Issuu

Geological Society of Australia Inc. Abstracts No. 81

Supercontinents and Earth Evolution Symposium Fremantle, Western Australia

26-30 September 2005 Maritime Museum, Victoria Quay

program and abstracts

hosted by the Tectonics Special Research Centre, the Geological Society of Australia (WA Division) and the Geological Survey of Western Australia in collaboration with the International Geological Correlation Program

Supercontinents and Earth Evolution Symposium Fremantle, Western Australia 26-30 September 2005 Maritime Museum, Victoria Quay Hosted by the Tectonics Special Research Centre the Geological Society of Australia (WA Division) and the Geological Survey of Western Australia in collaboration with the International Geological Correlation Program

Program and Abstracts Geological Society of Australia Inc. - Abstracts, No. 81

PUBLISHED BY PROMACO CONVENTIONS PTY LTD

FIRST PUBLISHED 2005 BY PROMACO CONVENTIONS PTY LTD FOR THE Tectonics Special Research Centre Supercontlnents and Earth Evolution S y m p o s i u m 2005 Edited by Wingate, M.T.D. and Pisarevsky, S.A. This volume is a pre-conference publication of the presentations to be given at the Tectonics Special Research Centre, Supercontinents and Earth Evolution Symposium to be held in Fremantle, Western Australia in September 2005. The contents and any opinions expressed represent the views of the authors only. This publication is copyright. Apart from fair dealing for the purpose of private study, research, criticism or review as permitted under the Copyright Act, no part may be reproduced without prior written permission from the Publisher. Additional copies may be obtained from the Secretariat: Promaco Conventions Pty Ltd PO Box 890 Canning Bridge WESTERN AUSTRALIA 6153 Tel: +61 8 9332 2900 Fax: +61 8 9332 2911 Email: promaco@promaco.com.au

ISBN: 186308 120 8

Suggested Citation W i n g a t e , M . T D . a n d Pisarevsky, S.A. (Editors) 2005. S u p e r c o n t i n e n t s a n d Earth E v o l u t i o n S y m p o s i u m 2005. G e o l o g i c a l S o c i e t y of A u s t r a l i a Inc. A b s t r a c t s 81, 177p.

Contents Progrann Overview Pre-Conference Workshops Social Program General Information Map of Fremantle

Symposium Program........

Poster Presentations Post-Conference Field Trip Abstracts Author Index

175

Program Overview Pre-Conference Workshop 1 Palaeomagnetism for Dummies

Sunday 25 Sept

Pre-Conference Workshop 2 Issues with the Precambrian Time Scale

Social Program Sunday 25 September 2005 Ice-breaker Reception - 1730 Maritime Museum, Victoria Quay

Evening

Registration & Ice-breaker Reception

An Ice-breaker Reception will be held at the Maritinne Museum, Victoria Quay. Join us for a special opportunity to renew old acquaintances or make some new friends.

Monday-Friday 26-30 Sept

Supercontinents and Earth Evolution Symposium

Cost: Included for fulltime delegates. Additional tickets A$33 inc GST

Scientific program will cover supercontinent history, global processes and earth evolution, assembling Australia, orogenic processes and imaging orogens Thursday evening

Conference Dinner

1-6 Oct

Post-Conference Field Trip 1 Albany-Fraser Orogen: Mesoproterozoic collision between the WA and SA cratons

Thursday 29 September 2005 Conference Dinner -1900 Fremantle Sailing Club An optional conference dinner will be held at the Fremantle Sailing Club. This is a time to network with your peers and relax. Cost: All tickets A$60.50 inc GST Dinner cost only, beverage extra.

The following Pre-Conference Workshops were held.

Post-Conference Field Trip

Pre-Conference Workshop 1

The Post-Conference Field Trip to Albany will depart on Saturday 1 October at the conclusion of the Symposium. See page xi for full details.

Sunday 25 September 2005 Palaeomagnetism

for Dummies

Venue: The University of Western Australia Time: 0900 - 1630 This workshop was designed to provide participants a basic understanding of the fundamentals of palaeomagnetism and its tectonic applications, with a practical session illustrating procedures of sampling, data analysis, and construction of palaeogeographic maps. The workshop was conducted by Professor Michael W McElhinny, Professor Rob Van der Voo and researchers at UWA.

Pre-Conference Workshop 2

Symposium Committee Members Kevin Cassidy Peter Cawood Alan Collins Ian Fitzsimons Zheng-Xiang Li

Sunday 25 September 2005 Issues with the Precambrian

Places might still be available if there are last-minute cancellations - enquire at the registration desk.

Time Scale

Venue: Esplanade Hotel, Fremantle Time: 0900 - 1600 The Subcommission on Subdivision and Calibration of the Precambrian Time Scale met for an informal workshop. The aim was to bring together a multidisciplinary and international group of scientists to review new developments and tackle remaining issues with the Precambrian time scale. The Precambrian spans 88 percent of Earth history - yet, we do not have a complete and comprehensive time scale to help analyse, describe, and communicate this part of Earth history.

Sergei Pisarevsky Steve Reddy Ian Tyler Michael Wingate Geoff Wood

III

General Information Location of Sessions Symposium sessions will be held in the Lecture Theatre of the Maritime Museum. Morning and afternoon teas and lunches will be served in the function room on Level amongst the posters. Registration Desk The registration desk w i l l be serviced each day of the conference from 0800 - 1530. Speaker Preparation A l l s p e a k e r s are r e q u e s t e d t o c h e c k t h r o u g h t h e i r presentations, in advance of their timesloL in the lecture theatre. Please check with the registration desk for technical assistance. Dress Code Smart casual is suggested for all conference sessions. Messages A notice board will be in the registration area for messages to delegates. Mobile Telephones As a courtesy to speakers, mobile telephones are to be turned off within the lecture rooms during all sessions. Name Badges Each attendee of the conference will be issued a name badge at registration. The badge is the official pass and must be worn at all times. Delegates not wearing badges may not be admitted to social functions, lunches or conference sessions.

Accommodation H

Fremantle E s p l a n a d e Hotel

B a c k p a c k e r s Inn Freo

D a n u m House

Flying A n g e l C l u b

H a r b o u r Village Quest Apartments

Kilkelly's B & B

Pier 21

Tradewinds

Bus Stop Locations (Free CAT Service) o o e o o

o o o

Fremantle Train Station

South Terrace Fisheries Department

High Street Kings Square

South Terrace Wardie Street

High Street CBC Fremantle

South Terrace Orient Street

Ord Street Frenianfle Park

Marine Terrace Douro Road

Queen Street Shopping

South Terrace Charles Street

Precinct

Fremantle Sailing Club

Centre Beach Street James Street

Marine Terrace Capo D'Orlando

Cheviot Hotel

Drive

Marine Terrace Free parking

Centre

Marine Terrace Free parking

Fishing Boat Harbour

Fremantle Hospital

Maritinfie Museum

South Terrace Wray Avenue

Phillimore Street

Cantonment Street Woo/stores Shopping

® Cappuccino Strip

Fremantle Markets

Marine Terrace

Fremantle Arts & Leisure

Esplanade Hotel

Victoria Quay

M a p prepared by Promaco Geodraft, 2005

Program and Abstracts Additional copies of the abstracts may be purchased f r o m the registration desk during the conference for $27.50 (inclusive of GST). Parking Paid parking is available to delegates in the carpark immediately in front of the Museum. Taxis Swan Taxis 13 1330 Transport Trains depart from the Fremantle Station, a short walk from the M u s e u m approximately every 15 minutes. Bus and timetable information can be obtained by telephoning 13 6213. Shopping Hours Shops are open f r o m 8.30am to 5.30pm, Monday to Friday, 8.30am to 5.00pm on Saturday and from 12.00 noon to 5.00pm on Sunday in Fremantle. Medical Emergency Numbers Fremantle Hospital 9431 3333 Royal Perth Hospital 9224 2244 Sir Charles Gairdner 9346 3333 Dental 9220 5777 Doctor 9328 7111 Pharmacy 9335 9633 Smoking Policy For the comfort and health of all conference attendees, the Maritime Museum is a non-smoking venue.

Symposium Program

NB: The presenting author is shown with the paper title. All authors are listed on the abstracts. The abstracts are included in day program order from page 1. Poster astracts are listed in alphabetical order from page 116. An author index can be found at the back of the book.

Symposium Program Precambrian Geomagnetic Field: Intensity, Morphology and Secular variation Updated Palaeoproterozoic Apparent Polar Wander Paths for Laurentia and Australia and their Implication for the Palaeogeography of Nuna Paleomagnetism of the Lower Waterberg Group of South Africa: Towards a Better Defined Apparent Polar Wander Path for the Paleoproterozoic Kaapvaal Craton ra u

Johanna Salminen

Paleomagnetic and Rock Magnetic Data From V a l a a m Sill, Lake Ladoga, Russian Karelia - Implications for Supercontinent Hudsonland

KEYNOTE: Russell Korsch

Assembling Australia

Mark Barley

Contrasting Tectonic Histories of Australia's Pilbara, Yilgarn and Gawler Cratons: Key Pieces of the Late Archean to Early Palaeoproterozoic Tectonic Puzzle

Catherine Spaggiari

Proterozoic Deformation of the Northwestern Yilgarn Craton, Western Australia

Steve Sheppard

Does the CA 1800 MA Capricorn Orogeny Mark Collision of the Yilgarn and Pilbara Cratons?

Hidekl Masago

Metamorphic Evolution of the Northwestern Capricorn Orogen, Western Australia

MORNING TEA

c •c to

Sandi Occhipinti Keith Sircombe

Mesoproterozoic to Neoproterozoic Reworking in the Capricorn Orogen, Western Australia: Evidence From "^^Ar/^^Ar Dating Reconnaissance Geochronology of the Ashburton Basin, Western Australia

DISCUSSION

Ian Tyler

Tectonic Significance of Detrital Zircon Age Profiles Across Palaeoproterozoic Orogens in the Kimberley Region of Northern Australia

S i m o n Bodorkos

Constraints on Proterozoic Cooling and Exhumation of the Halls Creek Orogen , WA

Ian Scrimgeour

Palaeo - To Mesoproterozoic Tectonics of the Arunta Region, Central Australia

Dot Close

Late Palaeoproterozoic Oblique Accretion of a 1690-1660 Ma Magmatic Arc onto the North Australian Craton

Narelle N e u m a n n

New Shrimp Geochronology for the Western Fold Belt of the Mount Isa Inlier: Developing a 1800-1650 MA Event Framework

Karin Barovich

Links Between Proterozoic Australia: Geochemical Provenance of Paleoproterozoic Metasedimentary Rocks From the Mount Isa Inlier and the C u r n a m o n a Province

n m

AFTERNOON TEA

5 U

Justin Payne

Provenance of Sequences in the Northern Gawler Craton: Re-Assembling Proterozoic Australia and Dispersing the Gawler Craton

Michael S z p u n a r

Creating a Coherent Tectonic Synthesis Between the Gawler Craton and C u r n a m o n a Province

Caroline Forbes

Dating Prograde and Peak Metamorphism Within a Complexly Deformed Terrane Using In Situ U-Pb Monazite Geochronology

Greg Swain

Proterozoic Magmatic Arcs and Oroclines: St Peter Suite, Gawler Craton, S A

Ben Wade

Geochemistry and Provenance of a Mesoproterozoic (1.4Ga) Eastern Musgrave Block Basin: Buddying up to the Belt-Purcell Basin

Heather Howard DISCUSSION POSTER SESSION - Drinks provided END OF DAY

' A Preliminary LIthological and Tectonic Chronology for the West Musgrave J Complex

Symposium Program Wednesday 28 September 2005 KEYNOTE: Paul Hoffmann

Nature and Origin of Neoproterozoic Glacials

Yiefel Jia

Evolution of Atmospheric

Qi >

Lena Evins

Linking Cambrian Volcanism to Marine Anoxia and Mass Extinction: Clues from Sulphur Isotope Geochemistry

and Continental Growth

Kath Grey

Neoproterozoic Correlations, Insights and Impediments

Marie Corkeron

'Cap Carbonates' and Why the Walsh Tillite Is Marlnoan

Tim Raub

High-Resolution Magnetostratigraphy of Australia's Marlnoan Cap Carbonates

KETNOTE: Michael Brown

Metamorphic Patterns in Accretlonary and Colllsional Orogens - Secular Variation in Metamorphic Regimes and Punctuated Tectonic Evolution of Earth

Steven Reddy

Palaeoproterozolc Eclogite Exhumation & Implications for Ancient Tectonic Processes

Craig Buchan Alfred Kroner Alan Smith

Linking Supercontinent Assembly, Subduction Initiation and Accretlonary Orogenesis Accretlonary Growth in the Central Asian Orogenic Belt of Mongolia and Kasakhstan and Comparison with the Arabian -Nubian Shield Emplacement of the Western Tethyan Ophiolites and Atlantic Ocean-Floor Spreading

DISCUSSION

KEYNOTE: David Groves

The Temporal Distribution of Mineral deposits: A Strong Reflection of Tectonic and Lithospheric Evolution

KEYNOTE: Onno Oncken

What Controls Orogeny at a Convergent Margin - The Andean Case

Ramon Carbonell

Crustal Structure of the Uralide Orogen

Kate Selway

Magnetotelluric Imaging of the Central Australian Lithosphere

Walter Mooney

Orogenic Processes of the Northeast Tibetan Plateau, China, From Deep Crustal Seismic Profiling Data

David Gray

Assembly of West Gondwana - Perspective from the Damara Orogen Colllsional Triple Junction

Michael Mints

Crustal Structure of the East-European Craton: Imaging Early Precambrian Accretlonary and Colllsional Orogens

KEYNOTE: Larry Brown

Deep Reflections, Lithospheric Structure And Supercontinent Evolution

y if)

AFTERNOON TEA

DISCUSSION 18:30

POSTER SESSION - Drinks provided END OF DAY

vii

Symposium Program Paleomagnetism and Supercontlnents: Resolving the RodinIa Puzzle Karl Karlstrom

Progressive Proterozoic Development of Southern Laurentia: New Map And Linked Databases

Phillip Schmidt

A Revised Pole For The '^1070(7) Ma Alcurra (Kulgera) Intrusions

Laurl Pesonen

New Paleomagnetic Data of the 1.1 G A Diabases of Arizona - Implications for the Assembly of Rodinia

Natalia Lubnina

Paleomagnetism of Middle Riphean Dykes from Ladoga Lake Region (Northern Karelia)

Valery Vernikovsky

Meso-Neoproterozoic Tectonic and Magmatic Events Along the Western Margin of Siberia: Aspects Related and Not Related to Rodinia A s s e m b l y

Jim Sears

Stratigraphic Tests of Proterozoic SE Siberia - S W Laurentia Connection

S b. oJ P

0) if)

MORNING TEA

KEYNOTE: Richard Hanson

Rodinia: The African Perspective

Joachim Jacobs

From Proto-Kalahari to Greater Kalahari: Significance for the Reconstruction of Rodinia

J5 < L

Richard Hanson

Relations of the Kalahari Craton to Rodinia: Paleomagnetic Evidence

David Cornell

Timing of Collisions and Thermal Events In the '>'1.0 G A Namaqua-Natal Province of Southern Africa

"ro

DISCUSSION

KEYNOTE: Relnhardt Fuck

Rodinia Descendants in South America

S i m o n Johnson

Mesoproterozoic Supra-Subduction Magmatism in the Southern Irumide Belt, Central Southern Africa: Implications for the Congo Craton in Rodinia Reconstructions

Bert De Waele

Episodic Crustal Recycling in the Metacratonic Irumide Belt: Geochemlcal and Isotopic Evidence

Xian-hua Li

The Formation of the South China Block: Evidence from SIbaoan Orogenic Magmatism

Matthew Greentree

Late Mesoproterozoic - Early Neoproterozoic Basin Record of the Sibao Orogenesis in Western South China Block and its Relationship to the A s s e m b l y of Rodinia

Jo-Anne Wartho

A r - A r Dating of 1 Ga Sibao Orgenic Events in South China: the 'Missing Link' In the A s s e m b l y of Rodinia?

AFTERNOON TEA

Glen Phillips

Relicts of A CA. 1100 - 1000 MA Mobile Belt Within Neoproterozoic Passive Margins Attributed to the Break-Up of Rodinia: Evidence From the Ruker Province, East Antarctic Shield

Kamal S h a r m a

A s s e m b l y and Dispersal of Rodinia: Evidences From the Northwestern Indian Shield

Manoj Pandit

A Comprehensive Paleomagnetic Study of the Vindhyan Supergroup, Central India: Preliminary Results

Ananya Biswas

Mesoprterozolc Nallamalai Sedimentation in the Cuddapah Basin: A Link Between India and East Antarctica in the Rodinia Supercontinent

Zheng Xiang LI

Synthesis on the A s s e m b l y and Break-Up History of Rodinia: Results From IGCP 4 4 0

DISCUSSION 17:15

E N D OF D A Y I Conference Dinner at Fremantle Sailing Club

via

Symposium Program

•fo u

Marcelo Martins-Nito

The Sao Francisco Basin, Eastern Brazil, The Record of Rodinia (?) Breakup and Gondwana Assembly

Fernando A l k m i m

Nutcracker Tectonics During the Neoproterozoic A s s e m b l y of West G o n d w a n a : The Development of the Confined Aracuai-West Congo Orogen (Brazil/Africa)

Denny Loose

Pan-African Crustal Thickening and Eclogite Facies Metamorphism in Cameroon

Volker Schenk

Pan-African Reworking of the Northeastern Corner of the Congo Craton in Uganda

Ian Fitzsimons

Detrital Zircon Provenance of Central Madagascar and Neoproterozoic Terrane Transfer Across the Mozambique Ocean

Neils Jons

P-T Evolution of the Bemarivo Belt (Northern Madagascar): The Final A s s e m b l y of Gondwana

Talari Chetty

The Cauvery ShearZone, Southern Granulite Terrain, Southern India: A Crustal-Scale Flower Structure

Masaaki Owada Dan Dunkley Alan Collins

Metamorphic and Magmatic Processes in Dronning Maud Land During the Pan-African Event, East Antarctica In-Situ Shrimp Dating of Zircon in Mlgmatite: Resolving the Pan-African Development of the Lutzow-Holm Complex, East Antarctica The East African Orogen and the A m a l g a m a t i o n of Gondwana - A Palaeogeographic Perspective

DISCUSSION LUNCH

Pradip Dasgupta

The Pan-African Orogeny in Northwestern Himalaya: The Missing Link in the Transition From Rodinia to Gondwana

Songlnian Lu

Evidence From Breakup of Rodinia to A s s e m b l y of Gondwana In China Continent

Chris Kirkland u

Vicky Pease

Detrital Zircon and Intrusive Geochronology of the Kalak Nappe Complex, Arctic Norway: A Neoproterozoic Accretionary Complex A Baltica Provenance for Detrital Zircons from Early Paleozoic Sediments of Southern Novaya Zemlya, Arctic Russia

AFTERNOON TEA & CLOSE

Post Conference Field Trip Albany-Fraser Orogen: Mesoproterozolc Collision Between the WA and S A Cratons

Symposium Posters in alphabetical order Benjamin Bley de Brito Neves

A First Tentative Classification of Rodinia's Descendants in South America

D Brown

Crustal Structure of the Uralide Orogen

Craig Buchan

Linking Supercontinent Assembly, Subduction Initiation and Accretionary Orogenesis

Peter Cawood

Trends in Sedimentary Provenance Along East Laurentia, Baltica and Amazonia During Assembly and Breakup of Rodinia

Alan Collins

Depositional Age and Provenance Record of Ultra-High Temperature Metasediment Protoliths of Southern Madagascar and India

Laura Crossey

In Situ U-TH-PB Microprobe Dating of Early Diagenetic Monazite in Neoproterozoic Black Shales of the Western United States

Huntly Cutten

Tectonics of Eastern Africa: The Mozambique Belt and Its Regional Setting

Michiel De Kock

Paleomagnetism of Borehole GKP-Olof the Agouron-Griqualand Paleoproterozoic Drilling Project, South Africa

Bert De Waele Gabriela Depine David Evans

Volgo-Uralia: SHRIMP Evidence of Strong Palaeo-Proterozoic Reworking of Archaean Crust ; Origin of Gondwana Margin Permo-Jurassic Granite-Rhyolite Provinces: The South American Choiyoi Example Rodina Radically Revised?

Richard Ernst

Lips Through Time

Ian Fitzsimons

Three Stages of Mesoproterozoic Metamorphism During Crustal Shortening in the Albany-Fraser Orogen of SW Australia

George Gibson

Syn- and Post-Extensional Orogenesis in the Palaeoproterozoic Rift Sequences of Broken Hill and Mt Isa and Implications for Reconstructions of Rodinia

Sarah Goss

Unravelling the New England Orocline and Implications for EndPalaeozoic To Early Mesozoic Orogenesis Along the Pacific Margin of Gondwana

David Gray

Assembly of West Gondwana - A perspective from the Damara Orogen collisional triple junction

Niels Jons

Relics of the Mozambique Ocean: Geochemistry of the Vohibory Block (Madagascar)

JR Kayal

Deep Crustal Structure of the Indian Lithosphere: Seismic Imaging

Rajat Mazumder

Continental Freeboard and Geological Evolution: A Precambrian Perspective

Michael Mints

3D Model of Crustal Structure of the Southeastern Fennoscandian Shield: Evidence From Reflection Siesmic and Geological Data

Walter Mooney

How Thick is the Earth's Crust?

Walter Mooney

Cool Cratons and Thermal Blankets: How Continents Affect Mantle Convection

Nobuhiko Nakano

Permian Continental Collision Event in Indochina Region

Sandra Occhipinti

The Palaeoproterozoic To Neoproterozoic Evolution of the Capricorn Orogen, Western Australia: A Summary in Space and Time

Yasuhito Osanai

Regional Division of Metamorphic Processes in the Napier Complex, East Antarctica As A Final Event of Continent Collision During Archean

B S Paliwal

Tectonics of the North Western Indian Peninsular Shield and its Significance

B S Paliwal

Pre-Tertiary Glacial Events in Northwestern India and Their Significance in the Assembly and Breakup of Rodinia and Growth of Asia

Symposium Posters in alphabetical order continued Asa Pettersson

; The Koras Group, South Africa: A Tectono-Stratigraphic Marker for the Namaqua Collision Event?

