S1? which is subparallel to S 0 ) around the megafold. COMPARISON WITH THE COFFS HARBOUR BEDS AND GUNDAHL COMPLEX

Korsch (1973, 1975) and Fergusson (1982, 1984) identified two main deformation events in the along-strike

166

P. G. L E N N O X A N D P. G. F L O O D

sequences to the Texas beds, namely the Coffs Harbour beds and Gundahl Complex. The Gundahl Complex exhibits variably plunging, close, chevron to rounded ¥} folds with a steeply inclined to vertical axial surface and a spaced cleavage. The F2 folds are open, kink-like folds with rounded or angular hinges with steeply inclined axial surfaces and fold axes. The first deformation in the Coffs Harbour beds was characterised by a slaty,

disjunctive to transposition cleavage and related to variably plunging, tight to open, folds with or without an axial-surface cleavage (Fergusson 1982). The second deformation was characterised by close to open, dextral or sinistral kinks in the bedding and cleavage, moderate to vertical fold axes and a vertical axial surface. The cleavage morphology and limited mesoscopic folds and kink data in the Texas beds are similar, but

Figure 5 Map of the Texas megafold with representative cleavage trends in the Texas beds and Permian slope basins. This map clearly shows a dominant cleavage defining the megafold, another cleavage almost at right angles to this cleavage [i.e. northwesttrending on the northwest limb (Domains 3 & 5) and northeasttrending on the northeast limb (Domain 4)] and an almost east-west cleavage especially northwest of Stanthorpe in Domain 4. These cleavage trends are consistent with megafolding of an early bedding-subparallel cleavage and fanning of an megafold axial-surface cleavage (see the text for further discussion).

TEXAS MEGAFOLD, NSW not the same as the Coffs Harbour beds and Gundahl Complex. There are no exposures with both fold generations and there are limited data on cross-cutting foliations in the Texas megafold, so the temporal relationships are problematic. The dominant foliation in the Texas beds varies from a slaty cleavage to a transposition cleavage (Dl?) and the moderately to steeply plunging folds usually do not have an axialsurface cleavage (Forster 1991; Cohen 1991). The kinking of this foliation is consistent with the second generation structures observed in the Coffs Harbour beds.

Figure 6 Mostly contoured, stereographic projections of the poles to cleavage in five of the six domains which cover the Texas megafold. The dominant cleavage population is northwest—southeast striking, the secondary cleavage population is either n o r t h e a s t southwest striking (Domain 3) or east-west striking (Domains 2 & 4). The pole to the best-fit great-circle girdle for each plot is oriented: (a) 82°->274°; (b) 17°-»331°; (c) 87°-> 171°; (d) 76°->297°; (e) 40°->313°. The near-coincidence of the derived fold axis from the bedding data and the cleavage data is due to the dominant cleavage being subparallel to bedding prior to megafolding (cf. Figure 3).

167

Fergusson (1982) recognised that over large areas of the Coffs Harbour sequence the slaty to spaced cleavage is 10-20° clockwise divergent from bedding and locally folds are transected by the same cleavage. In the Texas megafold, a similar cleavage is 24° anticlockwise divergent from bedding in Domains 1 and 4 and clockwise divergent from bedding in Domain 2 (5°) and Domain 3 (27°). It is difficult to account for the difference in vergence between Domain 4 and the along-strike Coffs Harbour beds; but the vergence change between Domains 1 and 2 is consistent with the presence of the Texas megafold.

168

P. G. L E N N O X A N D P. G. F L O O D

PERMIAN SLOPE BASINS Permian slope basins are located on the northeast limb (Pikedale, Alum Rock), or near the hinge zone (Terrica, Silver Spur, Glenmore) of the megafold (Figure 2). No data were compiled from the Glenlyon or Mole Valley basins.

The Terrica basin (Figure 2) contoured plot of poles to bedding features a southwest concentration corresponding to northwest-southeast-striking beds dipping moderately northeast (Figure 9a; Cohen 1991; Flood 1991). The dominant cleavage strikes northwestsoutheast parallel to the structural grain (Figure 1 la) to observed mesoscopic folds and to the derived axial

! V V V V V V V V V V V V V V v>57 V Y

v v\v\;

/ 7 V V ^ V V * 7 V V

P

V V V Vs d V V V V p ^ f ^ f v

v_v

v/ft v v v " 'V ^ 37

V V V V V V K 7 V *

V V V V / 15 , 1 1 30 v v V V V V V v \ v

Figure 7 Map of the beddingcleavage intersection lineations in the Texas beds and Permian slope basins in the megafold. There are lineations parallel to the megafold shape defined by bedding (cf. Figure 2) and the dominant cleavage (cf. Figure 5). The presence of minor lineations oriented northwestsoutheast on the northwest limb and northeast-southwest on the northeast limb and some approximately east—west oriented lineations is consistent with a secondary foliation at a high angle to the dominant bedding-subparallel cleavage.

TEXAS MEGAFOLD, NSW surface from the plot of poles to the bedding. The bedding at Alum Rock consists of a single point population (Figure 9b; Berg 1973; Cohen 1991). Cohen (1991) identified two cleavages from the Permian; one striking north-south (S2) and another striking northwest-southeast (S b subparallel to bedding) (Figure lib). The bedding in the Silver Spur basin contains three main populations corresponding to north-south, northwest-southeast and minor almost east-weststriking bedding (Figures 9c, 10). Derived fold axes from bedding and cleavage in the Permian sequences are moderately northwest (F 2 ) and north-northeast (Fj?)

Figure 8 Mostly contoured, stereographic projections of the bedding-cleavage intersection lineations in the six domains of the Texas megafold. Domains 1 - 4 and to a lesser extent 5 are characterised by steeply mainly northwest-plunging lineations. This is in contrast to the fold axis derived from the bedding data which plunges steeply south-southeast, due mainly to the preponderance of Domain 4 data which overall have an overall south-southeast-derived fold axis (Figure 3d). The average intersection lineations in some domains are: (b) 85°->346°; (c) 7 5 ° ^ 2 3 9 ° ; (d) 42°->314°.

169

plunging respectively (Figures 9c, 1 lc). The Silver Spur basin sequence exhibits a very-well-developed cleavage, with macroscopic and mesoscopic folds trending dominantly northwest-southeast and rarely northeastsouthwest or east-west (Figure 10; Oxley 1972). There are two comparable cleavage trends (northwestsoutheast and north-northeast-south-southwest) but no evidence of an east-west cleavage (Figure 11c). Whereas folding in the Permian sequence is gently plunging, close to open in style with moderate to steeply dipping axial surfaces, that in the Texas beds is moderate to steeply plunging with a vertical axial surface (F2 northwest to north-northwest;

170

P. G. L E N N O X A N D P. G. F L O O D

Fj east-northeast). Korsch and Harrington (1981) considered that the Texas beds were deformed at least three times prior to Permian deposition and at least twice subsequently. No data are presented to support their contention. The bedding at Pikedale is less steeply dipping than the Texas beds within this domain, but overall defines a similar fold axis (Figure 9d, cf. Figure 3d). The Pikedale basin contains the same three cleavages identified from within the Texas beds from Domain 4: (i) northwest-southeast (S l5 bedding subparallel); (ii) north-northeast-south-southwest (S 2 ); and (iii) eastwest (S 3 ; Forster 1991) (Figure lid). Folding in these Permian sediments is gently plunging, open with a

moderately inclined axial surfaces, whereas folding in the Texas beds is steeply plunging to vertical on steeply inclined to vertical axial surfaces (Foster 1991). Non-bedding foliations in the Glenmore basin were not observed, so analysis of the structures relies upon analysis of bedding (McCarthy 1971; Moore 1982; Boxall 1972) (Figure 9e). The outcrop pattern in some parts of the Glenmore basin is consistent with east-west folding. The Texas beds have folds of similar orientation to the overlying Permian volcanics of the Wandsworth Volcanic Group. Texas beds folds are asymmetric, east-verging, shallowly, southeastplunging, open folds; whereas the Permian basin contain shallowly, northwest-plunging folds.

Terrica

Alum Rock

Silver Spur

Pikedale

120

Glenmore Station

Figure 9 Contoured, stereographic projections of the poles to bedding from the Permian slope basins within the Texas megafold. New data were collected in the Terrica, Glenlyons and Glenmore Station basins, whilst appropriate data was extracted from Honours theses maps for Alum Rock (Berg 1973), Silver Spur (Oxley 1972), Pikedale (Forster 1991) and Glenmore Station (Moore 1982). The overall orientation of bedding reflects in large part the position and outcrop pattern of the particular Permian slope basin within the megafold. Thus the thin, elongate, regional strike-parallel Alum Rock basin on the northeast limb of the megafold yields a point population, whereas the megafold hinge zone Terrica and Silver Spur basins yield more complicated patterns within which the broad megafold trend can be discerned. The pole to the bestfit great-circle girdle for each plot is oriented: (a) 44°->346°; (b) 3°-»321°; (c) 61°->321°; (d) 22°->172°; (e) 23°->086°.

TEXAS MEGAFOLD, NSW Table 2 summarises the orientation of the dominant, secondary and minor cleavage populations within the Permian basins of the megafold. The northwestsoutheast trend of the dominant cleavage at Silver Spur and Glenmore is parallel to the secondary foliation formed during megafolding, whereas elsewhere in the Permian basins this cleavage is parallel to the dominant, pre-megafold foliation. DISCUSSION Comparisons of foliations between the Texas beds and Permian bodies The bedding within the Texas beds is notably more steeply dipping (Figure 12d; Table 1) compared with the bedding in the Permian (Figure 12b). The derived fold axes from the bedding plots plunges very shallowly (4°-> 155°, Permian) to steeply (76°H> 164°, Texas beds) southeast.

Figure 10 A modified version of Oxley's (1972) map of the Silver Spur area near Texas. N o t e the f o l d i n g of the Permian—Texas beds unconformity by a northwest-plunging, macroscopic fold labelled A and the southeast-plunging, macroscopic fold labelled B in the underlying Texas beds to the east. These folds are oriented parallel to the s e c o n d a r y megafold-related foliation. The northeast-trending foliation dominant near the Silver Spur mine would correspond to the dominant bedding-subparallel foliation on the northwest-limb of the megafold.

171

The plots of cleavage contain two common point populations corresponding to steeply southwest-dipping, northwest-striking and northeast-striking cleavage (Figure 12a, c; Table 1). The Texas beds have a minor population of east-west striking cleavage. This eastwest cleavage corresponds to that measured in the Texas area near the hinge zone of the megafold. The spread of the dominant cleavage readings around great circles on these plots is consistent with refolding of the cleavage. The map of the cleavage pattern and field observations indicate that the bedding-subparallel cleavage in the Texas beds is folded about the megafold. Thus any derived fold axis would in part correspond to the megafold axis. BEDDING-FOLIATION LINEATIONS AND MESOSCOPIC FOLDS

The limited bedding-foliation intersection lineation data for the Permian bodies is more gently northwest-

172

P. G. L E N N O X A N D P. G. F L O O D

Table 1 Orientations of different foliation populations in Texas beds from Figure 12.

Frequency

Dominant population

Secondary population

Minor population

Bedding

83°->249°

84°->314°

82°->358°

Cleavage

87°->245°

90°-»288°

71°->358°

plunging compared with the Texas beds lineation data (Figure 13a, c) reflecting the deposition of Permian onto already steeply dipping Texas beds and their later megafolding. The limited mesoscopic fold data from the Permian sequence correspond to shallow to moderately plunging folds consistent with the intersection lineation data (Figure 13b, cf. Figure 13a). The plot of all mesoscopic folds observed in the field in the Texas beds is very similar to the intersection lineation data with the average mesoscopic fold orientation being very similar to the overall megafold orientation (Figure 13d, cf. Figure 13c).

Timing of megafolding Various ages have been suggested for megafold formation. They vary from Late Carboniferous (Murray et al. 1987), late Early Permian (Korsch & Harrington 1987; Korsch in Murray et al. 1987) and pre-Late Permian (Flood & Fergusson 1982; Fergusson & Flood 1984) to Late Permian (Collins et al. 1993). Recent palaeontological investigations by Briggs (in press) has dated the sedimentary sequence in outcrops of Permian in the Silver Spur basin as late Sakmarian to middle Artinskian. Since both the Texas beds and Permian in this area have been folded and cleaved during the megafolding event, the timing of this event must be post-Artinskian. Aubourg et al. (1994) and C. Klootwijk (pers. comm. 1996) considered on the basis of palaeomagnetic studies of the Alum Rocks volcanics that they were rotated about 40° prior to megafolding and 80° after megafolding. Since D. J. C. Briggs (pers. comm. 1996) reported an Asselian to Sakmarian fauna for Alum Rock Permian and Roberts et al. (1995, 1996) assigned an age of about 294 Ma for volcanics in this sequence, this provides a constraint on the beginning of megafold formation. Thus megafolding must have already started prior to 294 Ma and finished after the middle Artinskian, a period of about 15 million years.

Figure 11 Uncontoured and contoured, stereographic projections of the poles to cleavage from the Permian slope basins within the Texas megafold. Overall the derived fold axis plunges moderately to steeply in the n o r t h w e s t e r n quadrant similar to the megafold trends. This would be expected if the d o m i n a n t c l e a v a g e in the Permian slope basins was folded during megafolding. The pole to the best-fit great-circle girdle for each plot is p l u n g i n g : (a) 75°->288°; (b) 31°->337°; (c) 73°-»026°; (d) 82°->349°.

TEXAS MEGAFOLD, NSW

173

Table 2 Orientation of cleavage populations in Permian basins, Texas megafold.

Permian slope basin Pikedale Alum Rock Terrica Silver Spur Glenmoret

Dominant cleavage population

Secondary cleavage population

Minor cleavage population

Derived fold axis*

NE-SW N-S N-S to NNE-SSW (NE-SW)

E-W E-W Texas - late (E-W)

22°->172° n/a 44°->346° 61°-»321° 15°->310°

NW-SE NW-SE NW-SE NW-SE Texas - early (NW-SE)

*

from the plot of poles to bedding except in the case of Glenmore when the pole to the great circle girdle of the dominant population of poles was used to derive the fold axis, t () assuming cleavage is axial surface to folds defined from plots of poles to bedding.

This timing is in agreement with that suggested by Korsch and Harrington (1987), Flood and Fergusson (1982) and Fergusson and Leitch (1993). Structural synthesis and model The dominant foliation in the field and thin-section is Permian

Texas

Figure 12 Contoured stereographic projections of the poles to the total bedding and cleavage datasets for the Texas beds and Permian slope basins within the Texas megafold. Note the Texas beds contain somewhat more steeply dipping cleavages and bedding than the Permian reflecting their prePermian deformation. The pole to the best-fit great-circle girdle for each plot is oriented: (a) 83°->335°; (b) 4°->155°; (c) 82°->316°; (d) 76°-+164°.

(c)

beds

subparallel to bedding and is clearly folded around a structure with an axial-surface trace oriented overall north—south but with northwest- and north-northeasttrending sections (Flood & Fergusson 1982; Fergusson & Flood 1984). This cleavage is considered to represent a foliation formed probably during subduction-accretion (Sj; Figure 14a). Lineament and palaeomagnetic studies Cleavage

Cleavage

Permian

Texas

beds

Bedding

Bedding

174 P. G. L E N N O X AND P. G. F L O O D (Vinayan et al. 1993; Aubourg et al. 1994) proposed that this macroscopic fold is more complicated with some sections displaying an axial-surface trace northsouth, whilst other sections are oriented north-northwest to north-northeast. The later, patchily developed, cross-cutting foliations (S ) are crenulation cleavages, kinks or C shears and are fanned around the megafold. They probably developed during the early stages of the megafold formation and have subsequently been fanned as the megafold tightened (Figure 14 b, c). The east-west oriented foliation in the Permian outliers and within the Texas beds is poorly developed and although it cannot be shown to overprint the other foliations, analysis of folding in the Texas area suggests it was the last cleavage (S ) and formed after the megafolding event in this area (Figure 14d). 2

3

Tectonic model

The New England region was part of a east-facing, almost meridional subduction system, with the Texas beds considered to represent part of an accretionsubduction complex in the Late Devonian to Early Carboniferous (Murray et al. 1987). Changes in plate vectors at the end of the Carboniferous led to oblique subduction and backstepping of the system so that the New England area was affected by a dextral transpressional system for a short interval (Figure 14b; Murray et al. 1987). This was resolved in the Texas area

by the formation of a foliated, Early Permian megafold. This foliation in the developing megafold was oriented northwest-southeast (Figure 14b; S ). Continuing tightening of this structure led to the fanning of this foliation about the megafold (Figure 14c). The presence of both northeast-southwest, east-west and northwest-southeast oriented foliations in the Pikedale basin suggests that megafolding occurred after the deposition of this basin. Folding of the unconformity separating Permian and Texas beds at Silver Spur occurred during development of the megafold. Thus megafold formation has deformed late Sakmarian to middle Artinskian Silver Spur rocks (Briggs in press) indicating an upper bound for the period of megafold development. The late Asselian—Sakmarian Alum Rock basin was deposited upon an already partly rotated Texas bed sequence indicating megafolding started pre-late Asselian (C. Klootwijk pers. comm. 1996). Megafolding must have been initiated pre-late Asselian and continued until at least post-middle Artinskian to account for the geological, faunal and palaeomagnetic data. The event is Early Permian and not Late Carboniferous as suggested by numerous other workers. Therefore any tectonic model for the southern New England Orogen must include Early Permian subduction. Perhaps megafold formation represents the termination of subduction prior to backstepping some 250 km eastward in the New England area (Collins et al. 1993). 2

Figure 13 Contoured and uncontoured stereographic projections of the beddingcleavage intersection lineation data and mesoscopic fold axes data for the Permian slope basins (a, b) and Texas beds (c, d) respectively. The intersection lineations and fold axes of the Texas beds are more steeply plunging compared with the Permian fault block data. This reflects the pre-Permian tilting and mesoscopic folding (?) of the Texas beds. The dominant Permian lineation plunges 28°->315° compared with the dominant Texas bed lineation plunge which is almost vertical (88°-»000°) and average Texas bed mesoscopic fold which plunges 85°->162°.

TEXAS MEGAFOLD, NSW Strike-slip ^/faults

(a)

175

(b)

A I £ 5

'l 1 ..'J-

UJ

I

O

"(••IN

LL

'1 | 1

I I .

S-| subduction - accretion foliation

/

f

Sisubduction

I

-

accretion

foliation

Figure 14 A model for the development of the main foliations in the Texas megafold. (a) Late Devonian (?) to Early Carboniferous subduction with an east-verging complex with subduction-accretion foliation (Sj). (b) Permian dextral transpression of eastern Australia with b a c k s t e p p i n g of the subduction zone leading to megafold initiation and foliation development (S 2 ). (c) Fanning of the megafold-related cleavage (5 2 ) during further tightening of the megafold. (d) Possible latestage north-south shortening leading to the development of an east-west oriented cleavage (53).

ACKNOWLEDGMENTS

REFERENCES

We are indebted to students who carried out mapping projects within the Texas megafold. Amongst these are H. Berg from the University of Queensland; A. Boxall, H. Butler, B. McCarthy, T. Moore, G. Oxley and D. Warner from the University of New England; and L. Cohen and P. Forster from the University of New South Wales. Bill Collins, Chris Fergusson, Russell Korsch and Dick Glen are thanked for constructive comments on various versions of this paper. Field and laboratory work was financed by two Special Research Grants, Faculty of Applied Science, University of New South Wales (PGL) and an ARC Grant 940175 (PGL, PGF and Terry Farrell).

AITCHISON J. C. & FLOOD P. G. 1990. Early Carboniferous

radiolarian ages constrain the timing of sedimentation within the Anaiwan terrane, New England orogen, eastern Australia. Neues Jahrbuch fur Geologie und Palaontologie Abhandlungen 180, 1-19. AITCHISON J. C . , FLOOD P. G . & SPILLER F. C . P.

1992.

Tectonic setting and palaeoenvironment of terranes in the southern New England orogen, eastern Australia as constrained by radiolarian biostratigraphy. Palaeogeography, Palaeoclimatology, Palaeoecology 94, 31—54. AUBOURGC., KLOOTWIJKC. & KORSCH R . J. 1 9 9 4 . M a g n e t i c

Fabric and Palaeomagnetic studies in the Texas and Coffs Harbour Blocks, New England Orogen. Australian Geological Survey Organisation Record 1994/58.

176

P. G. LENNOX AND P. G. FLOOD

BARNES P. M. & KORSCH R. J. 1990. Structural analysis of a

middle Cretaceous accretionary wedge, Wairarapa, New Zealand. New Zealand Journal of Geology & Geophysics 33, 355-375. BERG H. R. 1973. Geology of the Alum Rock area near Stanthorpe, Southeast Queensland. BSc (Hons) thesis, University of Queensland, Brisbane (unpubl.). BOXALL A. M. 1972. The Geology of an area to the SouthEast of Bonshaw, North-eastern New South Wales. BSc (Hons) thesis, University of New England, Armidale (unpubl.). BRIGGS D. J. C. Permian Productidina and Strophalosiidina from the Sydney-Bowen Basin and the New England Orogen: systematics and biostratigraphic significance. Association of Australasian Palaeontologists Memoir (in press).

New England Fold Belt. In: Herbert H. K. & Rynn J. M. W. eds. 1984 Field Conference, pp. 1-19., Geological Society of Australia, Queensland Division, Brisbane. FORSTER P. 1991. Geology of a transect through the Pikedale Area, Border Rivers Area, Queensland. BSc (Hons) thesis, University of New South Wales, Sydney (unpubl.). GRAY D. R. 1978. Cleavages in deformed psammitic rocks from southeastern Australia: their nature and origin. Geological Society of America Bulletin 89, 577—590. KJYOKAWA S. 1992. Geology of the Idonnappu Belt, Central Hokkaido, Japan: evolution of a Cretaceous accretionary complex. Tectonics 11, 1180-1206. KORSCH R. J. 1973. Structural analysis of the Palaeozoic sediments in the Woolgoolga District, North Coast, New South Wales. Journal and Proceedings of the Royal Society of New South Wales 106, 98-103.

BUTLER H. R. 1974. The Geology of the Stanthorpe-Leyburn Mineral District, Queensland and New South Wales. BSc (Hons) thesis, University of New England, Armidale (unpubl.). COHEN L. 1991. The Geology of the Terrica to Stanthorpe Adamellite Transect, Border River Area, Queensland. BSc (Hons) thesis, University of New South Wales, Sydney (unpubl.).

KORSCH R. J. 1975. Structural analysis and geological evolution of the Rockvale-Coffs Harbour region, northern New South Wales. PhD thesis, University of New England, Armidale, (unpubl.).

COLLINS W . J., OFFLER R . , FARRELL T . R. & LANDENBERGER

fragmentation and deformation of terranes in the New England Orogen, Eastern Australia. In: Leitch E. C. & Scheibner E. eds. Terrane Accretion and Orogenic Belts, pp 129-139. American Geophysical Union Geodynamics Series 19.

B. 1993. A revised Late Palaeozoic-Early Mesozoic tectonic history for the southern New England Fold Belt. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 69-84. Department of Geology & Geophysics, University of New England, Armidale. CROSS K . C . , FERGUSSON

C. L. & FLOOD

P. G .

1987.

Contrasting structural styles in the Paleozoic subduction complex of the Southern New England Orogen, Eastern Australia. In: Leitch E. C. & Scheibner E. eds. Terrane Accretion and Orogenic Belts, pp. 83-92. American Geophysical Union Geodynamics Series 19. FERGUSSON C. L. 1982. Structure of the Late Palaeozoic Coffs Harbour Beds, northeastern New South Wales. Journal of the Geological Society of Australia 29, 25-40. FERGUSSON C. L. 1984. The Gundahl Complex of the New England Fold Belt, eastern Australia: a tectonic melange formed in a Palaeozoic subduction complex. Journal of Structural Geology 6, 257-271. FERGUSSON C. L. & FLOOD P. G. 1984. A late Palaeozoic

subduction complex in the Border Rivers area of southeast Queensland. Proceedings of the Royal Society of Queensland 95, 47-55. FERGUSSON C. L. & LEITCH E. C. 1993. Late Carboniferous to

Early Triassic Tectonics of the New England Fold Belt, eastern Australia. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 53—59. Department of Geology & Geophysics, University of New England, Armidale.

KORSCH R . J. & HARRINGTON H . J. 1 9 8 1 . S t r a t i g r a p h i c a n d

structural synthesis of the New England Orogen. Journal of the Geological Society of Australia 28, 205-226. KORSCH R. J. & HARRINGTON H. J. 1987. Oroclinal bending,

KORSCH R . J., JOHNSTONE D . W . & W A K E - D Y S T E R K . D .

1997. Crustal architecture of the New England Orogen based on deep seismic reflection profiling. Geological Society of Australia Special Publication 19, 29-51. LEITCH E. C. 1974. The geological development of the southern part of the New England Fold Belt. Journal of the Geological Society of Australia 21, 133—156. LUCAS K. G. 1960. The Texas area. Journal of the Geological Society of Australia 7, 229-235. MCCARTHY B. H. 1971. The Permian Fauna in the 'Glenmore' Homestead area, Bonshaw, NSW. BSc (Third Year Project) University of New England, Armidale (unpubl.). MOORE T. W. 1982. The stratigraphy and setting of a Late Permian volcanic sequence in northern New England. BSc (Hons) thesis, University of New England, Armidale (unpubl.). MURRAY C . G . , FERGUSSON C . L., FLOOD P. G . , WHITAKER W .

G. & KORSCH R. J. 1987. Plate tectonic model for the Carboniferous evolution of the New England Fold Belt. Australian Journal of Earth Sciences 34, 213—236. OLGERS F. & FLOOD P. G. 1974. Palaeozoic geology of the

Warwick and Goondowindi 1: 250 000 sheet areas, Queensland and New South Wales. Bureau of Mineral Resources Report 140.

FLOOD P. G. 1991. Geology of the Treverton Creek Salinity Project (Connors Creek Catchment). Inglewood Land Care Committee (unpubl.).

OXLEY G. W. 1972. The stratigraphy and economic geology of the Silver Spur area, southern Queensland. BSc (Hons) thesis, University of New England, Armidale (unpubl.).

FLOOD P. G . & FERGUSSON C . L. 1 9 8 2 . T e c t o n o - s t r a t i g r a p h i c

ROBERTS J., CLAOUE-LONG J. C . , JONES P. J. & FOSTER C . B .

units and structure of the Texas-Coffs Harbour Region. In: Flood P. G. & Runnegar B. eds. New England Geology, pp. 71-78. Department of Geology, University of New England and AHV Club, Armidale.

1995. SHRIMP zircon age control of Gondwanan sequences in Late Carboniferous and Early Permian Australia. In: Dunnay R. E. & Hailwood E. A. eds. Nonbiostratigraphical Methods of Dating and Correlation, pp. 145-174. Geological Society of London Special Publication 89.

F L O O D P. G . & FERGUSSON C . L .

1984. T h e

geological

development of the northern New England Province of the

TEXAS MEGAFOLD, NSW

111

1996.

VINAYAN P. K . , TAYLOR G . R . , BALIA L. M . & LENNOX P. G .

SHRIMP zircon dating of the Permian System of eastern Australia. Australian Journal of Earth Sciences 43,

1993. The integration of vector and raster-based remotely sensed data for geological exploration. In: Masters E. G. & Pollard J. R. eds. Advanced Remote Sensing Conference, vol. 1, pp. 349-358. School of Geography, University of New South Wales, Sydney.

ROBERTS J., CLAOUE-LONG J. C . &

FOSTER C . B .

401-421.

VAN NOORD K. A. A. 1993. Structure and stratigraphy of the Early Devonian Silverwood Group: implications for the tectonic development of the New England Orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 587-589. Department of Geology & Geophysics, University of New England, Armidale.

WELLMAN P . , WILLIAMS J.

W.

&

MAHER

A. R.

1994.

Interpretation of gravity and magnetic anomalies in the Clarence-Moreton Basin region. Australian Geological Survey Organisation Bulletin 241, 217-229. (Received 24 April 1996; accepted 2 January 1997)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 178-187.

Lithostratigraphy of the Gamilaroi terrane, upper Barnard region, northeastern New South Wales J. M. C. STRATFORD1 AND J. C. AITCHISON2 1

1 Department of Geology and Geophysics, University of Sydney, NSW 2006, Australia. Department of Earth Sciences, University of Hong Kong, Pokfulam Road, Hong Kong, China.

A succession of approximately 4200 m of sedimentary and volcanic rocks within the Gamilaroi terrane in the upper Barnard region of northeastern New South Wales contains five mappable units. The Silurian-Lower Devonian Pitch Creek Volcanics lies at the faulted base of the succession and consists primarily of felsic volcanic rocks and associated tuffs. It is overlain by volcaniclastic sandstone and fine-grained tuffs of the Lower to Middle Devonian Frog Hollow Formation, which envelop the laterally discontinuous Bennys Tops Limestone Member. Overlying this unit are volcaniclastic conglomerate, sandstone and mudstone of the Captain Rocks Formation. The uppermost units in the upper Barnard region are Middle Devonian fine-grained volcaniclastic sandstone, mudstone and tuffs of the Curricabark Formation and intercalated massive and pillow basalt and dolerite of the Folly Volcanics. The stratigraphic sequence is broadly similar to that of the Gamilaroi terrane in the Tamworth-Nundle area. Comparison can also be made with the general composition and distribution of lithologies in other areas of Gamilaroi terrane outcrop including the Hastings Block and northern New England Orogen. Unequivocal lithological correlations with existing lithostratigraphic units are problematical. Differences in the stratigraphic successions throughout the Gamilaroi terrane are interpreted as being related to original lithofacies variability inherent in intra-oceanic island arc environments. Key words: Devonian, Gamilaroi terrane, lithostratigraphy, New England Orogen

INTRODUCTION

The Gamilaroi terrane (Flood & Aitchison 1988, 1992), the westernmost terrane in the southern New England Orogen, has recently been recognised as an example of an ancient intra-oceanic island arc rift sequence (Aitchison & Flood 1995; Stratford & Aitchison 1996). Rocks correlated with this tectonostratigraphic unit can be recognised from as far north as Rockhampton in the northern New England Orogen to south of Taree near the New South Wales coast in the southern New England Orogen (Stratford et al. 1996). In the southern part of the southern New England Orogen (Figure 1), Gamilaroi terrane rocks crop out in a series stream basin exposures separated by remnants of Tertiary basalt flows. From north to south these are referred to as the Barry, Glenrock and Pigna Barney areas and are named after the main 1:25 000 map sheets. Together, these three areas are referred to as the upper Barnard region after the deeply incised Barnard River. In this paper we describe the local lithostratigraphic succession within the Gamilaroi terrane. Most previously published studies of Gamilaroi terrane rocks in the southern New England Orogen have been undertaken in the Attunga—Tamworth-Nundle region 80-100 km northwest of Glenrock. In this region Benson (1913a, b, c; 1915a, b; 1918a) named and mapped rocks of the Tamworth Series (revised to the Tamworth Group by Voisey 1958). In subsequent lithostratigraphic studies a number of formations within

the Tamworth Group (Crook 1961; Cawood 1983) were defined. The general appearance and approximate stratigraphic position of many of the volcanic and volcaniclastic rocks in the Tamworth Group are similar to those observed in the upper Barnard region. However direct lithostratigraphic correlation between these two areas is difficult (Stratford & Aitchison 1996) because of inferred lateral facies variation of key lithologies and discontinuous exposure between Nundle and the upper Barnard region. Potential lithostratigraphic correlations between the Tamworth-Nundle and upper Barnard region are discussed later in this paper. All grid references are from the Central Mapping Authority New South Wales 1: 25 000 topographic map sheets Barry (9134-I-N), Curricabark (9234-4-S), Glenrock (9134-I-S), Nundle (9135-111-S) and Pigna Barney (9234-III-N). Where referred to in the text, the title of the relevant map sheet is given at the end of the grid reference notation. AGE OF THE GAMILAROI TERRANE

Volcaniclastic rocks of the Murrawong Creek and Pipeclay Creek Formations contain Cambrian trilobite, inarticulate brachiopod and conodont faunas (Cawood 1976; Stewart 1995) and possibly form a localised ancient basement to the predominantly mid-Palaeozoic rocks of the Gamilaroi terrane. Their outcrop has so far been reported only from the Nundle district (Cawood 1976, 1983). Of the rocks which comprise the bulk of

GAMILAROITERRANE STRATIGRAPHY, NSW the terrane the oldest in situ igneous rocks are Silurian felsic volcanic rocks (Kimbrough et al 1993; Aitchison & Ireland 1995). Overlying sedimentary rocks are mainly Early to Middle Devonian (Hill 1942; Pedder 1967; Mawson & Talent 1994; Furey-Greig 1995; Mawson et al 1995) and possibly as young as Frasnian in age (McMinn 1982). Traditionally, ages assigned to the sedimentary sequence have been based on those of limestone bodies scattered throughout the sequence. Radiolarians are abundant in tuffs and tuffaceous mudstones, and as limestones are often not present in Gamilaroi terrane exposures, they have the potential to provide increased age control throughout the terrane. Since the pioneering work of Hinde (1899) recent work has established a number of radiolarian assemblages (Aitchison 1988, 1993; Aitchison et al 1992; Aitchison & Stratford 1994; Stratford & Aitchison 1994a, 1997) which indicate an Early (?Pragian) to Middle Devonian age range for the bulk of the sedimentary rocks in the Gamilaroi terrane. STRATIGRAPHY The geology of the upper Barnard region was first studied by Benson (1918b). Apart from the work of Offler (1982), most subsequent geological mapping in the region has been limited to unpublished student theses. These sources (Mayer 1972; Cross 1983; Weekes 1984; Allan 1987) have provided lithostratigraphic terms that have migrated to the formal literature

179

(Morris 1988; Kimbrough et al 1993; Dongal 1995). Detailed mapping undertaken at Glenrock (Stratford 1993, 1995; Stratford & Aitchison 1995, 1996) (Figure 2), combined with extensive reconnaissance geological surveys by the authors (Aitchison et al 1997) have enabled a succession of distinctive mappable lithostratigraphic units to be recognised throughout the region. This paper describes the lithostratigraphy of the Gamilaroi terrane in the upper Barnard region and formalises previous nomenclature. Details of Gamilaroi terrane radiolarian biostratigraphy (Stratford & Aitchison 1994a, b, 1997) and sedimentology (Stratford & Aitchison 1996) at Glenrock are published elsewhere. Pitch Creek Volcanics (revised definition) The representative section of the Pitch Creek Volcanics in the Pigna Barney River originally proposed by Mayer (1972) is adopted as the type section. The name of the unit is taken from Pitch Creek, a tributary of the Pigna Barney River. The type section extends from the base at GR 697818 downstream to GR 717808 (Pigna Barney). An additional reference section is present in Orham Creek from the faulted base at GR 569949 to GR 561946 (Glenrock). The maximum thickness of the Pitch Creek Volcanics in the type section was estimated at 2200 m (Mayer 1972) whereas the thickness of the reference section is approximately 1200 m (Figure 3). The base of this unit is locally faulted against rocks of

Northern NEO Southern NEO 0 km 50

relative location of geological maps shown on Fig. 2.

V77Z7777777777T

Peel-Manning Fault system 31°

151 "WE,

Barry. Glenrock" J• Pigna Barney. B u n d o ^ W

Port Macquarie

cover strata

151°

•

152°

M s m 153°E

TAMWORTH BELT

Undifferentiated Tamworth Belt strata (Upper Devonian-Carboniferous) Gamilaroi terrane (Silurian-Upper Devonian)

HASTINGS BLOCK

Undifferentiated strata - includes Gamilaroi terrane equivalent rocks (Middle Devonian-Carboniferous)

Undifferentiated rocks of the New England orogen Figure 1 (a) Tectonostratigraphic map of the southern New England Orogen showing the location of the places mentioned in the text (after Aitchison & Flood 1995). (b) Location of geological maps shown on Figure 2.

180 J. M. C. S T R A T F O R D AND J. C. A I T C H I S O N either the Weraerai or Djungati terranes (Flood & formation is most likely a function of original irreguAitchison 1988). A transition from lavas in the lower larities in the ground surface surrounding volcanic parts to predominantly pyroclastic rocks in the upper edifices. Large portions of the formation are poorly parts is typical of most sections. The upper contact of exposed and defy straightforward mapping. It is the Pitch Creek Volcanics is locally faulted or possible that blocks of volcanic rocks up to several conformable, and is marked by the appearance of a hundred cubic metres in size, intermixed with other significant thicknesses (> 15 m) of bedded volcaniclastic lithologies, represent remnants of widespread pyrosedimentary rocks. Upper levels of the Pitch Creek clastic debris avalanche deposits that developed in Volcanics may possibly be intercalated with response to the collapse of nearby volcanic edifices. volcaniclastic sandstone of the Frog Hollow Formation The Pitch Creek Volcanics is composed of a varied along strike. Relief on the upper surface of the suite of predominantly volcanic and minor sedimentary Barry

Tertiary cover rocks Manning facies Anaiwan terrane Djungati terrane Weraerai terrane Location of sections

Gamilaroi terrane Folly Volcanics Curricabark Formation Captain Rocks Formation Frog Hollow Formation Pitch Creek Volcanics

Pigna Barney Bennys Tops Limestone Member

151°40'E 1

Figure 2 Geological maps showing the distribution of Gamilaroi terrane stratigraphic units and approximate location of type and reference sections in the Barry, Glenrock and Pigna Barney areas. Relationships between the adjacent terranes in the upper Barnard region are seen in greater detail in Aitchison et al (1997).

GAMILAROITERRANE STRATIGRAPHY, NSW rocks. Felsic lavas and tuffs are most common, forming from 40 to 50% of the exposures, although this high figure may indicate the relative resistance of these lithologies to weathering compared with associated basaltic rocks. Autoclastic volcanic breccia and very

Glenrock

Barry

Nundle

approximate position of stages

181

large-scale volcanic debris avalanche deposits are also present. Felsic volcanics range in colour from pale green to pale grey in fresh exposures, to dirty brown-red when weathered. Most felsic lavas are porphyritic with quartz phenocrysts up to 5 mm across. Crystal tuffs Pigna Barney

^iooom/ Parry Group

Famennian

nature of contact uncertain

nature of contact uncertain

Frasnian Folly Volcanics Folly Volcanics

Curricabark Formation

Givetian Yarrimie Formation

Eifelian

^ Loomberah Limestone Silver Gully Formation

\

I Captain Rocks Formation

•v-V-V-v V'

issa

CL/"

X s *

/

>> /

cza

V V * }I \

V / V/ i Wogarda Argillite Emsian

IHB

Northcotte Formation

m Frog Hollow Formation

IS3

Vv V V VDrik Drik Formation Y y • M i l

Pragian

Nemingha Limestone

VV >Tvv Copes Creek v v vV V V VKeratophyre member

Silurian

B/7^ /V A VV V> V <S V V V > VV VV VV V> <S V VV V VPitch Creek VV V V V V Vs> <S V V V Volcanics V V V V> ^VVVVVVV

VVVV , VVVV v^ •V VVVVV VV V V VVVH - v V v vH V VWV V W V VV VVVVV *V VVVv ^V VV VV VVVV^

Lithologies

VV f V s Vv V S 'V' V VVV VV> VVVVVVVV ^ /VVVVVVV V>V X V V V V V V V SVVVVVVVV •VVVVVVVV y v v v ^ VVVV S V V V > V /V VVVV VVV> /VV V V V > *

massive and pillowed basalts unconformity

Ordovician

ZZZZ1

Haedon Formation

dolerite and gabbro

unconformity

Cambrian

s s

Pipeclay Creek Formation Murrawong Creek Formation

predominantly coarse-grained

•"•V'J volcaniclastic sedimentary rocks limestone fine-grained laminated volcaniclastic sedimentary rocks interbedded tuffs, mudstone and sandstone interbedded sandstone mudstone and conglomerate

VV VV basaltic breccia, tuffs and VVVV

f

felsic volcanic rocks

Figure 3 Composite columns showing the generalised lithostratigraphy of the Gamilaroi terrane at different locations in the upper Barnard region compared with the Tamworth Group (Crook 1961; Cawood 1983) in the Nundle area. The approximate age of the stratigraphic units is also given.

182

J. M. C. S T R A T F O R D A N D J. C. A I T C H I S O N

commonly associated with the lavas are characterised by coarse to very coarse sand-sized euhedral feldspar crystals and quartz grains. Fine-grained felsic tuffs and tuffaceous mudstone are present throughout the unit, especially at Barry. Felsic intrusive bodies within the formation (GR 412067, Barry) are rare. Conglomerate, thinly bedded intercalated volcaniclastic sandstone, mudstone and basalt are minor lithologies in the Pitch Creek Volcanics. Conglomerates are typically polymict with angular basalt, dolerite and minor andesite clasts set in a characteristically red hematitic matrix. Blocks of thin-bedded black mudstone, grey tuffaceous mudstone and lenses of grey recrystallised limestone, with maximum dimensions up to 20 x 10 m, are rarely associated with the breccias. Mineralised (Cu, Au) dolerites (Gilligan & Brownlow 1987) are present at Barry. Discontinuous zones of serpentinite throughout the Pitch Creek Volcanics commonly mark fault zones which cut through the unit. U—Pb dates from felsic volcanic rocks in the Pitch Creek Volcanics at Glenrock (zircon SHRIMP 421 ±10 Ma: Aitchison & Ireland 1995) and Pigna Barney (mass spectrometry 436±9 Ma and 411 ±5 Ma: Kimbrough et al. 1993) give a maximum Silurian age for this formation. Identifiable fossils have not yet been recovered from this formation which may range into the Lower Devonian. Frog Hollow Formation (new formation) The type section for the Frog Hollow Formation is located along the Pigna Barney River extending from the base at GR 717808 downstream to GR 746811 (Pigna Barney). The Frog Hollow Formation includes the Bennys Tops Limestone Member and is named from Frog Hollow, a tributary of the Pigna Barney River. At Glenrock and Barry the Frog Hollow Formation is present within two blocks (northeastern and southwestern blocks) on either side of the PeelManning Fault System. Sections within these blocks can be correlated biostratigraphically with the complete section at Pigna Barney and are considered to represent an essentially continuous stratigraphic section (Stratford & Aitchison 1996). A composite reference section at Glenrock is divided by the Peel—Manning Fault System. The lower part of the reference section is located along Orham Creek in the northeastern block at GR 557945 upstream to GR 547938 (Glenrock) where the section is truncated. The upper part of the reference section in the southwestern block extends from a faulted base at GR 487941 to GR 481945 (Glenrock). Maximum thickness of the Frog Hollow Formation is estimated at 13001400 m. The base of the formation is defined as the first appearance of volcaniclastic sandstone and mudstone above the primarily volcanic rocks of the Pitch Creek Volcanics. The top of the Frog Hollow Formation is gradational with the overlying Captain Rocks Formation, the base of which is defined by the incoming of massive conglomerate horizons. Two lithofacies associations are present in the Frog Hollow Formation. The first consists of fine- to

medium-grained volcaniclastic sandstone interbedded with dark mudstone and minor pebble conglomerate. Clasts in the conglomerates consist primarily of tuffaceous mudstone intraclasts and basic to felsic volcanic rocks. Coarse sandstone beds less than 6 m thick which commonly contain rip-up clasts of grey mudstone are also present. The second association consists of predominantly of thinly bedded green tuff and tuffaceous mudstone which contain abundant, but variably preserved, radiolarians and rare pillow lavas and volcaniclastic sandstone. The volcaniclastic sandstone facies is common at the base of the formation. Limestone bodies are present in both lithofacies. They are typically elongate and laterally discontinuous, up to 50 m wide and 300 m long, although most form smaller pods between 20 and 40 m in length. The character of the tuffaceous sedimentary rocks appears to change slightly above the main limestone exposures to include interbedded tan-coloured, radiolarian-bearing tuffaceous mudstone and fine-grained sandstone. Limestones in the Frog Hollow Formation at Barry contain poorly preserved heliolitid corals which indicate an Early to Middle Devonian age. Emsian conodonts have also been extracted from limestones in a fault zone at the base of the Frog Hollow Formation at Glenrock (Metcalfe et al. 1997). Limestones from Limestone Creek at Pigna Barney have yielded Pragian conodonts (Dongal 1995). The Early Devonian Helenifore laticlavium radiolarian assemblage is present in tuffaceous mudstone toward the top of the Frog Hollow Formation in the northeastern block at Glenrock (Stratford & Aitchison 1997; Metcalfe et al. 1997), and adjacent to the Bennys Tops Limestone Member at Pigna Barney (Dongal 1995). The Emsian to Eifelian Circulaforma admissarius radiolarian assemblage is also present above limestones at the base of the southwestern Block at Glenrock (Stratford & Aitchison 1997). BENNYS TOPS LIMESTONE MEMBER (NEW MEMBER)

Limestones in the upper Barnard region, like those elsewhere throughout the Gamilaroi terrane, are laterally discontinuous. One distinctive horizon of limestone bodies, herein referred to as the Bennys Tops Limestone Member, is present within the upper sections of the Frog Hollow Formation at Pigna Barney where it can be traced along strike in an arcuate pattern crossing the Pigna Barney and Manning Rivers. Individual limestone bodies vary in size from 400 x 100 m to 2 x 2 m in dimension. Similar corals to those in the Bennys Tops Limestone Member, including Pseudamplexus, Acanthophyllum, Phillipsastrea, Yacutiopora and various heliolitids and favositids, occur in limestones at the base of the southwestern block at Glenrock (Stratford & Aitchison 1994b; Stratford 1995). However until taxa with more restricted age ranges can be found, detailed correlations between the Bennys Tops Limestone Member and the limestones at Glenrock are not possible. The member is named from Bennys Tops at Pigna Barney. Large bodies of limestone up to 900 m long and

GAMILAROITERRANE STRATIGRAPHY, NSW 183 100 m thick crop out adjacent to volcaniclastic content containing both igneous and sedimentary clasts. sandstone and mudstone and minor volcanic rocks at Volcaniclastic sandstone is commonly interbedded with Limestone Creek 13 km west of Bennys Tops. As they the conglomerates. Thin-bedded (2—5 cm) green and red are not continuous with the limestones at Bennys Tops tuffaceous mudstone and black mudstone up to 30 m these particular limestones are not included with the thick are intercalated with the coarser lithologies. Rare Bennys Tops Limestone Member. The Limestone Creek crystal tuffs, fine-grained siliceous tuffs and peperites exposures have yielded Pragian conodonts (Dongal are exposed at the base of the Captain Rocks Formation 1995) which contrasts with the Emsian age (Dongal at Glenrock. Basalt and basaltic-andesite are present within the formation at Barry and there are minor basalt 1995) of the Bennys Tops Limestone Member. flows at Glenrock. The Captain Rocks Formation at Glenrock contains Captain Rocks Formation (new formation) the Helenifore pilosidiscus radiolarian assemblage The type section of the Captain Rocks Formation is which is interpreted as Eifelian in age (Stratford & located in Spring Creek from the base at GR 487932 Aitchison 1997). upstream to GR 465919 (Glenrock). The formation has a laterally variable thickness ranging from Curricabark Formation (new formation) approximately 1200 m at Glenrock and Barry to 100 to 200 m at Pigna Barney (Figure 4). The formation is The type section of the Curricabark Formation is named after Captain Rocks Creek, a tributary of Spring located in the upper catchment of Spring Creek from the base at GR 465919 upstream to GR 459899 (Glenrock). Creek. The base of the Captain Rocks Formation is defined Thick sections of basalt associated with the Folly by the appearance of thick conglomerate and volcani- Volcanics are present within this formation. Maximum clastic sandstone which are distinct from the underlying thickness of the Curricabark Formation is approximatetuffaceous mudstone and fine sandstone of the Frog ly 1200-1400 m. The formation is named after the Hollow Formation. The top of the Captain Rocks parish of Curricabark. The Curricabark Formation is the uppermost Formation is gradational and defined by the incoming of thick (>25 m) fine-grained volcaniclastic sedimentary sedimentary unit recognised in the Gamilaroi terrane in the upper Barnard region. The lower contact with the rocks of the Curricabark Formation. The Captain Rocks Formation consists predominantly Captain Rocks Formation is gradational and is marked of massive conglomerate and coarse volcaniclastic by the appearance of a significant thickness of fine sandstone with minor mudstone. The conglomerates are sandstone and mudstone and a significant decrease in either matrix- or clast-supported. Matrix-supported the proportion of conglomerate within the section. The lithologies of the Captain Rocks Formation usually Curricabark Formation at Pigna Barney is overlain by contain clasts of limestone or grey thinly bedded to Upper Devonian-Lower Carboniferous rocks although massive mudstone up to 2 m in diameter within a the exact position and character of the upper contact is medium-grained volcaniclastic sandstone matrix. Clast- uncertain. The upper contact at Glenrock and Barry is supported conglomerates have a more diverse clastic obscured under the Tertiary basalt flows. Glenrock

24 km -

Pigna Barney

|Folly Volcanics| / / / / / / / / / / / / SSS/S//SSSSS S/SSS/SS/SSS

'S/SSSSS/SSSSSS/SSSSS 'SSSSSSSfSSS/SSSSSSSS 'SSSSSSSSSSSftSSSSSSS

SSSSSSSSSSSSVSSSSfSSSSSSSSfSSSS/S

^Curricabark Formation''////, S r r r r r f r f f f f r f r r r r r f r f r r r r r f S S S S S SS/SSSSSSSSS 'SSSSSSSSSSSS/SSSSSSS SSSSSSSSSSSS 'SSSSSSSSS///SSSSSSSS S/SSS/SSSSSS 'SSSSSSSSSSSSSSSSSSSZ SSSSSSSSSSSS 'SSSSSSSSSSSSS/jJ^-^"* ssssss/ssssj. i i • • •

=Captain Rocks Formation

Figure 4 Inferred lateral relationships between Gamilaroi terrane lithostratigraphic units at various localities in the upper Barnard region. Exposure of lithostratigraphic units between the areas represented by these columns is poor to non-existent as Tertiary basalt flows obscure older rocks. PMFS marks the approximate position of the Peel-Manning Fault system.

(medium grained tuffaceous fades) PMFS Frog Hoi ow Formation ! (fine grained tuffaceous fades)

'Pitch Creek Volcanics

Benriys Tops Limestone Member

184

J. M. C. S T R A T F O R D A N D J. C. A I T C H I S O N

Predominant lithologies in the Curricabark Formation are green to blue fine-grained volcaniclastic sandstone and interbedded black to green mudstone. Some volcaniclastic sandstones contain crystal-rich beds up to 3 cm thick. Thinly bedded (3-8 cm) red-green structureless radiolarian-bearing mudstones in exposures up to 10 m thick and minor conglomerates (<10%) are present throughout the formation. There is an apparent overall fining-up trend within this formation. Both the Protoholoeciscus hindea and Ceratoikiscum regalinodus radiolarian assemblages are present in the Curricabark Formation at Glenrock indicating a probable Eifelian to Givetian age (Stratford & Aitchison 1997). Folly Volcanics (revised definition) The Folly Volcanics were informally described as the Folly Spilite by Glenton (1979). Use of the name volcanics more accurately reflects the presence of a number of different igneous rock types. The type section for the Folly Volcanics is located along an unnamed creek between GR 253196 and GR 258191 (Nundle). A reference section is given in the upper Barnard region along Ben Halls Creek from the base at GR 354010 to GR 349000 (Barry). The Folly Volcanics are laterally discontinuous throughout the upper Barnard region and vary from a maximum thickness of 1000 m in the type section (Glenton 1979) to less than 200 m at Pigna Barney (GR 725796, Pigna Barney) (Figure 4). Basalts of the Folly Volcanics occur within the upper stratigraphic units of the Gamilaroi terrane (Curricabark and Yarrimie Formations). Contacts with bounding sedimentary rocks are poorly exposed. The Folly Volcanics are named from Folly Creek (GR 255202, Nundle) near the type section. Massive and pillowed basalts, previously referred to as spilites because of a pervasive but variable degree of sodic alteration (Vallance 1960, 1968), have long been reported from upper portions of the Gamilaroi terrane from Attunga to Bundook (Benson 1915a, 1916, 1918a; Sussmilich 1921; Voisey 1939; Crook 1961; Vallance 1968; Cawood 1983; Weekes 1984; Morris 1988). These basic igneous rocks appear to form an identifiable stratigraphic horizon, but the paucity of exposure prevents confirmation of possible lateral continuity of individual basalt/dolerite horizons. The Folly Volcanics consist of massive basalt, pillow lava and minor dolerite. Volcanic rocks are generally exposed as thick (>20 m) flows of massive blue-grey basalt with subordinate intercalated intervals of basaltic pillow lava. The basalts are tholeiitic containing plagioclase and minor clinopyroxene phenocrysts. Rare basaltic breccias are intercalated with the pillow basalts. COMPARISON WITH OTHER GAMILAROI TERRANE SUCCESSIONS Successions within the Gamilaroi terrane have been studied in several separate areas including the Tamworth—Nundle district (Crook 1961; Cawood 1983), southern Hastings Block (Roberts et al. 1995) and the

Calliope volcanic assemblage in the northern New England Orogen (Kirkegaard et al. 1970; Morand 1993). In the western Hastings Block the basal unit is the Bitter Ground Volcanics, a complex succession of basic to acidic volcanic rocks, tuff, volcaniclastic sandstone and siltstone. Radiolarians from this unit include Helenifore laticlavium (Ishiga et al. 1988) which has also been reported from Lower and Middle Devonian rocks in the upper Barnard region (Stratford & Aitchison 1997). This unit may be older than the Frasnian age previously suggested (Roberts et al. 1995), and it is suggested this unit is part of the Gamilaroi terrane. On the basis of a similar sequence of lithologies the pre-Famennian Birdwood Formation and its lateral equivalent the Ellenborough Volcanics (Roberts et al. 1995) may be tentatively correlated with the Gamilaroi terrane although more detailed geological information is required. Based on their age, Middle Devonian volcaniclastic sedimentary rocks of the Mile Road beds in the eastern Hastings Block (Roberts et al. 1995) may also be part of the Gamilaroi terrane. It can be seen from these areas that while the overall mix of Gamilaroi terrane lithologies is similar throughout the northern and southern New England Orogens (Aitchison & Flood 1995; Stratford 1995; Stratford et al. 1996), detailed correlation of individual formations is more problematic. Such differences are seen to be the result of the formation of the Gamilaroi terrane in an intra-oceanic island arc setting. In this type of setting it is envisaged that volcanic centres distributed along the arc influenced local sedimentation patterns and possibly lead to the formation of localised basins (Stratford & Aitchison 1996). Evidence for this is seen in the significant development of local stratigraphic variations within the terrane. The Tamworth Group and constituent formations in the Nundle district (Benson 1913b; 1915a; Crook 1961; Cawood 1980, 1983) contain an assemblage of intraoceanic rocks similar to that seen in other parts of the terrane. This well established Gamilaroi terrane lithostratigraphy can be usefully compared with that of the upper Barnard district. Despite idealised stratigraphic columns (Cawood 1983) which illustrate continuous sections, exposure in the Nundle district is remarkably poor. Silver Gully, the type section for the Silver Gully and Yarrimie Formations, contains only scattered outcrops and contacts between formations are either poorly exposed or are totally obscured. The best correlation between the upper Barnard and NundleTamworth sections is between felsic volcanics and tuffs at the base of the Gamilaroi terrane (Figure 3), however, no direct lithologic connection exists between these two regions. Cambrian and Ordovician rocks which underlie the Drik-Drik Formation in the Nundle district (Cawood 1983) are unknown in the upper Barnard region. The overlying Northcotte Formation in the Tamworth Group consists of sandstone, granule-pebble conglomerate, vitric crystal tuff, limestone and minor siltstone (Cawood 1980, 1983). In the upper Barnard region there is also a general change to epiclastic sedimentary rocks overlying the basal volcanics. However differences between the lower Frog Hollow Formation and the

GAMILAROI TERRANE STRATIGRAPHY, NSW Northcotte Formation, include an absence in the former of red-breccia fragments, mass-flow structures and lesser amounts of detrital quartz in the volcaniclastic sandstones [3% average quartz content in the Frog Hollow Formation (Stratford 1995); 10% average quartz content in the Northcotte Formation (Cawood 1983)]. The Wogarda Argillite at Nundle is poorly exposed and consists of greenish-grey siltstone and minor sandstone (Crook 1961; Cawood 1980, 1983). In contrast upper sections of the Frog Hollow Formation are well exposed, with a more diverse range of lithologies including greenish-blue and brown radiolarian tuffs, rare basalts and massive and bedded limestones. We cannot exclude the possibility that the lower Frog Hollow Formation and Northcotte Formation, and the Wogarda Argillite and upper Frog Hollow Formation are lateral correlatives. Distinctive limestone conglomerates, such as those found in the Silver Gully Formation (Crook 1961), also occur in the Captain Rocks Formation. However, limestone conglomerates represent a particular distinctive sedimentary facies which reoccurs throughout the Gamilaroi terrane, also appearing in the Frog Hollow Formation in the upper Barnard region and the Drik-Drik and Northcotte Formations (Cawood 1980), and as such they are a poor basis for lithostratigraphic correlations. The Yarrimie Formation is characterised by distinctive black and white banded radiolarian-bearing tuffaceous mudstones with greenish-grey or green cherty argillites lower in the unit and minor greywacke and limestone (Crook 1961). In contrast, the Curricabark Formation contains a significantly greater proportion of green-blue fine-grained volcaniclastic sandstone. Clearly both areas were receiving a similar influx of volcaniclastic and biogenic material. Radiolarian-bearing tuffaceous mudstones of the Curricabark Formation lack the banded black and white motif of the tuffaceous mudstones that characterise the Yarrimie Formation in areas in the Peel River catchment. The overall appearance of the Yarrimie and Curricabark Formations indicates there was an overall fining- and thinning-upward trend in sedimentation within the Gamilaroi terrane at this time. The finingand thinning-upward trend may be related to a deepening of the basin associated with arc rifting (Stratford & Aitchison 1996). Basic volcanic rocks are widespread from Nundle southwards. The geochemical affinities of these rocks provide further evidence of arc rifting (Aitchison & Flood 1995) and can be clearly correlated from Nundle to the upper Barnard region. Overall stratigraphic similarities between exposures of Gamilaroi terrane can be seen to reflect distinct phases in the evolution of the arc in which these rocks formed (Stratford & Aitchison 1996). The presence of voluminous felsic volcanic rocks at the base of the section (Pitch Creek Volcanics and Drik-Drik Formation) are related to initial formation of arc basement, commencing in the Silurian. This is followed by a period of arc-basin sedimentation including a wide range of rocks from deep water radiolarian-bearing rocks to shallow water limestones and minor volcanics.

185

The presence of voluminous mass-flow deposits (Captain Rocks and Silver Gully Formations) higher in the section indicates the possibility of localised deepening associated with initial rifting and subsequent filling of the rift basin (Curricabark and Yarrimie Formations), with rift-related basalts (Folly Volcanics) also forming at this time. Although broad correlation between Nundle and the upper Barnard regions is possible, we suggest the differences between local stratigraphies reflect variations in intra-basinal sedimentation patterns. We suggest that local stratigraphic schemes are perhaps more illuminating when interpreting geological history of the terrane than correlations which might oversimplify the complexity inherent in depositional settings such as those in which the Gamilaroi terrane developed. ACKNOWLEDGMENTS We gratefully acknowledge the financial assistance of the Australian Research Council Large Grant A39331590. We also thank Bruce McNaughton and the staff at Glenrock, the manager of the Goonoo Goonoo Pastoral Company (Barry), Bruce Moore and family on the 'Osland' property in the Back River and various property owners in the Pigna-Barney district for permitting access to these important geological sections. Thanks also to John Roberts and Peter Cawood for their constructive reviews of this paper. REFERENCES AITCHISON J. C. 1988. Radiolaria from the southern part of the New England Orogen, eastern Australia. In: Kleeman J. D. ed. New England Orogen Tectonics and Metallogenesis, pp. 49-60. Department of Geology, University of New England, Armidale. AITCHISON J. C. 1993. Albaillellaria from the New England orogen, Eastern NSW, Australia. Marine Micropaleontology 15, 353-368. AITCHISON J. C. & FLOOD P. G. 1995. Gamilaroi terrane: a

rifted Devonian intra-oceanic island-arc assemblage, NSW, Australia. Geological Society of London Special Publication 81, 155-168. AITCHISON J. C . , FLOOD P. G . & SPILLER F. C . P.

1992.

Tectonic setting and paleoenvironment of terranes in the southern New England orogen as constrained by Radiolarian biostratigraphy. Palaeogeography, Palaeoclimatology, Palaeoecology 94, 31-54. AITCHISON J. C . & I R E L A N D T . R .

1995. A g e profile

of

ophiolitic rocks across the Late Palaeozoic New England Orogen, New South Wales: implications for tectonic models. Australian Journal of Earth Sciences 42, 11—23. AITCHISON J. C . & STRATFORD J. M . C . 1 9 9 4 . E a r l y

and

Middle Devonian radiolarian biostratigraphy of the Gamilaroi terrane, NSW, Australia. In: Interrad VII conference abstracts, Osaka Japan, October 1994, pp. 10. Interrad Organising committee, Osaka. AITCHISON J. C . , STRATFORD J. M . C . & BUCKMAN S. 1 9 9 7 .

Geology of the Upper Barnard region: evidence of Early Permian oblique-slip faulting along the Peel-Manning Fault System. Geological Society of Australia Special Publication

19, 188-196.

186

J. M. C. STRATFORD AND J. C. AITCHISON

ALLAN A. D. 1987. Geology and tectonic significance of Devonian and Permian sequences in the Upper Barnard River district. BSc Hons thesis, University of Sydney, Sydney (unpubl.).

In: Kleeman J. D. ed. New England Orogen Tectonics and Metallogenesis, pp. 7-11. Department of Geology, University of New England, Armidale. FLOOD P. G. & AITCHISON

J. C. 1992. Late Devonian

BENSON W. N. 1913a. The geology and petrology of the Great Serpentine Belt of New South Wales. Part i. Introduction. Proceedings of the Linnean Society of New South Wales 38, 490-517.

accretion of the Gamilaroi terrane to eastern Gondwana: provenance linkage suggested by the first appearance of Lachlan Fold Belt-derived quartzarenite. Australian Journal of Earth Sciences 39, 539-544.

BENSON W. N. 1913b. The geology and petrology of the Great Serpentine Belt of New South Wales. Part ii. The geology of the Nundle district. Proceedings of the Linnean Society of New South Wales 38, 569-596.

FUREY-GREIG T. 1995. The "Nemingha" and "Loomberah" limestones (Early Devonian; Emsian) of the NeminghaNundle area, northern New South Wales: conodont data and inferred environments. Courier Forschungsinstitut Senckenberg 182, 217-233.

BENSON W. N. 1913C. The geology and petrology of the Great Serpentine Belt of New South Wales. Part iii. Petrology. Proceedings of the Linnean Society of New South Wales 38, 662-724. BENSON W. N. 1915a. The geology and petrology of the Great Serpentine Belt of New South Wales. Part iv. The dolerites, spilites and keratophyres of the Nundle district. Proceedings of the Linnean Society of New South Wales 40, 121-173. BENSON W. N. 1915b. The geology and petrology of the Great Serpentine Belt of New South Wales. Part v. The geology of the Tamworth district. Proceedings of the Linnean Society of New South Wales 40, 540-624. BENSON W. N. 1916. The general geology of the Gloucester area. Proceedings of the Linnean Society of New South Wales 50,31-35. BENSON W. N. 1918a. The geology and petrology of the Great Serpentine Belt of New South Wales. Part vii. The geology of the Loomberah district and a portion of the Goonoo Goonoo Estate. Proceedings of the Linnean Society of New South Wales 43, 320-384. BENSON W. N. 1918b. The geology and petrology of the Great Serpentine Belt of New South Wales. Part viii. The extension of the Great Serpentine Belt from the Nundle district to the coast. Proceedings of the Linnean Society of New South Wales 43, 593-599. CAWOOD P. A. 1980. The geological development of the New England Fold Belt in the Woolomin-Nemingha and Wisemans Arm regions: the evolution of a Palaeozoic fore-arc terrain. PhD thesis, University of Sydney, Sydney (unpubl.). CAWOOD P. A. 1976. Cambro-Ordovician strata in northern New South Wales. Search 7, 317-318. CAWOOD P. A. 1983. Modal composition and detrital clinopyroxene chemistry of lithic sandstones from the New England Fold Belt (east Australia): a Palaeozoic forearc terrane. Bulletin of the Geological Society of America

94, 1199-1214.

CROOK K. A. W. 1961. Stratigraphy of the Tamworth Group (Lower and Middle Devonian), Tamworth-Nundle District, N.S.W. Journal and Proceedings of the Royal Society of New South Wales 94, 173-188. CROSS K. D. 1983. The Pigna Barney ophiolitic complex and associated basaltic rocks, northeastern New South Wales, Australia. PhD thesis, University of New England, Armidale (unpubl.). DONGAL G. M. S. 1995. Early Devonian (Pragian and early Emsian) fauna from the eastern Tamworth terrane, New South Wales. Memoirs of the Association of Australasian Palaeontologists 18, 131-142. FLOOD P. G . & AITCHISON J. C . 1 9 8 8 . T e c t o n o s t r a t i g r a p h i c

terranes of the southern part of the New England Orogen.

GILLIGAN L. B . & B R O W N L O W

J. W .

1987.

Tamworth-

Hastings 1:250000 Metallogenic Map. Mineral deposit data sheets and metallogenic study. New South Wales Department of Mineral Resources, Sydney. GLENTON P. N. 1979. Stratigraphy and sedimentology of Middle Devonian to Early Carboniferous of the Nundle district. BSc Hons thesis, University of Sydney, Sydney (unpubl.). HILL D. 1942. The Devonian rugose corals of the Tamworth District, N.S.W. Journal and Proceedings of the Royal Society of New South Wales 76, 142-164. HINDE G. J. 1899. On the radiolaria in the Devonian rocks of New South Wales. Quarterly Journal of the Geological Society of London 55, 38-64. ISHIGA H., LEITCH E. C., WATANABE T . , NAKE T . & IWASAKE

M. 1988. Radiolarian and conodont biostratigraphy of siliceous rocks from the New England Fold Belt. Australian Journal of Earth Sciences 35, 73—80. KIMBROUGH D . L., CROSS K . C . & KORSCH R . J. 1 9 9 3 . U - P b

isotopic ages for zircons from the Pola Fogal and Nundle granite suites, southern New England orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 403-412. Department of Geology and Geophysics, University of New England, Armidale. KIRKEGAARDA. G . , SHAW R . D . & MURRAY C . G . 1 9 7 0 .

Geology of the Rockhampton and Port Clinton 1:25 000 Sheet areas. Geological Survey of Queensland Report 38. MAWSON R. & TALENT J. A. 1994. The T a m w o r t h G r o u p

(mid-Devonian) at Attunga, New South Wales: conodont data and inferred ages. Courier Forschungsinstitut Senckenberg

168, 37-59.

MAWSON R . , TALENT J.

A.

&

FUREY-GREIG, T.

1995.

Coincident conodont faunas (late Emsian) from the Yarrol and Tamworth belts of northern New South Wales and central Queensland. Courier Forschungsinstitut Senckenberg

182, 421-445.

MAYER W. 1972. Palaeozoic sedimentary rocks from southern New England: a sedimentological evaluation. PhD thesis, University of New England, Armidale (unpubl.). MCMINN A. 1982. The age and facies distribution of the Yarrimie and Baldwin Formations in the Manilla District. In: Flood P. G. & Runnegar B. New England Geology, pp. 113-120. University of New England and AHV Club, Armidale. METCALFE I., AITCHISON J. C . & STRATFORD J. M . C . 1 9 9 7 .

Lower Devonian (Emsian) microfauna from the Gamilaroi terrane at Glenrock Station in the southern New England orogen, New South Wales. Proceedings of the Linnean Society of New South Wales 118, 123-130. MORAND V. J. 1993. Stratigraphy and tectonic setting of the Calliope Volcanic Assemblage, Rockhampton area,

GAMILAROITERRANE STRATIGRAPHY, NSW

187

Journal of Earth Sciences 40,

Ecostratigraphy and Palaeontology and Association of Australasian Palaeontologists, Sydney.

MORRIS P. 1988. Pedogenesis of fore-arc metabasites from the Palaeozoic of New England, eastern Australia. Mineralogy and Petrology 38, 1-16.

STRATFORD J. M. C. & AITCHISON J. C. 1995. Lower Permian

Queensland. Australian 15-30.

OFFLER R. 1982. Geochemistry and tectonic setting of igneous rocks in the Glenrock Station area, N.S.W. Journal of the Geological Society of Australia 29, 443-455. PEDDER A. E. H. 1967. The Devonian system of New England, New South Wales, Australia. In: Oswald D. H. ed. International Symposium on the Devonian System, Volume 2, pp. 135-142. Alberta Society of Petroleum Geologists, Calgary. ROBERTS J., LEITCH E. C . , LENNOX P. G . & OFFLER R. 1 9 9 5 .

Devonian—Carboniferous stratigraphy of the southern Hastings Block, New England Orogen, eastern Australia. Australian Journal of Earth Sciences 42, 609-634. STEWART I. 1995. Cambrian age for the Pipeclay Creek Formation, Tamworth Belt, northern New South Wales. Courier Forschungsinstitut Senckenberg 182, 565-566. STRATFORD J. M. C. 1993. Lithostratigraphy of the Gamilaroi terrane at Glenrock Station, Scone, N.S.W. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 247-253. Department of Geology and Geophysics, University of New England, Armidale. STRATFORD J. M. C. 1995. The formation of an intra-oceanic island arc rift: a case study based on the Siluro-Devonian Gamilaroi terrane at Glenrock Station, southern New England orogen, New South Wales. PhD thesis, University of Sydney, Sydney (unpubl.). STRATFORD J. M. C. & AITCHISON J. C. 1994a. Middle

Devonian radiolarians from the Gamilaroi terrane, Glenrock Station, NSW, Australia. Interrad VII Conference abstracts, Osaka Japan, October 1994, p. 112. Interrad Organising committee, Osaka. STRATFORD J. M . C . & A I T C H I S O N

J. C .

1994b.

New

Palaeozoic faunas from Glenrock Station, southern New England orogen, NSW. Abstracts and Programme, Australasian Palaeontological Convention, Sydney, February 1994, p. 91. Macquarie University Centre for

fauna from Manning facies rocks along the Peel-Manning Fault System, Glenrock Station, southern New England orogen. Proceedings of the Linnean Society of New South Wales 114, 239-246. STRATFORD J. M. C. & AITCHISON J. C. 1996. Devonian intra-

oceanic island arc rift sedimentation: facies development in the Gamilaroi terrane, New England orogen, eastern Australia. Sedimentary Geology 101, 173-193. STRATFORD J. M. C. & AITCHISON J. C.1997. Lower to Middle Devonian radiolarian assemblages from the Gamilaroi terrane, Glenrock Station, NSW, Australia. Marine Micropaleontology 30, 225—250. STRATFORD J. M . C . , AITCHISON J. C . & FLOOD P. G . 1 9 9 6 .

Correlation of fragments of a Devonian intra-oceanic island arc accreted to the New England orogen, eastern Australia. Geological Society of Australia Abstracts 41, 420. SUSSMILICH C. A. 1921. The geology of the Gloucester district of NSW. Journal and Proceedings of the Royal Society of New South Wales 60, 234-262. VALLANCE T. G. 1960. Concerning spilites. Proceedings of the Linnean Society of New South Wales 85, 8-52. VALLANCE T. G. 1968. Recognition of specific magmatic character in some Palaeozoic mafic lavas in New South Wales. Geological Society of Australia Special Publication2,

163—167.

VOISEY A. H. 1939. The geology of the lower Manning district. Proceedings of the Linnean Society of New South Wales 64, 394-407. VOISEY A. H. 1958. The Manilla Syncline and associated faults. Journal and Proceedings of the Royal Society of New South Wales 91, 209-214. WEEKES G. J. 1984. The stratigraphy, sedimentology and structure of the Bo Bo Creek district, Taree, NSW. BSc Hons thesis, University of Sydney, Sydney (unpubl.). (Received 10 April 1996; accepted 15 July 1996)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 188-196.

Geology of the Upper Barnard region: evidence of Early Permian oblique-slip faulting along the Peel-Manning Fault System J. C. AITCHISON,1 J. M. C. STRATFORD2 AND S. BUCKMAN1 1

Department of Earth Sciences, University of Hong Kong, Pokfulam Road, Hong Kong, China 2 Department of Geology and Geophysics, University of Sydney, NSW 2006, Australia

Palaeozoic Weraerai, Djungati and Anaiwan terranes of the New England Orogen are well exposed in the Upper Barnard River catchment. These terranes are juxtaposed against each other along various strands of the PeelManning Fault System which transects the area. The presence of voluminous felsic arc volcanics and subvolcanic intrusives in the easternmost exposures of the Gamilaroi terrane is inconsistent with earlier interpretations that this terrane developed in association with an east-facing subduction system. Lower Permian sediments accumulated rapidly in transtensional basins that developed at a major bend in the Peel-Manning Fault System in response to oblique slip along this structure. Breccias and diamictites were deposited in narrow elongate basins which were gradually displaced relative to source terranes during their development. A substantial tuffaceous component in Manning facies rocks indicates contemporaneous subduction-related volcanism. Evidence of significant Cenozoic landscape evolution is seen in major relief at an Eocene angular unconformity above which thick Cenozoic fluvial sediments and widespread alkaline basalts accumulated. Subsequent dissection of the landscape resulted in development of present day topographic relief. Key words: New England Orogen, oblique-slip faults, Permian, structural terranes.

INTRODUCTION

The Upper Barnard valley provides some of the best exposures of the rocks which are juxtaposed along the Peel—Manning Fault System in eastern Australia. Unlike most of the New England Orogen, considerable topographic relief in this area permits close examination of the lithologies and structures within several key New England terranes. Details of terrane interrelationships provide constraints on the timing and nature of tectonic assembly of this region. The orientation of the PeelManning Fault System changes markedly through the area and detailed mapping (Bird 1993; Buckman 1993; Stratford 1995; Aitchison & Stratford unpubl. data) provides clues as to the nature of this fault system. In this paper, we describe the nature and distribution (Figures 1, 2) of the lithotectonic units which outcrop in the area. Assessment of their spatial relationships permits development of a new understanding of this section of the complex Peel-Manning Fault System. Over the past decade, considerably more information has become available on the New England Orogen (Flood & Runnegar 1982; Kleeman 1988; Flood & Aitchison 1993). New age constraints invalidate many earlier models for tectonic evolution of this region. Models, adapted from elsewhere in the world (Beck 1985), upon which those for the New England Orogen are based (Murray et al. 1987), have been discarded suggesting that reassessment of New England Orogen models is required. However, despite a proliferation of models in the 1980s (Cawood 1982b; Flood 1988;

Harrington & Korsch 1985; Murray et al 1987) relatively little new detailed mapping upon which models can be based has been published. Since an initial reconnaissance survey of this region (Benson 1918) the majority of geological mapping in the Upper Barnard region, with a few exceptions (Offler 1982a), has been limited to unpublished student theses (Allan 1987; Cross & Fergusson 1983; Mayer 1972). The Geological Survey of New South Wales has produced a compilation of local student thesis and undergraduate project mapping which is incorporated on its 1:250 000 metallogenic map sheet (Gilligan & Brownlow 1987). Several different geological units are present in the Upper Barnard region. Older units in the area are tectonostratigraphic entities or terranes and original relationships between them remain uncertain. Younger units constitute overlap or successor assemblages which developed after the initial amalgamation/accretion of the various terranes. The Upper Barnard region is named after the Barnard River which cuts across the area. It is presently split into a series of watershed exposures separated by erosional remnants of Cenozoic basalt flows, which from north to south are referred to as Barry, Glenrock and Pigna Barney. All grid references are from the Central Mapping Authority, New South Wales 1:25 000 topographic map sheets Barry (9134-I-N), Curricabark (9124-IV-S) and Glenrock (9134-I-S). Where referred to in the text, the title of the relevant map sheet is given at the end of the grid reference.

BT

^

CR -T C ® O' 3 O

** £

^ G

3 T

8 N

£ ^

S.dQ

O CD RT- O SRG O 3 O O EL C/I 3

3

OQ"^ $3 O I—T-

§

C

O

I S

HRI TN P 2-

M

o o R

W 3 3 P

O

•-A OKPS ' F G P

O

C _

cr DO

M ^

P

CD CD R

W

>

DO p M CD

>

P

g 3

O ^

CD

M

O

HH

O O N O Z OO

T3 BCL O CD £ "O O 3 EO CTQ 151°30'E

H51°40'E

00

190

J. C. AITCHISONETAL

GEOLOGICAL UNITS Anaiwan terrane The Anaiv^an terrane (Flood & Aitchison 1988) is characterised by voluminous arc-derived volcaniclastic turbiditic sediments (Flood & Fergusson 1982; Korsch 1984), tuffs, minor cherts and rare basalts. The lithologies present and their spatial distribution, have led to widespread interpretation of the terrane as a subduction complex with which we concur. Using radiolarian biostratigraphy, Aitchison et al. (1992) suggested an idealised relict stratigraphic succession of base-faulted slices of basalt overlain by predominantly

light-coloured, ribbon-bedded chert, interbedded tuffs and volcaniclastic greywackes. This stratigraphy is repeated by thrust faults that are interpreted as having developed during subduction (Cawood 1982a; Aitchison etal. 1992). Although the overall lithological content of the Anaiwan terrane is similar to that of the Djungati terrane, the relative proportions of constituent lithologies, and details of geochemistry, petrography and ages, differ (Aitchison et al. 1992). The Djungati terrane is dominated by red ribbon-bedded chert which is different to the chert in the Anaiwan terrane (Aitchison et al. 1992). Furthermore, the Anaiwan terrane is dominated by volcaniclastic greywackes (Flood & Fergusson 1982). These major differences are

Figure 2 Interpretive crosssections through the basement geology of the Upper Barnard region. Symbols are as in Figure 1. Contacts between all lithologies are faulted. Dips on faults have been inferred from the nature of their surface traces. Interpretation at depth is more conjectural but is based on what might be expected given the surface distribution of lithologies, overall tectonic setting, and deep seismic data from elsewhere along the Peel-Manning Fault System (Korsch et al. 1993).

GEOLOGY, UPPER BARNARD REGION, NSW

191

readily recognisable in the field. Anaiwan terrane rocks occur along the eastern edge of the area mapped and have been thrust westwards to structurally overlie other units (Figure 1). Excellent exposures occur in the bed of the Barnard River (GR 582970-610968 Curricabark) where they are easily accessed in the vicinity of the Elcom Barnard-Hunter Valley water scheme. Radiolarians recovered from cherts and tuffaceous cherts at Glenrock and Barry indicate a Devonian range similar to, or slightly older than, that reported elsewhere from the southwestern parts of this terrane (Aitchison et al. 1992). Coral faunas from limestone lenses and within greywackes constrain trench sedimentation to the Early to Middle Devonian (Stratford & Aitchison 1997a). Although previously referred to en masse as melange (Cross et al. 1987) neither the Anaiwan terrane nor the Djungati terrane consist entirely of block-in-matrix melange. There are localised melange zones where a variety of blocks occur in a matrix but elsewhere many units are merely internally disrupted and exotic lithologies are absent. Such units are best referred to as broken formation (sensu Hsu 1974). Within both terranes individual lithological packages such as chert horizons are mappable over considerable (several kilometres) distances.

Devonian coral and conodont fauna (Bird 1993). Associated volcanic and pyroclastic rocks range from alkaline basalt to more evolved compositions. A thick succession of hyaloclastites is intercalated with, and overlies, the basalt. Other similar alkaline basalts, of probable hot-spot origin, are known from elsewhere in the terrane (e.g. Monkey Creek; Cawood 1982a, 1984) and are interpreted as accreted seamounts (Flood et al. 1994). The relatively thin stratigraphic succession of the Djungati terrane is structurally repeated by numerous thrust faults. The terrane is further characterised by mesoscopic isoclinal folding of strata (Blake & Murchey 1988) and unless facing can be determined it is often difficult to discriminate between the effects of isoclinal folding and thrust faulting. However, basalt/ limestone successions at Barry and basalt/chert successions at Glenrock form distinctive mappable horizons that locally aid in the detection of large-scale folds. Present knowledge of radiolarian biostratigraphy (Stratford & Aitchison 1997b) indicates that the radiolarian cherts formed during the interval Silurian to Middle Devonian. Tectonic assembly of the sequence of thrust repetitions of the stratigraphic succession of chert and volcaniclastic sandstone, which now characterise the terrane, occurred between the Middle and Late Devonian.

Djungati terrane

Weraerai terrane

The Djungati terrane (Flood & Aitchison 1988) is characterised by erosion-resistant horizons of SilurianDevonian red ribbon-bedded cherts (Aitchison et al. 1992). In many areas these cherts can be traced for considerable distances along strike. The cherts commonly overlie base-faulted metabasaltic units of either MORB or alkaline affinity (Offler 1982a). The youngest rocks in the terrane are felsic arc-derived volcaniclastic sandstones (Cawood 1983) interbedded with chaotic olistostromal units which contain blocks of all other Djungati lithologies. These rocks comprise a relatively thin succession (Aitchison et al. 1992) that is repeated numerous times along thrust faults in a style similar to that seen in the Marin Headlands terrane of western North America (Murchey 1984). As many of these faults are at a very low angle to bedding they are almost impossible to detect without the detailed biostratigraphic control provided by radiolarians. Most of the Djungati terrane between Barry and Pigna Barney comprises thrust repetitions of red ribbonbedded chert. Massive grey tuffaceous cherts are locally common in the vicinity of Bralga Gap (GR 527949 Glenrock). Olistostromal units are present around Meerschaum Creek and include some large limestone olistoliths (Bird 1993). Within the bed of the Upper Barnard River upstream of Barry Homestead and along Wild Cattle Creek, red, green and grey tuffaceous radiolarian cherts are intercalated with subordinate finegrained felsic volcaniclastic sandstones. A distinctive subunit occurs within the Djungati terrane at Barry Station. This unit is characterised by locally pillowed alkaline basalt (Flood et al. 1994), which is overlain by fossiliferous limestone containing a

The Weraerai terrane (Flood & Aitchison 1988) comprises the remnants of an Early Cambrian ophiolite which outcrops along the Peel-Manning Fault System and is described in detail elsewhere (Aitchison et al. 1994; Aitchison & Ireland 1995). An extensive zone of Weraerai terrane rocks, up to 1 km wide, outcrops in the eastern part of the area. The ophiolite is dismembered but all elements of the former stratigraphic sequence — tectonised harzburgite (GR 525946 Glenrock), gabbro (GR 519979 Glenrock), dolerite and pillow basalts (GR 380103 Barry), as well as plagiogranite (GR 424066 Barry) and chert (GR 432068 Barnard River) — are present. Plagiogranites (tonalites) are extensively developed in the Barry Property where they intrude gabbroic rocks. Gabbros in the Schofields Creek region (GR 525975 Glenrock) have been affected by hightemperature amphibolite facies metamorphism (Offler 1982b). Several subparallel zones of more extensively disrupted Weraerai terrane rocks outcrop to the west of the main zone of Weraerai terrane, along the Peel, Bralga and Barnard River Faults. Weraerai terrane rocks have been reduced to serpentinite-matrix melange and are discussed further in the Peel-Manning Fault System section below. Radiometric dating of plagiogranites from outside the study area indicates an Early Cambrian age of formation for the Weraerai terrane ophiolite (Aitchison & Ireland 1995). The first evidence of the presence of Weraerai terrane rocks in New England is seen as detrital clasts in Lower Permian Manning facies sedimentary rocks which crop out along the PeelManning Fault System (Aitchison & Flood 1992). The original location of formation of the Weraerai terrane remains conjectural but we note similarities to

192

J. C. A I T C H I S O N ET AL.

Cambrian ophiolites of the Lachlan Orogen in western Victoria and Tasmania (Aitchison et al. 1994). Gamilaroi terrane Subvolcanic felsic calc-alkaline intrusives and extrusive lavas and their pyroclastic and volcaniclastic cover, together with minor island-arc tholeiites (both hypabyssal intrusives and eruptive flows), represent fault-bounded remnants of a Devonian intra-oceanic island arc (Aitchison & Flood 1994; Stratford & Aitchison 1996). A 4200+ m-thick section of intercalated sedimentary and volcanic rocks is present within the Gamilaroi terrane (Flood & Aitchison 1988) in the Upper Barnard region (Stratford 1993a, b, 1995; Stratford & Aitchison 1994, 1996, 1997b). Detailed geological mapping permits the recognition of five mappable units (Stratford 1995; Stratford & Aitchison 1996, this volume). The Pitch Creek Formation lies at the faulted base of the succession and consists primarily of felsic volcanic rocks and associated tuffs. It is overlain by volcaniclastic sandstones and fine-grained tuffs of the Frog Hollow Formation, which envelope the laterally discontinuous Bennys Tops Limestone Member at Pigna Barney. Overlying this unit are volcaniclastic conglomerates, sandstones and mudstones of the Captain Rocks Formation. The uppermost units in the Upper Barnard region are fine-grained volcaniclastic sandstones, mudstones and tuffs of the Curricabark Formation and intercalated massive and pillow basalts and dolerites of the Folly Volcanics. Eastern outcrops of the terrane are dominated by felsic intrusive and extrusive rocks overlain by volcaniclastic sediments including thick volcarenites generated from submarine high-density mass flows. Radiolarian-rich tuffaceous mudstones are also common especially to the west of the area. Thick lenses of tholeiitic pillow basalts occur high in the stratigraphic succession at several localities. Radiometric dating indicates that the volcanic basement of the Gamilaroi terrane may be as old as Silurian (Aitchison & Ireland 1995; Kimbrough et al. 1993). Detailed radiolarian biostratigraphy allows correlation between zones of Gamilaroi terrane outcrop and indicates that the sedimentary assemblages accumulated between the Early and Middle Devonian. Rich Lower Devonian coral faunas occur in the Glenrock area (Stratford 1995) and Lower to ?Middle Devonian brachiopods and conodonts in Gamilaroi limestones at Barry (GR 409069 Barry; Buckman 1993). Although the stratigraphic sequence is broadly similar to that of the Gamilaroi terrane in the Tamworth-Nundle area, unequivocal lithological correlation with the Tamworth Group is problematic. Differences in the stratigraphic successions in these two areas are interpreted as being related to original lithofacies variability inherent in intra-oceanic island arc environments such as the one in which the Gamilaroi terrane formed. The Gamilaroi terrane has long been interpreted as a forearc sequence which developed above a west-dipping subduction system and was associated with a now obscured or 'hidden' magmatic arc (Leitch 1974). The

recognition of the extensive zone of extrusive and intrusive arc volcanic and subvolcanic rocks which outcrops from Barry to Pigna Barney, a strike length of over 40 km, necessitates reassessment of the evolution of the terrane. Recent interpretations of the Gamilaroi terrane are as an intra-oceanic island arc, which developed above an east-dipping plate and was accreted to the Gondwana margin in the Late Devonian (Aitchison & Flood 1994; Stratford 1995; Stratford & Aitchison 1996). Manning facies The Manning facies (Manning Group, Mayer 1972; Manning facies usage after Flood & Aitchison 1988) comprises pebbly mudstones, conglomerates, sandstones and siltstones locally derived from nearby terranes, which crop out along the Peel—Manning Fault System and other major faults throughout the southern New England orogen (Aitchison & Flood 1992). These sediments are widely developed in the area mapped where they occur along the main trace of the PeelManning Fault System and are extensively developed along other associated faults in the Back River (Buckman 1993) and Upper Barnard (Allan 1987; Allan & Leitch 1990) districts. Lower Permian faunas occur in the Glenrock (Stratford & Aitchison 1995), Upper Barnard (Allan & Leitch 1990) and Pigna Barney (Cross 1983) districts and provide constraints on the timing of their deposition. At Back River the presence of horizons of comminuted shell fragments, an unidentified gastropod fossil and siliceous sponge spicules also indicates deposition in a marine environment. Analysis of Manning facies demonstrates several critical factors about the depositional setting (Aitchison & Flood 1992). Sandstone compositions vary both laterally and vertically according to their position within particular basins (Backman 1993). The presence of large angular clasts of serpentinite within conglomerates and breccias at Glenrock and Back River indicates (i) that Weraerai terrane serpentinites were exposed by the time of Manning facies deposition and (ii) proximity of the depocentre to the sediment source area. Breccias distributed along the present-day margins of Manning facies outcrop are commonly juxtaposed against country rocks that could not possibly have been their sources (Buckman 1993). Such geological mismatches are a characteristic of sediments which develop in basins associated with oblique-slip tectonics (Christie-Blick & Biddle 1985). Volumetrically subordinate, but significant, horizons of lithologies present amidst the Manning facies at Glenrock include a thick, resistant horizon of feldspathic crystal tuff together with thin horizons of enigmatic weathered basic extrusive rocks. Thin-section examination of Manning facies rocks elsewhere in the southern New England Orogen (J. C. Aitchison & C. A. Landis unpubl. data) indicates that a devitrified tuffaceous matrix is common in Manning facies diamictite units suggesting that the contribution of active volcanoes to the formation of this unit may have been greater than previously estimated.

GEOLOGY, UPPER BARNARD REGION, NSW Much of the sedimentation within the various Manning facies basins occurred from high-density mass flows. These flows generated a thick pile of amalgamated diamictites and coarse-grained turbidites. Individual bed thicknesses of 2-3 m are common and widespread dewatering features attest to rapid loading of pre-existing thicknesses of waterlogged unconsolidated sediments. Diamictites dominate the stratigraphic section in almost all the remnants of Manning facies basins. The section in the Back River area (Buckman 1993) typifies initial basin sedimentation conditions. Accumulation of proximal fluvial breccias was typically followed by a rapid up-section transition to diamictite deposition from high-density mass flows in deep water. Where preserved, younger sediments indicate gradual infilling of individual basins. The distribution of Manning facies sediments is indicative of their having been deposited rapidly in narrow elongate basins that developed in response to oblique-slip faulting. Such faulting and associated basins characterise much of the Early Permian of the southern and possibly northern New England orogens (Aitchison & Flood 1992). In most areas containing outcrops, Manning facies rocks are very steeply dipping, basin margins are truncated by faults and sediment thicknesses only represent a fraction of the original sediment fill of individual basins. Although it has been suggested that Manning facies rocks indicate deposition within a series of basins controlled by oblique-slip faulting (Aitchison & Flood 1992) it has also been suggested that they were once deposited in a single basin which developed in response to an extensional tectonic regime (Barnard Basin: Leitch 1988). However, individual facies packages cannot easily be traced between basins and, given the distribution of basins and their sedimentary style (Vickers & Aitchison 1992, 1993, 1994), it is more likely that basins developed as independent entities in response to widespread oblique-slip faulting throughout the New England Orogen during the Early Permian (Aitchison & Flood 1992). The recognition of a considerable tuffaceous component to many Manning facies sediments indicates that although basin development occurred in response to oblique-slip fault movement the prevailing tectonic setting may not have been purely transform. In a highly oblique convergent setting, such as that recently postulated to be associated with the San Andreas Fault system in western North America (Jones et al. 1994) development of oblique-slip-related sedimentary basins may be concomitant with subduction-related volcanism. In any plate margin where convergence is oblique, a component of that motion is transferred to the overriding plate (Jarrard 1986) and commonly results in development of oblique-slip basins. The widespread existence of Early Permian subduction-related intrusive rocks in the southern New England Orogen (Shaw & Flood 1981) may also indicate ongoing plate convergence. It is, therefore, surprising that the likelihood of oblique-sliprelated basin development has received little previous attention.

193

Felsic intrusives Rare, fine-grained felsic intrusive rocks which transect Palaeozoic basement terranes are interpreted as dykes which may be related to Late Permian felsic intrusion of the Barrington Tops or Nundle Intrusive suites. Cenozoic Evidence for an extensive, deeply incised Palaeogene fluvial drainage system is well-preserved in the Upper Barnard area. Flat-lying fluvial conglomerates and sandstones up to 50 m-thick overlie an erosional surface with up to 360 m palaeorelief. These rocks are most extensively developed in the upper Schofields Creek area on Glenrock Station. Widespread alkaline olivine basalt flows, probably associated with Eocene basalts from the Mt Royal Range (Wellman & McDougall 1974), cap the Palaeogene fluvial rocks and are related to Palaeogene intraplate volcanism along the eastern margin of Australia. Outcrop of the lowermost basalt flows at different elevations throughout the study area is related to initial surface relief and the flow of basalt lavas down the palaeodrainage system rather than posteruption tectonic adjustment in the New England region. Peel-Manning Fault System Much of the Peel-Manning Fault System is characterised by zones of serpentinite-matrix melange. This melange is dominated by blocks derived from dismemberment of the Weraerai terrane ophiolite and the matrix was most likely generated through serpentinisation of olivine-rich ultramafic mantle rocks of the ophiolite. Given the low (1.8 g/cm 3 ) density of the serpentinite and the likelihood that other terranes have been thrust over the Weraerai terrane it is a realistic hypothesis to interpret much of the serpentinitematrix melange as having risen along portions of the Peel-Manning Fault System that were in a tensional setting. Other lithologies entrained in the melange include fragments characteristic of various local terranes as well as exotic high P/T metamorphic blocks such as eclogite (Allan & Leitch 1992) and blueschist (Offler 1982b; Fukui et al. 1995). Because of their interesting mineralogies and metamorphic histories these blocks are somewhat disproportionately represented in the literature relative to other New England Orogen rocks. Nevertheless they apparently represent evidence of an Ordovician (Fukui et al. 1995) metamorphic event for which other similarly aged subduction-related rocks are unknown in the New England Orogen. DISCUSSION The Peel-Manning Fault System is very complex in the Upper Barnard region. Terranes, themselves bounded by faults, are repeated across several strands of the fault. Most elements of the Peel-Manning Fault System are near vertical with kinematic indicators showing that

194

J. C . A I T C H I S O N

ETAL.

strike-slip displacements were predominant. However, the Barnard River Fault, which may also be part of the Peel-Manning Fault System, is a southwest-directed northeast-dipping thrust fault with Anaiwan terrane thrust over other terranes and Manning facies rocks. This thrust fault post-dates earlier subduction-related thrust faulting which affected the entire Anaiwan terrane (Cross et al. 1987). Overall, the distribution of Early Permian rocks and structures in the Upper Barnard region is similar to that which would be expected to have developed in an environment which was dominated by oblique-slip tectonics. A major swing in the orientation of the PeelManning Fault System from north-south to northwestsoutheast occurs in the Upper Barnard region and the spatial distribution of faults which form part of the Peel-Manning Fault System in this district shows a strong similarity to what might be expected to be observed at a bend along a strike-slip fault. Development of localised zones of transtension and transpression associated with flower structures is likely to occur anywhere that the trace of a strike-slip fault is not linear (Christie-Blick & Biddle 1985). Closely spaced zones of extension and compression may all develop in such areas. At Barry/Back River the disposition of secondary faults and their orientation with respect to the main Peel-Manning Fault System is suggestive of an overall dextral displacement. The sigmoidal (lazy-z shaped) distribution of Manning facies and basin boundary faults mimics what would be expected in a zone of dextral strike-slip displacement (Mann et al. 1983). However, structural analysis of serpentinites at Glenrock (Offler & Williams 1987) indicates that deformation of Peel-Manning Fault System serpentinites has produced fabrics indicative of sinistral strike-slip movement over a wide area. An oblique-slip model can account for almost all of the features of the Peel—Manning Fault System (Aitchison & Flood 1992) in the Upper Barnard region, including the local thrusting of the Anaiwan terrane over the Djungati terrane (Figures 1, 2). Strike-slip displacement is also a potential explanation for repetition of various terranes across the Peel—Manning Fault System. Entirely dissimilar terranes are juxtaposed across steep faults and proximal sediments and their source rocks have been separated. The basal section of the Gamilaroi terrane (Stratford & Aitchison 1996), which is seen in the Barry, Glenrock and Pigna Barney districts, is absent in areas to the northwest. A potential explanation is that the zone of Gamilaroi terrane rocks present in the east of the Upper Barnard region may have been dextrally translated from an original location closer to Nundle/ Tamworth or other points further to the north along the Peel-Manning Fault System. Alternatively, these rocks could have been translated sinistrally from near Pigna Barney. The absence of recognisable offset marker horizons precludes determination of the total displacement along the Peel-Manning Fault System and farther detailed structural and sedimentological analyses of Early Permian features in this region are clearly warranted.

ACKNOWLEDGMENTS We would like to thank landowners in the Upper Barnard region who allowed access in the process of mapping this region. In particular we would like to thank Bruce Moore (Osland), and the manager of the Goonoo Goonoo Pastoral Company (Barry Station) and Bruce McNaughton (Glenrock Station). Discussions about New England Orogen geology and manuscript reviews by Peter Flood and Robin Offler are also appreciated. ARC financial assistance with the costs of fieldwork is gratefully acknowledged. Numerous Geology III students who participated in University of Sydney field mapping camps held at Barry and Osland added to our experience. Robert Broug, Richard Jacques and Phil Manning are thanked for the many ways in which they assisted in the smooth running of field camps. SB gratefully acknowledges the company and veterinary skills of Damon Bird during his honours mapping in this area. REFERENCES AITCHISON J. C . , BLAKE M . C . JR., FLOOD P. G . & JAYKO

A. S. 1994. Paleozoic ophiolitic assemblages within the southern New England orogen of eastern Australia: implications for growth of the Gondwana margin. Tectonics 13, 1135-1149. AITCHISON J. C. & FLOOD P. G. 1992. Early

Permian

transform margin development of the southern New England Orogen, eastern Australia (eastern Gondwana). Tectonics

11, 1385-1391.

AITCHISON J. C. & FLOOD P. G. 1994. Gamilaroi terrane: a

Devonian rifted intra-oceanic island arc assemblage, NSW, Australia. In: Smellie J. L. ed. Volcanism associated with extension at consuming plate margins, pp. 155—168. Geological Society of London Special Publication 81. AITCHISON J. C . , FLOOD P. G . & SPILLER F . C . P.

1992.

Tectonic setting and paleoenvironment of terranes in the southern New England orogen as constrained by radiolarian biostratigraphy. Palaeogeography, Palaeoclimatology, Palaeoecology 94, 31-54. AITCHISON J. C . & I R E L A N D T . R .

1995. A g e profile

of

ophiolitic rocks across the Late Palaeozoic New England Orogen, New South Wales: implications for tectonic models. Australian Journal of Earth Sciences 42, 11—23. ALLAN A. D. 1987. Geology and tectonic significance of Devonian and Permian sequences in the Upper Barnard River district. BSc (Hons) thesis, University of Sydney, Sydney (unpubl.). ALLAN A. D. & LEITCH E. C. 1990. The tectonic significance

of unconformable contacts at the base of Early Permian sequences, southern New England Fold Belt. Australian Journal of Earth Sciences 37,43-49. ALLAN A. D. & LEITCH E. C. 1992. The nature and origin of

eclogite blocks in serpentinite from the Tamworth Belt, New England Fold Belt, eastern Australia. Australian Journal of Earth Sciences 39, 29-35. BECK M. E. J. 1985. Model for Late Mesozoic-Early Tertiary tectonics of Coastal California and western Mexico and speculations on the origin of the San Andreas Fault. Tectonics

1, 4 9 - 6 4 .

GEOLOGY, UPPER BARNARD REGION, NSW

195

BENSON W. N. 1918. The geology and petrology of the Great Serpentine Belt of New South Wales. Part viii. The extension of the Great Serpentine Belt from the Nundle district to the coast. Proceedings of the Linnean Society of New South Wales 43, 593-599.

FLOOD P . G . , AITCHISON

BIRD D. 1993. The geology and tectonic history of the subduction complex rocks at Barry Station, southern New England orogen, New South Wales. BSc (Hons) thesis, University of Sydney, Sydney (unpubl.).

FLOOD P. G . & FERGUSSON C . L . 1982. T e c t o n o - s t r a t i g r a p h i c

BLAKE M. C. & MURCHEY B. L. 1988. A California model for

the New England Fold Belt. In: Kleeman J. D. ed. New England Orogen Tectonics and Metallogenesis, pp. 20-31. University of New England, Armidale. BUCKMAN S. 1993. The geology north of the Barnard River at Barry Station: evidence of Early Permian strike-slip faulting and basin development. BSc (Hons) thesis, University of Sydney, Sydney (unpubl.).

J. C . & LENNOX P. G .

1994.

Limestone-capped seamounts within the Late Palaeozoic accretionary prism of the New England orogen, eastern Australia. Geological Society of Australia Abstracts 37,

110-111.

units and the structure of the Texas-Coffs Harbour region. In: Flood P.G. & Runnegar B. eds. New England Geology, pp. 71-78. University of New England and AHV Club, Armidale. FLOOD P. G. & RUNNEGAR B. 1982 (eds.). New

England

Geology. University of New England and AHV Club. FUKUI S., WATANABE T . , ITAYA T . & LEITCH E . C . 1 9 9 5 .

Middle Ordovician high P / T metamorphic rocks in eastern Australia—evidence from K-Ar ages. Tectonics 14, 1014-1020. GILLIGAN L. B. & B R O W N L O W J. W . 1 9 8 7 .

Tamworth

CAWOOD P. A. 1982a. Structural relations in the subduction complex of the Paleozoic New England Fold Belt, eastern Australia. Journal of Geology 90, 381—392.

-Hastings 1:250 000 Metallogenic Map. Mineral deposit data sheets and metallogenic study, Geological Survey of New South Wales, Sydney.

CAWOOD P. A. 1982b. Tectonic reconstruction of the New England Fold Belt in the Early Permian: an example of development of an oblique-slip margin. In: Flood P. G. & Runnegar B. eds. New England Geology, pp. 25—34. University of New England and AHV Club, Armidale.

HARRINGTON H. J. & KORSCH R. J. 1985. Tectonic model for

CAWOOD P. A. 1983. Modal composition and detrital clinopyroxene chemistry of lithic sandstones from the New England Fold Belt (east Australia): a Paleozoic forearc terrane. Bulletin of the Geological Society of America

94, 1199-1214.

the Devonian to middle Permian of the New England Orogen. Australian Journal of Earth Sciences 32, 163—179. Hsu K. J. 1974. Melanges and their distinction from olistostromes. In: Dott R. H. Jr. & Shaver R. H. eds. Modern and ancient geosynclinal sedimentation, pp. 321— 333. Society of Economic Paleontologists and Mineralogists Special Publication 19. JARRARD R. D. 1986. Terrane motion by strike-slip faulting of forearc slivers. Geology 14, 780-783.

CAWOOD P. A. 1984. A geochemical study of metabasalts from a subduction complex in eastern Australia. Chemical Geology 43, 29-47.

JONES D . L., GRAYMER R . , WANG C . , MCEVILLY T . V . &

CHRISTIE-BLICK N . & BIDDLE K . T . 1 9 8 5 . D e f o r m a t i o n a n d

KIMBROUGH D . L., CROSS K . C . & KORSCH R . J. 1 9 9 3 . U - P b

basin formation along strike-slip faults. In: Christie-Blick N. & Biddle K. T. eds. Strike-slip deformation, basin formation, and sedimentation, pp. 1-34. Society of Economic Paleontologists and Mineralogists Special Publication 37.

isotopic ages for zircons from the Pola Fogal and Nundle granite suites, southern New England orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 403-^12. Department of Geology and Geophysics, University of New England, Armidale.

CROSS K. C. 1983. The Pigna Barney ophiolitic complex and associated basaltic rocks, northeastern New South Wales, Australia. PhD thesis, University of New England, Armidale (unpubl.). CROSS K. C. & FERGUSSON C. L. 1983. Contrasting styles of

melange development in the New England Orogen. Geological Society of Australia Abstracts 9, 35-36. CROSS K . C . , FERGUSSON C . L . & F L O O D

P. G .

1987.

Contrasting structural styles in the Paleozoic subduction complex of the southern New England Orogen, Eastern Australia. In: Leitch E. C. & Scheibner E. eds. Terrane accretion and orogenic belts, pp. 83—92. American Geophysical Union Geodynamics Series 19. FLOOD P. G. 1988. New England Orogen: geosyncline, mobile belt and terranes. In: Kleeman J. D. ed. New England Orogen tectonics and Metallogenesis, pp. 1-6. University of New England, Armidale. FLOOD P. G . & AITCHISON J. C. 1 9 8 8 . T e c t o n o s t r a t i g r a p h i c

terranes of the southern part of the New England Orogen. In: Kleeman J. D. ed. New England Orogen Tectonics and Metallogenesis, pp. 7-11. University of New England, Armidale. FLOOD P. G. & AITCHISON J. C 1993 (eds.). New

LOMAX A. 1994. Neogene transpressive evolution of the California Coast Ranges. Tectonics 13, 561-574.

KLEEMAN J. D. 1988 (ed.). New England Orogen and Metallogenesis. University of New Armidale.

Tectonics England,

KORSCH R. J. 1984. Sandstone compositions from the New England Orogen, eastern Australia: implications for tectonic setting. Journal of Sedimentary Petrology 54, 192-211. KORSCH R . J., WAKE-DYSTER K . D . & JOHNSTONE D .

W.

1993. The Gunnedah Basin—-New England Orogen deep seismic reflection profile: implications for New England tectonics. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 85-100. Department of Geology and Geophysics, University of New England, Armidale. LEITCH E. C. 1974. The geological development of the southern part of the New England Fold Belt. Journal of the Geological Society of Australia 29, 83—89. LEITCH E. C. 1988. The Barnard Basin and the Early Permian development of the southern part of the New England Fold Belt. In: Kleeman J. D. ed. New England Orogen Tectonics and Metallogenesis, pp. 61-67. University of New England, Armidale.

England

MANN P., HEMPTON M . R . , BRADLEY D . C . & BURKE K . 1983.

Orogen, eastern Australia. Department of Geology and Geophysics, University of New England, Armidale.

Development of pull-apart basins. Journal of Geology 91, 529-554.

196

J. C. A I T C H I S O N E T A L .

MAYER W. 1972. Palaeozoic sedimentary rocks from southern New England: a sedimentological evaluation. PhD thesis, University of New England, Armidale (unpubl.). MURCHEY B. L. 1984. Biostratigraphy and lithostratigraphy of cherts in the Franciscan Complex, Marin Headlands, California. In: Blake M. C. Jr. ed. Franciscan Geology of Northern California, pp. 51—70 Society of Economic Paleontologists and Mineralogists Special Publication 43. MURRAY C . G., FERGUSSON C . L., FLOOD P. G., WHITAKER W .

G. & KORSCH R. J. 1987. Plate tectonic model for the Carboniferous evolution of the New England Fold Belt. Australian Journal of Earth Science 34, 213-236. OFFLER R. 1982a. Geochemistry and tectonic setting of igneous rocks in the Glenrock Station area, N.S.W. Journal of the Geological Society of Australia 29, 443-455. OFFLER R. 1982b. The origin of exotic blocks in serpentinites, Peel Fault System, Glenrock Station, N.S.W. In: Flood P. G. & Runnegar B. eds. New England Geology, pp. 43-52. University of New England and AHV Club, Armidale. OFFLER R. 1984. Subcalcic, Fe-rich amphiboles in metadolerites, Glenrock Station, NSW, Australia. Mineralogical Magazine 48, 47—52. OFFLER R. & WILLIAMS A. J. 1987. Evidence for sinistral

movement on the Peel Fault System in serpentinites, Glenrock Station, N.S.W. In: Leitch E. C. & Scheibner E. eds. Terrane accretion and orogenic belts, pp. 141—152. American Geophysical Union Geodynamic Series 19. SHAW

S. E . & F L O O D R . H .

1981. T h e N e w

England

Batholith, Eastern Australia: geochemical variations in time and space. Journal of Geophysical Research 86, 10530-10544. STRATFORD J. M. C. 1993a. Intra-oceanic island arc rocks of the Gamilaroi terrane, Glenrock Station, New England orogen, NSW. In: Ancient volcanism and modern analogues, p. 108. International Association of Volcanology and Chemistry of the Earth Interior, General Assembly, Canberra Abstracts. STRATFORD J. M. C. 1993b. Lithostratigraphy of the Gamilaroi terrane at Glenrock Station, Scone, N. S. W. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 247—254. Department of Geology and Geophysics, University of New England, Armidale. STRATFORD J. M. C. 1995. The formation of a rifted intraoceanic island arc: a case study based on the Siluro-

Devonian Gamilaroi terrane at Glenrock Station, southern New England orogen, New South Wales. PhD thesis, University of Sydney, Sydney (unpubl.). STRATFORD

J. M . C . & AITCHISON J. C .

1994.

Middle

Devonian radiolarians from the Gamilaroi terrane, Glenrock Station, NSW Australia. INTERRAD VII Osaka Abstracts, 112. STRATFORD J. M. C. & AITCHISON J. C. 1995. Lower Permian

fauna from Manning facies rocks along the Peel-Manning Fault System, Glenrock Station, southern New England orogen. Proceedings of the Linnean Society of New South Wales 115, 239-246. STRATFORD J. M. C. & AITCHISON J. C. 1996. Devonian intra-

oceanic arc rift sedimentation — facies development in the Gamilaroi terrane, New England orogen, eastern Australia. Sedimentary Geology 101, 173-193. STRATFORD J. M. C. & AITCHISON J. C. 1997a. Devonian

corals from the Anaiwan terrane, southern New England Orogen, Australia. Neues Jahrbuch fur Geologie und Palaontologie Abhandlungen 201, 289—301. STRATFORD J. M. C. & AITCHISON J. C. 1997b. L o w e r to

Middle Devonian radiolarian assemblages from the Gamilaroi terrane, Glenrock Station, NSW, Australia. Marine Micropaleontology 30, 225-250. VICKERSM. D. & AITCHISON J. C. 1992. Lower Permian

Manning Group: tectonic implications of sedimentary architecture and provenance studies in the NowendocCooplacurripa area. 26th Annual Symposium on advances in the study of the Sydney Basin Abstracts, 31—37. VICKERSM. D. & AITCHISON J. C. 1993. Lower Permian

Manning Group at Kangaroo Tops, southern New England orogen: tectonic implications of basin studies. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 309-314. University of New England, Armidale. VICKERS M. D. & AITCHISON J. C. 1994. Fossil constraints

on the development of Early Permian Khatambuhl Association rocks, Southern New England Orogen. Australasian Palaeontological Convention Abstracts, 92. WELLMANP. & MCDOUGALLI. 1974. Potassium-argon ages

of the Cenozoic volcanic rocks of New South Wales. Journal of the Geological Society of Australia 21, 247-272. (Received 15 May 1996; accepted 14 September 1996)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 197-211.

Geochemistry of the mafic and ultramafic complexes of the northern Great Serpentinite Beit, New South Wales: implications for first-stage melting K. YANG1 AND P. K. SECCOMBE2 1

CSIRO Division of Exploration and Mining, North Ryde, NSW 2113, Australia. 2 Department of Geology, University of Newcastle, NSW 2308, Australia.

The Great Serpentinite Belt in the southern New England Fold Belt of eastern Australia comprises mafic and ultramafic rocks of a disrupted ophiolitic association. In the northern portion of the Great Serpentinite Belt, chromites from the mantle harzburgite show Cr# [Cr/(A1+Cr)] in the range 0.35-0.67; the overlying cumulate sequence displays an order of crystallisation commencing with olivine + chromite and followed by clinopyroxene and plagioclase, which is similar to that of mid-ocean ridge basalts; dolerites from the uppermost parts of the ophiolitic igneous sequence have geochemical affinities transitional between mid-ocean ridge basalts and islandarc basalts. All these features are consistent with a model of single-stage melting for the generation of the ophiolite and thus point to geochemical differences with the more refractory ophiolitic complexes encountered in the southern portion of the Great Serpentinite Belt, which are considered to have been generated by second-stage melting. The geochemical distinction between the two portions of this ophiolite belt may suggest that the two portions formed in different tectonic settings. The geochemical characteristics of the northern Great Serpentinite Belt are compatible with a backarc basin as a likely tectonic setting for their generation. Key words: geochemistry, mafic rocks, New England Fold Belt, ophiolite, ultramafic rocks.

INTRODUCTION

Magmatism is known to vary both vertically and laterally within a single ophiolite belt, resulting in the coexistence of two or more types of basalts, such as boninite and alkali basalt (Laurent & Hebert 1989) or boninite, MORB (mid-ocean ridge basalts) and islandarc tholeiite (Coish 1989) as illustrated by the Appalachian ophiolites; boninite and tholeiites in the New Caledonia ophiolites (Cameron 1989; Meffre et al. 1996); MORB and boninites in the mafic-ultramafic complexes from western Tasmania (Crawford et al 1989); and MORB, island-arc tholeiite and boninite in the Pindos ophiolite of Greece (Jones et al. 1991). Each of the basalts may have formed within one tectonic environment either at different stages of melting from the same source (Crawford et al. 1989; Crawford 1993) or by partial melting from different local sources in a heterogeneous upper mantle (Van Der Laan et al. 1989). Alternatively, each may have been generated in different tectonic settings, but emplaced along the same major fault structure. Therefore, a single ophiolitic belt may record a complex tectonic history and it is necessary to identify the various magmatic associations in the one belt in order to interpret the tectonic significance of the ophiolite. The Great Serpentinite Belt comprises mafic and ultramafic rocks that are recognised as a dismembered ophiolitic association in the southern New England Fold Belt of eastern Australia (Crook & Felton 1975; Pooley 1979; Cross 1983; Rogers 1986). Since Cross (1983)

identified the low-Ti refractory nature of the Pigna Barney ophiolite complex in the southern portion of the Great Serpentinite Belt, the belt as a whole has been cited as a refractory island-arc ophiolite (Flood & Aitchison 1993; Aitchison & Ireland 1995). However, the more extensive and continuous ophiolitic complexes of the northern belt, as will be discussed in this paper, display a different geochemical affinity that is indicative of first-stage melting without a typical island-arc signature. In this paper, we present geochemical data obtained from the northern portion of the Great Serpentinite Belt to highlight the contrasts with the ophiolitic complex of the southern portion, and use these data to discuss the likely differences in the tectonic setting of the ophiolites. GEOLOGICAL SETTING

The New England Fold Belt has been interpreted as originating from subduction of the Pacific Plate beneath/onto the Australian continent during Palaeozoic-Early Mesozoic time (Aitchison et al. 1992; Flood & Aitchison 1993). In the southern New England Fold Belt, the mafic and ultramafic rocks forming the Great Serpentinite Belt crop out along the PeelManning Fault System for more than 300 km, with the outcrop width varying from less than 10 m up to approximately 2 km (Figure 1). To the west of the fault system, the Tamworth Belt has been considered to be a forearc basin sequence by Leitch (1974) or an island-arc sequence by Aitchison et al. (1992). To the east of the

198 K. YANG AND P. K. SECCOMBE fault system is the subduction complex of the Central Block (Leitch 1974; Cawood 1982a) comprising rocks ranging in age from Silurian to Late Carboniferous (Aitchison 1990). The Peel-Manning Fault probably originated during subduction (Collins 1991), as implied by the presence of exotic blocks of blueschist in the fault system, but has since experienced at least two stages of strike-slip movement as evidenced by the fabrics preserved in serpentinised harzburgite (Offler & Williams 1987). Late Palaeozoic to Mesozoic migration of the magmatic activity in the arc resulted in intrusion of granites into the subduction complex (Cawood 1984). During Early Permian, backarc extension resulted in the

development of cover basins such as the Sydney Basin, which partly overlies the Tamworth Belt, and the Nambucca Basin that overlies part of the subduction complex (Cawood 1982b; Collins et al 1993). To the southwest of the New England Fold Belt occur multiply deformed, low-grade metamorphic rocks of the Early Palaeozoic Lachlan Fold Belt. The Lachlan Fold Belt is separated from the New England Fold Belt by the Permo-Triassic Sydney Basin. Seismic data suggest that rocks of the Lachlan Fold Belt may underlie the eastern part of the Sydney Basin and probably extend further beneath the Tamworth Belt of the New England Fold Belt (Korsch et al 1993).

.. \\ \ \TTVX

L I \\V\ > \ -Jt/sN .\\ \\ \\\\\\\ \\ \ ovs^Xi-V^, . \ \ vy\ \ \ \ \ \ \ \ \ w \ \\ \\s .. \w\ \w\ \w\ \w\ \w\ \w\ \w\ w\ \ w\ \ ^> .WWWWWWWWWM .>\\\\\\\\\\\\\\\\\\\\\\\\ \\ \\ \\ \\ \\ \\ \) l —s \x \w\ w\ \w\ \\w\ \s\\' ^\o^ \ \J .\ \

Tamworth Belt I Subduction comple:

1 Central Block

Y / / / A L ^ e P-T granites ^ / / / & volcanics [ * * * Late C-Early P granites

Figure 1 Geological map of the southern New England Fold Belt (modified from Collins et al. 1993). Outcrop width of ophiolite is slightly exaggerated. 1, Bingara; 2, Upper Bingara; 3, Paling Yard; 4, Crown Mountain; 5, Glenrock; 6, Pigna Barney; PMFS, Peel-Manning Fault System.

GEOCHEMISTRY, GREAT SERPENTINITE BELT, NSW The ophiolite sequence of the Great Serpentinite Belt has been intensely deformed, resulting in internal fault imbrication, loss of section, and faulted contacts between the various ophiolitic members. The ophiolite complex in the northern portion of the Great Serpentinite Belt is relatively well preserved and is marked by continuous exposure for approximately 70 km from Bingara to Crown Mountain (Figure 1). Southwards, outcrops of the ophiolite become discontinuous, comprising separate blocks that are dominated by serpentinite. Based on the geochemical distinctions identified in this paper, the ophiolitic rocks in the Great Serpentinite Belt may be assigned to two types, representing the northern and southern portions of the belt. The boundary between the two portions cannot be resolved from the available geochemical data, but probably lies between the Glenrock and Pigna Barney segments (Figure 1). By this division, the northern portion occupies a major part of the belt (approximately along the Peel Fault) but the southern portion is distributed only along that part of the Peel-Manning Fault System where orientation of the fault system changes to a westnorthwest direction (approximately along the Manning Fault). As the Manning Fault comprises a series of splay faults (Leitch 1974), blocks of ophiolitic complexes are found to be associated with these small faults adjacent to the main fault plane (Figure 1). Whereas the northern portion of the ophiolite belt has been investigated by Pooley (1979), Offler (1982), Rogers (1986), Offler and Williams (1987), and Blake and Murchey (1988), with the study areas covering almost the entire northern portion from Bingara to Glenrock, the only detailed work in the southern portion was carried out by Cross (1983) in the Pigna Barney area, a section of the ophiolite belt approximately 20 km Southern GS (after Cross 1<

Northern GSB

F Dolerite

Low-Ti basalt

F Low-Ti dolerite Gabbro

: OI-Chr

F Low-Ti gabbro

199

long. The area where our data were collected is from Upper Bingara to Paling Yard (Figure 1), which represents the best exposure of the northern belt. In the northern belt, the mantle harzburgite is almost completely serpentinised. The crustal succession of the ophiolite is dominated by a gabbroic complex and dolerite sills. Lenses of olivine-chromite cumulate commonly lie at the base of the gabbro complex. In the gabbroic complex, layered gabbro is predominant, but isotropic gabbro is locally present. Basalt in the northern belt is rare and has not been found as an individual unit in the ophiolite sequence. Due to later tectonic disturbance, no consistent facing direction of the ophiolite sequence has been observed. A simplified stratigraphy of the northern belt, together with that of the southern belt, is shown in Figure 2.

ANALYTICAL METHODS Because the mafic and ultramafic rocks are extensively altered, we derive geochemical features of the ophiolite based only on chemical composition of relict magmatic minerals and ratios of relatively immobile trace elements. Chemical analysis of single minerals was performed with a JSM-840 scanning electron microscope fitted with a Tracor Northern 5500 EDS system, using natural and synthetic minerals as standards. The operating conditions were 2 juA beam current, 15 kV accelerating voltage and 60 seconds counting time. The detection limits for most elements are around 0.1 wt%, and the estimated error is <5%. Trace element concentrations of whole-rock samples were measured on pressed powder discs by a Philips PW 1404 automatic sequential XRF spectrometer, using the X44 software package and US Geological Survey and South African standards. The analytical error for those elements used in this paper (Cr, Ni, Zr and Y) is estimated <20%. Whole-rock rare-earth elements (REE) were commercially analysed at the Sheen Analytical Laboratories, Perth, with ICP-mass spectrometry. The detection limits are 0.2 ppm for Gd, 0.05 ppm for La, Ce, Nd, Sm, Dy, Er and Yb and 0.02 ppm for Pr, Eu, Tb, Ho, Tm and Lu. The error, estimated by duplicate samples, is <15 % for most of the elements.

RESULTS AND INTERPRETATION Harzburgite

F Low-Ti norite

cumulate

SILICATE CHEMISTRY

F Serpentinised harzburgite

Serpentinised harzburgite

F

Figure 2 Simplified ophiolitic stratigraphy of the Great Serpentinite Belt (GSB). The southern belt stratigraphy is from Cross (1983).

The primary assemblage of the harzburgite from the northern belt is 5 5 - 6 5 % olivine, 3 0 - 4 0 % orthopyroxene, 1 - 3 % chromite and 1—2% clinopyroxene, displaying a polygonal granular or pseudoporphyritic texture, with orthopyroxene varying in grainsize from 2 to 8 mm and olivine up to 2 mm in diameter. The harzburgite is highly serpentinised (>80%), but most chromite has survived serpentinisation. Relict olivine and orthopyroxene show F090.2-92.5 and

Table 1 Composition of representative olivine.

N> o o

Sample

T8911 hz

T8911 hz

T8911 hz

T89112 hz

T89112 hz

T89112 hz

8902410 8902410 8901900 8901900 8901900 8901900 8901921 8901921 8901921 hz hz oJ-chr ol-chr ol-chr ol-chr ol-chr ol-chr ol-chr

8902160 8902160 ol-chr ol-chr

8902160 ol-chr

Si0 2 AI2O3

40.77 nd

40.38 nd

40.62 nd

40.82 nd

41.98 nd

41.36 nd

41.44 nd

41.09 nd

40.32 nd

40.25 nd

40.20 nd

40.15 nd

41.31

nd

1.32

8.41

1.33

8.37 nd

1.26

8.55 nd

7.95

8.59

11.07 nd

nd

10.88 nd

11.23 nd

10.90 nd

9.11

0.16

7.69 0.12

9.04

0.16

8.81 nd

10.91

nd

8.35 nd

1.31

O >

FeO MnO

40.98 1.29

nd

50.42

50.24

nd

50.09

8.99 nd

50.42

8.78 0.16

O

MgO

9.06 nd

49.82

48.22

nd nd

48.93

nd

nd

nd

nd nd

99.08

99.79

100.85

100.46

99.55

nd 99.02

nd nd

98.97

1.30 100.02

nd

99.60

1.50 101.73

98.57

nd nd 99.21

nd

nd

nd nd

49.23 nd

49.59

0.09

49.49 nd

49.28

nd nd

47.86 nd

47.83

nd nd

48.16 nd

47.49

nd nd

50.35 nd

49.88

CaO

50.76 nd nd

99.53

101.35

100.78

100.96

nd 100.9

nd nd 100.14

0.99

1.00

0.99

1.01

1.00

0.97

0.98

1.00

1.00

1.00

1.00

1.00

A1

0.99 -

1.00

1.00

Fe Mn

0.17 -

0.17

0.17

0.17

1.00 0.04

0.04

0.04

0.99 0.04

0.04

Mg Ca Na

1.83 -

1.84

1.83

1.83

Mg#

0.91

0.92

>

Na 2 0 Total

40.41

41.50

41.12

41.06

Atom (4 oxygens)

Si

0.16

0.17

0.00

0.00

1.82

1.82

0.17

0.15

0.23

0.23

0.00 1.74

1.77

1.78

1.77

0.89

0.89

1.77

1.77

1.78

1.00

0.18

0.19

0.18

0.18

0.18

-

-

-

-

0.00

1.78

1.78

1.78

1.79

1.780

0.91

0.91

0.91

0.91

0.91

0.00 0.91

0.91 2+

2+

0.92

0.91

0.11

0.10

0.91

0.92

nd, not detected. Mg# = Mg/(Mg+Fe ), Fe is assumed to be equal to total Fe. hz, harzburgite; ol-chr, olivine-chromite cumulate.

GO w o o

o

dd w

201

G E O C H E M I S T R Y , GREAT S E R P E N T I N I T E BELT, N S W Table 2 Composition of representative orthopyroxene in harzburgite.

Sample

8902410

8902410

8902410

T8911

T8911

T8911

Si0 Ti0 AI O Cr 0 FeO MnO MgO CaO Na 0 Total

55.80 nd 2.01 0.80 5.27 nd 33.93 1.25 0.26 99.31

57.34 nd 1.35 0.55 5.57 nd 33.92 1.42 0.30 100.40

57.32 nd 0.42 0.26 5.80 nd 34.52 0.14 1.16 99.62

55.28 nd 1.98 0.94 5.40 0.21 33.41 0.90 1.22 99.34

57.84 nd 0.78 0.52 5.73 0.17 33.86 1.34 1.34 101.60

57.32 nd 0.98 0.54 5.60 0.12 33.80 1.53 1.43 101.30

2

2

2

3

2

3

2

Atom (6 oxygens) 1.96 Si Ti 0.04 Al(IV) Al(VI) 0.01 Cr 0.16 Fe _ Mn 1.77 Mg 0.05 Ca 0.01 Na 0.92 Mg# -

1.99

1.98

1.94

1.97

1.96

-

-

-

—

0.01

0.04

-

-

—

—

—

0.02 0.01 0.01 0.16

-

0.00 0.17

—

—

1.78 0.01 0.06 0.91

1.75 0.05 0.02 0.92

0.01 0.16 0.01 1.75 0.03 0.07 0.92

0.02

0.02

0.00

0.00

0.16

0.16 0.01 1.72 0.05 0.07 0.91

0.00 1.72 0.06 0.07 0.92

nd, not detected. Erio.913—0.920? respectively (this study; Pooley 1979; representative analyses in Tables 1, 2). If compared to other ophiolites, the compositions of olivine and orthopyroxene from the northern belt are more refractory than lherzolites from the Ronda ophiolite (average Fo : Dick & Bullen 1984) and the Northern Appennines ophiolites (Fo _ , En o_ for olivine and orthopyroxene respectively: Hebert et al 1989), but are similar to peridotites from many other ophiolites, such as dunites of the Twin Sisters ophiolite (average Fo Dick & Bullen 1984), harzburgite of the Yakuno ophiolite (Fo . _ i.i and En . _ i. for olivine and orthopyroxene respectively: Ishiwatari 1985) and lherzolite of the Lanzo massif (Fo _ and En _ for olivine and orthopyroxene respectively: Bodinier et al 1986). Also, the compositions are close to those of the harzburgite from the Izu-Bonin-Mariana forearc (Fo i_ .4 and En _ . for olivine and orthopyroxene respectively: Parkinson et al 1992), which is the residue after boninite generation. The very similar composition for olivine and orthopyroxene in a spectrum of mantle ultramafic rocks, which have undergone various degrees of melting, is probably due to the small value of K d 2 + between olivine and melt (Kd = 0.3: Roeder & Emslie 1970) and between orthopyroxene and 89 8

88

90

9

92

91

90 3

9

90 7

9

90

91

92

9L5

Fe

/Mg

92 5

7

92

89

92

melt (0.23-0.30: Barnes 1986). Such a small Kd value determines a slow (and small) change in Fe /Mg for both olivine and orthopyroxene in mantle rocks, compared with the change in Fe /Mg of the melt, as melting proceeds. It appears that, chemically the relict silicate minerals from the northern belt are not significantly distinct from the peridotite of the southern belt or many other ophiolites. 2+

2+

CHROMITE CHEMISTRY

Disseminated spinels in mantle peridotites may be used as an indication for the degree of partial melting (Dick & Bullen 1984). Dick (1977) and Dick and Bullen (1984) have noted that Cr# [Cr/(A1+Cr)] of chromites from alpine peridotites increase with the degree of partial melting and the latter two authors have classified three types of alpine peridotites on the bases of chromite Cr#: type 1 with Cr# <0.6 defining a mid-ocean ridge setting, type 3 with Cr# >0.6 indicative of subvolcanic arc, and type 2 with Cr# transitional between types 1 and 3, which could be produced in a variety of tectonic settings. Similarly, Hebert et al (1989) have showed that oceanic harzburgites possess higher chromite Cr# than oceanic lherzolites. Takahashi (1991) also found a

202 K. Y A N G A N D P. K. S E C C O M B E progressive decrease in Cr# of chromite from harzburgite through lherzolite to plagioclase lherzolite, which corresponds with a decrease in the degree of partial melting for generating these rocks.

mantle rocks that show varying degrees of depletion, with the less-depleted peridotite being the residue of an earlier stage melting. Cumulates CRYSTALLISATION SEQUENCE

0.9

The crustal sequence of the northern belt is dominated by a gabbroic complex. The base of the complex is 6+ s marked by lenses of olivine—chromite cumulate 5 0.5 (< 100 m thick), which in places are in fault contact with •if 6 the underlying harzburgite. At the present exposure 0.3 level, gneissic gabbro, in which the 'gneissosity' is displayed by oriented clinopyroxene, forms the main part of the complex, with isotropic gabbro present 0.3 0.4 0.5 0.6 0.7 locally. The olivine-chromite cumulate consists of Fe /(Mg+Fe ) cumulus olivine+chromite (65-75%) and intercumulus plagioclase+clinopyroxene (35-25%). The gneissic Figure 3 Composition of chromites from harzburgites of the gabbro is dominated by clinopyroxene+plagioclase, Great Serpentinite Belt. The shaded area is the composition of together with <0.5% chromite. The isotropic gabbro is chromites from the cumulate sequence of the northern belt also composed of clinopyroxene+plagioclase, accom(data from Yang & Seccombe 1993). Data for the northern panied by accessory ilmenite and magnetite, but lacking belt (circles) are from this study, for Glenrock (squares) from R. Offler (unpubl.), and for the southern belt (triangles) from chromite. This magmatic sequence reveals the order of Cross (1983). crystallisation olivine+chromite followed by clinopyroxene and plagioclase in the magma chamber. As Figure 3 depicts the composition of chromite from rock units with plagioclase or clinopyroxene as the sole both the northern and southern portions of the Great cumulus phase have not been observed in the northern Serpentinite Belt (representative analyses in Table 3). belt, it is not clear whether or not there was a The harzburgitic chromites in the northern belt have Cr# crystallisation stage of plagioclase or clinopyroxene that that lie in the range 0.35-0.67. The maximum value of followed olivine-chromite but preceded clinopyroxene— the Cr# of 0.67 for these chromites is only slightly plagioclase cotectic. higher than the values (Cr# <0.6) for both abyssal The gabbro complex is moderately to highly altered, peridotites and the type 1 alpine peridotites defined by with complete transformation of plagioclase to albite Dick and Bullen (1984), and thus the harzburgite from + clinozoisite + prehnite + pumpellyite or (hydro)the northern belt may be grouped into the type 2, grossular + prehnite. Clinopyroxene is little or weakly indicating a low to moderate degree of partial melting. altered. However, harzburgitic chromites of the Pigna Barney One of the mineralogical features of the cumulate complex in the southern belt, for which only a few sequence of the northern belt is the absence of orthoanalytical data can be cited from Cross (1983), show a pyroxene as a major phase. Orthopyroxene is a common wide range of Cr# from low to high values. While the cumulus phase crystallised from second-stage, higher Cr# (>0.7) for some of the southern belt refractory magmas (Hamlyn & Keays 1986; Crawford chromites indicate probably a higher degree of partial 1993), such as in the Bushveld Complex (Hatton & melting, those chromites with lower Cr# but high Mg# Sharpe 1989), the Stillwater Complex (Irvine et al [Mg/(Mg+Fe )] in the same area may suggest partial 1983) and cumulates of the western Tasmanian ophimelting under higher pressure (Jaques & Green 1980) olites (Crawford & Berry 1992). Also, orthopyroxene probably at a lower degree of melting. The cause for the occurs commonly as phenocrysts in boninitic lavas [e.g. highly variable spinel Cr# in the Pigna Barney complex DSDP Site 458 of the Mariana arc (Hickey-Vargas is not clear, although the low Cr# does not appear to be 1989); the Troodos Ophiolite, Cyprus (Rogers et al compatible with the refractory nature of the overlying 1989); and the New Caledonia ophiolite (Cameron magmatic rocks. The wide range of Cr#, which may 1989)]. Hence, lack of orthopyroxene in the magmatic designate the host harzburgite also as a type 2 cumulate of the northern belt may indicate that peridotite, may indicate a complicated melting history crystallisation took place from a first-stage magma. In (Dick & Bullen 1984) for the southern belt and two terms of early crystallisation of clinopyroxene, the stages of partial melting are possible. Considering the northern belt cumulate resembles MORB, but Mariana forearc as a modern analogue to the second- considering that most MORB feature a crystallisation stage melting setting, where chrome spinels from the sequence olivine-plagioclase-clinopyroxene (Bryan & recovered harzburgite (Leg 125) are 0.38-0.83 in Cr# Moore 1977; Wilson 1989; Elthon 1991), a small (Ishii et al 1992), the wide range of Cr# for the difference exists and it is probably due to the slightly southern belt is not unusual. In fact, in a tectonic higher degree of melting for the harzburgite of the environment where more than one stage of melting has northern belt in comparison with the mantle rocks of occurred, it may be rather usual to find a set of residual MORB-source (see below). second stage melting

0.7

first stage meltina

..^.p, • \

oO

0.1

0.2

2+

2+

2+

0

Table 3 Composition of representative chromite.