Robert Ralnbird

; The Basal Paleoproterozoic Cover Sequences of Laurentia: From the Break-Up of Kenorland To the Assembly of Nuna

Timothy Raub

Detrital Zircon Geochronology of the Australian Marinoan Glacial Interval

Volker Schenk

Gondwana Formation: A Metamorphic View From the Margins of the , Congo Craton

Randell Stephenson

Crustal Structure and Phanerozoic Orogenic Processes On and Near the Southern Margin of Baltica

X Wang

S LA-ICPMS U-Pb Zircon Geochronology of Neoproterozoic Igneous Rocks From Northern Guangxi, South China: Implications for Petrogenesis and Tectonic Evolution

Michael Wingate

Two Large Igneous Provinces in Late Mesoproterozoic Australia

Post-Conference Field Trip 1 Saturday 1 October to Thursday 6 October Albany-Fraser Orogen: Mesoproterozoic collision between the WA and SA cratons The Albany-Fraser field trip will concentrate on well exposed coastal outcrops of metasedimentary and metaigneous gneisses of the Biranup and Nornalup Complexes that preserve beautiful examples of intensely deformed high-grade metamorphic rocks that were deformed and metamorphosed in the Mesoproterozoic. We will also examine metasedimentary schists and quartzites of the Palaeo-Mesoproterozoic Mount Barren Group The aim of the trip is to provide an overview of this little-studied orogen and highlight the implications that a better understanding of its tectonic evolution has for unravelling the Mesoproterozoic amalgamation of both Australia and Rodinia. The fact that the region to be traversed contains a World Biosphere Reserve, a number of national parks and is famous in gastronomic circles for boutique wine and seafood will ensure a successful trip. Leaders: Ian Fitzsimons and Craig Buchan (Curtin University of Technology). Trip details: Depart early morning Saturday 1 October from hotels in Fremantle for Wave Rock and Hopetoun. Field trip will run for 6 days 1-6 October ending in Perth on Thursday evening. All meals to lunchtime Thursday 6 October provided. Travel will be by coach and accommodation will be in twin rooms in 2-3 star hotels and motels. Cost: A$803 inc GST (A693 inc GST student rate)

Field Trip 1 Images: Albany Tourist Centre

Field Guides will be provided by the Geological Survey of Western Australia. XI

Notes

Fremantle Images: Fremantle Chamber of Commerce

Xll

Disclaimer The information contained in this publication is correct at the tinne of printing. The connmittee reserves the right to alter or delete items from the program or tours. The organisers shall not be responsible for any costs or damages arising from any action based on the information contained herein. E.&.O.E.

Supercontinents and Earth Evolution Symposium 2005

MONDAY 26 SEPTEMBER 2005

Supercontinents and Earth Evolution Symposium 2005

RODINIA IN 2005 Ian WD Dalziel University of Texas Institute for Geophysics, 4412 Spicewood Springs Road, Building 600, Austin, Texas, 78759-8500, USA (ian@ig.utexas.edu)

Current thinking about pre-Pangea supercontinents began with the work of Gerard Bond, Michelle Konnintz and Peter Nickeson at the then Lamont-Doherty Geological Observatory of Columbia University in the nnid-1980's. 'Backstripping' the late Precambrian-early Paleozoic sedinnentary successions of North and South America that appeared to them to reflect rift-todrift transitions and passive margin subsidence, they concluded that Laurentia, the cratonic core of present-day North America, must have broken out of a late Precambrian supercontinent. Noting the similarity in the subsidence curves of protoAppalachian Laurentia and the Precordillera of northwest Argentina, and the presence of the characteristically Laurentian Olenellid benthic trilobite fauna in the Lower Cambrian strata of the Precordillera, they suggested that the present Atlantic margin of Laurentia and the present Pacific margin of South America might have been conjugate rifted margins. Critically, this broke away from the Wilsonian view of prePangea plate tectonics, which held that the early Paleozoic lapetus Ocean had opened and closed between Laurentia and the European margin in the same site relative to those cratons as the Mesozoic-Cenozoic Central and North Atlantic Ocean basins. Rather, they suggested that Laurentia originated within today's southern continents. The early 1990's saw the publication by Eldridge Moores of the Southwest United StatesEast Antarctica (SWEAT) hypothesis, its refinement by me, and the suggestion by Paul Hoffman and myself that Laurentia might have been 'sandwiched' between the cratons of what became East and West Gondwanaland. The late Chris Powell was working on the assembly of Gondwanaland at the time, and immediately saw and grasped the opportunity to establish a research centre dedicated not only to the study

of late Precambrian paleogeography, but indeed of all the supercontinents of which Australia had ever been a part. It was this broad vision that brings us to Fremantle in 2005. First to follow were refinements: the idea that the Greenland-Scotland-Labrador promontory of Laurentia might have rifted from the ancestor of the Arica embayment in proto-Andean South America; the suggestion that the Precordillera might represent a 'tectonic calling card' from Laurentia transferred during a collision or 'near miss' rather than its conjugate margin; and the hypothesis that the South China block might have been located between the Pacific margins of Laurentia and East Gondwanaland. Then came radically different ideas: Australia against the western margin of the US (AUSWUS); northeast Australia against Mexico (AUSMEX); and a revival of the older hypothesis by Ray Price and Jim Sears that Siberia might have been juxtaposed with the Pacific margin of Laurentia until the late Precambrian. Still others have followed, and the paleogeography has meantime taken on a new importance with the revival and refinement of Sir Douglas Mawson's and Brian Harland's idea of a global late Precambrian glaciation, in the form of the snowball Earth hypothesis of Joe Kirschvink, Paul Hoffman and their colleagues. In this presentation, I will not try to defend the original SWEAT hypothesis, or to justify any specific revision. Rather I will review what I regard as the most important attributes that any viable hypothesis regarding the make up of late Precambrian supercontinent must possess. I will go on to present new data and ideas that appear to me to be potentially fruitful in the quest for 'the perfect Rodinia,' and hence at least one of the ancient supercontinents so avidly sought by Chris Powell.

Supercontinents and Earth Evolution Symposium 2005

EVIDENCE FOR EARLY SUPERCRATONS, SOME RULES OF THE GAME, AND THE QUICKEST WAY FROM A TO B Wouter Bleeker Geological Survey of Canada, 601 Booth Street, Ottawa, Ontario, Canada, K1A 0E8 (wbleeker@nrcan.gc.ca) Annong the most fundamental contributions the geological sciences have made to the overall body of scientific knowledge are: 1) the concept of deep time; 2) a detailed record of biotic evolution, as preserved in the fossil record; 3) a detailed and mobilistic view of the workings of planet Earth, e.g. plate tectonics; and 4) a record of its evolving paleogeography through time. The latter is very much a work in progress, known in detail only back to ca. 250 Ma, the time of "maximum packing" of supercontinent Pangaea. Prior to 250 Ma, the paleogeographic record of Earth's continental crust becomes increasingly speculative, although there is growing optimism that this problem may be tractable, in principle, back to ca. 2.5-2.6 Ga, the age of "cratonization" of a considerable fraction of extant crust. Older crustal fragments are either too few in number or too reworked to allow meaningful pre-2.7 Ga global reconstructions, although isolated reconstructions can be suggested for some of the better preserved crustal fragments (e.g. the Vaalbara hypothesis, or qualitative docking histories within individual Archean cratons). Earth's evolution is that of a broad cyclicity superimposed on monotonic evolution driven by declining thermal budget and irreversible chemical differentiation. The first-order cyclicity of the supercontinent cycle, perhaps the dominant signal since about 2.5 Ga, involves episodic break-up and re-aggregation of continental crustal fragments at a time scale of about 900 to 300 million years. This time scale decreased with growth in average plate size. Among the many extant crustal fragments ca. 35 are Archean cratons. These ancient "pieces of the puzzle" are of special interest because they have the longest memory. Over time, with successive break-up events, they can only get smaller; some fragments may have been erased completely from the record, almost certainly biasing the geological record. Every global break-up event should have increased the number of Archean fragments. Alternatively, going back in time, every valid

supercontinent reconstruction should reduce the number of Archean fragments in play. Indeed, detailed Pangaea reconstructions immediately reduce the number of Archean fragments. In contrast, few Rodinia reconstructions pass this qualitative test. Many well-known Archean cratons are surrounded by Paleoproterozoic rifted margins (e.g., the Superior, Slave, etc.), and thus must have broken out of late Archean "supercratons" that formed in response to a singular peak in crustal growth around 2.7 Ga. Whether these supercratons ever formed a transient late Archean supercontinent remains unproven. I will present growing evidence for one late Archean supercraton, Superia, the ancestral landmass from which the Superior craton originated. Other cratonic fragments that may trace their origin back to 2.6-2.4 Ga Superia are Karelia, Yilgarn, Hearne, and Wyoming cratons. Break-up of Superia likely involved unzippering along an Atlantic-style oceanic rift propagating from one plume centre to the next starting at ca. 2505 Ma. Break-up of any large late Archean landmass must have liberated both "internal pieces" and "external pieces", the latter characterized by long term, ca. 2.7-2.4 Ga, active and/or passive margins. In my judgement such external pieces are under-represented in the geological record, reinforcing the notion of preservational bias. At present, all reconstructions, including that of Rodinia, remain severely under-constrained. Short-lived mafic magmatic events and their dyke swarms can now be routinely dated with high precision and can constrain identity ("bar codes"), relative orientation, latitude, and polar wander paths for continental pieces. The most efficient route to more robust reconstructions, benefiting many scientific disciplines, would be a concerted international effort to date all mafic magmatic events around the world in the next decade or so. The cost of such a programme would be modest compared to other "big science" projects.

Supercontinents and Earth Evolution Symposium 2005

CACHE CREEK TERRANE, CANADIAN CORDILLERA: IMPLICATIONS FOR PLATE TECTONICS OF PANGEA AND PANTHALASSA, AND FOR PALEOGEOGRAPHIC RECONSTRUCTIONS OF RODINIA Gilles D Borel\ Stephen T Johnston^ ^Museum of Geology, BFSH-2, CH-1015, Lausanne, Switzerland (Gilles.Borel@unil.ch) ^SEOS, Univ. of Victoria, PO Box 3055, STN CSC, Victoria, BC, V8W 3P6, Canada The assembly of supercontinents is a manifestation of plate tectonic processes. The application of plate tectonic boundaries to models of supercontinent assembly and evolution can, therefore, provide an efficient means of testing paleogeographic models. In the case of pre-Paleozoic supercontinents, such as Rodinia, paleomagnetic data provides the main method of constraining the relative positions of the continental blocks that comprise the supercontinents. In the case of the Late Paleozoic to Early Mesozoic Pangea supercontinent, further paleogeographical constraints are provided by faunal data. Here we make use of the Cache Creek terrane, a Tethyan oceanic assemblage within the Canadian Cordillera, to constrain plate tectonic models of Pangea and its surrounding super-ocean, Panthalassa. The Cache Creek terrane consists in part of shallow water carbonates that overlie OIB basalts indicating deposition on seamounts or oceanic plateaux. Equatorial carbonate deposition spans the Late Carboniferous into the Lower Triassic, a time during which Earth's continental blocks resided within a supercontinent (Pangea) configuration. Established paleogeographic and plate tectonic models show Pangea surrounded by a global ocean (Panthalassa), and characterized by an internal oceanic domain (Tethys) that opened eastward between Laurasia to the north and Gondwana to the south. Terranes of the Indo-China area are thought to have formed the eastern limit of the Tethyan domain. Cache Creek fauna are Tethyan and contrast with coeval North American fauna. For example, Tethyan Verbeekinid (Yabeina) fusilinids in Early to Late Permian thick subtidal carbonates contrast with the exclusively Swagerinid fusilinids of the Americas, and the only non-Asian occurrence of the Late Permian conodont Wardlawella, is in the Cache Creek terrane. Fauna remain distinctively Tethyan until the Middle Triassic. Carnian (230 Ma) accretion of the Cache Creek seamounts is

recorded by onlap of arc-derived detritus onto the carbonates, and by inflection of the arcs against which Cache Creek terrane accreted by Tethyan fauna. The Cache Creek terrane therefore provides key plate tectonic constraints - the seamounts had to remain within the tropical belt during Permian to Triassic translation, and appropriate plate boundaries are required to facilitate translation eastward across the entire Panthalassa. If we assume a starting point for the Cache Creek seamounts near the Indo-Chinese terranes of easternmost Tethys, a plate velocity of 20 cm/y (for all models) and seamount accretion at 230 Ma, accretion is limited to have occurred in the middle of Panthalassa. A northsouth trending ridge opening about a polar Euler pole with no subduction beneath western Panthalassa, forcing the spreading ridge to migrate to the east, could account for the predicted translation. This model is, however, problematic as the arcs against which Cache Creek accreted are commonly interpreted as having American affinities. Alternatively, if the arc terranes lay adjacent to the Americas, with a 2000 km wide back-arc, the origin of the Cache Creek is restricted to central Panthalassa. This model is inconsistent with the strong Tethyan character of Cache Creek fauna. If Cache Creek originated in Tethys, and the arcs to which it accreted are American, then unrealistic rates of translation are required (40 cm/y for 50 million years). These restraints make plate tectonic scenarios for accommodating translation of Cache Creek difficult, and call into question our current understanding of Pangea and Panthalassa. Faunal constraints such as those provided by the Cache Creek terrane are unavailable for most of Earth history, limiting our ability to test most proposed supercontinent reconstructions. Placing models of super-continent assembly and dispersal within a plate tectonic framework can, however, help to distinguish between the plausible and the implausible.

Supercontinents and Earth Evolution Symposium 2005

SUBDUCTION EROSION DESTROYS CONTINENTAL CRUST AT A RATE THAT IF APPLIED TO THE PAST BRINGS A CHALLENGE TO SUPERCONTINENT RECONSTRUCTION David W Scholl\ Roland von Huene^ ^Department of Geophysics, Stanford University, Stanford, CA 94305, USA (dscholl@pangea.stanford.edu) Department of Geology, University of California, Davis, CA 95616, USA

INTRODUCTION At subduction zones the underthrusting ocean plate conveys large quantities of continental debris toward the mantle. Transport occurs in the subduction channel separating the two plates and is effected by two kindred processes: 1) sediment subduction, and 2) subduction erosion. Subduction erosion is the important process with respect to continental reconstructions because it removes crustal material from the submerged forearc. This results in a loss of crust at the seaward edges of continents and island arcs, the long-term impact of which is the permanent and time-progressive destruction of material useful in guiding continental reassemblies, in particular those of older age. EVIDENCE OF SUBDUCTION EROSION Subduction erosion, a form of tectonic erosion, reduces the width of the submerged margin and dramatically thins its rock framework. The principal evidence for thinning is the recovery near the trench of deeply subsided (4-5 km) shoreline or non-marine sediment and rock. Subsidence of this magnitude requires crustal reduction on the order of 10-15 km. Other manifestations of forearc truncation and foreshortening is a long record (>10-20 Myr) of inboard migration of the arc and the coastal and offshore occurrence of Mesozoic arc magmatic rocks that are typically emplaced 150-200 km inboard of a trench. RATES AND VOLUMES For each lateral km of margin, the average solid volume transport of eroded material toward the mantle is estimated at --45 km^/million years. This volume rate, based on measurements along ten subduction zones, corresponds to a truncation rate (landward trench migration) of 2-3 km/million years. This is only about a thirtieth of the orthogonal rate of convergence. Globally, the annual transport of subducted sediment and eroded forearc is -2.5 km^, of which 2/3 is tectonically eroded debris. The thickness of the material in the subduction channel averages --1

km. Because arc rocks are chiefly mantle melts and the global volume of exposed but once deeply subducted material is small, we presume that >95 % of subducted crustal matter is recycled to the mantle. IMPLICATIONS FOR SUPERCONTINENT RECONSTRUCTIONS If modern rates of recycling are applied to the past 2.5 billion years, during which subduction has been ongoing, then a volume of continental crust roughly equal to the standing volume (7 X 10^ km^) has been flushed to the mantle. Destruction of this amount of crust, in particular the 2/3 or perhaps more erased by subduction erosion, must have been importantly involved jn shaping the Earth's existing framework of basement rock. For example, just since the breakup of Pangea, the estimated subduction erosion of the Peru-Chile forearc has removed a ~6000-km-long, ~500-km-wide coastal strip of Mesozoic arc and older cratonic crust from western South America. Since Rodinian rifting of parts of South America from Laurentia, It is likely that subduction erosion demolished an even greater width of western Gondwanan South America. Similar truncation scars must have commonly formed elsewhere in the past. The older cratonic crust is, the greater is the likelihood that rock of this age has been exposed to subduction zone destruction. Older continental reconstruction must thus be attempted with increasingly less remaining material to guide the restoration. The perspective is gained that the relatively small volume of exposed Archean crust is more a consequence of what remains than all that was originally made. No Archean craton borders a modern subduction zone. Preservation of ancient crust in a subduction zone environment is favoured by formation of a protecting offshore arc-backarc basin system; for example, as the Japan and Okhotsk Seas shield Archean terranes of western China from western Pacific subduction zones.

Siipercontinents and Earth Evolution Symposium 2005

THE LOWER MANTLE TWINS: SUPERPLUMES AND POSTPEROVSKITE PHASE TRANSITION David A Yuen Dept. of Geology and Geophysics and Minnesota Superconfiputing Institute, University of Minnesota, Minneapolis, MN 55455-0219, USA (davey@krissy.geo.umn.edu)

The formation of supercontinents must be intimately related to the dynamics of the lower mantle, because of the amount of material which must be brought up from the transition zone. The plume dynamics emerging from the transition zone are influenced by the underlying superplumes. Supercontinents are related to the formation of superplumes in the lower mantle, whose existence was unveiled around 20 years ago by Adam Dziewonski's pioneering work in seismic tomography. Two large structures with girths exceeding 1000 km were observed in the lower mantle under the central Pacific and Africa. They have been invoked by the Japanese school under Shige Maruyama in 1994 as a viable agency for precipitating singular tectonic events, such as continental breakup. Within the past two years, there has been an important discovery of a new phase occurring in the deep mantle, called the post-perovskite phase. This phase transition has been uncovered by both X-ray observations and first principles calculations using local density functional theory. This novel phase change takes place at a depth of about 200 km above the core-mantle boundary at temperatures of around 2700 to 2800 K. It is exothermic in character, leading to many small-scale

instabilities at the base of the mantle. In order for superplumes to develop successfully in the lower mantle under the severe circumstances imposed by the post-perovskite phase transition, one must invoke some sort of radiative heat transfer in the lower mantle. Recent work on the high spin to low spin electronic transition of ferrous iron in wustite has lent support to this possibility for enhanced radiative transfer in the deep mantle. This is also corroborated by Mossbauer spectroscopy under high pressure conditions. In the last year since its initial discovery, rapid developments in mineral physics, seismology, and geodynamics have taken place, and a genuine revolution in our understanding of the lower mantle and its dynamics is developing. Because of its positive Clapeyron slope, the post-perovskite transition would not occur in a hot young Earth but would only develop some time afterwards. Therefore the timing of superplumes and supercontinents must happen later in Earth's history, because of the encroachment by the post-perovskite phase in the deep mantle along with its geophysical, geochemical and geological ramifications. In this talk I will bring the audience up to speed on this new movement in the Earth sciences.

Supercontinents and Earth Evolution Symposium 2005

AGGREGATION AND DISPERSAL OF SUPERCONTINENTS IN GLOBAL MANTLE CONVECTION MODELS Benjamin R Phillips\ Hans-Peter Bunge^ ^Department of Geosciences, Princeton University, Princeton, NJ 08544, USA (benp@princeton.edu) ^Geophysics, Departnnent of Earth and Environnnental Sciences, Munich University, Theresienstr. 41, 80333 Munich, Germany

In 1966, J. Tuzo Wilson suggested that the Atlantic Ocean basin had closed and then reopened, a process now commonly termed the Wilson Cycle. Since then, numerous paleomagnetic studies have shown that Wilson's original idea may be extended to describe a global cycle, punctuated by the periodic formation of supercontinents such as Pangea, Rodinia, and Columbia, separated by time scales of several hundred million years. It is generally accepted that these motions are coupled to large scale mantle convection. Early two dimensional (2D) mantle convection models demonstrated the dynamic feasibility of such supercontinent cycles. Here we present the first ever high resolution, 3D spherical mantle convection models with multiple continents. We study

models incorporating three to six continents in a predominantly radiogenically heated mantle with radially stratified viscosity. From these models we find that continents aggregate and disperse cyclically, with variable periods of several hundred million years. We also find that continental and oceanic velocities vary in association with these cycles. Finally, the digressions in mantle temperature that we record suggest that the coupled continent-mantle system leads to transient heating and cooling throughout the mantle, both below continents and below oceans. These results agree well with geologic and geophysical observations and place dynamic constraints on global mantle flow models.

Supercontinents and Earth Evolution Symposium 2005

TWO-CELL MANTLE CONVECTION AND THE DISTRIBUTION OF THE CONTINENTS IN THE PROTEROZOIC David Giles School of Geosciences, Australian Crustal Research Centre, Monash University, Melbourne VIC (giles@mail.earth.monash.edu.au)

The present Earth can be simplified to a two-cell convection system with poloidal equatorial upwelling in the Pacific and beneath Africa and a complimentary girdle of downwelling surrounding the Pacific Ocean. The continents are contained entirely within one cell (Pangea cell) whereas the other cell is entirely oceanic (Pacific cell). This appears to be a relatively stable arrangement that has existed since the formation of the Pacific Ocean in the Neoproterozoic, and (we speculate) may have been a first-order control on the position of the continents throughout the Proterozoic. In this talk, I will discuss the implications of two-cell convection for the Proterozoic Earth, using the Phanerozoic distribution and motions of the continents as a template. An important aspect of two-cell convection is that the downwelling girdle, once established, should be relatively stable and should act as a long-term (100s of millions of years) stress guide on the planet. Continental material should be attracted to the girdle and develop long-lived convergent margins. These external continents should be able to move about the girdle in a toroidal way but their motions and interactions with the other continents will be governed by their position on the girdle. They might episodically converge or diverge with their neighbours but should maintain their relative position within the cell. The first order configuration of the continental cell should have a limited range of possibilities. This implies that the wholesale redistribution of the continents requires destruction of the twocell configuration. One possibility is that the continental cell expands as the oceanic cell is consumed - effectively evolving toward a transient period of single-cell convection. In this case, the distribution of the continents would be controlled by the consumption of the oceanic cell and (once again) there should be a limited range of possible configurations. In the simplest case.

the continental cell would effectively be turned inside out and the long-lived convergent margins would become collision belts that occupy the interior of a new continental cell established on the opposite side of the globe. The new continental cell would be bisected by a long-lived collision zone of hemispheric proportions. This geometry could be argued for the ca 1.31.0 Ga assembly of Rodinia (although if we use the Grenvillian belts as pinning points in our reconstruction the argument is somewhat circular). However it is not a particularly good match for the ca 1.9-1.8 Ga or the ca 0.6-0.5 Ga orogenic belts, which have also been linked to major periods of continental amalgamation. The Pan African and Trans-Hudson systems consist of a complex interconnected array of relatively short orogenic belts, many of which have a limited history of convergence. This geometry is more consistent with a within-cell continental reconfiguration than with a period of global-scale cell-busting tectonics. The Pan African example provides a useful study. Whilst the assembly of Gondwana undoubtedly involved major convergence, the survival of significant lengths of Neoproterozoic (Rodinian?) convergent margins into the Phanerozoic (e.g. Siberia, China, the Cadomian terranes) suggests that the Neoproterozoic oceanic-cell was not entirely consumed during the Pan African. The ultimate fate of that ocean can be surmised from the distribution of the surviving Neoproterozoic convergent margins on the Phanerozoic Earth - decorating the margins of the Tethys. We propose that the Pan African belts formed by the consumption of internal oceans within the Neoproterozoic continental cell (the Rodinia cell) as a first response to the opening of the Pacific. The Neoproterozoic super-ocean, rather than being entirely destroyed, became an internal ocean in the new Pangea cell (the Tethys), and was gradually consumed during the Phanerozoic.