Sample

T8911 hz

T8911 hz

T8911 hz

T8911 hz

8902363 8902363 T89112 T89112 hz hz hz hz

T8972 ol-chr

T8972 ol-chr

T8972 ol-chr

T8972 ol-chr

T8972 ol-chr

T8972 ol-chr

T8972 ol-chr

8901951 8901951 8901951 8901951 8901951 8901951 gab gab gab gab gab gab

Si0 2

nd

nd

nd

nd

0.29

0.12

nd

nd

nd

nd

nd

nd

nd

nd

nd

0.14

nd

0.31

0.47

0.72

Ti02

nd

nd

nd

nd

nd

nd

0.11

nd

1.04

0.99

1.00

0.93

1.02

0.87

1.22

1.24

nd

0.42

0.53

1.29

0.30 0.72

A1 2 0 3

19.82

20.20

19.91

21.88

33.42

36.84

18.81

19.37

17.77

18.04

17.72

17.50

18.30

17.93

16.85

13.43

26.01

18.54

14.43

10.60

20.49

v2o3

0.10

0.23

nd

0.13

nd

nd

nd

0.24

nd

0.26

0.23

nd

nd

0.45

0.14

0.68

0.40

0.36

0.68

0.75

0.29

Cr 2 0 3

49.00

46.96

46.04

44.51

34.69

30.95

49.83

49.82

44.48

44.58

44.09

43.81

43.80

43.30

43.96

32.67

34.19

36.94

38.23

41.80

38.19

SFeO

17.44

17.97

19.36

18.67

18.85

17.55

20.68

19.52

25.03

23.98

26.65

27.35

26.33

27.09

25.10

43.42

28.90

33.59

40.23

37.14

28.75

MnO

nd

nd

nd

nd

0.42

0.43

0.25

nd

nd

nd

nd

nd

nd

nd

nd

0.40

nd

0.60

1.24

0.88

0.50

MgO

12.23

11.71

10.71

11.24

12.77

13.85

11.27

11.80

10.83

11.14

9.33

8.50

9.34

8.91

10.43

5.81

8.77

7.20

2.24

3.06

8.42

CaO

0.10

nd

0.34

0.39

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

0.68

0.26

0.64

0.38

0.69

0.40

ZnO

nd

0.25

0.26

nd

nd

0.46

nd

nd

nd

nd

nd

nd

nd

nd

nd

0.32

nd

nd

nd

0.44

nd

Total

98.68

97.32

96.62

96.82

100.44 100.20 100.95 100.75

99.15

98.99

99.02

98.09

98.79

98.55

97.70

98.79

98.53

98.60

98.43

97.37

98.06

0.08

0.13

0.20

0.08

Atom (32 oxygens) 0.03

—

—

—

0.02

-

0.20

0.19

0.19

0.18

0.20

0.17

0.24

0.25

-

0.08

0.11

0.27

0.14

6.08

6.59

9.24

10.02

5.54

5.69

5.35

5.42

5.40

5.41

5.57

5.49

5.17

4.24

7.70

5.70

4.69

3.52

6.24

—

0.03

—

-

0.00

0.05

-

0.05

0.05

-

-

0.09

0.03

0.15

0.08

0.07

0.15

0.17

0.06 7.80

-

Ti

—

—

—

A1

5.90

6.09

V

0.02

0.05

Cr Fe 3 + Fe

2+

0.04

0.07

Si

9.78

9.50

9.44

9.00

6.44

5.64

9.85

9.81

8.98

8.99

9.01

9.08

8.95

9.05

6.92

6.79

7.62

8.33

9.30

0.31

0.36

0.48

0.38

0.18

0.28

0.56

0.45

1.27

1.16

1.16

1.14

1.09

1.17

1.27

4.11

1.43

2.28

2.35

2.06

1.46

4.19

5.62

4.65

5.05

6.92

6.68

4.75 0.11

3.37

3.49

3.72

3.61

3.52

3.10

8.90

3.77

3.62

4.08

3.96

4.60

4.86

4.60

4.72

Mn

—

—

—

—

0.08

0.08

0.05

-

-

-

-

-

-

-

-

0.09

-

0.13

0.29

0.21

Mg

4.60

4.47

4.14

4.28

4.47

4.76

4.20

4.38

4.12

4.23

3.59

3.32

3.60

3.45

4.05

2.32

3.28

2.80

0.92

1.28

3.24

Ca

0.03

—

0.09

0.11

—

-

-

-

-

-

-

-

-

-

-

0.20

0.07

0.18

0.11

0.21

0.11

Zn

—

0.05

0.05

—

-

0.08

-

-

-

-

-

-

-

-

-

0.06

-

-

-

0.09

-

Mg#

0.58

0.56

0.53

0.54

0.56

0.61

0.53

0.55

0.50

0.52

0.44

0.41

0.44

0.42

0.49

0.29

0.41

0.36

0.12

0.16

0.41

Cr#

0.62

0.61

0.61

0.58

0.41

0.36

0.64

0.63

0.63

0.62

0.62

0.63

0.62

0.62

0.64

0.62

0.47

0.57

0.64

0.73

0.56

nd, not detected. Fe 3 + and Fe 2+ calculated from total Fe. hz, harzburgite; ol-chr, olivine-chromite cumulate; gab, gabbro. Cr# = Cr/(A1+Cr).

o m o o ffi m g

cn H

< O

V m >

H C/3 W

& tt

Z

H — IH z H—1 H tfl w M r H

zCO

204 K. YANG AND P. K. SECCOMBE Church and Riccio (1977) classified ophiolites on the view of the much higher Mg# for the post-cumulus basis of crystallisation sequence of magmatic cumu- clinopyroxene from the olivine-chromite cumulate than lates, and they suspected that the various crystallisation that of clinopyroxene from the dolerite, we doubt that sequences in ophiolites might reflect varying degrees of the clinopyroxene in the olivine-chromite cumulate partial melting in the mantle. This suspicion was could have crystallised from a 'closed system' of interconfirmed by Ishiwatari (1985), who has shown that the stitial melts. To produce the post-cumulus clinopyroxcrystallisation sequence for magmatic cumulates of an ene in olivine-chromite cumulate with a Mg# as high as ophiolite is linked to the degree of partial melting of that of the cumulus clinopyroxene in layered gabbro, the the mantle source. Ishiwatari (1985) thus classified interstitial melts must have had composition exchange three types of ophiolites (the plagioclase, clinopyroxene with the main body of magma during crystallisation of and orthopyroxene types) generated by low (<15%), post-cumulus clinopyroxene. Alternatively, because of moderate and high (>30%) degrees of partial melting, the decrease in Kd 2 M between clinopyroxene and respectively. He postulated that for each, the primary olivine with decrease in temperature (Obata et al. 1974, magma after olivine crystallisation was in the cited from Deer et al. 1982) Mg-Fe exchange between plagioclase, clinopyroxene and orthopyroxene field, the clinopyroxene and cumulus olivine probably ocrespectively, so that each type had a different crystal- curred at subsolidus conditions. Similarly, the decrease lisation sequence. Accordingly, the crystallisation in Cr# of clinopyroxene from the olivine-chromite sequence of the northern belt would suggest the cumulate, through gabbro to dolerite also suggests plagioclase- or clinopyroxene-type ophiolite and it then fractional crystallisation within the magma chamber suggests a low to moderate degree of partial melting. (Figure 4). Compared with the compiled data of By contrast, in the Pigna Barney complex of the Ishiwatari (1985), clinopyroxene from the cumulates of southern belt, norite (with more orthopyroxene than the northern belt is similar in Ti0 content (0.17-0.82% clinopyroxene) occurs as a major cumulate unit in the for the majority of clinopyroxene; Figure 4) to other magmatic section (Figure 2). The norite underlying clinopyroxene-type ophiolites such as the Yakuno and gabbro indicates an orthopyroxene-type crystallisation the Canyon Mountain. The moderate level of Ti0 also sequence of olivine-orthopyroxene-clinopyroxene, im- agrees with a moderate degree of melting to generate plying a relatively higher degree of melting than that of the primary magma of the northern belt. the northern belt. This crystallisation sequence is somewhat similar to the Papua, Betts Cove and Thetford o ophiolites (Church & Riccio 1977). Furthermore, the ° ° *£ B*an norite is characterised by low Ti-content, suggesting a ± 0.2 i refractory nature (Cross 1983), which is consistent with < the low-Ti nature of the overlying basalts (see following section). Fe

V

g

2

2

a

S

CHROMITE AND CLINOPYROXENE CHEMISTRY

Chromites from the olivine-chromite cumulate of the northern belt show a range in Cr# from 0.56 to 0.68 (Figure 3; Table 3). This Cr# range indicates a moderate Cr content in the primary melts, and matches the Cr# of chromite in the mantle harzburgite. As these chromite grains occur in the base of the magmatic sequence and thus represent one of the earliest crystallised phases, their Cr# value may be regarded as the indication for the Cr content of the primary magma that had not been affected by fractionation in the magma chamber. Comparing the Cr# value of the northern belt with the data compiled by Crawford et al. (1989), we see that the Cr# of the primary magma of the northern belt is lower than that of boninitic liquids (Cr# >0.7), but just beyond the range for MORB (Cr# <0.65) (Figure 3). Composition of clinopyroxene in the mafic and ultramafic rocks from the northern belt is plotted in Figure 4 (representative analyses in Table 4). The clinopyroxene displays a decrease in Mg# from harzburgite (near 1.00), olivine-chromite cumulate (0.91-0.98), layered gabbro (0.88-0.94) to dolerite (0.68-0.80). A process of fractional crystallisation could be implicated by the high Mg# of clinopyroxene from both the ultramafic and gabbroic cumulate in contrast with the low Mg# of clinopyroxene in dolerite. However, in

0.1 H

0.0 J

o<? o 6b&®>

1.2

i t 0.8 CM

o

0.0 0.7

0.8

0.9

Mg/(Mg+Fe2+)

Figure 4 Composition of clinopyroxene from the northern belt. Clinopyroxene of harzburgite (diamonds), of olivinechromite cumulate (squares), of gabbro (triangles) and of dolerite (circles).

Dolerite geochemistry Dolerite is one of the main ophiolitic members in the northern belt. The dolerite shows ophitic or intergranular texture and contains clinopyroxene and plagioclase as the main phases, accompanied by accessory Ti-magnetite that is now replaced by an

Table 4 Composition of representative clinopyroxene.

Sample 8902410 8902410 8902410 T8911 T8911 hz hz hz hz hz

T8911 T8972 hz ol-chr

T8972 T8972 8902160 8902160 8902160 8901760 8901760 8901951 8901951 8901951 8902202 8900703 8900703 8900703 8900740 8900740 8900740 ol-chr ol-chr ol-chr ol-chr ol-chr gab gab gab gab gab gab dol dol dol dol dol dol

Si0 2

53.72

54.16

54.23

54.69 54.91

54.41

52.16

52.42

52.67

52.56

51.78

52.22

51.68

52.27

52.03

53.94

53.36

52.02

53.34

53.87

52.57

53.46

53.19

53.08

Ti02

nd

nd

nd

nd

nd

nd

0.50

0.43

0.45

nd

0.46

0.75

0.25

0.27

0.23

0.13

0.13

0.49

0.28

nd

0.21

0.11

nd

0.34

AI 2 O 3

1.65

1.35

1.37

0.54

0.88

0.89

2.60

2.32

2.10

2.16

2.22

2.53

2.94

2.70

2.61

2.36

2.59

2.42

1.14

0.34

1.31

0.64

0.86

1.32

Cr 2 0 3

0.77

0.75

0.73

0.57

0.29

0.70

0.98

1.27

1.14

1.28

1.41

1.43

0.74

0.71

0.56

0.67

0.73

0.88

nd

nd

nd

0.12

nd

nd

IFeO

1.57

1.48

1.62

1.41

1.42

1.64

3.25

2.85

2.77

2.61

2.77

2.64

4.47

4.32

4.51

3.20

3.28

3.04

10.35

9.60

10.7

10.51

10.62

10.81

MnO

nd

nd

nd

nd

0.10

nd

0.13

nd

nd

nd

nd

nd

nd

nd

nd

0.14

nd

nd

0.22

0.19

0.24

0.30

0.28

MgO

17.90

17.28

17.93

17.54 17.76

17.52

16.76

16.85

17.11

16.62

16.67

16.27

16.04

16.14

15.99

16.39

16.39

15.91

14.21

14.00

13.70

13.68

13.82

13.39

23.15

22.57

23.43

22.34

22.64

22.75

23.83

23.98

23.79

22.45

23.52

22.19

22.60

22.43

22.63

nd

nd

nd

0.57

0.50

0.50

nd

nd

nd

0.59

0.26

0.44

0.50

0.43

0.54

nd

nd

nd

nd

nd

nd

nd

CaO

23.86

25.07

23.91

25.22 25.55

25.26

22.93

22.61

22.84

Na20

0.64

0.60

0.63

0.63

0.89

0.99

0.49

nd

nd

K20

nd

nd

nd

nd

nd

nd

Total 1100.10

100.69 100.40 100.6 101.80 101.41

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

nd

99.80

98.75

99.08

98.37

99.88 100.17

99.03

99.56

99.18

100.66

100.46

0.33

98.55 102.58 101.78 101.36 101.92 101.62 102.44

O m O n K m g GO H

O m >

Atom (6 oxygens) Si

1.96

1.97

1.97

1.98

1.96

1.96

1.94

1.97

1.97

1.98

1.98

1.96

1.94

1.95

1.95

1.99

1.97

1.97

1.95

1.98

1.96

1.97

1.97

1.96

Ti

-

-

-

-

-

-

0.01

0.01

0.01

-

0.01

0.02

0.01

0.01

0.01

0.00

0.00

0.01

0.01

-

0.01

0.00

-

0.01

0.03

0.03

0.01

0.02

0.02

0.01

0.03

0.03

0.02

0.03

0.04

0.06

0.05

0.05

0.01

0.03

0.04

0.03

0.01

0.03

0.01

0.02

0.03

-

-

-

-

-

0.05

0.02

0.01

0.03

0.02

0.02

0.01

0.01

0.01

0.04

0.03

0.02

-

-

-

-

-

-

0.02

0.02

0.02

0.02

0.01

0.01

0.01

0.01

0.01

0.01

-

-

0.00

-

-

-

-

-

0.06

0.04

0.05

-

-

-

0.04

0.02

0.04

0.03

0.04

0.04

0.09

0.10

0.09

0.10

0.10

0.10

0.28

0.27

0.29

0.29

0.29

0.29

-

-

-

0.00

-

-

0.01

0.01

0.01

0.01

0.01

0.01

Al(IV)

0.04

Al(VI)

-

H CO W

W hd W

H i—(

Cr

0.01

0.01

0.01

0.01

0.00

0.01

0.01

0.02

Fe 3 +

0.05

0.04

0.05

0.04

0.04

0.05

-

-

Fe 2 +

-

-

-

0.00

-

-

0.10

0.09

0.09

0.08

0.09

0.08

Mn

-

-

-

-

0.00

-

0.00

-

-

-

-

-

Mg

0.97

0.94

0.97

0.95

0.94

0.94

0.93

0.94

0.95

0.93

0.94

0.91

0.90

0.90

0.89

0.90

0.90

0.90

0.78

0.77

0.76

0.75

0.76

0.74

Ca

0.93

0.98

0.93

0.98

0.98

0.97

0.91

0.91

0.92

0.94

0.92

0.94

0.90

0.91

0.91

0.94

0.95

0.96

0.88

0.93

0.88

0.89

0.89

0.89

H M W tn r H

0.02

0.03

Z

Na Cr# Mg#

0.04 0.24 1.00

0.03 0.27 1.00

0.04 0.27 1.00

0.04 0.41 1.00

0.05 0.18 1.00

0.06 0.34 1.00

0.03 0.20 0.90

-

-

-

-

-

-

0.03

0.27

0.27

0.28

0.30

0.27

0.14

0.91

0.92

0.92

0.92

0.92

0.91

nd, not detected; hz, harzburgite; ol-chr, olivine-chromite cumulate; gab, gabbro; dol, dolerite. Al(IV) and Al(VI) calculated from total Al, and Fe 3 + and Fe 2+ from total Fe based on stoichiometry.

-

0.03

-

-

-

0.03

0.01

0.03

0.03

0.15

0.13

0.16

0.16

0.20

0.00

0.00

0.00

0.11

0.00

0.00

0.90

0.91

0.90

0.90

0.90

0.74

0.74

0.72

0.72

0.73

0.71

0.03

z

C/3

K> O

206

K. Y A N G A N D P. K. S E C C O M B E

aggregate of sphene in a skeletal framework of ilmenite. Alteration has resulted in partial replacement of plagioclase by (clino)zoisite + albite ± chlorite, and of clinopyroxene by chlorite + epidote. To identify the geochemical affinity of the parental magma, relatively immobile elements including REE, Y, the high-field-strength elements Ti and Zr, and the transition elements Ni and Cr, are used to avoid any possible effects of low-temperature metamorphism on the composition of dolerite (Coish 1977; Laurent & Hebert 1989; Saunders & Tarney 1991).

before, the slightly more refractory harzburgite of the northern belt, compared with abyssal peridotites, indicates a higher melting degree than that for the MORB source. On the other hand, basalts of the southern belt are more refractory than typical arc lavas. Compared with typical MORB, most dolerites of the northern belt are slightly lower in Y (Figure 6; Table 5). A difference exists in the Cr-Y plot between the northern dolerites and the southern basalts, with lower Cr for the former. In this plot, the northern dolerites show largely an IAB (island-arc basalt) signature. 10<

? ?

1

/ A * 4 A MORB\ ° a a°O\\ V ^ o O \\ ^ CP 8 o o° cW \ • • 1 13 • / o \\ \

E 102

O 0.1

°°o ARC

\

10

100

10

500

10

Zr (ppm)

Figure 5 Ti-Zr plot for dolerites and basalts from the Great Serpentinite Belt. Circles, dolerites of the northern belt (data from this study; Pooley 1979); triangles, basalts of the southern belt (data from Cross 1983); squares, basalts of the subduction complex (data from Offler et al. 1988). Fields for basalts after Pearce (1982).

As indicated by the Ti-Zr plot (Figure 5), dolerites from the northern belt plot between fields defining MORB in the subduction complex of the Central Block and the refractory basalts from the southern belt. As the abundance of incompatible elements in a melt is an indication of the degree of melting (Pearce 1982), the relatively high Ti and Zr contents in the northern dolerites suggest that the magma was generated by a lower degree of melting or from a less refractory mantle source than the southern basalts. The Ti/Zr for the northern dolerites ranges from 75 to 101, close to the values of 94—111 obtained by Offler et al. (1988) for the MORB lavas from the nearby subduction complex, but significantly lower than the Ti/Zr values (120-180) of the southern belt basalts. For degrees of partial melting >10%, the ratio of two incompatible elements in a melt should be independent of melting degree, but similar to that of the source rock (Gill 1981). Hence, the low Ti/Zr of the dolerites in the northern belt suggests derivation from a low Ti/Zr source, probably a MORBtype mantle, whereas the basalts of the southern Great Serpentinite Belt must have been derived from a more refractory mantle source with higher Ti/Zr values. The Ti and Zr contents for some of the northern belt dolerites are lower than typical MORB and the basalts from the subduction complex of the Central Block (Figure 5). We believe that this depletion relative to typical MORB is due to a slightly higher degree of partial melting than that of the melting process taking place in a mid-ocean ridge. As already mentioned

\ WPB

\ ,H

A/

°

\

20

L

50

1 100

Y (pprn)

Figure 6 Cr-Y plot for dolerites and basalts from the Great Serpentinite Belt. Legends and data source as for Figure 5. Fields for basalts after Pearce (1982).

In the Zr/Y-Zr plot (Figure 7), the northern dolerites clearly show affinities between MORB and IAB and they are, once again, geochemically distinct from the southern basalts. As already pointed out and further supported by REE data (see below), the apparent arc influence on the northern dolerites is probably due to slightly higher degrees of melting than MORB-source mantle. 10 5

£ N 1

0.5

0.2

aaa^A A / 10

50

Zr (ppm)

100

500

Figure 7 Z r / Y - Z r plot for dolerites and basalts from the Great Serpentinite Belt. Legends and data source as for Figure 5. Fields for basalts after Pearce and Norry (1979).

In the Ni-Cr plot (Figure 8), samples from the northern and southern belt are confined to different groups. Dolerites from the northern belt define a trend with Ni/Cr around 0.4, similar to the MORB lavas from the adjacent Central Block. The southern belt basalts

GEOCHEMISTRY, GREAT SERPENTINITE BELT, NSW

207

Table 5 Whole-rock trace-element composition of dolerite (in ppm).

Sample

8900860 sill

8900740 sill

8900660 sill

8900703 sill

8900710 sill

8902101 sill

8900890 dyke

8900080 dyke

Rb Sr Th Pb Nb Ba Zr Hf La Ga Y U Ce Cr V Cu Zn As Ni Bi Cd Sb S Ta Se F CI Ge Cs

0.6 287.1 4.0 13.2 6.0 122.5 94.7 3.1 33.7 21.8 33.5 3.0 58.8 9.5 222.0 289.4 132.3 1.5 10.9 0.4 1.6 3.3 402.1 1.3 nd nd 56.2 nd nd

0.8 108.9 1.9 4.6 4.7 6.2 60.9 1.4 1.7 14.3 27.3 nd 5.1 301.5 142.7 74.6 71.1 0.3 118.3 0.1 0.7 3.7 18.6 nd nd nd 33.1 0.4 nd

2.6 148.6 1.6 4.1 4.4 nd 75.7 2.0 9.6 17.7 31.0 nd 1.7 159.2 190.4 83.9 77.7 nd 65.8 0.1 0.6 2.3 385.3 nd nd nd 221.4 nd 27.3

4.2 186.7 2.6 4.9 6.0 123.7 49.3 nd nd 13.9 18.1 nd 14.6 67.9 251.0 113.0 77.4 1.1 41.2 3.6 0.9 nd nd nd nd 1357.8 92.4 nd 4.4

4.8 312.5 1.1 2.1 6.0 135.2 46.0 1.2 nd 11.5 17.5 nd 5.0 114.0 265.0 120.8 57.0 1.0 43.0 6.4 3.4 6.5 nd 1.2 nd 1270.0 128.5 nd nd

3.9 191.3 3.0 1.8 6.2 88.5 61.8 0.1 1.2 14.7 17.5 nd 15.0 106.9 291.8 169.6 80.0 1.4 52.9 3.0 4.8 5.8 nd 0.8 nd 1206.1 299.1 nd nd

0.9 159.9 3.3 5.2 3.1 31.3 27.2 2.0 12.7 15.1 14.7 1.6 22.9 77.5 374.4 150.7 94.0 1.4 43.1 0.1 0.9 2.1 219.5 nd nd nd 70.7 nd nd

nd 342.6 4.5 8.9 3.4 219.6 39.7 3.6 25.8 15.8 16.1 0.2 12.3 103.2 288.7 177.6 67.1 1.4 53.3 0.1 nd 0.7 15.9 nd nd nd 103.7 0.8 10.5

nd, not detected.

contain higher Cr and Ni contents than the northern belt dolerites, but define a trend with a Ni/Cr of approximately 0.12. Since both Ni and Cr are compatible with mantle rocks, their abundance and ratio in melts are controlled by the minerals that have been melted. An obvious increase in abundance of compatible elements

such as Cr and Ni is only possible when high degrees of melting are attained (Pearce 1982). Therefore, we interpret differences in both the Ni/Cr ratio and the abundance of Ni and Cr between the two groups as a probable result of partial melting from different mantle sources, that is a MORB-like source for the northern 10

o

SZ

O 10 <y

^

Q.

O

E

sC

o

0

100

200

300

Cr (ppm)

400

500

600

Figure 8 Ni-Cr plot for dolerites and basalts from the Great Serpentinite Belt. Legends and data source as for Figure 5.

La Ce Pr

Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Figure 9 Normalised REE pattern of dolerites from the northern belt. REE values of CI chondrite from Evensen et al (1978). The shaded area represents the composition of basalts from the subduction complex (data from Offler et al. 1988).

208

K. YANG A N D P. K. SECCOMBE

Table 6 Whole-rock REE composition of dolerite (in ppm).

Sample

La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

8900660 sill

8900740 sill

8900890 dyke

30.00 8.50 1.50 8.00 2.80 1.10 4.00 0.74 5.10

2.60 7.20 1.20 6.50 2.30 0.84 3.20 0.56 4.00 0.78 2.30 0.32 1.90 0.24

5.90 12.00 1.70 7.40 1.70 0.52 1.80 0.28 1.90 0.40 1.20 0.18 1.20 0.18

1.00 2.90 0.42 2.70 0.36

Measured by ICP-MS.

belt but a depleted source (after extraction of MORB or IAB melts) for the southern belt. REE abundances of basalts from arc systems are highly variable, ranging from a flat, unfractionated chondritic pattern (Baker 1982) to LREE-enriched or LREE-depleted patterns (Gill 1981; Brouxel et al 1989). These variations probably depend largely on the composition of the melting source and the extent of contamination of the source by subducting slab-derived materials or arc crust. The flat REE patterns without an Eu anomaly for two of the northern dolerite samples (Figure 9; Table 6) are between N- and P-type MORB, and similar to the basalts from the backarc basin of the East Scotia Sea (Hawkesworth et al. 1977). The overall REE abundances (approximately 10 times chondrite) and the lack of differentiation between LREE (light rare-earth elements) and HREE (heavy rare-earth elements) suggest that these samples represent melts derived from a MORB mantle source with little contamination. However, the third sample is slightly lower in REE abundances (approximately 8 times chondrite) and displays a moderate fractionation between LREE and HREE, with relative HREE depletion. LREE enrichment as shown by this third sample is common for backarc or island-arc basalts, but the lack of a negative Nb anomaly may suggest an affinity with backarc basalts (Saunders & Tarney 1991). The low Ti, Zr and Y for this sample confirm that it might have crystallised from a melt which was different from that of the other two samples. The LREE enrichment relative to HREE and the lowering of all REE compared to MORB define an 'island-arc chemical signature' resulting from slab-derived components (Saunders & Tarney 1991), and thus this sample shows more significant influences from the subducted material than the other two. Overall, the REE patterns indicate that the melting sources for these dolerites are MORB-like, but that slab-derived components have variably affected the melts.

DISCUSSION

The geochemical features of the northern Great Serpentinite Belt clearly suggest that this ophiolite sequence has been derived fromfirst-stagemelting. The primary melts demonstrate geochemical affinities with MORB or between MORB and IAB. Basalts transitional between MORB and IAB are likely to be generated in convergent plate boundaries (Saunders & Tarney 1991), although the possibility of a mid-ocean ridge setting cannot be eliminated (Walker 1989). If formed in an arc system, melting must have taken place well away from the subduction zone so that the melts were not contaminated significantly byfluidsderived from the subducting slab. Backarc basalts are clinopyroxene-bearing, and show MORB-like spinel Cr# (Dick & Bullen 1984; Saunders & Tarney 1991). Baker (1982) and Saunders and Tarney (1991) have concluded that backarc basalts have geochemical affinities either close to MORB or transitional between MORB and IAB, and Wilson (1989) proposed that the transition might be due to the involvement of subduction-derived fluids. The melting source should have been so undepleted as to supply melts with moderate levels of incompatible elements such as Ti, Zr and REE, but with relatively low contents of compatible elements like Cr and Ni. The source therefore must be similar to MORB-type mantle, probably from the backarc spreading centre. However, the melting degree is probably slightly higher than that for typical MORB, as indicated by the slightly lower Ti and Zr in the northern dolerites, compared with MORB. Cross (1983) concluded that the low-Ti basalts and norite of the southern portion of the Great Serpentinite Belt were crystallised from refractory melts generated from a depleted mantle source. The various data as presented in this paper have further indicated the geochemical and mineralogical differences between the northern and the southern portions of the Great Serpentinite Belt. To produce refractory magmas by second-stage melting, low pressures and high temperatures, a previously depleted mantle source and high H 2 0 activity are needed (Van Der Laan et al. 1989). Based on modern analogues (e.g. western Pacific) for the generation of boninitic melts, the tectonic setting is likely to be an island arc or forearc. The depletion in incompatible elements like Ti, Zr and Y in the mantle source was probably due to preliminary extraction of MORB or IAB liquids. CONCLUSIONS

The mafic and ultramafic complexes generated by firststage melting predominate in the Great Serpentinite Belt. The northern belt probably represents oceanic crust formed in an open oceanic basin (a backarc basin?), whereas the southern belt is likely to be generated in an island arc or forearc. The genetic linkage between the northern and the southern complexes is not clear. Later thrusting along the PeelManning Fault System was responsible for emplacement of the two types of ophiolitic complexes into the same fault system.

GEOCHEMISTRY, GREAT SERPENTINITE BELT, NSW ACKNOWLEDGMENTS This study was supported by a University of Newcastle Postgraduate Research Scholarship (to KY) and grants from the University of Newcastle Research Management Committee and the Australian Institute of Nuclear Science and Engineering (to PKS). Thanks are due to R. Offler for providing his unpublished data, R. Bale and D. Todd for help with XRF analyses, D. Phelan for assistance with electron microprobe analyses, and E. Krupic for thin-section preparation. Discussions with R. Offler and W. J. Collins and reviews by P. A. Cawood and Y. Niu have led to great improvement of the manuscript. REFERENCES AITCHISON J. C. 1990. Significance of DevonianCarboniferous radiolarians from accretionary terrains of the New England Orogen, eastern Australia. Marine Micropalaeontology 15, 365-378. AITCHISON J. C . & I R E L A N D T . R .

1995. A g e profile of

ophiolitic rocks across the Late Palaeozoic New England Orogen, New South Wales: implications for tectonic models. Australian Journal of Earth Sciences 42, 11—23. AITCHISON J. C . , FLOOD P. G . & SPILLER F. C . P.

1992.

Tectonic setting and paleoenvironment of terranes in the southern New England orogen, eastern Australia as constrained by radiolarian biostratigraphy. Palaeogeography, Palaeoclimatology, Palaeoecology 9, 31-54.

BAKER P. E. 1982. Evolution and classification of orogenic volcanic rocks. In: Thorpe R. S. ed. Andesites, pp. 11-23. John Wiley, New York. BARNES S. J. 1986. The distribution of chromium among orthopyroxene, spinel and silicate liquid at atmospheric pressure. Geochimica et Cosmochimica Acta 50, 1889— 1909.

BLAKE M. C. JR & MURCHEY B. L. 1988. A California model

for the New England fold belt. In: Kleeman J. D. ed. New England Orogen Tectonics and Metallogenesis, pp. 20-31. Department of Geology & Geophysics, University of New England, Armidale. BODINIER J .

L.,

GUIRAUD

M.

&

DOSTAL

J.

1986.

Geochemistry of basic dikes in the Lanzo massif (western Alps): petrogenetic and geodynamic implications. Tectonophysics 128, 77—95. BRYAN W. B. & MOORE J. G. 1977. Compositional variations

of young basalts in the mid-Atlantic Ridge rift valley near 36°49'N. Geological Society of America Bulletin 88, 556-570. BROUXEL M . , LECUYER C . & LAPIERRE H . 1 9 8 9 . D i v e r s i t y o f

magma types in a lower Paleozoic island arc-marginal basin system (Eastern Klamath Mountains, California, USA). Chemical Geology 77, 251-264. CAMERON W. E. 1989. Contrasting boninite-tholeiite associations from New Caledonia. In: Crawford A. J. ed. Boninites, pp. 314-338. Unwin Hyman, London. CAWOOD P. A. 1982a. Structural relations in the subduction complex of the Paleozoic New England Fold Belt, Eastern Australia. Journal of Geology 90, 381-392. CAWOOD P. A. 1982b. Tectonic reconstruction of the New England fold Belt in the Permian: an example of development at an oblique-slip margin. In: Flood P. G. &

209

Runnegar B. eds. New England Geology, pp. 25-35. Department of Geology & Geophysics, University of New England, Armidale. CAWOOD P. A. 1984. The development of the S.W. Pacific margin of Gondwana: correlations between the Rangitata and New England Orogens. Tectonics 3, 539-553. CHURCH W. R & RICCIO L. 1977. Fractionation trends in the

Bay of Islands ophiolite of Newfoundland: polycyclic cumulate sequences in ophiolites and their classification. Canadian Journal of Earth Sciences 14, 1156-1165. COISH R. A. 1977. Ocean floor metamorphism in the Betts Cove ophiolite, Newfoundland. Contributions to Mineralogy and Petrology 60, 255—270. COISH R. A. 1989. Boninitic lavas in Appalachian ophiolites: a review. In: Crawford A. J. ed. Boninites, pp. 264-287. Unwin Hyman, London. COLLINS W. J. 1991. A reassessment of the 'Hunter-Bowen Orogeny': tectonic implications for the southern New England Fold Belt. Australian Journal of Earth Sciences 38, 409-423. COLLINS W . J., OFFLER R . , FARRELL T . R . & LANDENBERGER

B. 1993. A revised Late Palaeozoic—Early Mesozoic tectonic history for the southern New England Fold Belt. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 69-84. Department of Geology & Geophysics, University of New England, Armidale. CRAWFORD A. J. 1993. Late Proterozoic-Cambrian tectonic evolution of the Lachlan Fold Belt in SE Australia, South Island New Zealand and northern Victoria Land, Antarctica. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 147—155. Department of Geology & Geophysics, University of New England, Armidale. CRAWFORD A. J. & BERRY R. F. 1992. Tectonic implications

of late Proterozoic-early Palaeozoic igneous rock associations in western Tasmania. Tectonophysics 214, 37-56. CRAWFORD A . J., FALLOON

T . J. & G R E E N D . H .

1989.

Classification, petrogenesis, and tectonic setting of boninites. In: Crawford A. J. ed. Boninites, pp. 2-49. Unwin Hyman, London. CROOK K. A. W. & FELTON E. A. 1975. Tasman Geosyncline

greenstones and ophiolites. Journal of the Society of Australia 22, 117-131.

Geological

CROSS K. C. 1983. The Pigna Barney ophiolitic complex and associated basaltic rocks, northeastern New South Wales, Australia. PhD thesis, University of New England, Armidale (unpubl.). D E E R W . A . , HOWIE R . A . & ZUSSMAN J. 1 9 8 2 .

Rock-forming Minerals Longman, London.

Vol

1A (Second

Orthosilicates,

Edition).

DICK H. J. B. 1977. The Josephine Peridotite, a refractory residue of the generation of andesite. EOS 56,464. DICK H. J. B. & BULLEN T. 1984. C h r o m i a n spinel as a

petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology 86, 54-76. FLOOD P. G . & AITCHISON J. C . 1 9 9 3 . U n d e r s t a n d i n g

New

England Geology: the comparative approach. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 1-10. Department of Geology & Geophysics, University of New England, Armidale.

210

K. YANG AND P. K. SECCOMBE

ELTHON D. 1991. Experimental phase petrology of mid-ocean ridge basalts. In: Floyd P. A. ed. Oceanic basalts, pp. 94— 115. Blackie, Glasgow.

LEITCH E. C. 1974. The geological development of the southern part of the New England Fold Belt. Journal of the Geological Society of Australia 21, 133-156.

EVENSEN N . M . , HAMILTON P. J. & O'NIONS R . K. 1978. R a r e

LEITCH E. C. 1980. The Great Serpentinite Belt of New South Wales: diverse mafic-ultramafic complexes set in a Palaeozoic arc. In: Panayiotou A. ed. Proceedings International Ophiolite Symposium Cyprus 1979, pp. 637— 648. Geological Survey Department, Cyprus.

earth abundances in chondritic meteorites. Geochimica et Cosmochimica Acta 42, 1199-1212. GILL J. 1981. Orogenic Andesites Springer-Verlag, Berlin.

and Plate

Tectonics.

HAMLYN P. R. & KEAYS R. R. 1986. Sulfur saturation and

second-stage melts: application to the Bushveld platinum metal deposits. Economic Geology 81, 1431-1445. HATTON C. J. & SHARPE M. R. 1989. Significance and origin

of boninite-like rocks associated with the Bushveld Complex. In: Crawford A. J. ed. Boninites, pp. 174-207. _ Unwin Hyman, London. HAWKESWORTH C . J., O ' N I O N S R . K . , PANKHURST R. J., HAMILTON P. J. & EVENSEN N . M . 1977. A g e o c h e m i c a l

study of island arc and back-arc tholeiites from the Scotia Sea. Earth and Planetary Science Letters 36, 253-262. HEBERT R., SERRIG. & HEKINIAN R. 1989. Mineral chemistry

of ultramafic tectonites and ultramafic to gabbroic cumulates from the major oceanic basins and Northern Appennines ophiolites (Italy): a comparison. Chemical Geology 77, 183-207.

HICKEY-VARGAS R. 1989. Boninites and tholeiites from DSDP Site 458, Mariana forearc. In: Crawford A. J. ed. Boninites, pp. 339-356. Unwin Hyman, London. IRVINE T . N . , KEITH D. W . & TODD S. G . 1 9 8 3 . T h e J - M

platinum-palladium reef of the Stillwater Complex, Montana: 2. Origin by double-diffusive convective magma mixing and implications for the Bushveld Complex. Economic Geology 78, 1287-1334. ISHII T . , ROBINSON P. T . , MAEKAWA H . & FISKE R .

1992.

Petrological studies of peridotites form diapiric serpentinite seamounts in the Izu-Ogasawara-Mariana forearc, Leg 125. In: Fryer P., Pearce J. A., Stokking L. B. et al. eds. Proceedings of the Ocean Drilling Program, Scientific Results 125, pp.445-485. College Station, TX (Ocean Drilling Program). ISHIWATARI A. 1985. Igneous petrogenesis of the Yakuno ophiolite (Japan) in the context of the diversity of ophiolites. Contributions to Mineralogy and Petrology 89, 155-167. JAQUES A. L. & GREEN D. H. 1980. Anhydrous melting of

peridotite at 0-15 kb pressure and the genesis of tholeiitic basalts. Contributions to Mineralogy and Petrology 73, 287-310.

MEFFRE S., AITCHISON J. C . & C R A W F O R D A . J.

1996.

Geochemical evolution and tectonic significance of boninites and tholeiites from the Koh ophiolite, New Caledonia. Tectonics 15, 67-83. OFFLER R. 1982. The origin of exotic blocks in serpentinites, Peel Fault System, Glenrock Station, NSW. In: Flood P. G. & Runnegar B. eds. New England Geology, pp. 43-51. Department of Geology & Geophysics, University of New England, Armidale. OFFLER R. & WILLIAMS A. 1987. E v i d e n c e for sinistral

movement on the Peel Fault System in serpentinites, Glenrock Station, NSW. In: Leitch E. C. and Scheibner E. eds. Terrane Accretion and Orogenic Belts, pp. 141—151. American Geophysical Union. OFFLER R . , GARRAD D . , FARDY J. & SECCOMBE P. K . 1 9 8 8 .

Accretion of N-type ORB basalts in the Nangahrah Formation, Woodsreef area, NSW. In: Kleeman J. D. ed. New England Orogen: Tectonics and Metalio genesis, pp. 99-104. Department of Geology & Geophysics, University of New England, Armidale. PARKINSON I. J., HALL G . E . M . & PEARCE J. A .

1992.

Palladium, platinum, and gold distribution in serpentinite seamounts in the Mariana and Izu—Bonin forearcs: evidence from Leg 125 fluids and serpentinites. In: Fryer P., Pearce J. A., Stokking L. B. et al. eds. Proceedings of the Ocean Drilling Program, Scientific Results 125, pp. 507-518. College Station, TX (Ocean Drilling Program). PEARCE J. A. 1982. Trace element characteristics of lavas from destructive plate boundaries. In: Thorpe R. S. ed. Andesites, pp. 525-547. John Wiley & Sons, New York. PEARCE J. A. & NORRY M. J. 1979. Petrogenetic implications

of Ti, Zr, Y and Nb variations in volcanic rocks. Contributions to Mineralogy and Petrology 69, 33-47. POOLEY G. D. 1979. Petrological studies in the Great Serpentinite Belt of NSW near Upper Bingara. PhD thesis, University of New South Wales, Sydney (unpubl.). ROEDER

P.

L. &

EMSLIE

R. F.

1970.

Olivine-liquid

equilibrium. Contributions to Mineralogy and Petrology 29, 275-289.

JONES G . , ROBERTSON A . H . F. & CANN J. R 1991. G e n e s i s

and emplacement of the Supra-subduction zone Pinos ophiolite, northwestern Greece. In: Peters Tj., Nicolas A. & Coleman R. G. eds. Ophiolite genesis and evolution of the oceanic lithosphere, pp. 771-799. Kluwer Academic Publishers, Dordrecht. KORSCH R. J., W A K E - D Y S T E R K . D . & JOHNSTONE D. W .

1993. The Gunnedah Basin-New England Orogen deep seismic reflection profile: implications for New England tectonics. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 85-100. Department of Geology & Geophysics, University of New England, Armidale. LAURENT R. & HEBERT R. 1989. The volcanic and intrusive

rocks of the Quebec Appalachian ophiolites (Canada) and their islands arc setting. Chemical Geology 77, 287-302.

ROGERS G. P. 1986. Geology of the northern part of the Great Serpentinite Belt: Bingara to Upper Bingara. BSc (Hons) thesis, University of New England, Armidale (unpubl.). ROGERS N . W . , MACLEOD C . J. & M U R T O N B . J. 1 9 8 9 .

Petrogenesis of boninitic lavas from the Limassol Forest Complex, Cyprus. In: Crawford A. J. ed. Boninites, pp. 288—313. Unwin Hyman, London. SAUNDERS A. & TARNEY J. 1991. Back-arc basins. In: Floyd P.A. ed. Oceanic basalts, pp. 219-263. Blackie, Glasgow. TAKAHASHI N. 1991. Evolutional history of the uppermost mantle of an arc system: petrology of the Horoman peridotite massif, Japan. In: Peters Tj., Nicolas A. and Coleman R. G. eds. Ophiolite genesis and evolution of the oceanic lithosphere, pp. 195-205. Kluwer Academic Publishers, Dordrecht.

GEOCHEMISTRY, GREAT SERPENTINITE BELT, NSW V A N DER LAAN S. R . , FLOWERS M . F . J. & KOSTER VAN

GROOS A. F. 1989. Experimental evidence for the origin of boninites: near-liquidus phase relations to 7.5 kbar. In: Crawford A. J. ed. Boninites, pp. 112-147. Unwin Hyman, London. WALKER C. 1989. North Atlantic ocean crust and Iceland. In: Floyd P. A. ed. Oceanic basalts, pp. 311-352. Blackie, Glasgow.

WILSON M. 1989. Igneous London.

pedogenesis.

211

Unwin Hyman,

YANG K. & SECCOMBE P. K. 1993. Chemical variation of

chromite in the ultramafic cumulates of the Great Serpentinite Belt, Upper Bingara to Doonba, New South Wales, Australia. Canadian Mineralogist 31, 75-87. (Received 10 April 1996; accepted 23 October 1996)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 212-225.

Silica—carbonate alteration zones and gold mineralisation in the Great Serpentinite Belt, New England Orogen, New South Wales P. M. ASHLEY

Department of Geology & Geophysics, University of New England, Armidale, NSW 2351, Australia.

Bodies of silica-carbonate rock (listwaenite) are common along tectonic contacts in the Great Serpentinite Belt (Weraerai terrane) in the New England Orogen of northeastern New South Wales. In places they host vein and disseminated gold deposits and their degradation has led to the formation of eluvial and alluvial gold concentrations. Hydrothermal replacement of serpentinite produced a ferroan magnesite + quartz (± fuchsite + chlorite ± dolomite + FeNiCoAs sulfide) assemblage, which commonly contains anomalous Au, As, Sb and Hg values. Silica-carbonate rock development involved addition of C0 2 , Si0 2 , CaO, K 2 0, Ba, Sr, As, Sb, Hg, Au and S, loss of H 2 0, MgO and Ni and reduction of Fe 2 0 3 to FeO. The isotopic compositions of carbonates (513C = -8.0 to 1.3%o, 5 1 8 0 = 12.2-19.6%o), quartz (5 18 0 = 15.7-21.8%o) and pyrite (534S = 0.8-3.8%o) are consistent with a deep-seated metamorphic (homogenised crust) and/or magmatic source. Fluid-inclusion data indicate that hydrothermal fluids were C0 2 -bearing, low salinity (maximum 6.1 equivalent wt% NaCl), reduced, near-neutral and had minimum homogenisation temperatures between 165° and 255°C. Gold precipitation resulted from carbonatisation and sulfidation reactions, leading to decrease in fluid pH and / S 2 and increase in / 0 2 . Hydrothermal activity, focused along the Peel Fault, occurred in the Late Permian to Early Triassic, in the terminal orogenic stage of the southern New England Orogen and was the same as that responsible for the formation of widespread structurally controlled mesothermal Au and Sb-Au deposits throughout the region. Key words: geochemistry, gold, Great Serpentinite Belt, mineralisation, New England Orogen, silicacarbonate rocks.

INTRODUCTION Bodies of quartz-carbonate altered ultramafic rock are common in disrupted ophiolite complexes worldwide (Barnes et al. 1973; Buisson & Leblanc 1987; Bohlke 1989; Pohl 1990; Madu et al. 1990; Auclair et al. 1993). They are also developed within ultramafic flow and cumulate sequences in Archaean greenstone belts, especially in association with major structural discontinuities (Fyon et al. 1983; Pearton & Viljoen 1986; Kishida & Kerrich 1987). In many occurrences there is a relationship between the alteration zones and hydrothermal Au, Sb or Hg mineralisation of mesothermal to epithermal character (Barnes et al. 1973; Coveney 1981; Pearton & Viljoen 1986; Buisson & Leblanc 1987; Kishida & Kerrich 1987; Bohlke 1989; Madu et al. 1990; Peters 1991, Tiiysuz & Erler 1993) and therefore may constitute attractive exploration targets. Carbonate-quartz-rich alteration zones in ophiolitic ultramafic rocks have commonly been termed 'listwaenite' (Buisson & Leblanc 1987; Madu et al. 1990; Auclair et al. 1993). They grade into intensely silicified serpentinite ('birbirite': Auclair et al. 1993) and into talc-carbonate altered serpentinite. The rocks represent hypogene replacement of variably serpentinised ultramafic rock and have a dominant mineral assemblage of ferroan magnesite and quartz. Ferroan dolomite and calcite occur in places, with minor chromian muscovite (fuchsite/mariposite), chlorite,

relict chromian spinel and traces of disseminated sulfide minerals. This paper describes the occurrence of alteration zones in the Great Serpentinite Belt of northeastern New South Wales, their relationship to gold mineralisation and implicit exploration potential. Altered ultramafic rocks were first reported by Benson (1913) and termed 'silica-carbonate rocks' by Ashley and Brownlow (1993) due to their mineralogical constitution. It is emphasised, however, that the term is synonymous with listwaenite. GEOLOGICAL SETTING OF SILICACARBONATE ROCKS Great Serpentinite Belt In the southern portion of the New England Orogen in northeastern New South Wales, the Great Serpentinite Belt (Weraerai terrane of Flood & Aitchison 1988) represents a structurally dismembered ophiolite sequence. It crops out largely along the PeelManning Fault system for more than 300 km between Warialda and Yarras (Figure 1), separating SiluroDevonian terranes of different character: the Gamilaroi terrane of oceanic island-arc affinities to the west and the subduction complex rocks of the Djungati terrane to the east (Figure 2) (Flood & Aitchison 1988; Aitchison et al. 1992, 1994; Aitchison & Ireland 1995). At least part of the ophiolite was generated in the

SILICA-CARBONATE ROCKS, NSW

213

31 ° S -

v^/ty*

^^^Macquarie 153°E

Figure 1 Location of the Great Serpentinite Belt along the Peel-Manning Fault system in the southern part of the New England Orogen.

Cambrian, based on U-Pb geochronology of zircons from plagiogranite (Aitchison et al 1992; Aitchison & Ireland 1995). It was subsequently technically emplaced as serpentinite-matrix melange, plus less disrupted ophiolite segments, along the Peel-Manning Fault system by the Late Carboniferous to Early Permian. Ophiolitic rocks have been contact metamorphosed by the Bundarra Plutonic Suite granitoids (Herbert 1987), the latter yielding ages of 286 ± 13 Ma and 287 ± 10 Ma (Flood & Shaw 1977; Hensel et al 1985). Nephrite reaction zones in serpentinite have been dated at 279-286 Ma (Lanphere & Hockley 1976) and fossiliferous Early Permian sandstones adjacent to the ophiolite locally contain serpentinite detritus (Cross 1983). In the Attunga and Nundle areas, the Great Serpentinite Belt has been intruded and metamorphosed by Early Triassic I-type granitoids (Figure 2). In the dismembered ophiolite, the most common rock

is serpentinite melange, derived from a tectonite harzburgite precursor. In places, the ophiolite also contains disrupted zones and tectonic blocks of partly serpentinised cumulate wehrlite and rare dunite, variably altered gabbro, dolerite and minor to rare basalt, chromitite, clinopyroxenite, diorite and plagiogranite (Yang & Seccombe 1993; Aitchison et al 1994 ). Field aspects of silica-carbonate rocks Silica-carbonate rocks are scattered along the length of the Great Serpentinite Belt in the southern New England Orogen (Figure 2). All occurrences are along tectonic contacts, with the majority being between sedimentary rocks of the Gamilaroi terrane and the western margin of the Great Serpentinite Belt. At a few locations, however, silica-carbonate rock (but no unaltered serpentinite) is juxtaposed between the Gamilaroi and

214

P.M.ASHLEY

Djungati terranes and represents the position of the Peel Fault. Occurrences range from lensoid masses a few metres long to elongate lenses and tabular bodies up to several hundred metres long and many tens of metres wide. All bodies have steep dips, and strike parallel to the Peel Fault (340° ± 10°). Lamprophyre dykes, with variable carbonate alteration, occur within and adjacent to several silica-carbonate masses. Elsewhere in the southern New England Orogen, Late Permian lamprophyres are commonly associated with

mesothermal Au (-Sb) deposits (Henley 1991; Ashley etal. 1994; Kent 1994). Silica-carbonate rocks form distinctive, massive, dark brown to orange-brown outcrops, more resistant to erosion than associated serpentinite. Parental ultramafic rock textures are commonly preserved and include orthopyroxene pseudomorphs ('bastite'), relict chromian spinel aggregates and zones of prior foliation. Many silica-carbonate rocks contain small green clots of fuchsite or chlorite. There is a gradation from

T.N.

' Mesothermal gold ^J vein clusters B Bingara UB Upper Bingara PY Paling Yard W Woodsreef CM Crow Mountain A Attunga N Nundle

Main deposits of silica-carbonate rock 1 Whitlow Creek 2 Bingara 3 Spring Creek 4 Stoddarts Valley 5 Barrack Creek 6 Upper Bingara 7 Piedmont 8 Piedmont South 9 Basin 10 Scotia 11 Paling Yard 12 Crow Mountain 13 Halls Creek 14 Bowling Alley Point T5 Trevena 16 Hanging Rock

Mesozoic and Tertiary cover rocks l-type granitoids (—245-255 Ma) X v X |

\

S-type granitoids (-290 Ma) 1 Great Serpentinite Belt (Weraerai terrane) 1 Accretionary complex (Djungati terrane) Accretionary complex (Anaiwan terrane) Accreted oceanic island arc (Gamilaroi terrane)

Figure 2 The major segment of the Great Serpentinite Belt (Weraerai terrane of Flood & Aitchison 1988) between Warialda and Nundle, southern New England Orogen, showing location of main occurrences of silica—carbonate rock and mesothermal gold vein clusters.

SILICA-CARBONATE ROCKS, NSW

carbonate-rich into local quartz-rich rock and all may be cut by late, extensional fracture systems occupied by veins, stockworks and breccia masses of quartz and/or magnesite. Where strongly weathered, silica-carbonate rocks have been reduced to porous quartz-hematite-goethite leached cappings. At several places, e.g. Crow Mountain, Paling Yard, Trevena mine (Figure 2), these contain anomalous As, Sb and Au values and eluvial and alluvial Au workings occur nearby. In addition, cinnabar was reported in association with silicacarbonate rocks at Spring Creek and Crow Mountain (Barnes et al 1988). Age relationships The timing of silica-carbonate rock formation throughout the Great Serpentinite Belt can, in general, be only inferred indirectly. All occurrences of silicacarbonate rock are massive and earlier foliations are pseudomorphed; their formation has therefore post-dated the development of serpentinite-matrix melange and emplacement age of the ophiolitic rocks (Late Carboniferous-Early Permian). In the Nundle area (Figure 2), silica-carbonate rocks have replaced antigorite metaserpentinite developed in the contact metamorphic aureole of the Duncans Creek Trondhjemite, an Early Triassic member of the New England Batholith with Rth-Sr biotite ages of 246-248 ± 3 Ma (Shaw & Flood 1993). In this area, therefore, silica-carbonate rock development is Early Triassic or younger. This inference has been confirmed at the Trevena mine and Hanging Rock in the Nundle area, where whole rock K-Ar dates on fuchsite-bearing silica-carbonate rocks have yielded ages of 243 ± 2 Ma and 251 ± 2 Ma respectively (Ashley & Brownlow 1993). A temporal link is thus implied between at least some of the silica-carbonate rock formation in the Great Serpentinite Belt and the intrusion of Late Permian to Early Triassic granitoids of the New England Batholith and lamprophyre dykes. Economic significance Historical gold production of at least 11.2 t has been derived from the Bingara to Nundle segment of the Great Serpentinite Belt and immediately adjacent terranes (Markham 1975). Alluvial and eluvial sources produced most gold, but at least 2.5 t has been mined from hard-rock mesothermal deposits (Figure 2), which, when degraded, have supplied some of the surficial gold. At many places, silica-carbonate rock crops out. It is from these, and locally associated altered mafic rocks, that eluvial and alluvial gold is being shed at BingaraSpring Creek, Borah Creek, Barrack Creek, Paling Yard, Emello, Crow Mountain, Trevena mine and Hanging Rock (Figure 2). Cinnabar occurs in association with serpentinite at Spring Creek and Crow Mountain (MacNevin 1974) and has been found in silica-carbonate rock at the former location (Peter English, pers. comm. 1996). The Piedmont magnesite deposit (Figure 2) occurs in a fractured silica-carbonate

215

body (Brownlow & Ashley 1991) and consists of hydrothermal magnesite (-quartz-dolomite) veins and breccia infillings with locally anomalous values of As, Hg and Au. Vein and breccia textures are consistent with those at epithermal precious metal (Dong et al. 1995) and vein-type magnesite deposits (Pohl 1990; Abu-Jaber & Kimberley 1992). PETROGRAPHY AND MINERALOGY

Petrographic features The most common host for silica-carbonate rock formation in the Great Serpentinite Belt is massive to schistose, completely serpentinised harzburgite. Cumulate ultramafic rocks of wehrlitic affinity have also been local precursors. Serpentinites largely consist of lizardite, with minor chrysotile, magnetite, ferrichromite, relict chromian spinel and traces of awaruite, pentlandite, millerite and heazlewoodite. Relict olivine (Fo ) and enstatite (En ) are rare and antigorite is mostly restricted to contact metamorphic aureoles near Nundle and Attunga. Contacts between silica-carbonate rock and protolith ultramafic rocks are commonly sharp, but may grade over a few metres from unaltered serpentinite through a zone of carbonate porphyroblast development into massive silica-carbonate rock. The latter is pale grey or buff when fresh and dominated by an aggregate of ferroan magnesite and quartz. Proportions vary, and although magnesite is typically more abundant than quartz (70:30 ratio), types with up to 95 vol.% quartz also occur. Both minerals mimic parental serpentinite texture and there is preservation of relict chromian spinel (including delicate symplectites), rimmed by ferrichromite. Other minor to trace phases include ferroan dolomite, fuchsite, talc, chlorite, graphite and sulfide minerals. Talc-bearing rocks are uncommon and chlorite occurs in two different assemblages. A bluegreen CrNi-bearing chlorite occurs sparsely in silicacarbonate rocks in the northern part of the Great Serpentinite Belt, whereas a dark green Cr-bearing chlorite is found in the Trevena mine and Hanging Rock areas, with the latter rocks displaying relict textures from an ultramafic cumulate precursor. Fuchsite is found as aggregates intergrown with quartz, carbonate or chlorite, in part replacing relict chromian spinel. Sulfides contain Fe, Ni, Co, Cu, As, Sb and Hg, with disseminated pyrite being the most common. As-bearing phases, including gersdorffite and arsenopyrite, occur sparsely and there is replacement of rare pyrrhotite by marcasite, and pentlandite by millerite or violaritepolydymite phases. Vein, stockwork and breccia fillings occupying brittle fractures in silica-carbonate rocks include two main types: (i) quartz-rich types, generally up to 10 cm wide, with variable amounts of magnesite, dolomite, talc, fuchsite and pyrite, and (ii) uncommon elongate zones up to several metres wide of massive cryptocrystalline low-Fe magnesite, with minor quartz and dolomite. The latter type is well exemplified at the Piedmont magnesite deposit. 90

90

216

P.M.ASHLEY

Table 1 Summary of carbonate, silicate and sulfide mineral chemical data from silica-carbonate rocks in the Great Serpentinite Belt. Carbonates Magnesite (n = 71). CaC0 3 = 0.0-1.2, MgC0 3 = 71.9-98.5, FeC0 3 = 1.5-27.9 (molecular %) Dolomite (n = 35). CaC0 3 = 46.7-53.9, MgC0 3 = 35.0-49.6, FeC0 3 = 1.2-14.3 (molecular %) Magnesite contains up to 0.66 wt% NiO and 0.47 wt% MnO Mica (n = 40) All chromian phengite (fuchsite) with Si lv = 6.276-7.116 atoms per formula unit. Cr 2 0 3 values range from 0.21-9.05 wt% and mg* values from 69.0-95.2. Chlorite Nundle area (n = 21). All pynochlorite with mg = 76.3-80.0 and Cr 2 0 3 = 0.16-2.35 wt%. Paling Yard (n = 8). All penninite with mg= 90.8-96.9, Cr 2 0 3 = 2.37-6.17 wt% and NiO = 1.31-8.70 wt%. Sulfides Pyrite (n = 13). 0.13-2.91 wt% Ni, 0.04^0.19 wt% Co, 0.06-0.53 wt% As. Gersdorffite (n = 2). Nio g^^Feoo^ogCoo.oi Aso ^ ^ S b o o j ^ ^ S ] 06_j 10 Millerite. Ni0.85Fe0.07Co0.03Sj.04 Cobaltian violarite. Ni, 77Feo83Coo.32S4o7 Cobaltian polydymite. Ni2.6iCo0 pFeooyS^ Arsenian vaesite. Ni0.93Fe0.03As0.25S, 79 * mg = 1 OOMgO/(MgO + IFeO) atomic. Electron microprobe analyses performed on ETEC and JEOL instruments at Macquarie University and University of New England respectively. Analysts: P. M. Ashley and R. B. Breyley.

Mineral chemistry

analyses qualify as fuchsite, rather than mariposite (Max et al. 1983). Dark-green chlorite in altered ultramafic ?cumulate rocks from the Nundle area is pycnochlorite, whereas the blue-green chlorite from Paling Yard is penninite with high Cr 2 0 3 and NiO contents. The higher mg values of the latter reflect a harzburgitic precursor for the host silica-carbonate rock. Analyses of several sulfide phases (Table 1) show minor substitution of Ni, Co and As into pyrite and Fe, Co and Sb into gersdorffite. Millerite, violarite, polydymite and vaesite also display minor element substitutions. Discrete particles of gold have only been noted in supergene oxidised material (e.g. at Hanging Rock). In unoxidised rock, gold may be cryptically held in sulfides and a pyrite concentrate from the Trevena mine contained 13.5 ppm Au.

Mineral compositions in silica—carbonate rocks were obtained by electron microprobe analysis and results are summarised in Table 1, with data for carbonates also shown on Figure 3. Magnesite ranges from near endmember (98.5 mol% MgC0 3 ) to strongly ferroan (up to 27.9 mol% FeC0 3 ) compositions. Dolomite compositions also range from near end-member to ferroan types (up to 14.3 mol% FeC0 3 ). Although there is a wide range in magnesite compositions {mg - 72.1-98.5, where mg = 100MgO/(MgO + IFeO) atomic), the distribution is bimodal. Magnesite with mg values >87 (average 93) occurs in altered serpentinised harzburgite, whereas that with mg values <87 (average 77) is restricted to altered ultramafic cumulate rocks. Clearly, magnesite compositions mimic the pre-alteration protolith composition. The conspicuous green mica in many silica-carbonate rocks is chromian phengite. Cr 2 0 3 contents are mostly between 1 and 3 wt%, but attain 9.05 wt%. Molecular Si per formula unit is commonly between 6.3 and 7.0, with Al iv >1.0 and thus most

GEOCHEMISTRY Serpentinised harzburgite, the most common protolith for silica-carbonate rock in the Great Serpentinite Belt,

30 mol % FeCOs + MnCOs

magnesiteferroan magnesite (71 analyses)

dolomiteferroan dolomite (35 analyses)

CaC03

Dolomite

Figure 3 Compositions of carbonate minerals from silicacarbonate rock in the Great Serpentinite Belt in part of the system CaC0 3 -MgC0 3 -<FeC03 MgCOs + MnC0 3 ).

S I L I C A - C A R B O N A T E ROCKS, NSW

217

Table 2 Chemical compositions of ultramafic and silica-carbonate rocks from the Great Serpentinite Belt

Si0 2 Ti0 2 A1203 Fe 2 0 3 FeO MnO MgO CaO Na 2 0 K20

P2O5 LOI Total Ba Rb Sr Y Zr Nb U Th Ce Nd La Ga Sc V Ni Cr Cu Pb Zn As Sb Hg Au mg Fe 2 0 3 /Fe0 SG

1

2

3

4

5

6

7

40.02 0.03 0.73 4.92 2.02 0.10 38.44 0.22 0.08 0.03 0.01 13.08 99.46

37.63 0.09 5.99 5.23 3.15 0.13 32.66 3.23 0.51 0.03 0.01

30.31 0.02 0.67 0.99 3.55 0.08 27.34 2.96 0.04 0.01 <0.01 33.86 99.83

96.39 0.02

34.80 0.02 1.15 1.32 3.61 0.09 25.69 2.21 0.34 0.14 0.02 30.54 99.93

51.59 0.05 4.56 0.77 5.19 0.16 10.31 5.89 0.05 1.19 0.02 19.16 98.94

43.87 0.07 5.54 1.36 6.13 0.15 13.47 6.77 0.04 1.14 0.01 20.78 99.33

30 6 110 <2 2 <2 <2 <2 7 <2 <2 <2 2 26 1407 2028 11 2 23 109 12 0.44 0.082

49 23 131 <2 12 <2 <2 <2 5 <2 4 3 17 97 149 1360 37 4 76 28 8 na 0.54

50 16 164 2 12 <2 na na <2 na 2 3 6 84 449 2116 22 4 54 20 na na 0.25

90.5 0.37 2.87

75.7 0.15 2.82

76.5 0.22 2.88

10 <2 2 <2 <2 <2 <2 <2 <2 <2 3 <2 7 28 2286 2953 2 <2 39 5 na na na 91.4 2.44 2.50

11.28 99.94 <5 <2 29 <2 <2 <2 <2 <2 <2 <2 4 4 11 38 1316 2514 50 2 44 3 na na na 88.1 1.66 2.65

7 <2 118 <2 <2 <2 <2 <2 <2 <2 5 <2 5 32 1182 2411 10 2 23 473 25 0.05 0.035 91.6 0.28 2.91

1.00 0.08 0.09 0.01 0.34 0.04 <0.01 0.23 0.02 0.59 98.81 na 14 16 4 3 <2 na na na na na na na 30 855 1356 12 22 37 325 na na na 79.2 0.89 2.66

Most major elements and trace elements analysed by XRF at the University of New England, together with FeO (titrimetry) and LOI (gravimetry). Analyst: J. Bedford. Hg and Au analysed at Australian Laboratory Services, Brisbane, using ICP and AASgraphite furnace methods, respectively. LOI = loss on ignition (includes H 2 0, C0 2 , S); na = not analysed; mg= l()OMgO/(MgO + XFeO) atomic; SG = specific gravity in g/cm 3 . 1. Average of 14 serpentinised harzburgites from the Nundle to Bingara segment of the Great Serpentinite Belt. 2. Average of 7 partly serpentinised ultramafic cumulates (wehrlite) from the Upper Bingara region, Great Serpentinite Belt. 3. Representative low-Al silica-carbonate rock (magnesite + quartz + dolomite + chlorite), Spring Creek, Bingara. GR 9038688884. Sample 820939. 4. Representative quartz-rich altered serpentinite, Mummel River goldfield, GR 9234-003063. Sample R59629. 5. Average of 17 low-Al silica-carbonate rocks from Great Serpentinite Belt. 6. Representative high-Al silica-carbonate rock (dolomite + chlorite + quartz + fuchsite + pyrite), Hanging Rock/GR 9135274156. Sample HR1. 7. Average of 9 silica-carbonate rocks (high-Al type) from the Hanging Rock district, Great Serpentinite Belt.

218

P.M.ASHLEY

has a composition typical of ophiolitic serpentinites worldwide, with mg = 91.4, Si0 /Mg0 and Ni/Cr close to unity, high MgO/CaO, Fe 0 /Fe0 and H 0, and low A1 0 and high field strength elements (Table 2, column 1). Alteration of this precursor has led to the formation of iow-Al' silica-carbonate rocks. However, in places (e.g. Nundle area), other types of ultramafic rocks, such as cumulate wehrlite (Table 2, column 2) have been altered, giving rise to 'high-Al' silica-carbonate rocks. In contrast to harzburgite-derived serpentinite, wehrlite contains higher A1 0 , CaO, Na 0, Sr, Sc, V and Cu, and lower MgO and Ni (Table 2). The replacement of ultramafic rocks by silicacarbonate rock involves large chemical changes (Table 2), which have been explored using a mass-balance approach. Average compositions of serpentinised harzburgite and partly serpentinised wehrlite were compared against average compositions of their interpreted alteration derivatives, the low-Al and high-Al silicacarbonate rocks, respectively (Table 2). Major element oxide and trace-element concentrations for each protolith-derivative pair were plotted on to isocon diagrams of Grant (1986). Results show that for the alteration of serpentinised harzburgite to low-Al silicacarbonate rock (Figure 4a), Ti0 , MnO, ZFeO, MgO, Si0 and Cr have remained relatively immobile, whereas for the alteration of partly serpentinised wehrlite to high-Al silica-carbonate rock, Ti0 , A1 0 , ZFeO and Cr have remained relatively immobile (Figure 4b). Gains and losses of components may be calculated (Table 3), using the relatively immobile element isocons as reference lines, and are graphically represented on Figure 4. From these results, it is evident that FeO, CaO, K 0, LOI (C0 ), Sr and As (and by inference, Au, Sb and S) have been strongly enriched during alteration. There have also been modest gains of Si0 , A1 0 , MnO and V. Na 0, Cu and Zn display erratic behaviour, with Na 0 and Cu being enriched in low-Al silica-carbonate rocks, but depleted in the highAl type, relative to their protoliths. Zn displays opposite behaviour. MgO is weakly depleted by alteration, but losses of Ni, and Fe 0 in particular, are more substantial. The latter point illustrates the fact that alteration to form silica-carbonate rocks is reductive, leading to a decrease in the Fe 0 /Fe0 ratio (Table 2). The apparent mobility (gain) of A1 0 , a component commonly considered to be immobile, during alteration of serpentinised harzburgite may be due to the relatively small sample base and variability of low A1 0 , concentrations (~1% ± 0.5%), rather than being a genuine addition. The bulk chemical changes and observed element behaviour in the conversion of ultramafic rock to silicacarbonate rock in the Great Serpentinite Belt are similar to those found in other ophiolites (Bohlke 1989; Auclair et al. 1993; Tuysiiz & Erler 1993) and in Archaean greenstone belts (Kishida & Kerrich 1987). They are also analogous to wallrock alteration patterns about mesothermal lode gold deposits (Bohlke 1989; Groves & Foster 1991). From the exploration geochemistry viewpoint, it is significant that silica-carbonate rock is anomalous in 2

2

2

2

Low-AI protolith (serpentinized harzburgite)

30 •

(b)

2

3

2

2

JsC

o

0) 2 2 0 W — Ca < o O) CO X V(0 o

3

3

2

4 Si02/2

Cr/100

2

*Sr/10

lOOMnOA

a V / 5

^

^

A 2CaO A2FeO 2AI203

k 10K2O

A MgO/2

2

5Fe203 A

50Na20 A

2

30

High-AI protolith serpentinized wehrlite)

(partly

2

2

3

2

2

2

Figure 4 Isocon plots after Grant (1986) showing chemical effects of silica-carbonate rock alteration of ultramafic rocks in the Great Serpentinite Belt, (a) Major element oxide (in %) and trace element (in ppm) concentrations in average serpentinised harzburgite (horizontal axis) plotted against concentrations in low-Al silica-carbonate rocks. Circled components (Ti0 , MnO, ZFeO, MgO, Si0 , Cr) are interpreted to be relatively immobile and define an isocon line, with altered rock/protolith (A/P) = 0.78, implying a mass increase of 28% and a volume increase of 23% using equations of Grant (1986). (b) Major element oxide (in %) and trace element (in ppm) concentrations in average partly serpentinised wehrlite (horizontal axis) plotted against concentrations in high-Al silica-carbonate rocks. Circled components (Ti0 , A1 0 , ZFeO, Cr) are interpreted to be relatively immobile and define an isocon line with A/P = 0.86, implying a mass increase of 16% and a volume increase of 7%. Note that components above isocon lines were gained during alteration and those below were lost. The isocon lines were also used as references to calculate the gains and losses shown in Table 3. 2

2

2

2

3

3

2

3

2

3

2

3

2

3

SILICA-CARBONATE ROCKS, NSW

219

Table 3 Calculated weight percent component gains and losses in the conversion of serpentinised harzburgite to low-Al silicacarbonate rock and partly serpentinised wehrlite to high-Al silica-carbonate rock using method of Grant (1986).