Supercontinents and Earth Evolution Symposium 2005

THE LARGE IGNEOUS PROVINCE (LIP) RECORD OF RODINIA Richard E Ernst\ Kenneth L Buchan\ Michael TD Wingate^, Zheng-Xiang Li^ ^Geological Survey of Canada, Natural Resources Canada, 601 Booth St, Ottawa, Ont., K1A 0E8, Canada (rernst@NRCan.gc.ca) ^Tectonics Special Research Centre, University of Western Australia, Crawley, WA 6009, Australia

Large igneous provinces (LIPs) represent high volume, generally short duration pulses of magnnatisnn that punctuate Earth's history perhaps as often as every 10 to 20 million years. Most are dominated by mafic magmatism, and many have been linked with mantle plumes. In addition, they are commonly associated with continental breakup, can cause regional domal uplift, may have a significant effect on climate, and are highly prospective for Ni-Cu-PGE mineral deposits. In this contribution, we focus on the implications of LIPs for the reconstruction of Rodinia and its breakup history. During the period ca. 1.7 to 0.5 Ga, which is relevant to the life cycle of Rodinia, significant periods of LIP magmatism occurred at: 1.67-1.57 Ga (northern Baltica, northeastern and southern Australia, northeastern Laurentia), 1.46 Ga (eastern and western Laurentia, northwestern Australia, northwestern Baltica), 1.38-1.35 Ga (western Laurentia, Siberia, Antarctica, Baltica, Congo?), 1.28-1.26 Ga (northern and eastern Laurentia, northwestern Baltica), 1.25-1.23 Ga (eastern Laurentia, Antarctica), 1.21 Ga (western Australia), 1.11-1.07 Ga (central and southwestern Laurentia, south-central Africa, Antarctica, central and western Australia), 0.97-0.94 Ga (northwestern Baltica), 0.83-0.80 Ga (southern Australia, South China, Congo?),

0.78-0.75 Ga (western and southeastern Laurentia, South China, India, western Australia, Seychelles, southwestern Asia?, southwestern Africa?), 0.72 Ga (northern Laurentia, southwestern Africa?), 0.62-0.55 Ga (eastern Laurentia, northwestern Baltica, Rio de la Plata Craton?), 0.5 (western Australia, southern Siberia) [In the above list, note that northeastern and eastern Laurentia include Greenland]. This LIP record can be used in reconstructing Rodinia and identifying the timing and patterns of breakup events: 1) LIPs are excellent targets for integrated studies of geochronology and paleomagnetism. "Key" (well-defined and precisely-dated) paleopoles are necessary for rigorous testing of continental reconstructions. 2) "Bar codes" (term coined by W. Bleeker) of LIP events can be constructed for each craton; cratons with similar bar codes may have been in close proximity to one another. 3) Giant linear and radiating dyke swarms can be used as piercing points for reconstructing cratons. 4) Margins associated with similar aged LIP events may represent formerly conjugate margins.

Supercontinents and Earth Evolution Symposium 2005

KIMBERLITE MAGMATISM AND SUPERCONTINENT CYCLES Chris Hatton De Beers Exploration, PO Box 82232, Southdale 2135, South Africa (chris.hatton@debeersgroup.conn)

Peaks in kimberlite activity coincide with major episodes of continental growth and plate reorganisation. These episodes may be related to the interaction of shallow and deep mantle convective systems. Shallow mantle convection is regulated by variations in lithospheric thickness. A model in which the base of the small scale convecting layer is at the 410 km discontinuity accurately accounts for the observed lithospheric geothermal gradients. A thin convecting layer necessarily underlies thick lithosphere, and because heat flow is proportional to layer thickness, thick lithosphere is associated with low heat flow and low geotherms. Because the base of the lithosphere is isothermal, with a temperature of 1350°C, the temperature at the 410 km discontinuity is consequently much colder (>230°C colder according to the model,) beneath thick cratonic lithosphere than beneath thin oceanic lithosphere. For thermal stability, therefore, continents must overlie zones of the deep mantle which are colder than average. Two major warm upwellings, sometimes referred to as superplumes, and an intervening girdle of cold downwelling mantle, control temperature variations in the deep mantle. One upwelling, broadly associated with the African plate, is centred at 0°N and 10°E and spreads over a zone of 60° radius. The Pacific upwelling, also of 60° radius, is centred at 0°N, 190°E. The intervening girdle hosts western Australia, most of Asia, all of North America but for a sliver to the west of the Sierra Nevada, most of South America, and all of Antarctica. Most continents therefore conform to the proposal that cratons should be preferentially located in a zone associated with colder lower mantle. Those portions of continents lying above the upwellings are associated with tectonic and

magmatic activity. An extensive zone of magmatism runs along the rift valleys of Africa. Alkali basalts are widespread in eastern Australia, and the west coast of North America is the site of transcurrent faulting and magmatism. A plausible hypothesis to account for the above observations is that when continents move from the cooler girdle into the zone of deep mantle upwellings, the increase in temperature at the 410 km discontinuity thermally destabilises the cratons, initiating a cycle which culminates in widespread magmatism. The episodic nature of crustal magmatism might then reflect the periodic drift of continents from the cooler girdle into the upwelling zones. This drift could be driven by rotation of the deep mantle convection system relative to the shallow mantle, or by growth of an ocean on one upwelling, driving the continents to the other upwelling. Increases in kimberlite activity might also have a direct relation to the movement of continents from the cooler girdle to the warmer upwellings. However Tertiary kimberlite activity in Canada is one example of kimberlite activity within the colder girdle, and points to an alternative cause of kimberlite magmatism: hotspots. Most hotspots are located within the warm upwellings, but several lie within the cooler girdle, so kimberlite activity would then be expected to increase when continents drift into upwelling zones, but would not be confined to these periods. While the linear character of many kimberlite provinces is probably controlled by the shallow small scale convective rolls, which are elongated in the direction of plate motion, relatively stationary hotspots account for the age progression that is observed in these linear provinces.

Supercontinents and Earth Evolution Symposium 2005

THE PALAEOPROTEROZOIC ERA (2500-1600 Ma): GLOBAL PALAEOGEOGRAPHrS FINAL FRONTIER David AD Evans Department of Geology & Geophysics, Yale University, PO Box 208109, New Haven, CT 06520, USA (dai.evans@yale.edu)

Is there a supercontinent cycle? Have its characteristic periodicities been speeding or slowing through Earth history? Do supercontinents "introvert" or "extrovert?" What (if any) is the general relationship between palaeogeography and global climate? How can the search for supercontinents guide mineral exploration on global (temporal) and regional scales? These questions can only be addressed through a multi-gigayear perspective, whereby reconstructing Pangaea and Rodinia constitutes merely the starting point of a greater endeavor. Such an effort will require global advances in tectonostratigraphy, geochronology, and palaeomagnetism. In a deep-time context, the latter field currently lags behind the other two, but the basic physics of magnetism in rocks provides no hindrance to investigations of even our oldest geological record. Instead, three geodynamic factors appear to place limitations on our abilities to reconstruct the most ancient supercontinents. First, we expect that the earliest Earth lacked a strong internal magnetic field. Determining when the geodynamo attained a palaeogeographically useful intensity is a daunting task in itself, but current estimates typically range within the interval 3000-2500 Ma. By this factor alone, palaeomagnetism should be able to constrain all post-Archaean supercontinents. Second, global rock preservation dwindles with age. The total extent of non-reworked Archaean crustal area would fit comfortably within the bounds of a single present continent, Eurasia. Only a handful of Archaean regions of stratified rocks (allowing precise control of palaeohorizontal attitudes for palaeomagnetic study) document sub- to lowestgreenschist metamorphic grade that, in most instances, forms the upper limit to preservation of a pre-metamorphic magnetic remanence. Third, extant Archaean continental lithosphere appears to have been less rigid than Proterozoic plates, as evidenced by lack of abundant dyke swarms and a smaller characteristic wavelength of deformation in granite-greenstone terrains. This means the select handful of low-grade rock

exposures will bear little utility for palaeogeographic reconstructions of substantially larger regions. These three factors, in combination, suggest that the Palaeoproterozoic Era (2500-1600 Ma) will be the final frontier to global palaeogeography, as long as palaeomagnetism remains the only quantitative tool for Precambrian plate reconstructions. In contrast, the Palaeoproterozoic geodynamo was at least sporadically if not persistently strong, plates were stiff and internally rigid over larger areas, and rocks from many regions of the world are well preserved. High-precision dating of this more extensive geological record, using a variety of novel techniques, advances satisfactorily. Palaeomagnetic reconstructions that cover a large portion of the Earth's surface back to -2200 Ma are foreseeable within this century. A late Palaeoproterozoic supercontinent, known as Nuna (or Columbia, Hudsonland, etc.) appears to have been assembled by 1900-1800 Ma, as documented by a worldwide network of orogenic belts from that age interval. The geometry of Nuna will be constrained in large part by a "correct" Rodinia configuration, the latter entity currently enjoying intense debate. Many proposed Rodinia connections lack a "Grenvillian" suture and therefore must have been inherited from earlier assemblies; in this way, once Rodinia is solved then the search for Nuna will be well underway. Prior to Nuna's assembly, different regional perspectives led to different views of supercontinental episodicity. One may see the ca. 2100-Ma Eburnian and Transamazonian orogenies as precursors to Nuna's assembly, but during the same period numerous mafic dyke swarms manifest the end of a protracted early Palaeoproterozoic interval of rift-related magmatism. Broader spatial contexts of such events are needed. Defining Sclavia, Superia, Kenorland, and Vaalbara—as supercratons or supercontinents—will present a great challenge for deep-time palaeogeographers.

Supercontinents and Earth Evolution Symposium 2005

SPATIO-TEMPORAL PATTERNS OF OROGENESIS IN THE SUPERIOR PROVINCE, CANADA Paul M Evins, William J Collins School of Earth Sciences, James Cook University, Townsville, QLD 4811, Australia (paul.evins@jcu.edu.au) The range of postulated orogenic processes in the Archean is surprising compared to the relatively simple array of dominantly subductiondriven orogenic processes active today. Mechanisms ranging from mantle plumes to multiple reversals in subduction polarity to amalgamation and occasional subduction of soft continental microplates to backarc basin development have been invoked to explain geological, geochemical, and geochronological observations from Archean cratons. The Superior Province is the largest exposed Archean craton on Earth. The wealth of data from the Lithoprobe and NATMAP projects makes it also one of the best studied cratons. Recent studies have focused on whether the Neoarchean volcanosedimentary packages were deposited on a common basement (autochthonous) or evolved disparately as microplates (allochthonous with intermittent subduction zones) and were subsequently

amalgamated. Both models require a rather convoluted sequence of orogenic processes that only partially explain the temporal and spatial patterns of magmatism, metamorphism, deformation, and basin formation observed in the province. Can these patterns be explained simply by subduction related processes, or are processes required that were unique to the Archean? Both published and new orogenic models will be tested for the Superior Province using newly constructed time-space diagrams for both the eastern and western Superior Province, based mainly on the major N-S trending western Superior and the Abitibi Lithoprobe transects and an up-to-date synthesis of the geochronology. The analysis will be augmented by preliminary estimates of crustal paleothickness variations through the orogen, based on greenstone geochemistry.

Supercontinents and Earth Evolution Symposium 2005

TWO NEOARCHEAN SUPERCONTINENTS REVISITED: NEW CONSTRAINTS ON RECONSTRUCTIONS OF KENORLAND AND NUNA FROM THE CORE OF LAURENTIA Sally J Pehrsson, Robert H Ralnbird, Robert G Berman, William J Davis Geological Survey of Canada, 601 Booth Street, Ottawa, Ontario, K1A 0E8, Canada (Pehrsson@nrcan.gc.ca)

INTRODUCTION The past decade has seen several different reconstructions for one or more late Neoarchean to early Paleoproterozoic super-continents, largely based on improved U-Pb geochronology of sedimentary basins and orogens, key tie points in any reconstruction. Models favouring a single supercontinent, Kenorland, classically include at least Superior, Wyoming, Churchill and Karelia. A second Neoarchean supercontinent, Zimbalvara, comprising the Zimbabwe, Kaapvaal, and Pilbara cratons, has also been proposed. Neoarchean reconstructions involving the Yilgarn, North China, Slave, Nain, Congo and Sao Francisco cratons vary widely, but it is broadly accepted that all were amalgamated in a single Nuna supercontinent by ca 1.8 Ga. A central feature of Kenorland reconstructions is the coherence of the Superior, Wyoming and Churchill cratons in a single block. RESULTS Utilizing an expanding geochronological dataset for the Churchill, we consider the evidence for various Kenorland reconstructions and highlight that the Churchill (comprising the Rae and Hearne domains) may not have amalgamated until the assembly of Nuna. Recognition of a major trans-Rae ca 2.3 Ga orogenic event, a

2.55-2.45 Ga magmatic/ collisional event, and younger rift-cover sequence provides critical piercing points for reconstruction. Differences in their pre-2.1 Ga history suggest the Rae domain, unlike the Hearne, was not attached to the Superior-Wyoming margin in Kenorland. The Rae's distinctive 2.3-2.5 Ga magmaticorogenic history, comparatively rare in the worldwide record, is similar to that of the North China craton, to which it may have been connected in the late Neoarchean-earliest Paleoproterozoic. The vergence and distribution of a 1.9-1.8 Ga Trans-Hudson orogen across the Rae and intriguing similarities between a ca. 1.77-1.75 Ga mafic igneous event and associated sedimentary basins/ rapakivi granite intrusions, allow that the North China craton lay to the north of the Rae by at least Nuna time (present co-ordinates). Other Archean cratons that share this distinctive 2.62.3 Ga history, but similarly lack the 2.4-2.2 Ga glaciogenic sequence that overlies the Superior, Wyoming and Hearne, include the Sao Francisco craton and the poorly-known Sask craton. CONCLUSION Collectively, these data lend additional support for a second Neoarchean supercontinent and highlight alternative reconstructions for the assembly of Nuna.

Supercontinents and Earth Evolution Symposium 2005

BALTICA BEFORE RODINIA: SEPARATE EXCURSIONS OF THREE CRUSTAL SEGMENTS Svetlana V Bogdanova Dept. of Geology, Lund University, Solvegatan 12,22362 Lund, Sweden (Svetlana.Bogdanova@geol.lu.se)

LITHOSPHERIC SEGMENTS OF PROTEROZOIC BALTICA The continental core of the Baltica proto-plate is the East European Craton (EEC). That craton originated at ca. 1.8-1.7 Ga by collision of three crustal segnnents named Fennoscandia, Sarnnatia and Volgo-Uralia, each of which had a different Archaean and Palaeoproterozoic evolution. The Archaean crust of Fennoscandia is made up of several terranes with ages not exceeding 3.5 Ga. Major stages of accretion and gradual growth of the continental crust occurred at 3.1-2.9, 2.92.85, 2.85-2.75, and 2.75-2.65 Ga. Between 2.5 and 2.0 Ga, the Archaean crust of Fennoscandia was affected by mantle-plume events that were accompanied by rifting and fragmentation, and even the opening of small "intra-continental" oceans. These were closed between 1.9 and 1.8 Ga, semi-simultaneously with the Svecofennian orogeny, which took place off the western margin of the Archaean core and included several accretionary (subductional-collisional) events. Orogenic processes also occurred later, e.g. at 1.81.1.78 and 1.77-1.75 Ga, and after the creation of the EEC at 1.70-1.65, 1.60-1.55 Ga and 1.501.45 Ga, and at 1.2-1.0 Ga. Sarmatia is built up of several formerly independent Archaean domains. Its main crustforming processes occurred at 3.7-2.9, 3.2-3.0 and ca. 2.7(?) Ga. Ca. 2.2-2.0 Ga orogenic belts separate these domains, whereas a large continental magmatic arc, ca. 2.0-1.95 Ga in age, follows its NW margin. Characteristic are large 1.79-1.74 Ga gabbro-anorthosite-rapakivi plutons, occurring in the Ukrainian-Shield part of Sarmatia. No other Proterozoic crustal events have so far been detected in that segment. The least known crustal segment is VolqoUralia, which is buried beneath Proterozoic to Phanerozoic cover and is a coherent Meso- to Neoarchaean high-grade domain. It is marked by unique circular patterns of magnetic and gravity anomalies interpreted as large Palaeoproterozoic domes developed some time before 2.1 Ga. Was the doming a reflection of mantle plume tectonics between 2.6 and 2.1? After that Volgo-Uralia

underwent a crustal evolution similar to that of Sarmatia. INTERSEGMENT SUTURE ZONES These comprise remnants of Palaeoproterozoic oceanic island arcs, back-arcs, and active as well as passive continental margins. As is seen from geophysical surveys, intense collisional tectonics characterised the suture zones. Meso- to Neoproterozoic transcratonic rifting also followed their courses. Convergence took place at various times. Sarmatia and Volgo-Uralia docked at ca. 2.1-2.0 Ga, forming the larger continent of VolgoSarmatia. The final assembly of the EEC occurred between 1.8 and 1.7 Ga, when Fennoscandia joined in. After that, all the segments shared a common evolution, but at the western EEC margin, accretion periodically lasted until Rodinia time, whereas in the east, rifting and the formation of a passive margin occurred from ca. 1.6-1.5 Ga onwards. BALTICA IN RECONSTRUCTIONS OF A PALAEOPROTEROZOIC SUPERCONTINENT In Palaeoproterozoic palaeomagnetic and palaeogeographic reconstructions, Baltica has mostly been treated as a uniform large continent. This is not correct and contradicts the geological evidence of segments with different stories of crustal evolution. Also their small dimensions must be considered when modelling a supercontinent. Until ca. 1.8 Ga, Sarmatia and Volgo-Uralia were part of a lithospheric plate different from that which contained Fennoscandia. The latter had much in common with the Late Archaean Laurentia-West Greenland "Kenorland" amalgamation, while the former resemble the Congo, West Africa, Sao Francisco and partly the Amazonia and Australia cratons. Separate assemblies are thus required. After creation of the EEC, good fits of western Baltica to Amazonia and Laurentia-Greenland are obvious for most of the time between ca. 1.8 and 1.35 Ga. That assembly remained stable until it drifted from equatorial to polar positions. Which of the continents neighboured eastern Baltica but was rifted off in the Meso- to Neoproterozoic remains a problem. South Australia is a good candidate.

Supercontinents and Earth Evolution Symposium 2005

1.83 - 1.77 Ga EXTENSION OF THE NORTH CHINA CRATON Guiting Hou^'^, Jianghai Li\ Chuancheng Wang\ Xianglin Qian^ ^School of Earth and Space Sciences, Peking University, Beijing 100871, PR China (guiting.hou@anu.edu.au) Research School of Earth Sciences, Australian National University, ACT 0200, Australia

INTRODUCTION Late Proterozoic extensional structures of the North China Craton (NCC) and their significance to reconstruction of a pre-Rodinia supercontinent are discussed based on research into mafic dyke swarms and aulacogens. RESULTS The NCC was stabilised at 1.85 Ga after the last orogeny and metamorphism along its northern margin (Inner Mongolian-Northern Heike Orogen, and the northern Central Orogen Zone). Crustal extension then started, as demonstrated by the Xiong'er - Zhongtiao aulacogen, Yanliao aulacogen and extensive mafic dyke swarms. The Xiong'er-Zhongtiao aulacogen developed Xiong'er Group volcanic rocks stretching to interior from the southern NCC as a triple junction(1.83 Ga) while the Yanliao aulacogen developed anorogenic plutons (1.8-1.7 Ga) and Tuanshanzi-Dahongyu Formations (1.7-1.6 Ga). The mafic dykes cut all the Neoarchean and early Proterozoic terranes and range in age from 1.83 - 1.77 Ga based on recently obtained precise U-Pb zircon SHRIMP ages and "^^Ar-^^Ar ages. For example, the zircon U-Pb SHRIMP dating of dykes in Shandong and Shanxi Provinces gave weighted mean ^ ages of 1837 ± 18 Ma and 1778 ± 3 Ma. The ages of 1804 ± 16 Ma and 1780-1765 Ma are given in Shanxi Province. The average east-west crustal extension given by the dykes across the entire NCC is only 0.35%. Most extension in the NCC is contributed by late Paleoproterozoic aulacogens. Based on the tectonic setting, the NCC was subject to compression from the north and south margins of the craton. 2-D finite element modelling of the late Paleoproterozoic stress field across the NCC is carried out using ANSYS software. The NCC is regarded as a triangular plate and tectonic

forces are applied along the north and south margins of the craton whereas the east margin is a slip fault (Tanlu Fault) which already existed at the end of Archean. The modelling results show that the orientations of principle compressive stress are consistent with the dyke swarms' orientations, which indicates that the mafic dyke swarms in Neoarchean units from the Western Block, Central Orogen Zone and Eastern Block were emplaced in the same stress field. The match between the principle compressive stress and dyke orientation suggests the modelling results are reasonable. DISCUSSION The mafic dyke swarms and aulacogens are not synorogenic events because they occurred later than the last orogeny (1.85 Ga), and they are not post-orogenic episodes because their strikes are not consistent with the orogen belt strike. The geochemistry of the mafic dyke swarms also indicates that they were developed in an intraplate rifting setting. In fact, the dykes and aulacogens are the first extensional features developed in the same stress field relative to continent break-up after NCC cratonization. CONCLUSION The late Proterozoic mafic dyke swarms and aulacogens in the NCC mark the beginning of a pre-Rodinia supercontinent break-up episode. The 1.90-1.85 Ga Andean-type orogen along the north margin of the NCC should be studied and compared with orogens in other cratons, and comparison of NCC dyke swarms with those from other cratons may help to reconstruct a paleo-mantle plume position and reveal the mechanism of supercontinental break-up. Paleomagnetic work on the dykes is presently underway.