Protolith: average serpentinised harzburgite (Table 2, column 1). Altered rock: average low-Al silica-carbonate rock (Table 2, column 5). Best-fit isocon, with A/P = 0.78, using Ti0 2 , MnO, IFeO, MgO, Si0 2 , Cr.

Si0 2 Ti0 2 A1203 Fe 2 0 3 FeO IFeO MnO MgO CaO Na 2 0 K20 LOI Ba Sr Sc V Ni Cr Cu Zn As

+11% -14% +102% -66% +129% -5% +15% -14% +1190% +440% +500% +200% +280% +6940% -

+19% -21% -12% +600% -25% +2690%

Au, As, Sb and Hg (Table 2). Quartz-rich veins in silica-carbonate rock commonly contain Au values in the range 0.2-2 ppm and porous ferruginised leached capping over silica-carbonate rock at Crow Mountain and Paling Yard (Figure 2) contains average values of Au of 14 ppb and 26 ppb, As 377 ppm and 474 ppm, and Sb 61 ppm and 62 ppm, respectively (Lawie 1993). The degradation of this material has led to the formation of eluvial and alluvial gold deposits at several locations. FLUID-INCLUSION OBSERVATIONS No fluid inclusions of size large enough for microthermometric measurement were noted in typical silicacarbonate rock. However, they do occur in cross-cutting vein quartz and carbonates. As veins contain the same hydrothermal minerals as the enclosing rocks and alteration assemblages are zoned about veins in cumulate ultramafic rocks, it is argued that they formed by the same hydrothermal process and at a similar or slightly later time. Fluid-inclusion data from veins may thus give a guide to silica-carbonate rock formation conditions. A limited set of homogenisation temperature (Th) and freezing point depression (T m ) determinations were obtained on four samples from the Nundle area (Trevena mine and Hanging Rock), using an SGE heating-freezing stage. Fluid inclusions were chosen for measurement on the basis of petrographic criteria indicating likely primary characteristics (Roedder

Protolith: average partly serpentinised wehrlite (Table 2, column 2). Altered rock: average high-Al silica-carbonate rock (Table 2, column 7). Best-fit isocon, with A/P = 0.86, using Ti0 2 , A1203, IFeO, Cr.

+32% -10% +4% -70% +126% +5% +30% -52% +140% -91% +4310% +110% -

+550% -37% +150% -60% -2% -49% +38% +670%

1984). All inclusions measured were in quartz; they are up to 40 jim across and are simple two-phase types with relatively constant liquid:vapour volume proportions of 80:20 to 90:10, homogenising to the liquid phase. Results (Figure 5) indicate a total range of T h from 156° to 308°C, with mean T h s of individual samples being between 190°C and 243°C and the majority of data falling in the range 165°-255°C. From the freezing point depression of inclusion liquids (-2.2° to -3.8°C), calculated salinities range between 3.7 and 6.1 equivalent wt% NaCl, using the equation of Bodnar (1992). Since the veins and wallrocks contain carbonates, it is assumed that the fluids are C0 2 -bearing and are not simple H 2 0-NaCl solutions. Although no clathrates formed in fluid inclusions during freezing experiments, other spatially and temporally related Au-quartz veins in the Nundle district contain three-phase, liquid-C0 2 bearing inclusions (Ashley & Hartshorn 1988). Therefore, the apparent salinities may be too high due to dissolved C0 2 (Hedenquist & Henley 1985; Bodnar et al. 1985). Estimation of depth and pressure of formation for these veins is difficult due to lack of evidence for fluid immiscibility. A minimum depth of formation may be constrained by boiling point-depth curves, but these are dependent on the proportion of dissolved C 0 2 and could range for dilute NaCl solutions at 200°-250°C from as little as about 150 m (if C0 2 -free) to about 1100 m with 0.5 m dissolved C0 2 (Bodnar et al 1985). In any case, veins in silica-carbonate rocks throughout the Great Serpentinite Belt fill brittle fractures, show

220

P.M.ASHLEY

5

R57911

5_

T4

n = 14

mean = 190°C

n = 18 mean = 243 43°C

I J

• •

5

n = 25 mean = 216°C

820852

15

—•

HR 1

n = 83 mean = 207°C

10 _

5 _

150

Th °C open-space filling textures (vughs) and locally contain banded chalcedonic quartz and carbonate, characteristics typical of relatively shallow crustal depths for hydrothermal activity. Due to these uncertainties, no pressure correction is applied to the fluid-inclusion homogenisation data and therefore measured T h values must represent minimum trapping temperatures. STABLE ISOTOPE RESULTS Stable isotope data on carbonates, quartz and pyrite in silica—carbonate alteration zones in the Great Serpentinite Belt have been obtained using standard gas extraction and mass spectrometric procedures at the Centre for Isotope Studies, CSIRO, North Ryde. Results for 1 8 0 / 1 6 0 , 1 3 C/ 1 2 C and 3 4 S / 3 2 S , using the conventional 5 notation relative to VSMOW, VPDB and CDT standards, are presented in Table 4 and 5 1 8 0 and 5 13 C data for carbonates are plotted on Figure 5. Carbonates show a range of 5 13 C of -8.0 to 1.3%o and 5 1 8 0 of 12.2—19.6%o, quartz 5 1 8 0 ranges from 15.7 to 21.8%o and pyrite 5 34 S from 0.8 to 3.8%o. If an average minimum temperature of 225°C is assumed from fluidinclusion data (see above) for fluids responsible for formation of silica-carbonate rocks in the Great Serpentinite Belt, calculated 5 1 8 0 H 2 o values are in the range 2.2— 10.4%o (mean 6.0%o). The results are based on measured 5 l s O values in quartz and carbonate (Table 4) and calculated respectively using quartz-water

Figure 5 Histograms of homogenisation temperature (T h ) of fluid inclusions from quartz (—carbonate) veins in silica-carbonate rocks at the Trevena mine and Hanging Rock areas, Nundle goldfield. On the right-hand side are histograms showing equivalent weight percent NaCl of inclusion fluids determined from ice melting temperatures (Tm) using the equation of Bodnar (1992). In sample R5791 1 (Trevena mine), Th range is 160°—205°C, mean Th = 190°C (n = 14); Tm range is -3.1° to -3.4°C (5.0-5.5 equivalent wt% NaCl) (n = 5). In sample T4 (Trevena mine), Th range is 220°—271 °C, mean Th = 243°C (n = 18). In sample 820852 (Hanging Rock), Th range is 156°—262°C, mean Th = 216°C (n = 25); Tm range -3.4° to -3.7°C (5.5-6.0 equivalent wt% NaCl) (n = 4). In sample HR1 (Hanging Rock), Th range is 168°—308°C, mean Th = 207°C (n = 83); Tm range is -2.2° to -3.8°C (3.7-6.1 equivalent wt% NaCl) (aI = 25).

fractionation data in Matsuhisa et al. (1979) and extrapolated magnesite-water fractionation data in Aharon (1988). The 5 1 8 0 values and 5 1 8 0 H 2 q range are compatible with either or both a magmatic or metamorphic fluid source and also with l 8 0 data from carbonates and quartz associated with mesothermal vein Sb-Au and Au deposits in the southern New England Orogen (Robinson & Farrand 1982; Ashley et al. 1994), and mesothermal Au and Sb-Au deposits and silicacarbonate rocks elsewhere (Bohlke & Kistler 1986; Nesbitt et al. 1989; Madu et al. 1990; Nesbitt 1991; Auclair et al. 1993). Carbonate 8 1 3 C in silica-carbonate rock from the Great Serpentinite Belt has a mean value of -4.0%o (Table 4). Using the approximations calculated for magnesite-C0 2 fractionation by Zachmann and Johannes (1989), the C 0 2 responsible for formation of silica-carbonate rock at 225°C has an average § 1 3 C value of -5.4%0, consistent with homogenised crustal, magmatic or mantle sources (Ohmoto 1986; Golding et al. 1989; Rock 1991). Carbonate 6 1 8 0 and 5 13 C data overlap the field of carbonates from mesothermal S b Au and Au vein deposits in the southern New England Orogen (Figure 6) implying a commonality of fluid source. Although the trend to isotopically lighter 5 13 C values in vein deposits in the region (Figure 6) may reflect acquisition of carbon from a reduced sedimentary (organic) source (Ashley et al. 1994), such a source is less likely for the silica-carbonate rocks. Similarly, the

SILICA-CARBONATE ROCKS, NSW

221

Table 4 Stable isotope results from silica-carbonate rocks in the Great Serpentinite Belt.

Sample

Location

5 18 O qz

HR1 820852 820922 820923 820925 820052 82 820854 T1 T2 T4 T5 T6 T6WR 820758 820759 820776 R62052 R62241 R62243 R62658 820950 820951 R58038 R58040 R58043 R58050 820939 820940 820937 820938

Hanging Rock Hanging Rock Hanging Rock Hanging Rock Hanging Rock Trevena Trevena Trevena Trevena Trevena Trevena Trevena Trevena Trevena Munro Creek Munro Creek Alliance Crow Mountain Crow Mountain Crow Mountain Crow Mountain Paling Yard Paling Yard Spring Creek Spring Creek Spring Creek Spring Creek Spring Creek Spring Creek Bingara Bingara

16.7 21.4

Means

-

15.9 — -

15.7 15.7

8 13 C cb

5 18 O cb

-

-

—

-5.3 -4.4 -4.4 -3.6 -3.5

16.4 14.9 14.7 15.9 12.9 0.8 14.8 13.3

2.4 3.1 1.9 2.1

-

-4.1 -3.7 -

-

—

-5.8 -4.7 -4.0 -4.3 -5.6 -5.3 -2.3 -1.9 -2.1 -2.5 -4.9 -4.9 -4.1 -2.8 -8.0 -5.8 1.3 -5.2 -5.3 -2.6 -3.3

12.2 13.8 14.5 14.1 14.0 15.5 15.7 19.3 19.1 18.3 15.4 18.2 18.6 18.0 19.6 17.2 17.4 18.8 14.6 17.1 17.5

18.1

-4.0

16.1

-

17.3 20.3 -

21.8 -

8 34 S py

—

—

— —

—

3.8 — —

—

2.4

Results in %o. Analyses obtained at Centre for Isotope Studies, CSIRO, North Ryde. Oxygen for isotopic analysis was generated from quartz by reaction with BrF5 at 550°C (Clayton & Mayeda 1963). Carbonates (ferroan magnesite and ferroan dolomite) were reacted with '103 %' phosphoric acid at 100°C for four hours to generate C 0 2 (Aharon 1988). Pyrite was burnt with excess Cu 2 0 at 1000°C to produce S0 2 (Robinson & Kusakabe 1975). Isotopic analyses were carried out on modified Micromass 602 mass spectrometers. 513C values are relative to VPDB, 5 1 8 0 relative to VSMOW and 534S relative to CDT.

5 13 C values do not point to a significant C0 2 source from marine carbonates in the adjacent Gamilaroi terrane. Disseminated pyrite in several silica-carbonate rock occurrences has a limited range of 534S values (Table 4) and a mean 5 34 S value of 2.4%o. At 225°C, the H 2 S responsible for pyrite precipitation has a 5 34 S value of 0.8%o, using the pyrite-H 2 S fractionation data of Ohmoto and Rye (1979). The Great Serpentinite Belt results may indicate a magmatic or homogenised crustal sulfur source (Ohmoto 1986; Nesbitt 1991) and differ significantly from the isotopically heavy pyrite (534S values of 7.0-19.l%o) from silica-carbonate rocks reported by Auclair et al. (1993), for which a marine sulfate or sedimentary rock source for sulfur was favoured.

DISCUSSION Structural control and relationship with southern New England Orogen mesothermal gold deposits Silica-carbonate rocks in the Great Serpentinite Belt occur at tectonic contacts and therefore it is implied that there has been structural focusing of causative hydrothermal fluids. The majority of occurrences are located along the western contact of the Great Serpentinite Belt and as this is also the western margin of the Peel Fault system (Figure 2), the contact has been the locus for periodic reactivation and hydrothermal fluid channelling. Replacement of serpentinite has led to development of massive silica—carbonate rock which subsequently exhibited a competency contrast with

222 P.M.ASHLEY adjacent serpentinite-matrix melange. Brittle fracturing phenomena dominate in silica-carbonate rocks, with these structures controlling the late hydrothermal deposition of (Au-bearing) quartz and carbonate veins. Many bodies of silica-carbonate rock in the Great Serpentinite Belt contain anomalous Au, As, Sb and Hg (Table 2) and Au-bearing veins, and are located adjacent to mesothermal vein Au deposits hosted in diverse rock types (Figure 2). These phenomena imply that the fluids which formed the silica-carbonate rocks are the same as those responsible for the formation of mesothermal Au and Sb-Au deposits elsewhere in the southern New England Orogen. The link is consistent with similarities in wallrock alteration styles, stable isotopic data, structural controls and age constraints (Barnes et al 1988; Ashley & Hartshorn 1988; Ashley et al. 1994). Hydrothermal alteration of ultramafic rocks in the Great Serpentinite Belt is simply a variation on the overall theme of alteration about mesothermal Au deposits. Carbon and oxygen isotopic results from silica-carbonate rocks and southern New England Orogen mesothermal Sb-Au and Au deposits indicate a commonality of hydrothermal fluid (Figure 6). Fracture porosity along the tectonic margins of serpentinites may be a critical factor in silica-carbonate rock development, by the processes of fluid focusing and contact with the chemically reactive ultramafic rocks. Regionally, the timing of silica-carbonate alteration in the Great Serpentinite Belt can only be constrained as following emplacement of serpentinite-matrix

25 n

melange along the Peel Fault (i.e. probably post-Late Carboniferous). However, at Nundle, geochronological data indicate that alteration occurred at about 245250 Ma (Ashley & Brownlow 1993), demonstrating temporal equivalence with the Late Permian to Early Triassic formation of mesothermal Sb-Au and Au deposits in the southern New England Orogen (Barnes et al. 1988; Ashley et al. 1994). This period was late in the tectonic development of the southern New England Orogen, post-dating regional metamorphism and deformation, and overlapping temporally with regional uplift (Roberts & Engel 1987), intrusion of voluminous I-type granitoids (Shaw & Flood 1993) and lamprophyre dykes (Henley 1991; Ashley et al. 1994). It is interpreted that the silica-carbonate rocks and mesothermal Au and Sb-Au deposits in the southern New England Orogen are simply the relatively shallow crustal manifestations of deeply sourced hydrothermal systems developing in a former accretionary complex in a regime of rapid uplift and high heat flow, perhaps analogous to younger examples in New Zealand (Craw & Koons 1989; Craw & Norris 1991) and California (Barnes et al. 1973; Peters 1991). Conditions of silica-carbonate rock formation Fluid-inclusion data from Great Serpentinite Belt silicacarbonate rocks indicate minimum fluid temperatures in the range of 165°-255°C, conditions which overlap those determined from southern New England Orogen Sb-Au and Au veins (range 100°-350°C: Comsti & Taylor 1984; Ashley & Hartshorn 1988; author's unpubl. data). Fluids are dilute, low-NaCl (<6 equivalent wt%) with less than 0.85 m C0 (Hedenquist & Henley 1985). Veins and breccia zones contain openspace-filling and crustiform banded textures implicit of shallow formation depths (e.g. <5 km). Therefore, pressures are inferred to have been below 100— 200 MPa. The assemblage magnesite + quartz typifies silica-carbonate rocks, and being stable at all but very low mole fractions Xco (e.g. <0.03) at this pressure, and temperatures below about 350°C (Johannes 1969), is consistent with the fluid-inclusion data. Two reactions have occurred to form silica-carbonate rock from serpentinite: (i) serpentine + magnetite + C0 ± H S ferroan magnesite + quartz ± sulfides, and (ii) serpentine + magnetite + C0 ± H S -> ferroan magnesite + talc ± quartz ± sulfides. Reaction (ii) has limited manifestation (e.g. Trevena mine) and may reflect higher temperatures of formation than reaction (i) (Johannes 1969). The presence of Al in protolith material has influenced formation of chlorite and introduction of K in hydrothermal fluids has been responsible for fuchsite crystallisation. The fluids also sourced S, As, Sb, Hg and Au, causing sulfidation reactions. Fe, Ni and Co were sequestered from serpentine minerals and magnetite to form sulfides. Fluids must have been relatively reducing as whole-rock Fe 0 /Fe0 ratios were diminished (Table 2) and magnetite destroyed. 2

2

-15

-10 )

j

-5

1

0

5

Figure 6 Oxygen and carbon isotopic compositions of vein and alteration carbonates from silica-carbonate rock and mesothermal Au and Au-Sb deposits in the southern New England Orogen. Field 1 shows the extent of isotopic values from 42 carbonates (ankerite, dolomite and calcite) from mesothermal vein deposits (Ashley et al. 1994). Field 2 outlines individual carbonate isotopic values of ferroan magnesite and ferroan dolomite (x) from silica-carbonate rock in the Great Serpentinite Belt (see also Table 4). Field 3 represents the isotopic range of carbonatites from Rock (1991).

2

3

2

2

2

2

S I L I C A - C A R B O N A T E ROCKS, N S W 223 In summary, it is estimated that silica-carbonate rock C0 , Si0 , CaO, K 0, Ba, Sr, As, Sb, Hg, Au and S, formation occurred at minimum temperatures of 165°- loss of H 0, MgO and Ni and reduction of Fe 0 to 255°C from fluids which had a deep-seated source of FeO. Mass and volume increases are implied during possible metamorphic (homogenised crustal material) alteration. (4) Fluid inclusions in vein quartz in silica-carbonate and/or magmatic derivation. Fluids were C0 -bearing, of low salinity, reducing (probably HS- as main S rocks indicate that the fluids had minimum temperatures species) and probably close to neutral (Craw 1989). in the range 165°-255°C, were of low salinity and Reducing conditions are consistent with the serpentinite probably C0 -bearing. Mineralogical criteria also show environment (Frost 1985) and favour transportation of that the fluids were relatively reducing and near neutral. Au as a bisulfide complex whose solubility reaches a (5) Isotopic compositions of carbonates (5 C = -8.0 maximum in near-neutral conditions (Seward 1989; to 1.3%o, 5 0 = 1 2 . 2 — 1 9 . 6 % o ) , quartz (5 0 = 1 5 . 7 Peters 1991). Multiple gold-precipitation mechanisms 2 1 . 8 % o ) and pyrite (5 S = 0 . 8 - 3 . 8 % o ) are consistent are likely and may have included temperature decrease with a deep-seated metamorphic (homogenised crust) and wallrock reactions such as: (i) C0 fixation by Mg and/or magmatic source for hydrothermal fluids. and Fe in serpentinite, leading to precipitation of (6) Gold enrichment in silica-carbonate rock was ferroan magnesite and decrease of fluid pH; (ii) related likely to be due to precipitation of gold from bisulfide increase in fluid f due to solution of magnetite; and complexes as a result of carbonatisation and sulfidation (iii) sulfidation reactions involving Fe (NiCo) in reactions, leading to decreases in fluid pH and / . serpentine minerals and magnetite causing lowering of (7) Hydrothermal fluid focusing along the Peel Fault / Each of these parameters is effective in reducing Au led to silica-carbonate rock formation at relatively high solubility as bisulfide complex in hydrothermal fluids crustal levels at about the Permian-Triassic boundary, (Seward 1989). in the terminal orogenic stage of the southern part of the New England Orogen, correlating with regional uplift Exploration implications and intrusion of I-type granitoids and lamprophyre dykes. Silica-carbonate rock formation and associated Silica-carbonate rocks in the Great Serpentinite Belt mineralisation is another manifestation of the widehost low-grade gold disseminations and local higher spread structurally controlled mesothermal Au and Augrade concentrations in vein, stockwork and breccia Sb vein deposits in the region. zones. Several bodies have been explored for gold potential by drilling. It is proposed that high reactivity ACKNOWLEDGMENTS of serpentinite with hydrothermal fluids has caused gold to be strongly dispersed and therefore pervasive high Much of the data acquisition for this work was grades would not be expected. Values of 10-100 ppb supported by Australian Research Council grant Au, 10-500 ppm As and 5-50 ppm Sb are typical of A38715500. Field and laboratory studies by Jeff silica-carbonate rock. However, two situations may Brownlow, Greg Hartshorn, Dave Lawie, Greg Rogers, occur to develop higher grades of gold: (i) where there Steve Nano, St John Herbert and Richard Breyley added has been strong fracturing of silica-carbonate rock, gold to the knowledge of the silica—carbonate rocks. may have been hydrothermally remobilised and Anita Andrew and Andrew Bryce of the Centre for precipitated into vein, stockwork and breccia systems; Isotope Studies, CSIRO are thanked for assistance in and (ii) where weathering processes have caused obtaining the stable isotope results and Riki Davidson supergene enrichment of gold above silica-carbonate expertly performed fluid-inclusion determinations. rocks, forming eluvial and leached capping-hosted I thank Rob Barnes, Ed Mikucki, John Ridley, John deposits, further reworking by fluvial action will lead to Walshe and Kai Yang for their comments on drafts of this paper. the formation of placer deposits. 2

2

2

2

2

3

2

2

13

18

18

34

2

2+

0 l

S2

S r

CONCLUSIONS

REFERENCES

(1) Bodies of hydrothermal silica-carbonate rock are common along tectonic contacts in the Great Serpentinite Belt (Weraerai terrane) in the New England Orogen of northeastern New South Wales. They have been derived by replacement of serpentinite melange (dominantly serpentinised harzburgite, but also including wehrlitic protoliths). (2) Silica-carbonate rocks are pervasively anomalous in As and Sb, and contain low-grade Au values and traces of cinnabar. Weathering and erosion have led to formation of several eluvial and alluvial gold deposits proximal to silica-carbonate masses. (3) Replacement of ultramafic protolith by a ferroan magnesite + quartz (± fuchsite ± chlorite ± dolomite ± FeNiCoAs sulfide) assemblage involved addition of

ABU-JABER N. S. & KIMBERLEY M. M. 1992. Origin of

ultramafic-hosted vein magnesite deposits. Ore Geology Reviews 7, 155-191.

AHARON P. 1988. A stable isotope study of magnesites from

the Rum Jungle uranium field, Australia: implications for the origin of strata-bound massive magnesites. Chemical Geology 69, 127-145.

AITCHISON J. C., IRELAND T. R., BLAKE M . C. JR. & FLOOD P.

G. 1992. 530 Ma zircon age for ophiolite from the New England orogen: oldest rocks known for eastern Australia. Geology 20, 125-128.

AITCHISON J. C., BLAKE M . C. JR., FLOOD P. G. & JAYKO A.

S. 1994. Paleozoic ophiolitic assemblages within the southern New England orogen of eastern Australia: implications for growth of the Gondwana margin. Tectonics 13, 1135-1149.

224

P.M.ASHLEY

AITCHISON J. C . & I R E L A N D T . R .

1995. A g e profile of

CLAYTON R. N . & MAYEDA T . K . 1 9 6 3 . T h e u s e o f b r o m i n e

ophiolitic rocks across the Late Palaeozoic New England Orogen, New South Wales: implications for tectonic models. Australian Journal of Earth Sciences 42, 11-23.

pentafluoride in the extraction of oxygen from oxides and silicates for isotopic analysis. Geochimica et Cosmochimica Acta 27, 43-52.

ASHLEY P. M . & BROWNLOW J. W . 1 9 9 3 . S i l i c a - c a r b o n a t e

COMSTI E. C. & TAYLOR G. R. 1984. Implications of fluid

alteration zones in the Great Serpentinite Belt, southern New England Orogen: their nature and significance. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 197—214. Department of Geology and Geophysics, University of New England, Armidale.

inclusion data on the origin of the Hillgrove goldantimony deposits. N.S.W. Proceedings of the Australasian Institute of Mining and Metallurgy 2 8 9,

ASHLEY P. M . , COOK N . D. J., HILL R. L. & KENT A . J. R.

1994. Shoshonitic lamprophyre dykes and their relation to mesothermal Au-Sb veins at Hillgrove, New South Wales, Australia. Lithos 32, 249-272. ASHLEY P. M . & HARTSHORN G . K . 1 9 8 8 . G e o l o g i c a l a n d

geochemical characteristics of lode gold deposits in the Nundle district, northeastern New South Wales. In: Kleeman J. D. ed. New England Orogen—tectonics and metallogenesis, pp. 249-263. Department of Geology and Geophysics, University of New England, Armidale. AUCLAIR M . , GAUTHIER M . , TROTTIER J., JEBRAK M .

&

CHARTRAND F. 1993. Mineralogy, geochemistry, and paragenesis of the Eastern Metals serpentinite-associated Ni—Cu-Zn deposit, Quebec Appalachians. Economic Geology 88, 123-138. BARNES I., O'NEIL J. R . , RAPP J. B. & WHITE D . E. 1 9 7 3 .

Silica-carbonate alteration of serpentine: wall rock alteration in mercury deposits of the California Coast Ranges. Economic Geology 68, 388-398. BARNES R. G . , BROWN R . E., BROWNLOW J. W . , GILLIGAN L. B . , KRYNEN J. & WILLIS I. L. 1 9 8 8 . A r e v i e w o f t h e

mineral deposits of the New England Orogen in New South Wales. In: Kleeman J. D. ed. New England Orogen—tectonics and metallogenesis, pp. 211-227. Department of Geology and Geophysics, University of New England, Armidale. BENSON W. N. 1913. The geology and petrology of the Great Serpentine Belt of New South Wales. Part i. Introduction. Proceedings of the Linnean Society of New South Wales 38, 4 9 0 - 5 1 7 .

BODNAR R. J. 1992. Revised equation and table for determining the freezing point depression of H 2 0-NaCl solutions. Geochimica et Cosmochimica Acta 57, 683-684. BODNAR R. J., REYNOLDS T . J. & KUEHN C. A . 1985. F l u i d

inclusion systematics in epithermal systems. In: Berger B. R. & Bethke P. M. eds. Geology and geochemistry of epithermal systems, pp. 73-97. Reviews in Economic Geology 2. BOHLKE J. K. 1989. Comparison of metasomatic reactions between a common C0 2 -rich fluid and diverse wall rocks: intensive variables, mass transfers, and Au mineralization at Alleghany, California. Economic Geology 84, 291-327.

195-203.

COVENEY R. M. 1981. Gold quartz veins and auriferous granite at the Oriental mine, Alleghany district, California. Economic Geology 76, 2176-2199. CRAW D. 1989. Shallow-level, late-stage gold mineralisation in Sawyers Creek, Shotover valley, northwest Otago, New Zealand. New Zealand Journal of Geology and Geophysics 32, 385-393. CRAW D. & KOONS P. O. 1989. T e c h n i c a l l y

induced

hydrothermal activity and gold mineralization adjacent to major fault zones. In: Keays R. R., Ramsay W. R. H. & Groves D. I. eds. The geology of gold deposits: the perspective in 1988, pp. 471^478. Economic Geology Monograph 6. CRAW D . & NORRIS R. J. 1 9 9 1 . M e t a m o r p h o g e n i c

Au-W

veins and regional tectonics: mineralisation throughout the uplift history of the Haast Schist, New Zealand. New Zealand Journal of Geology and Geophysics 34, 373-383. CROSS K. C. 1983. The Pigna Barney ophiolitic complex and associated basaltic rocks, northeastern New South Wales, Australia. PhD thesis, University of New England, Armidale (unpubl.). DONGG., MORRISONG. & JAIRETH S. 1995. Quartz textures in

epithermal veins, Queensland—classification, origin, and implication. Economic Geology 90, 1841-1856. FLOOD P. G . & AITCHISON J. C . 1 9 8 8 . T e c t o n o s t r a t i g r a p h i c

terranes of the southern part of the New England Orogen. In: Kleeman J. D. ed. New England Orogen—tectonics and metallogenesis, pp. 7-10. Department of Geology and Geophysics, University of New England, Armidale. FLOOD R. H. & SHAW S. E. 1977. Two 'S-type' granite suites with low initial 87 Sr/ 86 Sr ratios from the New England Batholith, Australia. Contributions to Mineralogy and Petrology

61, 163-173.

FROST B. R. 1985. On the stability of sulfides, oxides, and native metals in serpentinite. Journal of Petrology 26, 31-63. FYON J. A . , CROCKET J. H . & S C H W A R C Z

H . P.

1983.

Magnesite abundance as a guide to gold mineralization associated with ultramafic flows, Timmins area. Journal of Geochemical Exploration 18, 245—266. GOLDING S. D . , MCNAUGHTON N . J., BARLEY M . E., GROVES D . I., H o S. E., ROCK N . M . S. & TURNER J. V . 1989.

BROWNLOW J. W. & ASHLEY P. M. 1991. Piedmont magnesite

Archean carbon and oxygen reservoirs: their significance for fluid sources and circulation paths for Archean mesothermal gold deposits of the Norseman-Wiluna Belt, Western Australia. In: Keays R. R., Ramsay W. R. H. & Groves D. I. eds. The geology of gold deposits: the perspective in 1988, pp. 376-388. Economic Geology Monograph 6.

deposit — a hydrothermal vein system in the Great Serpentinite Belt. New South Wales Geological Survey Quarterly Notes 82, 1-20.

GRANT J. A. 1986. The isocon diagram — a simple solution to Gresens' equation for metasomatic alteration. Economic Geology 81, 1976-1982.

BUISSON G. & LEBLANC M. 1987. Gold in mantle peridotites from upper Proterozoic ophiolites in Arabia, Mali and Morocco. Economic Geology 82, 2091-2097.

GROVES D. I. & FOSTER R. P. 1991. Archaean lode gold deposits. In: Foster R. P. ed. Gold metallogeny and exploration, pp. 63—103. Blackie, Glasgow.

BOHLKE J. K . & KISTLER R . W . 1 9 8 6 . R b - S r , K - A r , a n d

stable isotope evidence for the ages and sources of fluid components of gold-bearing quartz veins in the northern Sierra Nevada foothills metamorphic belt, California. Economic Geology 81, 296-322.

SILICA-CARBONATE ROCKS, NSW HEDENQUIST J. W. & HENLEY R. W. 1985. The importance of

C 0 on freezing point measurements of fluid inclusions: evidence from active geothermal systems and implications for epithermal ore deposition. Economic Geology 80, 1379-1406. HENLEY H. F. 1991. Lamprophyres and gold in the New England Fold Belt. New South Wales Geological Survey Quarterly Notes 83, 7-24. 2

HENSEL H.-D., MCCULLOCH M . T. & CHAPPELL B. W . 1985.

The New England Batholith: constraints on its derivation from Nd and Sr isotopic studies of granitoids and country rocks. Geochimica et Cosmochimica Acta 49, 3 6 9 - 3 8 4 . HERBERT ST J. S. G. 1987. The geology of the Crow Mountain-Warrabah Creek district, 35 km southeast of Barraba. BSc (Hons) thesis, University of New England, Armidale (unpubl.). JOHANNES W. 1969. An experimental investigation of the system M g 0 - S i 0 - H 0 - C 0 . American Journal of Science 267, 1083-1104. KENT A. J. R. 1994. Geochronology and geochemistry of Palaeozoic intrusive rocks in the Rockvale region, southern New England Orogen, New South Wales. Australian Journal of Earth Sciences 41, 3 6 5 - 3 7 9 . KERRICH R. & WYMAN D. 1990. Geodynamic setting of mesothermal gold deposits: an association with accretionary tectonic regimes. Geology 18, 882—885. KISHIDA A. & KERRICH R. 1987. Hydrothermal alteration zoning and gold concentration at the Kerr-Addison Archean lode gold deposit, Kirkland Lake, Ontario. Economic Geology 82, 649-690. LANPHEREM. A. & HOCKLEY J. J. 1976. The age of nephrite occurrences in the Great Serpentine Belt of N.S.W. Journal of the Geological Society of Australia 23, 15-17. LAWIE D. C. 1993. A geochemical and multivariate statistical evaluation of gossans and ironstones from the southern New England Orogen and Lachlan Fold Belt in New South Wales. BSc (Hons) thesis, University of New England, Armidale (unpubl.). MACNEVIN A. A. 1974. Great Serpentinite Belt. In: Markham N. L. & Basden H. eds. The mineral deposits of New South Wales, pp. 392-^4-03. Geological Survey of New South Wales, Sydney. 2

2

2

MADU B. E., NESBITT B. E. & MUELENBACHS K. 1990. A

mesothermal gold-stibnite-quartz vein occurrence in the Canadian Cordillera. Economic Geology 85, 1260-1268. MARKHAM N. L. 1975. Gold deposits of the New England Fold Belt. New South Wales Geological Survey Report GS1975/378 (unpubl.). MATSUHISA Y., GOLDSMITH J. R. & CLAYTON R. N. 1979.

Oxygen isotopic fractionation in the system quartz-albiteanorthite-water. Geochimica et Cosmochimica Acta 43,

1131-1140. MAX M. D., TRELOAR P. J., WINCHESTER J. A. & OPPENHEIM

M. J. 1983. Cr mica from the Precambrian Erris Complex, NW Mayo, Ireland. Mineralogical Magazine 47, 359-364. NESBITT B. E. 1991. Phanerozoic gold deposits in technically active continental margins. In: Foster R. P. ed. Gold metallogeny and exploration, pp. 104—132. Blackie, Glasgow.

225

NESBITT B. E., MUELENBACHS K. & MUROWCHICK J. B. 1989.

Genetic implications of stable isotope characteristics of mesothermal Au deposits and related Sb and Hg deposits in the Canadian Cordillera. Economic Geology 84, 1489—

1506. OHMOTO H. 1986.

Stable isotope geochemistry of ore

deposits. Reviews in Mineralogy 16, 4 9 1 - 5 5 9 . OHMOTO H. & RYE R. O. 1979. Isotopes of sulfur and carbon. In: Barnes H. L. ed. Geochemistry of hydrothermal ore

deposits, pp. 5 0 9 - 5 6 7 . Wiley, New York. PEARTON T. N. & V I L J O E N M. J. 1986.

Antimony mineralization in the Murchison greenstone belt—an overview. In: Anhaeusser C. R. & Maske S. eds. Mineral deposits of southern Africa, pp. 2 8 3 - 3 2 0 . Geological Society of South Africa, Johannesburg. PETERS E. K. 1991. Gold-bearing hot spring systems of the northern Coast Ranges, California. Economic Geology 86, 1519—1528. POHL W. 1990. Genesis of magnesite deposits—models and

trends. Geologische Rundschau 79, 291-299.

ROBERTS J. & ENGEL B. A. 1987. Depositional and tectonic

history of the southern New England Orogen. Australian Journal of Earth Sciences 34, 1—20. ROBINSON B. W. & FARRAND M. G. 1982. Sulfur isotopes and the origin of stibnite mineralisation in New England, Australia. Mineralium Deposita 17, 161-174. ROBINSON B. W. & KUSAKABE M. 1975. Quantitative preparation of sulphur dioxide, for S / S analyses, from sulphides by combustion with cuprous oxide. Analytical Chemistry 47, 1179-1181. ROCK N. M. S. 1991. Lamprophyres. Blackie, Glasgow. ROEDDER E. 1984. Fluid inclusions. Reviews in Mineralogy 34

32

12.

SEWARD T. M. 1989. The hydrothermal chemistry of gold and

its implications for ore formation: boiling and conductive cooling as examples. In: Keays R. R., Ramsay W. R. H. & Groves D. I. eds. The geology of gold deposits: the perspective in 1988, pp. 3 9 8 - ^ 0 4 . Economic Geology Monograph 6. SHAW S. E. & FLOOD R. H. 1993. A compilation of Late Permian and Triassic biotite Rb-Sr data from the New England Batholith and areas to the southeast. In: Carr P. F. ed. Centre for Isotope Studies Research Report 1991-92, pp. 151-155. Centre for Isotope Studies, CSIRO Minerals Research Laboratories, North Ryde. TUYSUZN. & ERLER A. 1993. Geochemistry and evolution of listwaenites in the Kagizman region (Kars, NE Turkey). Chemie der Erde 53, 315-329. YANGK. & SECCOMBE P. K. 1993. Platinum-group minerals in the chromitites from the Great Serpentinite Belt, NSW, Australia. Mineralogy and Petrology 47, 2 6 3 - 2 8 6 . ZACHMANN D. W . & JOHANNES W . 1989. Cryptocrystalline magnesite. In: Moller P. ed. Magnesite. Geology, mineralogy, formation of Mg-carbonates, pp. 15—28. Gebruder Borntraeger, Berlin. (Received 10 April 1996; accepted 16 July 1996)

Tectonics and Metallogenesis

of the New England Orogen. Geological Society of Australia Special Publication 19, 226-241.

Tectonic significance of veins and associated structures at Woodsreef, New South Wales R. OFFLER, 1 D. S. O ' H A N L E Y 2 * A N D P. G. L E N N O X 3 1

Department of Geology, University of Newcastle NSW 2308, Australia. 2 Royal Ontario Museum, Toronto, Ontario, Canada M5S. 3 Department of Applied Geology, University of New South Wales, Sydney NSW 2052, Australia. The rocks in the Woodsreef chrysotile asbestos deposit record a progressive deformation history which commenced with the emplacement of dolerites into meridional and east-west-trending shear zones located within harzburgites. The contacts between these two rock types acted as conduits for fluid and as planes of slip during subsequent deformation events, resulting in the development of veins, shear bands, slickenfibres and slickenlines. Initially, several types of veins developed in partially serpentinised harzburgite adjacent to major meridional and north-northeast-trending shear zones that are exposed in the deposit. They are more numerous near these zones forming a network of regular, tabular veins and en echelon arrays which become less well-developed and eventually disappear to the east away from these structures. The veins show asymmetric tips, bridges, and some, features characteristic of crack-seal deformation indicating an extensional origin for them; their close spatial relationship to shear zones suggests they are fault related. The majority of veins dip east to southeast at moderate to steep angles with slight variations occurring between pits. A similar variation in orientation exists for each vein type suggesting they have experienced a similar tectonic history. The variability in orientation shown by the veins is attributed to transient changes in near-field stress orientations during shearing. Slickenfibres and shear bands formed after the veins and show left lateral, right lateral, reverse and normal senses of movement. Stress tensor analysis of these structures suggests that the maximum principal stress (<7]), varied from east-west to meridional to north-northwest during shortening events and vertical during extension. The apparent change in near-field principal stress directions indicated by the extreme variability in shear sense and slip direction shown by slickenfibre lineations and shear bands, is attributed to rotation of the planes containing them. Slip on faults bordering the Woodsreef Serpentinite produced the rotation required to bring the planes into orientations which allowed the different senses of movement to take place.

Key words: asbestos, kinematic history, serpentinite, vein deposits, Woodsreef.

INTRODUCTION The Woodsreef chrysotile asbestos deposit is located near the village of Woodsreef 15 km northeast of Barraba, New South Wales (Figure 1). Several geological studies have been carried out, most of which have dealt with the origin of the asbestos-bearing veins (Proud & Osborne 1952; Glen & Butt 1981; Katz 1986). Katz (1986) believed that the veins resulted from the deformation of the Woodsreef Serpentinite during dextral-transtensional movement on the Peel Fault System. This is contrary to the view of Glen and Butt (1981) who considered that there is a genetic relationship between the veins and the Slice, a north-northeasterly trending offshoot of the main serpentinite body (Figure 1). Further, it does not accord with the interpretation of Proud and Osborne (1952) who related the intrusion of dolerites to the formation of the asbestos fibre. From this, it is clear that the origin of the veins in the Woodsreef asbestos deposit has yet to be clarified. In an attempt to construct a tightly constrained model, we have carried out an extensive analysis of the veins in the opencuts developed during the mining of the

asbestos. Detailed geometric analyses of slickenfibres and slickenlines have also been carried out as these structural elements have been used by Katz (1986) in a reconnaissance study to determine the kinematic history of the deposit. The interpretation of these analyses and a model for the formation of the veins are given in this paper. It will be shown that the rocks in the Woodsreef deposit record a progressive deformation history which commenced with the formation of high-temperature shear zones in harzburgite and culminated with the production of asbestos and other types of veins adjacent to shear zones at dolerite-harzburgite contacts. Subsequent deformation led to the formation of slickenfibres, slickenlines and shear bands of varying orientation. The study reveals a far more complicated tectonic history than previously reported for the Woodsreef asbestos deposit. GEOLOGIC SETTING The Woodsreef asbestos deposit is located in the Woodsreef Serpentinite, a lensoid north-northwest*Present address: Trinity School, River Ridge, 2300 East 88th Street, Bloomington, MN 55425-2187, USA.

227

VEINS AND ASSOCIATED STRUCTURES, WOODSREEF Sediments

CAINOZOIC

LOWER PERMIAN G3Arenite E. CARBONIFEROUS [ ^ C a r o d a Fm.

M. DEVON- E. CARB. K iNangahrah Fm. Woodsreef Serpentinite harzbur^gife r P e n t ' n ' S G C ' t = j Massive serpentinite I [Schistose serpentinite HDolerite, gabbro and rodingite Geological Boundaries Position accurate Fault, position accurate /

Fault, position approximate

Road

O

I

150 300

I

I

450

I

Figure 1 Geology of the Woodsreef Serpentinite and approximate outline of pits (modified from Glen & Butt 1981). Inset in upper left-hand corner shows the location of Woodsreef; inset in the lower left-hand corner shows the shape of the Woodsreef Serpentinite (shaded area) and its setting with respect to the Peel Fault System. See Figure 2 for more detail.

striking body with a length of 8 km and a width of 2.4 km (Proud & Osborne 1952; Figure 1). It is bounded on the west by the Peel Fault and on all other sides by the Nangahrah Formation (Blake & Murchey 1988). Dips at all contacts are vertical or steep to the east or west. The Woodsreef Serpentinite is one of many serpentinites that delineate the Peel-Manning Fault System, a suture extending for approximately 270 km near the western margin of the southern New England Fold Belt (Figure 1). The serpentinites form part of a dismembered, earliest Cambrian ophiolite referred to as the Weraerai terrane (Aitchison et al. 1992). The PeelManning Fault System separates Devonian to Carboniferous forearc basin sequences of the Tamworth Belt to the west from Silurian to Early Carboniferous accretionary-subduction sequences of the Tablelands Complex to the east. The in situ exposure within the openpits of the Woodsreef mine is limited because of slumping and the presence of backfill. However, sufficient outcrop is present in all pits to allow a detailed investigation to be made (Figures 2, 3a). Partially serpentinised

harzburgite, massive and cleaved serpentinite as well as rare dunite, occur in different parts of the deposit. Several shear zones containing dolerites or lenticular rodingitised, boudinaged dolerites bordered by pervasively cleaved serpentinite, are exposed in the pits; they are commonly curvilinear in plan and section. The most strongly developed shear zone occurs in the Hardie and Centre pits and is oriented 020/70SE (hereafter referred to as the N20E shear zone). East of the N20E shear zone is a network of shear zones developed in massive serpentinite and several types of veins, most of which are localised in partly serpentinised harzburgite (Figure 2). Outcrops adjacent to the Hardie pit indicates that they die out to the east. Similar features are apparent in the North pit and appear to be related to a meridional-trending shear zone on the western wall where numerous rodingite pods crop out. However, massive serpentinite and small shear zones are more abundant in this pit. Alteration of the peridotite Microscopic studies reveal that the protolith of the

228

R. O F F L E R £ T ^ Z .

serpentinite was a harzburgite containing orthopyroxene + enstatite + olivine + chromite (O'Hanley & Offler 1992). No variation in the olivine or orthopyroxene composition occurs across the body (Glen 1971). Harzburgite textures are primarily Type 1 protogranular textures (Mercier & Nicholas 1975) with wormy chromite grains that are intergrown with enstatite; cumulate textures have also been recognised. Locally, near a shear zone exposed in the southern margin of the Wunderlich pit, the harzburgite consists of kinked olivine and orthopyroxene grains with neoblasts of olivine, indicating that a high-temperature anhydrous deformation, possibly localised on shear zones, affected the harzburgite. More widespread than the products of the high-temperature deformation, but still restricted to rocks adjacent to shear zones in the Hardie, Wunderlich and North pits, is a lower temperature, greenschist facies assemblage consisting of tremolite + chlorite + talc + antigorite (tremolite assemblage). Brittle deformation occurred subsequently and resulted in a myriad of closely spaced fractures of varying orientation in some samples and the fracturing of tremolite in one sample. All of these assemblages are overprinted by lizardite which formed during partial serpentinisation of the harzburgite. The textural and mineralogical features described above are significant because they are observed in altered harzburgite adjacent to shear zones containing dolerite and rodingite. Importantly, olivine shows evidence of dynamic recrystallisation and later brittle failure suggesting that these zones were formed very early in the deformation history of the deposit and were reactivated subsequently when temperatures were lower. In addition to the alteration that produced the tremolite assemblage, two other assemblages have been identified, based on cross-cutting relationships in outcrop and on mineral assemblages, textures, and 5 1 8 0 values (O'Hanley & Offler 1992). The tremolite assemblage (5 1 8 0 = 7.6, 7.7%0) is overprinted by a spatially widespread, but incomplete, development of

lizardite + brucite (3.5<5 1 8 0<5) during the first stage of serpentinisation. During the second stage of serpentinisation, texturally different lizardite + brucite formed (4<5 1 8 0<6). Subsequently, this assemblage was replaced by lizardite and chrysotile (3.7<5 1 8 0<8.6) associated with the formation of chrysotile asbestos veins (6<5 1 8 0<7). Temperatures during the second serpentinisation event, calculated from 8 l s O values from serpentine-magnetite mineral pairs, are 350-220°C, indicating that this event occurred while temperature was dropping, consistent with increasing 5 1 8 0 values in successively younger assemblages. The data obtained from the serpentinemagnetite pairs also suggests that moderately hot, hydrothermal fluids fluxed into the host rocks from the shear zones at the time serpentine minerals were formed during the second event. Subsequent to the formation of the veins, major influxes of fluid along the shear zones caused extensive alteration of the rocks adjacent to them. This alteration partly to completely obliterated pre-existing textures and minerals formed earlier in the second serpentinisation event. Rodingites and dolerite Rodingites, and less commonly dolerites, occur in the harzburgite, the latter exceeding 3 m in width in some cases. Most trend meridionally (Figure 5), which accords with the studies carried out by Glen and Butt (1981) in the Woodsreef area. The rodingites form pinch-and-swell structures which are mantled by sheaves of black, chlorite-rich, anastomosing, cleaved, highly polished and striated aggregates which in turn are surrounded by highly deformed serpentinite. In some pits, dunite dykes occur which have a similar trend to the dolerite dykes. No dolerite dykes occur in the Nangahrah Formation suggesting that they were emplaced before the harzburgite and the associated rocks reached their present crustal position.

Figure 2 Map showing the distribution of rock types, shear zones, rodingites and dolerite dykes in the pits. Clear areas labelled H, C, W and N are filled with water.

VEINS AND ASSOCIATED STRUCTURES, WOODSREEF VEINS Vein types As mentioned previously a variety of serpentine veins occur in partly serpentinised harzburgite in the pits adjacent to major shear zones. In the host rocks closest to the shear zones, a kernel pattern is developed. This is defined by intersecting sets of asbestos veins mantled by completely serpentinised harzburgite, surrounding a core of partly serpentinised harzburgite ( O ' H a n l e y & Offler 1992; Figures 2a, 3; Figures 3a, b). This pattern occurs in rocks referred to as massive serpentinite by Glen and Butt (1981; Figure 1). Glen and Butt (1981) defined six types of veins in the W o o d s r e e f Serpentinite based on their m e s o s c o p i c

Figure 3 (a) View to the south of the main shear zone in the Hardie pit, illustrating the variation in serpentinisation and recrystallisation with distance from the shear zone (SZ). Area labelled 4 B' is backfill. Kernel pattern (KP) is represented in the photograph by the white kernel rims. Kernel pattern dies out near 'E' where only black veins are present. The asbestos veins adjacent to the rims of the kernels dip to the east. R is a rodingite pod in the shear zone. Dashed line is the border of the shear zone. Bench height is 10 m. (b) Kernel defined by two intersecting veins. P, partly serpentinised harzburgite; R, ribbon vein; F, fibre vein; B, black vein. Lens cap in northeast corner of kernel. Diameter of lens cap is 5 cm.

229

Table 1 Vein types in the Woodsreef Serpentinite.

Glen & Butt (1981)

This study

Type 1 Type 2

Mantled veins Stockwork veins

Black

Type 6 Type 3

Schistose veins Ribbon veins

Fibre

Type 5 Type 4

Thread veins Kernel-rim veins

Ribbon Combination of Ribbon & Fibre Green

No equivalent

230

R. O F F L E R £ T ^ I .

appearance. We recognised these vein types in outcrop but, based on petrographic studies (O'Hanley & Offler 1992) and cross-cutting relationships among the veins, we have reclassified them into four vein types (Table 1). The veins commonly form en echelon arrays which occur at low angles to the zone containing them. Further, they show asymmetric tips (Figure 4a), have vein overlap ratios of 0.4 and well-developed bridges, features which are characteristic of extension arrays (Nicholson & Pollard 1985). Black veins are readily recognised because of their colour which contrasts with the brown colour of partially serpentinised harzburgite produced during weathering. They represent the first expression of the second episode of serpentinisation (O'Hanley & Offler 1992) and are found further from the shear zones than other veins (Figure 3a); microscopically they consist of lizardite ± magnetite. Ribbon veins occur adjacent to shear zones and consist of parallel veins of asbestos less than 0.3 cm wide and 2—3 m long (Figure 3b). Replacement of ribbon veins during recrystallisation near shear zones produces the thread veins of Glen and

Butt (1981) which consist of isolated lenses of asbestos fibre in bastite. Fibre veins are found further from the shear zones than the ribbon veins and consist of a single asbestos-filled fracture generally wider than 0.5 cm and longer than 1 m; microscopically, they are associated with chrysotile-bearing textures. Green veins are morphologically similar to fibre veins, but under the microscope they consist of bands of lizardite and polygonal serpentine. All types of asbestos veins and green veins may occur in a kernel rim. Cross-cutting relationships among the veins indicate that green veins post-date ribbon veins but pre-date fibre veins. Crackseal deformation features (Ramsay 1980) have been recognised in many vein types supporting their extensional origin (Figure 4b). Many of the veins have been recracked during subsequent deformation events resulting in the formation of slickenlines and slickenfibres defined by coarse, green fibres of chrysotile (Figure 4d). Some are partially to completely obliterated by subsequent shearing or by fluids responsible for the development of massive serpentinite.

Figure 4 (a) Chrysotile asbestos-bearing extension veins showing asymmetric tips cutting a matrix of completely serpentinised harzburgite. Specimen no. W44, from east side of Centre pit. Length of bar 0.4 mm. (b) Fibre vein containing extended magnetite (M) and fibrous chrysotile (C). S is completely serpentinised harzburgite. Specimen no. W7, from southwest side of Hardie pit. Length of bar 0.4 mm. (c) Slickenfibre defined by fibrous chrysotile on vertical shear joint. Note strike-slip and oblique-slip direction of movement shown by fibres. Arrows indicate orientation of fibres. Road cut 50 m east of Hardie pit. Length of bar 3 cm. (d) Slickenfibre (arrow) developed in recracked black vein. P is partly serpentinised harzburgite. East side of Centre pit. Scale in centimetres (top) and inches (bottom).

231 VEINS A N D ASSOCIATED STRUCTURES, WOODSREEF such a sequential order of formation does not appear to Serpentinised zone width versus vein width exist for veins in every outcrop. The only consistent The asbestos veins are bordered by zones of completely observation is that veins with poles plotting near 'A' are serpentinised harzburgite called the kernel rim. The cut by veins plotting near 'B' (Figure 8e) suggesting width of this rim is called the serpentinised zone width that two generations of veins formed at different times. (SW) which is measured by obtaining the width of the This is in accord with the interpretations of Glen and kernel rim and subtracting from it the width of all veins Butt (1981); however, they recognised a more complex in the rim. The total width of all measured veins is chronology involving the sequential formation of veins orientated 360/45E, 090/50S and 145/50SW. called the vein width (VW). Dresser (1917) first calculated the ratio SW/VW using measurements from mines in southeastern Cleavage Quebec. He found that each mine had a specific value for this ratio which varied from 3 to 8. Similar results Cleavage is restricted to shear zones that occur were obtained by Keith and Bain (1932) from the throughout the deposit as curviplanar zones varying in Belvedere deposit, southeast Quebec. However, in the Thetford Mines deposit to the east, Cooke (1936) found no specific value. In their study of the asbestos deposits of southeastern Quebec, Cogulu and Laurent (1984) concluded that the ratio SW/VW is a function of the degree of serpentinisation adjacent to the vein. Thus increasing serpentinisation would give rise to a decreasing ratio. In the Woodsreef mine, the ratio SW/VW is generally >5:1 (Figure 6). In detail however, the ratio varies between pits (Wunderlich and North pits) and within pits (Hardie). The highest values for SW are from ribbon veins or from a combination of ribbon and fibre veins. This observation and the fact that the ribbon veins are in close proximity to the shear zones, indicates that the shear zones may have been the source of the fluids responsible for the alteration, a conclusion previously arrived at by O'Hanley and Offler (1992). Geometry and relationships among the vein types The orientation of veins in each pit is shown in Figure 7a-f. The majority of the veins dip east to southeast at moderate to steep angles although slight variations exist among the vein maxima for each pit (Table 2). The orientation of the veins obtained in the North pit are similar to those recorded by Glen and Butt (1981). When the different types of veins are plotted separately, they show a similar variation in orientation (Figure 8ae), suggesting that most of the veins experienced a similar tectonic history. In general, most veins are orientated 008/60E and 044/40SE. On the basis of observations in outcrop and thinsection, the veins formed in the sequence (from oldest to youngest) black, ribbon, green and fibre. However,

270°

• 90°

Table 2 Orientation of veins in the Woodsreef Serpentinite. Pit

Orientation of maxima

Hardie Wunderlich North Pit (S zone) North Pit (N zone) Centre Pit All veins

051/48SE 018/90 04/44E, 313/60SW 326/90, 338/78NE 338/90 008/60E, 044/40SE

Figure 5 (a) Poles to rodingites and dolerites (triangles), and dunites (squares), (b) Rose diagram showing the orientation of dunite and partially and completely rodingitised dolerite dykes.

232 R. O F F L E R E T A L . width, strike and dip (Figure 2). It appears as an anastomosing curviplanar structure, the surface of which is commonly polished and striated or shows slickenfibre development. Microscopically it is defined by slip chrysotile and serpentinite clasts, features characteristic of incohesive, serpentine cataclasites (Norrell et al. 1989). In some specimens, it wraps around static alteration products associated with the second stage of serpentinisation indicating that it has formed after this event. When visible, the boundary between the foliated serpentinite in the shear zone and massive serpentinite is sharp. In some instances, kernels are dragged by normal slip movement into conformity with the shear zones (e.g. Hardie pit), and within these

25 23

zones are totally altered and enclosed by an anastomosing cleavage. A similar anastomosing affect is seen in shear zones containing rodingite showing pinch-andswell structure. The cleavage occurs in a chlorite-rich mantle around the rodingite and microscopically consists of foliated blades of chlorite and talc wrapping around aggregates of massive prehnite cut by veins of diopside, hydrogrossular and chlorite. This suggests that the cleavage developed subsequent to the alteration responsible for rodingitisation. Close examination of the shear zones reveals that one or more sets of shear bands are present, which vary in orientation and intensity of development. In some outcrops, shear bands with different orientation but the same sense of movement can be recognised, in others, the sense of movement is different. Further, in some outcrops three shear band C-surfaces (S , S , S ), which appear to have developed sequentially as a result of continued movement on these zones, can be identified. A cleavage (SO which is deformed by S is also present but it is not known whether it is a C-surface or a plane of flattening. S\ is finely penetrative in hand sample and has a highly variable orientation with maxima forming at 335/79W, 190/83W, 250/64N, 052/78N and 208/71NW (Figure 9b). Overprinting it is S which is also finely penetrative in hand sample and commonly forms a small angle to Sj; most are orientated 022/85W and 053/85SE (Figure 9c). S is spaced 1-2 cm apart and is more variable in orientation than S , most planes trending north-northeast and north-northwest which is the orientation of most shear zones in the pits (Figures 2, 9d). S is widely spaced and uncommon. Glen and Butt (1981) recognised two cleavages in the serpentinites, Sj and S , however their orientation is different from S] and S in this study. In general, the cleavage they refer to as Sj shows less variability in strike and a shallower dip than the S\ that we have recognised. Further, their S is similar in orientation to the S and S that we have noted in this study. In many shear zones, only one set of shear bands is present; for convenience the planes defining these shear bands are referred to as C and 'S' because their relative ages are unknown. The use of the term 'S' is not meant to imply that it is a plane of flattening as is the common situation in mylonites, merely that it is an earlier cleavage dragged into conformity with C. C has an orientation similar to S with most planes trending meridionally and dipping steeply west; less common are planes striking northwest and northeast (Figure 9a). Analysis of the relationship between C and 'S' and between shear band C-surfaces in outcrops showing several sets of shear bands, reveals that right-lateral, left-lateral, reverse and normal slip movements have taken place. Adjacent to the Peel Fault, to the west of the North pit, the shear bands trend approximately meridionally and show a left-lateral sense of movement similar to that recorded at other locations on this structure (Offler & Williams 1987; Offler et al 1989). These shear bands displace northeast-trending shear bands and become less well-defined towards the deposit. 2

•R

3

2

21 /•R

19

2

3

o 15 co o 13 0 £O

11 —

RA

2

4

2

2

1° Hf,G

5 3

2

"V.X

>2 9 — C£O 7

4

PIT /F,G,F,F,

• Hardie

• Centre

• Wunderlich • South Zone, North Pit a North Zone, North Pit

1

1 2 3 4 5 6 7 8 vw - width of vein (cm)

Figure 6 Plot of serpentinised zone width (SW) vs vein width (VW). F, fibre; R, ribbon; G, green. See text for discussion.

2

3

3

Lineations

233 curved fibres similar to those described by Twiss and Gefell (1990) occur and features characteristic of crack-seal deformation have been observed in fibres exposed on some planes. Difficulty was found in determining the time relationships between the various generations of fibres. However, in the Centre pit, planes orientated 046/60NW and showing a down-dip sense of movement are displaced by slickenfibre-bearing planes orientated 020/45E and 346/70E, the latter having a reverse-slip sense of movement. In general, slickenfibres are most strongly developed on steep east-northeast- or west-southwest-dipping planes, and less commonly on south-southeast- or north-

VEINS AND ASSOCIATED STRUCTURES, WOODSREEF

In all pits, planes showing slickenfibre and slickenlines are present. The former are commonly found in massive serpentinites and less commonly in the centre of black and fibre veins (Figure 4d) but are less well-preserved in the shear zones where reworking has occurred. Further, they are defined by prisms of chrysotile, the orientation of which may vary in a single outcrop and from pit to pit. Two and rarely three generations of slickenfibres of different orientation and showing similar or opposing senses of movement, may be developed on the one plane or adjacent planes (Figure 4c). More rarely,

b) WUNDERLICH

a) HARDIE PIT

2,4,6,8% c) NORTH PIT

n = 125 d) NORTH PIT

SOUTH ZONE

NORTH ZONE

ft

2,4,6%

2,4%

n = 85

e) CENTRE PIT

Figure 7 Poles to veins from individual pits, (a) Hardie. (b) Wunderlich. (c, d) North, (e) Centre, (f) Maxima from 7ae. 4% contour (b-e), 6% contour (a).

2,4%

n = 85 I •

I North

OHlHarclie

Centre H

Wunderlich

n = 33

234

R.O¥¥LERETAL.

northwest-dipping planes (Figure 10a, b). Apart from the northwest-dipping planes, all other planes are of similar orientation to the veins suggesting that the latter represent shearing of earlier extension veins. Most axes of slip fibres plunge steeply to the northwest and westsouthwest, and more gently to the northwest (Figure 10c, d); the sense of movement shown by them is left lateral, right lateral, reverse and normal slip, the last of these being the most common. Slickenfibres showing the same sense of movement occur on planes with a wide variety of orientations suggesting that the principal stresses have varied during the deformation of the deposit. These features indicate that a far more complex kinematic history is recorded in the deposit than suggested by Katz (1986). a) Fibre veins

b) Black veins

2,4,6%

2,4,6 %

c) Green veins

d) Ribbon veins

2,4,6%

In contrast to slickenfibres which are present in all parts of the Woodsreef deposit, slickenlines are more commonly found on highly polished surfaces on C planes within and adjacent to shear zones. However, they have also been observed in association with one or more slickenfibre lineations elsewhere in the pits and are clearly later as they smear out and overprint these structures. In some outcrops, they appear on a veneer of serpentine covering slickenfibre lineations formed previously. Normally, only one generation is present but rare occurrences of two have been noted. They are strongly developed on planes striking east-northeast and dipping steeply northwest (Figure lOe, f) and indicate either dip-slip or strike-slip movement (Figure lOg).

59 e) All veins

4,8,12%

50

422

Figure 8 Poles to serpentine veins plotted by type of vein, (a) Fibre veins, (b) Black veins, (c) Green veins, (d) Ribbon veins, (e) All veins.

Origin of veins

235 O'Hanley 1987). Many workers have identified asbestos veins as mode-1 fractures (Riordon 1955; Laubscher 1964; Laurent 1975; Glen & Butt 1981; O'Hanley 1987) but the genetic relationship between faults and veins was not recognised. Glen and Butt (1981) considered that fibre growth in veins took place in massive serpentinite during the formation of their S and S in the surrounding schistose serpentinite. Although some fibre growth took place at this time, there is clear evidence that the event producing the massive serpentinite is after vein formation because: (i) ghost outlines of kernel structure can be seen within it; and (ii) cleavage defining S] and S wrap around the kernel structure. Further, the model proposed by Glen and Butt (1981) does not explain the variable orientation of veins that can be observed adjacent to shear zones. We believe the development of these veins is in some ways similar to that proposed for Au-bearing vein systems associated with faults showing fault-valve behaviour (Cox 1995) because: (i) the shear zones are steeply inclined and have been the focus of high fluid flow as indicated by the extensive alteration within and adjacent to them, and associated with the ribbon veins; (ii) the 5 0 values reported by O'Hanley and Offler (1992) require high fluid fluxes; (iii) the veins are of extensional origin and in some cases show evidence for repeated crack-seal deformation; and (iv) the veins are clearly associated with the shear zones and die out away from them. However, the geometry of the vein

VEINS AND ASSOCIATED STRUCTURES, WOODSREEF

Any explanation for the origin of the veins must take into consideration: (i) their mesoscopic and microscopic features; (ii) the variability in their orientation within and between pits; (iii) the decrease in abundance and width they show away from the shear zones; and (iv) the wider zones of alteration associated with ribbon veins adjacent to shear zones. Evidence was presented above which suggested that the veins were of extensional origin. Features observed microscopically support this interpretation as magnetite, chromite and bastite fragments and older veins can be matched across many veins and show little offset parallel to the vein (Figure 4b). Further, the veins commonly exhibit features characteristic of crack-seal deformation as well as bridges and asymmetric tips. These observations indicate that the veins are mode-1 fractures (Pollard & Aydin 1988) and are similar to vein systems described by Nicholson and Pollard (1985) and Nicholson and Ejiofor (1987). Veins are more numerous near the shear zones, and disappear to the east away from them. Moreover, they are restricted in their orientation as indicated by the maxima in the different pits. The spatial association of the veins with the shear zones suggests that they are faulted related. This association has been observed in other asbestos mines (southeastern Quebec, Cooke 1936; Cassiar mine, north-central British Columbia,

x

2

2

18

a)

2,4,6

2,4,%

c)

F i g u r e 9 Poles to planes observed in shear zones from all pits, (a) C; n = 192 (b) S,; n = 56 (c) S ; n = 70 (d) S ; n = 65. All planes were measured in high-strain zones in which phacoids of massive serpentinite were absent or minimal. 2

3

2,4,6

2,4,6,%

236

R.

OYYLERETAL.

n = 173

n = 173

1%, 2 %

C.

n = 173

1%, 2%, 4%

n= 173

e.

2%, 4%

n = 62

n = 62

Figure 10 (a, b) Orientation of planes containing slickenfibre lineations: (a) rose diagram; (b) contoured poles to planes. (c) Slickenfibre lineations. (d) Slickenfibre lineations and planes containing them, (e, f) Orientation of planes containing slickenlines: (e) rose diagram; (f) contoured poles to planes. (g) Slickenline lineations and planes containing them.

VEINS AND ASSOCIATED STRUCTURES, WOODSREEF

237

Table 3 Palaeostress tensors.

N

Sense

6 5 13 7 11

Dextral

15

Thrust Normal

Slickenfibre Population c^

1 2 1 2 1

ESE-WNW NW-SW E-W, ENE N-S vertical

NNE-SSW NE-SW N—S variable steep W E-W

2

vertical

N-S

N

Sense

15 13 27 10 34

Dextral

23 19 13

Thrust Normal

Sinistral

Planes Population G{

1 2 1 2 1

shallow E shallow N E-W, ENE NNW-SSE vertical

N-S E-W steep E steep E variable WNW -SSW

1 2 3

E-W N-S NW-SE variable

SSW shallow E-W NE-SW variable

N, number of shear joints that fit tensor configuration on the same line; a,, maximum principal stress; a 3 , minimum principal stress.

arrays are far more complex than those associated with Au-bearing systems. Further, the host rocks are serpentinites which are rheologically very different to the rocks that host Au-bearing veins, thus aseismic creep may have occurred on the shear zones rather than episodic slip which is required in the fault-valve model. The repeated phases of crack-seal deformation evident in some veins requires that fluid pressure (Pf) cyclically exceed the sum of the near-field <?3 and the tensile strength of the rock and that fluid be fluxed into the host periodically. Such high Pf and migration of fluid can only occur if the shear zones have been a focus of high fluid flow and if Pf in the zones exceeds that in the host. This is believed to have been the situation at Woodsreef. A decrease in P f away from the shear zones could explain the decrease in size and abundance of the veins more distant from them. The pressures (P) operating during vein formation are difficult to estimate because of the lack of critical assemblages formed during alteration. However, studies done on asbestos deposits in Canada allow some constraints to be placed on P because they show a similar alteration regime and structure to those in the Woodsreef deposit. For example, in the Bowman asbestos mine, Ontario, serpentine recrystallisation took place at 290±40°C and < 100 MPa (Schandl et al 1989) and 300±50°C and <50 MPa at the Cassiar asbestos mine, British Columbia (O'Hanley et al 1992). The temperature range at Woodsreef (350-220°C; O'Hanley & Offler 1992) is similar. Assuming that PH2Q was approximately the same as in other asbestos mines and knowing that P H20 must exceed lithostatic pressure Pj + tensile strength of the rock for fracturing to occur (Etheridge 1983), then P is < 100 MPa indicating depths of ~2-3 km. At these shallow depths, we believe that the partially serpentinised harzburgites were deformed during shearing on pre-existing faults to form the veins.