Supercontinents and Earth Evolution Symposium 2005

BORDERING SEGMENTS OF THE EASTERN GHATS BELT, INDIA, INCLUDED IN UR SUPERCONTINENT S Bhattacharva Indian Statistical Institute, 203 B T . Road, Calcutta, 700108, India (sannar@isical.ac.in)

The Ur supercontinent (-3000 Ma) consisted of four nnajor cratons: Western Dharwar and Singhbhum cratons of India, the Kaapvaal craton of southern Africa and the Pilbara craton of western Australia (Rogers & Santosh, 2003). On the other hand, the Eastern Ghats granulite terrane, according to Harley (1992), is a Proterozoic terrane, "typified by preservation of relatively uniform granulite assemblages, implying similar physical conditions over very large areas, and a lack of gradual transition to lower grade areas". However, recent isotopic data for the Eastern Ghats terrane imply that Proterozoic granulite terrane includes most of the internal segments, while bordering segments to the north and west are Archaean domains (Rickers et al., 2001). Moreover, these bordering

segments record Archaean granulite facies events (Bhattacharya et al., 2001; Bhattacharya et al., 2004, submitted). Hence, the border segments of the Eastern Ghats terrane could have been parts of the Ur supercontinent. The 'Rengali' domain adjoining Singhbhum craton to the north, with -3500 Ma continental crust, recorded a granulite facies event at 3000 Ma, followed by collisional granitic magmatism at 2800 Ma (Bhattacharya et al., 2001; Misra et al., 2000). The Western charnockite domain bordering Bastar craton, with 3460 to 3730 Ma continental crust, recorded a 3000 Ma granulite facies event, followed by collision and TTG magmatism at 2500 Ma in the cratonic section.

Siipercontinents and Earth Evolution Symposium 2005

THE CHOTONAGPUR GNEISS-GRANULITE COMPLEX AND ASSOCIATED SUPRACRUSTAL BELT, INDIA: PART OF THE COLUMBIA SUPERCONTINENT Rajesh Mukherjee Department of Geology, University of Calcutta, India (rajesh_geol@rediffmail.conn)

The Chotonagpur Gneiss-Granulite Complex (CGGC) has been mainly covered by geological reconnaissance but its role in the crustal evolution of the Indian peninsula remains to be studied in detail. WSW-ENE striking supracrustal rocks and gneisses of the Satpura Fold Belt cover a vast area of the Satpura region, including the CGGC (93°30' E to 72E; 2 r N to 24°N) and the Sausar Group. The CGGC extends onto the Shillong plateau, through ChotonagpurSinghbhum, to form a separate Precambrian province. The CGGC is characterised by distinctive granulite-amphibolite fades, politic metasedimentary assemblages along with impure calcareous metasediments, as well as calc-alkaline rocks, anorthosite, and relict Archaean enclaves. Owing to its unique position in the tectonic mosaic of the Indian shield, with numerous associated continental fragments (cratons), annealed sutures, and evidence of marine metasediments, the question arises whether there was any possible link between the CGGC and any supercontinent (cf. Unrug, 1992). In addition, large continental masses, including CGGC, record ages > -1.5 Ga. Continuous reorganization of few large Paleo- to Mesoproterozoic continental masses led to formation of the Columbia supercontinent. An attempt has therefore been made to determine whether the CGGC was a part of the Palaeo- to Mesoproterozoic Columbia supercontinent. Most Indian crust was formed prior to Palaeoproterozoic time. Ancient crust was remobilized at least twice during the Meso- and Neoproterozoic in Eastern Indian shield. Despite no detailed structural mapping, field relations show that the CGGC is in the north of Singhbhum Fold Belt (SFB), whereas the SFB girdles the Singhbhum craton in the south. Furthermore, the CGGC is demarcated regionally in the north by the Son Narmada South fault and in the south by the Central Indian Tectonic Zone (CITZ). This tectonostratigraphic framework can be ascribed to the opening and closing of a sea, documented by deposition of the CGGC and progressive cratonisation when

considered from the coupling between CGGC and SFB, over the Singhbhum craton. Coalescence of terrains related spatially and temporally with the CGGC, suggestive of complex geological history, might have extended the westerly crustal segment in the Satpura Fold Belt (SFB). The CGGC, north of the SFB, exhibits a long Proterozoic history of sedimentation and felsic plutonism, having undergone major thermal events at -1.6 Ga (high-grade metamorphism), -1.35 Ga (widespread felsic plutonism), and - 1 . 0 Ga (minor felsic plutonism). Reset isotopic signatures in the CGGC reflect intermittent granitic magmatism between ca. 1.7 and 0.7 Ga. U-Pb and Sm-Nd isochron ages of 940 ± 5 and 934 ± 9 Ma, respectively, with sNd = -8.5, suggest magma derivation from a LREEenriched crustal protolith with a crustal residence age of 1.8 to 2.5 Ga. A Rb-Sr age of 1635 ± 67 Ma (IR = 0.715 ± 0.01) provides additional support for derived TDM model ages. In conclusion, it can be assumed that the supercontinent Columbia accreted between ca. 1.9 and 1.5 Ga. One suture lay between western and northern Laurentia and an orogenic belt along the margins of eastern India, western Australia, and attached parts of coastal East Antarctica. Columbia must have formed along at least two major accretionary zones, somewhat earlier than 1.5 Ga, perhaps beginning about 1.9 Ga in some areas. In present coordinates, one was between western North America (part of Nena) and eastern India, western Australia, and attached East Antarctica (part of expanded Ur). The eastern margin of India was well defined by the presence of the Singhbhum orogen beneath the CGGC. In this context, one is tempted to conclude that the CGGC, within the tectonic mosaic of peninsular and central India, formed part of the Columbia supercontinent. Separation, beginning at -1.6 Ga, formed correlative rift valleys in India and Laurentia and ultimately led to fragmentation of Columbia and reorganization to form Rodinia.

Supercontinents and Earth Evolution Symposium 2005

LATE MESOPROTEROZOIC (1.65 Ga) PALAEOMAGNETIC RECORD FROM DYKES AROUND CUDDAPAH BASIN, INDIA: IMPLICATIONS FOR TRACING THE SOUTH INDIAN SHIELD IN THE GLOBAL CONTINENTAL ASSEMBLY Tallavaihala Radhakrishna, NR Krishnendu, G Balasubramonian Centre for Earth Science Studies, Trivandrum 695 031, India (tradha1@rediffmail.conn)

The Cuddapah Basin of southern India, elongated N-S for 400 km and 150 km wide, is one of the world's largest Proterozoic sedimentary basins. Profuse dyke magmatism occurred all around the convex western basin boundary within the 2.6 Ga eastern Granite Greenstone Complex. These dykes are the main source of information for understanding Palaeoproterozoic mantle processes and the geodynamic evolution of the south Indian shield and for tracing its location within the early global continental reconstructions. However, the available palaeomagnetic data from these dykes are inadequate, in terms of number of samples or detailed demagnetizations, to isolate reliable Characteristic Remanent Magnetisations (ChRM). The geochemical studies are restricted to characterise magma types essentially based on major elements. Our investigations on these dykes involve 850 samples from 85 sites (dykes) for palaeomagnetism. Trace element geochemistry was conducted on over 200 samples. One of the significant observations is that a steep ChRM was isolated in 18 sites from all around the Cuddapah Basin although its statistical abundance is quite low (one in eleven dykes obtained ChRM grouping) on the extreme southern border. This direction compares and correlates very well with steep directions determined in our earlier studies on dykes in other parts of the Dharwar craton and also from the dykes in the transition zone granulites of Tiruvannamalai area. Combining all these data yields a precise pole (Long. = 254°; Lat. = 2°; A95 = 7.4°; n = 72 dykes) for the south Indian shield. Although reliable U-Pb, Sm-Nd or Rb-Sr isotopic ages are lacking, a coherent group of 13 of 19 K-Ar whole rock ages (younger ages are virtually absent) indicate a 1.65 Ga age for this pole and requires revision of earlier imprecise estimates of 2.4 Ga, based on Rb-Sr whole rock data. Although the pole overlaps the younger 800 Ma Harohalli pole and also the 1.0 Ga pole

from kimberlite, these younger thermal events are totally absent in the entire area of dyke sampling in south Indian shield, except in Harohalli area. Furthermore, where 800 Ma alkaline magmatism is widespread (Dharmapuri area), the dykes registered the thermal event in the Ar-Ar age spectrum, but the ChRMs differ considerably from those of the corresponding younger age. These results testify that the ChRM directions recorded by these dykes reflect primary magnetizations. The remarkable agreement and correlation between the ChRMs of these dykes across a large area straddling the eastern and western Granite Greenstone regions of the Dharwar craton and into granulite region in the south allow us to conclude that the terrains had united and retained their relative proximity without large scale displacements at least since late Palaeoproterozoic times (1.65 Ga). The reported 1.65 Ga palaeomagnetic pole is the first major quantitative result to trace the Indian shield within Palaeoproterozoic global reconstructions. Palaeolatitude estimates from the 1.65 Ga dykes of south India are compatible with the suggestion of proximity of the Australia and Indian shield as predicted by the "Ur" proposition at the end of the Paleoproterozoic (1.65 Ga). However, the geological record has been recently interpreted to suggest distal positions of Pilbara and Kaapvaal cratons with respect to India in early Palaeoproterozoic times. Whereas our palaeomagnetic results from the older dykes around Cuddapah basin will be useful to ascertain these configurations, in the light of these interpretations, we suggest that the Indian shield might have amalgamated into an "Ur" configuration during assembly of a Hudsonland/Nuna supercontinent during 2.0-1.8 Ga orogenic events. The widespread 1.65 Ga dyke magmatism in south India, however, suggests major plume activity causing the rifting and dispersal of supercontinental fragments that later accreted to Rodinia around 1 Ga.

Supercontinents and Earth Evolution Symposium 2005

DIAMICTITE AT THE BASE OF THE SAUSAR GROUP, INDIA: IMPLICATIONS FOR PALAEOPROTEROZOIC SUPERCONTINENT ASSEMBLY Saradaprasad Mohantv\ Anup K Prasad^, Aruna K Sahu^ ^Department of Applied Geology, Indian School of Mines, Dhanbad, 826004, India (nriohantysp@yahoo.com) Indian Institute of Technology, Kanpur, India ^Oil and Natural Gas Corporation, Dehra Dun, India

The Paleoproterozoic Sausar Group of central India has unconformable relation with Tirodi gneiss (Archean basement gneiss). Although involved in superposed deformation and metamorphism in amphibolite fades, sedimentological features of the Sausar Group are well preserved. Glaciogenic origin of diamictites in the lower part of the Sausar Group is attested by the presence of a polished and striated basement, poorly sorted clasts of different shapes, sizes and lithology, and association of dropstones and splash up structures. The glaciogenic rocks occur as discontinuous lenses along the basement contact and represent proximal glaciomarine facies. Wherever the glaciogenic rocks are

absent a carbonate unit comes in direct contact with the basement. Deglaciation was responsible for marine transgression and deposition of the carbonate unit (as a cap carbonate), followed by fine clastics with manganese ore. The lithological association is comparable with the enigmatic carbonate-tillite association of the Huronian Supergroup of Canada. Equivalent stratigraphic positions and ages of the Aravalli SupergroupSausar Group-Gwalior Group of India with the Hurwitz Group and Huronian Supergroup, and occurrence of glaciogenic rocks in the Sausar Group, prompts us to speculate for an unified Paleoproterozoic assembly for these provinces.

Siipercontinents and Earth Evolution Symposium 2005

ASSEMBLY, GROWTH AND BREAKUP OF A PALEOMESOPROTEROZOIC SUPERCONTINENT Guochun Zhao\ Simon A Wilde^ Min Sun\ Sanzhong Lp ^Department of Earth Sciences, The University of Hong Kong, Pokfulam Road, Hong Kong (gzhao@hkucc.hku.hk) Department of Applied Geology, Curtin University of Technology, GPO Box U1987, Perth, WA 6845, Australia ^College of Earth Sciences, Ocean University of China, Qingdao, 266603, China

Geological and paleomagnetic data support the hypothesis that a Paleo-Mesoproterozoic supercontinent, referred to as Columbia, existed before the formation of Rodinia. This pre-Rodinia supercontinent was assembled along globalscale 2.1-1.8 Ga collisional orogens and contained almost all of Earth's continental blocks. Following its final assembly at -1.8 Ga, the supercontinent Columbia underwent longlived (1.8-1.3 Ga), subduction-related growth via accretion at key continental margins, forming a 1.8-1.3 Ga large magmatic accretionary belt along the present-day southern margin of North America, Greenland and Baltica. It includes the 1.8-1.7 Ga Yavapai, Central Plains and Makkovikian Belts, 1.7-1.6 Ga Mazatzal and Labradorian Belts, 1.5-1.3 Ga St. Francois and Spavinaw Belts and 1.3-1.2 Ga Elzevirian Belt in North America; the 1.8-1.7 Ga Ketilidian Belt in Greenland; and the 1.8-1.7 Transscandinavian Igneous Belt, 1.7-1.6 Ga Kongsberggian-Gothian Belt, and 1.5-1.3 Ga Southwest Sweden Granitoid Belt in Baltica. Other cratonic blocks also underwent marginal outgrowth at about the same time. In South America, a 1.8-1.3 Ga accretionary zone occurs along the western margin of the Amazonia Craton, represented by the Rio Negro, Juruena and Rondonian Belts. In Australia, 1.81.5 Ga accretionary magmatic belts, including the Arunta, Mt. Isa, George-town, Coen and Broken Hill Belts, occur surrounding the southern and eastern margins of the North Australia

Craton and the eastern margin of the Gawler Craton. In China, a 1.8-1.4 Ga accretionary magmatic zone, called the Xiong'er belt (Group), extends along the southern margin of the North China Craton. Fragmentation of this supercontinent began about 1.6 Ga ago, associated with continental rifting along the western margin of Laurentia (Belt-Purcell Supergroup), southern margin of Baltica (Telemark Supergroup), southeastern margin of Siberia (Riphean aulacogens), northwestern margin of South Africa (Kalahari Copper Belt), and northern margin of North China (Zhaertai-Bayan Obo Belt). The fragmentation corresponded with widespread anorogenic magmatic activity, forming anorthosite-mangerite-charnockite-granite (AMCG) suites in North America, Baltica, Amazonia and North China, and continued until the final breakup of the supercontinent at about 1.3-1.2 Ga, marked by the emplacement of the 1.27 Ga Mackenzie and 1.24 Ga Sudbury mafic dike swarms in North America. This research was financially supported by Hong Kong RGC Grants (7055/03P, 7048/03P and 7058/04P) and HKU Seed Funding for Basic Research Program (200411159122).

Supercontinents and Earth Evolution Symposium 2005

TUESDAY 27 SEPTEMBER 2005

Supercontinents and Earth Evolution Symposium 2005

PRECAMBRIAN GEOMAGNETIC FIELD: INTENSITY, MORPHOLOGY AND SECULAR VARIATION Alexei Smirnov, David AD Evans Department of Geology and Geophysics, Yale University, PO Box 208109, New Haven, CT, 06520, USA (aleksey.smirnov@yale.edu)

Paleomagnetic data have been instrumental in studying Precambrian geological history (e.g. continental reconstructions, evolution of supercontinents, and true polar wander). Interpretation of the data, however, is ultimately based on our understanding of the characteristics of geomagnetic field during that time period. Unfortunately, our knowledge of the Precambrian geomagnetic field is far from complete. However, for at least part of the Precambrian period, the space-time characteristics of the field could have been significantly different from their Phanerozoic counterparts (e.g., Kent and Smethurst, 1998). Theoretically, the transition from the primordial to modern compositionally-driven geodynamo could have occurred any time during a broad interval, with estimates ranging from the Archean through Paleozoic (e.g., Bloxham, 2000; Labrosse, 2003). While the timing and duration of this event is unknown, it raises questions about the nature of the field before and during the transition. One important problem here is the relative significance of the dipole and non-dipole components of the geomagnetic field throughout Precambrian time. Another problem is the field stability; it has been suggested that the hotter Earth and the absence of the inner core could have resulted in higher variation of the field, including more frequent excursions (e.g. Gubbins, 1999). In the absence of strict theoretical constraints, paleomagnetic data become a principal source of information about the Precambrian field. Paleointensity data may provide an important time constraint on the onset of the modern geodynamo, which could have been accompanied by a sharp increase in the field strength. However, many Precambrian rock sequences have been affected by alteration, which hinders the preservation and measurement of paleointensity using bulk rock samples. Alternatively, single feldspar crystals may be used as paleointensity recorders (e.g. Cottrell and Tarduno, 1999). Such crystals are

much less susceptible to alteration in nature and during paleointensity experiments. This approach was applied to plagioclase crystals derived from --2.45 Ga border dykes of the Burakovka layered intrusion (Russia). Fifteen determinations from 4 dykes yielded an average virtual dipole moment of 8.43 ±2.11 • 10^^ Am^ (Smirnov et al., 2003). Although this result is unlikely to adequately represent the time-averaged field, the mean and range of values are similar to those of the present-day field. The coeval Matachewan dykes (Canada) are, in general, more altered than the Karelian rocks, yet preserve directional information useful for constraining paleosecular variation (PSV). Preliminary results indicated a variation that is not grossly different from that of the present-day field (Smirnov and Tarduno, 2004). Hence, the available paleointensity and secular variation data are consistent with a field through the Archean-Proterozoic transition that is similar to that of the modern field. We will discuss estimates of PSV for other time intervals in the Precambrian. Some recent efforts to model the geodynamo using a small inner core yielded a geomagnetic field with a stronger dipole component than today's field (e.g. Glatzmaier and Roberts, 1995). On the other hand, relatively large octupole contributions to the geomagnetic field have been suggested based on the distribution of Precambrian inclinations (e.g., Kent and Smethurst, 1998). Important inferences on this problem can be derived from both paleodirectional and paleointensity data. For example, coeval values for sites formed at different paleolatitudes can be compared with theoretical models of the latitudinal dependence of the field, developed for different dipole/nondipole ratios (e.g. Merrill et al., 1996). Similar tests can be performed on dykes characterized by long (several hundreds of kilometers) paleolongitudinal extent, such as the --500 km Binneringie dyke of the Yilgarn craton, or the -530 km Great Dyke of Zimbabwe.

Supercontinents and Earth Evolution Symposium 2005

UPDATED PALAEOPROTEROZOIC APPARENT POLAR WANDER PATHS FOR LAURENTIA AND AUSTRALIA AND IMPLICATIONS FOR THE PALAEOGEOGRAPHY OF NUNA Theresa MP Raub\ David AD Evans\ Robert H Rainbird^ Vale University, PO Box 208109, New Haven, CT, 06520, USA (theresa.raub@yale.edu) ^Geological Survey of Canada, 601 Booth St., Ottawa, Ontario, K1A0E8, Canada

Palaeogeographic reconstructions of the Palaeoproterozoic supercontinent Nuna currently are based either on elements of poorly constrained Rodinia reconstructions or on comparisons of single palaeopoles from multiple continents. New palaeomagnetic results from the Dubawnt Supergroup, Canada, and the Lawn Hill Platform, Australia, provide the opportunity to increase both the resolution and reliability of the Palaeoproterozoic apparent polar wander paths (APWPs) for Laurentia and Australia, allowing for better constraints on their relative palaeogeography in the Proterozoic. Contradicting data exist in the ^1.9-1.6 Ga Laurentian palaeomagnetic data set. TransHudson Orogen (THO) palaeomagnetic studies yield higher-latitude palaeopoles than that obtained in the 1970s from the Dubawnt Supergroup in the Western Churchill Province. A critical review of the THO data reveals that the ages of magnetisation in these rocks are up to 100 million years younger than originally presumed. Streaked palaeomagnetic data from many studies indicate mixing of components and

likely varying amounts of overprinting throughout the orogen. Eight hundred samples spanning deposition of the Dubawnt Supergroup offer the chance to test the reliability of the original Dubawnt pole and obtain a new stratigraphic succession of poles constrained by abundant palaeomagnetic field tests on red beds and volcanic rocks throughout Baker Lake Basin. Two rhyolite flows (--1755 Ma) from the Pitz Formation in the Wharton Group yield distinct directions. The VGP calculated from the lower flow falls close to the 1740 Ma Cleaver dykes pole but does not overlap it, yet it does overlap the Athabasca sandstone 'A' pole. Along with preliminary results from the correlative Thelon sandstone, this suggests that a diagenetic fluid-flow event may be the cause of overprinting in the Pitz Formation. Increased understanding of magnetisation ages and diagenetic history in the Baker Lake Basin gained from these results allows for comparison with an updated Australian APWP and increased understanding of Palaeoproterozoic palaeogeography.

Supercontinents and Earth Evolution Symposium 2005

PALEOMAGNETISM OF THE LOWER WATERBERG GROUP OF SOUTH AFRICA: TOWARDS A BETTER DEFINED APPARENT POLAR WANDER PATH FOR THE PALEOPROTEROZOIC KAAPVAAL CRATON Michiel O de Kock\ HC Dorland\ NJ Beukes\ DAD Evans^ ^Department of Geology, University of Johannesburg (Aucklandpark Kingsway Campus), PO Box 524, Aucklandpark, 2006, South Africa (email) ^Department of Geology and Geophysics, Yale University, PO Box 208109, New Haven, CT 065208109, USA

The Waterberg Group in South Africa (WG hereafter) is an erosional remnant of a once extensive redbed succession that developed during the Paleoproterozoic on the Kaapvaal craton. Well-defined paleomagnetic poles are few and far between when it comes to the Apparent Polar Wander Path (APWP) of the Paleoproterozoic Kaapvaal craton, and the few poles that do exist are often the subject of much debate. Previous paleomagnetic work on the WG is hampered by the fact that blanket demagnetization methods were employed, they lack statistically robust field tests and until recently high uncertainties existed in the numerical age of the WG. Within the last year though, a precise zircon SHRIMP U-Pb age of 2054±4 Ma has been reported from quartz porphyritic lava from near the base of the WG. We studied the Swaershoek Formation of the Lower WG with the goal in mind to define a new paleomagnetic pole for the Kaapvaal craton at -2.05 Ga. RESULTS AND DISCUSSION Three ancient magnetic components were observed. Two mesoscale and one megascale fold tests indicate that the first of these were acquired after folding of the Lower WG. Similarity to directions observed (during this study) within a dolerite sill that cuts the sequence at Bakers Pass and directions obtained recently from other post-Waterberg dolerites suggest magnetization at -1.87 Ga during magmatism associated with the development of the Soutpansberg basin. The second component to unblock was also observed as a component of magnetization of the post-Waterberg dolerite at Bakkers Pass. It is directed steep upwards in a southwesterly

direction, but (regarding its origins) remains unidentified as yet. The most stable component unblocks above 590°C and was observed in 57 samples from 8 sampling sites. This component, directed NE at a moderate angle downwards, pre-dates the folding of the Lower WG (Indicated by positive meso- and megascale fold tests, with best groupings achieved at 80-85% unfolding.). It is therefore believed to represent a primary magnetization. Six out the eight Swaershoek sampling sites were used to calculate the pole position (Only those sites with a95<20°and k>10 were used.). The -2.05 Ga Lower WG pole (21.1°N, 077.5°E, K = 18.90, A95 = 18.00, Q = 5/7) is more or less consistent with previous work, but since blanket demagnetization methods were often used our results were not combined with existing data. The Lower WG pole is also relatively consistent with poles obtained from the Bushveld Complex, which is in spite of high variability in obtained ages considered to have intruded the Kaapvaal craton between 2.06-2.05 Ga. A wide distribution of poles creates further doubt as to how well the Bushveld Complex pole can be established at present. In spite of this the poles of the Bushveld Complex has been widely used in the development of the Paleoproterozoic APWP of the Kaapvaal craton. The validity of which can now be tested and or contested. CONCLUSION The new -2.05 Ga Lower WG pole might be a better representation of the position of the Kaapvaal craton during the closing stages of the intrusion of the Bushveld Complex, and provides a step towards a better constrained APWP for southern Africa in the Late Paleoproterozoic.