The question now remains as to why the veins vary in orientation. The possible explanations for this variation are that: (i) the veins formed during a non-coaxial deformation event which resulted in the incremental finite extension direction rotating during shearing (Choukroune et al 1987; Ratschbacher et al 1993); (ii) variations in far-field principal stress directions occurred throughout the deformation history of the deposit; and (iii) variations in near-field principal stress orientations and shear stress occurred during faulting (Cox 1995). The first explanation requires that each vein set forms sequentially with continued shearing and that evidence for this can be observed in the field. In the Woodsreef deposit, no such evidence is apparent. Careful examination of the particular vein sets (e. g. asbestos veins) and their relationships did not reveal a consistent chronology which argues against non-coaxial deformation being responsible for the vein arrays in the deposit. Also we would argue against far-field stress variations being responsible for the variable geometry of the veins because the vein sets developed in one pit are either poorly or not developed in another (compare Figure 7a with 7c). These observations also would not support the proposal by Katz (1986) that the veins are the result of movement on the Peel Fault. The interpretation we favour requires transient rotations of near-field stress orientations during faulting, as has been proposed by Cox (1995) to explain the development of steeply dipping and subhorizontal extension veins adjacent to the Wattle Gully Fault, Victoria. However, the situation at Woodsreef is more complex than the relationships reported by Cox (1995) because more vein systems are present. They require more than a flip through 90° of G\ as he has suggested to explain the formation of the two sets of veins. It is not known what caused these transient rotations of the principal stress directions.

238 R. 0 F F L E R £ T , 4 Z . PRINCIPAL STRESS ANALYSIS

Slickenfibres and shear bands

From the previous discussion, it is clear that the Woodsreef deposit has experienced a complex deformation history which has involved fundamental changes in the orientation of the principal stress axes with time during the deformation of the Woodsreef asbestos deposit. Evidence supporting this is the: (i) variable orientation of the veins; (ii) different directions and senses of movement shown by the slickenfibre lineations and shear bands; (iii) presence of two or more lineations on a single shear plane; (iv) curvature exhibited by many fibres indicating that the shear displacement vector has changed in orientation; (v) reorientation of kernels adjacent to some faults; (vi) displacement of one shear plane by another showing a different sense of movement; (vii) variation in orientation shown by planes containing slickenfibre lineations indicating the same sense of movement; and (viii) recracking and displacement of the asbestos veins during the formation of the slickenfibre lineations. To determine the orientation of these axes, the geometrical relationship between the veins and shear zones in the pits have been examined and programs involving the analysis of fault striae have been applied (Marrett & Allmendinger 1990; Sperner et al 1993). In order to simplify the analysis, the fault-striae data have been divided into four sets containing planes showing the same sense of movement. In most sets, at least two populations of faults appear to exist. From the original database, several populations have been culled and stress tensors calculated using the 'grid search method' of Hardcastle and Hills (1991). These are summarised in Table 3. Veins The variation in the orientation of the veins at Woodsreef is considerable. However, clearly defined maxima are present which are distributed along a great circle the axis of which plunges 50° in a direction 150° (Figure I f ) . This indicates that a has varied substantially during deformation as the pole to the veins is the minimum principal stress axis. For the area as a whole, the main maxima in the synoptic diagram (Figure 8e) indicate that a varied from 30 -> 278 to 50 314. Since the veins appear to be fault related, the orientation of the stress field during their formation can be determined. Thus, in the Hardie pit where most veins are orientated 054/60SE and the shear zone 020/76E, a slip direction of 49 -> 049 and principal stress directions of a, = 18 -> 065, a = 5 -> 180 and a = 30 -> 323 are indicated. The orientation of the principal stress directions in other pits is difficult to define because it is not clear which shear zones are responsible for the vein arrays (e.g. Wunderlich, Centre) or because the orientation of the shear zone has not been well defined (e.g. North). What is apparent however, is that the principal stress directions are different in the North pit because it shows vein maxima which are poorly developed or not present in the other pits (e.g. 004/56E, 313/70SW, 090/82W).

The shortening and extension axes calculated from the programs of Marrett and Allmendinger (1990) and Sperner et al. (1993), for slickenfibres and shear bands showing sinistral, dextral, normal and thrust movement, are shown in Figures 11 and 12. In these calculations, 0 = 45° where 0 is the angle between shortening axis and the shear plane; maximum shear stress is thus imposed on the shear plane in this case (Marrett & Extension Axes

Shortening Axes

3

3

2

3

n=4

Figure 11 Shortening and extension axes from slickenfibre data, (a, b) Reverse slip, (c, d) Normal slip, (e, f) Right-lateral slip, (g, h) Left-lateral slip.

VEINS AND ASSOCIATED STRUCTURES, WOODSREEF

Allmendinger 1990). A more appropriate angle may be 33°, the coefficient of internal friction for dry, nonschistose serpentinite (Raleigh & Paterson 1965). However, the orientation of the shortening and extension axes did not show any significant change when this value was used. Examination of the shortening and extension axes obtained from the slickenfibre data reveals that a single stress field cannot be responsible for the movement patterns. Application of the 'grid search method' of Hardcastle and Hills (1991) confirms this and shows that <j varied from east-west to meridional to north-northwest during the contraction events and vertical during extension at the time the slickenfibre lineations were formed; a was east-west in the

239

latter event (Table 3). Similar orientations are obtained from the C-'S' data. However, this type of analysis does not indicate whether we are dealing here with variations in far-field or near-field stress directions. This is a far more complex pattern than envisaged by Katz (1986) who proposed that a single transtensional event was responsible for the faults, fractures, schistosities and asbestos veins.

STRESS CONDITIONS—SLICKENFIBRES AND SHEAR BANDS? Evidence has been presented to show that the principal stress directions varied during the formation of these structures. The question arises as to whether the slickenfibres and shear bands, and the movement patterns shown by them, are the result of far-field stresses whose orientation has changed in time or a single far-field event associated with local, transient changes in nearfield stress conditions. We believe that major and probably rapid changes in far-field stress conditions are an unlikely explanation for the complex kinematic history recorded by the shear bands and slickenfibre lineations in the Woodsreef deposit. Rather, it is the result of a single contractional, far-field event activating on a local scale, faults and fractures of the appropriate orientation to produce the various types of movement indicated by the slickenfibres and shear bands. Such complex movement patterns have been identified by Zoback and Beroza (1993) in their analysis of aftershocks produced after the Loma Prieta, California earthquake on the San Andreas fault system. They were able to show that the sense of movement depended on the orientation of the aftershock plane, those striking at an acute angle to the main shock plane showing either or left-lateral movement, and normal or reverse 3,6,12% rightslip when inclined either slightly steeper or more gently than the main shock plane. The maximum horizontal principal stress responsible for the main and aftershocks was ~80°—85° to the strike of the San Andreas fault zone indicating that motion on this structure is in response to very small shear stresses. Such low shear stresses can activate high-angle faults if P is high (Zoback & Beroza 1993; Cox 1995). Low shear stresses may have operated at the time the slickenfibres and shear bands were developed as high fluid fluxes occurred at this time and earlier in the deformation history of the Woodsreef deposit. The high P resulting from the focusing of fluids along the major shear zones would cause a reduction in effective normal stress and resistance to sliding. Movement on major zones such as the Peel Fault Zone or those in the pits, all of which dip at high angles, may have resulted in the aftershocks necessary for the formation of shear bands and slickenfibres on planes trending north-northwest, northeast and northwest (Figure 10a, b) suitably orientated for slip. Previous studies have shown that reverse followed by left-lateral slip have occurred on the Peel Fault System (Offler & Williams 1987; Offler et al. 1989; Cao & Durney 1993). Since the present study Figure 12 Shortening (right-hand column) and extension axes suggests that shear bands showing sinistral movement (left-hand column) from shear-band data, (a, b) Reverse slip, (c, are later than all other structures observed in the d) Normal slip, (e, f) Right-lateral slip, (g, h) Left-lateral slip. }

3

f

f

240

R. O F F L E R £ T ^ Z .

deposit, the far-field stress responsible for the earlier thrusting on the Peel Fault may also have been responsible for the small-scale, strike-slip, reverse and normal slip indicated by the slickenfibre lineations and shear bands. At first this seems to be a plausible explanation for the different senses and direction of movement observed at Woodsreef. However, it does not take into consideration the considerable variation in a j shown by the principal stress analysis and the curvature shown by slickenfibre lineations on many planes, some changing in orientation by as much as 90°. The latter implies that there has been a major and progressive change in the near-field principal stress directions. Such a change does not have to occur if the slickenfibre-bearing planes undergo rotation in time. With rotation, planes can be reactivated even if they are not optimally orientated, and slip directions change on them (e.g. the West Transverse Ranges, southern California: Scotti et al. 1991). In this way, with (j\ near-parallel, oblique or almost at right angles to the rotating slip plane at different times, rake can change dramatically. Although there is no direct evidence for rotation of the slickenfibre-bearing planes, it is possible that it occurred, as the deposit is situated between the Peel Fault to the west and the fault-bounded contact with the Nangahrah Formation to the east. Movement on these faults could have produced the rotation required to bring the fibre-bearing planes into orientations which allowed the different senses of movement to take place. CONCLUSIONS The Woodsreef deposit records a kinematic history far more complex than previously reported. It commenced with the formation of high-temperature shear zones in the harzburgite along which dolerite dykes were subsequently emplaced. Influx of hydrothermal fluids along the contact between the harzburgite and dolerite then took place resulting in the exchange of elements between the two rock types and the formation of rodingitised dolerite and serpentinised harzburgite. Subsequently, a variety of veins of extensional origin developed in partially serpentinised harzburgite adjacent to major shear zones localised at rodingite-harzburgite contacts. The shear zones focused further influxes of fluid leading to periodic local increases in P and growth of the veins. Transient changes in near-field stress directions associated with movement on major shear zones caused several generations of veins of different orientation to form. Major influxes of fluid took place later bringing about the formation of massive serpentinite and destruction of veins in some areas. Deformation after the veins were formed led to them being recracked and to the development of slickenfibres within them; slickenfibres also formed in fractures within massive serpentinite, as well as shear bands, at this time. Movement on faults bordering the deposit, rotated the fibre-bearing planes resulting in major changes in sense and direction of slip on them and formation of additional shear bands. Further deformation at low temperatures led to the reactivation of pre-existing f

fractures and shear zones producing the highly polished and striated surfaces seen in all pits. ACKNOWLEDGMENTS We thank Stephen Cox and Dave Durney for constructive comments on the manuscript; Rod Holcombe for providing the software Georient used to obtain the lower-hemisphere equal-area stereographic projection plots and rose diagrams in Figures 7, 8, 9, 10, 11 and 12; Blanka Sperner for providing fault-striae analysis software; Esad Krupic for preparation of thin-sections; Zhiyu Jiang and Hope Ruming for drafting; Geraldine MacKenzie and Sharon Francis for typing various sections of the manuscript; Martin Hand for assistance in the field; and Stephen Cox for making many helpful comments about various aspects of the study. This project was supported by Research Management Committee Grants, University of Newcastle to RO. REFERENCES AITCHISON J. C., IRELAND T. R., BLAKE M . C. JR & FLOOD P.

G. 1992. 530 Ma zircon age for ophiolite from the New England Orogen: Oldest rocks known from eastern Australia. Geology 20, 125-128. BLAKE M. C. JR & MURCHY B. L. 1988. A Californian model for the New England fold belt. Geological Survey of New South Wales Quarterly Notes 16, 7 9 3 - 7 9 5 . CAO X. N. & DURNEY D. W. 1993. Folding and cleavage in the eastern Tamworth belt, Manilla. In: Flood P. ed. Proceedings of the New England Orogen, eastern Australia Conference, pp. 255—256. University of New England, Armidale. CHOUKROUNEP., GAPAIS D. & MERLE O. 1987. Shear criteria and structural symmetry. Journal of Structural Geology 9, 525-530. COGULU E. & LAURENT R. 1984. Mineralogical and chemical variations in chrysotile asbestos veins from the asbestos belt of southern Quebec. Canadian Mineralogist 22, 173-183. COOKE H. C. 1936. Asbestos deposits of Thetford district, Canada. Economic Geology 31, 355-376. Cox S. F. 1995. Faulting processes at high fluid pressures: an example of fault valve behaviour from the Wattle Gully Fault, Victoria, Australia. Journal of Geophysical Research 100, 12841-12859. DRESSER J. A. 1917. Preliminary report on the serpentine and associated rocks of southern Quebec. Geological Survey of Canada Memoir 22. ETHERIDGEM. A. 1983. Differential stress magnitudes during regional deformation and metamorphism: upper boundary imposed by tensile fracturing. Geology 11, 231—234. GLEN R. A. 1971. The geology of the Woodsreef serpentinite, near Barraba, New South Wales. BSc (Hons) thesis, University of Sydney, Sydney (unpubl.). GLEN R. A. & BUTT B. C. 1981. Chrysotile asbestos at Woodsreef, New South Wales. Economic Geology 76,

1153—1169. HARDCASTLE K. C. & HILLS L. C. 1991. BRUTE3 and

SELECT: QuickBASIC 4 programs for determination of stress tensor configurations and separation of

241

VEINS AND ASSOCIATED STRUCTURES, WOODSREEF heterogeneous populations of fault-slip data. and Geosciences

Computers

17, 2 3 - 4 3 .

KATZ M. B. 1986. Tectonic analysis of the faulting at Woodsreef asbestos mine and its possible relationship to the kinematics of the Peel Fault. Australian Journal of Earth Sciences 33, 99-105. KEITH S. B. & BAIN G. W. 1932. Chrysotile asbestos:

I. chrysotile veins. Economic Geology 27, 169-188. LAUBSCHER D. 1964. The occurrence and origin of chrysotile asbestos and associated rocks, Shabani, southern Rhodesia. In: Haughton S. H. ed. The geology of some ore deposits on southern Africa II, pp. 593-625. Geological Society, Johannesburg. LAURENT R. 1975. Petrology of the alpine-type serpentinites of Asbestos and Thetford Mines, Quebec. Schweizerische Mineralogische und Petrographische Mitteilungen 5 , 4 3 1 455. MARRETT R .

&

ALLMENDINGER R . A . 1 9 9 0 .

Kinematic

O'HANLEY D . S. & OFFLER R . 1 9 9 2 . C h a r a c t e r i z a t i o n

of

multiple serpentinisation, Woodsreef, New South Wales. Canadian Mineralogist 30, 1113-1126. O'HANLEY D . S., SCHANDL E. S. & WICKS F . J. 1 9 9 2 . T h e

origin of rodingite from Cassiar, British Columbia, and their use to estimate T and P(H 2 0) o f serpentinisation. Geochimica et Cosmochimica Acta 56, 97—108. POLLARD D. D. & AYDIN A. 1988. Progress in understanding

jointing. Geological 1181-1204.

Society

of America

Bulletin 100,

PROUD J. S. & OSBORNE G . D. 1952. S t r e s s - e n v i r o n m e n t in

the genesis of chrysotile, with specific reference to the occurrence at Woodsreef, near Barraba, New South Wales. Economic Geology 47, 13—23. RALEIGH C. B. & PATERSON M . S. 1 9 6 5 .

Experimental

deformation of serpentinite and its tectonic implications. Journal of Geophysical Research 70, 3965-3985.

Geology

RAMSAY J. G. 1980. The crack-seal mechanism of rock deformation. Nature 284, 135-139.

MERCIER J-C. C. & NICHOLAS A. 1975. Textures and fabrics

RATSCHBACHER L., FRISCH W . , LINZER H - G . ETAL. 1993. T h e

of upper-mantle peridotites as illustrated by xenoliths from basalts. Journal of Petrology 16, 454^487.

Pieniny Klippen Belt in the West Carpathians of northeastern Slovakia: structural evidence for transpression.

analysis of fault-slip data. Journal of Structural 12, 9 7 3 - 9 8 6 .

NICHOLSON R. & POLLARD D. D. 1985. Dilation and linkage

of echelon cracks. Journal

of Structural

Geology 7,

583-590. NICHOLSON R . & EJIOFOR I. B. 1987. T h e t h r e e - d i m e n s i o n a l

morphology of arrays of echelon and sigmoidal, mineralfilled fractures: data from north Cornwall. Journal of the Geological Society of London 144, 79-83. N O R R E L L G . T . , ALEXANDER R . J. & HARPER G. D . 1989.

Planar and composite fabrics in ultramafic and mafic cataclastic rocks from the Josephine Ophiolite, NW California. EOS 70,462-463. OFFLER R. & WILLIAMS A. 1987. Evidence for sinistral

movement on the Peel Fault System in serpentinites, Glenrock station, NSW. In: Leitch E. C. & Scheibner E. eds. Terrane Accretion and Orogenic Belts, pp. 141-151. American Geophysical Union. OFFLER R., O'HANLEY D . S. & LENNOX P. G. 1989. K i n e m a t i c

Indicators in serpentinites — the Peel Manning Fault System, a test case. In: Tectonics and Structural Geology Specialist Group Conference, Kangaroo Island, South Australia, pp. 110-111. Geological Society of Australia. O'HANLEY D. S. 1987. The origin of the chrysotile asbestos veins, southern Quebec. Canadian Journal of Earth Sciences 24, 1—9.

Tectonophysics

226, 471-483.

RIORDON P. H. 1955. The genesis of asbestos in ultrabasic rocks. Economic Geology 50, 67-83. SCHANDL E. S., O'HANLEY D . S. & W I C K S F. J.

1989.

Rodingites in serpentinised ultramafic rocks of the Abitibi greenstone belt, Ontario. Canadian Mineralogist 27, 579-692. SCOTTI O . , NUR A . & E S T E V E Z

R.

1991.

Distributed

deformation and block rotation in three dimensions. Journal of Geophysical Research 96, 12225-12243. SPERNER B . , RATSBACHER L . & OTT R . 1 9 9 3 . F a u l t - s t r i a e

analysis: a Turbo Pascal program package for graphical presentation and reduced stress tensor calculation. Computers and Geosciences 19, 1361-1388. Twiss R. J. & GEFELL M. J. 1990. Curved sU^kenfibers: a new brittle shear sense indicator with application to a sheared serpentinite. Journal of Structural Geology 12, 471—481. ZOBACK M. D. & BEROZA G. C. 1993. Evidence for near-

frictionless faulting in the 1989 (M6.9) Loma Prieta, California, earthquake and its aftershocks. Geology 21, 181-185. (Received 15 May 1996; accepted 26 October 1996)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 242—253.

Ar/ Ar ages of Clarence River Supersuite intrusions from the northern portion of the New England Batholith, southern New England Orogen 40

39

C. J. BRYANT, M. A. COSCA AND R. J. ARCULUS Department of Geology, Australian National University, Canberra, ACT 0200, Australia. Universite de Lausanne, Institut de Mineralogie, BFSH 2, CH-1015 Lausanne, Switzerland. 1

1 2

2

1

Contrasting emplacement ages for Clarence River Supersuite intrusions in the northern part of the New England Batholith have been determined by Ar/ Ar dating of magmatic hornblende. The Towgon Grange, Dumbudgery Creek and Koreelan Creek Granodiorites give ages of 257.5, 253.6 ± 1.0 to 256.9 ± 1.8, and 250.5 ± 1.4 Ma respectively, consistent with previously obtained Late Permian-Early Triassic Rb/Sr biotite ages (255—244 Ma). With an age of 293.1 ±1.8 Ma, the Kaloe Granodiorite is the oldest known I-type intrusion in the southern New England Orogen, being temporally intermediate between those of the S-type Hillgrove and Bundarra Supersuites. The distinct geochemical and magnetic signature of the Kaloe Granodiorite, and the presence of east—westtrending fractures that are absent from other Clarence River Supersuite intrusions, are consistent with an older age. The 293 Ma age obtained for the Kaloe Granodiorite provides strong evidence that the Yugambal terrane is Silurian to Devonian in age. The tectonic events that culminated in the older igneous activity, whether related to backarc/intra-arc rifting, remnant arc magmatism or megafolding, remain unclear. 40

39

Key words: argon dating, Clarence River Supersuite, New England Batholith. INTRODUCTION

The Clarence River Supersuite (Clarence River Suite of Shaw & Flood 1981) is one of three major I-type supersuites in the New England Batholith. It is composed of 12 intrusions, generally < 100 km , that occur at the northeastern (Figure 1) and southern extensions of the batholith. Despite their geographical separation, the northern and southern intrusions are broadly similar, having lower abundances of the alkalis, P, Nb, Ba, light REE, Pb, Th and U than other I-type suites in the New England Batholith (Bryant et al 1997). Their primitive isotopic characteristics (typically initial Sr/ Sr = 0.7031-0.7042 and 8 +1.8 to +6.1: Bryant et al. 1997) imply that they were derived from relatively young, isotopically primitive, but heterogeneous materials including isotopically primitive lower crustal and possibly mantle-derived melts. More rarely the plutons have incorporated a large-ion lithophileenriched and isotopically evolved upper crustal component. Numerous studies have determined the emplacement ages of the Clarence River Supersuite intrusions in the southern portion of the New England Batholith, including Hensel et al. (1985), Roberts and Engel (1987), Collins et al. (1993), Kimbrough et al. (1993), and Shaw and Flood (1993). In contrast, there are few published ages for the northern Clarence River Supersuite intrusions. Using Rb/Sr dating of biotites, Shaw and Flood (1993) indicated that they were emplaced between 255 and 244 Ma, as determined for other intermediate New England Batholith I-type 2

87

86

Nd

intrusions (excluding the leucogranites and small intrusions along the coast, which appear to be younger). More specifically, Shaw and Flood (1993) reported Rb/Sr (biotite) ages of 249 and 250 Ma for the Dumbudgery Creek Granodiorite. An unpublished K-Ar age of 250 Ma referred to by Herbert (1984) for an unspecified intrusion of the Clarence River Supersuite is consistent with the ages outlined above. In addition, McKenzie (1972) obtained a K-Ar (biotite) age of 210 ± 11 Ma for the Koreelan Creek Granodiorite. Here we present Ar/ Ar ages for hornblendes from Clarence River Supersuite intrusions in the northern portion of the New England Batholith, including the Kaloe, Towgon Grange, Dumbudgery Creek and Jenny Lind plutons, together with a Ar/ Ar age for hornblende from the Koreelan Creek Granodiorite. The classification of the Koreelan Creek intrusion is unknown, as it has geochemical characteristics intermediate between the low-K Clarence River and high-K Moonbi supersuites (Shaw & Flood 1981). It is included in this paper, as at the start of this dating program it was classified as a Clarence River Supersuite intrusion (Shaw & Flood 1981). We demonstrate that, with the exception of the Kaloe Granodiorite, the Clarence River Supersuite intrusions were emplaced during a narrow time interval, almost certainly related to the same tectonic episode. The Kaloe Granodiorite is shown to be geochemically, structurally, and temporally distinct from the other intrusions, and can be related to earlier tectonic events within the Yugambal terrane. 40

39

40

39

CLARENCE RIVER SUPERSUITE, NSW REGIONAL GEOLOGY AND TECTONIC DEVELOPMENT The New England Orogen is an accreted continental margin, generated during Devonian and Carboniferous subduction, along or adjacent to the Gondwanan margin (Aitchison et al. 1992; Fergusson & Leitch 1993). The New England Orogen comprises numerous tectonic terranes formed in a variety of subduction-related

243

settings, incorporating intra-oceanic arc, continental arc, forearc and accretionary wedge components (Flood & Aitchison 1988). The detailed geology of the New England Orogen is complex and there is little consensus about the tectonic development (see Aitchison et al. 1992; Collins et al. 1993; Fergusson & Leitch 1993). It is, however, generally agreed that an initial period of island arc activity during the Devonian was followed by Andeanlv v vl Permian I-type volcanic rocks i-mm ' — * nTTI I-type intrusions^" 1:111 S-type intrusions CRS intrusions ^ I I NEO

°

0

Bundjalung Terrane

• •

Anaiwan Terrane Yugambal Terrane (Willowie Creek beds) Gidabal Terrane (Emu Creek Formation) | CRS Intrusions

50 km

Fault River Highway

Other NEB Intrusions Permian I-type volcanic rocks Tertiary volcanic rocks 10

15

20

25 km

Figure 1 (a) Locality map and (b) map showing the distribution of the northern Clarence River Supersuite intrusions. CRS, Clarence River Supersuite; NEO, New England Orogen; NEB, New England Batholith.

244 C. J. BRYANT ETAL. style continental margin volcanism. Models vary as to whether this occurred as a result of a long-lived westerly dipping Benioff zone (Leitch 1974; Cross et al. 1987), or was associated with a change in subduction polarity (Aitchison et al. 1992). Most authors recognise a major change in the tectonic pattern at approximately 300 Ma, although there is little agreement about tectonic events following this change. Aitchison and Flood (1992) argued that during the Late CarboniferousEarly Permian there was a change from high-angle to oblique convergence, resulting in the formation of numerous narrow, rapidly filled sedimentary basins along the Peel-Manning fault system. Following Murray et al. (1987), Fergusson and Leitch (1993) suggested that the convergence was replaced by extension after subduction of a mid-ocean ridge and associated transform faults, converting eastern Australian into a dextral transform margin. Fergusson and Leitch (1993) indicated that subsequent dextral movement at ca 290 Ma, possibly along the Gogango— Baryugil Fault, promoted large-scale folding, which resulted in ~500 km of lateral displacement of the subduction complex and the formation of the TexasCoffs Harbour Megafold. According to their model, this tectonic regime was relatively short-lived, being replaced by a new convergent margin and associated Andean-style volcanism at ca 280 Ma. They proposed that, during this period, the southern New England Orogen was in a backarc setting. A contrasting model is given by Collins et al. (1993) who proposed that backarc extension immediately followed the Late Carboniferous Andean-style volcanism due to steppingout of the arc. They did not recognise a major phase of strike-slip motion, or megafolding between 300 and 270 Ma, arguing instead that the Texas-Coffs Harbour Megafold was produced during the Hunter-Bowen Orogeny at ca 260 Ma and was immediately followed by a period of extension and I-type volcanism and plutonism. LOCAL GEOLOGY AND GENERAL CHARACTERISTICS The Kaloe Granodiorite is a predominantly tonalitic intrusion that forms a narrow, elongate body (2-4 km wide, ~30 km long) near the western margin of the Yugambal terrane. It is chemically distinct from other members of the Clarence River Supersuite, containing higher abundances of Al, Na, Sr and Ga and lower Fe, Sc, and V at equivalent Si0 (Bryant et al. 1997). It is characterised by a very low magnetic signature, consistent with the lower Fe 0 /Fe0 ratios observed for this intrusion. In contrast to other Clarence River Supersuite intrusions, it contains abundant closely spaced and approximately east-west-trending fractures. The Towgon Grange Granodiorite also intrudes the Yugambal terrane, occurring near the western margin of the Clarence-Moreton Basin directly south of the Dumbudgery Creek Granodiorite (Figure 1). This intrusion consists predominantly of quartz diorite and tonalite, and has depleted geochemical and primitive isotopic characteristics (Bryant et al. 1997). 2

2

3

The Dumbudgery Creek Granodiorite was intruded along the margin of the Yugambal and Bundjalung terranes, forming a stitching pluton between the two (Flood & Aitchison 1988). Along its northern margin it intrudes the Permian Drake Volcanics (Stewart 1977). The Dumbudgery Creek Granodiorite is a compositionally and geochemically variable intrusion. Two chemically distinct units are present; (i) a tonalite-trondhjemite; and (ii) a high-K granodioriteleucomonzogranite (most abundant). Additionally, enclosed within the intrusion are several elongate zones of mafic rocks including dolerite, porphyritic hornblende gabbro, biotite-free quartz gabbro, and foliated or massive hornblende ± biotite quartz diorite. The margins of the dolerite zones are characterised by abundant, small, angular microgranular enclaves and/ or a narrow zone of hybridised diorite and tonalite. The abundance and angularity of the enclaves implies that they were formed by magma mingling close to their original site of formation. The Jenny Lind and Koreelan Creek Granodiorites intrude the Emu Creek Formation, with the former intrusion being unconformably overlain by the Late Triassic {ca 200 Ma) Laytons Range Conglomerate, a lower unit of the Clarence-Moreton Basin succession. The Koreelan Creek Granodiorite is intruded by the Rivertree Monzogranite. PETROGRAPHY AND SAMPLE SELECTION The Clarence River Supersuite granites form mediumgrained, equigranular intrusions. In most samples quartz and K-feldspar form irregular interstitial grains between subhedral to euhedral crystals of plagioclase, hornblende, biotite ± clinopyroxene ± orthopyroxene. Accessory phases include magnetite, ilmenite, apatite, zircon, titanite (typically secondary) and allanite. The Clarence River Supersuite intrusions contain early formed pyroxenes, followed by magmatic hornblende, then biotite. The occurrence of early formed pyroxenes in all plutons implies that the magmas were initially undersaturated with respect to water. Hornblende barometry calculations indicate that they only became water-saturated at shallow depths (pressures <0.24 GPa: Bryant et al. 1997). The clinopyroxene is variably replaced by patchy green actinolitic hornblende (bluish-green) or a fibrous amphibole (colourless-pale green), possibly actinolite. Magmatic hornblende can be discerned from the actinolitic hornblende on the basis of colour, pleochroism and birefringence. Samples chosen for this study were selected on the basis of: (i) a high proportion of isolated hornblende grains as glomerophyric aggregates, as these tend to contain abundant intergrowths with both biotite and pyroxene; (ii) a high proportion of magmatic hornblende to actinolitic amphibole, as discerned in thin-sections; and (iii) freshness. In most cases, samples that fill these criteria are characterised by low clinopyroxene abundances (Table 1), consistent with hornblende being on the liquidus for some time. The hornblende grains range in size from 0.5 to 3 mm.

C L A R E N C E RIVER S U P E R S U I T E , N S W

245

Table 1 Modal analyses of Clarence River Supersuite samples.

Quartz Plagioclase K-feldspar Hornblende Biotite Pyroxene Other

R72006* Tonalite

R72177 Tonalite

R70221 Quartz diorite

R72270 Quartz diorite

R72334 Quartz diorite

R72358 Granodiorite

26.3 58.7 3.2 1.7 9.9 np 0.2

24.2 56.8 1.8 6.4 9.7 0.1 1.0

6.7 24.2 2.5 31.6 np 34.6 0.4

4.8 57.9 6.1 25.1 2.2 np 3.9

10.2 59.8 0.4 26.2 2.0 0.1 1.3

15.6 46.2 8.5 11.5 18.0 0.1 0.1

*R72006, Kaloe (454600E, 6736730N); R72177, Towgon Grange (464830E, 6743940N); R70221, Dumbudgery Creek (451250E, 6767660N); R72270, Dumbudgery Creek (452110E, 6765580N); R72334, Jenny Lind (445870E, 6823030N); R72358, Koreelan Creek (431750E, 6833040N).

In the case of the Dumbudgery Creek Granodiorite, it was not possible to find a suitable sample from either the tonalitic or granodioritic unit. The two samples chosen for analysis (R70221, R72270) come from the margins of the mafic zones. On the basis of field relations, these samples are interpreted to have been formed by mixing of felsic and mafic magmas. Although these samples are not geochemically representative of the predominant high-K, LILE and LREE phase of the intrusion, they should adequately represent the age of the Dumbudgery Creek Granodiorite. Samples R72270 and R70221 do not fit the general petrographic description outlined above. Sample R72270 is a foliated quartz diorite that contains abundant euhedral hornblende and plagioclase, with titanite being the most abundant accessory phase. Actinolitic hornblende is absent and biotite is very minor (Table 1). Sample R70221 is an appinite that is characterised by abundant hornblende megacrysts (<3 cm), which contain small inclusions of pyroxene but essentially no actinolitic hornblende. Abundant pyroxene (Table 1) that occurs as both inclusions in anorthitic plagioclase oikocrysts and within the groundmass pyroxene (different population), is characterised by remarkably little alteration. Notably, the latter coexists with groundmass hornblende (1-2 mm). In all cases the rock samples chosen are either tonalite or quartz diorite in composition and, possibly with the exception of R72334 (Jenny Lind Granodiorite), all contain abundant magmatic hornblende (Table 1). METHODS Preparation The Ar/ Ar ages came from typically pure mineral separates of hornblende obtained by crushing handsized samples, then sieving. The 120-140 mesh fraction was selected for further concentration, as it reduced contamination from intergrowths by pyroxene and/or its alteration products during subsequent hand-picking. This fraction was washed with distilled water, and oven40

39

dried at 80°C. The hornblende was then concentrated by magnetic separation and ultrasonic washing in distilled water. Blocky and visually homogeneous grains were then hand-picked to produce a typically high purity hornblende separate and to minimise the amount of actinolitic amphibole. This material was finally washed in distilled water and then dried at 80°C. With the exception of sample R70221 and a duplicate analysis of R72270, the sample weights ranged between 32.1 and 49.0 mg. 40

Ar/ Ar analyses 39

The Ar/ Ar analyses were made at the Universite de Lausanne. Samples, together with the standards, were irradiated for 60 MW hours in the central thimble position of the TRIGA reactor in Denver, USA (Dalrymple et al. 1981). The samples were incrementally heated in the furnace and the gas was expanded and purified using activated Zr/Ti/Al getters and a metal cold finger maintained at liquid nitrogen temperatures. All analyses were made using a low blank, double vacuum resistance furnace and metal extraction line connected to an MAP 250-50 mass spectrometer equipped with an electron multiplier. Time zero regressions were fitted to data collected from seven scans over the mass range 40 to 36. Peak heights above backgrounds were corrected for mass discrimination, isotopic decay, and interfering Ca-, K- and Cl-derived isotopes of argon. Blanks were measured at several temperatures and subtracted from the sample signal. For mass 40, blank values ranged from 4xl0~ mol below 1350°C to 9xl0" mol at 1650°C. Blank values for masses 36—39 were below 2xl0" mol for all temperatures. Isotopic production ratios for the TRIGA reactor were determined from analyses of irradiated CaF and K S 0 and found to be within error of those reported in Dalrymple et al (1981). Correction for the neutron flux was determined with an intralaboratory precision of 0.25% using the standard MMHB-1, assuming an age of 520.4 Ma (Samson & Alexander 1987). A mass discrimination correction of 1.008 amu 40

39

15

15

17

2

2

4

246

C.J. BRYANT ETAL.

Table 2 Argon release data for Clarence River Supersuite hornblendes.

T(°C)

40Arxloi6

mol

39

Arx 1017 mol

38

Arx 1017 mol

37

Arx 1017 mol

36

Arx 1018 mol

% 40 Ar*

39 Ar (% of total)

40

Ar*/ Ar

Age±2a

8.18 8.33 10.00 10.59 10.79 10.75 10.63 10.69 10.65 10.62 10.84 10.80 10.78 10.67 10.77 10.63 10.78 11.70

195.5 ±5.9 199.0 ±5.1 236.2 ±4.6 249.3 ±3.0 253.7 ±1.8 252.7 ±1.4 250.2 ±1.3 251.4 ± 1.2 250.5 ±1.2 249.8 ±1.3 254.7 ±1.3 253.8 ±1.3 253.5 ±1.2 250.9 ±1.2 253.2 ±2.3 250.0 ±1.7 253.4 ±2.1 273.5 ±14.8

39

R72358 Hornblende, 42.67 mg, J = 0.01399 + 0.25% 800 850 875 900 925 950 975 985 1000 1015 1025 1050 1075 1100 1150 1200 1400 1601

8559 + 15 4928 ±12 545 ±9 4949 + 12 3901 ±10 553 ±4 3878 + 11 3055±12 428 ±4 5008 ±8 4031 ±19 805 ±8 11512 ± 8 9465 ±21 2883±13 54100 ±44 46662±56 18410 ±40 173729 ±203 156176 ±204 65741±152 135704±170 122791 ±119 53478 ±71 13913 +147 128458±167 56730 ±131 102448 ±140 95242±171 42689 ±88 55472 ±73 51149 ±81 22762 ±58 67735 ±86 62479 ±89 27918 ±71 102477 ±96 94079 ±93 42365 ±87 93233 ±94 86233 ±96 39241 ±85 39087 ±193 35519 ±100 15761 ±87 29115 ±21 26646 ±46 11655 ±42 18785 ±15 16990±36 7364±14 2484 ±8 1692 ±13 682 ±6

12575 ±64 1566 ±21 19733 ±74 629 ±12 20443 ±71 337 ±11 35905±126 353 ±10 91224 ±129 700 ±16 407732 ±499 2501 ±24 1286476 ±2039 6275 ±37 984883 ±1310 4328 ±27 1001028±1126 3658 ±37 737199 ±787 2557 ±15 395852 ±482 1142 + 12 490999 ±514 1498 ±13 768897 ±983 2554 ±16 703220±755 2438 ±14 286757±1247 1100 ±13 217760 ±424 893 ±13 144490 + 190 573 ±12 14100 ±92 211+15

47.0 65.5 78.4 84.8 88.2 92.2 95.1 96.2 97.8 98.2 99.5 99.1 98.5 98.2 97.4 96.8 97.0 79.3

0.5 0.4 0.3 0.4 1.0 4.9 16.4 12.9 13.5 10.0 5.4 6.6 9.9 9.1 3.7 2.8 1.8 0.2

Age Plateau (975-1015°C) = 250.5 ±1.4 Ma Total fusion age = 251 Ma R72006 Hornblende, 32.14 mg, J = 0.01399 ±0.25% 800 850 875 900 925 950 975 985 1000 1015 1025 1050 1075 1100 1150 1200 1400 1601

9268 ±12 4104 ±25 4516 ±5 2705 ±16 3500 ± 6 2235 ± 7 5361 ±10 3530 ±12 16142 ±13 11033 ±25 53597 ±63 38598 ±65 110021 ±272 81786 ±143 50121 ±60 65656 ±61 52396 ±86 40275 ±71 38333±230 29650 ±127 22922 ±15 17745 ±38 27276 ±80 21046 ±40 49170 ±52 38049 ±69 43288 ±169 33427 ±148 14031 ±23 10445 ±25 9582 ±21 6876 ±25 10852 ±20 7356 ±21 2088 ± 5 947 ±7

263 ±10 183 ±5 180 ±5 433 ±3 1774 ±14 6433 ±27 13142 ±48 8186+31 6473 ±25 4856±50 2912 ±12 3440±17 6426 ±25 5931 ±35 1922 ±14 1214 ±12 1248 ±10 172 ±5

19037 ±91 2615 ±30 17012 ±50 565 ±12 17213 ±72 284 ± 11 39295 ±108 380 ±10 154122 ±154 1146 ±20 556150 ±727 3209 ±33 1163656±1888 5652 ±35 704763 ±1201 2948 ±26 560688 ±1099 2216 ±26 405537±1921 1468 ±34 244164 ±255 1016 ± 12 298117±256 1114 ± 12 556139±1006 1934 ±24 513359±2073 1877 ±26 156562 ±275 734 ± 11 100982±211 547 ± 11 108266±197 807 ±19 13404 ±70 302 ±14

18.2 65.9 79.9 84.7 86.5 90.4 93.1 95.1 95.8 96.9 95.2 96.5 97.2 96.4 93.2 91.3 85.8 62.3

1.0 0.7 0.6 0.9 2.8 9.7 20.5 12.5 10.1 7.4 4.4 5.3 9.5 8.4 2.6 1.7 1.8 0.2

4.13 11.05 12.57 12.96 12.77 12.67 12.64 12.57 12.58 12.64 12.41 12.62 12.68 12.61 12.65 12.85 12.78 13.87

101.3 ± 11.7 259.3 ±7.3 292.3 ±3.9 300.7 ±3.1 296.4 ±1.9 294.4 ±1.7 293.7 ±1.8 292.3 ±1.5 292.5 ±1.5 293.8 ±3.4 288.8 ±1.9 293.2± 1.8 294.6 ±1.7 293.2 ±2.5 293.9 ±2.6 298.3 ±2.1 296.7 ±3.6 320.0 ±17.2

Age plateau (950-1015°C) = 293.1 ±1.8 Ma Total fusion age = 291.4 Ma R72270 Hornblende, 49.01 mg, J = 0.01417 ±0.25% 800 850 875 900 925 950 975 985 1000 1015

6860 ±10 2553 ± 6 2037 ±7 1870 ±5 3053 ±5 7015 ± 12 20032 ±25 28111 ±214 46272 ±40 88960 ±84

2879 ±15 1564 ±9 1303 ± 6 1388 ±9 2412 ±9 5762 ±19 17669 ±42 25441 ±134 42254 ±40 81685 ±124

658 ±8 262 ±4 216 ±5 293 ± 3 786 ±6 2151 ±13 6578 ±25 9221 ±30 15985 ±29 31820 ±79

18736 ±78 6384 ±53 5339 ±37 7135 ±56 15702±58 39539 ±119 126888±244 185977±251 317022±405 609658 ±870

2314 ±28 292 ±17 282 ±14 188 +14 110 ± 10 233 ± 11 494 ±18 690 ±12 1123 ±14 2041 ±27

2.4 68.2 61.2 73.3 93.4 94.6 97.7 97.9 98.2 98.6

1.6 0.9 0.7 0.8 1.3 3.2 9.7 14.0 23.2 44.8

0.58 11.16 9.59 9.91 11.87 11.56 11.12 10.87 10.80 10.78

14.7 ± 15.2 264.8 ±10.1 229.9 ±10.6 237.1 ±13.2 280.4 ±3.3 273.7 ±2.4 264.0 ±1.9 258.3 ±2.1 256.9 ±1.4 256.5 ±1.3

Age plateau (985-1015°C) = 256.9 ± 1.8 Ma Total fusion age = 254.3 Ma

CLARENCE RIVER SUPERSUITE, NSW

247

Table 2 continued

T(°C)

40

Arx 1016 mol

39

Arx 1017 mol

38

Arxl0 1 7 mol

37

Arxl0 1 7 mol

36

Arxl0 1 8 mol

% 40 Ar*

12636±129 2378±28 12422±153 357 ±19 10971 ±131 234 ±27 17482±130 253 ±10 56674 ±331 672 ±19 271290±2810 2429 ±17 826542 ±413 5774 ±28 1375782±1301 4256 ±26 1086761 ±230 3290±18 489578±1705 1866 ±42 227253 ±1306 978 ±13 215037±1695 873 ±12 393577 ±4439 1586 ±29 763610 ±860 2591 ±21 370628 ±4424 1713 ±30 916 ±12 144572 ±708 244319 ±4740 1627 ±26 6628 ±78 193 ± 13

12.0 73.2 76.7 79.9 77.1 82.3 88.3 97.6 98.1 96.0 94.5 95.4 95.7 97.1 93.3 88.4 87.2 63.5

39

Ar(% 40Ar*/ 39 of total) Ar

Age±2a

R72177 Hornblende, 32.81 mg, J = 0.01408 ± 0.25% 800 850 875 900 925 950 975 985 1000 1015 1025 1050 1075 1100 1150 1200 1400 1601

7879 ±12 3580 ±5 2592 ±5 3032 ±8 6750 ±15 28587 ±273 90539±155 77259 ±52 65682 ±69 42930 ±67 20449 ±21 19532 ±17 37735±176 57888 ±42 32529 ±78 13594 ±15 22662 ±168 1423 ±5

4035 ±17 2463 ±15 1765 ±16 2083 ±8 4849±15 22561 ±170 76011 ±104 67128 ±52 57476 ±31 37738 ±67 18204 ±44 17196 ±33 32957±176 51653 ±60 28387 ±106 11281 ±23 18332 ±101 577 ±8

1183 ±13 717 ± 10 528 ±13 853 ±5 3174 ±18 17803 ±133 62141 ±83 55163 ±58 48057 ±69 31963 ±81 15072 ±33 14175 ±25 27930 ±178 44155 ±52 24839 ±146 9975 ±21 16160 ±191 457 ±6

0.9 0.5 0.4 0.5 1.1 5.0 16.7 14.8 12.6 8.3 4.0 3.8 7.2 11.4 6.2 2.5 4.0 0.1

2.35 10.67 11.30 11.69 10.81 10.51 10.59 11.39 11.35 11.02 10.70 10.92 11.04 10.98 10.78 10.74 10.87 15.77

58.8 ±10.7 252.6 ±8.8 266.4 ±15.8 274.8 ±3.9 255.7 ±5.3 248.9 ±4.3 250.7 ±1.8 268.3 ±1.5 267.5 ±1.4 260.2 ±2.0 253.3 ±2.7 258.0 ±2.2 260.7 ±3.2 259.4 ±1.5 255.0±2.6 254.0 ±2.7 256.9 ±2.1 361.8 ± 14.1

No age plateau Total fusion <age = 257.5 Ma R72334 Hornblende, 33.,16 mg, J = 0.01408 ±0.25% 875 900 925 950 975 985 1000 1015 1025 1050 1075 1100 1150 1200 1400

1944 ±4 1627 ±7 4199 ±13 12229 ±9 31538 ± 117 28658±185 47128±42 69521 ±61 38628±230 57757 ±73 46440 ±46 49417 ±76 55218 ±65 17064 ±12 39279 ±50

1033 ±7 1125 ±12 3191 ±13 9829 ±20 26465 ±87 24778 ±110 41003 ±46 61442 ±69 34073 ±174 52303 ±77 41869 ±48 44570 ±71 49650 ±67 15174 ±33 34894 ±52

460 ±6 696 ± 11 3198 ±18 10435 ±29 28742±112 27568±194 46194 ±94 68959 ±77 37484 ±246 54859 ±114 42925 ±69 48034 ±56 53450 ±87 16230 ±44 37928±54

363 ±16 14369 ±128 164 ±22 20585 ±145 60028±332 446 ± 11 154878 ±630 1091 ±13 375572 ±1382 1951 ±25 327804±3664 1640 ±15 609306 ±403 2820 ±22 802011 ±927 3771 ±24 458710 ±5424 1640 ±15 768180 ±524 2754 ±22 48756 ±270 2096 ±20 258181 ±542 2293 ±18 424663 ±300 2609 ±18 203419 ±839 825 ±13 470943 ±1640 1831 ±32

50.5 80.0 79.8 83.5 91.0 92.0 92.4 93.0 96.7 96.3 87.5 90.4 92.0 95.0 95.6

0.2 0.3 0.7 2.2 6.0 5.6 9.3 13.9 7.7 11.8 9.5 10.1 11.2 3.4 7.9

9.60 11.71 10.63 10.50 10.94 10.73 10.72 10.61 11.06 10.73 9.70 10.05 10.29 10.78 10.85

228.6 ±16.8 275.4 ±8.1 251.5 ±5.0 248.7 ±2.1 258.6 ±2.7 253.9±2.8 253.7 ±1.6 251.1 ±1.5 261.1 ±2.8 253.9 ±1.6 231.1 ± 1.5 238.8 ±1.7 244.1 ±1.5 254.9 ±1.6 256.5 ±1.8

No age plateau Total fusion age = 249.7 Ma All isotopes corrected for blanks, mass discrimination and radioactive decay. The 2cy errors on ages do not include the 0.25% error on J.

was determined by online measurement of air and applied to the data. RESULTS The results of the 4 0 Ar/ 3 9 Ar step heating experiments are presented in Table 2 and Figures 2-4. All ages and trapped 40 Ar/ 36 Ar ratios are reported with 2 a errors. Age plateaus were determined using the critical value test (Cosca et al. 1991). The 3 9 Ar/ 4 0 Ar vs 36 Ar/ 40 Ar isochrons (Turner 1971) were calculated using a least squares fit with correlated errors (York 1969). Most of the samples yielded relatively flat 40 Ar/ 39 Ar age spectra

with good agreement between the isochron, integrated total fusion and plateau ages, indicating a lack of any disturbance to the K/ Ar system. For all samples except R72334 and R72270, the heating steps forming the isochrons are clearly defined and the linear regressions yielded trapped 40 Ar/ 36 Ar ratios consistent with that of modern atmosphere (295.5). Sample R72006 has plateau and isochron ages of 293.1 ± 1.8 and 291.2 ± 2 . 2 Ma, respectively, while samples R72358, R70221, R72270 and R72177 all have plateau and isochron ages of approximately 250 Ma (Table 2; Figures 2, 3). An analytical problem occurred during the analysis of sample R72270, where several

248

C.J. B R Y A N T

ETAL. Sample R72334 yielded an 40Ar/39Ar age spectra and Ca/K ratios consistent with a compositionally complex hornblende. Four heating steps between 975 and 1015°C (26% 39Ar) yielded the most precise isochron (age = 230 ± 7.2 Ma), but the relatively tight clustering of the data points results in large uncertainties for both the age and trapped 40 Ar/ 36 Ar ratio (Figure 3b).

heating steps were lost. Clustering of data points results in large uncertainties in the age and the trapped 40 Ar/ 36 Ar value for this sample. Despite the loss of some gas, the 'plateau' age of 256.9 ± 1.8 Ma is similar to another analysis of this sample that gave a total fusion age of 254.0 Ma (Figure 4) and is interpreted as a reliable cooling age for this sample.

ir

400

"15 *

"10 ^

•

Plateau steps Age = 253.5 ±4.7 Ma MSWD = 0.8 40Ar/36Artr = 541 ± 320

0.0030

a s 360 <D <bO

< O

R72270 Hornblende

0.0020

320

Ch <Ch 280

- 256.9 ± 1 . 8 Ma -

0.0010

0

0 0

o

240

(b)

Integrated Age = 254.3 Ma

0.0000

200 0

10

20

30

40

50

60

70

80

90

100

0.030 0.040 0.050 0.060 0.070 0.080 0.090

Cumulative % 39Ar Released

f

300

39Ar/40Ar

20 10 u 0

L

a

0.0035

- 253.6 ± 1 . 0 Ma-

<D

250 " <W)

<

-

<

R70221 Hornblende

£ CU200 <Oi

*

0.0030-

Nw

Steps 1100°-1250°C Age = 252.2 ± 2.8 Ma MSWD = 0.5 40Ar/36Artr = 260 ± 58

0.0025Nv

0.0020-

x Nv f1

<n 0.00150.0010-

150 "

0.0005Integrated Age = 250.9 Ma

0

0.0000 0

10

20

30

40

50

60

70

80

90

(d) 0.0050

100

0.01

50

a

200

Steps 1050°-1400°C Age = 260.0 ±1.2 Ma MSWD = 4.9 40Ar/36Artr = 252±18

0.0024

Uh

320 G <D

240

0

o

<

Oh 280

0.0032

<1-H

R72177 Hornblende

360

bO

<

^

U

400 s

0.020

0.015

39Ar/40Ar

Cumulative % 39Ar Released

< 0.0016

Integrated Age = 257.5 Ma

L

0

(e) 0

rv

0.0008

10

20

30

40

50

60

70

80

Cumulative % 39Ar Released

90

100

0.0000

(f) 0.040

0.050

0.060

0.070

0.080

0.090

39Ar/40Ar

Figure 2 (a, c, e) 4 0 Ar/ 3 9 Ar age spectra and Ca/K plots, and (b, d, f) 3 6 Ar/ 4 0 Ar vs 3 9 Ar/ 4 0 Ar isotope correlation diagrams for hornblendes, (a, b) R72270, Dumbudgery Creek Granodiorite. (c, d) R70221, Dumbudgery Creek Granodiorite. (e, f) R72177, Towgon Grange Granodiorite.

CLARENCE RIVER SUPERSUITE, NSW

249

all of the samples dated, the hornblende 40 Ar/ 39 Ar ages are interpreted as the time the rock cooled below the argon closure temperature. For typical compositions, the closure temperatures in hornblende grains with diffusion radii of 80 |Lim vary between ~580 and 490°C for cooling rates of 500 to 5°C/10 6 y (Harrison 1981).

Because of the uncertainties with the isochron calculation, the integrated age is considered the most reliable estimate of the cooling age. All samples except R72334 have relatively constant Ca/K ratios but record low Ca/K ratios in the early degassing steps, which are typical of minor contamination, such as intergrowths of biotite or other phyllosilicates. For r50 u

0.0016

400

)

Steps 975°-1015°C

1

1V

Age = 230 ± 7 Ma MSWD = 0.4

R72334 Hornblende 360

0.0012

40Ar/36Artr = 646±110

<H 320 0.0008

Integrated Age = 249.7 Ma

o

280

°

0.0004 240 200

(aI 0

(b) 0.0000

10

20

30

40

50

60

70

80

90

-

0.040

100

0.050

0.060

0.070

0.080

-

0.090

39Ar/40Ar

Cumulative % 39Ar Released 0.0020

Plateau steps Age = 250.1 ±1.4 Ma

400

0.0016

MSWD = 2.0

R72358 Hornblende

360

40Ar/36Artr = 311 ±52

< § 0.0012 <

320

0.0008 280

• 250.5 ± 1.4 M a i 0.0004

240 200

Integrated Age = 251.0 Ma

(C) 0

10

20

30

40

50

60

70

80

90

0.0000

100

0.060

0.070

400 •

40

^

20

U 0.0028

0

Age = 291.2 ±2.2 Ma

R72006 Hornblende

MSWD = 0.6 0.0020

40Ar/36Artr = 328±27

-293.1 ± 1 . 8 M a — | o

0.0012 a.

<

0.100

Plateau steps

0

<3 S 360 <D bO < 320

0.090

0.080 39Ar/40Ar

Cumulative % 39Ar Released

280 240 200

0.0004 (f) 0

10

20

30

40

50

60

70

80

Cumulative % Ar Released

90

100

0.040

0.050

0.060

0.070

0.080

39Ar/40Ar

Figure 3 (a, c, e) 4 0 Ar/ 3 9 Ar age spectra and Ca/K plots, and (b, d, f) 3 6 Ar/ 4 0 Ar vs 3 9 Ar/ 4 0 Ar isotope correlation diagrams for hornblendes, (a, b) R72334, Jenny Lind Granodiorite. (c, d) R72358, Koreelan Creek Granodiorite. (e, f) R72006, Kaloe Granodiorite.

250

C. J. B R Y A N T ETAL.

equivalent, the Dumbudgery Creek and Towgon Grange Granodiorites probably represent two different U intrusions. 400 The A r / A r spectrum for the Jenny Lind Granodiorite (R72334) is internally discordant with an R72270 Hornblende integrated age of 249.7 Ma. The similarly discordant 300<D Ca/K ratios for this sample are indicative of a <bfl compositionally complex amphibole separate. In this pluton it was difficult to find a sample that did not a. contain abundant glomerophyric aggregates and <a* actinolitic amphibole. As previously discussed, the 100 isochron for this sample is poorly constrained and Integrated Age = 254 Ma the integrated age of 249.7 Ma is the most reliable age estimate. Typically, the I-type intrusions from the three 0 10 20 30 40 50 60 70 80 90 100 main supersuites have ages of 255-244 Ma (Shaw & Flood 1993), with younger I-type intrusions in Cumulative % A r Released the northern portion of the New England Batholith being exclusively leucogranites. The 249.7 Ma age Figure 4 A r / A r age spectrum and Ca/K plot for the duplicate analysis of R72270. determined for the Jenny Lind Granodiorite is consistent with those results. However, another dating technique may be required to more precisely determine DISCUSSION its age. The overlapping plateau age, isochron, and integFor the purposes of this discussion the ages obtained in rated ages (250 ± 1 Ma) from the Koreelan Creek this study are divided into two groups—Late Granodiorite (R72358) indicates crystallisation at the Carboniferous-Early Permian and Late Permian-Early Late Permian-Triassic boundary (Jones 1996). This Triassic—similar to the two groups identified by Shaw Ar/ Ar age is considerably older than the K-Ar age and Flood (1981). of 211 ±11 Ma obtained by McKenzie (1972). Thomson (1976) indicated that the K-Ar age represents Late Permian-Early Triassic the minimum emplacement age, and is most probably too young given that the Laytons Range Conglomerate, The Ar/ Ar dates obtained from the Dumbudgery a lower unit of the Clarence-Moreton Basin succession, Creek Granodiorite (R70221, R72270), Towgon Grange is ca 200 Ma. The Koreelan Creek Granodiorite is Granodiorite (R72177), Jenny Lind Granodiorite intruded by the Rivertree Monzogranite, which has been (R72334), and Koreelan Creek Granodiorite (R72358), dated at 215 ± 11 Ma (K-Ar) by McKenzie (1972), all fall within this group. although this too represents the minimum emplacement The A r / A r hornblende plateau dates of age. The Rivertree Monzogranite forms a texturally and 253 ± 1.0 Ma and 256.9 ± 1.8 Ma obtained for the geochemically distinct phase that is similar to the Dumbudgery Creek Granodiorite samples R70221 and adjacent Stanthorpe Monzogranite, being separated R72270 are similar to the Rb/Sr (biotite) ages of 249 from the latter by a major shear zone. The Rb/Sr dates and 250 Ma reported by Shaw and Flood (1993) for this for the Stanthorpe Monzogranite indicate it has an intrusion. These results not only confirm the Late age of 244—242 Ma, within the 244-236 Ma range Permian-Early Triassic age of the Dumbudgery Creek observed for leucogranites in the northern portion of Granodiorite, but indicate that the mafic rocks cooled the New England Batholith (Shaw & Flood 1993). The and were probably intruded at a similar time as the date of 250.1 ± 1.4 Ma implies the Koreelan Creek main granodioritic phase. These ages are consistent Granodiorite was intruded prior to the Rivertree with the stitching of the Yugambal and Bundjalung Monzogranite and is consistent with field observations. terranes in the Late Permian, although terrane accretion This date is equivalent with the ages of other Clarence could have occurred anytime between Devonian and River Supersuite intrusions and those of the nearby Late Permian, depending on the actual age of the Moonbi Supersuite intrusions, namely the Bungulla and Yugambal terrane. Undercliffe Falls Monzogranites, which have dates The 257.5 Ma age obtained for the Towgon Grange of 243-248 and 245 Ma respectively (Shaw & Flood Granodiorite is similar to that of the Dumbudgery Creek 1993). Granodiorite, confirming that the two intrusions were The ages determined for the Dumbudgery Creek, emplaced during the same magmatic episode. Barnes Towgon Grange, Jenny Lind and the Koreelan Creek and Willis (1987) speculated that both granodiorites are Granodiorites (257-250 Ma) are similar to those the exposed portions of a single high-level intrusion. determined for the majority of I-type intrusions in the Although they may be in contact at depth, and were southern New England Orogen. This is consistent with intruded at similar times, the two plutons are geo- them being generated during the same tectonic episode, chemically distinct and have incorporated different possibly during a major phase of extension that led to source components during in their genesis (Bryant the formation of the Clarence-Moreton Basin (Collins et al. 1997). Consequently, although temporally etal 1993). 40

H

39

40

39

40

40

39

40

39

39

39

CLARENCE RIVER SUPERSUITE, NSW Late Carboniferous-Early Permian The good agreement between the plateau, total fusion and isochron ages from the Kaloe Granodiorite (R72006) implies that the date obtained for this intrusion is a valid emplacement age. The 293.1 ± 1.8 Ma date for the Kaloe Granodiorite is the oldest recorded for an I-type intrusion in the southern New England Orogen, being intermediate between those of the S-type Hillgrove (302 ± 4 Ma, Collins et al 1993; 303 ± 3 Ma, Kent 1994) and Bundarra Supersuites (280-270 Ma, Shaw & Flood 1982). This age overlaps the range of K-Ar (biotite, hornblende) ages obtained from the I-type granites of the Urannah Batholith in the northern New England Orogen (Webb & McDougall 1968; Gust et al 1993), although U/Pb dating indicates that most crystallisation ages may be - 2 0 million years older (Black 1994; C. M. Allen pers. comm. 1996). The interpretation of events that led to the emplacement of the Kaloe Granodiorite intrusion is problematical. This results principally from the lack of consensus about the tectonic development of the southern New England Orogen during this period. On the basis of current models there are two possibilities. (1) The Kaloe Granodiorite was produced within, and as a consequence of, the Texas-Coffs Harbour megafolding event. This model requires that the megafolding event was initiated prior to 291 ± 2 Ma. The Kaloe Granodiorite is not foliated, being characterised by abundant parallel fractures, implying brittle rather than ductile deformation. However, this is not necessarily inconsistent with the model, as brittle versus ductile structures are potentially controlled by their level of emplacement, the stress regime, temperature of the crust, and the mode of emplacement (Fowler & Lennox 1992; Petford 1996). Alternatively, the intrusion may not have been deformed during megafolding, but was fractured during a subsequent deformational event. (2) The Kaloe Granodiorite was intruded prior to the Texas-Coffs Harbour megafolding event and subsequently incorporated into that structure. This scenario is possible in the models of both Collins et al (1993) and Fergusson and Leitch (1993), although in the latter case, a marginally younger initiation age for the folding event would be required. Collins et al. (1993) have proposed that the thermal perturbation required for the formation of the Bundarra and Hillgrove Supersuites resulted from either remanent arc magmatism following the eastward migration of the arc at the cessation of subduction/accretion in the southern New England Orogen, or backarc extension with the arc being located farther east (Dirks et al. 1992). Collins et al. (1993) favoured the former, arguing that bimodal volcanic rocks indicative of a riftrelated setting are not evident until ca 280 Ma. Additionally, they argued that if the Roberts et al. (1991) calibration of the Carboniferous-Permian boundary at ca 275 Ma is correct, then the GympieCamboon arc and associated intra-oceanic rifting, as

251

evidenced by the Early Permian Gympie Volcanics (Sivell & Waterhouse 1988), did not begin until at least 20 million years after the intrusion of the Hillgrove Supersuite. This interpretation depends critically on both the age of the Carboniferous-Permian boundary and the age of the Gympie Volcanics. In contrast to Roberts et al. (1991), the Australian Geological Survey Organisation places the Carboniferous-Permian boundary at ca 298 Ma (Jones 1996). Additionally, although the Lizzie Creek Volcanics are dated at 280270 Ma (Webb & McDougall 1968), part of the Camboon Andesite may be older (294 Ma and 281 Ma: Runnegar 1979). If this is the case, and the Camboon Andesite is truly a product of intra-arc rifting, then such an event cannot be discounted as a potential heat source for the formation of the S-type granites and consequently the Kaloe Granodiorite. On the basis of our current understanding of New England Orogen geology, it is difficult to evaluate the cause of the 'Kaloe' partial melting event. The restricted compositional range, combined with REE characteristics and isotopic variation in this intrusion argue against its formation simply by the fractional crystallisation of an arc-derived magma (Bryant et al. 1997). Most notably, both s N d and initial 87 Sr/ 86 Sr ratios decrease sharply with increasing Si0 2 , with the most felsic endmembers requiring a component with time-integrated low Sm/Nd and Rb/Sr, as observed in some lower crustal granulites. On the basis of the geochemical and isotopic characteristics Bryant et al. (1997) argued that these granites were possibly derived by partial melting of relatively young lower to middle crustal materials at temperature-pressure conditions very close to the amphibole-out curve, where some magmas cross over from corundum to diopside normative (925-1000°C: Beard & Lofgren 1991; Rushmer 1991; Wolf & Wyllie 1994). Although this intrusion is not interpreted to be a direct fractionate of an arc basalt, this does not exclude such a magma from being the source of heat required for crustal anatexis. However, intra-arc rifting provides an equally suitable source of heat. Atherton (1990, 1993) concluded that the predominantly tonalitic Coastal Batholith in Peru was derived from the rapid recycling of relatively young basaltic crust within a rifted continental margin. He reported that during this event, the Huarmey marginal basin locally attained geothermal gradients as high as 300°C/km. The answer to the backarc versus arc setting may lie in the gabbroic stocks that intrude the Kaloe Granodiorite. As yet these intrusions have not been dated or undergone detailed trace element or isotopic investigation. The date obtained for the Kaloe Granodiorite constrains the age of the Yugambal terrane to being greater than 293.1 ± 1.8 Ma. Such an age implies that the proposal by Korsch and Harrington (1981) and Harrington and Korsch (1985) that this unit represents a Late Carboniferous to Early Permian backarc basin sequence behind an Early Permian arc ('Gympie arc system') is unlikely. The age obtained for the Kaloe Granodiorite is more consistent with this terrane being of Siluro-Devonian age as proposed by Day et al. (1978), Leitch (1975) and Fergusson (1988).

252

C.J. B R Y A N T

ETAL.

CONCLUSIONS

greenstones and amphibolites at 1, 3, and 6.9 kb. Journal of Petrology 32, 365-401.

New dates obtained f r o m the Towgon Grange, Dumbudgery Creek, Jenny Lind and Koreelan Creek Granodiorites are 257-250 Ma, within error of the 255— 244 Ma range proposed by Shaw and Flood (1993) for these plutons. The age of 293.1 ± 1.8 Ma obtained for the Kaloe Granodiorite is considerably older than other Clarence River Supersuite intrusions. This is the oldest recorded age for an I-type intrusion in the southern New England Orogen. The unusual geochemical and isotopic characteristics are consistent with this pluton being derived from different source rocks, during a different melting episode. Whether this melting event occurred in response to backarc/intra-arc rifting, remnant arc magmatism or in response to megafolding is unclear. The presence of abundant approximately east-westtrending fractures that are absent from other Clarence River Supersuite intrusions is consistent with this intrusion having experienced a period of deformation that is not recorded in other intrusions. It is unclear if these structures were in response to the megafolding event, or to stresses imparted during the Hunter-Bowen Orogeny, assuming that these are two distinctly different tectonic events (Collins et al. 1993; Fergusson & Leitch 1993). This age implies that the Yugambal terrane is most probably of Siluro-Devonian age. ACKNOWLEDGMENTS

BLACK L. P. 1994. U-Pb zircon ion-microprobe ages from the

northern Drummond Basin, northeastern Queensland. Australian Geological Survey Organisation Record 1994/34. C. J. 1 9 9 2 . Clarence River Suite granitoids: petrography and geochemistry. BSc (Hons) thesis, University of New England, Armidale, (unpubl.).

BRYANT

BRYANT C. J., ARCULUS R. J. & CHAPPELL B. W . 1997. T h e

Clarence River Supersuite: a 250 Ma Cordilleran tonalitic I-type intrusion in Eastern Australia. Journal of Petrology 38, 9 7 5 - 1 0 0 1 . COLLINS W . J., OFFLER R . , FARRELL T. R. & LANDENBERGER

B. 1993. A revised Late Palaeozoic-Early Mesozoic tectonic history for the southern New England Fold Belt. In: Flood P. G. & Aitchison J. C. eds, New England Orogen, Eastern Australia, pp. 69-84. Department of Geology and Geophysics, University of New England, Armidale. COSCA M. A., SUTTER J. F. & ESSENE E. J. 1991. Cooling and

inferred uplift/erosion history of the Grenville Orogen, Ontario: constraints from 40 Ar/ 39 Ar thermochronology. Tectonics 10, 959-977. CROSS K . C . , FERGUSSON C . L . & F L O O D

P. G .

1987.