Supercontinents and Earth Evolution Symposium 2005

PALEOMAGNETIC AND ROCK MAGNETIC DATA FROM VALAAM SILL, LAKE LADOGA, RUSSIAN KARELIA: IMPLICATIONS FOR SUPERCONTINENT HUDSONLAND Johanna Salminen, Lauri J Pesonen Division of Geophysics, PO Box 64, 00014 University of Helsinki, Finland

It has been proposed that Laurentia and Baltica remained intact for more than 600 million years during 1.83 - 1.26 Ga forming the core of the supercontinent Hudsonland. To test this hypothesis paleomagnetic data are needed from this age interval from both continents. The Valaam gabbro-monzonite-syenite sill in Lake Ladoga, Russian Karelia, is most suitable for this purpose since its age of 1459 ± 3 Ma matches well with that of the St. Francois igneous rocks (1.47 Ga) of Laurentia, and because the sill is in its original nearly horizontal position. We present preliminary paleomagnetic and rock magnetic results of the Valaam sill as based on 19 (out of 46) samples from various phases of the sill, collected during the 2003 and 2004 field seasons. MEASUREMENTS Paleomagnetic and rock magnetic measurements were carried out at the Solid Earth Geophysics Laboratory of the University of Helsinki (Finland). Each specimen was demagnetized by alternating fields up to 160 mT. The remanence components were isolated with multicomponent analysing techniques. Hysteresis properties were measured using a vibrating sample magnetometer and thermomagnetic measurements were carried out using a KLY-3 kappabridge. Petrophysical properties (susceptibility and density) were measured using standard equipments. RESULTS Density, susceptibility and NRM values of gabbro-monzonite are higher than in syenites.

Moderate Q-values (gabbro-monzonites 0.512.31, syenites ca. 1.24) suggest that the remanence is carried by large PSD grains. Hysteresis data also point to PSD grain sizes, close to MD area. The slightly irreversible thermomagnetic curves indicate that at least three magnetic phases (titanomagnetite, nearly pure magnetite and hematite) are present. The sill carries a high coercivity remanent magnetization (D = 45.3°, I = -17.7°, k = 485; ags = 3.5°, 19 samples), which is regarded as primary, although no field test is yet available. The majority of the samples show also two low coercivity components. The viscous component (D = -58°, I = -75°) is interpreted to be caused by the present Earth's magnetic field, whereas the other one (D = 44.4°, I = 27.6°, k = 76; ags = 10.6°, 13 samples) gives a paleomagnetic pole at Lat. = -33.7°, Long. = 156.0°, dp = 6.3° and dm = 11.6° and suggests a Permian (?) overprint. The high coercivity component yields a pole Lat. = 11°, Long. = 165°, dp = 2°, dm = 4° placing Baltica at equatorial latitudes of -10°S. The results suggest a magnetization age between Subjotnian (1.60 Ga) and Jotnian (1.27 Ga), based on the APWP of Baltica. CONCLUSION Results from this study places Baltica in low latitudes at 1.46 Ga. We will discuss the implications of this result in terms of the Hudsonland supercontinent and the positions of Laurentia and Baltica at ca. 1.46 Ga.

Supercontinents and Earth Evolution Symposium 2005

ASSEMBLING AUSTRALIA Russell J Korsch, Natalie Kositcin Geoscience Australia, GPO Box 378, Canberra ACT 2601 Australia (Russell.Korsch@ga.gov.au)

Fundamentally, the geology of Australia consists of an Archaean western part, a Proterozoic central part and a Phanerozoic eastern part that have been assembled progressively over a period of greater than 2.6 billion years to form the present continent of Australia. A key element in understanding this evolution is the Proterozoic, because it forms the "glue" between the Archaean cratons in the west and the eastern Phanerozoic accretionary orogens that are related to convergent plate margin processes. In terms of the assembly during the Proterozoic, there is still much debate on the interpretation of intracratonic settings versus accretionary or collisional orogens. Much of this relates to potential plate margin processes that result in intraplate events that manifest themselves in a variety of ways. Thus, fundamental questions can be asked: Which blocks were amalgamated and at what time? How far apart were the various blocks prior to amalgamation? What was the timing of accretion of these blocks to ProtoAustralia? What is the nature of the boundaries (sutures) of the blocks? What can be inferred about the operating tectonic processes? Our approach is to use time-space plots, constructed in a consistent fashion, to provide a realistic comparison between the various Proterozoic elements. Within Geoscience Australia, a database is being developed to store interpretations on geodynamic settings and geological events, with links to track back to the raw data. This information allows the dynamic generation of time-space plots, including displaying the tectonic settings of the blocks, and the present day distribution and geometry of the blocks can be compiled. Palinspastic reconstructions for the times of assembly, constrained if possible by palaeomagnetic data, would further enhance tectonic interpretations. Using time-space plots and APWP data, several key intervals of time for deformational and/or magmatic events in Proterozoic Australia appear to be 1880-1860 Ma, 1820-1800 Ma, -1740 Ma, 1700-1690 Ma, -1670 Ma, -1640 Ma, 16001590 Ma and-1550 Ma. Another way of contributing to our understanding on the assembly of Australia is to use deep seismic reflection profiles. These provide information on the geometry, in the third

dimension, of the suture zones and boundaries, and also on the internal geometry of the Proterozoic building blocks. For example, based on crustal reflectivity patterns, we are able to partition the crust in the Arunta and Musgrave provinces in central Australia into several faultbounded packages which are cut by planar, trans-crustal north-dipping thrusts, many supporting large hanging wall anticlines. Some of the key faults forming the boundaries to the packages are the Ormiston Thrust Zone, the Redbank Thrust Zone and the Desert Bore Thrust. At Mount Isa, the deep seismic transect show that the province consists of three distinct crustal regions: a central buttress block beneath the Kalkadoon Block, an eastern thin-skinned region corresponding to the Eastern Succession and a thick-skinned Western Succession. At Broken Hill, major D3 structures imaged in the seismic data, including the retrograde shear zones, dip southeast rather than to the northwest. Many extend to middle and lower crustal levels and probably served as important conduits for fluid flow. Some upper crustal shear zones south of Broken Hill are listric and sole into detachments in the mid-crust. Seismic reflection data from the Olympic Dam region of the Gawler Craton show that the ArchaeanProterozoic basement is dominated by fold and thrust belt structures, with thrust-wedges in the upper and lower crust and duplexing in a midcrustal layer. In the southern McArthur Basin, the Tawallah Fault is a west-dipping thrust, forming part of a major east-directed fold-thrust belt that was unknown prior to the acquisition of the seismic data. Uplift of the Roper Group along thrust ramps indicates that this event is younger than 1490 Ma. The deep seismic data also allow some of Australia's world class mineral deposits to be placed in a whole-of-crust context. Results from Mount Isa, McArthur River, Broken Hill and Olympic Dam are challenging our current understanding of the geology, ore deposit models and mineral potential of the Proterozoic of Australia. In summary, both time-space plots and deep seismic reflection data from critical areas are contributing to our understanding of the assembly of Proterozoic Australia.

Supercontinents and Earth Evolution Symposium 2005

CONTRASTING TECTONIC HISTORIES OF AUSTRALIA'S PILBARA, YILGARN, AND GAWLER CRATONS: KEY PIECES OF THE LATE ARCHEAN TO EARLY PALAEOPROTEROZOIC TECTONIC PUZZLE Mark Barlev\ Bryan Krapez\ Greg Swain^ ^School of Earth and Geographical Sciences, The University of Western Australia, 35 Stirling Highway, Crawley WA 6009, Australia (mbarley@segs.uwa.edu.au) ^School of Earth and Environmental Sciences, The University of Adelaide, SA 5005, Australia

The late Archaean record comprises 35 cratons described as "a puzzle in 35 pieces" by Wouter Bleeker in his 2003 review. Most of these display Proterozoic rifted margins suggesting that they were fragments of larger continents or a Late Archaean supercontinent. Although it is possible to group cratons with similar tectonic histories, fundamental differences between the tectonic histories of some of the better known cratons suggest they evolved separately. Consequently a more complete history of Late Archaean tectonics may be recorded by the contemporary, but contrasting tectonic regimes preserved by individual cratons or groups of cratons. Australia's Pilbara, Yilgarn and Gawler cratons have contrasting tectonic histories that record different tectonic environments and stages of a full Late Archaean global tectonic cycle. The Pilbara Craton acted as stable continental lithosphere by 2.8 Ga and may have been part of the continent Vaalbara. The Kaapvaal Craton has a similar 2.8 to 2.6 Ga history and together these cratons provide evidence for Late Archaean continental rifting and breakup enhanced by mantle plume magmatism. The 2.59 to 2.40 Ga tectonic histories of these cratons reflect the conversion from passive margins of an internal ocean, with deposition of banded iron formations during a period of mantle plume activity, to foreland basin sedimentation culminating in collision with other cratons or terranes and continental stability. Granitoid-greenstone terranes world-wide record one of the most prodigious periods of generation and stabilisation of new continental crust preserved in the geological record, between 2.8 and 2.6 Ga. The eastern Yilgarn Craton, and many other Late Archaean terranes, show histories of magmatic arc and mantle

plume magmatism and associated sedimentation culminating in orogeny, widespread granitoid emplacement and terrane accretion, that parallel those of marginal basins to the Pacific during the Mesozoic breakup of Pangea and Cretaceous mantle plume activity. The formation of Kenorland by -2.6 Ga is evidence that cratons started to aggregate at that time. The Gawler Craton, provides evidence for a second cycle of convergent margin tectonics and collision between 2.6 and -2.4 Ga. The Gawler Craton contains 2.56 to 2.5 Ga ultramafic to felsic volcanic rocks (including -2.51 Ga komatiites), metasedimentary rocks, and granitoids with compositions that are typical of Archean granitoid-greenstone terranes interpreted to have formed at convergent continental margins. These were deformed, intruded by granitoids and metamorphosed to high grade (up to granulite fades) during the 2.48 to 2.42 Ga Sleafordian orogenic cycle, that also affected equivalent rocks in east Antarctica. Terranes in east Antarctica, India and China have similar histories culminating with orogeny and high-grade metamorphism after 2.5 Ga corresponding to the aggregation of Indian cratons within a larger continent. The tectonic histories of the Pilbara, Yilgarn and Gawler cratons represent the evolution of different tectonic environments and stages of a full 2.8 to 2.4 Ga global tectonic cycle involving the breakup of a pre-existing continent accompanied by growth and aggregation of cratons to form new continents and followed by continued convergence and collision of cratons and continents culminating in the possible formation of the Earth's first supercontinent by 2.4 Ga. The Earth's first widespread glaciation and oxidation of the atmosphere accompanied supercontinent formation.

Supercontinents and Earth Evolution Symposium 2005

PROTEROZOIC DEFORMATION OF THE NORTHWESTERN YILGARN CRATON, WESTERN AUSTRALIA Catherine V Spaqgiari Geological Survey of Western Australia, 100 Plain Street, East Perth, WA 6004, Australia (Catherine.Spaggiari@doir.wa.gov.au)

The tectonic evolution of the Yilgarn Craton is generally thought of in ternns of Archaean granite-greenstone formation, with little Proterozoic modification except for on its extreme margins (e.g. Errabiddy Shear Zone, Albany-Fraser Orogen, Darling Fault). Recent work in the Jack Hills Belt shows that Proterozoic tectonics has played a major role in shaping the northwestern part of the craton. The Jack Hills Belt is a greenstonemetasedimentary belt located within the southern part of the Narryer Terrane, one of Earth's oldest crustal fragments. It lies adjacent to the Murchison Terrane to the south, which forms the west-central part of the craton. The nature of the boundary between the two terranes is not clear, particularly in terms of location and tectonic evolution, and it is even questionable as to whether the terrane concept is appropriate. This study is currently examining structural relationships between the two terranes in an effort to understand the extent of Proterozoic reworking, and to determine whether there are major regional differences in tectonic evolution. This has implications for understanding Archaean geology, such as the source and distribution of >4.0 Ga detrital zircons, for which the Jack Hills Belt is famous, and which are now known to exist to the southeast in the Southern Cross Terrane. Understanding ore mineral potential in this relatively under-explored part of the craton is also a fundamental issue being addressed. RESULTS AND DISCUSSION There are distinct differences between the Jack Hills Belt (and southern Narryer Terrane) and the Murchison Terrane. Much of the sequence within the Jack Hills Belt occurs as fault lenses on scales up to 10s of km in length, with some relatively more complete sections preserved in lower strain zones. Detailed fieldwork coupled with satellite imagery and geophysical data interpretation show that the Jack Hills Belt has undergone a long and complex deformation history, and has been cut by major, pre-

dominantly east-west transpressive shearing. The belt as a whole has a pronounced sigmoidal curvature suggestive of dextral movement, and the majority of kinematic indicators in the field match this interpretation. Ar-Ar dating of white micas indicates that this shear-related deformation occurred during the Capricorn Orogeny (-1830-1780 Ma), a widespread event that recorded the final amalgamation of the Pilbara and Yilgarn Cratons. Earlier deformation is only present in some of the sequence, and it is now clear that not all the of rocks in the belt are Archaean. Shearing in the belt is overprinted by low grade, predominantly brittle deformation, possibly related to the early stages of the Edmundian Orogeny. The main deformation is also cut by gabbroic and doleritic dikes dated at --1200 and 1075 Ma, respectively (MTD Wingate, pers. comm., 2004). Thus the Jack Hills records a particularly long geological history spanning at least 2 billion years. The Proterozoic shear-dominated deformation is not present in the Weld Range greenstone belt, located -100 km south. The Weld Range has the same overall east-northeast trend as the Jack Hills Belt, but is truncated at its northeastern end by the Meekatharra Belt. The Meekatharra Belt is a wide north-northeast trending belt that includes at least one major zone of shearing with complex kinematics, and probably others that extend southwards for at least 200 km. Given the structural differences in these belts, it is possible that Proterozoic deformation related to the Capricorn Orogeny only extends as far south as the Jack Hills Belt, and possibly marks a major Palaeoproterozoic fault contact with the Murchison Terrane. However the age and nature of deformation in the Meekatharra Belt is poorly constrained, and may also contain relatively young elements. CONCLUSION This study has shown that Proterozoic deformation extends further south than previously recognised, and has played a major role in northern Yilgarn Craton tectonics.

Supercontinents and Earth Evolution Symposium 2005

DOES THE CA. 1800 Ma CAPRICORN OROGENY MARK COLLISION OF THE YILGARN AND PILBARA CRATONS? Steve Sheppard Geological Survey of Western Australia, 100 Plain Street, East Perth, WA 6004, Australia (steve.sheppard@doir.wa.gov.au)

Orogenic activity and magmatisnn in nnany parts of the world at 1850-1800 Ma reflect annalgannation of nunnerous cratonic fragments (e.g. Trans-Hudson Orogen, Svecofennian). Orogenic belts of this age are common in Australia, and include the Capricorn Orogen in Western Australia. The Capricorn Orogen has juxtaposed two distinctly different Archaean cratons, both of which are exposed. The two cratons, Yilgarn and Pilbara, are widely regarded as having collided during the 1830-1780 Ma Capricorn Orogeny; however, views on the timing of rifting and convergence, and the polarity of subduction vary widely, in part because these models are based on interpretations of poorly dated metasedimentary successions in the northern part of the orogen. Few, if any, models have taken account of the medium- to high-grade metamorphic rocks and granites of the Gascoyne Complex in the western part of the orogen. Structures, metamorphic assemblages, and granite intrusions throughout the western part of the Capricorn Orogen, including the Gascoyne Complex, generally have been attributed to the Capricorn Orogeny. However, recent U-Pb SHRIMP geochronological studies show that the Gascoyne Complex was shaped by four separate orogenic events: the 2005-1970 Ma Glenburgh Orogeny, the 1830-1780 Ma Capricorn Orogeny, the 1680-1620 Ma Mangaroon Orogeny, and the Neoproterozoic Edmundian Orogeny. The Glenburgh Orogeny is interpreted to reflect an Andean-type margin along the southern margin of the Gascoyne Complex and its subsequent collision with the Yilgarn Craton. Rocks older than 1970 Ma and the effects of the Glenburgh Orogeny are confined at surface to the southernmost zone of the complex. It is possible that these rocks extend farther north as basement to younger rock packages. Deformation, metamorphism, and extensive granitic intrusions related to intracratonic reworking during the Mangaroon Orogeny are

also developed across much of the Gascoyne Complex. Confusion over the tectonic history of the Gascoyne Complex is in part attributable to the linking together of structures formed during the separate Capricorn and Mangaroon orogenies. Nowhere in the Gascoyne Complex are igneous rocks present that could be interpreted as either an island arc or continental arc that evolved before a postulated continent-continent collision during the 1830-1780 Ma Capricorn Orogeny. Indeed, there are no rocks with depositional or crystallization ages between c. 1960 and c. 1830 Ma. The paucity or absence of volcanic rocks does not reflect the current level of exposure, because metasedimentary rocks are abundant throughout the complex. Granitic rocks intruded during the Capricorn Orogeny are either silicic l-types formed by remelting of older crust, or S-types. These compositions, and the lack of associated gabbros, distinguish the granites from batholiths formed at Andean-type margins and at island arcs. Therefore, it is very unlikely that granites intruded during the Capricorn Orogeny reflect subduction of oceanic crust. The 'Capricorn' granites in the Gascoyne Complex have broad similarities with Phanerozoic granites interpreted to have formed in post-collisional settings (e.g. Hercynides, Caledonides, Lachlan Fold Belt); however, the Gascoyne Complex, unlike these Phanerozoic orogens, apparently had no prior convergent margin activity. If there was no subduction prior to, and during, the Capricorn Orogeny, then the Yilgarn and Pilbara Cratons must have been more or less in their current relative positions (other than probable strike-slip movement during the Capricorn and Mangaroon orogenies). If this is correct, then the earlier Glenburgh Orogeny, rather than marking accretion of a Gascoyne microcontinent to the Yilgarn Craton, may record collision of a combined Pilbara-Gascoyne ('Pilboyne') Craton with the Yilgarn Craton.

Supercontinents and Earth Evolution Symposium 2005

METAMORPHIC EVOLUTION OF THE NORTHWESTERN CAPRICORN OROGEN, WESTERN AUSTRALIA Hideki Masago\ Shigenori Maruyama^ ^Center for Deep Earth Exploration, Japan Agency for Marine-Earth Science and Technology, 3173-25 Showa-machi, Kanazawa-ku, Yokohama, Kanagawa 236-0001, Japan (masagoh@jamstec.go.jp) ^Department of Earth and Planetary Sciences, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro-ku, Tokyo 152-8551, Japan

INTRODUCTION The early Proterozoic Capricorn orogen, located between the Pilbara and the Yilgarn cratons in western Australia, is regarded as one of the oldest continental collision zones in the world. The essence of orogeny is denudation of the regional metamorphic belt. Nevertheless, in the Capricorn orogen, its metamorphic aspects are relatively poorly studied compared to the detailed structural and geochronological studies. GEOLOGY We have studied a 80 km (E-W) x 100 km (N-S) area in the northwestern Capricorn orogen covering the northern Gascoyne Complex and the southern Ashburton Basin. We have investigated this area total 50 days, and have collected about 2,000 samples throughout the area for petrological study. The Gascoyne Complex in the studied area is composed mainly of coarse-grained arkosic gneiss with minor politic gneiss and amphibolite. The Ashburton Group is composed of schists of turbidite origin, including psammitic, politic and conglomerate schists. A NW-SE striking, steeply-dipping foliation is well developed throughout the area. Mineral lineation is poorly developed suggesting a flattening-dominant strain regime in general, except in the vicinities of the boundary between the Gascoyne Complex and the Ashburton Group, and the southern margin of the Gascoyne Complex. The boundary between the Gascoyne Complex and the Ashburton Group is mostly covered by mid-Proterozoic rift sediments of the Bangemall Supergroup. Although the exact contact of these two bodies could not found, both the Gascoyne Complex and the Ashburton Group rocks have undergone cataclasis near the boundary suggesting the boundary is a fault. METAMORPHIC ZONATION Six representative mineral assemblages were observed in the Gascoyne Complex gneiss by

microscopic observation: 1) Bt + Ms, 2) Bt + Ms + Grt, 3) Bt + Ms + Grt + Sil, 4) Bt + Ms + Sil, 5) Bt + Ms + Grt + Sil + And, and 6) Bt + Ms + Kfs, with excess quartz and plagioclase in all assemblages. Assemblage (6) has no systematic distribution. This Bt + Kfs-bearing assemblage is inert for the change of pressure-temperature conditions. Excluding this assemblage, the other five assemblages show a systematic regional distribution, which enables to define three mineral zones: Bt-zone, Grt-zone and Sil-zone. The Ashburton schist has two representative mineral assemblages: 1) Bt + Chi + Ms, and 2) Grt + Bt + Chi + Ms except for the assemblage of And + Grt + Bt found only in a contact areole of Boolaloo batholith. THERMOBAROMETRY Mineral chemistry was analysed by electron probe. Garnet commonly has a reverse chemical zoning. It is more distinctive in Grt-zone samples. Fe - Mg partition coefficient (Kp) was calculated between carefully chosen compositions of garnet and coexisting biotite. KD suggests a continuous increase of temperature from Grt-zone to Silzone along the representative sampling route. Application of conventional Grt Bt geothermometry and Grt - PI - aluminosilicate quartz geobarometry yielded 2.7 - 4.1 kbar, 600 - 620 °C for the thermal core in Sil-zone. CONCLUSION Metamorphic thermal structure of the northern Capricorn orogen has revealed. It has a thermal axis in its core part and does not show constant nor gradual northward decreasing pattern as is said in the previous studies. This thermal structure is similar to those of modern collisionrelated metamorphic belts, although its metamorphic facies series has lower P-T features than Phanerozoic ones.

Supercontinents and Earth Evolution Symposium 2005

MESOPROTEROZOIC TO NEOPROTEROZOIC REWORKING IN THE CAPRICORN OROGEN, WESTERN AUSTRALIA: EVIDENCE FROM ''Ar^Ar DATING Sandra A Occhipinti, Steven M Reddy Tectonics Special Research Centre, Curtin University of Technology, PO Box 1987, Perth, WA 6004, Australia (S.Occhipinti@curtin.edu.au)

The Capricorn Orogen is a Palaeoproterozoic to Neoproterozoic zone that lies between the Archaean Pilbara and Yilgarn Cratons in Western Australia. It contains several tectonic domains, each with distinguishable rock types and tectonic histories. Detailed and regionalscale mapping in the southern part of the Capricorn Orogen, combined with geochronological dating, has led to a better understanding of the tectonic amalgamation of the Yilgarn and Pilbara cratons, the formation of the West Australian Craton, and it's subsequent tectonic evolution. An extensive study of "^Âr/^Âr age data from micas in the southern Capricorn region of Western Australia indicates a pervasive thermal event took place between 950 and 850 Ma. These data are at odds with the pervasive greenschist facies deformation event that took place throughout the region during the 1830 1780 Ma Capricorn Orogeny. However, northerly trending kinks and easterly, and northerly trending faults that cut the southern Capricorn Orogen, may be Neoproterozoic in age. The 950 - 850 Ma event is likely related to a thermal event resulting in the exhumation of the southern Capricorn Orogen in response to the Edmundian and Pinjarra Orogenies. The "^Âr/^Âr ages become systematically older to the east indicating exhumation is concentrated on the western margin of the West Australian Craton. Modelling diffusion of muscovite, and comparison of the results to empirically measured "^Âr/^Âr age profiles from in situ ultraviolet laser probe analyses of muscovite from the Capricorn Orogen shows that apparent "^Âr/^Âr ages from terranes that have had complicated thermal histories, could contain components of argon gas derived from earlier

cooling events. Furthermore the shape of measured "^Âr/^Âr age profiles may be an indication of the presence or absence of earlierderived argon gas with relatively flat age profiles indicating either simple thermal histories, or near-complete to complete resetting within a grain. If the presence of older argon gas is indicated by empirically measured "^Âr/^Âr age profiles, then when interpreting "^Âr/^Âr age data all parts of a samples thermal history have to be accounted for, not just the last thermal event. In this study "^Âr/^Âr diffusion modelling has shown that the thermal event in the Capricorn Orogen was hot enough to only partially reset argon at c. 900 Ma. Given the pervasive and regional nature of this event, we interpret the data to reflect collision of a continent onto western margin of the West Australian craton. The timing of this event is broadly constrained to be during the amalgamation of the supercontinent Rodinia. Coincident with exhumation and probable uplift of the Capricorn region was the development of early Neoproterozoic basins, such as the Officer Basin (Centralian Superbasin). Detritus in these sediments include granitic and sedimentary components, some of which include the underlying Bangemall Supergroup. Palaeocurrent data from sediments within the Officer Basin (Sunbeam Group) indicating that their detritus was largely sourced from the west, so may have come from the uplifted West Australian Craton margin. In addition, the age of the upper part of the Sunbeam Group is between 900-750 Ma. This supports that exhumation and uplift of the southern Capricorn Orogen resulted in the deposition of early Neoproterozoic basins, such as the Officer Basin (Centralian Superbasin).

Supercontinents and Earth Evolution Symposium 2005

RECONNAISSANCE GEOCHRONOLOGY OF THE ASHBURTON BASIN, WESTERN AUSTRALIA Keith Sircombe Geoscience Australia, GPO Box 378, Canberra, ACT 2601, Australia (Keith.Sircombe@ga.gov.au)

The Palaeoproterozoic Ashburton Basin records the evolution of the Capricorn orogeny on the southern margin of the Pilbara Craton. Numerous tectonic models have been proposed but debate has been hampered by a lack of absolute ages for the succession, especially in the thick, regionally extensive and largely homogeneous turbiditic Ashburton Formation. The Ashburton Formation is a 5-12 km thick turbiditic succession of mudstone and immature sandstone with minor amounts of conglomerate and volcanic rocks. Based on field, modal and palaeodirection data the succession has been interpreted as a longitudinal deep-marine basin with detritus fed from the southeast. As convergence continued submarine fan systems prograded northwards from the southern margin and ultimately culminated in the fluvial dominated sediments preserved in unconformably overlying basins (Mt. Minnie, Blair and Bresnahan). Two aspects of the Ashburton Basin's geochronology have been examined. Firstly, zircons from rare volcanic sequences within the Ashburton Formation have been used to geochronologically constrain the succession. Secondly, detrital zircon geochronology has been used to examine the evolution of the basin. DATED VOLCANIC CONSTRAINTS In the northwest basin, the June Hill Volcanics underlie the Ashburton Formation and have previously been used with an overlying tuff further south in the Capricorn Group to constrain deposition to a MO million year period in the Palaeoproterozoic. However, recent SHRIMP reexamination of the June Hill Volcanics has shown that the U-Pb age is significantly younger and remarkably similar to the overlying tuff. This suggests that the Ashburton Formation may be diachronous across the region and youngs westward. The hypothesis of diachroneity is further reinforced by a significantly older age for

the Mt. Boggola volcanics in the southeastern basin. Detrital zircon from the Ashburton Formation in the northwest also has a younger age component than seen elsewhere in the basin. DETRITAL ZIRCON GEOCHRONOLOGY Several samples of the Ashburton Formation were collected in a spatial and temporal transect across the central and eastern basin. Samples from lower in the succession tend to have polymodal and wide-ranging age components, whereas higher succession samples tend toward younger and unimodal components. Whereas the polymodal samples contain individual grain ages that potentially match the younger range of known ages in the Pilbara and Yilgarn cratons, there are no prominent clusters in this age range and Archaean ages are uncommon. All samples have prominent late Palaeoproterozoic components that are interpreted as having a provenance in similarly aged granites known from the Gascoyne Complex to the southwest. The recent development of statistical tools now provide quantification of the similarities between these - and other regional - detrital zircon datasets. CONCLUSIONS Detrital zircons from the central Ashburton Formation basin indicate an evolution from a polymodal Archean-Palaeoproterozoic provenance to singular Palaeoproterozoic provenance. The ubiquitous presence of Palaeoproterozoic detritus in the Ashburton Formation suggests that the depocentre was linked to the Gascoyne Complex during amalgamation between the Pilbara and Yilgarn. New dates for volcanism related to the Ashburton Formation indicates that deposition within the depocentre was diachronous, migrating from the southeast to northwest over - 3 0 million years.

Supercontinents and Earth Evolution Symposium 2005

TECTONIC SIGNIFICANCE OF DETRITAL ZIRCON AGE PROFILES ACROSS PALAEOPROTEROZOIC OROGENS IN THE KIMBERLEY REGION OF NORTHERN AUSTRALIA Ian M Tvler\ Steve Sheppard\ Simon Bodorkos\ Rod W Page^ ^Geological Survey of Western Australia, 100 Plain Street, East Perth, WA 6004, Australia ifian.tyler@doir.wa.gov.au) Research School of Earth Sciences, Australian National University, Canberra, ACT 0200, Australia

The pattern of differing ages and tectonic settings of metasedimentary rocks across orogenic belts gives important clues to their evolution. In the Halls Creek and King Leopold Orogens in the Kimberley region of northern Australia, terrane accretion onto the Kimberley Craton between 2500 and 1850 Ma was followed by collision with the North Australian Craton at 1820 Ma. Turbiditic metasedimentary rocks are present in three tectonostratigraphic terranes, each with histories that reflect disparate tectonic settings. The Western Zone formed between 1870 and 1850 Ma as a rift, marginal to the Kimberley Craton, filled with low- to high-grade turbiditic metasedimentary rocks of the Marboo Formation. SHRIMP U-Pb dating of detrital zircons from two samples show profiles derived from older accreted terranes dominated by ages ranging from 2500 to 1910 Ma. A population of zircon cores at ca. 1872 Ma in the Mount Joseph Migmatite provides a maximum depositional age, while rims at ca. 1861 Ma give the age of metamorphism. The Marboo Formation was deformed and metamorphosed, and then intruded by granitic and mafic rocks of the 18651850 Ma Paperbark Supersuite during the accretionary Hooper Orogeny. It is unconformably overlain by the Whitewater Volcanics, co-genetic with the supersuite. In the Central Zone an oceanic island arc developed at ca. 1865 Ma. Turbiditic metasedimentary rocks, together with predominantly mafic volcanic and volcaniclastic rocks, form the amphibolite to granulite fades Tickalara Metamorphics. Detrital zircons from five samples all show age profiles dominated by a single population giving a maximum depositional age of ca. 1865 Ma. A granitic sheet intruded at ca. 1863 Ma, suggesting that development of the arc took place rapidly. Maficultramafic intrusions occurred at ca. 1855 Ma, with the emplacement of tonalite sheets at ca. 1850 Ma. Peak metamorphism, with the formation of migmatitic rocks, took place at ca.

1845 Ma, coincident with further mafic-ultramafic intrusions. Rifting of the arc produced felsic volcanism and associated VHMS mineralization of the Koongie Park Formation. The oldest rocks in the Eastern Zone are ca. 1910 Ma felsic volcanics and associated granites. They are overlain by the Saunders Creek Formation at the base of the Halls Creek Group. In contrast to the active turbiditic environments of the Western and Central Zones, a detrital zircon age profile from this passive continental margin fluviatile unit is Archaean, ranging from a population at 2512 Ma to one at ca. 3600 Ma. It is overlain by ca. 1880 Ma passive margin basalts of the Biscay Formation. Lowto medium-grade turbiditic metasedimentary rocks of the overlying Olympic Formation can be divided into upper and lower units separated by phases of alkaline volcanism at ca. 1857 Ma and ca. 1847 Ma. The Olympic Formation represents a transition from a passive to an active margin setting, with the development of a foreland basin. A single population of detrital zircons at ca. 1874 Ma dominates three samples from the lower unit, while a single sample from the upper unit is dominated by a ca. 1847 Ma population. Further granites and gabbros of the Sally Downs Supersuite intruded mainly the central zone between 1835 and 1805 Ma, representing syn- to post-collisional settings during the Halls Creek Orogeny. Deformed and metamorphosed Olympic Formation is intruded by ca. 1820 Ma post-collisional granite. The Halls Creek and King Leopold Orogens may be part of a larger, diverse collisional orogen on a scale similar to the present AlpineHimalayan Orogen. This is consistent with suggestions that the collision drove intracratonic deformation in the Tanami and Arunta regions of the North Australian Craton, and that uplift provided the source for detrital zircons in post1835 Ma turbiditic meta-sedimentary rocks in those regions.

Siipercontinents and Earth Evolution Symposium 2005

^^Arl^^Ar CONSTRAINTS ON PROTEROZOIC COOLING AND EXHUMATION OF THE HALLS CREEK OROGEN, WA Simon Bodorkos\ Steven M Reddy^ ^Geological Survey of Western Australia, 100 Plain Street, East Perth, WA 6004, Australia (sinnon.bodorkos@doir.wa.gov.au) Tectonics SRC, Curtin University of Technology, GPO Box U1987, Perth, WA 6845, Australia

The Palaeoproterozoic Halls Creek Orogen preserves the most complete record of "Barramundi" (1880-1820 Ma) tectonism in northern Australia, but appears unaffected by the 1750-1500 Ma tectonomagmatic activity that pervades other parts of the North Australian Craton. However, reliable geochronological data in the Halls Creek Orogen are largely restricted to U-Pb SHRIMP zircon and monazite determinations, and the history of post-tectonic cooling in the crystalline basement remains poorly understood. We have undertaken a reconnaissance "^Âr/^Âr study in the northern central part of the orogen, where the protoliths of the Tickalara Metamorphics underwent granulite facies metamorphism at 1850-1845 Ma, prior to the emplacement of the Mabel Downs Tonalite over the interval 1835-1825 Ma. This mid-crustal segment was subsequently cut by the post-1820 Ma Highway Shear Zone, which developed under greenschist facies conditions and juxtaposed packages of Tickalara Metamorphics with slightly different prograde metamorphic histories, implying a net vertical displacement of 2-6 km across the structure. On the higher-pressure eastern side of the Highway Shear Zone, hornblende and biotite "^Âr/^Âr results from the Mabel Downs Tonalite yield plateau ages of 1705 ± 5 Ma and 1546 ± 19 Ma respectively, interpreted to record cooling through the temperatures 540 ± 40 "C and 315 ± 45 °C respectively. These values yield timeaveraged cooling rates of 1.5 ± 0.5 °C Myr^ (1845-1705 Ma) and 1.4 ± 0.4 °C Myr' (17051545 Ma). Both results imply very slow cooling of the mid-crust following pluton emplacement: however, neither violates any existing constraint on the exhumation of the mid-crustal segment currently exposed. Retrogressed Tickalara pelitic migmatite and sheared leucocratic orthogneiss were subjected to muscovite and biotite "^Âr/^Âr analysis, to constrain the latest episode of greenschist facies along the Highway Shear Zone. The two minerals from both rocks were characterised by variably discordant data that proved difficult to

interpret. However, none record Palaeoproterozoic isotopic closure, despite ubiquitous evidence for significant excess argon in the ^Âr release spectra. The data are more consistent with "^Âr/^Âr resetting during Neoproterozoic (900-800 Ma) recrystallisation at low temperatures. The tectonic implications of these results are difficult to assess, due to the small number of samples, their restricted spatial distribution, and the paucity of reliable geochronological controls on Meso- and Neoproterozoic tectonism in northwestern Australia. In the central Kimberley region, the intracratonic Yampi Orogeny deformed ca. 1700 Ma rocks of the Kimberley Group, and multi-grain biotite K-Ar ages in the range 1475-1000 Ma have previously been used to infer the timing of this event. However, if our Highway Shear Zone samples record Yampi-age tectonism, the combination of our pseudoplateau ages (900-800 Ma) and our total gas ages (in the range 1400-800 Ma) strongly suggest that the central Kimberley K-Ar data reflect the presence of significant unrecognised excess argon. Consequently, the 1000 Ma K-Ar age should be considered a maximum for biotite recrystallisation, raising the possibility that Yampi tectonism took place at ca. 900 Ma. An independent minimum age constraint is provided by post-orogenic sedimentary rocks, correlated with Super-sequence 1 of the Centralian Superbasin, that contain pre-800 Ma stromatolites. The wider significance of these results remains debatable; however, similar (1000-900 Ma) "^Âr/^Âr mica ages have recently been reported from the Capricorn Orogen, where crustal-scale Palaeoproterozoic shear zones were reactivated during the Neoproterozoic Edmundian Orogeny. We therefore suggest that the Yampi and Edmund tectonism reflect farfield, intracratonic jostling of the basement blocks constituting Rodinia during the final stages of supercontinent amalgamation, prior to the initiation of rifting, mafic magmatism and Rodinia breakup at ca. 830 Ma.

Siipercontinents and Earth Evolution Symposium 2005

PALAEO- TO MESOPROTEROZOIC TECTONICS OF THE ARUNTA REGION, CENTRAL AUSTRALIA Ian Scrimgeour Northern Territory Geological Survey, GPO Box 3000, Darwin, NT 0801, Australia (ian.scrimgeour@nt.gov.au)

INTRODUCTION The Arunta Region occurs at the southern margin of the North Australian Craton (NAC) in central Australia, and underwent prolonged tectonic activity in a series of events in the period 1810-1560 Ma. These events have varying structural, igneous and metamorphic characteristics, reflecting changing responses to plate margin processes. The southern part of the Arunta is characterised by high-grade, medium-

to high-P metamorphism, with evidence for significant burial and exhumation of rocks in a plate margin setting. In comparison, the northern Arunta is dominated by a series of lower grade events with localised high-T, low-P metamorphism, reflecting intraplate responses to events occurring to the south. This contrast provides an opportunity to compare the structural and thermal response to the same events in areas both proximal and more distal to the convergent margin. EVENTS IN THE ARUNTA The earliest major event in the Arunta Region is the Stafford Event at 1810-1800 Ma. In the eastern Arunta, this time period involved volcaniclastic sedimentation in a possible backarc environment, whereas the northern Arunta unden/vent bimodal magmatism and localised high-T, low-P metamorphism up to granulite facies. Widespread felsic and lesser mafic magmatism accompanied by compressional deformation occurred during the Yambah Event at 1780-1770 Ma. Metamorphism in this event was generally low- to medium-grade, and may be interpreted as closure of the back-arc basin and

development of a continental arc. Tectonic activity in the period 1735-1690 Ma was largely limited to the eastern Arunta, where granulite facies metamorphism occurred during the Strangways Orogeny. Intraplate responses to this event include localised granite intrusion and high-T, lowP metamorphism to the north and west. In the Warumpi Province (southwestern Arunta Region), voluminous magmatism at 16901660 Ma is interpreted to have formed outboard of the North Australian Craton, and thus has no known response in the craton. Transpressional accretion of the Warumpi Province onto the NAC is believed to have occurred in the Liebig Orogeny at 1640 Ma. This accretion was accompanied by deep-crustal granulite facies metamorphism, magmatism and exhumation in parts of the Warumpi Province, with broadly contemporaneous, localised extension and mafic magmatism in the NAC. Following this event was north-directed high-strain and orogenesis in the southern and central Arunta during the 15901560 Ma Chewings Orogeny. In the southern Arunta, this event involved Barrovian-style metamorphism, whereas localised higher-T metamorphism occurred in the central Arunta in the vicinity of highly radiogenic granites. The Chewings Orogeny represents a period of significant tectonism throughout much of Proterozoic Australia and appears to reflect an intraplate response to the collision of the NAC with the South Australian Craton. This event effectively cratonised the Arunta, until a series of intraplate reworking events in the Palaeozoic.

Supercontinents and Earth Evolution Symposium 2005

LATE PALAEOPROTEROZOIC OBLIQUE ACCRETION OF A 1690-1660 Ma MAGMATIC ARC ONTO THE NORTH AUSTRALIAN CRATON Dorothy Close\ Ian Scrimgeour\ Christine Edgoose\ Michael TD Wingate^, Kate Selway^ ^Northern Territory Geological Survey, GPO Box 3000, Darwin, NT 0801, Australia (dorothy.close@nt.gov.au) Tectonics Special Research Centre, University of Western Australia, WA 6009, Australia ^Continental Evolution Research Group, Geology and Geophysics, University of Adelaide, SA 5005, Australia

The Warunnpi Province is a 1690-1610 Ma easttrending elongate terrane on the southwestern margin of the Arunta Region, that is both isotopically distinct and has different protolith ages from the North Australian Craton (NAC). This terrane is interpreted as an exotic terrane that accreted onto the NAC around 1640 Ma. The Warumpi Province can be divided into three distinct domains. In the south, the amphibolite fades Haasts Bluff Domain comprises 1690-1660 Ma granites and volcanics that are interpreted to have formed in an arc environment outboard from the North Australian Craton. The Yaya Domain in the north is dominated by 1660-1650 Ma metasediments that were metamorphosed to granulite facies (>850°C, 9-10 kbar) and intruded by voluminous granites and lesser gabbros during the Liebig Orogeny at 1640-1635 Ma. This event has been interpreted to reflect the accretion of the Warumpi Province onto the NAC. Evidence for accretion at this time includes rapid, deep burial and exhumation of the Yaya Domain, a linear belt of calc-alkaline felsic magmatism, and a hairpin bend in the apparent polar wander path for northern Australia. In stark contrast to the Yaya Domain, 1690-1670 Ma granites and 1630 Ma volcanics of the Kintore Domain in the west of the Warumpi Province are only weakly deformed under greenschist facies conditions. Immediately north of the Warumpi Province, 1850-1830 Ma metasediments of the Aileron Province (NAC) were intruded by layered mafic rocks of the Andrew Young Igneous Complex at 1633 Ma. A granulite facies contact aureole to these mafic rocks suggests intrusion at 10-15 km depth. The presence of 1633 Ma layered intrusions implies extensional tectonics at the time, which appears to conflict with the

interpretation that the Liebig Orogeny resulted in the accretion of the Warumpi Province onto the NAC. Further evidence for late Palaeoproterozoic extension in the southwestern Arunta is a swarm of north-trending dykes in the Yaya Domain that truncate 1640-1635 Ma charnockites but which were deformed during the 1590-1570 Ma Chewings Orogeny. This dyke suite, named the llpilli Dolerite, has a precise SHRIMP U-Pb zircon age of 1633 Ma. This implies that a period of extension occurred at 1635-1630 Ma, immediately post-dating the Liebig Orogeny in the Yaya Domain. This age is also identical to that of the Andrew Young Igneous Complex. The existing data suggest that in the period 1640-1630 Ma, localised deep burial and exhumation occurred in some parts of the Yaya Domain, whereas other parts of the province underwent extension or remained at shallow crustal levels. The partitioning of the area into discrete zones with broadly synchronous extension and compression is suggestive of an oblique accretional orogen. The Yaya Domain reflects a transpressional orogen of limited extent, whereas the Andrew Young Complex and llpilli Dolerite reflect localised transtension. A significant strike-slip component to the suture is supported by a recent magnetotelluric survey that shows a dramatic conductivity contrast between the NAC and Warumpi Province. The suture is imaged as being a steep to vertical structure that underwent reworking by the northdipping Redbank Thrust during the Mesoproterozoic and Palaeozoic. Therefore, the suture of between the Warumpi Province and NAC is interpreted to reflect rapidly evolving oblique accretion of an arc terrane in the period 1640-1630 Ma.

Supercontinents and Earth Evolution Symposium 2005

NEW SHRIMP GEOCHRONOLOGY FOR THE WESTERN FOLD BELT OF THE MOUNT ISA INLIER: DEVELOPING A 1800-1650 Ma EVENT FRAMEWORK Narelle Neumann, Peter Southgate, George Gibson Minerals Division, Geoscience Australia, GPO Box 378, Canberra, ACT 2601, Australia (Narelle.Neunnann@ga.gov.au)

The Proterozoic Mount Isa Inlier records an extensive history of sedimentation, magnriatism, tectonisnn, and nnineralisation. SHRIMP geochronology has been integrated with sequence stratigraphy and facies analysis to develop a regional chronostratigraphic framework for sedimentary packages of the Leichhardt and Calvert Superbasins in the Western Fold Belt. Determining the depositional age and regional extent of these packages is important because they are possible source regions for younger Pb mineralising systems, and because the geometries of these basins have an important control on the migration of fluids and on the development of the overlying ca. 1670 to 1575 Ma Isa Superbasin. New SHRIMP geochronology integrated with basin analysis recognises three supersequences in the Leichhardt Superbasin: The Guide Supersequence spans the interval -1800-1785 Ma and includes the Bottletree Formation and the Mount Guide Quartzite. Sequence relationships suggest that this package represents an asymmetric second order cycle, recording a thickened transgressive suite and a condensed interval. The overlying Myallv Supersequence spans the interval -1780-1765 Ma and includes the Eastern Creek Volcanics and syn-depositional Lena Quartzite, and the Myally Subgroup. This package represents a second order supersequence cycle in which mafic volcanism was initiated during a phase of east-west extension. Following the cessation of volcanism, transgression led to the deposition of the Alsace Quartzite and deeper water Bortala Formation. An increase in the rate of sediment supply over accommodation resulted in progradation and deposition of the Whitworth Quartzite and red bed playa facies of the Lochness Formation as accommodation closed.