Contrasting structural styles in the Palaeozoic subduction complex of the southern New England Orogen, eastern Australia. In: Leitch E. C. & Scheibner E. eds. Terrane accretion and Orogenic Belts, pp. 8 3 - 9 2 . American Geophysical Union, Geodynamics Series 19. DALRYMPLE G . B., ALEXANDER E. C., LANPHERE M . A . &

We thank Dave Foster and Stirling Shaw for their helpful reviews, and Heather Davies for improving the presentation of this paper. Financial support for the geochronology was provided by the Swiss National Science Foundation. CJB was funded by a Postgraduate Research Award, with further funding provided by ARC. This is publication number 71 in the Key Centre for the Geochemical Evolution and Metallogeny of Continents. REFERENCES

1992.

Tectonic setting and palaeoenvironment of terranes in the southern New England Orogen, eastern Australia as constrained by radiolarian biostratigraphy. Palaeogeography, Palaeoclimatology, Palaeoecology 94, 3 1 - 5 4 . ATHERTON M. P. 1990. The Coastal Batholith of Peru: the

product of rapid recycling of 'new' crust formed within rifted continental margin. Geology Journal 25, 337-349. ATHERTON M. P. 1993. Granite magmatism. Journal of the Geological Society of London 150, 1009—1023.

G. & WILLIS I. L. 1987. The geology of the Grafton-Maclean 1 : 2 5 0 , 0 0 0 sheet areas. Geological Survey of New South Wales Report GS1989/117. BEARD J. S. & LOFGREN G. E. 1991. Dehydration melting and water-saturated melting of basaltic and andesitic BARNES R .

DAY R . W . , MURRAY C. G . & WHITAKER W . G . 1 9 7 8 . T h e

eastern part of the Tasman orogenic zone. Tectonophysics 48, 3 2 7 - 3 6 4 . DIRKS P. H. G . M . , LENNOX P. G . & SHAW S. E . 1 9 9 2 .

Tectonic implications of two Rb/Sr biotite dates for the Tia Granodiorite, southern New England Fold Belt, NSW, Australia. Australian Journal of Earth Sciences 39, 111-114. FERGUSSON C. L. 1988. Tectonostratigraphic terranes in the

J. C. & F L O O D P. G. 1 9 9 2 . Early Permian transform margin development of the southern New England Orogen, eastern Australia (eastern Gondwana). Tectonics 11, 1 3 8 5 - 1 3 9 1 .

AITCHISON

AITCHISON J. C . , FLOOD P. G . & SPILLER F. C. P.

KRAKER G. P. 1981. Irradiation of samples for 40 Ar/ 39 Ar dating using the Geological Survey TRIGA reactor. US Geological Survey Professional Paper 1176.

central Coffs Harbour Block, northeastern New South Wales. In: Kleeman J. D. ed. New England Orogen Tectonics and Metallogenesis, pp. 42-48. Department of Geology and Geophysics, University of New England, Armidale. FERGUSSON C. L. & LEITCH E. C. 1993. Late Carboniferous to

Early Triassic tectonics of the New England Fold Belt, eastern Australia. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 53-67. Department of Geology and Geophysics, University of New England, Armidale. FLOOD P. G. & AITCHISON J. C. 1988.

Tectonostratigraphic terranes of the southern part of the New England Orogen. In: Kleeman J. D. ed. New England Orogen Tectonics and Metallogenesis, pp. 7-10. Department of Geology and Geophysics, University of New England, Armidale.

T. J. & L E N N O X P. G. 1992. Low-pressure deformation of a small high-level pluton: the syn-tectonic Davys Creek Granite from the southeastern Lachlan Fold Belt. Tectonophysics 214, 293-309.

FOWLER

CLARENCE RIVER SUPERSUITE, NSW G U S T D . A . , STEPHENS C . J. & G R E N F E L L A . T .

1993.

Granitoids of the northern New England Orogen: their distribution in time and space and their tectonic implications. In: Flood P. G. & Aitchison J. C. eds, New England Orogen, eastern Australia, pp. 565-571. Department of Geology and Geophysics, University of New England, Armidale. HARRINGTON H. J. & KORSCH R. J. 1985. Tectonic model for

the Devonian to middle Permian of the southern New England Orogen. Australian Journal of Earth Sciences 32, 163—179.

HARRISON T. M. 1981. Diffusion of 40 Ar in hornblende. Contributions to Mineralogy and Petrology 78, 324—331. HENSEL H. D. 1982. The mineralogy, petrology and geochronology of granitoids and associated intrusives from the southern portion of the New England Batholith. PhD thesis, University of New England, Armidale, (unpubl.). HENSEL H. D . , MCCULLOCH M . T . & CHAPPELL B. W . 1985.

The New England Batholith: constraints on its derivation from Nd and Sr isotopic studies of granitoids and country rocks. Geochimica et Cosmochimica Acta 49, 369-384. HERBERT H. K. 1984. Granites of the Stanthorpe-DrakeEmmaville region, north-eastern New South Wales and south-eastern Queensland. In: Herbert H. K. & Rynn J. M. W. eds. Volcanics, granites, and mineralisation of the Stanthorpe—Emmaville—Drake region, pp. 55-65. Geological Society of Australia, Queensland Division, Brisbane, Field Conference Guide. JONES P. J. (Compiler) 1996. AGSO Phanerozoic 1995. Oxford University Press, Melbourne.

Timescale

KENT A. J. R. 1994. Geochronology and geochemistry of Palaeozoic intrusive rocks in the Rockvale region, southern New England Orogen, New South Wales. Australian Journal of Earth Sciences 41, 365-379. KIMBROUGH D . L., CROSS K . C & KORSCH R . J. 1993. U - P b

isotopic ages for zircons from the Pola Fogal and Nundle granite suites, southern New England Orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 403-412. Department of Geology and Geophysics, University of New England, Armidale. KORSCH R . J. & HARRINGTON H . J. 1981. S t r a t i g r a p h i c a n d

structural synthesis of the New England Orogen. Journal of the Geological Society of Australia 28, 205-226. LEITCH E. C. 1974. The geological development of the southern part of New England Fold Belt. Journal of the Geological Society of Australia 21, 133-56. LEITCH E. C. 1975. Plate tectonic interpretation of the Paleozoic history of the New England Fold Belt. Geological Society of America Bulletin 86, 141-144. MCKENZIE A. J. 1972. Potassium-argon dating of ten samples submitted by the Department of Mines, Sydney, Australia (Robertson Research International Ltd, Technical Report 723). Geological Survey of New South Wales Report GS1972/224 (unpubl.). MCPHIE J. 1987. Permo-Carboniferous silicic volcanism and palaeogeography on the western edge of the New England Orogen, north-eastern New South Wales. Journal of the Geological Society of Australia 31, 133-146.

PETFORD N. 1996. Dykes or diapirs? Transactions of the Royal Society of Edinburgh, Earth Sciences 87, 105-114. ROBERTS J., CLAOUE-LONG J.

C.

&

JONES P. J.

1991.

Calibration of the Carboniferous and Early Permian of the southern New England Orogen by SHRIMP ion microprobe zircon analysis. In: Advances in the Study of the Sydney Basin. Twenty fifth Newcastle Symposium, pp. 38-43. Department of Geology, University of Newcastle, Newcastle. ROBERTS J. & ENGEL B. A. 1987. Depositional and tectonic

history of the southern New England Orogen. Australian Journal of Earth Sciences 34, 1—20. RUNNEGAR B. 1979. Ecology of Eurydesma and the Eurydesma fauna, Permian of eastern Australia. Alcheringa 3, 261—285. RUSHMER T. 1991. Partial melting of two amphibolites: contrasting experimental results under fluid-absent conditions. Contributions to Mineralogy and Petrology 107,41-59. SAMSON S. D . & ALEXANDER E. C . 1 9 8 7 . C a l i b r a t i o n o f t h e

interlaboratory 4 0 Ar/ 3 9 Ar dating standard, MMhb-1. Chemical Geology 66, 27-34. SHAW

S. E . & F L O O D R . H .

1981. T h e N e w

England

Batholith, eastern Australia: geochemical variations in time and space. Journal Geophysical Research 86, 10530— 10544.

SHAW S. E. & FLOOD R. H. 1982. The Bundarra Plutonic

Suite: a summary of new data. In: Flood P. G. & Runnegar B. eds, New England Geology, pp. 183-192. Department of Geology and Geophysics, University of New England, Armidale. SHAW S. E. & FLOOD R. H. 1993. A compilation of Late

Permian and Triassic biotite Rb-Sr from the New England Batholith and areas to the southeast. Centre for Isotope Studies Research Report 1991-1992, 151-155. SIVELL W . J. & WATERHOUSE J. B. 1 9 8 8 . P e d o g e n e s i s o f t h e

Gympie Group volcanics: evidence of an Early Permian volcanic arc in eastern Australia. Lithos 21, 21—95. STEWART J. I. 1977. The economic geology of the Gordonbrook Serpentinite Belt, New South Wales. BSc (Hons) thesis, University of New England, Armidale, (unpubl.). THOMSON J. 1976. Geology of the Drake 1:100 000 sheet 9340. Geological Survey of New South Wales, Sydney. TURNER G. 1971. 40 Ar/ 39 Ar ages from the Lunar Maria. Earth and Planetary Science Letters 11, 159—191. WEBB A . W . & MCDOUGALL 1. 1 9 6 8 . T h e g e o c h r o n o l o g y o f

the igneous rocks of eastern Queensland. Journal of the Geological Society of Australia 15, 313-346. W O L F M . B. & WYLLIE P. J. 1994. Dehydration-melting of

amphibolite at lOkbar: the effects of temperature and time. Contributions to Mineralogy and Petrology 115, 369-383. YORK D. 1969. Least squares fitting of a straight line with correlated errors. Earth and Planetary Science Letters 5, 320-324.

MURRAY C . G . , FERGUSSON C . L., FLOOD P. G . , WHITAKER W .

G. & KORSCH R. J. 1987. Plate tectonic model for the

253

Carboniferous evolution of the New England Fold Belt. Australian Journal of Earth Sciences 34, 213-236.

(Received 24 April 1996; accepted 7 March 1997)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 254-265.

Timing of thermal and mineralisation events associated with the Mole Granite, New South Wales J. D. KLEEMAN, 1 I. R. PLIMER, 2 J. LU, 2 D. A. FOSTER 3 AND R. DAVIDSON 1 1

Department of Geology and Geophysics, University of New England\ Armidale, NSW 2351, Australia. 2 Victorian Institute of Earth and Planetary Sciences, School of Earth Sciences, University of Melbourne, Parkville, Vic. 3052, Australia. 3 Victorian Institute of Earth and Planetary Sciences, School of Earth Sciences, La Trobe University, Bundoora, Vic. 3083, Australia. The K-Ar, 40 Ar- 39 Ar and Rb-Sr radiogenic isotope data are used to calculate absolute ages for several stages in the sequence from emplacement of the Mole Granite at 246 ± 2 Ma to low temperature geothermal activity at 231 ± 2 . Samples selected for analysis are well related to detailed field mapping that outlined the major events. The granite is interpreted as a sill-like mass only a few kilometres thick, emplaced at a high level in the crust. Crystallisation inward produced a pressure-quenched, megacrystic granite carapace, pegmatite bodies were formed and microgranite was extensively emplaced as dykes in the fractured carapace and overlying country rock. Sheeted vein systems formed at 246 ± 1 Ma above ridges in the roof of the granite by hydrofracturing of contact rocks by a magmatic aqueous hydrothermal fluid. An aqueous magmatic fluid utilised the same structures as the microgranite and, at 246 Ma, reacted in places with the granite to form silexite, a quartz-topaz rock. The fractured roof zone of the cooling granite was the locus for structurally controlled meteoric water dominated fluid circulating nearsurface geothermal system for the next 3 to 4 million years and large quartz vein systems were precipitated. Two whole-rock isochrons and a cluster of mineral ages at 243 ± 2 Ma mark the closure of widespread circulation of fluids at about 300°C and its accompanying re-equilibration of isotopic systems. A geothermal system operated in some parts of the upper portion of the granite for at least a further 10 million years and possibly for longer. Key words: argon-argon dating, mineralisation, Mole Granite, potassium-argon dating, rubidiumstrontium dating, tin deposits.

INTRODUCTION

The leucocratic Mole Granite in northern New South Wales is an internationally studied system with a complex set of hydrothermal events (Weber 1974; Eadington 1983; Plimer & Kleeman 1985; Sun & Eadington 1987; Heinrich & Ryan 1992; Rankin et al. 1992; Heinrich et al 1992). Multiphase hydrothermal imprints render radiometric dating of the granite somewhat equivocal because biotite is variously recrystallised, chemically re-equilibrated, chloritised or interlaminated with fluorite and the feldspars are altered (Plimer & Kleeman 1986). K-feldspar is commonly partially altered to sericite and, in places, is altered to dickite or kaolinite. Fresh plagioclase is not present in the Mole Granite. On the basis of earlier radiometric dating of stratigraphically analogous granites, Leitch (1969) suggested that the Mole Granite was Late Permian to Early Triassic. A 235 ± 6 Ma K-Ar biotite age by Kleeman (1982) from The Gulf and the four Rb-Sr biotite ages from the Mole Granite of 237 Ma and 242 Ma by Shaw and Flood (1993) suggested an Early Triassic emplacement age for the Mole Granite. However, the preliminary 40 Ar- 39 Ar data of Plimer et al. (1995) showed that the initial and major hydrothermal event associated with the Mole Granite was at 246 Ma and

that there was thermal resetting of the argon isotopes at 242 Ma. In this paper we present new K-Ar, 40 Ar- 39 Ar and Rb-Sr data from the Mole Granite, from the altered phases of the Mole Granite, from hydrothermal veins associated with the Mole Granite and from the thermal metamorphic aureole. Our aim was to determine the age of emplacement of the Mole Granite, to quantify the timing of major hydrothermal events that produced the tin, tungsten, base-metal, topaz and beryl deposits and to understand the history of fluid-rock interaction. G E O L O G I C A L SETTING

The Mole Granite (centred on about 29°14'S, 151°30'E, see Figure 1) intrudes Permo-Carboniferous sediments and Permian felsic volcanics and has an outcrop area of 650 km2, although Kleeman (1982) argued that the subsurface and surface area is 1800 km2 because the roof dips shallowly outward at 7-15° and there is a thermal metamorphic aureole up to 10 km in width. On the basis of this larger area of the Mole Granite, heatflow measurement from a drillhole and heat production from U, Th and K, Kleeman (1982) calculated the thickness (of the order of 1 km) and volume of the Mole Granite (2000 km3). It was concluded that the Mole granite is a sill-like mass which had been emplaced at a

DATING MINERALISATION, MOLE GRANITE, NSW mid-Permian unconformity above which the rocks have a lower density than that estimated for the Mole Granite magma (Kleeman 1982). Subsequently, a diamond drillhole in metasediments at the Taronga Prospect to the southwest of the exposed contact intersected Mole Granite at depth and this is evidently a 060°-trending ridge of granite. A subsurface granite cupola occurs at the intersection of the 060°-trending structure and an Fj fold axis suggesting that the upper surface of the granite is undulating and that granite emplacement was strongly structurally controlled (Stegman 1982). The major hydrothermal systems are concentrated above the subsurface granite ridges or within the granite mass beneath the projection of the subsurface granite ridges (Plimer & Kleeman 1985). The northeast strike of joints, hydrothermal veins, pegmatite and microgranite suggest that the 060°-trending structures were influential in the granite emplacement, the orientation of major jointing, emplacement of microgranite soon after solidification of the granite carapace, and the positions of hydrothermal plumbing systems. At the southern and western contacts, the Mole Granite has intruded undifferentiated pelitic Palaeozoic metasediments whereas at the eastern contact the granite has intruded the Emmaville Volcanics and the Dundee Rhyodacite (Plimer & Kleeman 1985). The Permian Bondonga beds were intruded by the Mole Granite at the northern contact. The 249-246 Ma Rb-Sr date for the Dundee Rhyodacite by Shaw and Flood (1993) constrains an upper age limit of emplacement of the Mole Granite. The granite has a number of roof pendants, the largest of which is the metapelitic Torrington roof pendant. The altitude and the central location of the Torrington roof pendant in the Mole Granite suggests that the granite

Figure 1 Geological setting and sample locations of the Mole Granite, New South Wales (modified from Baillie 1983, Brodie 1983, Cozens 1984, Stegman 1983, and the 1:250 000 series geological maps of the Geological Survey of New South Wales). Symbols are: +, Mole Granite; x, adamellites and granodiorites; v, Emmaville Volcanics and other Permian felsic volcanics; T, Tertiary basalt; D, Dundee Rhyodacite; P, Tent Hill Porphyrite; the clear areas are Permian and Carboniferous sedimentary rocks. The letters in circles represent the locations of samples for which data are reported in Table 1.

255

roof collapsed. Smaller roof pendants of I-type granite occur near the eastern margin of the mass. Rare granodioritic enclaves are present near the Butler Mine. They are subspherical, contain K-feldspar and euhedral quartz megacrysts, biotite clots and rare biotite schlieren. Three variants of granite were described by Kleeman (1982). The upper surface of the granite is defined by a 20-50 m-thick megacrystic granite, previously called a porphyritic granite by Kleeman (1982). The megacrystic granite occurs at the granite contact and at the highest topographic points of the granite mass suggesting that the granite has been only partially unroofed. The megacrysts comprise K-feldspar to 3 cm, bipyramidal quartz and subsolidus biotite in a groundmass of quartz, feldspars and interstitial magmatic biotite (Plimer & Kleeman 1986). The megacrystic granite acted as a pressure-quench carapace, constrained fluid in subsurface granite cupolas and ridges and, in places, has been fractured and hydrothermally altered. The megacrystic granite grades into a coarse-grained seriate variety which constitutes the bulk of the granite mass. A number of minor pegmatite bodies are present in metapelites at the southern contact and in the Torrington roof pendant. The third major type of granite is microgranite, which commonly occurs as dykes in the megacrystic granite and in associated metapelites at the southern contact as well as in the Torrington roof pendant. In the upper segment of the seriate granite, 060°trending joint-controlled microgranite dykes occur. The microgranite is well exposed in the Fielders Hill and Wild Kate openpit mines in the Torrington roof pendant. The microgranite emplacement is influenced by bedding and jointing with dykes varying from millimetres to metres in width. Microgranite dykes

256

J. D. K L E E M A N

ETAL.

Table 1 Summary of samples analysed, methods utilised and dates in Ma.

Sample*

Number

AMG co-ords

Phase

Method1"

Age (Ma)

Emplacement of granite Dundee Rhyodacite (in aureole) [K] Dundee Rhyodacite (in aureole) [K]

R64659 R64660

747493 746504

Biotite Biotite

K/Ar K/Ar

245 + 2 244 + 2

R62448

672427

Whole rock

K/Ar

244 ± 2

R62491

589471

Muscovite

K/Ar

247 ± 2

Sheeted vein mineralisation Selvage to quartz-cassiterite-sulfide vein in Emmaville volcanics, Great Britain mine [I] Sediments adjacent to quartz-cassiteritesulfide vein, Taronga prospect [D] Sediments adjacent to quartz-cassiteritesulfide vein, Taronga prospect [D] Quartz-cassiterite-muscovite vein in metapelite, Taronga prospect [D]

Early veins Quartz-cassiterite-fluorite vein, Gulf fluorite mine [A] Quartz-cassiterite-fluorite vein, Stormers mine [C]

Pegmatite Quartz-feldspar-biotite-beryl pegmatite, Fielders Hill [F] Quartz-feldspar-kaolinite-beryl-biotitesulfide pegmatite, Emerald mine [H] Silexite Quartz-topaz-biotite selvage to silexite vein Fielders Hill mine [F] Quartz-topaz-biotite rock, Fielders Hill mine [F] Quartz-biotite vein, McCowans mica lodes [E] Quartz-biotite vein, McCowans mica lodes [E] Quartz-topaz-biotite vein, McCowans mica lodes [E] Quartz-topaz-biotite vein, McCowans mica lodes [E] Microgranite-silexite transition, Fielders Hill mine [F] Microgranite Incipiently altered microgranite Biotite-quartz-topaz greisen Quartz-biotite-topaz greisen

R62491

589471

Muscovite

Rb/Sr

246 + 2

MG1-1 MG1-2

589471 589471

Muscovite Muscovite

40

Ar/ 39 Ar P 40 Ar/ 3 9 Ar P

245 ± 1 246 ± 2

MG7 MG8-1 MG8-2

495610 588582 588582

Muscovite Muscovite Muscovite

Ar/ 3 9 Ar T F A r / 3 9 Ar P 40 Ar/ 3 9 Ar T F

245 + 1 250 ± 2 250 ± 2

R62440 MG17

671627 668505

Biotite Biotite

40

K/Ar Ar/ 39 Ar P

243 ± 2 242 ± 1

R62441

671627

MG4 671627 R62474 650542 R62474 650542 650542 MG2 650542 MG3 Following 4 samples R64694 671627 671627 R64698 671627 R64695 671627 R64696

Large fracture-controlled veins and associated hydrothermal alteration 684551 Green sericite greisen selvage to veins, Butler mine [G] MG11 R62455 684551 Seriate granite near Butler mine [G] 726559 R62483 Coarse granite, Torrington [J] Fluid circulation in granite Whole rock samples shown below Selvage to vein, Great Britain mine [I] Seriate granite, The Gulf [B] Megacrystic granite, Butler mine [G] Microgranite, Butler mine [G] Coarse granite, Torrington [J] Megacrystic granite, Taronga prospect [D] Microgranite, Fielders Hill [F]

Following 7 samples R62448 672427 R62464 495610 R62465 684551 R62466 684551 R62483 726559 R64666 684551 R64694 671627

Capital letters in square brackets refer to the locations shown circled in Figure 1. t

40Ar/ 39 Arp, plateau age; 4 0 Ar/ 3 9 Ar T F , total fusion age.

40 40

Biotite

K/Ar

241 ± 2

Biotite Biotite Biotite Biotite Biotite Isochron

40

Ar/ 3 9 Arp K/Ar Rb/Sr 40 Ar/ 3 9 Ar P 40 Ar/ 3 9 Ar P Rb/Sr

241 ± 3 243 ± 2 231 ± 2 243 1 2 243 + 2 242 ± 1

Sericite Biotite Biotite

40

Ar/ 3 9 Ar T F K/Ar Rb/Sr

243 ± 1 231 ± 2 232 ± 1

Isochron

Rb/Sr

243 ± 1

DATING MINERALISATION, MOLE GRANITE, NSW contain angular metapelite enclaves of variable size and joint-controlled blocks of the overlying metapelite many of which have shapes that can be fitted back to their place of origin. The leucocratic Mole Granite contains magmatic Kfeldspar, quartz, plagioclase and biotite and accessory ilmenite, monazite and topaz (Kleeman 1985). However, in many places the magmatic mineral assemblage has been overprinted by K-feldspar, albite, biotite, quartz, topaz, tourmaline, muscovite, dickite and kaolinite (Plimer & Kleeman 1985, 1986). The Mole Granite plots at the ternary minimum of the normative An-Or-Ab and Q-Ab-Or projections of the Q-Ab-OrAn—H20 system at 100 MPa PH2o- All three types of the granite are very similar chemically and are characterised by high Si0 2 , alkalis, Rb, LIL, HREE and Sn and low CaO, MgO, FeO, MnO, Fe 2 0 3 , Fe 2 0 3 /Fe0, Sr, F and Eu. Although Mole Granite rock samples have a low F content, the F/OH ratios of interstitial biotite suggests that the melt originally had elevated aF_ (Plimer & Kleeman 1986). However, the microgranite contains slightly elevated Si0 2 , K 2 0, F, B and Li and depleted FeO, Na 2 0, CaO and Eu compared with the seriate and megacrystic varieties suggesting that this later phase might have slightly fractionated. On the basis of the stratigraphy of the intruded rocks, the low P H 2 0 and high aF„, and the experimental studies by Burnham (1979), it was proposed by Kleeman (1982) that the Mole Granite intruded to a high level in the crust, probably no more than 2 km in depth, a conclusion supported by Eadington (1983) on the basis of fluidinclusion data. MULTIPHASE MAGMATIC AND HYDROTHERMAL ACTIVITY Mapping by Baillie (1983), Brodie (1983), Cozens (1984) and Stegman (1983) has detailed the distribution of the three variants of the Mole Granite, the granite contact, the structure, the pegmatite and microgranite dykes, and the hydrothermal alteration and associated mineral deposits. This evidence has provided a relative chronology of magmatic and hydrothermal events (Plimer & Kleeman 1985), and the present work has the benefit of a mapped sequence of sequentially overprinted hydrothermal events. Megacrystic and seriate granite contains poikilitic Kfeldspar with a K-feldspar overgrowth. In places, the Kfeldspar overgrowth is overgrown by albite. Interstitial magmatic biotite in megacrystic granite and biotite clots in the megacrystic granite, seriate granite, roof pendants, enclaves and proximal contact rocks are all the same composition (Plimer & Kleeman 1986). Field mapping, by Baillie (1983), Brodie (1983), Cozens (1984) and Stegman (1983), and the authors' observations of superposed magmatic and hydrothermal events (Plimer and Kleeman, 1985), indicate the following relative chronology of events: (i) emplacement of granite; (ii) sheeted-vein mineralisation; (iii) pegmatite formation; (iv) microgranite intrusion; (v) silexite formation; (vi) fracture-controlled vein mineralisation; and (vii) hydrothermal alteration of granite.

257

Sheeted veins After emplacement of the mass of the Mole Granite comprising megacrystic and seriate granite, sheetedvein systems comprise the earliest mineralisation event. The Vegetable Creek tin field at Emmaville has produced 25 0001 of cassiterite from alluvium underlying Eocene to Miocene basalt (David 1887). Most of the cassiterite-bearing alluvium was derived from erosion of sheeted-vein systems (e.g. Taronga Prospect, Halls Grampians, Great Britain Mine) hosted by metamorphic rocks on the southern contact of the Mole Granite. Sheeted-vein systems occur in the thermal metamorphic aureole distal from granite. Sheeted-vein systems are not present at the northern contact of the granite. The sheeted-vein systems are steeply dipping with a northeast strike. At the Taronga Prospect, the sheetedvein system occurs in metapelites 100-300 m above a northeast-striking granite ridge. The vein system has an inverted cone shape and occupies an elliptical outcrop. It occurs at the intersection of 060°-trending joints and a northwest-trending fold axis (Stegman 1983). The vein density increases towards the focus of the sheeted-vein system, where up to 150 veins per linear metre have been counted, and vein systems decrease in width, vein density and vein width towards the granite. Individual veins are 0.1—3 cm wide. At Taronga, the veins are hosted by predominantly by metapelites and rarely by metapsammites, calcareous metasediments and metadolerite. The petrological description by Lawrence (1960a) suggests low P - high T thermal metamorphism on the basis of calcareous assemblages comprising quartz-plagioclase-titanite + tremolite or quartz-plagioclase-titanite + andalusite overprinted by epidotechlorite + axinite + adularia + manganoan calcite. Quartz predominates in the veins with variable amounts of cassiterite, arsenopyrite, pyrite, galena, sphalerite, argentian tetrahedrite, silver sulfosalts, tourmaline, topaz, muscovite, calcite and stilbite. At Taronga the mineral assemblages in the veins are weakly zoned from Sn-As-Cu near the granite to Cu-As, As-Pb-Zn and Pb-Zn-Ag with increasing distance from the granite. Associated with the sheeted veins are areas of replacement of the host rocks. Rare metadolerites contain a cassiterite-amphibole-sulfide + tourmaline + fluorite assemblage and Lawrence (1952) reports cassiterite pseudomorphs after crinoid stems and the gastropod Ptycomphalina in the Permian pelitic metasediments. Quartz from sheeted vein systems at Taronga has fluid-inclusion homogenisation temperatures of 280420°C, fluid inclusion salinities of 20-40 equiv. wt% NaCl and 5 1 8 0 of+7.5 to +9.5%o (Plimer et al. 1995). Pegmatite Rare pegmatites are present in the thermal aureole and in the Torrington roof pendant. At the Emerald Mine, pegmatite transgresses a sheeted-vein system and at Fielders Hill, pegmatite is transgressed by microgranite. Pegmatites strike 060-070°, are 10-100 cm wide and have a great diversity of mineral assemblages. In most

258 J. D. K L E E M A N ETAL. pegmatites, the primary magmatic mineral assemblage comprises perthite - quartz-muscovite - albite - beryl which have been partially replaced by Li-biotite, topaz, tourmaline, fluorite, muscovite, arsenopyrite, cassiterite and kaolinite. At Bismuth in the Torrington roof pendant, the complex pegmatite comprises perthiteLi-biotite-quartz-bismuth-bismuthinite-Bi-sulfosaltswolframite - topaz - fluorite - base-metal sulfides tourmaline - Co Fe Ni-As Sb S minerals - uraninite - monazite (Lawrence & Markham 1963). Fluid inclusions in quartz and beryl show evidence of boiling, have homogenisation temperatures of 510— 580°C and a salinity of 50-60 equiv. wt% salts and 5 0 of +9 to + 10%o for quartz from the complex pegmatite at Bismuth, which are in accord with a magmatic origin (Sun & Eadington 1987; Plimer et al. 1995). 18

Microgranite Fresh to altered dykes of microgranite occur in proximity to the Torrington roof pendant and in proximity to the southern contact of the granite. Microgranite and its alteration products are rare within the mass of the granite and at the northern contact. Microgranite transgresses metasediments, megacrystic and seriate granite and pegmatite and, in places, is altered to silexite. Silexite The alteration of granite to silexite is common near the contact of the Mole Granite. Eadington and Nashar (1978) argued that silexite is of intrusive magmatic origin, citing the proximity of their studied occurrences to the pegmatite at Bismuth and the presence of strongly saline fluid inclusions with very high homogenisation temperatures. However, Kleeman (1985) demonstrated that silexite is the product of hydrothermal alteration of already solidified granite, most frequently microgranite but also of seriate and megacrystic granite, and more rarely, of pegmatite. Textures and structures within microgranite, seriate granite or porphyritic granite, or pegmatite, are pseudomorphed as silexite textures. Topaz containing very-high-temperature fluid inclusions reported by Eadington and Nashar (1978) is interpreted as relict accessory topaz from predecessor microgranite, while later evidence shows additional fluid inclusions with homogenisation temperatures more consistent with activity by hot highly saline Si- and F-rich aqueous fluids responsible for the hydrothermal alteration to silexite (see below). Silexite is a quartz—topaz rock with various amounts of minor wolframite, Li-siderophyllite, F-schorl, Fmuscovite, beryl, molybdenite, chalcopyrite, bismuth, bismuthinite and dickite (Plimer et al. 1995). Fluidinclusion studies by Eadington (1983) and Plimer et al. (1995) show that magmatic topaz in microgranite and some topaz grains in silexite have homogenisation temperatures of 570-620°C. Another generation of fluid inclusions in quartz, topaz and beryl show evidence of boiling and have homogenisation temperatures of 517580°C while a third generation of fluid inclusions have

homogenisation temperatures of 300-500°C (Plimer et al. 1995). The oxygen and hydrogen isotope data of both Sun and Eadington (1987) and Plimer et al. (1995) are consistent with a magmatic origin for the fluids which altered microgranite to silexite. Large veins Large multistage veins in the uppermost 100 m of the seriate granite trend 045-065° and are controlled by joints, faults, shears and the walls of microgranite dykes. The economically most significant underground hard-rock mines in the district occur on several large vein systems (e.g. Butler Mine, McKinnon's Mine, Dutchman Lode, Marvin Mine, Wallaroo Lode, Curnow's Mine, McGowans Mine). Fault- and shear zone-hosted deposits are multiply veined and brecciated. Joint-controlled deposits are zoned with an innermost zone comprising quartz crystals covered with palebrown sphalerite with exsolved chalcopyrite, chalcopyrite, galena-^hlorite intergrowths and monazite. In places, the innermost zone is transgressed by adularia veinlets, partially pseudomorphed by manganoan calcite or transgressed by quartz-brick-red adularia-hematite veinlets (e.g. Curnow Mine). The innermost zone abuts an outer zone comprising quartz crystals, chlorite books and tabular wolframite. In places, the quartz crystals are pseudomorphed by cassiterite (e.g. Dutchman Lode: Lawrence 1960b). A vein selvage of quartz, chlorite and cassiterite is in contact with granite pseudomorphed by green muscovite-quartz-chlorite. Intersection of 060°-trending large veins with 320°trending joints at Silent Grove has led to the formation of pipe-like quartz veins in variably altered granite. These veins ('bungs') contain quartz veins overgrown with cassiterite, vugs with unconsolidated cassiterite sand at the base and cylindrical cavities lined with quartz crystals. The granite associated with bungs exhibits 060° joint-controlled green muscovite-quartz alteration overprinted by brown resinous metamict monazite. Quartz from the large veins has fluid inclusions with a great range of homogenisation temperatures (220350°C), fluid inclusion salinity of 28-32 wt% salts and S 0 of -4.4 to -13.8%o interpreted as derived from a dominantly meteoric fluid (Eadington 1983; Sun & Eadington 1987; Plimer et al. 1995). Quartz from the transgressive quartz-adularia-hematite veins has a 6 0 of-4.0 to -6.0%o (Sun & Eadington 1987; Plimer et al. 1995). 18

18

ISOTOPE AGE DATING Samples were selected for analysis from sites that have good field evidence showing the relationship between stages of magmatic and hydrothermal activity. Radiogenic isotope studies used Rb-^Sr, Ar- Ar and K-Ar methods. Table 1 summarises the sample data and the location of samples is shown in Figure 1. Details of the analytical methods, tables of analytical data and sample locations by grid reference are given in Appendices 1-4. 40

39

DATING MINERALISATION, MOLE GRANITE, NSW DISCUSSION The following discussion follows the sequence of events that can be identified in the field, and are following the order described above. Emplacement of granite Direct dating of the intrusion age of the Mole Granite is not easy. With the Rb-Sr system, mineral ages and whole-rock isochrons returned ages of 243 Ma or less. Fortunately, the emplacement age of the Mole Granite can be bracketed tightly: the Mole Granite is younger than the Dundee Rhyodacite (249-246 Ma: Shaw & Flood 1993) and older than the first mineralisation produced by it, the sheeted veins which have a mean age 246 ± 1 Ma. The outer limits of the bracketed age of the Mole Granite are therefore 249-246 Ma. However, there is no evidence to suggest that there is a time gap of 3 million years between emplacement and the sheeted-vein stage of mineralisation. The upper limit of the bracketed age (249 Ma) is therefore a logical extreme, and the bulk of the evidence is consistent with and age of 246 ± 2 Ma. In an attempt to determine the Mole Granite age more directly, biotite was separated from the Dundee Rhyodacite within the metamorphic aureole of the Mole Granite on the expectation that the biotites may be reset by heating associated with the intrusion of the granite. The ages of 244 ± 2 and 245 ± 2 Ma as minimum ages are consistent with an intrusion age of 249-246 Ma. Sheeted-vein mineralisation The results confirm the field evidence that this is the earliest recognised episode of mineralisation associated with the Mole Granite. Table 1 lists six determinations in the range of 247 ± 2 to 244 ± 2 Ma using Rb-Sr, 40 Ar—39Ar and K-Ar methods. The mean of 40 Ar- 39 Ar and Rb-Sr (which have slightly higher closure temperatures than K-Ar) determinations is 246 ± 2 Ma which is indistinguishable from the age of the granite. The determinations on the muscovite from the Taronga prospect are the most interesting, with three different methods in close agreement in the range 247-245 Ma. The muscovite data yield apparent ages that indicate cooling below 350-400°C (McDougall & Harrison 1982; Lister & Baldwin in press). Early vein mineralisation The exact stratigraphic relationships of the two samples in this group are unclear. Field relationships suggest that both are related to early episodes of mineralisation, with the sample MG8 from Stormers Mine (250 ± 2 Ma) being equated by Stegman (1983) with events of similar age to the sheeted-vein mineralisation, and the Fluorite Mine (sample MG7, 246 ± 1 Ma) is interpreted as a basal portion of a vertical sequence of mineralisation produced by F-rich fluids, with affinities to the sheetedvein episode. Considering the error bars of these two apparent ages, they are consistent with both the age of

259

emplacement of the granite and the sheeted-vein mineralisation, and indicate that all three events overlapped within the accuracy of our measurements. Pegmatite Two samples of pegmatite yield biotite ages of 243242 Ma, recording cooling below about 300-330°C (Harrison et al. 1985). As described below, this age is interpreted as a later stage in the hydrothermal history than the events clustered about 246 ± 1 Ma. Field evidence shows that pegmatite formed immediately after the sheeted-vein mineralisation (Stegman 1983) and it is not likely that such a high temperature event would be delayed by 3-4 million years following sheeted-vein mineralisation, nor is it considered that there would be a discernible apparent age difference resulting form varying closure temperatures between K— Ar (300-330°C) and 40 Ar- 39 Ar (350-400°C). The Mole Granite is apparently a thin body, almost like a sill (Kleeman 1982) and would cool comparatively rapidly, especially with venting of fluids and the rapid introduction of meteoric water (Plimer et al. 1995). While its high U, Th and K concentrations would lead to high heat production, the dimensions of the granite would facilitate heat loss. Microgranite intrusion The age of this event has not been determined isotopically and it is doubtful that any age other than 243-242 Ma would be obtained. Whole-rock points of microgranite plot on the 242.8 ± 0.5 Ma isochron and therefore may be dominated by that hydrothermal episode. However, it is inferred that the microgranite event was very soon after the emplacement of the granite. It is difficult to find a model that can accommodate a delay of 4 million years between emplacement of the main mass of the Mole Granite and the intrusion of a small volume of what is apparently a late pulse of residual magma from the same body. Silexite formation Silexite is a quartz-topaz rock formed by alteration of granite (Kleeman 1985) involving F-rich fluids. The whole-rock isochron of microgranite, incipiently altered microgranite, biotite greisen and biotite-rich greisen was constructed (Figure 2) in an attempt to date the event of silexite formation. Some caution must be retained in interpreting this apparent age of 242.3 ± 0.5 Ma, because the spread in Rb-Sr is largely due to variable degrees of hydrothermal alteration. A mixing line produced during the alteration process could form the basis for an erroneous age. However, the agreement between this calculated age and the 40 Ar- 39 Ar and K-Ar results grouped under this heading in Table 1 gives, on balance, the benefit of any doubt to the 242.3 ±0.5 Ma Rb-Sr isochron. Sample R62474 is from McCowan's mica lodes and is interpreted as associated with the episode of silexite formation on the basis of field evidence (Stegman 1983). However the Rb-Sr biotite

260

J. D. K L E E M A N

ETAL. subsequent hydrothermal episode which is most likely related to the 517-580°C second generation of fluid inclusions noted above. Fracture-controlled vein mineralisation One sample of green sericite greisen selvage from the Butler Mine (MG11) yielded an 4 0 Ar- 3 9 Ar age of 243 ± 1 Ma. This formation of greisen is consistent with many other ages grouped around 243-242 Ma in diverse rock types and therefore indicative of a widespread and influential episode of hydrothermal reequilibration at that time. The large range of temperatures from fluid-inclusion studies and the superposition of mineralisation from medium temperature (300-350°C) to low temperature (150°C or less) implies a substantial history of hydrothermal activity later than 242 Ma, especially focusing on major fractures as conduits.

87

Rby®6Sr

Figure 2 Plot of whole rock 87 Sr/ 86 Sr vs 87 Rb/ 86 Sr for a profile from microgranite to silexite. The age calculated is 242.7 ± 0.3 Ma, with (87Sr/86Sr)i = 0.7075 ± 0.0012.

age on this sample (230.8 ± 1.7 Ma) is interpreted as a late re-equilibration of biotite by warm fluids capable of removing 87Sr from the mica, but which did not result in loss of radiogenic 40Ar. Silexite formation is most likely a moderately high temperature event soon after granite emplacement at 246 Ma. The ages grouped around 242 Ma are interpreted as indicating a widespread

87

Rb/ 86 Sr

Figure 3 Plot of whole rock 87 Sr/ 86 S vs 87 Rb/ 86 Sr for granite and wall-rock alteration. The age calculated is 242.8 ±0.5 Ma, while (87Sr/86Sr)i = 0.7067 + 0.0016. The fluorite point at (87Sr/86Sr) = 0.7065 on the Y-axis is not used to calculate the age or initial ratio, but is shown for comparison.

Hydrothermal alteration of granite The data presented in Table 1 under this event type show a dominant age group centred on 243-242 Ma and several individual mineral ages down to 231 Ma. The whole-rock isochron 242.8 ± 0.5 Ma (Figure 3) is clearly not the granite crystallisation age because this has been established above at 246 Ma. Again, technically some caution must apply to an isochron that derives from an alteration process, but there is a number of ages presented in Table 1 clustered about 243-242 Ma. It is suggested that this period represents a widespread hydrothermal event, homogenising whole-rock samples and resetting mica ages with respect to both Rb-Sr and K-Ar isotope systems (Table 1). Shaw and Flood (1993) also showed biotite ages of 237 Ma (both near the point S on Figure 1) and 242 Ma (both near the point J on Figure 1) for the Mole Granite. Ages of 237 Ma presented by these authors and the biotite ages of R62474 (230.8 ± 1 . 7 Ma by Rb-Sr), R62483 (232.2 ± 1.2 Ma by Rb-Sr) and R62455 (231 ± 2 Ma by K-Ar) as well as the 236 ± 5 Ma previously determined (Kleeman 1982) all illustrate a period of protracted isotopic re-equilibration that may well represent warm fluid-dominated phase alteration by chemical means rather than temperature-dominated closure of isotopic systems. Contrasting ages from the biotite of R62483 are interesting. The rock is a coarse seriate granite near Torrington, not especially close to any mapped fracture zone or mineralisation. Biotite separated from this rock gives 232.2 ± 1.2 Ma by Rb-Sr and 241 ± 2 Ma by K-Ar and the whole-rock point falls very well on the 242.8 ± 0.5 Ma whole-rock isochron. This implies isotopic development of the rock, but that localised alteration persisted until 232 Ma, perhaps involving isotopic exchange between biotite and other phases within the rock with fluid movement on a very limited scale. Mineralogical and stable isotopic evidence (Sun & Eadington 1987; Plimer et al. 1995) and the presence of low homogenisation temperature fluid inclusions are

DATING MINERALISATION, MOLE GRANITE, NSW all consistent with a substantial period of moderate- to low-temperature hydrothermal circulation after 242 Ma. CONCLUSIONS We have successfully calculated absolute ages for three stages in the isotopic evolution that has accompanied the physical and chemical evolution of a granite-related hydrothermal system. A set of age determinations at 246 ± 2 Ma apply to the emplacement of the Mole Granite as a high-level sill-like mass; and also apply directly to the sheeted-vein mineralisation that followed immediately after granite crystallisation. We infer that the next events — pegmatite emplacement, microgranite intrusion and silexite formation — all occurred within a time period that would be indistinguishable from 246 ± 2 Ma, but these events are all overprinted by widespread fluid activity that gives a cluster of ages at 243 ± 2 Ma. Several mineral ages relate the next stage of mineralisation, the large fracture-controlled vein systems, at 243 ± 2 Ma. While there is no defining evidence that formation of these vein deposits did not commence earlier, the combination of these mineral ages and evidence of medium-temperature fluids (300350°C, from fluid-inclusion studies) leads to the conclusion that most fluid movement and mineralisation activity of this stage occurred at 243 ± 2 Ma. Finally, biotite mineral ages at 232 ± 2 Ma provide evidence for prolonged medium- to low-temperature hydrothermal movement in large fracture systems at the close of the system. ACKNOWLEDGMENTS This project was financially supported by an Australian Research Council Grant to IRP and JDK. Some of the field mapping was undertaken by Ian Baillie, Ross Brodie, Greg Cozens and Craig Stegman as part of the requirements for a BSc (Hons) at the University of New England under the supervision of JDK and IRP. Some mineral separation and the K-Ar dating was funded by small grants to JDK by the University of New England and a small grant from the University of Newcastle funded some mineral separation and fluid-inclusion studies. The Rb-Sr isotopic determinations were undertaken at the Centre for Isotope Studies at North Ryde and the assistance and encouragement by Dave Whitford is gratefully acknowledged, as is the support for that facility by both the Commonwealth Scientific and Research Organisation and the Australian Research Council. Irradiations for the 40 Ar- 39 Ar work were performed at the IRR-1 reactor at the Soreg Nuclear Research Facility and staff are thanked for their assistance. The authors are grateful for the criticism of the manuscript by Stirling Shaw, Chris Heinrich and Robert Barnes.

261

BRODIE R. 1983. Geology and mineralization of the Mole River-Silent Grove area, near Tenterfield, northern N.S.W. BSc (Hons) thesis, University of New England, Armidale (unpubl.). BURNHAM C. W. 1979. Magmas and hydrothermal fluids. In: Barnes H. L. ed. Geochemistry of hydrothermal ore deposits, pp. 71-136. John Wiley and Sons, New York. COZENS G. J. 1984. The geology and mineralisation of the southwest Mole Granite region, northeastern New South Wales. BSc (Hons) thesis, University of New England, Armidale (unpubl). DAVID T. W. E. 1887. The Vegetable Creek tinfield, N.S.W. New South Wales Department of Mines Monograph 116. EADINGTON P. A. 1983. A fluid inclusion investigation of ore formation in a tin-mineralized granite, New England, New South Wales. Economic Geology 78, 1204-1221. EADINGTON P. A. & NASHAR B. 1978. Evidence for magmatic

origin of quartz-topaz rocks from the New England batholith, Australia. Contributions to Mineralogy and Petrology 67, 433-438. HARRISON T .

M.,

DUNCAN

I. & MCDOUGALL I.

1985.

Diffusion of 40 Ar in biotite: temperature, pressure and compositional effects. Geochimica et Cosmochimica Acta 49, 2 4 6 1 - 2 4 6 8 . HEIMANN A . , STEINITZ G . & ZAFRIR H . 1 9 9 2 . I r r a d i a t i o n o f

samples for 4 0 Ar/ 3 9 Ar dating using the Soreq Nuclear Research Center IRR-1 reactor, Israel. Nuclear Geophysics 6, 273—286. HEINRICH C. A. & RYAN C. G. 1992. Mineral paragenesis and

regional zonation of granite-related Sn-As-Cu—Pb-Zn deposits: a chemical model for the Mole Granite district, Australia based on PIXE fluid inclusion analysis. In: Kharaka Y. K. & Maest A. S. eds. Water Rock Interaction, pp. 1583-1588. Balkema, Rotterdam. HEINRICH C . A . , R Y A N C . G . , M E R N A G H T . P. & EADINGTON

P. J. 1992. Segregation of ore metals between magmatic brine and vapor: a fluid inclusion study using PIXE microanalysis. Economic Geology 87, 1566-1583. KLEEMAN J.D. 1982. The anatomy of a tin-mineralizing Atype granite. In: Flood P. G. & Runnegar B. eds. New England Geology, pp. 327-334. Department of Geology, University of New England, Armidale. KLEEMAN J. D. 1985. Origin of disseminated wolframitebearing quartz-topaz rock at Torrington, New South Wales, Australia. In: High Heat Flow Granites Production (HHP) granites, hydrothermal circulation and ore genesis, pp. 197-201. Institution of Mining and Metallurgy, London. LAWRENCE L. J. 1952. The replacement of crinoid stems and gastropods by cassiterite at Emmaville, N.S.W. Proceedings of the Royal Society of New South Wales 86, 119—122. LAWRENCE L. J. 1960a. Cassiterite mineralization at Hall's Grampians, Emmaville, New South Wales. Neues Jahrbuch fir Mineralogie Abhandlungen 94, 150-161. LAWRENCE L. J. 1960b. A cassiterite pseudomorph after quartz from Torrington, New South Wales. American Mineralogist 45, 715-717.

REFERENCES

LAWRENCE L . J. & MARKHAM N . L. 1 9 6 3 . T h e p e t r o l o g y a n d

BAILLIE I. D. 1983. The geology and mineralisation of the Emmaville district, N.S.W. BSc (Hons) thesis, University of New England, Armidale (unpubl.).

mineralogy of the pegmatite complex at Bismuth, Torrington, N.S.W. Journal of the Geological Society of Australia 10, 343-364. LEITCH E. C. 1969. Igneous activity and diastrophism in the

262

J. D . K L E E M A N

ETAL.

Permian of New South Wales. Geological Australia Special Publication 2 , 2 1 - 3 7 .

Society

of

& Zak K. eds. Mineral Deposits, pp. 497-500. Balkema, Rotterdam.

LISTER G. S. & BALDWIN S. L. M o d e l l i n g t h e e f f e c t o f

RANKIN A . H . , RAMSEY M . H . , COLES B . , V A N LANGEVELDE

arbitrary P - T - t histories on argon diffusion in minerals using the MacArgon program for the Apple Macintosh. Tectonophysics (in press).

F. & THOMAS C. R. 1992. The composition of hypersaline, iron-rich granitic fluids based on laser-ICP and Synchrotron-XRF microprobe analysis of individual fluid inclusions in topaz, Mole, Granite, eastern Australia. Geochimica et Cosmochimica Acta 56, 67-79.

M C D O U G A L L I . & HARRISON T . M . 1 9 8 8 . Geochronology

Thermochronology Press, New York.

by the 40Ar—39Ar

Method.

and Oxford

SHAW S. E. & FLOOD R. H. 1993. A c o m p i l a t i o n o f L a t e

MCDOUGALL I. & ROKSANDIC Z. 1974. Total fusion 40 Ar/ 39 Ar ages using HI FAR reactor. Journal of the Geological Society of Australia 2 1 , 8 1 - 8 9 . PLIMER

I. R . &

KLEEMAN

J.

D.

1985.

Mineralization

associated with the Mole Granite, Australia. In: High heat production granites (HHP), hydrothermal circulation and ore genesis, pp. 563-569. Institution of Mining and Metallurgy, London. PLIMER I. R. & KLEEMAN J. D. 1986. M a j o r - and m i n o r -

element chemistry of biotites in Mole Granite, New South Wales, Australia. Transactions of the Institution of Mining and Metallurgy 95, B1—5. PLIMER I. R. & KLEEMAN J. D. 1991. G e o l o g y , g e o c h e m i s t r y

and genesis of the Sn-W deposits associated with the Mole Granite, Australia. In: Pagel. M. & Leroy J. L. eds. Source, transport and deposition of metals, pp. 7 8 5 - 7 8 9 . Balkema, Rotterdam. PLIMER I. R., KLEEMAN J. D. & L u J. 1992. T r a c e and rare

earth elements in cassiterite sources of components for deposits of the Mole Granite, Australia. Mineralium Deposita 26, 267-274. PLIMER I. R . , L u J . , FOSTER D . & K L E E M A N J . D .

Permian and Triassic biotite Rb-Sr data from the New England Batholith and areas to the southeast. In: Carr P. F. ed. Centre for Isotope Studies Research Report 1992-1993, pp. 151-155. Centre for Isotope Studies, North Ryde. STEGMAN C. L. 1983. The Mole Granite and its Sn-W-basemetal mineralisation—a study of its southern-central margin. BSc (Hons) thesis, University of New England, Armidale (unpubl.). SUN S-S. & EADINGTON P. J. 1987. Oxygen isotope evidence for the mixing of magmatic and meteoric waters during tin mineralization in the Mole Granite, New South Wales, Australia. Economic Geology 82, 43-52. WEBER C. R. 1974. Woollomin-Texas Block: plutonic rocks and intruded sediments. In: Markham N. L. & Basden H. eds. The Mineral Deposits of New South Wales, pp. 3 5 1 391. Geological Survey of New South Wales, Sydney. W I L L I A M S I. S . , C O M P S T O N

W.,

CHAPPELL

B.

1995.

Ar- 3 9 Ar dating of multiphase mineralisation associated with the Mole Granite, Australia. In: Pasava J., Kribek B.

40

(Received 27 June 1996; accepted 12 August 1997)

Appendix 1 Potassium—argon analyses.

Sample

%Kt

40

Ar*(x 10"10mol/g)

40

Ar*/40Artotal

A ge#

R62440

8.023, 8.012

36.148

0.964

243 ± 2

R62441

8.044, 8.065

36.072

0.986

241 ± 2

R62448

3.620, 3.610

16.415

0.982

244 ± 2

0.977

231+2

R62455

5.715, 5.704

24.407

R62474

7.958, 7.963

35.900

0.973

243 + 2

R62483

7.720, 7.719

34.622

0.974

241 ± 2

R62491

8.631, 8.597

39.601

0.977

247 ± 2

R64659

6.966, 6.989

31.779

0.980

245 ± 2

R64660

6.291, 6.292

28.501

0.976

244 + 2

* The mean K value was used in the age calculation. Radiogenic 40 Ar. Age in Ma with error limits given for the analytical uncertainty at one standard deviation. Constants: 40 K = 0.01167 atom%; Xb = 4.962 x l O ' V ; A,e = 0.581xl0- 10 y _1 . Argon isotopic analyses were performed by the Australian Mineral Development Laboratories. #

W.

&

SHIRAHASE T. 1975. Rubidium-strontium age determinations on micas from a geologically controlled, composite batholith. Journal of the Geological Society of Australia 22, 497-505.

DATING MINERALISATION, MOLE GRANITE, NSW Appendix 2

40

Step

36Ar

263

Ar/39Ar results.

mol

40

Ar mol

%40Ar*

3.130E-14 1.105E-13 1.772E-13 6.956E-14

93.4 97.7 98.3 98.1

19.255 18.584 18.596 18.669

253.5 + 5.6 7.46 245.2 + 1.1 36.01 245.4 ± 1.6 82.03 246.2 + 1.2 100.00 TFA - 246.1 ± 1.7 PA(2-4) = 245 + 1

2.590E-16 6.301E-15 1.74 IE-15 1.091E-15 1.588E-15

5.589E-15 1.196E-13 3.306E-14 2.108E-14 3.028E-14

91.5 98.0 97.7 98.6 97.4

19.766 18.616 18.551 19.045 18.587

259.7 ± 3.4 2.36 59.74 245.6 ± 0.9 244.8 + 2.4 75.60 85.54 250.9 ± 2.4 245.2 ±2.4 100.00 TFA = 246.3 + 1.5 PA(l-2) = 246 + 2

3.310E-15 2.300E-15 1.165E-15 9.187E-16

6.255E-14 4.315E-14 2.191E-14 1.732E-14

97.6 97.8 94.3 95.4

18.438 18.354 17.744 17.982

243.4 ±2.0 43.03 72.92 242.4 ± 1.6 234.8 ±1.6 88.06 237.8 ±2.9 100.00 TFA =: 241.1 ±2.0 PA(l-2)i = 243 ± 2

1.760E-15 7.004E-15 7.679E-15 2.329E-15 1.845E-15

8.568E-14 1.821E-13 1.693E-13 4.685E-14 3.659E-14

39.1 71.0 83.8 91.2 92.2

19.042 18.454 18.470 18.342 18.296

250.8 ±3.4 8.54 243.6 ±2.2 42.51 243.8 ±3.5 79.76 242.2 ±1.4 91.05 241.6 ±0.8 100.00 TFA = 244.0 ± 2.7 PA(2-5) = 243 ± 2

2.683E-14 2.457E-14 1.121E-14 6.881E-15 3.659E-15

97.4 97.4 97.7 98.3 82.0

18.716 18.262 18.318 18.705 15.907

246.8 ±2.0 36.22 241.2 ±3.3 70.22 241.9 ± 1.9 85.73 246.7 ± 2.5 95.11 211.9 ±8.9 100.00 TFA = 242.5 ± 2.7 PA(2-3) = 241 ± 3

1.554E-13 3.224E-13 2.293E-13 9.817E-14

85.2 98.1 97.6 99.9

18.791 18.629 18.414 18.881

247.7 ±2.1 17.03 245.8 ± 0.7 58.06 87.44 243.1 ±1.1 248.9 ± 1.0 100.00 TFA = 245.7 ± 1.1

6.169E-14 1.502E-13 4.880E-14 3.413E-14 1.45 IE-14

91.6 99.0 99.4 97.2 96.8

18.638 18.957 18.955 18.444 18.671

245.9 ± 1.4 18.97 249.8 ±1.7 68.03 249.8 ±2.1 84.04 243.5 ± 1.7 95.30 100.00 246.3 ± 2.2 TFA = 248.1 ±1.7 PA(2-3) = 250 ± 2

39

37

Ar mol

Ar mol

40

39

Ar*/ Ar(k)

Cumulative 39 Ar (%)

Age Ma± l a

MG1-1, muscovite; J = 7.8338E-3, 1 grain 1

2 3 4

7.066E-18 8.552E-18 1.028E-17 4.383E-18

4.583E-16 9.525E-16 1.441E-15 9.057E-16

1.518E-15 5.810E-15 9.364E-15 3.658E-15

MG1-2, muscovite; J = 7.8338E-3, 2 grains 1

2 3 4 5

1.699E-18 7.915E-18 2.502E-18 1.011E-18 2.657E-18

4.300E-16 1.206E-15 1.137E-16 1.813E-16 5.290E-16

MG2, biotite; J = 7.8338E-3, 1 grain 1 2 3 4

5.164E-18 3.126E-18 4.210E-18 2.710E-18

7.157E-16 2.417E-16 2.360E-16 2.294E-16

MG3, biotite; J = 7.8338E-3, 1 grain 1 2 3 4 5

1.765E-16 1.788E-16 9.288E-17 1.398E-17 9.574E-18

4.068E-16 9.011E-16 9.349E-16 4.192E-16 2.530E-16

MG4, biotite; J = 7.8338E-3, 2 grains 1 2 3 4 5

2.420E-18 2.086E-18 8.453E-19 4.477E-19 2.308E-18

5.559E-16 3.779E-19 3.779E-19 2.522E-16 3.566E-16

1.397E-15 1.310E-15 5.978E-16 3.617E-16 1.888E-16

MG7, muscovite; J = 7.8338E-3, 1 grain 1 2 3 4

7.785E-17 2.028E-17 1.847E-17 4.240E-20

1.120E-15 1.199E-15 2.377E-15 2.271E-18

7.046E-15 1.698E-14 1.216E-14 5.194E-15

MG8-1, muscovite; J = 7.8338E-3, 1 grain 1 2 3 4 5

1.75 IE-17 4.820E-18 8.848E-19 3.172E-18 1.623E-18

8.885E-16 3.414E-16 2.649E-16 3.019E-16 3.192E-16

3.034E-15 7.844E-15 2.560E-15 1.800E-15 7.524E-16

264

J. D. KLEEMAN

Appendix 2 continued

36

40

ETAL.

Ar/39Ar results

37

39

40

Ar mol

% 40 Ar*

MG8-2, muscovite; J = 7.8338E-3, 2 grains 1 2.073E-17 4.216E-16 3.256E-15 2 4.783E-18 2.008E-16 2.497E-15 3 5.910E-18 3.633E-19 5.953E-16 4 1.438E-18 1.765E-17 6.893E-16

6.850E-14 4.880E-14 1.269E-14 1.345E-14

91.0 97.1 86.2 96.8

19.149 18.970 18.369 18.884

46.27 252.2 + 2.5 81.74 250.0 ± 1.5 90.20 242.5 ± 1.8 100.00 248.9 ± 1.6 TFA =: 250.0 ± 2.0

MG11, muscovite; J == 7.8338E-3, 3 grains 1 5.833E-18 7.568E-17 U00E-15 2 3.777E-18 6.445E-16 5.927E-15 3 7.489E-18 3.472E-16 6.392E-15 4 1.694E-18 3.395E-16 2.059E-15 5 1.775E-18 5.030E-16 1.110E-15

2.279E-14 1.112E-13 1.187E-13 3.845E-14 2.089E-14

92.4 99.0 98.1 98.7 97.6

19.139 18.565 18.214 18.427 18.367

6.63 252.312.7 42.36 245.0+ 1.1 80.90 240.6 + 0.6 93.31 243.3 + 2.0 100.00 242.5 ± 2.4 TFA = 243.3 ± 1.2 PA(2-5) = 243 ± 1

MG17, biotite; J = 7.8338E-3, 1 grain 1 6.743E-18 1.986E-17 1.994E-16 2 9.297E-18 2.672E-16 1.493E-15 3 1.767E-17 9.504E-16 5.217E-15 4 6.614E-18 5.067E-16 4.384E-15 5 2.356E-19 3.779E-19 2.936E-15 6 3.997E-18 3.643E-15 1.142E-16 7 3.639E-18 4.516E-16 1.141 E-15

5.975E-15 3.013E-14 1.007E-13 8.208E-14 5.497E-14 6.826E-14 2.161E-14

66.6 90.9 94.8 97.6 99.8 98.2 95.1

19.963 18.334 18.311 18.270 18.682 18.400 18.008

1.05 262.1 +46.1 8.90 242.1 ±4.1 36.34 241.8+1.0 59.39 241.3+1.7 74.84 246.4 ±2.1 94.00 242.9 ± 2.6 100.00 238.1 ±2.2 TFA = 242.6 ± 2.4 PA(2-7) = 242 ± 2

Step

Ar mol

Ar mol

Ar mol

40 39

Ar*/ Ar(k)

Cumulative 39 Ar (%)

Age Ma ± l a

Notes 1 40Ar* includes line blank. TFA, total fusion age; PA, plateau age (steps used for calculation). 2 40Ar/39Ar analyses were performed at La Trobe University following methods of D. A. Foster & C. M. Fanning (pers. comm. 1996) Mineral separates and the flux monitor GA1550 biotite (97.9 Ma: McDougall & Roksandic 1974) were irradiated in a core position at the IRR-1 reactor, Soreq Nuclear Center, Israel (Heimann et al 1992) for 30 hours with 0.1 mm of Cd shielding. After irradiation 1-3 grains were loaded into pits in a Cu disk and placed in a stainless steel vacuum chamber with a glass view port. Gas was extracted from these samples using a computer-controlled 6 W argon ion laser with stepwise heating accomplished by varying the power output of the laser and using defocused beam to ensure roughly uniform heating across the grains. Gas was expanded into a stainless steel clean-up line and purified with two 10 1/s Zr-Ti getters. Argon isotopes were measured using a VG3600 mass spectrometer with a Daly photomultiplier operating at a sensitivity of 1 x 10"3 amps/torr. The data were corrected for machine background determined by measuring system blanks, and mass discrimination determined by analysing atmospheric argon. Correction factors for interfering isotopes were determined by analysing K 2 S0 4 and CaF2 salts irradiated with the samples and the following values were used: 40Ar/39Ark = 1.60E-2, 39 Ar/ 37 Ar=940E-4, 36Ar/37Ar=2.38E-4. Plateau ages calculated by the size-of-step method.

DATING MINERALISATION, MOLE GRANITE, NSW

265

Appendix 3 Strontium and rubidium isotope analyses.

Phase analysed

Sample no

Samples used in Figure 2

87Sr/86Sr

87

Rb/ 86 Sr

0.0 0.0

Fluorite Fluorite Whole rock Whole rock Whole rock Whole rock Whole rock Whole rock Whole rock Whole rock Whole rock

R64666 R64647 R64666 R62465 R62464 R62465 R64694 R62466 R62466 R62448 R62483

0.7065 0.7065 0.8698 0.9143 0.9177 0.9228 0.9936 1.7479 1.7981 1.7999 1.8234

47.3 60.0 61.4 62.7 82.6 300.9 315.4 316.5 323.5

Samples used in Figure 3 Whole rock Whole rock Whole rock Whole rock

R64696 R64698 R64694 R64695

2.46 1.9685 0.9936 4.9562

508.8 365.1 82.6 1230.1

Data for mineral ages in Table 1 Biotite Muscovite Biotite

R62483 R62491 R62474

60.9935 1.41614 46.7962

18257.0 203.0 14039.0

Analyses were undertaken at the Centre for Isotope Studies, CSIRO, North Ryde. Methods followed the procedures described by Williams et al. 1975, and used a VG54E mass spectrometer. 87 Sr/ 86 Sr is normalised to 86 Sr/ 88 Sr = 0.1194; X87Rb = 1.42 x lO^y" 1 ; NBS 987 = 0.710235 ± 0.000046. Mineral ages assumed that the initial 87 Sr/ 86 Sr = 0.705. The two fluorite values were not used in the regressions to obtain ages or initial 87 Sr/ 86 Sr and are shown for comparison.

Appendix 4 Sample locations. R62440 R62441 R62448 R62455 R62464 R62465 R62466 R62474 R62483 R62491 R64647 R64659 R64660 R64666 R64694 R64695 R64696 R64698 MG1.1 MG2 MG3 MG4 MG7 MG8 MG11 MG17

Quartz-feldspar-biotite-beryl pegmatite Quartz-topaz-biotite selvage to silexite Silicified selvage to veins Seriate granite Seriate granite Megacrystic granite Microgranite Quartz-biotite vein Coarse granite Muscovite bearing pelite Quartz-cassiterite-fluorite vein Dundee Rhyodacite Dundee Rhyodacite Seriate granite in borehole Microgranite Biotite-quartz-topaz greisen Quartz-biotite topaz greisen Incipiently altered micogranite Quartz-cassiterite-sulfide-muscovite vein in Quartz-topaz-biotite rock Quartz-topaz-biotite rock Quartz-topaz-biotite rock Quartz-cassiterite-fluorite vein Quartz-muscovite-tourmaline vein Green sericite greisen Quartz-feldspar-kaolinite-beryl

Fielders Hill Mine Fielders Hill Mine Great Britain Mine Butler Mine The Gulf Butler mine Butler mine McCowans Mica Lodes Torrington Taronga prospect Gulf Fluorite Mine In aureole In aureole Taronga prospect Fielders Hill Mine Fielders Hill Mine Fielders Hill Mine Fielders Hill Mine Taronga Prospect McCowans Mica Lodes McCowans Mica Lodes Fielders Hill Mine Gulf Fluorite Mine Stormers Mine Butler Mine Emerald Mine

The AMG grid coordinates refer to the Ashford (9139) and Clive (9239) 1:100 000 map sheets.

LN671627 LN671627 LN672427 LN684551 LN495610 LN684551 LN684551 LN650542 LN726559 LN589471 LN495610 LN747493 LN746504 LN589471 LN671627 LN671627 LN671627 LN671627 LN589471 LN650542 LN650542 LN671627 LN495610 LN588582 LN684551 LN668505

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 266-271.

Dumboy-Gragin Granite, northeastern New South Wales: age and compositional affinities N. M. VICKERY1*, P. M. ASHLEY1 AND C. M. FANNING2 1 2

Department of Geology & Geophysics, University of New England, Armidale, NSW 2351, Australia. Research School of Earth Sciences, Australian National University, Canberra, ACT 0200, Australia.

The Dumboy-Gragin Granite is classified as a member of the leucogranite suite of the New England Batholith, based on a strong radiometric signature, age and composition. It was previously grouped with the PermoCarboniferous, S-type Bundarra Plutonic Suite, but clearly differs from the latter by its Early Triassic U-Pb zircon age of 243.2 ± 3.7 Ma. The relatively high Si0 2 , K 2 0, Rb, Nb, Th, U, Y, Ce and Pb, and low Ti0 2 , total Fe, MgO, CaO, P 2 0 5 , Ba, Sr and Cu demonstrate the strong compositional affinities between the Dumboy-Gragin Granite and fractionated I-type leucogranites with associated Sn, W, base metal, Ag and U mineralisation elsewhere in the southern part of the New England Orogen. Key words: Dumboy-Gragin Granite, New England Orogen, zircon dating.

INTRODUCTION The southern part of the New England Orogen in northern New South Wales and southern Queensland is characterised by large volumes of granitic rocks of the New England Batholith. These Late Palaeozoic and Early Mesozoic plutons are grouped into several suites, based on their compositional and temporal relations (Shaw & Flood 1981, 1993; Hensel et al. 1985; Chappell 1994). A distinctive suite of leucogranites, emplaced late in the tectonic history of the region, has been long recognised, in part because of a spatial and genetic relationship to Sn, Mo, W, Bi, Ag, Au, U and base-metal mineralisation (Weber 1974; Plimer & Kleeman 1985; Barnes et al 1988; Blevin & Chappell 1993; Brown & Stroud 1993). During remapping and data compilation for the Inverell 1:250 000 metallogenic sheet, Stroud (1992) defined several new or previously poorly described granitic plutons, one being the Dumboy-Gragin Granite, cropping out in the DelungraWarialda area, west of Inverell (Figure 1). This pluton, originally informally named by Carne (1911), was subsequently included as part of the extensive PermoCarboniferous S-type Bundarra Plutonic Suite and not distinguished from the latter on published geological maps by the New South Wales Geological Survey or in regional studies of New England granitoids (Shaw & Flood 1981). On the basis of petrographic and chemical data, and strong radiometric signature, the DumboyGragin Granite was shown to be distinct from the Bundarra Plutonic Suite by Stroud (1992), who proposed it was more likely to be a member of the leucogranite suite of the New England Batholith. Rb-Sr biotite ages of Shaw and Flood (1993) indicated that the leucogranite suite is of Early Triassic age. In this paper, we present new radiometric age and chemical data on the Dumboy-Gragin Granite that confirm the proposals of Stroud (1992). The data

indicate an Early Triassic age for the pluton and strong petrologic and metallogenic links to certain leucogranites with fractionated I-type affinities and Sn, W, base metal, Ag and U mineralisation in the southern New England Orogen, for example the Mole, Gilgai, Elsmore and Webbs Consols Granites (Plimer & Kleeman 1985; Stroud 1992; Brown & Stroud; 1993, Blevin & Chappell 1993). FIELD AND PETROGRAPHIC CHARACTERISTICS The Dumboy-Gragin Granite is the most westerly pluton in the southern New England Orogen. It occurs west of the Bundarra Plutonic Suite and east of the Peel Fault, with the nearest outcrops of leucogranite (Gilgai Granite) being about 30 km to the southeast (Figure 1). On the southwestern margin of the Dumboy-Gragin Granite, there are intrusive contacts with metasedimentary rocks of the Anaiwan terrane of Flood and Aitchison (1988), but most of the pluton occurs as a series of inliers, surrounded and disconformably overlain by Mesozoic sedimentary rocks of the SydneyBowen and Surat Basins, Tertiary basalt and regolith (Figure 1). Typically, the pluton is composed of pale pink to buff, medium- to coarse-grained biotite leucogranite, locally strongly porphyritic with Kfeldspar megacrysts up to 4 cm across. There are a few microgranite dykes and one occurrence of slightly more melanocratic, hornblende-bearing granite immediately west of Delungra. Leucogranite contains scattered miarolitic cavities, displays local granophyric and micrographic textures, and has modal compositions according with alkali feldspar granite and syenogranite (Vickery 1993). In addition to quartz, perthitic * Present address: Plutonic Operations Limited, PO Box 7467, Cloisters Square, Perth, WA 6850, Australia.

DUMBOY-GRAGIN GRANITE

267

Figure 1 Location and regional setting of the Dumboy-Gragin Granite, with distribution of Bundarra Plutonic Suite granites and Early Triassic leucogranites. Zr, location of sample NEB307.

orthoclase and sodic plagioclase, it contains minor annitic biotite (Mg/Mg + I F e = 0.13-0.25) and accessory tourmaline (near schorl end-member), topaz, allanite, fluorite, muscovite, zircon, titanite, apatite, rutile, magnetite and ilmenite (Vickery 1993). Hydrothermal alteration effects are pervasive, with weak sericitisation of feldspars and chloritisation of biotite; locally, however, there is strong quartztourmaline (-cassiterite) and quartz-topaz (-arsenopyrite) greisen veining. Several mineral occurrences associated with greisen have been exploited for Sn, As and Cu.

thirteen individual zircon grains were analysed (Table 1; Figure 2) and the data reduced in a manner similar to that described by Compston et al (1992) and Williams and Claesson (1987), with augmented

GEOCHRONOLOGY Due to uncertainty over its age, a sample (NEB307) of Dumboy-Gragin Granite from GR 6727500N, 272000E was obtained for U - P b zircon dating. The zircons separated from sample NEB307 are euhedral grains with pyramidal terminations and zoned from centre to rim. These are interpreted to have been formed during a single magmatic crystallisation event. Some grains appear to have structurally distinct centres, possibly indicating the presence of an inherited component overgrown by a magmatic zircon rim. The sectioned grains have been analysed using SHRIMP II ion microprobe at the Research School of Earth Sciences, Australian National University. Sixteen areas on

Figure 2 Tera-Wasserburg plot of SHRIMP ion microprobe U-Pb data for all zircon areas analysed from sample NEB307 of the Dumboy-Gragin Granite. Data are plotted, uncorrected for common Pb, with 1 a error bars. All analyses are within analytical uncertainty of the proposed mixing line between a radiogenic end member at ca 243 Ma and common Pb. The weighted mean of the radiogenic 206 Pb/ 238 U ratios gives an age of 243.2 ± 3.7 Ma.

268 N. M. V I C K E R Y ET AL. Table 1 Summary of SHRIMP U-Pb results for zircons from Dumboy-Gragin Granite sample NEB307. Grain spot

U (ppm)

Th (ppm)

Th/U

Pb* (ppm)

204 /206p

1.1 1.2 2.1 2.2 3.1 4.1 5.1 6.1 7.1 8.1 9.1 10.1 11.1 12.1 12.2 13.1

340 405 764 4071 471 1098 762 543 483 596 437 340 436 663 667 552

138 200 366 1283 168 467 369 248 218 271 186 137 163 364 235 276

0.41 0.49 0.48 0.32 0.36 0.43 0.48 0.46 0.45 0.46 0.43 0.40 0.37 0.55 0.35 0.50

13 15 31 166 18 43 29 22 19 25 17 13 17 27 27 22

0.000359 0.000205 0.001776 0.001736 0.000191 0.001280 0.000425 0.000098 0.000107 0.000051

p b

t t

0.000183 0.000952 0.000058 0.000568

b

206P5/238U

Age (Ma)

0.0379 ± 0.0011 0.0366 ±0.0011 0.0387 + 0.0011 0.0410 ±0.0012 0.0382 ±0.0011 0.0381 ±0.0011 0.0372 ±0.0010 0.0393 ±0.0011 0.0381 ±0.0011 0.0399 ±0.0012 0.0374 ±0.0011 0.0378 ±0.0011 0.0389 ±0.0011 0.0392 ±0.0011 0.0399 ±0.0014 0.0386 ±0.0011

239.6 ±6.7 231.8 ±6.5 244.9 ± 6.6 259.1 ±7.6 241.5 ±6.8 240.7 ± 6.5 235.6 ± 6.4 248.5 ± 6.9 241.1 ±6.7 252.4 ±7.1 236.6 ±6.8 239.0 ±6.7 245.8 ± 6.9 248.0 ±6.8 252.5 ± 8.7 244.4 ± 6.7

t No Pb detected. Uncertainties given at the 1 a level. Correction for common Pb made on the basis of extrapolation to concordia along a mixing line with common Pb, following Tera & Wasserburg (1972), as outlined in Compston etal. (1992). 204

uncertainties using software of T. R. Ireland (Australian National University). On a Tera and Wasserburg (1972) concordia plot of measured 38{j/ Pb Pb/ Pb (Figure 2), all analyses are within analytical uncertainty of a mixing line between a 240-250 Ma radiogenic end-member and common Pb. This includes analyses of what were considered to be structurally distinct, inherited centres to three grains. Many analyses cluster on or near the concordia curve and define the 38u/ Pb j f diogenic endmember (Figure 2). Correction for common Pb is therefore made using this Tera-Wasserburg mixing line following the procedure outlined by Compston et al. (1992). A weighted mean of the radiogenic corrected P b / U ratios for all 16 analyses has no excess scatter giving an age of 243.2 ± 3.7 Ma (95% confidence limit) and we interpret this to be the crystallisation age for the magmatic zircon from sample NEB307. 2

207

206

v e r s u s

206

2

206

206

r a t

0 0

ra

238

CHEMISTRY Data from 28 relatively unaltered samples of DumboyGragin Granite were reported by Stroud (1992) and Vickery (1993) and an average and range of compositions are shown in Table 2. Results indicate that the pluton is relatively homogeneous, with a limited Si0 range of 74.3-78.5%, high alkalies and low Ti0 , total Fe, MgO, CaO, P 0 , Ba and Sr. However, certain high-field-strength elements such as Th, U, Y, LREE and Sn are relatively enriched. Most samples are mildly peraluminous (average 0.98 wt%, but up to 1.83 wt% normative corundum), although a few are metaluminous. On the normative Q-Or-Ab plot at 200 MPa P 2

2

2

5

H20

(Holtz et al. 1992), samples of the Dumboy-Gragin Granite fall close to the ternary minimum. DISCUSSION AND CONCLUSIONS On the basis of a distinctive, strong radiometric signature and chemical data, Stroud (1992) proposed that the Dumboy-Gragin Granite was a discrete pluton and not part of the Bundarra Plutonic Suite. Field and mineralogical characteristics of the former differ from typical Bundarra Plutonic Suite granitoids by being miarolitic, containing less biotite and no cordierite or garnet, but locally containing hornblende and accessory magnetite, titanite and allanite. On the other hand, late magmatic/hydro thermal tourmaline is common, a typical characteristic of the Bundarra Plutonic Suite. Texturally and mineralogically, the Dumboy-Gragin Granite displays strong affinities with Early Triassic leucogranites elsewhere in the New England Orogen (Plimer & Kleeman 1985; Blevin & Chappell 1993; Brown & Stroud 1993), except for the common occurrence of tourmaline. The presence of greisen with Sn, As and Cu in the Dumboy-Gragin Granite is not diagnostic, as mineralisation of this type occurs in both the leucogranites and Bundarra Plutonic Suite (Barnes et al. 1988). Mineralogical criteria mentioned above imply that the Dumboy-Gragin Granite has I-type characteristics, but the occurrence of ilmenite and late magmatic/hydrothermal Al-rich minerals (tourmaline, topaz, muscovite) could indicate S-type affinities. However, the presence of ilmenite and annitic biotite is a measure of oxidation state and it is therefore interpreted that the Dumboy-Gragin Granite magma

DUMBOY-GRAGIN GRANITE

269

Table 2 Average chemical compositions of least-altered Dumboy-Gragin Granite, Gilgai Granite, Mole Granite and Bundarra Plutonic Suite.

Dumboy-Gragin* (average) (range)

Si0 2 Ti0 2 A1203 Fe 2 0 3 FeO MnO MgO CaO Na 2 0 K20 P205 LOI Total

76.75 0.13 12.27 0.64 0.61 0.02 0.18 0.39 3.24 4.94 0.04 0.96 100.17

74.30 - 78.50 0.03 - 0.25 11.30 - 13.00 0.08 - 1.69 0.10 - 1.31 <0.01 - 0.05 <0.01 - 0.43 0.15 - 1.02 2.40 - 3.83 3.23 - 5.52 <0.01 - 0.13 0.20 - 1.90

Li Ba Rb Sr Zr Nb Th U Y Ce Nd La Ga Sc V Cr Ni Cu Pb Zn As Sn

39 86 444 21 143 15 59 11 61 92 40 44 20 2 6 15 5 7 31 30 12 13

< 5 - 101 22- 211 319- 1182 4 - 58 82- 192 3 - 42 3 7 - 86 4 - 30 2 9 - 145 2 9 - 179 12- 92 2 0 - 69 17- 33 5 <2< 3 - 15 < 2 - 60 < 2 - 12 < 2 - 22 2 0 - 50 12- 47 2 - 35 4 - 27

Gilgai1" (average)

Mole# (average)

Bundarra§ (average)

76.54 0.11 12.22 0.71 0.57 0.02 0.19 0.53 3.53 4.77 0.04 0.51 99.74

75.68 0.12 12.56 0.89 1.64 0.05 0.21 0.60 3.49 4.70 0.07 0.72 100.73

74.48 0.23 13.22 1.10 1.44 0.05 0.40 1.01 3.12 4.74 0.14 0.94 100.87

68

125 204 598 51 105 23 41 19 93 99 45 28 21

53 391 266 84 141 9 32 6 36 75 35 35 17 5 16 9 4 33 22 34

-

-

355 35 124 22 42 7 109 91 72 63 21 6 5 7 4 12 79 94 14 10

— -

12 33 28 -

—

8

—

* Average composition and range of 28 samples of Dumboy-Gragin Granite, with data from Vickery (1993) and Geological Survey of New South Wales database. t Average composition of 50 samples of Gilgai Granite, with analyses from Geological Survey of New South Wales database. # Average composition of 68 samples of Mole Granite, with analyses from Geological Survey of New South Wales database. § Average composition of 92 samples of granitoids from the Bundarra Plutonic Suite, with analyses from Geological Survey of New South Wales database. Blanks in trace-element data indicate not analysed or insufficient results for meaningful determination of average.

was relatively reduced (Blevin & Chappell 1993). The occurrence of late/post-magmatic Al-rich minerals could be due to either or both feldspar fractionation and hydrothermal alteration. The U-Pb zircon radiometric age of 243.2 ± 3.7 Ma of the Dumboy-Gragin Granite confirms the proposal of Stroud (1992) that the pluton was likely to be of Early Triassic age. The result accords with Rb-Sr biotite ages of Shaw and Flood (1993) from southern New England Orogen leucogranites, including the Mole Granite (237242 Ma), Gilgai Granite (243-247 Ma) and Ruby Creek

Granite (239-241 Ma). Clearly, the emplacement of the Dumboy-Gragin Granite post-dated that of the PermoCarboniferous Bundarra Plutonic Suite for which ages of 286 ± 13 Ma and 287 ± 10 Ma were reported by Flood and Shaw (1977) and Hensel et al. (1985), respectively. Although it is possible that the eastern margin of the Dumboy-Gragin Granite intruded the Bundarra Plutonic Suite, inliers of metasedimentary rocks northeast of Delungra imply that the former is separated from the latter by a screen of Anaiwan terrane rocks (Figure 1).

270

N. M. V I C K E R Y ET

AL.

Chemically, the Dumboy-Gragin Granite has strong affinities with other leucogranites of the southern New England Orogen (Table 2) and all conform closely to the compositions of fractionated I-type granites from the Lachlan Fold Belt (Chappell & White 1992; Blevin & Chappell 1995). Although the Dumboy-Gragin Granite and other southern New England Orogen leucogranites are, on average, mildly peraluminous, they are not as peraluminous as the cordierite-bearing granitoids of the Bundarra Plutonic Suite. In contrast with the latter, the Dumboy-Gragin Granite contains low P 2 0 5 and clearly plots with fractionated I-type granites on the Rb-P 2 0 5 variation diagram of Chappell and White (1992). In addition, the average composition of the DumboyGragin Granite contains higher Si0 2 , K 2 0, Rb, Nb, Th, U, Y, Ce and Pb, and lower Ti0 2 , total Fe, MgO, CaO, Ba, Sr and Cu than average Bundarra Plutonic Suite (Table 2). Contents of Sn in least-altered DumboyGragin Granite accord with other southern New England Orogen leucogranites (Table 2) and with Lachlan Fold Belt fractionated I-type granites (Blevin & Chappell 1992, 1995; Chappell & White 1992). The enrichment of Sn by fractionation processes in the Dumboy-Gragin Granite was probably a prerequisite to its hydrothermal concentration in vein-hosted greisen deposits. In conclusion, it has been demonstrated that the Dumboy-Gragin Granite in the southern New England Orogen has strong temporal and petrologic links with other mineralised, fractionated I-type leucogranites in the region. The pluton clearly post-dates and is unrelated to the S-type granitoids of the Bundarra Plutonic Suite, confirming the proposal of Stroud (1992). ACKNOWLEDGMENTS

In: Flood P. G. and Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 423-429. Department of Geology and Geophysics, University of New England, Armidale. BLEVIN P. L. & CHAPPELL B. W. 1995. Chemistry, origin, and

evolution of mineralized granites in the Lachlan Fold Belt, Australia: the metallogeny of I- and S-type granites. Economic Geology 90, 1604-1619. BROWN R. E. & STROUD W. J. 1993. Mineralisation related to

the Gilgai Granite, Tingha-Inverell area. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 431-447. Department of Geology and Geophysics, University of New England, Armidale. CARNE J. E. 1911. The tin-mining industry and the distribution of tin ores in NSW. Geological Survey of New South Wales Mineral Resources 14. CHAPPELL B. W. 1994. Lachlan and New England: fold belts of contrasting magmatic and tectonic development. Journal and Proceedings, Royal Society of New South Wales 127, 47-59. CHAPPELL B . W . & WHITE A . J. R .

1 9 9 2 . I- a n d

S-type

granites in the Lachlan Fold Belt. Transactions of the Royal Society of Edinburgh, Earth Sciences 83, 1-26. COMPSTON W . , WILLIAMS I. S., KIRSCHVINKJ. L., ZHANG Z . &

GUOGAN M. A. 1992. Zircon ages for the Early Cambrian time-scale. Journal of the Geological Society of London 149,171-184. FLOOD P. G . & AITCHISON J. C . 1 9 8 8 . T e c t o n o s t r a t i g r a p h i c

terranes of the southern part of the New England Orogen. In: Kleeman J. D. ed. New England Orogen tectonics and metallogenesis, pp. 7-10. Department of Geology and Geophysics, University of New England, Armidale. FLOOD R. H. & SHAW S. E. 1977. Two 'S-type' granite suites with low initial 87 Sr/ 86 Sr ratios from the New England Batholith, Australia. Contributions to Mineralogy and Petrology

61, 163-173.

HENSEL H . - D . , MCCULLOCH M . T . & CHAPPELL B . W . 1 9 8 5 .

The sample for U-Pb zircon ion microprobe dating was collected by Charlotte Allen and funding for analysis came from a collaborative CSIRO-University of New England grant. Jim Stroud is thanked for providing geochemical data on New England leucogranites from the Geological Survey of New South Wales database. We thank Phil Blevin, John Kleeman and Nick Stephenson for reviews of the paper. REFERENCES BARNES R . G . , BROWN R. E., BROWNLOW J. W . , GILLIGAN L. B . , KRYNEN J. & W I L L I S I. L. 1 9 8 8 . A r e v i e w o f t h e

mineral deposits of the New England Orogen in New South Wales. In: Kleeman J. D. ed. New England Orogen tectonics and metallogenesis, pp. 211-227. Department of Geology and Geophysics, University of New England, Armidale. BLEVIN P. L. & CHAPPELL B. W. 1992. The role of magma

sources, oxidation states and fractionation in determining the granite metallogeny of eastern Australia. Transactions of the Royal Society of Edinburgh, Earth Sciences 83, 305-316. BLEVIN P. L. & CHAPPELL B . W . 1 9 9 3 . T h e i n f l u e n c e o f

fractionation and magma redox on the distribution of mineralisation associated with the New England Batholith.

The New England Batholith: constraints on its derivation from Nd and Sr isotopic studies of granitoids and country rocks. Geochimica et Cosmochimica Acta 49, 369-384. HOLTZ F., PICHAVANTM., BARBEY P. & JOHANNES W . 1 9 9 2 .

Effects of H 2 0 on liquidus phase relations in the haplogranite system at 2 and 5 kbar. American Mineralogist 77, 1223-1241. PLIMER

I. R . &

KLEEMAN J. D .

1985.

Mineralisation

associated with the Mole Granite, Australia. In: High heat production (HHP) granites, hydrothermal circulation and ore genesis, pp. 563-569. Institution of Mining and Metallurgy, London. SHAW

S. E . & F L O O D R . H .

1981. T h e N e w

England

Batholith, eastern Australia: geochemical variations in time and space. Journal of Geophysical Research 86, 10530-10544.

SHAW S. E. & FLOOD R. H. 1993. A compilation of late

Permian and Triassic biotite Rb-Sr data from the New England Batholith and areas to the southeast. In: Carr P. F. ed. Centre for Isotope Studies, Research Report 1991-92, pp. 151-155. CSIRO, North Ryde. STROUD W. J. 1992. New and revised geological units from the Inverell and Goondiwindi 1:250 000 sheet areas, New England Fold Belt. Geological Survey of New South Wales Quarterly Notes 87, 1-16.

DUMBOY-GRAGIN GRANITE TERA F. & WASSERBURG G. J. 1972. U - T h - P b systematics in

three Apollo 14 basalts and the problem of initial Pb in lunar rocks. Earth and Planetary Science Letters 14, 281-304. VICKERY N. M. 1993. The geology and geochemistry of the Dumboy-Gragin Granite, near Delungra, NSW. Graduate Diploma of Science thesis, University of New England, Armidale (unpubl.). WEBER C. R. 1974. Woolomin-Texas Block. Plutonic rocks and intruded sediments. In: Markham N. L. & Basden H.

271

eds. The mineral deposits of New South Wales, pp. 351— 391. Geological Survey of New South Wales, Sydney. WILLIAMS I. S. & CLAESSON S. 1987. Isotopic evidence

for the Precambrian provenance and Caledonian metamorphism of high grade paragneisses from the Seve Nappes, Scandinavian Caledonides. II. Ion microprobe zircon U-Th-Pb. Contributions to Mineralogy and Petrology 97, 205-217. (Received 10 April 1996; accepted 14 July 1996)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 272-289.

Vitrinite reflectance and isotopic evidence for epithermal mineralisation at the Copeland goldfield, southern New England Fold Belt, New South Wales J. H. GIBSON A N D P. K. SECCOMBE

Department of Geology, University of Newcastle, NSW 2308, Australia.

Gold-bearing quartz veins of epithermal type in a low metamorphic grade turbidite succession of Devonian age at Copeland formed in fractures and shear zones produced during sinistral strike-slip faulting on the Peel-Manning Fault System during the Hunter-Bowen Orogeny (265-250 Ma). Near-zero 8 34 S values for ore sulfides, a restricted range in 5 1 8 0 composition for quartz (17.2 to 18.0%o) and calcite (12.6 to 15.7%o), S13C composition of calcite (-1.2 to 4.7%o) and a mantle signature for lead isotopic data from veins and wall rock indicate that the gold deposits formed from hydrothermal fluids of magmatic origin, derived from a mantle or deep-crustal melt of an age and composition similar to the Barrington Tops Granodiorite. Late-stage fluids are influenced by meteoric water and isotopic exchange with organic carbon and sedimentary carbonate. SHRIMP II in situ analysis of framboidal pyrite in host-rock sequences clearly defines a biogenic source of sulfur (534S range -37 to -16%o) unrelated to gold deposition. Zones of gold deposition appear to be defined by: (i) a narrow range in peak temperatures near 350°C, as indicated by the mean random reflectance of dispersed organic matter; and (ii) sedimentological and mineralogical attributes of the host Devonian metasediments, especially intervals marked by the presence of fine-grained, carbonaceous siltstones. Key words: epithermal deposits, gold, lead isotopes, New England Fold Belt, stable isotopes, vitrinite reflectance.

INTRODUCTION

The Copeland goldfield incorporates a number of small auriferous vein systems developed in low metamorphic grade siltstones and sandstones of Palaeozoic age near the village of Copeland, situated 15 km west of Gloucester and 105 km north of Newcastle in the southern New England Fold Belt of New South Wales. Gold-bearing veins are mainly restricted to the Copeland inlier, a 17 km north-trending belt of gently folded sandstone, siltstone and minor conglomerate assigned to the Bowman beds of Late Devonian age, which generally defines an uplifted block of high relief to the northeast of the mountainous Barrington Tops plateau. Flanking the Copeland inlier to the south and west is a conformable sedimentary sequence of Carboniferous age, including the Berrico Creek, Carsonville, Copeland Road Formations and Wootton beds (Figure 1). The Late Devonian and Early Carboniferous rocks form part of the Tamworth Belt (Blake & Murchey 1988), which represents a weakly deformed, Late Palaeozoic forearc basin sequence east of a coeval magmatic arc and separated to the east from more highly strained rocks of an accretionary complex (Central Block) by the Peel-Manning Fault System. In the eastern part of the Copeland area (Figure 1), Upper Devonian metasediments are in fault contact with a melange of predominantly Permian Manning Group diamictites and siltstones within the Peel-Manning Fault System.

Alluvial gold was found within Copeland Creek in the first quarter of 1876 and by the last quarter of that year reef mining from the small and discontinuous reefs had begun. From an Au production of 1.2 kg from gangue in 1876, production rose to a peak of 311.5 kg in 1879, falling gradually to 240 kg in 1882, and then rapidly to 8.6 kg in 1887. Since 1887 only small and spasmodic production was recorded until all mining was abandoned in 1940. Total official gold production from the field amounts to 1738 kg; representing 1511 kg of lode gold and 227 kg produced from alluvial workings (Osbourne 1975). Apart from the review work of Gilligan and Brownlow (1987), a minor geological study of one of the lines of lode (Golden Spur North: Cameron 1981) and the works of Gibson (1993) and Gibson and Seccombe (1993, 1994), which have reported on aspects of the mineralisation and metamorphism in the region, relatively little attention has been paid to this dormant goldfield. As compiled by Osbourne (1975) earlier reports on the mining field stressed development within the mines and have largely ignored the geology of the region. Despite the small size of the goldfield, interest in the area arises because of the enigmatic nature of the mineralisation. Mineralisation is clearly structurally controlled, however, structures have formed under essentially a brittle regime and the timing of the structures and their relationship to major sutures such as the Peel-Manning Fault System have remained unclear.

MINERALISATION, COPELAND GOLDFIELD, NSW

273

-70

-60

\ » Anticline with plunge

Permian

\

p T H Manning Group

> Syncline with plunge

— ?— Figure 1 Regional geology and principal structural features of the Copeland goldfield. Mines are numbered: 1, Black Prince; 2, Bromley; 3, Centennial; 4, Golden Crystal; 5, Golden Spur; 6, Great Britain; 7, Hidden Treasure; 8, Melbourne; 9, Mint; 10, Morning Star; 11, Mountain Maid; 12, Prince Charlie; 13, Rainbow; 14, Rose & Thistle.

Geol. boundary-inferred

\

| Buckets Gap - Mclnnes Formations

I 1 Berrico Ck/Carsonville/Copeland Rd Li—21 Formations/Wootton beds

Strike a n d d i p of strata

Mine Production (Au) 0 +100 kg • 99-10 kg o 9-1 kg

Models for vein-style gold-quartz mineralisation such as the slate-belt type appear to be discounted due to the very low grade of metamorphism and low strain in the host rocks. By a similar argument, the lack of igneous

Carboniferous |

Fault accurate — — Fault inferred

Devonian 1 1 Bowman beds

0

Scale 2 km

activity in the area does not immediately suggest a link between gold mineralisation and magmatic activity. This paper addresses a number of these problems by integrating the regional geology and structure with a

274

J. H. G I B S O N A N D P. K. S E C C O M B E

study of the mineralogy and paragenesis of the major deposits in the field. Stable isotopes (S, C and O) are used together with Pb-isotopes, a limited fluid-inclusion investigation and vitrinite-reflectance geothermometry to constrain the ore-forming environment, to establish those factors that appear to be important in controlling ore deposition and to define a source of gold and other ore components in the district. STRATIGRAPHY The economic auriferous vein systems lie entirely within Late Devonian metasiltstones and metasandstones of the Bowman beds (Figure 1). Carboniferous strata are poorly mineralised and contain only one uneconomic deposit in sandstone units of Visean age, which comprise part of the Wootton beds. Two contrasting lithologies are evident in the Bowman beds (Gibson & Seccombe 1993, 1994). (1) Thick sequences of thinly bedded to laminated, medium- to fine-grained, pyritic dark-coloured sandstones, interspersed with minor siltstone and mudstone. Within these lithologies sedimentary structures are preserved, such as small-scale dewatering and slump structures, current ripples and tool marks. These units contain abundant dispersed organic matter; rarely beds contain numerous current-deposited fragments of Leptophloeum australe, the indicator fossil for the Late Devonian epoch (White 1986, 1988). The presence of this plant material suggests a nearshore marine environment of deposition, in contrast with a deeper marine environment protected from detrital input, represented by other parts of the sequence which contain abundant radiolarians. (2) Within the finer grained units are vertically extensive sequences of very thickly bedded, pale bluegrey, medium-grained and poorly sorted volcaniclastic sandstone, which grades to granule conglomerate. The volcaniclastic units contain little or no dispersed organic matter or pyrite and contain a detrital plagioclase of An30_40 composition, indicative of dioritic source rocks (Kerr 1977). Typically, the units are tens of metres thick and commence with an erosional base and a basal layer of boulder to granule, mud interclasts or rare sedimentary breccia. Since the units appear to have little lateral extent, they are interpreted to represent mass-flow units resulting from the remobilisation of pyroclastic material that was dumped quickly onto a marine shelf (Gibson & Seccombe 1993). Redistribution of the volcaniclastic material across a substrate of unconsolidated, finegrained carbonaceous sand, would allow disruption and incorporation of portions of the fine-grained lithologies (sequence 1) within the volcaniclastic mass-flow units. A third minor rock type is noted, consisting of rare, channelised, pebble polymictic paraconglomerates. The clasts contained within this unit are likely to have been derived from a Lachlan Fold Belt hinterland. INTRUSIVE ROCKS Within the map area intrusive rocks are absent, apart from the outcrop of a small basaltic dyke in the bed of

the Barrington River. The lack of metamorphism displayed by the dyke and the similarity of texture and mineralogy to other dated intrusions point to it being of Tertiary age (R. Offler pers. comm. 1994). A suggested location of a kimberlitic diatreme recorded in the Gloucester district (MacNevin 1977) is Prince Charlie Creek (Figure 1, location 12), but no evidence of such a diatreme in that area has been found in this study. To the west of the area lies the Barrington Tops Batholith, dated at 265 ± 8 Ma (Roberts & Engel 1987; Collins et al. 1993). Isotopic evidence from this isotopically primitive, orthopyroxene-bearing granitoid (Collins et al. 1993) suggests that it is entirely or partly mantle-derived (Hensen et al. 1992), being emplaced as a dry (<0.2% H 2 0) and hot magma (1000°C) (Hensen & Shaw 1990). STRUCTURAL SETTING Regionally, the structural style established in the Copeland area can be assigned to the Dj and D 4 deformation events of the four tectonic stages proposed by Collins (1991) for the mid-Permian, Hunter-Bowen Orogeny. Within the Copeland area, east-west compression of the Devonian and Carboniferous units has produced a major, meridionally trending F! fold. The fold style is gentle to open, upright (Fleuty 1964) and sub-cylindrical (Ramsay & Huber 1987). The fold axis has a plunge of about 10° towards 189° and a halfwavelength of approximately 10 km. Owing to the gentle nature of the folding, and the small degree of shortening, an axial plane cleavage has not developed. Only two types of cleavage or closely spaced jointing have been noted. One cleavage style is represented by a widely spaced fracture cleavage developed adjacent to some fault zones, that can be defined as a 'disjunctive cleavage' using the terminology of Kisch (1991). The second is a 'scaly cleavage' (Kisch 1991) which exhibits anastomosing, closely spaced and slickensided surfaces, found only in zones along the Boonara Fault. The D 4 deformational event was a sinistral shearing event, culminating in the development of the PeelManning Fault System (Collins 1991). D 4 deformation rotated ¥ { fold axes and truncated the Gloucester Syncline. Within the mineralised area in the central part of the Copeland goldfield (Figure 1), an equal-area stereographic projection of 115 poles to bedding indicates that the Fx fold axis has rotated to 168° and the plunge has steepened to 18°. The change in the orientation of the axial surface of this major structure is attributed to D 4 left-lateral movement on the PeelManning Fault System. A further response to D 4 movement adjacent to the Peel-Manning Fault System is evident from smaller scale, F 4 fault-folds developed within 4 km of the Boonara Fault in the eastern part of the map area (Figure 1). These F 4 folds are open, plunge gently to the south and have axial surfaces which are inclined steeply to the east. They have a halfwavelength of 0.5-1 km and an amplitude of 0.4 km. The Boonara Fault has a trend of 145-150°, dips west at about 60° and can be traced from the northern boundary of the mapped area to terminate near the southeastern

275 areas affected by hydrothermal alteration, the metamorphic grade increases to lower greenschist facies. A general lack of metamorphic index minerals has made the definition of metamorphic boundaries within areas of higher metamorphic grade difficult. For example, although most of the Late Devonian strata have undergone prehnite/pumpellyite facies metamorphism, pumpellyite is not a stable phase in these rocks. Instead, a common metamorphic assemblage comprises albite-epidote-^chlorite-sericite and minor calciteprehnite. Since only poor control of metamorphic facies is afforded by the silicate assemblages, peak temperatures achieved in the metasediments and alteration zones were established from the reflectance of dispersed organic matter particles.

MINERALISATION, COPELAND GOLDFIELD, NSW boundary. Together with the en echelon Bowman River Fault further east, these faults are segments of the PeelManning Fault System. GOLDFIELD STRUCTURE Local control on the orientation of the vein systems in the Copeland Goldfield is related to left-lateral shear on the Peel-Manning Fault System and subsidiary faults. A plot of the trend for 25 productive veins from the Copeland area (Figure 2), shows that the bulk of the veins trend towards 070°, which represents the R riedel shear direction (Ramsay & Huber 1987) assuming D sinistral movement on the Peel-Manning Fault System (Offler & Williams 1987; Offler et al 1989). Secondary vein orientations are (i) at 045°, parallel to the direction of X fractures (Bartlett et al 1981) or (ii) parallel to the trend of the main fault (150°) and the orientation of predicted Y fractures (Bartlett et al 1981). Figure 2 also indicates that the dip of the productive veins is generally in excess of 60°. The geometry of the quartz-filled fractures fits the conditions for transpression between right-hand en echelon, left-hand shear fault segments (Ramsay & Huber 1987). Transpression is accomplished between the Boonara Fault segment of the Peel-Manning Fault System in the northeast, which terminates approximately 8 km east of the Copeland goldfield, near the township of Barrington, and an unnamed fault in the extreme southwest corner of the map area (Figure 1) that terminates to the northwest of the region. METAMORPHISM 2

4

The rocks have undergone burial metamorphism at very low grades, ranging from zeolite facies in the strata of Middle Carboniferous to Early Permian age and progressing to prehnite/pumpellyite facies in the Late Devonian to Early Carboniferous sequences. Around the

Figure 2 Rose diagram giving directional trends of 25 reefs within the Copeland goldfield. The important economic veins and the dominant orientations of the reef systems coincide with the predicted riedel shear direction (R->) of Ramsay and Huber (1987) and X and Y fracture orientations of Bartlett et al (1981). Y fractures are parallel to the trend of the PeelManning Fault System. The near-field principal stress direction (a,) bisects the acute angle between the predicted orientation of riedel shear sets (Rj and R ). The inset diagram indicates that dips of the veins are generally greater than 60°. 2

MINERALISATION All the mines examined within the area have a similar mineralogy (Figure 3); variations are encountered only in the abundance and distribution of the minerals. Quartz is the most abundant gangue mineral, accompanied by calcite, sericite and rare chlorite and prehnite; the latter three vein minerals are found mainly within zones of vein breccia. Macroscopic vein textures, described using the terminology of Dowling and Morrison (1989) comprise anhedral and euhedral buck quartz, together with fine-grained and rarer coarsegrained comb structure. Whereas ribbon gangue is rare, laminations defined by wall-rock inclusions are common. Superimposed features include infill and aggregate vein breccia and less common polished slickensides and slickenlines. Textures described are verified at the microscopic scale, together with latestage, quartz--calcite spider veins within the brecciated veins. Sulfides amount to less than 1% by volume of the vein-fill and may be absent, or exist only in trace Copeland Fracture Zone

276

J. H. G I B S O N A N D P. K. S E C C O M B E STAGE

PARAGENETIC Mineralogy

1

quartz calcite sericite pyrite arsenopyrite

0

O OL a> chalcopyrite x pyrrhotite

LU

Q_

3

00

covellite gold

300

200

100

Estimated Temperature °C

amounts within many veins. Arsenopyrite and pyrite are the major sulfides; chalcopyrite and pyrrhotite are present only as minor phases. Sulfides are associated with included laminae of wall rock parallel to vein walls or are dispersed within the gangue mineral assemblage. Vein evolution follows two major stages (Figure 3). During Stage 1, early deposition of quartz and arsenopyrite, the latter showing minor inclusions of pyrite and chalcopyrite, is followed by a more complex mineralogy comprising quartz-calcite-arsenopyrite with minor pyrite, chalcopyrite, sericite and gold. The major period of gold deposition occurs early in Stage 2, when pyrite, rather than arsenopyrite, is the principal sulfide, and quartz or calcite is the major gangue mineral. Pyrrhotite is encountered as rare inclusions within large, late-stage pyrite grains. A supergene stage is identified during which secondary pyrite, covellite and calcite was deposited. In the larger mines, gold production was confined to distinct zones or 'short shoots' of limited horizontal or vertical extent (Kenny 1935) and most mines are represented by shafts which rarely exceed 100 m in depth. Of the seven mines with a recorded production of over 50 kg of gold, five attained a depth of between 100 and 190 m, whereas the depth of smaller mines commonly ranged between 20 and 30 m. Individual veins rarely have a strike length of greater than 10-20 m, although zones of fracturing in which the vein systems are confined may extend for hundreds of metres. Widths of auriferous veins can range from a few centimetres to over a metre, but the veins in more productive parts of the mines were seldom wider than 60 cm. Where thick veins were encountered, the gold tended to be confined to thin zones or 'shoots' adjacent to wider intervals of barren or low-grade gangue. Gold ranges from a sub-millimetre to decimetre grainsize and is restricted to small, discontinuous veins (2 cm—1.5 m wide) composed principally of quartz and calcite (Gibson & Seccombe 1993). Evidence of the late-stage nature of much of the gold comes from its

Figure 3 Mineral paragenesis diagram for the Copeland veins. Temperature estimates for each stage are constrained by vitrinite reflectance, mineralogy and limited fluid-inclusion data (see text).

distribution as films along grain boundaries, fracture planes and vein laminations and as dustings within vuggy cavities. Within the hydrothermally altered areas, especially adjacent to auriferous quartz veining, calcite, albite, pyrite and arsenopyrite dominate the alteration assemblages, accompanied by minor prehnite. Sericite is confined entirely to the veins, whereas pyrite and arsenopyrite are restricted to an alteration zone at the centimetre/decimetre scale from vein boundaries, and calcite and prehnite define alteration in the metasediments to a distance of several metres from vein contacts. All but the largest plagioclase clasts are extensively altered and are conspicuous when the replacing assemblages preserve relict outlines. In some instances calcite defines a foliation, but is more commonly present as rounded aggregates of likely hydrothermal origin. The sulfide mineralogy within alteration zones is similar to that within the veins, except that arsenopyrite is a minor phase and gold is not evident in heavymineral concentrates prepared from sulfide-rich alteration zones. The exception to this is where zones of brecciation have developed, commonly within the more porous sandstones, to produce abundant arsenopyrite. ANALYTICAL METHODS Vitrinite reflectance Selected samples of fine-grained, dark-coloured and generally well-laminated metasediments were cut normal to bedding and polished in accordance with standard techniques. These blocks were viewed in oil under a Leitz MPV1 incident-light microscope at a magnification of 320x to 500x and measurements of mean random reflectance (Rr) were performed on 40 detrovitrinite or vitrinite particles per sample. Reported results are the arithmetic mean of at least forty readings. Barker and Goldstein (1990) contend that sources of

MINERALISATION, COPELAND GOLDFIELD, NSW

277

On

km

10J N

A

Figure 4 Vitrinite reflectance for dispersed organic matter particles in the Copeland goldfield. (a) Mines are confined to a north-northwest-trend in vitrinite reflectance and cluster around the boundary corresponding to the transition between anchizone and epizone metamorphic conditions (mean Rr of 5%), equivalent to the 350°C vitrinite isotherm, (b) Enlargement of inset. A west-northwest-trending corridor of low vitrinite reflectance separates two lines of lode in the central part of the goldfield (see also Figure 1).

• Mine production over 10 kg

^

Contoured in % mean random reflectance of vitrinite.

1

' km

^

N

'

\ A

278

J. H. GIBSON AND P. K. SECCOMBE

Table 1 Mean random reflectance (/?,.) and calculated temperatures for vitrinite dispersed organic matter particles from the Copeland mining field. Mine

R

Bromley Golden Crystal Centennial Rose & Thistle No. 1W Rose & Thistle No. 2W Rainbow Prince Charlie Melbourne Morning Star Hidden Treasure Great Britain Mint

5.32 5.23 5.22 5.12 5.08 5.11 5.10 5.09 5.02 4.98 4.89 4.88

r

Temperature Range (°cr (°C) 361 359 359 357 356 356 356 356 354 353 351 351

328-372 343-370 349-369 348-373 330-368 350-374 335-374 345-370 343-371 340-372 337-367 333-373

* Temperatures calculated from Barker and Goldstein (1990). error in the approach include (i) vitrinite reflectance suppression in hydrogen-rich environments typically encountered in weathering profiles, and (ii) the timedependent nature of vitrinite reflectance, which may reduce the ability of the dispersed organic matter material to record brief periods of exposure to high temperatures. At Copeland, slight weathering had minimal to no effect on the suppression of the statistically derived reflectivity (e.g. R for weathered material = 4.79%; unweathered = 4.84%), whereas more intensely weathered material was avoided in the study (e.g. R for strongly weathered material = 4.65%; unweathered = 4.88%). Another potential source of error would be the misidentification of vitrinite, especially since mean random reflectance investigations are carried out without polarisers within the light path. Detro-inertinite was distinguished from vitrinite by its shape and higher reflectance. Dispersed organic matter particles exhibiting a platy or granular texture or the characteristic strong reflectance of transitional material (OkuyamaKusunose & Itaya 1987; Diessel et al. 1978) are more difficult to distinguish from vitrinite and were avoided. Reflectivities (Figure 4a) and calculated peak temperatures were established for 117 sites selected within an area of 625 km defined by Figure 1. Special attention was paid to the mines within the area, where suitable specimens were collected from mullock dumps (see Table 1). An emphasis was also placed on sampling close to the area of maximum gold production close to the centre of the goldfield (Figure 4b). r

r

2

Fluid inclusions Approximately 30 doubly polished plates were prepared from veins containing either quartz or calcite, collected during the regional sampling program. Most material proved barren of measurable fluid inclusions.

Homogenisation temperatures only were established for quartz from the Rose and Thistle deposit (Figure 1, location 14) using a Reynolds gas-flow heating-freezing stage mounted on a Leitz Laborlux-S microscope equipped with a 32x objective lens, 12.5x oculars and JVC video monitor, which permit T data to be collected on inclusions as small as 3 jum. h

Isotopes Isotopic analysis of a variety of vein and wall-rock minerals was undertaken to establish the source of the mineralisation, the nature of the ore-forming fluids and the conditions of ore deposition at Copeland. Regional sampling of vein and wall-rock material from abandoned ore dumps for several of the lines of lode resulted in 52 conventional and ion microprobe 5 S analyses of sulfides, 15 5 C and 5 0 of vein carbonates, five 5 0 analyses of vein quartz and 13 lead isotopic-ratio determinations on a variety of pyrite samples. Most isotope analyses were undertaken at the Centre for Isotope Studies, CSIRO, North Ryde, New South Wales. Some samples of framboidal pyrite encountered in the Devonian host rock were analysed by the SHRIMP II ion microprobe at the Research School of Earth Sciences, Australian National University, Canberra. The majority of pyrite and arsenopyrite samples collected for sulfur isotopic study were obtained from mullock and abandoned ore dumps. The selected samples were crushed in a jaw crusher and sieved to 1.0 and 0.5 mm size. Only recognisable crystalline pyrite and arsenopyrite were chosen from the sieved fractions for sulfur-isotope analysis. Three samples of the host rocks which contained in excess of 10% sulfide, were selected for whole-rock isotopic analysis. Two samples were obtained from the Devonian Bowman beds and one represents adjacent Carboniferous strata. The sulfides in these samples are predominantly pyrite. A single sample of pyrrhotite was also extracted magnetically from a finely crushed host rock sample. One sample of pyritic Devonian sandstone, reporting a 5 S value of -6.5%o by conventional analysis was analysed by the SHRIMP II ion microprobe. For this sample, in situ analyses were conducted on a number of pyrite framboids in a section prepared normal to bedding, to check for possible stratigraphic variations in the isotopic composition of biogenic sulfide. Samples of calcite and quartz were collected from a number of mines to test for potential district-scale variation. Complementary to this, a mine-specific study was carried out to sample paragenetically early and late calcite from shafts along the Mountain Maid line of reef (Figure 1, location 11). Early calcite is typically white, vein-fill material which is commonly brecciated. Latestage calcite is in two forms: (i) white calcite in latestage fractures; or (ii) clear subhedral to euhedral calcite in vugs. Lead isotopic analysis was undertaken at the Centre for Isotope Studies on 13 samples of pyrite crystals extracted from the same vein and wall rock samples as used for the sulfur-isotope study. 34

13

34

18

18

MINERALISATION, COPELAND GOLDFIELD, NSW RESULTS AND DISCUSSION Vitrinite reflectance survey Most sedimentary rocks contain dispersed organic matter particles that respond similarly to transport and deposition as other mineral and lithic detritus. These vegetal particles undergo the same post-depositional processes that lead to the formation of coal and comprise the same macerals as coal (Diessel & Offler 1975; Duba & William-Jones 1983; Barker & Goldstein 1990; Feinstein et al 1991). Vitrinite is the maceral usually employed in metamorphic investigations. Because of the small grainsize (generally <0.02 mm) of the dispersed organic matter particles, the maceral detrovitrinite forms the basis of this study. The degree of coalification and the reflectance of vitrinite dispersed organic matter particles from sub-bituminous to meta-anthracite rank is a sensitive recorder of the peak temperatures experienced by the rock mass (Barker & Goldstein 1990; Feinstein et al. 1991). Unlike metamorphic index minerals, dispersed organic matter particles are known to monitor changes in temperature without the aid of a fluid phase or the redistribution of mineral components. Since the coalification of dispersed organic matter particles is irreversible, they remain stable in retrograde metamorphic regimes and thus have the potential to retain signatures of a thermal peak. Early work on vitrinite reflectance in diagenetic to very low-grade metamorphic environments (Bostik 1971; Diessel & Offler 1975) has been expanded to identify the conditions under which rocks have attained diagenetic, anchizone and lower-epizone metamorphic conditions (Arkai 1991). Barker and Goldstein (1990) have further extended the use of vitrinite reflectance to erect an empirical geothermometer, which is particularly applicable to low P-T environments in which a suitable mineralogy for conventional metamorphic studies is lacking. Temperatures were calculated from vitrinite reflectance data using the empirical relationship established by Barker and Goldstein (1990), the calibration for which is provided by the homogenisation temperatures of fluid inclusions in calcite: In (« r ) = 0.0081 l(Tpeak)-1.26 where RR is the mean random reflectance of vitrinite in oil and T peak represents the peak temperature achieved in the rocks. Productive deposits have RR values in the range from 4.9 to 5.3%, corresponding to a peak temperature of 351-361°C, respectively (Table 1). The deposits are confined to a north-northwest-trending zone of high RV values (Figure 4a), which is centred on the axial trace of the regional anticline developed in the Copeland inlier (Figure 1). In detail, however, substantial variations in RT values are evident, such as in the central part of the field (Figure 4b) where a west-northwest-trending corridor of low RR values (as low as 4.4%) separates two lines of gold-bearing deposits (RV in the range

279

5.0-5.2%) spaced approximately 2 km apart. The link between high RR values and mineralisation is suggested by the sampling profile adjacent to the Bromley vein (Table 2) where reflectance increases by over 0.3 units within 40 cm of the vein contact. Table 2 Mean random reflectance (R r ) and calculated temperatures adjacent to the Bromley Vein.

Distance (cm) from vein contact 0 20 30 40

5.51 5.33 5.15 5.21

Temperature (°C)

Range (°C)

366 362 357 359

319-387 326-386 347-370 341-374

Fluid inclusions Additional temperature constraints on the mineralisation were gained from a fluid-inclusion investigation, which yielded data from only one sample of quartz (Rose & Thistle deposit: Figure 1, location 14). Homogenisation temperatures (T h ) ranging from 195—202°C for four primary two-phase (liquid—vapour) inclusions and from 149-181°C for seven secondary inclusions (Table 3) were obtained from late-stage, vuggy white quartz crystals, 5—10 mm in length and overgrown by coarsegrained gold. Table 3 Fluid-inclusion homogenisation temperatures.

Size (jLim)

Temperature (°C)

Primary inclusions 33 33 12 28

203.1 202.4 197.8 195.4

Secondary inclusions 14 20 6 13 14 20 14

181.2 172.0 165.1 156.8 156.2 152.0 148.6

Sulfur isotopes Results of the S-isotopic analyses are presented in Table 4 and Figure 5. Sulfides from veins and wall rock display a restricted range in 534S values of-1.8 to 3.2%o, and -1.8 to 2.7%o respectively. The limited range in 534S composition and similarity in mean isotopic values for

280

J. H. G I B S O N A N D P. K. S E C C O M B E

Table 4 Sulfur isotope composition of sulfide minerals from the Copeland goldfield.

Location

Mines Bromley Centennial

Golden Consul

Golden Crystal Golden Spur Great Britain

Hidden Treasure

Kelly's shaft Kin San Lady Belmore

Lady Lizzie Mint

Morning Star Mountain Maid shaft

Mountain Maid tunnel Prince Charlie tunnel Rainbow Rose & Thistle Rose & Thistle no. 2 East

Metasediments Mud Hut Road quarry (Carboniferous) Coneac Trail (Devonian) Station 44 (Devonian) Station P83 (Devonian)

SHRIMP II analyses Station 44 (Devonian)

Mineral

Sample type

534S (%o)*

pyrite pyrite arsenopyrite pyrite pyrite pyrite arsenopyrite arsenopyrite pyrite arsenopyrite pyrite pyrite pyrite pyrite pyrite pyrite arsenopyrite pyrite arsenopyrite pyrite pyrite pyrite pyrite pyrite pyrite pyrite pyrite arsenopyrite arsenopyrite pyrite pyrite pyrite arsenopyrite pyrite pyrite pyrite pyrite pyrite pyrite pyrite pyrite pyrite pyrite

wallrock vein vein wallrock vein wallrock vein wallrock stockwork stockwork vein vein vein vein vein vein vein wallrock wallrock vein wallrock vein vein vein vein vein vein vein vein vein vein vein vein wallrock wallrock wallrock vein vein stockwork wallrock wallrock wallrock wallrock

-0.9 3.2 3.2 2.7 1.2 1.1 0.6 0.8 -0.9 -0.3 1.0 0.8 3.1 3.0 3.0 2.5 2.2 2.0 2.1 -1.6 -0.9 3.0 2.0 1.6 -0.4 0.5 0.0 1.2 0.5 0.3 -0.5 -1.2 -1.8 -0.9 -0.1 -0.9 0.6 0.1 0.4 -0.6 -1.8 -1.2 -1.3

pyrite pyrite pyrite pyrite pyrrhotite pyrrhotite

metasiltstone (whole rock) metasiltstone (whole rock) metasandstone (whole rock) metasandstone (whole rock) metasandstone metasandstone

-7.5 -8.1 -7.4 -6.5 1.7 1.5

pyrite pyrite pyrite pyrite pyrite

50 |nm framboidal grain 50 \xm framboidal grain (+5 mm) 50 jim framboidal grain (+10 mm) 50 |im framboidal grain (+15 mm) 50 jum framboidal grain (+20 mm)

-32 -32 -37 -16 -31

* Analytical error (2a): conventional analysis, ± 0.2%o:; SHRIMP II analysis, + 2%o.

MINERALISATION, COPELAND GOLDFIELD, NSW 281 Mass-balance effects suggest strongly that the wholesulfides from either veins (1.0%o) or altered wall rock (0.2%o) implies a single source of sulfur for both vein rock signature (-6.5%o) is a mix of light sulfur derived and wall-rock sulfides, especially considering the from framboidal pyrite and a sulfur component with an regional nature of the survey. The closeness of both isotopic composition heavier than -5%o. Re-examination means to zero is suggestive of a magmatic origin for the of the sampling site, a road-side quarry which had been hydrothermal fluids; the slightly more negative value for reactivated between visits, yielded material from a the sulfides from altered wall rock is attributed to slight deeper level with sulfides filling early, fine-scale joints contamination of those sulfides by the sedimentary rock and fractures. These fine sulfide-filled fractures had sequences. Further, the restricted range of data near 0%o been noted from other sites through the district. We suggests that source has a strong mantle component. infer that hydrothermal fluids bearing a sulfur isotopic Ohmoto (1986) contends that rocks generated from composition typical of hydrothermal veins (1.0%o) had partial melts of the mantle and fluids associated with invaded joints and fractures to give a mixed isotopic those melts are likely to have a 5 S composition close signature near -6.5%o for the whole-rock analysis. to that of the parental mantle and in the range from -3 to Pervasive veining from the ore-forming hydrothermal event is likely to be responsible for similar 8 S values The whole-rock samples (mean 5 S = - 7 . 4 % ; range established for the country rock samples taken -8.1 to -6.5%o) are thought to represent sequences elsewhere in the region. As part of the sulfur-isotope study, a comparison was unaffected by hydrothermal activity, except for the sample which contains pyrrhotite. The latter sample made of the 5 S data for pyrite and arsenopyrite crystals extracted from the same crushed sample. The comes from an area close to a shear zone, as evidenced by some fine 'spider web' veining at the collection sulfides come from nine sites; five of which represent auriferous veins and two represent altered wall rock. site. The isotopic composition of the pyrrhotite The intermineral variation in 5 S between arsenopyrite ( 5 S = 1.5%o; repeat sample 1 . 7 % o ) indicates that fluids with a similar isotopic signature to the auriferous vein- and pyrite ranges from -0.3 to 0.6%o for the five vein samples and from -0.6 to 0.1 %o for the two wall-rock forming fluids had permeated through the site. In order to verify that the sulfides within the country samples. Microscopic examination of polished blocks rock samples were of a sedimentary origin, in situ and mineral concentrates indicate that arsenopyrite analysis of five separate pyrite framboids from a single preceded the formation of the large crystals of euhedral rock sample were undertaken by the SHRIMP II ion pyrite. Variations in the sulfur contribution from either a probe. The range of data (Figure 5) from SHRIMP hydrothermal or wall-rock source or fluctuations in the analysis ( 5 S = - 3 7 to - 1 6 % ; mean - 3 0 % o ) indicates isotopic composition of the hydrothermal source may clearly the biogenic nature of the pyrite framboids. account for these small differences in isotopic Additionally, it can be concluded that framboidal pyrite composition of the pyrite and arsenopyrite. cannot be the only sulfur component present in the country rocks, since a 5 S whole-rock value of - 6 . 5 % o Oxygen and carbon isotopes is obtained by conventional S-isotopic analysis for this same sample. We may also conclude that remobilisation The 5 0 composition of quartz (Stage 1) from four of pyritic sulfur from the country rocks was not the deposits (Table 5) has a very restricted range (17.2 to 18.0%o). Isotopic compositions of all early-stage calcites likely source of sulfur for the auriferous vein sulfides. 34

34

34

0

34

34

34

34

0

34

18

|

pyrite

f T j pyrite-framboidal lAl whole rock I

VEIN

| arsenopyrite

[•] pyrrhotite

ALTERATION ZONE

Figure 5 Histograms for sulfur isotopic distribution in sulfides from veins, altered wall rock and host metasediments from the Copeland goldfield. Data for pyrite framboids were obtained from in situ SHRIMP II analyses.

j2

n

w

SEDIMENT ,

-30

i -20

534S %O

r\

•H -10

0+

282

J. H. G I B S O N A N D P. K. S E C C O M B E

Table 5 Oxygen isotope composition of quartz from the Copeland goldfield.

Mine

Sample

5180 (%o)

Hidden Treasure Golden Spur Mint Centennial

JGS1 JGS3 JGS4 JGS5

17.5 18.0 17.2 17.9

S180(fluid)(%o) 200°C 250°C

5.4 5.6 4.8 4.8

7.9 8.6 7.7 7.7

are also very consistent. Stage 1 calcites, either from the Mountain Maid mine or from the four other deposits in the regional survey have 6 1 8 0 values in the range 12.6 to 15.7%o and 5 ,3 C compositions ranging from -1.2 to -4.7%o (Table 6; Figure 6). These results reinforce the view that the ore-forming fluids in the district were homogeneous and were derived from a single source. The C- and O-isotope signature of the calcites relating to ore formation appears significantly different from a sample of calcite obtained from a splay of the PeelManning Fault System sampled northeast of Copeland (Figure 6), implying that fluids focused by this major structure are not associated with the origin of the gold deposits. Fluid temperature prevailing during early quartz deposition (Stage 1) at Copeland is not well constrained, but is taken to be about 250°C. Upper temperature limits for this stage are derived from vitrinite reflectance; lower limits are constrained by Th data near 200°C for later (Stage 2) quartz. Further estimates come from application of the work of Heinrich and Eadington

(1986) on the Fe-As-S system to the mineral assemblages at Copeland, where calcite samples filling voids within quartz, containing both arsenopyrite and pyrite suggest a temperature range from 220 to 320°C. Using fractionation factors derived for quartz and water by Friedman and O'Neil (1977) at temperatures appropriate for quartz deposition at Copeland, calculated 518Ofluid values (Table 5) are near 5%0 (200°C) or 8%o (250°C). The 518Ofluid compositions in this range are consistent with a magmatic source (Rye & Sawkins 1974; Sheppard 1986) and data obtained on I- and Stype granitoids of the Lachlan and New England Fold Belts (O'Neil & Chappell 1977; O'Neil et al. 1977) The 5 1 8 0 composition of Copeland quartz is also similar to examples of sediment-hosted epithermal gold mineralisation, such as Cortez (Rye et al. 1974) and Carlin (Radtke et al. 1980), where either a direct magmatic fluid contribution is feasible or the range in 5 1 8 0 could be achieved by extensive fluid-rock interaction in a sedimentary environment at elevated temperatures. The range in 513C composition (-1.2 to -4.7%o) for all Stage 1 calcites at Copeland does not discriminate between a crustal or magmatic source of carbon (Ohmoto 1986). However, organic carbon can be discounted as a major component in the early ore fluids, since arsenopyrite is the dominant sulfide at this stage of the paragenesis and the ore fluids are likely to be reducing (Ohmoto & Rye 1979). A comparison of 518Ofluid values calculated at 250°C for Stage 1 quartz deposition (mean = 8.0%o) and Stage 1 calcite (mean = 12.3%0) suggests that these two minerals are not in isotopic equilibrium. This conclusion supports textural evidence that commonly indicates overprinting of earlier quartz by later calcite microveins, even in the same overall paragenetic stage. Since the 5l8Ofluid values

Table 6 Oxygen and carbon isotope composition of calcite from the Copeland goldfield.

Mine

Hidden Treasure Mint Black Prince Centennial Mountain Maid Stage 1

Stage 2

Peel—Manning Fault System calcite

513C (°/oo)

5180 (%o)

JGC1 JGC2 JGC4 JGC5

-2.1 -2.4 -1.8 -4.7

15.7 13.9 14.8 14.2

JGC3 JGC7 JGC10 JGC11 JGC12 JGC15 JGC8 JGC9 JGC13 JGC14 JGC6

-2.9 -2.3 -2.5 -2.4 -2.6 -1.2 -7.4 -2.6 -6.6 9.0 -•11.2

14.6 14.3 14.8 14.7 12.6 14.8 20.3 9.0 10.8 8.1 11.6

Sample

S180(fluid)(%o) 150°C 250°C

_ _ _ -

_ 13.2 1.9 3.7 1.0 4.5

13.6 11.8 12.7 12.1 12.5 12.2 12.7 12.6 10.5 12.7

_ _ _ _ 9.5

Figure 6 Carbon and oxygen isotopic compositions of vein carbonate from Copeland gold deposits, compared to fields for organic carbon and meteoric water. • , Stage-1 calcite from Mountain Maid mine; A, Stage-2 calcite from Mountain Maid mine; carbonates from all other deposits; +, calcite from the P e e l - M a n n i n g Fault System (F). N u m b e r e d samples and arrows indicate early- and late-stage calcite pairs f r o m Mountain Maid (see Table 6 for dataset).

M I N E R A L I S A T I O N , C O P E L A N D G O L D F I E L D , NSW 283 for calcite are consistent with metamorphic environ- A model for gold mineralisation at Copeland ments (Field & Fifarek 1985) the isotopic data could indicate that the Stage 1 calcite is derived from local The available evidence points to the Copeland goldfield originating from a Late Permian structurally controlled sedimentary rocks, rather than sourced from magmatic hydrothermal event. Isotopic evidence suggest that the fluids. ore fluids, especially in the early stages of The late-stage calcites from the Mountain Maid de- mineralisation a magmatic signature and that the posit have a wide scatter in 5 0 and 5 C composition lead and sulfurhave with a magmatic origin. (Figure 6). Three Stage 2 calcites show 5 0-depletion Biotite K/ Ar agesareforconsistent Barrington Tops Granodiorite ranging from about 3 to 7%o, consistent with mixing of in the range 269-262 the (Cooper et al 1963; Roberts the Stage 1 fluids and meteoric water. Enrichment in & Engel 1987) and aMabiotite Rb/Sr age of 262 Ma 5 0 noted in a single sample of late-stage calcite (Hensel et al 1985) permit the intrusion to be a source (JGC8) may be attributable to supergene effects or of hydrothermal activity for gold mineralisation during partial dissolution (see Figure 3). the Hunter-Bowen Orogeny (265-250 Ma), particularly C-depletion indicated in two of the Stage 2 calcites as the intrusion post-dates D folding (Collins 1991). (JGC8, JGC13; Figure 6) suggests a contribution of Recent U/Pb dating of zircons from the granodiorite organic carbon to the fluids at this stage of the (Kimbrough et al 1993) yields an average age of paragenesis. Since carbonaceous sediments are impor- 281 Ma, taken as the age of early crystallisation of the tant host rocks to the deposits and gold deposition pluton. The zircon and biotite ages appear compatible if takes place at Copeland early in Stage 2, the role of typical pluton cooling rates are considered (20-40°C per organic matter appears significant in gold precipitation. million years: Mattinson 1978), which suggest that Substantial C-enrichment is evident in sample JGC 14 hydrothermal activity developed late in the cooling (5 C = 9.0%o from replicate analyses). This value is history. among the highest noted for geological environments Although the Barrington Tops Granodiorite is a (Field & Fifarek 1985). The calcite may originate from candidate for the source of lead and sulfur in the dissolution of pre-existing sedimentary carbonate and Copeland goldfield, it originated as a hot (>1000°C) and reprecipitation from a partially oxidised fluid containing dry (<0.2% H 0) magma (Hensen & Shaw 1990). both CH and C0 species. Considering the anhydrous nature of the intrusion and the distance to the margin of the pluton, the Barrington Lead isotopes Tops Granodiorite is unlikely to be the immediate Lead-isotope analyses were used to test the conclusion source of the ore fluids. The existence of a generalised drawn from the sulfur- and oxygen-isotope studies that vitrinite reflectance 'low' between Copeland and the the ore fluids have, at least in part, a magmatic signature granitoid (Gibson & Seccombe 1993) with R values and that the sulfur is of mantle origin. It was also ranging from 3.0% to 4.5%, equivalent to a temperature envisaged that a comparison of lead-isotope data with differential of 180°C, supports the view that the other published results for eastern Australia might assist granodiorite is not the direct source of fluids. Finally, in identifying the possible source of the magmatic if the granitoid had acted as a heat engine, substantial isotopic variation should be encountered as a result fluids. major meteoric water circulation initiated in the Mean lead-isotope compositions for pyrite crystals of country rocks. For these reasons we suggest that from altered wall rock and pyrite crystals obtained from the hydrothermal fluids emanated from a local quartz veins appear identical (Table 7). We conclude of similar composition to the Barrington Tops that there is a common source of lead for both varieties pluton Granodiorite, which we envisage to have intruded of pyrite. beneath the Copeland district (Figure 8). The field for the lead-isotopic composition of the The geometry of the auriferous veins at Copeland fits Copeland samples (Figure 7) lies well below the crustal with a transpressive style of deformation as discussed growth curve of Cumming and Richards (1975), by Ramsay and Huber (1987). Deformation appears to suggesting a strong mantle contribution to the data. have developed a framework of sinistral shear, A comparison of the datasets confirms that the lead between two majorinstrike-slip the Boonara Fault, signature for pyrite from Copeland fits closely a splay of the Peel-Manningfaults, Fault System, and an the results obtained from the Barrington Tops unnamed sinistral fault in the southwest of the map area Granodiorite (Hensen et al 1992), which at its closest, (Figures 1, 8). Structures developed between two crops out some 20 km to the west-southwest of faults support this contention. These featuresthese Copeland. These signatures are quite different from (i) the Devonian horst-block and adjacent include: graben other intrusive bodies in the New England Fold Belt structures, all of which are outlined by sharp changes in (e.g. the Highland complex of the New England across the boundary faults (Figure 4); Batholith or the Attunga complex; Figure 7). Hensen R(ii) values north-northeast-trending fault-folding (F ) adjacent et al (1992) inferred from the lead-isotope signature of to both boundary faults but especially well-developed the Barrington Tops Granodiorite, together with south of the Boonara Fault; and (iii) anticlockwise Sr/ Sr ratio of 0.70375 and high e (+5.6 to 7.8) rotation of the inlier as evidenced the rotated Fj fold that the parental magma had mantle precursors or axis. Within this regime, conjugatebyriedel could developed from high degrees of partial melting of be expected to develop, as is suggested shearing by the major primitive crust. 18

13

18

18

13

]

13

13

4

2

2

r

r

4

87

86

Nd

284

J. H. G I B S O N A N D P. K. S E C C O M B E

Table 7 Lead isotope data for the Copeland goldfield. Mine Bromley Golden Consuls Golden Crystal Golden Crystal Golden Spur Kin San Kin San Mint Mint Mt Maid Shaft & Tunnel Mt Maid Shaft & Tunnel Mt Maid Shaft & Tunnel Prince Llewellen Rainbow Rose & Thistle Barrington Tops Granodiorite

Sample no.*

Sample Mineral type

208p /205p b

b

207p /206p b

b

206 /204 p b

p b

207p /204 b

pb

208p /204 b

pb

Pb (PP ; m

GL-20 wallrock pyrite 2.0706 GL-14 vein pyrite 2.0714 GL-10 stockwork arsenopyrite 2.0684 GL-10R stockwork arsenopyrite 2.0706 GL-15 vein pyrite 2.0714 GL-18 wallrock pyrite 2.0692 GL-18R wallrock pyrite 2.0699 GL-16 vein pyrite 2.0725 GL-16R vein pyrite 2.0740

0.8430 0.8428 0.8408 0.8414 0.8429 0.8427 0.8427 0.8442 0.8444

18.485 18.475 18.548 18.553 18.474 18.458 18.468 18.436 18.450

15.582 15.570 15.594 15.611 15.572 15.554 15.564 15.564 15.579

38.275 38.269 38.365 38.416 38.267 38.195 38.227 38.209 38.265

4 806 19 24 25 197 197 62 72

GL-11

vein

pyrite

2.0741

0.8445

18.437

15.570

38.240

8

GL-12

wallrock

pyrite

2.0733

0.8442

18.451

15.575

38.254

17

GL-19 GL-8 GL-9 GL-1 GL-17

wallrock pyrite 2.0747 vein pyrite 2.0709 stockwork pyrite 2.0723 vein pyrite 2.0741 microdiorite whole rock 2.1033 dyke

0.8442 0.8428 0.8438 0.8447 0.8592

18.468 18.480 18.439 18.432 18.185

15.591 15.576 15.559 15.570 15.624

38.317 38.270 38.210 38.229 38.249

193 69 285 71 22

* R, repeat chemistry. Pb contents determined by isotope dilution.

shearing along the Copeland fracture zone (Figures 1, 8) and the vein geometry as depicted in Figure 2. Mineral assemblages and paragenesis suggest that vein formation developed from a sulfur-poor, reduced hydrothermal fluid, and that mineral deposition took place in two main stages under the influence of a declining temperature regime (Figure 3). A peak temperature of 350°C for paragenetic Stage 1 is indicated by reflectance data for dispersed organic matter particles in wall rock; a minimum temperature of 220°C for this stage is estimated from arsenopyrite solubility considerations (Heinrich & Eadington 1986). Although quartz and calcite are common gangue minerals in Stage 1, textural evidence and a contrast in equilibrium 5 O compositions between quartz and calcite data suggest that the two minerals were not deposited simultaneously. Arsenopyrite is the dominant opaque mineral deposited in Stage 1; pyrite and gold are present only as minor phases. Temperatures applicable to paragenetic Stage 2 are constrained by T data near 200°C for primary inclusions in late-stage quartz, and T in the range 149— 181°C for secondary fluid inclusions within healed fractures in the same quartz (Table 3). Pyrite is the principal sulfide at this stage and commonly forms cubes which contain small inclusions of pyrrhotite. The bulk of the gold is clearly overprinting and formed during Stage 2 as fracture fill within arsenopyrite and late pyrite, and as coatings on vuggy quartz and calcite. 18

h

h

fluid

Factors including, vein texture and mineralogy, isotopic evidence, temperature data and the controls on vein development exercised by brittle structures, all suggest that the Copeland deposits have affinities with low-sulfur, epithermal systems in sedimentary environments (Bagby & Berger 1985). There is no indication that the veins formed close to the surface; crustal depths of 2-4 km seem compatible with temperature data, vein textures, the lack of substantial alteration and comparisons with shear-zone hosted, epithermal-style mineralisation developed elsewhere (e.g. Mt Aubrey, New South Wales; Hopf & Andrew 1992). Figure 8 summarises the model, whereby fluids of essentially magmatic origin, associated with the emplacement at shallow crustal levels of a pluton cogenetic with the Barrington Tops Granodiorite, utilise fractures opened in response to movement on the Peel-Manning Fault System and precipitate gangue minerals and gold in the temperature interval 350-150°C in secondary structures adjacent to domed, carbonaceous sediments. Late-stage fluids are influenced by meteoric water and isotopic exchange with organic carbon and sedimentary carbonate. CONTROLS ON GOLD DEPOSITION

Gold deposition within the Copeland district appears to be constrained by the degree of coalification attained by dispersed organic matter particles within the

MINERALISATION, COPELAND GOLDFIELD, NSW carbonaceous metasediments. Gold is only deposited within veins in which dispersed organic matter particles have attained anthracite rank, or R values above 4.9% (Table 1), equivalent to a total carbon content of >94% and a volatile component (principally hydrogen) of approximately 6% (Diessel 1992). At this composition, graphitisation has commenced as evidenced by mottled, high bireflectance. For the depositional sequence envisaged in Stage 1 and in the temperature range from 350-220°C, transient rupture of shears and secondary fractures would lead to decompression and phase separation (Sibson 1992). Under these conditions steam would react with graphitic metasediments to form hydrogen gas and carbon monoxide (Bowdon 1948): v

C + H 0 = C0 + H 2

285 (1)

2

CO may react further with steam in the presence of a catalyst such as iron oxides (Bowdon 1948) to produce more hydrogen: CO + H 0 = C 0 + H (2) 2

2

2

Arsenopyrite (and minor pyrrhotite) would form at this stage, since reduction of the fluids should destabilise the H As0 arsenic complex (Heinrich & Eadington 1986). Sulfidation of greywacke and siltstone wall rocks in the fluids would consume H S from the fluids and produce pyrrhotite as a precursor to 3

3

2

standard precision

15.65-

.Q Attunga—

n

Highlands

15.55-

Barrington Tops. x

mullock

• vein

18.25

_L

18.35

18.55

18.45

206ni^ Pb//204rPb

standard precision

38.5

n

OL x y

_Q Figure 7 Lead-isotope ratio QL

plots giving data for sulfides from veins and wall rock (mullock) at Copeland and comparative data for intrusions from the New England Batholith (Barrington Tops Granodiorite, Attunga, Highlands; Hensen et al. 1992). Crustal growth curve (dashed line) from Cumming & Richards (1975).

mullock

38.1

• vein

18.25

18.35

18.45

18.55

286

J. H. G I B S O N A N D P. K. S E C C O M B E Figure 8 Schematic west-east cross-section through the Copeland region at the time of emplacement of a shallow crustal pluton, comagmatic with the Barrington Tops Granodiorite. Fluids of magmatic origin rise to shallow crustal levels on north-northwest-trending fracture zones established by left-lateral movement on the Peel-Manning Fault System. Quartz-carbonate veins are established in subsidiary shears and gold precipitates below 325°C in response to decompression and reaction with carbonaceous zones in the Devonian strata.