The Quilalar Supersequence spans the interval -1755-1740 Ma and includes the Quilalar Formation in the Western Succession, and the Ballara Quartzite and Corella Formation in the Mary Kathleen zone. Sequence analysis indicates that this package represents a series of storm-, tide- and wave-dominated shelfal marine depositional systems. Although there are no new depositional age constraints for the younger Bigie Formation, field relationships suggest that it is coeval with the -1710 Ma Fiery magmatic event. Therefore, we have defined a separate supersequence for the Bigie Formation, the Big Supersequence, even though it may be more genetically related to the Fiery Event. The Big Supersequence, together with the -1690 Ma Prize Supersequence, comprise the Calvert Superbasin. The evolution of the Leichhardt, Calvert and Isa Superbasins is temporally and spatially related with magmatism. In particular, the -1740-1735 Ma Burstall Event represents a bimodal, dominantly intrusive event following sedimentation of the Quilalar Supersequence in the Mary Kathleen Zone and the Eastern Succession. The refined age for the Weberra Granite is within error of the age for the Fiery Creek Volcanics, indicating that they are both part of the -1710 Ma Fiery Event. New SHRIMP ages for the Sybella Granite confirm that this unit is coeval with the Carters Bore Rhyolite, suggesting that magmatism associated with this event is constrained to 1675-1670 Ma, associated with and followed by deposition of the Gun Supersequence. Combining the new geochronological constraints with previous work now provides a detailed stratigraphic event framework between 1800 and 1575 Ma for the Western Fold Belt of the Mount Isa Inlier, and allows detailed comparisons and correlations with the Eastern Succession and other Proterozoic terranes.

Supercontinents and Earth Evolution Symposium 2005

LINKS BETWEEN PROTEROZOIC AUSTRALIA: GEOCHEMICAL PROVENANCE OF PALEO-PROTEROZOIC METASEDIMENTARY ROCKS FROM THE MOUNT ISA INLIER AND THE CURNAMONA PROVINCE Karin Barovich\ Narelle Neumann^, Martin Hand^ ^Continental Evolution Research Group, School of Earth and Environmental Sciences, University of Adelaide, SA 5005, Australia (karin.barovich@adelaide.edu.au) ^Minerals Division, Geoscience Australia, GPO Box 378, Canberra, ACT 2601, Australia

INTRODUCTION Correlations of the north and south Australian cratons during the Proterozoic have been drawn by a number of workers, based largely on lithostratigraphic, metamorphic and metallogenic similarities. Some geometric reconstructions place the Curnamona Province immediately south of the Mount Isa terrain, thereby correlating sedimentary sequences of the Willyama basin with those of the Eastern Fold Belt. In this study we compare Nd isotopic and geochemical datasets for coeval sedimentary rocks from the Willyama basin and the Eastern and Western Fold Belts. Such datasets can be used to evaluate the provenance of sedimentary rocks, particularly in terms of questions regarding tectonic regime and plate reconstructions. RESULTS AND DISCUSSION Our geochemical constraints point to a dominant central/northern Arunta source for lower Willyama sediments (older than ca 1650 Ma) with only minor input from the now adjacent Gawler Craton. Elevated geochemical patterns indicate a highly enriched fractionated felsic upper crustal source. Initial 8Nd values range from - 6 to - 4 , allowing for a dominantly central/northern Australian source. This interpretation is supported by detrital zircon patterns with peaks around 1780 Ma and a paucity of ca 1850 Ma zircons from the voluminous eastern Gawler Craton Donnington Suite. In contrast, the upper Willyama sequence (between 1650 and 1600 Ma) is marked by a significant shift in isotopic character to more primitive values (initial CNCI values around 0) and geochemical patterns far more typical of other post-Archaean fine-grained sedimentary rocks around Australia and the world.

New geochemical and Nd isotopic data for sedimentary rocks from the Leichhardt, Calvert and Isa Superbasins of the Western Fold Belt also indicate an evolved source, isotopically indistinguishable from the lower Willyama. But in contrast, the geochemical patterns are similar to the upper Willyama sequence, with no reflection of the highly enriched source seen in the lower Willyama rocks. Finally, previous Nd isotopic data from coeval sediments in the Eastern Fold Belt indicate that these packages contain less evolved Nd isotopic signatures than those from the Western Fold Belt, suggesting they received significantly less Archaean source material than the Western Fold Belt sequences. However, more Nd data from the sedimentary packages is required to establish this relationship. But, as has been suggested, it appears the Eastern Fold Belt sediments may have formed as an isolated terrain, distant from the Archaean to the west of the Mt Isa terrain. CONCLUSION Our detailed geochemical and isotopic work in the Willyama Supergroup shows that the upper Willyama sediments represent a change in tectonic regime, with introduction of a significantly more primitive source region. Previous detrital zircon work has closely correlated the Urquhart Shale of the Mount Isa Group with the upper Willyama Supergroup in the Curnamona Province. Our geochemical data provide evidence that those upper sedimentary sequences are not only coeval, but that the upper Willyama Supergroup saw a sedimentary source change that directly links its development to the ore-bearing Mount Isa Group

Supercontinents and Earth Evolution Symposium 2005

PROVENANCE OF SEQUENCES IN THE NORTHERN GAWLERCRATON: RE-ASSEMBLING PROTEROZOIC AUSTRALIA AND DISPERSING THE GAWLER CRATON Justin L Payne, Karin Barovich, Martin Hand Continental Evolution Research Group, University of Adelaide, SA 5005, Australia (justin.payne@adelaide.edu.au)

The ca. 150,000 km Nawa Domain, northern Gawler Craton, is situated in a critical location for unraveling the evolution of the Australian Proterozoic. It is the interface between the Archaean lithologies of the central Gawler Craton and the relatively juvenile Mesoproterozoic Musgrave Block and subsequently the North Australian Craton. Despite this key location, its history of crustal growth, deposition and provenance of sequences, and subsequent deformation is virtually unknown. A paucity of outcrop requires utilization of geochemical, isotopic and geophysical imaging tools to develop a tectonic framework for the Nawa and associated domains. RESULTS Detrital zircon spectra (obtained using U-Pb LAICPMS) indicate a maximum depositional age for the Nawa Domain and adjacent Coober Pedy Domain meta-sediments of -1700 Ma. The vast majority of detrital zircon ages fall between 1700 Ma and 1840 Ma with significant peaks at 17301750, 1760-1770, 1820-1830 and lesser peaks at ca. 1900, 2050 and 2500 Ma. The minimum depositional age constraint is metamorphism at -1650 Ma in the far-western Nawa Domain. In this study, single metamorphic grains produced comparable ages between -1600-1650 Ma from samples elsewhere in the domain. Nd isotopic data yield 8Nd(1.6 Ga) values for metasedimentary lithologies from -7.0 to -3.6 and depleted mantle model ages (Tdm) from 2.34 to 2.68 Ga. The majority of samples are extremely enriched in REE with chondrite normalized values up to 400 for LREE elements and LaA^b(n) values ranging from 10-36, contrasting with other Proterozoic Gawler metasediments which show significantly less enriched REE patterns, similar to Post-Archaean Australian Shale (PAAS). PROVENANCE The relative lack of Archaean input and an absence of -1860-1850 Ma zircons from the voluminous Gawler Craton Donnington Suite, appears to preclude the Gawler Craton as a

source region for the Nawa Domain and Coober Pedy sedimentary protoliths. Instead, the isotopic and distinctive trace element characteristics and detrital zircon age spectra display similarities to the lower Willyama Supergroup in the Curnamona Province. This suggests a similar source for these sequences. Existing work indicates the lower Willyama is derived from central and northern Australia. The detrital age spectra obtained in this study are also consistent with a northern Australian source. The Strangways Metamorphic Complex in the southern Arunta Region contains U-Pb zircon ages to which the Nawa detrital spectrum can be correlated. The Strangways orogeny (1730-1690 Ma) exhumed mid-upper level crustal lithologies of the appropriate ages, also creating topography to drive denudation and subsequent sedimentation. Hence we nominate this terrain as the dominant source for these metasediments. Additionally, the geochemical patterns of the metasediments indicate a highly enriched upper crustal granitic to rhyolitic source, similar to upper crustal Arunta composition. RECONSTRUCTIONS The above interpretation indicates the Nawa and Coober Pedy domains were proximal to the southern Arunta Region at —1700-1650 Ma. The paucity of Gawler Craton age detrital zircons also implies the Nawa and Coober Pedy Domains were exotic to the Gawler Craton nucleus at this time. Further data are needed to confirm this notion, including information on magmatic lithologies, timelines, and imaging of the nature of the domain boundaries. Nonetheless, the data suggest the Archaean nucleus and southern Proterozoic portions of the Gawler Craton were outboard of the NAC during the period ca. 17001650 Ma.

CONCLUSION Detrital zircon and Nd isotope information indicate the Nawa and Coober Pedy Domains were proximal to the southern Arunta Region at sometime during the interval ^1700-1650 Ma placing constraints upon reconstruction models.

Supercontinents and Earth Evolution Symposium 2005

CREATING A COHERENT TECTONIC SYNTHESIS BETWEEN THE GAWLER CRATON AND CURNAMONA PROVINCE Michael Szpunar, Martin Hand, Karin Barovich Department of Geology and Geophysics, University of Adelaide, Adelaide, SA 5000, Australia (michael.szpunar@adelaide.edu.au)

INTRODUCTION The Southern Australian Craton is dominated by the late Archaean to Mesoproterozoic Gawler Craton and the Paleo-Mesoproterozoic Curnamona Province. Whereas there is an established tectonic framework for the Curnamona Province and a growing focus on the evolution of the Gawler Craton, there has been little attempt to systematically explore tectonic linkages between the two domains. This is due in part to a -200 km wide rift between the domains filled with -15 km of Neoproterozoic sediments. Therefore this investigation is focused on the margins of the cratons and the basement outcrops between. RECORD OF ASSEMBLY IN CRATONIC MARGINS The -1750 Ma Wallaroo Group forms part of the leading edge of the eastern Gawler Craton, a region interpreted to be involved in a collision with elements of a proto-Curnamona Craton during the -1700 Ma Kimban Orogen. Because the Wallaroo Group was deposited before the Kimban Orogeny, it should record some evidence of this collisional event. However, there are currently no tectonothermal data for the Wallaroo Group. Therefore constraining the timing of deformation in the Wallaroo Group can indicate the timing of a collisional event between the Gawler and Curnamona. On the far western edge of the Curnamona Province lies the Mt. Painter Inlier. Here, geological relationships suggest there is an older basement that is cross-cut by a younger Mesoproterozoic cover sequence, whilst detrital zircon age data indicate both sequences are Mesoproterozoic. Furthermore there is an absence of information regarding timing and style of Proterozoic metamorphism. By constraining maximum depositional ages and the timing and nature of metamorphism within the

Mt. Painter Inlier, conclusions can be drawn on the effects of a collisional event on the western Curnamona. BETWEEN THE GAWLER CRATON AND THE CURNAMONA PROVINCE The Barossa Complex is a series of basement inliers outcropping between the Gawler Craton and the Curnamona Province. It is not known whether the inliers form part of the Gawler or Curnamona, or are separate former basins. Limited data from the inliers suggest metamorphism occurred between -1620 Ma and 1590 Ma, whereas there are preliminary magmatic ages of -1590 Ma and -1715 Ma (comparable with the Curnamona). By constraining the timing and nature of tectonothermal events and comparing these with cratonic margins to the east and west, potential links can be made with adjacent craton margins. U-Pb detrital zircon and Sm-Nd data from Barossa Complex metasediments can provide an insight into sedimentary provenance, providing direct comparisons with adjacent terrains and helping to constrain the timing of amalgamation of the larger cratonic blocks. Importantly for the latter, if the Gawler and Curnamona were amalgamated at -1700 Ma then the Curnamona should have a significant eastern Gawler detrital input. Because the Curnamona contains no -1750 Ma (Wallaroo Group age) detrital zircons, it appears amalgamation did not happen until after the sediments in the Curnamona were deposited (-1650 Ma). CONCLUSION In terrains where craton boundaries are obscured by thick cover sequences, a multianalytical approach targeting craton margins and basement inliers can provide a coherent tectonic synthesis between geographically separate domains.

Siipercontinents and Earth Evolution Symposium 2005

DATING PROGRADE AND PEAK METAMORPHISM WITHIN A COMPLEXLY DEFORMED TERRANE USING IN SITU U-Pb MONAZITE GEOCHRONOLOGY Caroline J Forbes\ Dave Giles^, Pete G Betts^, Roberto Weinberg\ Pete D Kinny' predictive minerals discovery Cooperative Research Centre, School of Geosciences, Monash University, Melbourne, VIC, Australia (Caroline.Forbes@doir.wa.gov.au) ^School of Geosciences, Monash University, Melbourne, VIC, Australia ^Tectonics SRC, Department of Applied Geology, Curtin University of Technology, Perth WA, Australia

Characterisation of the earliest phases of deformation can be difficult within complexly deformed terranes that record multiple episodes of deformation and metamorphism. Within such terranes, early-formed shear zones can become critical structures in unravelling the terrane history because they are localised areas into which strain is partitioned during deformation, and can record evidence of events that may go unrecognised in less deformed rock packages. Providing timing constraints on recognised metamorphic and deformational events allows a more complete tectonothermal history to be constructed. However, within poly-deformed and metamorphosed terranes, this task can be difficult owing to the possibility of multiple mineral growth events and the requirement to use a suitable isotopic system with a closure temperature that will allow dating of events that may pre-date peak metamorphic conditions. In situ SHRIMP U-Pb analysis of monazite grains was used to constrain the timing of prograde amphibolite and peak granulite facies metamorphism from pelitic rocks within an earlyformed high-temperature shear zone within the complexly poly-deformed Broken Hill Block, Australia. This isotopic system was applied because it has a high closure temperature within some common U-bearing accessory phases, and the in situ nature of the analysis allows retention of the textural contexts of the grains and

identification of subtle complexities in grain populations that may otherwise go unnoticed. Geochronological analysis revealed two distinct age populations, ca. 1.62 Ga and 1.60 Ga, and were obtained from grains located in texturally distinct positions. The older ca. 1.62 Ga monazite age population was obtained from grains that occur as inclusions completely enclosed within coarse-grained K-feldspar and garnet grains. The host minerals comprise part of the peak granulite facies mineral assemblage of the pelites, and armoured the monazite inclusions against the effects of younger deformation and metamorphic events. The ca 1.62 Ga monazite population occurs as part of an amphibolite facies inclusion assemblage hosted within the peak granulite facies mineral assemblage, and constrains the timing of prograde amphibolite facies metamorphism within the Broken Hill Block. The younger -1.60 Ga monazite age population was obtained from grains within a pervasive shear fabric of the pelites, and from grains located adjacent to fractures or grain boundaries of host K-feldspar and garnet grains, and were not armoured against later events. The ca. 1.60 Ga age population is interpreted to represent a later stage of monazite growth and/or isotopic resetting during deformation along the hightemperature shear zone at peak granulite facies conditions.

Supercontinents and Earth Evolution Symposium 2005

PROTEROZOIC MAGMATIC ARCS AND OROCLINES: ST PETER SUITE, GAWLER CRATON, SA Greg Swain\ Karin Barovich\ Martin Hand\ Gary Ferris^ ^Continental Evolution Research Group, School of Earth and Environnnental Sciences, University of Adelaide, SA 5005, Australia (greg.swain@adelaide.edu.au) ^Office of Minerals and Energy Resources, PIRSA, GPO Box 1671, Adelaide, SA 5001, Australia

INTRODUCTION Australian Proterozoic magmatic rocks have historically been interpreted to have formed in intraplate settings, derived largely from melting of pre-existing continental crust. These conclusions have been based largely on studies of granitoid magmatism in northern and central Proterozoic Australian regions. Proterozoic granitoid magmatism in the Gawler Craton has been understudied to date. The St Peter Suite defines a ca. 1 6 2 0 - 1 6 0 8 M a mafic to felsic intrusive complex that dominates the southwestern and central Gawler Craton (GO) in southern Australia. Volumetrically dominant magmatic phases include diorite, tonalite, granodiorite and granite.

RESULTS The intermediate to felsic tonalitic to granodioritic rocks (Si02 ^ 60-73 wt%) of the St Peter Suite are metaluminous, have moderate to high LREE and Sr, low HREE concentrations, and display strong negative Nb and Ti anomalies. The mafic rocks (Si02 = ~49-55 wt%) intruded as med- to high-K -calc-alkaline diorites, have comparable moderate to high LREE, Sr and low HREE concentrations, and Nb and Ti depletions. SmNd isotopic signatures for the St Peter Suite are predominantly juvenile (8Nd(i62o Ma) " ~2 to +2), suggesting only minor input of evolved subducted sediment or crustal contamination during magma ascent. DISCUSSION The geochemical affinities of the St Peter Suite are characteristic of modern adakitic magmas (arc-andesites and dacites), which suggests generation in an arc environment. The relatively juvenile isotopic signature implies minor Archaean crust contamination, demonstrating that subduction was active outboard of the GC continental margin. Subduction-related magmatism of the ca. 1620-1608 Ma St Peter Suite supports the role of lateral accretionary tectonics during Proterozoic assembly of the Australian continent. However,

geometrically, the St Peter Suite appears as a central core within an arcuate terrane of older crustal domains (late Archaean Mulgathing and Sleaford Complexes). We investigate whether the Archaean-Palaeoproterozoic GC was a linear continental crustal fragment which underwent Mesoproterozoic oroclinal bending. South Australian palaeomagnetic poles are restricted to the southern GC (SA iron ore deposits, Tournefort Dykes, and Gawler Range Volcanics). These data remain equivocal with respect to both Proterozoic plate reconstructions and a ca. 1600-1550 Ma thermal overprint. The palaeomagnetic poles are also geographically clustered, and fail to constrain the relative geometric arrangement of the northwestern GC. Evidence for a previously linear ArchaeanPalaeoproterozoic ribbon continent includes geological constraints on both the east-west Tallacootra and north-south Kalinjala Shear Zone systems, which have consistent EPMA monazite ages of ca. 1680 Ma and dextral vergence directions. These shear zone systems, related to the late Kimban Orogen (ca. 1730-1700 Ma), are pervasive throughout Archaean crust in the GC, and may have formed as part of a transpressional regime involving the previously linear ribbon continent. CONCLUSION St Peter Suite island arc magmatism developed outboard of a Gawler Archean continental domain, coeval with high-grade metamorphism in the northern GC between ca. 1630-1600 Ma. This island-arc terrane was accreted and may have been folded during oroclinal bending, immediately followed by generation and emplacement of the ca. 1595-1575 Ma Hiltaba Suite. Internal reorganisation of the Gawler continued until ca. 1460 Ma, via strain partitioning into major shear zones (e.g. Yerda, Yarlbrinda and Coorabie), followed by stabilisation of the craton.

Siipercontinents and Earth Evolution Symposium 2005

GEOCHEMISTRY AND PROVENANCE OF A MESOPROTEROZOIC (1.4 Ga) EASTERN MUSGRAVE BLOCK BASIN: BUDDYING UP TO THE BELT-PURCELL BASIN Ben Wade, Karin Barovich, Martin Hand Continental Evolution Research Group, School of Earth and Environmental Sciences, University of Adelaide, SA 5005, Australia (benjamin.wade@adelaide.edu.au)

INTRODUCTION Mesoproterozoic reconstructions of Rodinia place an enigmatic continent outboard of the western margin of Laurentia. Candidates have included both South China and Siberia. However, for approximately 15 years, Australia has featured prominently in a collision-rift setting with Laurentia's western margin. Versions include the southwestern US - East Antarctica (SWEAT) model, based on observation of tectonic similarities extending from Antarctica into parts of southern Laurentia. Another is the Australia - Western US (AUSWUS) model, based largely on matching crystallisation and deformation ages. Recent acquisition of precise latest Mesoproterozoic palaeopoles have led others come to the conclusion that Australia was separate from Laurentia at 1.2 Ga, and possibly only came together by ca. 1.07 Ga in the AUSMEX configuration. However, owing to the sparseness of reliable and geochronologically constrained Australian Mesoproterozoic palaeopoles, it can be said that the relative configuration and proximity of both Australia and Laurentia between 1.5 and 1.2 Ga is not well known. Detailed geochemical and geochronological studies can allow the matching of sedimentary basins along the eastern margin of Proterozoic Australia with coeval basins along the western margin of Laurentia. Data from the Laurentian Belt Supergroup have led to suggestions that eastern Australia is a viable source for 1.6-1.57 Ga Belt Basin detrital zircons, because this period represents a magmatic gap in Laurentia. Similarly detailed studies along the eastern Australian margin are required to use sedimentary sequences as geologic piercing points in matching up continental margins. This study presents results from a Mesoproterozoic sedimentary basin in the eastern Musgrave Block, and proposes a correlation with the Belt Basin.

RESULTS AND DISCUSSION Ages of detrital zircons from metasediments in the eastern Musgrave Block cluster at 1.08-1.2 Ga, 1.40-1.49 Ga, 1.54-1.60 Ga, and 1.65-1.81 Ga, with the youngest detrital population of ca. 1.40-1.49 Ga providing a rough maximum depositional age. The samples exhibit moderately evolved to juvenile Nd isotope characteristics, with 8nci(1400 Ma) values ranging from -5.1 to-0.1. Basement to the metasediments are ca. 1.601.54 Ga felsic granulites with £Nd(1400 Ma) values ranging from -0.9 to -2.7. The age and juvenile Nd isotopic characteristics of these gneisses provides an excellent match for the source of the ca. 1.60-1.54 Ga detrital zircons and the juvenile nature of the sequences of the lower Belt Supergroup. Zircon sources falling in the age ca. 1.49-1.4 Ga are unrepresented in Australia, and fall within the Australian "magmatic gap", much like the North American "magmatic gap" of ca. 1.6-1.53 Ga. As such, the ca. 1.49-1.4 detrital zircons are interpreted to been derived from the voluminous 1.5-1.3 Ga "anorogenic" magmatic belt of Laurentia. CONCLUSIONS Detrital zircons and Nd isotopes exclude the presently exposed fragments of Australia as major contributors to the metasediments of the eastern Musgrave Block. A number of grains within the metasediments of an age within Australia's "magmatic gap" require proximity of a Laurentian-like source of ca. 1.49-1.40 Ga. Additionally, the Nd isotopically juvenile accretionary terranes of southern Laurentia provide an excellent isotopic match. Conversely, the presence of abundant ca. 1.60-1.54 Ga zircons within the isotopically juvenile lower sequences of the Belt basin could conceivably by sourced from the ca. 1.60-1.54 Ga felsic granulites of the Musgrave Block.