2.5

km

arsenopyrite formation: Fe + H S = FeS + 2H

CONCLUSIONS

(3)

By integrating the results of a mineralogical and geochemical study of the Copeland goldfield in the The limited amount of pyrrhotite likely to be southern New England Fold Belt of New South Wales, produced from S-deficient fluids by this process will we demonstrate that: react with arsenic complexes and H generated from (1) Gold-bearing quartz-carbonate veins, confined to equations (1) and (2) to form arsenopyrite: a Late Devonian sequence of low metamorphic-grade sandstones and carbonaceous siltstones, developed in a 2FeS + 2 H A S 0 + 3 H = 2FeAsS + 6 H 0 (4) transpressive regime during left-lateral movement on the Peel-Manning Fault System during the HunterThese reactions could account for the paucity of Bowen Orogeny (265-250 Ma). The orientation of the pyrite and gold in this early stage of deposition, principal veins (070°) fits the direction for R riedel since the formation of arsenopyrite consumes both shears in a shear couple accomplished by the Boonara hydrogen, a potential reductant for gold precipita- Fault segment of the Peel-Manning Fault System in the tion, and pyrrhotite, a precursor to pyrite formation northeast of the Copeland district and an unnamed fault at temperatures near 250°C (Schoonen & Barnes in the extreme southwest corner of the area. 1991). (2) Vein evolution follows two major stages. During During Stage 2, arsenopyrite did not form, because Stage 1, early deposition of quartz, arsenopyrite and As-transport is inadequate within the estimated minor pyrite and chalcopyrite, is followed by a more temperature range of 220-150°C applicable for this complex mineralogy comprising quartz-calcitestage (Heinrich & Eadington 1986). However, most arsenopyrite with minor pyrite, chalcopyrite, sericite gold, associated with minor pyrite, formed early in and gold. The major period of gold deposition occurs Stage 2. early in Stage 2, when pyrite, rather than arsenopyrite, Hayashi and Ohmoto (1991) have shown that, for a is the principal sulfide, and quartz or calcite is the major hydrothermal system in which the dominant sulfide gangue mineral. assemblage is pyrite-pyrrhotite-arsenopyrite, the most (3) Mean random reflectance for dispersed organic probable aqueous gold species is HAu(HS) . Their matter in the veins and host rocks highlights sites of experimental work has also established that the gold hydrothermal fluid flow in the district and regions of bisulfide complex is destabilised by the production of major gold production. Temperatures calculated from hydrogen, as in: vitrinite reflectance data using the empirical relationship established by Barker and Goldstein (1990) range (5) from 351° to 361°C for the most productive deposits, HAu(HS) + }H = Au + 2H S corresponding to R values in the range from 4.9% to Since hydrogen is not consumed by the production 5.3%, respectively. The deposits are confined to a northof arsenopyrite (reaction 4 above), gold precipitation northwest-trending zone of high R values, centred on is enhanced during Stage 2. Additionally, because the axial trace of the regional anticline developed in the pyrrhotite is not lost to the arsenopyrite reaction, Copeland inlier. pyrite becomes a stable phase, by progression along (4) A restricted range in 5 S values of-1.8 to 3.2%o the low-temperature pathway involving mackinawite and -1.8 to 2.7%o for sulfides from veins and wallrock, pyrrhotite -> pyrite (Schoonen & Barnes 1991). respectively, implies a single sulfur source of magmatic 2+

2

+

2

3

3

2

2

2

2

2(aq)

2(aq)

(s)

2

(aq)

r

r

34

MINERALISATION, COPELAND GOLDFIELD, NSW origin for district-scale mineralisation. Whole-rock S-isotope analysis of samples taken distant from the mineralised veins (mean 5 3 4 S = - 7 . 4 % 0 ; range - 8 . 1 to -6.5%o) represent sequences generally unaffected by hydrothermal activity. However, SHRIMP II ion microprobe analysis of five separate pyrite framboids from one of these samples (8 34 S = - 3 7 to - 1 6 % o ; mean 34 - 3 0 % o ) indicates that the whole-rock 5 S isotopic value of-6.5%o determined by conventional analysis is a mix of biogenic sulfur derived from framboidal pyrite and a hydrothermal sulfur component with an isotopic composition near 0%o. (5) The 5 1 8 0 composition of Stage 1 quartz from four deposits has a very restricted range (17.2 to 18.0%o). Stage 1 calcites have 5 1 8 0 values in the range 12.6 to 15.7%o and 5 1 3 C compositions ranging from -1.2 to -4.7%o. Using temperatures derived from vitrinite reflectance and a limited fluid-inclusion study, calculated 5 1 8 O f l u i d values near 5%o (200°C) or 8%o (250°C) are consistent with a magmatic source of fluid. Isotopic depletion in both 1 8 0 and 13 C ranging from about 3 to 7%o for Stage 2 calcite from the Mountain Maid deposit is consistent with entry of meteoric water to the hydrothermal system and the role of organic carbon in enhancing gold deposition. Mean lead-isotope compositions for sulfides extracted from veins and altered wall rock are identical, have a strong mantle component and suggest a genetic link to a magmatic source cogenetic with the Barrington Tops Granodiorite. (6) Vein texture, mineralogy, isotopic evidence, temperature data and the controls on vein development exercised by brittle structures, all suggest that the Copeland deposits have affinities with low-sulfur, epithermal systems in sedimentary environments. A crustal depth of 2-4 km is compatible with temperature data, vein textures, the lack of substantial alteration and comparisons with shear-zone hosted, epithermalstyle mineralisation developed elsewhere. Mineralisation developed from fluids of essentially magmatic origin, associated with the emplacement at shallow crustal levels of a pluton cogenetic with the Barrington Tops Granodiorite. Gangue minerals and gold precipitated over the temperature interval 350-150°C in secondary structures that opened in response to movement on the Peel-Manning Fault System, adjacent to domed, carbonaceous sediments. Late-stage fluids are influenced by meteoric water and isotopic exchange with organic carbon and sedimentary carbonate.

287

REFERENCES ARKAI P. 1991. Chlorite crystallinity: an empirical approach and correlation with illite crystallinity, coal rank and mineral facies as exemplified by Palaeozoic and Mesozoic rocks of northeast Hungary. Journal of Metamorphic Geology 9, 723-734. BAGBY W. C. & BERGER B. R. 1985. Geologic characteristics

of sediment-hosted, disseminated precious-metal deposits in the western United States. In: Berger B. R. & Bethke P. M. eds. Geology and Geochemistry of Epithermal Systems, pp. 169-202. Reviews in Economic Geology 2. BARKER C. E. & GOLDSTEIN R. H. 1990. F l u i d - i n c l u s i o n

technique for determining maximum temperature in calcite and its comparison to the vitrinite reflectance geothermometer. Geology 18, 1003—1006. BARTLETT W . L., FREIDMAN M . & LOGAN J. M .

1981.

Experimental folding and faulting of rocks under confining pressure. Part IX. Wrench faults in limestone layers. Tectonophysics 79, 255—277. BLAKE M. C. & MURCHEY B. L. 1988. A California model for

the New England fold belt. In: Kleeman J. D. ed. New England Orogen—Tectonics and Metallogenesis, pp. 2031. Department of Geology and Geophysics, University of New England, Armidale. BOSTIK N. H. 1971. Thermal alteration of clastic organic particles as an indicator of contact and burial metamorphism in sedimentary rocks. Geoscience and Man 3, 83-93.

BOWDON A. 1948. A Handbook of Chemistry. Shakespeare Head Press, Sydney. CAMERON R. G. 1981. The geology of the Golden Spur Northern mine, Upper Bowman, near Gloucester, NSW. Geological Survey of New South Wales Report GS1981 / 506. COLLINS W. J. 1991. A reassessment of the 'Hunter-Bowen Orogeny': tectonic implications for the southern New England Fold Belt. Australian Journal of Earth Sciences 38, 409-423. COLLINS W . J., OFFLER R., FARRELL T. R. & LANDENBERGER

B. 1993. A revised Late Palaeozoic-Early Mesozoic tectonic history for the southern New England Fold Belt. In: Flood P. G & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 69-84. Department of Geology and Geophysics, University of New England, Armidale. COOPER J. A., RICHARDS J. R. & WEBB A . W . 1963. S o m e

potassium-argon ages in New England, New South Wales. Journal of the Geological Society of Australia 10, 313316.

CUMMING G. L. & RICHARDS J. R. 1975. Ore lead isotope

ACKNOWLEDGMENTS The work presented here forms part of the MSc research program of JHG at the University of Newcastle. We thank Anita Andrew, Judy Dean and Graham Carr, CSIRO, North Ryde, for assistance with isotopic analyses. Hope Nesmith prepared the diagrams to a tight deadline. The paper benefited from discussions with Claus Diessel and Zhiyu Jiang. PKS acknowledges financial assistance for the project through grants from the Research Management Committee, University of Newcastle.

ratios in a continuously changing Earth. Earth and Planetary Science Letters 28, 155—171. DIESSEL C. F. K. 1992. Coal-bearing Depositional Systems. Springer-Verlag, Berlin. DIESSEL C. F. K. & OFFLER R. 1975. Change in physical

properties of coalified and graphitised phytoclasts with grade of metamorphism. Neues Jahrbuch fur Mineralogie und Monatshefte 1, 11-26. DIESSEL C. F. K., BROTHERS R. N . & BLACK P. M . 1978.

Coalification and graphitization in high-pressure schists in New Caledonia. Contributions to Mineralogy and Petrology 68, 63-78.

288

J. H. GIBSON AND P. K. SECCOMBE

DOWLING K. & MORRISON G. 1989. Application of quartz

textures to the classification of gold deposits using north Queensland examples. In: Keays R. R., Ramsay W. R. H. & Groves D. I. eds. The Geology of Gold Deposits: The Perspective in J988, pp. 342-355. Economic Geology Monograph 6. DUBA D. & WILLIAM-JONES A . E. 1983. T h e a p p l i c a t i o n o f

illite crystal linity, organic matter reflectance, and isotopic techniques to mineral exploration: a case study in southwestern Gaspe, Quebec. Economic Geology 7 8 , 1350-1363. FEINSTEIN S., WILLIAMS G . H., SNOWDON L. R., GOODARZI F.

& GENTZIS T. 1991. Thermal maturation of organic matter in the Middle Devonian to Tertiary section, Fort Norman area (central Mackenzie Plain). Canadian Journal of Earth Sciences 28, 1009-1018. FIELD C. W. & FIFAREK R. H. 1985. Light stable-isotope

systematics in the epithermal environment. In: Berger B. R. & Bethke P. M. eds. Geology and Geochemistry of Epithermal Systems, pp. 99-128. Reviews in Economic Geology 2. FLEUTY M. J. 1964. The description of folds. Proceedings of the Geological Association of London 75, 461^192. FRIEDMAN I. & O'NEIL J. R. 1977. Compilation of stable

isotope fractionation factors of geochemical interest. In: Fleischer M. ed. Data of Geochemistry. Sixth edition, Chapter KK. US Geological Survey Professional Paper 440-KK. GIBSON J. H. 1993. The use of dispersed organic matter in the investigation of hydrothermal fluids within the Copeland goldfields. In: Diessel C. F. K. ed. Proceedings of the Newcastle Symposium, Advances in the Study of the Sydney Basin, pp. 205—212. University of Newcastle, Newcastle. GIBSON J. H. & SECCOMBE P. K. 1993. A geological synthesis

of the Copeland gold district, southern New England Fold Belt. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 467-478. Department of Geology and Geophysics, University of New England, Armidale. GIBSON J. H. & SECCOMBE P. K . 1 9 9 4 . M i n e r a l i s a t i o n a n d

controls on gold deposition, Copeland goldfield. In: Diessel C. F. K. & Boyd R. L. eds. Proceedings of the Newcastle Symposium, Advances in the Study of the Sydney Basin, pp. 341-351. University of Newcastle, Newcastle. GILLIGAN L . B . & B R O W N L O W

J. W . 1 9 8 7 .

Tamworth-

Hastings 1:250 000 Metallogenic Map, Mineral Deposit Data Sheets and Metallogenic Study. Geological Survey of New South Wales, Sydney.

pp. 88-90. CSIRO Mineral Research Laboratories, North Ryde. HENSEN B. J., ZHOUB., SHAW S. E. & CARR G. R . 1992. L e a d

isotopic constraints on mantle contributions to granitoids of the New England Batholith. Centre for Isotope Studies Research Report 1991-1992, pp. 129-130. CSIRO Mineral Research Laboratories, North Ryde. HOPFS. & ANDREW A. S. 1992. Stable isotope signature of the Mount Aubrey epithermal system, NSW. Centre for Isotope Studies Research Report 1991-1992, pp. 33-39. CSIRO Mineral Research Laboratories, North Ryde. KENNY E. J. 1935. Gold mines and prospects, Commonwealth Creek, Gloucester District. New South Wales Department of Mines Annual Report for 1934, 81-83. KERR P. K. 1977. Optical Mineralogy. McGraw-Hill Inc., New York.

Fourth edition.

KIMBROUGH D. L., CROSS K. C. & KORSCH R . J. 1993. U - P b

isotopic ages for zircons from the Pola Fogal and Nundle granite suites, southern New England Orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 403-412. Department of Geology and Geophysics, University of New England, Armidale. KISCH H. J. 1991. Development of slaty cleavage and degree of very-low-grade metamorphism: a review. Journal of Metamorphic Geology 9, 735-750. MACNEVIN A. A. 1977. Diamonds in New South Wales. Geological Survey of New South Wales Mineral Resources 42. MATTINSON J. M. 1978. Age, origin and thermal histories of some plutonic rocks from the Salinian Block of California. Contributions to Mineralogy and Petrology 67, 233-245. OFFLERR. & WILLIAMS A. J. 1987. Evidence for sinistral

movement on the Peel Fault System in serpentinites, Glenrock Station. In: Leitch E. C. & Scheibner E. eds. Terrane Accretion and Orogenic Belts, pp. 141—151. American Geophysical Union Geodynamics Series 19. OFFLERR., O'HANLEY D. S. & LENNOX P. G . 1989. K i n e m a t i c

indicators in serpentinites: the Peel-Manning Fault System. Geological Society of Australia Abstracts 24, 110-111. OHMOTO H. 1986. Stable isotope geochemistry of ore deposits. Reviews in Mineralogy 16,491-559. OHMOTO H. & RYE R. O. 1979. Isotopes of sulfur and carbon. In: Barnes H. L. ed. Geochemistry of Hydrothermal Ore Deposits. Second Edition, pp. 509-567. John Wiley & Sons, New York. OKUYAMA-KUSUNOSE Y . & ITAYA I. 1987. M e t a m o r p h i s m o f

HAYASHI K.-I. & OHMOTO H. 1991. Solubility of gold in

carbonaceous material in the Tono contact aureole, Kitakami Mountains, Japan. Journal of Metamorphic

NaCl- and H 2 S bearing solutions at 250-350°C. Geochimica et Cosmochimica Acta 1,2111—2126.

O'NEIL J. R. & CHAPPELL B. W. 1977. Oxygen and hydrogen

HEINRICHC. A . & EADINGTON P. J. 1986. T h e r m o d y n a m i c

predictions of the hydrothermal chemistry of arsenic, and their significance for the paragenetic sequence of some cassiterite-arsenopyrite-base metal sulfide deposits. Economic Geology 81, 511—529. HENSEL H . - D . , MCCULLOCH M . T . & CHAPPELL B. W . 1985.

Geology 5, 121-139.

isotope relations in the Berridale batholith. Journal of the Geological Society of London 133, 559-571. O ' N E I L J. R., SHAW S. E. & FLOOD R. H. 1977. O x y g e n a n d

hydrogen isotope composition as indicators of granite genesis in the New England Batholith, Australia. Contributions to Mineralogy and Petrology 62, 313-328.

The New England Batholith: constraints on its derivation from Nd and Sr isotopic studies of granites and country rocks. Geochimica et Cosmochimica Acta 49, 369-384.

OSBOURNE I. 1975. Annual Report Compilation 18, Copeland Division-Hastings Sheet 1876—1973. Department of Mineral Resources, Sydney.

HENSEN B. J. & SHAW S. E. 1990. Isotopic study of pyroxenebearing granitoids of possible mantle derivation. Centre for Isotope Studies Research Report 1989-1990,

RADTKE A . S., RYE R. O . & DICKSON F. W . 1980. G e o l o g y

and stable isotope studies of the Carlin gold deposit, Nevada. Economic Geology 75, 641-672.

MINERALISATION, COPELAND GOLDFIELD, NSW RAMSAY J. G . & H U B E R M . I. 1 9 8 7 . The

Techniques

of

Modern Structural Geology. Volume 2 Folds Fractures. Academic Press, London.

and

ROBERTS J. & ENGEL B. A. 1987. D e p o s i t i o n ^ and tectonic

history of the southern New England Orogen. Australian Journal of Earth Sciences 34, 1-20. RYE R. O. & SAWKINS F. J. 1974. Fluid inclusion and stable isotope studies on the Casapalca Ag-Pb-Zn-Cu deposit, central Andes, Peru. Economic Geology 69, 181-205. RYE R. 0., DOE B. R. & WELLS J. D. 1974 Stable isotope and lead isotope study of the Cortez, Nevada, gold deposit and surrounding area. US Geological Survey Journal of Research

2, 13—23.

SCHOONEN M . A . A . & BARNES H . L . 1 9 9 1 . M e c h a n i s m s o f

pyrite and marcasite formation from solution. III. Hydrothermal processes. Geochimica et Cosmochimica

16.

SIBSON R. H. 1992. Earthquake faulting, induced fluid flow, and fault-hosted gold-quartz mineralization. In: Bartholomew M. J., Hyndman D. W., Mogk D. W. & Mason R. eds. Basement Tectonics 8, Characterization and Comparison of Ancient and Mesozoic Continental Margins, Proceedings of the 8th International Conference on Basement Tectonics, pp. 603-614. Kluwer Academic Publishers, Dordrecht, Netherlands. WHITE M. E. 1986. The Greening of Gondwana. Reed Books, Sydney. WHITE M. E. 1988. Australia's Fossil Plants. Reed Books, Sydney.

Acta 55, 3491-3504.

SHEPPARD S. M. F. 1986. Characterization and isotopic

289

variations in natural waters. In: Valley J. W., Taylor H. P. Jr. & O'Neil J. R. eds. Stable Isotopes, pp. 165-184. Mineralogical Society of America, Reviews in Mineralogy

(Received 15 May 1996; accepted 21 January 1997)

Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia Special Publication 19, 290-299.

Metal contamination at the abandoned Halls Peak massive sulfide deposits, New South Wales B. G. LOTTERMOSER,* P. M. ASHLEY, M. MULLER AND B. D. WHISTLER Division of Earth Sciences, University of New England, Armidale, NSW 2351, Australia.

Small lenses of Zn-Pb-Cu-Ag massive sulfides at Halls Peak in northeastern New South Wales crop out in steep terrain near the eastern edge of the New England plateau. Local relief of 700 m and slopes up to 50°, coupled with an average annual rainfall of about 1000 mm has led to local severe landsliding in previously mined areas. Waters emanating from the main landslide area, from underground workings and draining scattered unconfined stockpiles of sulfide ore (point-pollution sources) have pH values of 3.2 to 4.2 and contain elevated contents of As, Cd, Cu, Fe, Mn, Pb and Zn. Chemical mobility of heavy metals into nearby streams occurs under low pH conditions with Zn exhibiting the greatest dispersion. Increasing pH conditions downstream from the point-pollution sources cause precipitation of Fe phases and coprecipitation of heavy metals. Physical dispersion of secondary metal-bearing minerals from the landslide and ore stockpiles into the surrounding creeks results in elevated levels of Cu, Pb and Zn in the stream sediments thereby exceeding the local and regional background values by one to three orders of magnitude. Areas disturbed by mining are characterised by lack of native vegetation, possible biomagnification of heavy metals into algae and lower plants, and partial recolonisation by metal-tolerant grass species [snow grass (Poa sieberana), blady grass (Lomandra longifolia)]. Key words: acid mine waters, heavy metals, massive sulfide deposits.

INTRODUCTION

Acid mine drainage emanating from base and precious metal mining operations is a difficult environmental problem facing the mining industry in humid to semiarid climatic regions. From numerous acid mine drainage research studies conducted worldwide, it is generally accepted that the five major ingredients for the formation of acid mine drainage are (Kwong 1993): oxygen, water/moisture, sulfide minerals, absence of minerals with significant acid-neutralising capacity, and presence of sulfide-oxidising bacteria. Facilitated by bacterial decomposition, acid is produced when metal sulfides react with oxygen-rich water, forming metal ions, sulfate and hydrogen ions (Kelley & Tuovinen 1988). The generated acidic water is high in total dissolved solids and may seep from the oxidising mineralisation into adjoining groundwater and surfacewater flow systems (Davis et al. 1991; Gray et al. 1994). The formation of secondary minerals, particularly insoluble Fe(III) species, has been observed to produce protective coatings and layers thereby affecting the dissolution of sulfides and reducing the metal concentrations in the leachate. Among the reactions controlling metal ion attenuation, the precipitation of Fe-oxyhydroxide, jarosite and secondary heavy metal minerals (Chapman et al. 1983), the presence of acidneutralising minerals and the prevailing redox conditions appear to be important. Uncontrolled release of acid, metal-bearing waters may result in contamination of soils, stream sediments, floodplains, subsurface and surface waters and cause potential phyto- and zootoxicity.

This paper reports the results of a case study on the environmental impacts of former mining of massive sulfide Zn-Pb-Cu-Ag deposits in the Halls Peak area of northeastern New South Wales (Figure 1). It investigates aspects of the heavy-metal geochemistry of mine waste, soil, water and plant material. Major objectives were to evaluate metal loading in local streams draining the mining area, to elucidate the metal-transport mechanisms and mobility in local streams, to evaluate potential phytotoxicity within sites disturbed by former mining, and to point to potential rehabilitation measures. HALLS PEAK MINING A R E A

The Halls Peak massive sulfide Zn-Pb-Cu—Ag deposits are located in northeastern New South Wales, 500 km north of Sydney and 50 km southeast of Armidale at latitude 30°45'S, longitude 152°02'E. The area lies towards the eastern edge of the New England plateau on the steep upper slopes of the Chandler River valley (Figure 1). The latter is a tributary of the Macleay River and is incised to depths of over 700 m. The region has a humid temperate climate, with an average annual rainfall of about 1000 mm. Most of the region is covered by dry sclerophyll forest, but patches of dry rainforest occur in sheltered gullies. Soils range from skeletal on steeper slopes (where accumulations of scree are common) to well developed on the plateau. Several small lenses of massive sulfide ore were mined by * Present address: School of Earth Sciences, James Cook University, PO Box 6811, Cairns, Qld 4870, Australia.

METAL CONTAMINATION, HALLS PEAK NSW

291

Figure 1 Halls Peak area showing position of the landslide, sulfide stockpile and polluted area, drainage into Chandler River, and sampling sites of soils, landslide materials and stream sediments.

opencut and underground methods between the 1890s and 1960s and several small sulfide ore stockpiles remain. The ore lenses occur in steep country, with slopes locally up to 50° and at altitudes between 500 m and 900 m. In two of the mined areas, landsliding has resulted in large masses of unconsolidated mine waste extending downslope, the larger of which, at Gibson's Mine, has a vertical extent of 120 m. In this area, unconstrained drainage of heavy-metal-bearing acid waters occurs into several ephemeral tributaries of the

Chandler River, with sources including the landslide, underground mine openings and small stockpiles (up to several tonnes) of sulfide ore (Figure 1). At Gibson's Mine, ground disturbed by mining activities covers an area of about 80000 m2 and includes a collapsed opencut and underground workings, access tracks and landslide material (Figure 1). Little vegetation exists in the central part of the site, although encroachment of native grasses and shrubs has occurred at the periphery since mining ceased. Nearby, oxidation

292 B. G. L O T T E R M O S E R ET AL. of small ore stockpiles and resultant acid drainage has led to destruction of eucalypt forest downslope. The area affected by landsliding is about 50 000 m , with at least 100000 m of unconsolidated material on slopes of 15^0° (Figure 1). The slide remains active, especially after major rainfall events, and is composed of material with grainsizes ranging from clay to boulders 2—3 m across. Constituents include sedimentary and felsic volcaniclastic material, minor sulfide and gossan fragments, post-mine oxidation minerals and assorted mining waste. The active nature of landslide material allows access of air and rain, resulting in oxidative weathering which generates acidic metal-rich effluents (Muller 1994; Whistler 1995). 2

3

poitevinite [(Cu,Fe,Zn)S0 H 0] and melanterite (FeS0 7H 0). Such sulfates are well known to form under ambient temperatures in the near vicinity of oxidising pyrite or other sulfides (Williams 1990). They appear to be replaced by less soluble phases such as copiapite [Fe(II)Fe(III) (S0 ) 20H 0], gypsum, jarosite and plumbojarosite with time and increasing exposure to air. 4

4

2

2

4

4 6

2

METHODS Sampling and total element analyses

Over 100 stream sediment, landslide, ore stockpile, soil composite (A, B, C horizons), topsoil (A-horizon), soil LOCAL GEOLOGY B-horizon, gossan and sulfide ore samples were taken in the Halls Peak area. Local background stream sediments The Halls Peak Zn—Pb-Cu-Ag deposits are hosted by were collected upstream of abandoned mining activities. Early Permian felsic volcaniclastic and sedimentary Stream sediment and landslide samples were dried at rocks of the Ngaku terrane in the southern part of the room temperature, crushed with mortar and pestle, and New England Orogen and are considered to be of then sieved through a 180 mm-size nylon gauze (-80 submarine volcanogenic origin (Moody et al. 1993). mesh). Topsoil samples were sieved through a 1 mmPrior to mining, several occurrences of gossan cropped size nylon gauze in order to evaluate the possible out, or were covered by shallow soil (e.g. Khan's Creek transport of windblown material from Gibson's opencut deposit: Palethorpe 1980). Sulfide-rich lenses contained into the local region. Bulk landslide (-10 kg), ore up to several thousand tonnes of ore and ranged from stockpile (-10 kg), gossan, sulfide ore, topsoil and soil massive to banded. They are enclosed in shale and composite samples were dried at room temperature, volcaniclastic siltstone, with nearby diamictite and crushed with a jaw crusher and ground with a chromecoarse felsic volcaniclastic rocks. In the Gibson's Mine steel ringmill. Sample powders (approx. 0.2 g) were area, several stacked, moderately dipping, massive dissolved in a hot HF-HN0 -HC10 acid mixture sulfide lenses are enclosed in a broad hydrothermal (approx. 15 mL), refluxed with the acid mixture if alteration envelope dominated by sericite and quartz required, and analysed in the Division of Earth (± pyrite) (Moody et al. 1993). Massive sulfide ore is Sciences, University of New England by AAS dominated by sphalerite and galena, with subordinate (atomic absorption spectrophotometry) for their total chalcopyrite and pyrite, and traces of arsenopyrite and tetrahedrite. Gossan outcrop and subcrop are charac- Table 1 Average composition of massive sulfide and gossan. terised by relatively immature mineral assemblages ranging from goethite, hematite and plumbojarosite [PbFe (S0 ) (0H) ] [± malachite Cu C0 (0H) and Sulfide Gossan azurite Cu (C0 )2(0H)2] at the surface to goethite, (n = 6) (72 = 8) plumbojarosite, cerussite (PbC0 ), smithsonite (ZnC0 ), malachite, azurite, brochantite [Cu S0 (0H) ], and aurichalcite [(Zn,Cu) (C0 ) (0H) ] at depth. SG 4.4 gem" 2.2 gem" Metal grades at Gibson's Mine were very high (average 16.3% Zn, 18.5% Pb, 3.6% Cu, 356 ppm Ag: Fe 11.1 % 33.2 % Mn Gilligan et al. 1992) and evidently included supergene159 29 Cu 4.49 % 2.09 % enriched material. Representative data were acquired by Pb 10.16% 4.24 % one of the authors (PMA) on sulfide-rich material and Zn 19.5 % 1668 gossan from Gibson's Mine (Table 1). Mass-balance 319 284 Ag calculations reveal that Fe and Bi have been enriched As 973 708 and other metals have been leached during oxidation of Au 520 ppb 220 ppb the sulfide ore to gossan. Ba 860 105 Post-mine oxidation has resulted in an array of Bi 230 1405 secondary sulfate minerals coating surfaces of underCd 531 5 ground openings, ore stockpiles and landslide material. Co 33 38 Although some are forming directly on massive Cr 11 22 Sb 297 54 sulfides, others have been precipitated from surface Mo 39 36 and groundwater, or others are forming from the Ni 28 14 oxidation of pyrite-bearing altered host rock. Many of Sn <5 29 the minerals are soluble in water and redissolve during rain events. The most soluble minerals include goslarite (ZnS0 7H 0), halotrichite [FeAl (S0 ) 22H 0], Values in ppm, except where indicated. 3

6

4

3

4

12

2

3

2

3

3

3

4

5

4

2

4

3 2

4

6

6

3

2

4

2

2

3

METAL CONTAMINATION, HALLS PEAK NSW Ag, Cd, Cu, Fe, Mn, Pb and Zn contents. Duplicate analyses for soils and sediments indicate that the data are reproducible to within ± 5%. Heavy-metal analyses of the ore stockpile, gossan and sulfide ore samples were performed by divisional and commercial AAS (Analabs®, Brisbane). Total element contents were determined in order to evaluate the dispersion of heavy metals into the local drainage system, to compare the composition of the Halls Peak area stream sediments with the chemistry of the primary sulfide ore and gossan, to establish the natural background levels for heavy metals in local soils, to investigate the impact of wind blown topsoil originating from the opencut on the natural geochemistry of local soils, and to calculate the amount of heavy metals present in the residual solids from the sum of the eluates of the sequential extraction analysis.

293

for analysis of total (waters unfiltered) and dissolved metals (waters filtered below 0.45 ^m) and were preserved using HN0 3 . The ambient pH was determined in the field and chemical analyses (As, Cd, Cu, Fe, Mn, Pb, Zn) were performed by AAS and ASV (anodic stripping voltammetry). Biogeochemistry Eight biological samples were taken within and outside the Gibson's Mine landslide and a nearby ore stockpile (Figure 1). Sample materials included snow grass (Poa sieberana), tussock sedge (Carex appressa) and an algae sample (Kidbsormidium rivulare). Following collection, roots were trimmed and the samples were subsequently washed with tap and Millipore® water, dried, ashed at 550°C (grass samples), and dissolved in hot HC1. Solutions were analysed for Cu, Pb and Zn by AAS.

Sequential extraction technique HEAVY-METAL GEOCHEMISTRY Sequential extraction studies provide an insight into the location of elements in different particulate forms. A sequential extraction analysis was performed on the stream sediments and landslide material in order to determine the solid speciation of heavy metals. The analytical procedure follows that of Tessier et al (1979) and Salomons and Forstner (1980) and uses the following sequence of chemical reagents: approximately 10 g sample powder as starting material; 50 mL of 1 M MgCl2 agitated for 1 hour; 20 mL of 0.2 M NH4Ox/ HOx (ammonium oxalate/oxalic acid) agitated for 24 hours; 14 mL of 0.02 M HN0 3 , 8 mL of 30 % H 2 0 2 and 20 mL of H 2 0 for 4 hours at 85°C. The extractants were analysed for Ag, Cd, Cu, Fe, Mn, Pb and Zn by AAS. The wet chemical extraction procedure allowed the differentiation between exchangeable, Fe-Mn oxyhydroxide, oxidisable (sulfides, organic matter), and residual heavy-metal fractions (insoluble sulfates, silicates). The extractants of sequential leaching studies are assumed to remove individual phases of the material selectively, while causing relatively insignificant dissolution of other phases. However, it has also become obvious that these differentiated wet chemistry analyses have several problems (Rendell et al 1980; Tipping et al 1985; Forstner 1986) including: (i) the reactions involved in the leaching procedures are not selective; (ii) readsorption and precipitation processes may occur; and (iii) sample preparation commonly causes transformation of labile phases to more stable components. In spite of these limitations, sequential extraction analysis gives information on the relative speciation of metals within solid products (Belzile et al 1989).

Landslide geochemistry Total heavy-metal concentrations in landslide materials are high (Table 2). Bulk landslide material and its -80 mesh fraction possess similar Pb, Cu and Ag concentrations indicating that these heavy metals are not preferentially incorporated into the finer fraction (Table 2). However, Zn and Cd concentrations of the -80 mesh fraction are higher than those of the bulk landslide material indicating that Zn and Cd are readily mobilised into the finer fraction. The sequential extraction technique shows that in the bulk landslide material Ag, Cd, Cu, Pb and Zn are preferentially associated with the insoluble metal salt fraction (e.g. jarosite-type phases, sulfates) (Figure 2a). The exchangeable and Fe-Mn oxyhydroxide fractions generally carry significantly less amounts of Zn and Cd compared with the sulfide and insoluble metal salt extractable forms. In contrast, the -80 mesh fraction of the landslide material possesses more heavy metals, particularly Zn and Cd, incorporated in the exchangeable fraction and therefore higher percentages of labile heavy metals (Figure 2b). Thus physical erosion and acid sulfate weathering within the landslide is accompanied by the liberation of heavy metals, their subsequent incorporation into the adsorptive fraction of finer particles and the precipitation of insoluble metal salts. Zinc and Cd thereby exhibit the greatest tendency to become liberated from the sulfide minerals, to experience pronounced mobilisation during weathering and to be subsequently adsorped on to finer particles.

Hydrogeochemistry

Ore stockpile geochemistry

Water samples were taken from stagnant pools and mine seepages in the landslide and ore stockpile area and from the local drainage system. Sampling occurred at low-flow conditions during a prolonged period of dry weather (April 1994) and also after periods of average rainfall (March and June 1995). Samples were collected

Several small stockpiles of sulfide ore are located in the Halls Peak area. A stockpile near Gibson's Mine which has resulted in destruction of vegetation downslope was sampled (Figure 1). It consists of lumps of partly oxidised massive sulfides, loosely cemented by finegrained plumbojarosite and minor anglesite (PbS0 4 )

294

B. G. LOTTERMOSER ET

AL.

Table 2 Arithmetic mean and maximum and minimum concentrations for elements in soil composites, landslide materials, ore stockpile, Chandler River, Barkers Creek and Silver Gully sediments, and stream sediments representing local background samples.

Zn

Pb

Cu

Ag

Cd

Mn

Fe

Local soil composites (n = 32) minimum 20 maximum 350 arithmetic mean 82

1 4400 201

5 94

21

<3 111 19

<1.2 2.4 2.1

14 760 168

780 6540 2700

Landslide materials (bulk) (n = 11) minimum 1460 maximum 28600 arithmetic mean 8650

9200 47100 16900

1250 24600 6830

42 157

2.6

101

38 11

156 454 252

7780 63100 37300

Landslide materials (-80 mesh) (n = 9) minimum 3740 maximum 58200 arithmetic mean 14100

3640 21800 12600

4180 10400 6610

39 119 82

5.2 102 23

199 526 350

45700

Ore stockpile R75749

163200

32000

360

178

22 18

na na na

na na na

na na na

na na na

Chandler River sediment (below confluence with Barkers Creek) R76739 751 1040 254

10

<1.2

1040

24400

Barkers Creek sediments (draining landslide) (n = 9) minimum 1000 450 maximum 3650 19000 arithmetic mean 2090 11600

1500 4300 2580

20 68

<1.2 <1.2

45

195 1750 460

32200 45200 36600

Silver Gully sediments (draining stockpile) (n = 5) minimum 950 2040 maximum 7280 8770 arithmetic mean 2870 5200

1590 3600 2460

10 38 22

<1.2 11 3

45 125 78

12300 16500 13700

Dam sediment (draining stockpile) R75754 2750 HP2 1090

2650 1660

na

4 na

168 400

13000 36300

19 52

<3 3

<1.2 <1.2

836 966

18600 21400

86000

Chandler River sediments (regional background) upstream 60 30 upstream 76 23 upstream 74 20

879 1050

Local stream sediment samples (local background) R75756 650 56 R75735 305 81

20

35900

60000

All values given in ppm dry weight; na, not analysed.

(confirmed by X-ray diffraction). A little goethite is also present and efflorescences of goslarite occur in dry weather. The ore stockpile sample (sample R75749, Table 2) contains significantly lower Zn, Cu and Cd and higher Pb concentrations than massive sulfide lumps (Table 1). The stockpile has been at the site since the late 1960s and thus the observed mineralogical and chemical changes are likely due to destruction of sulfides and preferential leaching of Zn, Cu and Cd by recent weathering processes. Nearby soil composite and B-horizon soil samples and stream sediments of Silver

Gully draining the area possess heavy-metal contents two to three orders of magnitude lower than the stockpile sample (Table 2), but still two to three orders of magnitude higher than background samples. Stream sediment geochemistry Heavy metals are found in the adsorptive and residual insoluble metal salt fractions (e.g. jarosite-type phases, sulfates) in sediments downstream of the abandoned stockpile as indicated by the sequential extraction

METAL CONTAMINATION, HALLS P E A K N S W 100

100

Cu Pb Zn Ag Cd Fe Mn

Cu Pb Zn Ag Cd Fe Mn • 0 • •

Figure 2 Calculated proportions of Cu, Pb, Zn, Ag, Cd, Fe and Mn in fractions of (a) bulk landslide materials (n = 6); (b) -80 mesh landslide materials (n = 9); (c) Silver Gully stream sediments (-80 mesh) draining an abandoned ore stockpile (n = 4); and (d) Barkers Creek stream sediments (-80 mesh) draining the landslide (n = 9). Metal concentrations are based on the weight of the starting material (10 g) before extraction.

295

insoluble metal salt fraction organic matter / sulphide fraction Fe-Mn oxyhydroxide fraction adsorptive fraction

100 80

60

% 40 20 0 Cu Pb Zn Cd Ag Fe Mn

technique (Figure 2c) (confirmed by X-ray diffraction). The Fe-Mn oxyhydroxide fraction is generally a minor, yet significant host for the heavy metals (4-28%). Thus a proportion of the mobilised metals is immediately adsorbed or coprecipitated with secondary Fe phases under pH conditions (pH = 4) which are sufficiently high for Fe precipitation as oxyhydroxides and basic sulfates (Robinson 1981). Stream sediments taken in Barkers Creek at distances of up to 1.5 km downstream from Gibson's Mine have elevated metal contents (Table 2). Concentrations of Cu, Pb and Zn exceed regional background values by two to three orders of magnitude and local background values by one to two orders of magnitude (Table 2). Heavy metals are dominantly held in these stream sediments as insoluble metal salt fraction (e.g. jarosite-type phases, sulfates) (Figure 2d) (confirmed by X-ray diffraction). The steep relief, presence of secondary heavy-metal minerals in the landslide and stockpiles, and the relatively high pH of the stream waters (pH = 5.5-7.6), indicate that secondary heavy-metal-bearing minerals such as jarosite-type phases and anglesite reach the

Cu Pb Zn Ag Fe Mn

stream bed by erosion from the abandoned mines, sulfide stockpiles and landslide, rather than precipitating from solution in the water column. Several watercourses close to oxidising sulfidic materials have thin Fe oxide and algae coatings (Kidbsormidium rivulare) and bacterial slimes on their beds suggesting that biotically mediated oxidation of Fe and Mn resulted in precipitation of Fe-Mn oxyhydroxides. In addition, heavy-metal profiles of stream sediments are also characterised by increasing proportions of Cu, Pb, Zn and Ag in the Fe-Mn oxyhydroxide particulate form (5-10%; Figure 2c, d) when compared with the -80 mesh fraction of the landslide (1-7%; Figure 2b). Increasing pH conditions in creeks further downstream are likely to cause Fe precipitation and coprecipitation of heavy metals. Thus as the pH increases in the creeks through dilution and neutralisation of the acid discharge by more alkaline waters and through pH buffering due the reaction with silicate rock/sediment materials, the proportion of dissolved metals that precipitates increases at the expense of the dissolved fraction. 2+

2+

296

B. G. L O T T E R M O S E R ET

AL.

Soil geochemistry Soil composite and topsoil samples were collected along two major ridge traverses, along the plateau and across the ore stockpile (Figure 1). Prevailing wind directions include a strong westerly component (mainly winter) and topsoils taken downwind from Gibson's Mine possess heavy-metal concentrations two to three magnitudes lower than those within the mine. Thus no significant amounts of topsoil are transported from the abandoned mine downwind. Soil composites within the local region generally exhibit a restricted range of metal contents and possess arithmetic mean metal values by up to three orders of magnitude lower than the landslide materials (Table 2). Hydrogeochemistry Mine seepages in contact with sulfides have pH values of 3.2-4.2 and contain elevated As, Cd, Cu, Fe, Mn, Pb and Zn values (up to 50 ng/L As, 4 000 jug/L Pb, 6 350 fig/L Mn, 6 300 ng/L Cd, 25 000 |ig/L Fe, 475 000 jug/L Cu, 1 870 000 ^g/L Zn) (Figure 3). Filtered and unfiltered waters have similar metal concentrations indicating that the metals are dominantly transported in a dissolved form. The acidic leachate is gradually neutralised and reaches neutral pH about 1.5 km downstream from the landslide and before the confluence with the Chandler River. Reduction in metal contents downstream from the discharge points appears to be due to dilution and precipitation and adsorption to solid phases. However, near-neutral stream waters taken at distances of up 3 km from the mining areas remain anomalous in metal contents. For the streams recovering from acidification over a pH range of 3.2 to 8.1, Cd, Cu, Fe, Mn and Pb are removed first from solution, while Zn is most mobile (Figure 3). Seasonal variations (variable rainfall with prolonged periods of dry weather) are the most likely cause for the observed minor changes in pH and metal contents of mine seepages and waters downstream. Pools of mine seepages in contact with sulfides and taken under dry conditions have slightly elevated metal concentrations compared with those taken under wet conditions suggesting that significant evaporation causes concentration changes in stagnant pools. Also, hydrogeochemical investigations on the dam which was constructed during the former mining operations and is now located downstream from an ore stockpile (Figure 1), revealed fluctuations in metal contents and pH values. The dam may represent a wetland which reduces metal concentrations and acidity. In addition, samples obtained from local streams after rainfall events possess increased acidity and metal contents. Abundant secondary minerals (e.g. goslarite, halotrichite, poitevinite, melanterite) are forming presently due to the oxidation of sulfides exposed by mining. They form as efflorescences on the surfaces of mining-related rock debris and ore stockpiles by evaporation of surface waters during dry periods and redissolve during periods of rainfall. These secondary minerals may be partly responsible for the observed

increases in acidity and metal content of Halls Peak mine waters and thus may influence the drainage chemistry. Biogeochemistry Natural vegetation has been largely destroyed in several areas including the landslide, abandoned mines and ore stockpiles. After 25 to 30 years, minor peripheral revegetation is characterised by snow grass, blady grass (Lomandra longifolia), Acacia obtusifolia and Pultenaea sp. Barren areas are the result of several factors including: (i) high levels of dissolved heavy

10

N:43 •o

••

1 mg/L

8

.1

•

o<5

.01 .001

o

3

4

5

6

PH 10000 1000

•

(| IT* too o o (p , A,

7

8

9

local background

N:43 d» o

100

10 mg/L &

d"

°

1

.1

.01

-NITtVA^DQ:

3

4

5

6 7 8 9 local background PH

Figure 3 Total metal concentrations and pH of surface waters in the Halls Peak area (filtered below 0.45 jam).

METAL CONTAMINATION, HALLS P E A K N S W metals and low pH; (ii) lack of soil cover due to erosion and mining; (iii) low soil nutrient abundances; and (iv) slope instability, manifest in landsliding. Snow grass and Pultenaea sp. recolonisation is occurring around the toe of the main landslide below Gibson's Mine. Here, snow grass contains Cu, Pb and Zn contents one to two orders of magnitude higher (Cu, 0.4—4 ppm; Pb, 2-6 ppm; Zn, 2-6 ppm; ashed weight, n = 3) than snow grass on adjacent soils which have background metal values (Cu, 0.1 ppm; Pb, 0.1 ppm; Zn, 0.1 ppm; ashed weight). In the area downslope of the ore stockpile, snow grass and blady grass have colonised the edge of the contaminated zone and tussock sedge grows around the edge of the dam containing water and sediment from the abandoned stockpile (Figure 1; Table 2). Again, Cu, Pb and Zn contents of tussock sedge are one to three orders of magnitude higher (Cu, 11 ppm; Pb, 11 ppm; Zn, 33 ppm; ashed weight) than those from adjacent, metalpoor soils (Cu, 0.1, 0.2 ppm; Pb, 0.01, 0.1 ppm; Zn, 0.2, 11 ppm; ashed weight). In a small stream 100-200 m downslope from the ore stockpile, filamentous green algae {Kidbsormidium rivulare) grow in acid (pH 4.1), metal-bearing water. Kidbsormidium rivulare possesses elevated Cu, Pb and Zn concentrations (Cu, 15 ppm; Pb, 5 ppm; Zn, 21 ppm; dry weight) similar to those plants surviving in the metal-rich soils. Metal concentrations of the alga are higher than those of the water and suggest that biomagnification of heavy metals occurs at the polluted sites. METAL MOBILITY The Halls Peak massive sulfide ores have high concentrations of base metals in the order of Zn > Pb > Cu (Table 1). In contrast, the landslide materials have metal concentrations in the order of Pb > Zn > Cu, whereas the stream sediments draining the landslide and gossans have Pb > Cu > Zn and drainage waters have Zn> Cu> Pb (Tables 1, 2; Figure 3). Such distinctly different relative concentrations of heavy metals in various sample media collected from weathered sulfide ores have been related to the different mobility of heavy metals in surface environments (e.g. at Captains Flat, Sunny Corner and Woodlawn in New South Wales: Chapman et al 1983; Jacobson & Sparksman 1988). At Halls Peak, Zn displays the greatest mobility, whereas Cu and Pb show limited mobility possibly due to formation of insoluble precipitates, coprecipitation and adsorption onto particles. These results are in agreement with those of other workers (Giblin 1978; Mann & Deutscher 1980; Filipek et al 1987; Rampe & Runnells 1989) who demonstrated that Zn remained in solution compared with Fe, Cu and Pb, rather than precipitating or adsorbing to solid phases. POLLUTION AND REHABILITATION Investigations on the occurrence, significance, prevention and remediation of acid mine drainage from mine

297

sites in Australia have received an increasing degree of attention in recent years (e.g. Captains Flat, Sunny Corner, Woodlawn, New South Wales — Chapman et al 1983; Jacobson & Sparksman 1988: Conrad, New South Wales — Brooks & Mcllveen 1988: Mt Lyell, Tasmania — Wood 1991: Brukunga, South Australia — Smith & Hancock 1992: other recent publications in Grundon & Bell 1995). At Halls Peak significant acid mine drainage problems have resulted because sulfidic mine waste is being oxidised in a humid temperate climate zone with 1000 mm annual rainfall. The problem is exacerbated because the surrounding rocks are carbonate-poor clastic sediments and volcanics (no short-term pH buffering capacity), pyrite occurs in massive sulfide ores and alteration halos (high acid generating capacity), fine-grained sulfides are present in ore stockpiles and landslide materials (increased surface area for oxidative processes), and large amounts of sulfide-bound metals are available for surficial oxidation processes. In contrast to many documented mine sites affected by acid mine drainage (Merrington & Alloway 1993), at Halls Peak uncontrolled dynamic landsliding of unconsolidated, sulfidic material prevents the formation of cemented Fe-oxide layers which could moderate fluid flow and oxidative processes. This also allows oxidation of freshly exposed sulfidic material and the incorporation of abundant sulfide-oxidising bacteria due to continuously new supply of acidity, warmth, air and moisture (increased rate of chemical oxidation). Concerns continue because metal-bearing ore stockpile, stream sediment and landslide materials are exposed to oxygenated waters and these materials have long-term potential to generate additional acidand metal-rich waters. Risk assessment of environmental pollutants is frequently based on established lists of critical metal concentrations in soils and waters. These triggerconcentrations are the intervention values above which the sample medium requires restoration. Water quality degradation at Halls Peak is a result of continuous stream acidification, surface water runoff over ore stockpiles, and erosion of acidic landslide material into streams during heavy rain events. The impact of the latter on water chemistry depends on the intensity and duration, but due to their irregular occurrence, effects are difficult to quantify. Mine seepage waters in contact with sulfides contain Cd, Cu, Fe, Mn, Pb and Zn values well in excess of World Health Organisation recommended maximum acceptable concentrations in drinking waters and Australian and New Zealand Environment and Conservation Council water guidelines for domestic use (ANZECC 1992; WHO 1984: Table 3). Cadmium contents exceed Australian and New Zealand Environment and Conservation Council values by up to 1000 times. Lead concentrations of stream sediments, landslide materials and the ore stockpile, and Cd contents of the ore stockpile and some landslide materials from Halls Peak exceed the health investigation level guidelines for contaminated sites proposed by the Australian and New Zealand Environment and Conservation Council and National Health and Medical Research Council

298 B. G. L O T T E R M O S E R ET AL. Table 3 Range of metal concentrations within surface

waters taken from the Halls Peak area. Also shown are World Health Organisation (WHO) 1984 recommended maximum acceptable metal concentrations in drinking waters and Australian and New Zealand Environment and Conservation Council (ANZECC) 1992 water guideline values for domestic use.

Halls Peak (N = 4 3 )

Fe

Mn As Cu

Pb

Zn Cd

< 1 0 - 2 5 000 <10-6350 <5-50 < 1 0 - 4 7 5 000 <5 -4000 1 8 - 1 870 000 < 2 - 6 300

WHO

-

50 —

50 —

5

ANZECC 300 100 50 1000 50 5000 5

Values in ng/L.

(Cd, 20 ppm; Pb, 300 ppm: ANZECC 1992 guidelines). In addition, the heavy-metal abundances at Halls Peak exceed the Dutch and German soil quality assessment criteria and the Canadian and British environmental quality criteria for contaminated sites (listed in: Alloway 1990; Alloway & Ayres 1993). The Halls Peak area retains considerable exploration potential for volcanogenic massive sulfide deposits, but is located in a sensitive area adjacent to a national park. Although it is possible that a future discovery may result in new mining, eventual rehabilitation strategies may have to be undertaken by the State Government or mining lessee. Rehabilitation would be difficult due to the rugged terrain and the construction of wetlands for remediation of the heavy-metal-bearing effluents originating from the major pollutant source, the sulfidic landslide, is impossible. Thus in comparison with other abandoned deposits in the New England area, the abandoned Halls Peak deposits require the following rehabilitation measures. Initial stabilisation of the landslide must include benching and armouring of slopes by large boulders. This would be followed by impervious sealing of the sulfidic mine waste and control of runoff by drains and ponding. Ore stockpiles scattered throughout the area could be removed for processing or sealed by impervious layers. They should not be returned to the collapsed opencut as this would increase the already existing slope instability. Revegetation of stockpile sites, bench tops and slopes should occur using imported topsoil (from the plateau) and metal-tolerant, slowly growing native species such as snow grass, blady grass, Acacia obtusifolia and Pultenaea sp. (Ernst 1996). Selection of native plants has to be confined to grass and shrub species because roots of larger plants, particularly trees, could eventually penetrate the impermeable layer and provide access for oxygenated water to sulfidic material.

CONCLUSIONS In contrast to other documented acid mine drainage sites, at Halls Peak uncontrolled, dynamic landsliding of unconsolidated sulfidic material generates a continuous supply of acidic, metal-rich waters. Acid mine drainage, chemical mobility of As, Cd, Cu, Fe, Mn, Pb and Zn, and associated vegetation anomalies are primarily restricted to ore stockpiles, abandoned mines and the landslide. Erosion and physical transport of secondary heavy-metal-bearing minerals are responsible for elevated metal values in stream sediments of the local drainage system. Increasing pH conditions in creeks as a result of dilution from more alkaline tributaries and groundwaters, and pH buffering due the reaction with silicate rock/ sediment materials cause Fe precipitation, coprecipitation of heavy metals and attenuate the metal contents in waters over 1.5 km. Aqueous heavy-metal transport is largely controlled by pH, with Zn showing the greatest mobility. Phytotoxicity has not been established in the study area, however, vegetation anomalies and possible biomagnification of heavy metals occur at the contaminated sites. Rehabilitation measures will have to include stabilisation of the landslide, impervious sealing or removal of sulfidic material, and revegetation using imported topsoil and metal-tolerant, slowly growing native grass and shrub species. ACKNOWLEDGMENTS Jan Cook (University of New England) devoted much time and patience to the analyses of the stream sediments, soils and plants and the Environmental Chemistry Section of the Mineral Resources Development Laboratory, Lidcombe, New South Wales, performed some of the water analyses. Neva BeresfordSmith (University of New England) and Tim Antwistle (National Herbarium of Victoria) identified the plant specimens. Anita Andrew (CSIRO, Sydney) and Graham Taylor (CSIRO, Adelaide) are thanked for their constructive criticism of the manuscript. REFERENCES ALLOWAY B. J. (ed.) 1990. Heavy Metals in Soils. Blackie,

London.

ALLOWAY B. J. & AYRES D. C. 1993. Chemical Principles

of Environmental Pollution. Blackie, London. ANZECC 1992. Australian water quality guidelines for fresh and marine waters. Australian and New Zealand Environment and Conservation Council. BELZILE N., LECOMTE P. & TESSIER A. 1989. Testing readsorption of trace elements during partial chemical extractions of bottom sediments. Environmental Science & Technology 23, 1015-1020. BROOKS K. A. & MCILVEEN G. R. 1988. The impact of a derelict base metal mine on the aquatic environment. In: The Third International Mine Water Congress Proceedings, pp. 793—802. Melbourne. CHAPMAN B. M., JONES D. R. & JUNG R. F. 1983. Processes controlling metal ion attenuation in acid mine drainage

METAL CONTAMINATION, HALLS PEAK NSW streams. Geochimica et Cosmochimica Acta 47, 1957— 1973. DAVIS A . , OLSEN R . L. & WALKER D . R . 1991. D i s t r i b u t i o n o f

metals between water and entrained sediment in streams impacted by acid mine drainage, Clear Creek, Colorado, U.S.A. Applied Geochemistry 6, 333-348. ERNST W. H. O. 1996. Bioavailability of heavy metals and decontamination of soils by plants. Applied Geochemistry 11, 1 6 3 - 1 6 7 . FILIPEK L. H . , NORDSTROM D . K . & FICKLIN W . H . 1 9 8 7 .

Interaction of acid mine drainage with waters and sediments of West Squaw Creek in the West Shasta mining district, California. Environmental Science & Technology

21, 388-396.

FORSTNER U. 1986. Metal speciation in solid wastes—factors affecting mobility. In: Landner L. ed. Speciation of Metals in Water, Sediment and Soil Systems, pp. 13-41. Lecture Notes in Earth Sciences 11. Springer Verlag, Berlin. GIBLIN A. M. 1978. Experiments to demonstrate mobility of metals in waters near base-metal sulphides. Chemical Geology 23, 215—223. GILLIGAN L. B . , BROWNLOW J. W . , CAMERON R . G .

&

HENLEY H. F. 1992. Dorrigo-Coffs Harbour 1:250 000 Metallogenic Map SHI56-10, SHI56-11: Metallogenic Study and Mineral Deposit Data Sheets. Geological Survey of New South Wales, Sydney. GRAY J. E., COOLBAUGH M . F., PLUMLEE G . S. & ATKINSON

W. W. 1994. Environmental geology of the Summitville mine, Colorado. Economic Geology 89, 2006-2014. GRUNDON N. J. & BELL L. C. (eds.) 1995. Proceedings

of the

Second Australian Acid Mine Drainage Workshop, Charters Towers, Queensland. 28-31 March 1995. Australian Centre for Minesite Rehabilitation Research, Brisbane. JACOBSON G. & SPARKSMAN G. F. 1988. Acid mine drainage at Captains Flat, New South Wales. BMR Journal of Australian Geology & Geophysics 10, 391-393. KELLEY B. C. & TUOVINEN O. H. 1988. M i c r o b i o l o g i c a l

oxidations of minerals in mine tailings. In: Salomons W. and Forstner U. eds. Chemistry and Biology of Solid Waste, pp. 33-53. Springer Verlag, Berlin. KWONG Y. T. J. 1993. Minesite acid rock drainage assessment and prevention — a new challenge for a mining geologist. In: Robertson I., Shaw W., Arnold C. & Lines K. eds. Proceedings of the International Mining Geology Conference, Kalgoorlie-Boulder, Western Australia. 5-8 July 1993, pp. 213—217. Australasian Institute of Mining and Metallurgy, Melbourne. MANN A. W. & DEUTSCHER R. L. 1980. Solution geochemistry

of lead and zinc in water containing carbonate, sulphate and chloride ions. Chemical Geology 29, 293—311.

Permian Halls Peak Volcanics and associated massive sulphide deposits: implications for the southern New England Orogen. In: Flood P. G. & Aitchison J. C. eds. New England Orogen, eastern Australia, pp. 331-336. University of New England, Armidale. MULLER M. 1994. The environmental geochemistry of the Halls Peak VMS deposit, northern NSW. BSc Hons thesis, University of New England, Armidale (unpubl.). PALETHORPE C. 1980. Khan's Creek Cu-Pb-Zn prospect, Nambucca Block, N.S.W. Journal of Geochemical Exploration 12, 320-322. RAMPE J. J. & RUNNELLS D. D. 1989. Contamination of water

and sediment in a desert stream by metals from an abandoned gold mine and mill, Eureka District, Arizona, U.S.A. Applied Geochemistry 4, 445-454. RENDELL P. S., BATLEY G . E. & C A M E R O N A . J.

1980.

Adsorption as a control of metal concentrations in sediment extracts. Environmental Science & Technology 14,314-18.

ROBINSON G. D. 1981. Adsorption of Cu, Zn and Pb near sulfide deposits by hydrous manganese-iron oxide coatings on stream alluvium. Chemical Geology 33, 65—79. SALOMONS W. & FORSTNER U. 1980. Trace element analyses on polluted sediments. Part II: Evaluation of environmental impact. Environmental Technology Letters 1,506-511.

SMITH P. & HANCOCK S. 1992. Brukunga—the acid test. Waste Disposal and Water Management in Australia 19, 3-13. TESSIER A . , CAMPBELL P .

G.

C.

&

BISSON M .

1979.

Sequential extraction procedure for the speciation of particulate trace metals. Analytical Chemistry 51, 844—851. TIPPING E., HETHERINGTONN. B., HILTON J., THOMPSON J. W . , BOWIES E. & HAMILTON-TAYLOR J. 1985. A r t i f a c t s in t h e

use of selective chemical extraction to determine distribution of metals between oxides of manganese and iron. Analytical Chemistry 57, 1944-1946. WHISTLER B. D. 1995. Environmental and exploration geology of the Halls Peak massive sulphide ZnPbCuAg deposit, New South Wales. 3rd year research project report, University of New England, Armidale (unpubl.). WHO 1984. Guidelines for drinking water quality. World Health Organisation. WILLIAMS P. A. 1990. Oxide Zone Geochemistry. Horwood, London.

Ellis

WOOD I. B. 1991. The Mount Lyell Mining and Railway Company Limited—an environmental case study. In: Proceedings of the Environmental Workshop 1991, pp. 203-220. Australian Mining Industry Council, Dickson.

MERRINGTON G . & ALLOWAY B. J. 1993. E n v i r o n m e n t a l sig-

nificance of heterogeneous metal distribution in historical Pb-Zn mine tailings heaps. Transactions of the Institution of Mining and Metallurgy Section A 102, A71-A74.

299

MOODY T. C . , ASHLEY P. M . & FLOOD P. G . 1993. T h e E a r l y

(Received 10 April 1996; accepted 15 July 1996)

Index 197, 212, 226, 272, 290 178, 266

acid mine waters argon-argon dating asbestos

290 242, 254 226

New South Wales

backarc basalts Bowen Basin

148 148 80

Carboniferous Clarence River Supersuite continental accretion Copeland Goldfield copper deposits crustal extension crustal architecture crustal convergence

66 242 1,66 272 109 66 29 52

oblique-slip faults ophiolite ore genesis orogeny

188 197 109 1

palaeogeography Peel-Manning Fault System peperite Permian plate tectonics potassium-argon dating

80 188 109 52, 66, 80, 148, 188 1 254

deep seismic reflection profiling Devonian Dumboy-Gragin Granite

29 178 266

Queensland central southeast

96 80, 109 48

epithermal deposits

272

fold belt fold-thrust belt

1 52

rhyolite Rockhampton-Mackay region rubidium-strontium dating

109 80 254

Gamilaroi terrane geochemistry geology geosyncline Gogango Overfolded Zone gold deposits Great Serpentinite Belt Gympie Province

178 148, 197,212 128, 188 1 80 109,212, 272 197,212 128, 148

Halls Peak heavy metals Hunter-Bowen event

290 290 52

seismic profiles serpentinite shear zones silica-carbonate rocks stable isotopes Stanage Fault Zone Strathmuir Synclinorium stratigraphy structural analysis structural terranes

29 226 96 212 272 96 80 80, 128 161 96, 188

tectonics

lead isotopes lithostratigraphy

272 178

mafic rocks mass-flow deposits massive sulfide deposits metal contamination mineralisation

197 109 290 290 109, 128,212, 254, 272 154 109 96

tectonic evolution tectonic model Texas megafold thermal events The Shacks Mylonite Zone thrust faults tin deposits Triassic trondhjemite

1,29, 52, 66, 96, 128, 226 1, 52, 66 161 161 254 96 96 254 52 109

ultramafic rocks upper Barnard region

197 178, 188

vein deposits vitrinite reflectance volcanism

226 272 109

Woodsreef

226

zircon dating

266

Mole Granite Mt Morgan Au-Cu deposit mylonite neodynium isotopes New England Batholith New England Fold Belt northern southern New England Orogen southern

148 242 52, 66, 80, 96, 197 52, 66, 96 272 1,29, 178, 188, 212, 266 161,242

northeastern

301