Supercontinents and Earth Evolution Symposium 2005

A PRELIMINARY LITHOLOGICAL AND TECTONIC CHRONOLOGY FOR THE WEST MUSGRAVE COMPLEX Heather M Howard, RH Smithies, F Pirajno, S Bodorkos, IM Tyler Geological Survey of Western Australia, 100 Plain Street, East Perth, WA 6004, Australia (heather.howard@doir.wa.gov.au)

The Musgrave Complex, in central Australia, is a Mesoproterozoic orogenic belt which lies at the junction between the North, West and South Australian cratons. It has a connplicated metamorphic, intrusive, and deformational history that spans the earliest Mesoproterozoic (at least) to Lower Cambrian, and includes the ca. 1220-1150 Ma Musgravian Orogeny, the intracratonic ca. 1080-1060 Ma Giles Event and the ca. 570-530 Ma Petermann Orogeny. How this relates to other Proterozoic zones such as the South Arunta, Rudall and Albany-Frazer is unclear. So far our regional geological study of the western part of this complex has concentrated mainly on country rock to the ca. 1070 Ma mafic-ultramafic Giles intrusions, adjacent to the WA-NT and WA-SA borders. This area is characterised by a generally easterly- to northeasterly-trending series of anastomosing mylonite zones. Northward thrusting of the Musgrave Complex over the Amadeus Basin, during the Petermann Orogeny (570-530 Ma), is marked by the easterly-trending Woodroffe Thrust. Other major fault zones in this region, last active during the Petermann Orogeny, include the Mann-Hinckley Fault Zone, and a northwest-trending splay. The oldest exposed rocks are locally migmatitic, mafic and felsic, garnet and pyroxene-bearing granulitic gneisses that form rafts and xenoliths in younger ca. 1220-1150 Ma granites. A mafic granulite with no geochemical evidence for contamination contains rare intact zircon cores that yield a SHRIMP U-Pb age of 1596 ± 28 Ma. A well banded (possibly bedded?) felsic migmatitic gneiss gave a wide range of SHRIMP U-Pb zircon ages. The majority of zircon cores from the leucosome gave ages of ca. 1680-1550 Ma. Two cores gave Archaean ages of 2590 and 2740 Ma. Rare zircon rims showed syn-migmatitic ages of 1200-1190 Ma, more commonly reset during the Giles Event. Granites of the Pitjantjatjara Supersuite were intruded during the Musgravian Orogeny. They range from quartz monzodiorite to syenogranite, with the majority being monzogranitic and they contain clinopyroxene ± orthopyroxene, with garnet coronas. While they have an apparent

metamorphic mineralogy, they commonly preserve igneous textures, such as rapakivi. These granites show geochemical characteristics of A-type magmas such as high P, Ti, HFSE and alkalies, and like other similar A-type magmatic systems, have potential for Olympic Dam style mineralisation. U-Pb SHRIMP zircon data from Pitjantjatjara Supersuite granites from WA indicate cores as old as ca. 1220 Ma. This date is interpreted as the age of the intrusion, while the younger (<1180 Ma) rims are interpreted as a high-grade metamorphic overprint, coinciding with the intrusion of several other Pitjantjatjara Supersuite granites, with local development of gneissic fabrics. A 1188 ± 8 Ma age for a gabbro-anorthosite and a 1165 ± 6 Ma age for a mafic dyke indicate contemporaneous mantlederived magmatism. The granites are virtually free of inherited zircon except for rare 1600 Ma cores that reflect a much older source component (possibly the older gneisses). The lack of inherited zircons probably relates to the unusually high temperatures required to form Atype magmas. Nd-model ages (TDM ca. 2000 Ma) of these granites are older than any exposed crustal material and require an old (pre 1600 Ma) crustal component at some stage in their petrogenesis. The Musgravian Orogeny appears to have been a protracted series of events characterised by regionally hot and dry magmatic episodes that melted a refractory and relatively homogeneous basement. The Giles Event is characterised by voluminous mafic-ultramafic intrusions belonging to the Warakurna LIP. Particularly to the south of the Mann-Hinckley Fault Zone, the Giles Event gabbros show mingling textures with granites that are compositionally distinct from the Pitjantjatara Supersuite. The Giles intrusions were emplaced into the crust at a high metamorphic grade, and are locally migmatised and folded about southeast-plunging fold axes. To the north of the Mann-Hinckley Fault, easterly-trending mylonite and pseudotachylite zones are common. In these zones, pseudotachylite both cuts and is cut by mylonite, suggesting that conditions were around the brittle/ductile transition. The zones have also

Supercoiitinents and Earth Evolution Symposium 2005

locally been migmatised and folded. Similar structures have elsewhere been ascribed to the Petermann Orogeny and to the development of the Woodroffe Thrust system. However, the mylonite and pseudotachylites zones in this area show a consistent top to the south movement,

opposite to the main sense of movement along the Woodroffe Thrust. This movement might reflect late- to post-Petermann Orogeny collapse or relaxation along structures related to the Woodroffe Thrust.

Siipercontinents and Earth Evolution Symposium 2005

WEDNESDAY 28 SEPTEMBER 2005

Supercontinents and Earth Evolution Symposium 2005

NATURE AND ORIGIN OF NEOPROTEROZOIC GLACIALS Paul F Hoffman Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA 02138, USA (hoffman@eps.harvard.edu)

There were at least three major Neoproterozoic glaciations, the 'Sturtian' from <740 Ma until <710 Ma, the 'Marinoan' from <665 Ma until 635 Ma, and the 'Gaskiers' entirely in 580 Ma. The first two were global but the last appears limited to Avalonia-Cadomia, Baltica, NW China and NW Australia. In the type area (eastern Newfoundland), it lasted <1.0 million years. In contrast, the Sturtian and Marinoan glaciations lasted for millions to <30 million years according to stratigraphic, paleomagnetic and geochemical evidence. They occur in both clastic and carbonate-dominated successions, whereas the Gaskiers occurs in carbonate-poor areas only. Banded iron formations are intimately associated with Sturtian (Australia, Canada, Namibia, USA) and Marinoan (SW Brazil, SW Namibia) glacials, but are unknown in the Gaskiers. Post-glacial cap carbonates show that the ocean was most oversaturated after the Marinoan and least oversaturated after the Gaskiers. These observations suggest that the Sturtian and Marinoan glaciations were snowball events, but the Gaskiers was a regional glaciation and possibly diachronous. Boron isotope data suggest that seawater pH was lowered by perhaps one full pH unit during the Marinoan glaciation, implying a dramatic rise in atmospheric CO2 (which would have equilibrated with seawater on geological time scales even if air-sea contact was limited to crack systems). The transgressive nature of Marinoan cap carbonates is generally related to glacioeustatic rise, implying a time scale of the order of 2000 years according to GCM simulations of deglaciation. The only source of alkalinity capable of driving the acidic ocean to critical oversaturation on this time-scale is carbonate weathering; methane oxidation could only have been a quantitatively minor contributor given the average thickness of 10-20 m for Marinoan cap dolostones. On a snowball Earth, net sublimation would occur near sea level in the subtropics and net condensation in the tropics and mid-latitudes. This is consistent with the meridional distribution of glacial deposits given by paleomagnetic constraints. An important role for topography in ice-sheet development is indicated by the observed regional-scale correlation between

glacial sediment volume and contemporaneous tectonic activity. Within a region, glacial sediments are typically thickest on submarine slopes and in small basins, and thin or absent on paleotopographic highs (converse to cap carbonates). Thick deposits limited to broad paleovalleys suggest an important role for ice streams in ice-sheet drainage. Development of ice sheets on low-latitude carbonate platforms in intraplate settings implies that critical topographic relief was provided by large sea-level falls. Causal theories for Neoproterozoic glaciation postulate long-term chilling factors and proximal events. Among the former are a high degree of continental fragmentation, an apparent paucity of high-latitude land area, a preponderance of tectonic lands in the tropics, and the eruption or drift of flood basalts in the tropics. These factors all favor a cold global climate due to silicateweathering feedback. Paleomagnetic data from Neoproterozoic carbonates in Oman, South China and Svalbard show that they formed at low paleolatitudes, implying a normal meridional climate gradient contrary to the large orbital obliquity hypothesis of low-latitude glaciation. Inertial-interchange true polar wander around 800 Ma, based on paleomagnetic data from South China, and Svalbard, is postulated to have swung land areas rapidly into the tropics, initiating the cold global climate. Proximal triggers include large impacts (modeling suggests that a K-T size impact could trigger a snowball Earth if the ocean was cold like the Pleistocene, not warm like the Maastrichtian), astronomical encounters with giant molecular clouds (which could theoretically lower radiative forcing by <9 Wm^ on a time-scale that is short relative to silicate-weathering feedback) and various methane greenhouse destruction scenarios. The past year was one of major breakthroughs, including direct U-Pb dating of all three glacials, and the discovery of Ir spikes and boron isotope anomalies in cap carbonates in Congo-Zambia and Namibia, respectively. Recalibrating the central African Ir spikes with the Holocene Ir accumulation rate from Greenland ice cores gives - 2 4 million years duration for each of the Sturtian and Marinoan glaciations. Subject to confirmation in other

Supercontinents and Earth Evolution Symposium 2005

regions and additional testing (e.g. Os isotopes), the Ir spikes and boron isotope anomalies give new support to the snowball Earth hypothesis. Falsification is most likely to come from

geochronology, the short-lived Gaskiers glaciation for example, but only U-Pb thermalionization mass spectrometry (TIMS) has the required analytical precision.

Supercontinents and Earth Evolution Symposium 2005

EVOLUTION OF ATMOSPHERIC GROWTH

AND CONTINENTAL

Yiefei Jia Commonwealth Science and Industrial Research Organization, School of Geosciences, Monash University, PO Box 28E, VIC 3800, Australia (yiefei.jia@csiro.au)

The Earth is thought to have acquired its atmosphere-hydrosphere at ca. 4.5 Ga from impacts of carbonaceous chondritic meteorites (CI) and/or comets formed in the vicinity of Jupiter, with = +30 to +42%o, suggested on the basis of nitrogen isotopic compositions of 2.7 Ga Archean sedimentary rocks and crustal hydrothermal systems; they are characterized by a of 15 to 24%o, compared to existing data of 2 to 6%o their Phanerozoic counterparts, in early report by the author and co-worker, implying a secular decrease in of crustal rocks. Previous studies also show that the upper mantle has a mean of-5%o as inferred from modern mid-ocean ridge basalts and most diamonds, which are as old as 3.2 Ga. Some rare diamonds have down to -25%o and are thought to present an initial mantle value. The well-characterized nitrogen cycle demonstrates that the crustal nitrogen in both organic matter and ammoniated phyllosilicates result from the biological sequestration of atmospheric N2 by microorganisms (such as cyanobacteria). The biological assimilation of nitrogen, via the fixation of atmospheric N2, forms organic nitrogen compounds, so-called organic matter. The decomposition of marine organic matter due to thermal processes results in the partial release of the nitrogen as N H / , ultimately forming structurally robust potassium (K)-

silicates within crustal crystalline rocks. Consequently, shifts of atmospheric to its present value of 0%o can be accounted for by a combination of sequestration of atmospheric N2 into crustal rocks and ^^N-depleted upper mantle degassing by volcanism at island arcs, back arcs, and mid-ocean ridges. Shifts of the upper mantle of up to -5%o result from recycled Nenriched crustal rocks being incorporated into the mantle by subduction. The evolution of atmospheric N2 would therefore have been linked to the growth of continental crust, although this has not been explicitly stated in previous research. Models for growth of the continental crust range from early growth with rapid recycling of continental and oceanic crust into the mantle, to progressive growth and minimal recycling of continental crust. However, the fast shifts in the atmosphere from + 4 2 % o at 4 . 5 Ga to - 0 % o over 2 Ga, would support an early crustal growth model. This model postulates a very early growth with rapid recycling of continental and oceanic crust into the mantle. Also, recent studies on both the Nd isotopic composition of the 4 . 0 - 3 . 6 Ga Acasta gneiss, and the Nb/U and Nb/Th ratios of basalts erupted from Archean mantle plumes, support early growth and recycling of continental crust.

Supercontinents and Earth Evolution Symposium 2005

LINKING CAMBRIAN VOLCANISM TO MARINE ANOXIA AND MASS EXTINCTION: CLUES FROM SULPHUR ISOTOPE GEOCHEMISTRY M Hough\ GA Shields\ H Strauss^ LZ Evins\ RA Henderson\ S Mackenzie^ ^School of Earth Sciences, James Cook University, Townsville, QLD 4811, Australia (Lena.Evins@jcu.edu.au) Geologisch-Palaontologisches Institut und Museum, Westfalische Wilhelms-Universitat Munster, Corrensstrasse 24, 48149 Munster, Germany

It has previously been shown that there is an excellent temporal correlation between Large Igneous Provinces (LIPs) and episodes of dramatic global environmental change (mass extinctions and marine anoxia) from the Permian period onwards. New studies show that this correlation may be extended back in time to the Cambrian. Recently, Linda Glass (ANU) established that the Antrim Plateau basalts and their equivalents in northern Australia form part of a LIP, called the Kalkarindji Continental Flood Basalt Province, which erupted at ca. 510 Ma. This new age places the Kalkarindji province at the boundary between Early and Middle Cambrian, around the time of the Toyonian biotic crisis, a mass extinction that saw the collapse of metazoan reef communities worldwide. As with many other mass extinction events, it coincides with an episode of global marine anoxia. To explore further this Cambrian anoxiavolcanism-mass extinction connection, we seek to test the hypothesis that the environmental effects of the Kalkarindji eruptions can be seen in the sedimentary record. Our aim is to study the isotope stratigraphy of biostratigraphically constrained sedimentary sequences from around the Early-Middle Cambrian boundary. Here, we focus on the S isotopic composition of francolitebound sulphate from pyrite-free samples from the Middle Cambrian Duchess phosphorite deposit in Australia. Sulphur isotopic analyses were performed on a Finnigan MAT DeltaPlus mass spectrometer at the Geology-Paleontology Institute of the University of Muenster, Germany. The systematic behaviour of S during diagenesis allows us to backstrip the effects of early marine diagenesis to pinpoint seawater sulphate and SO4 concentration.

RESULTS The Australian marine phosphorites were deposited after the cessation of igneous activity at c. 510 Ma, following the post-extinction transgression, but earlier than 505 Ma. All samples so far analysed exhibit unusually high sulphate S^'^S values of between 45%o and 62%o, which are the highest values ever recorded in marine phosphorite. The systematic behaviour of S during diagenesis allows us to pinpoint seawater sulphate S^'^S at 50%o, which is considerably higher than ever reported for the Phanerozoic, and 15%o higher than during the preceding 100 million years of the Ediacaran Early Cambrian. A similar positive excursion has been recorded for the Permian-Triassic boundary and late Permian, and is thought to be a result of the global marine anoxia that accompanied the end-Permian and Frasnian-Famennian extinctions. We suggest that the anoxia recorded by the sulphur isotopes at the time of the Toyonian crisis was linked to global warming triggered by the Kalkarindji eruptions. CONCLUSION Our sulphur isotope studies point to a major positive excursion associated with the EarlyMiddle Cambrian transition. We propose that this excursion represents a sustained period of superanoxia and widespread pyritisation related to global warming induced by volcanism; a scenario similar to that associated with the endPermian extinction. Our study therefore confirms a possible superanoxia-volcanism-mass extinction connection for another Phanerozoic extinction and suggests a common cause for the Late Devonian, Permian-Triassic and Toyonian extinctions.

Supercontinents and Earth Evolution Symposium 2005

NEOPROTEROZOIC CORRELATIONS, INSIGHTS AND IMPEDIMENTS Kathleen Grev\ Clive R Calver^ ^Geological Survey of Western Australia, 100 Plain Street, East Perth, WA 6004, Australia ifkath.grey@doir.wa.gov.au) Mineral Resources Tasmania, PO Box 56, Rosny Park, TAS 7018, Australia

Ratification of the GSSP for the Ediacaran System and Period has revived interest in subdividing not only the Ediacaran, but also other parts of the Neoproterozoic. As well, it raises the issue of the most appropriate techniques to use for Proterozoic subdivision and correlation. As a result of extensive studies of the Cryogenian and Ediacaran successions, a correlation framework already exists in Australia. The scheme is based on integrated results from lithostratigraphy, event stratigraphy, sequence stratigraphy, seismic interpretation, biostratigraphy, and isotope chemostratigraphy. Results so far are encouragingly consistent. However, they also highlight contradictions in global correlations, especially between Australia and China. CRYOGENIAN Neoproterozoic correlations have advanced considerably through the study of continuously cored drillholes in the Officer and Amadeus Basins and Adelaide Rift Complex. As part of GSWA's evaluation of the hydrocarbon potential of the western Officer Basin, all available and potential biostratigraphic controls were assessed. Lack of outcrop, limited seismic coverage, and sparse drillholes impose constraints that prompted appraisal of Cryogenian successions elsewhere in Australia. The drilling of Empress 1/1A and Lancer 1 tested the correlation model based on palynology and acritarch biostratigraphy and matched it to parallel stable isotope studies. The lower Buldya Group (western Officer Basin) can be correlated with the Bitter Springs Formation (Amadeus Basin) and the Callanna Group (Adelaide Rift Complex). The Sturtian glaciation is marked by a hiatus in Western Australia, but it is probably about 700 Ma, based on detrital zircon ages. EDIACARAN Little is known about the biostratigraphy of the interval between the Sturtian and Marinoan glaciations, but the overlying Ediacaran is characterised by well-documented and wellpreserved palynomorph assemblages and a characteristic 5^^Corganic curve. Morphologically complex acritarchs are present in the middle

Ediacaran and four zones allow correlation across a wide area of Australia. The interval between the Marinoan glaciation and the first acanthomorph (spiny) acritarchs (at least 700 m thick) contains only leiospheres (simple spheres) and it is not until the second transgressive/regressive cycle after glaciation that acanthomorphs appear and diversify. This is inconsistent with Snowball Earth models that predict diversification during the first postglacial transgression. The biotic change may be more closely associated with the Acraman impact event than with the glaciation. Again, correlations based on integrated results from biostratigraphy and stable isotopes are highly consistent. DISCUSSION Acritarchs show good potential for Ediacaran subdivision and global correlation because several Australian species are also known from Siberia and China. However, attempts to match successions from the Australian Ediacaran and the Chinese Doushantuo Formation produces major inconsistencies. The anomalies apply to correlations based on both biostratigraphy and carbon isotope curves. These inconsistencies can be resolved if the Nantuo Tillite and Elatina Formation (the Marinoan glaciation) are different ages and if the Doushantuo represents a highly condensed succession and/or contains stratigraphic breaks. However, if the Nantuo Tillite really is the same age as the Marinoan, there must have been at least four glaciations in Australia; the Sturtian (c. 700 Ma), the Marinoan (c. 635 Ma), the Croles Hill Diamictite/Cottons Breccia (c. 580 Ma) and the Egan Formation (c. 600 Ma). CONCLUSIONS The corollary of this is that it raises doubts about correlations made solely on the isotope signal associated with the Marinoan glaciation, and even about whether the Marinoan glaciation was a single global event. These problems can be resolved through ongoing subdivisions and correlation studies, and by adequate dating of Australian Neoproterozoic glaciations, either directly, or (more feasibly) indirectly through detrital zircon analyses.

Supercontinents and Earth Evolution Symposium 2005

'CAP CARBONATES' AND WHY THE WALSH TILLITE IS MARINOAN Maree Corkeron School of Earth Science, James Cook University, Townsville, QLD 4811, Australia (maree.corkeron@jcu.edu.au)

Ice ages of possibly global proportions are commonly accepted as phenomena featuring in late Neoproterozoic history. Glacigenic deposits found world-wide are the primary evidence for large-scale glaciation, but a ubiquitous and more enigmatic sedimentary unit, the 'cap carbonate' is often interpreted as evidence for global synchronicity of glaciation. So distinct, sedimentologically consistent and globally widespread are these carbonate units, that the many models proposed for their formation commonly suggest that a global precipitation event was inherent in their genesis. The term 'cap carbonate', however, is somewhat loosely used. In fact it is feasible to readily characterise 'cap carbonates' by specific mineralogical, sedimentological, petrographic, geochemical and stratigraphic features and hence differentiate 'cap carbonates' from other carbonate units which may be associated with a glacigenic succession. If a reasonable assessment of carbonate units based on these characteristics is undertaken, it is possible to be more confident in the use of 'cap carbonates' as stratigraphic marker horizons. In Neoproterozoic successions where geochronological age control is poor or absent, the 'cap carbonate' unit may be the strongest circumstantial evidence available to provide stratigraphic constraint. The carbonate unit overlying the Walsh Tillite in the Kimberley region of Western Australia is an excellent example where 'cap carbonate' characteristics are clearly applicable, hence providing evidence for correlation of this unit both regionally and

globally and identifying it as a Marinoan equivalent, now constrained at -630 Ma. Palaeomagnetic evidence, however, is inconsistent with this age interpretation and as a consequence the Walsh Tillite is alternatively interpreted as recording glaciation at —750 Ma. Furthermore, this putative interpretation has been used to support models and timing for Rodinian reconstructions and break-up, specifically the relationship between the Northern Australia Block (NAB) and South China Block (SOB). In the light of new geochronological evidence from three independent research groups in China and the United States which confirms a Marinoan age of ~ 6 3 0 M a from 'cap carbonates' from South China, the South China Block-Northern Australia Block relationship and timing needs to be reassessed. Additionally, the assertion that the Walsh Tillite is a Sturtian equivalent is also undermined.

While a Marinoan age for the Walsh Tillite seems the most likely interpretation on sedimentological and geochemical grounds, one further conundrum from the palaeomagnetic data remains unresolved. Palaeomagnetic evidence from the Marinoan aged Elatina Formation in South Australia suggests a palaeolatitude of 6° at 630 Ma, a result that is incompatible with the Walsh Tillite palaeopole, if we assume the Australian craton was intact by this time. Clearly, further attention needs to be given to the palaeomagnetic results, perhaps from both South and Northwestern Australia, in order to understand their true meaning within the context of Rodinia and global glaciations.

Supercontinents and Earth Evolution Symposium 2005

HIGH-RESOLUTION MAGNETOSTRATIGRAPHY OF AUSTRALIA'S MARINOAN CAP CARBONATES Timothy D Raub, David AD Evans Yale University, PO Box 208109, New Haven, CT 06520, USA (timothy.raub@yale.edu) Within the central Flinders Ranges, the Nuccaleena cap dolostone clearly reveals ancient geonnagnetic reversals correlatable about 50 km along strike, in three stratigraphic sections. Polarity reversals are identified near the base of the unit, midway through the continuous peloidal carbonate; and within the mixed transition between cap carbonate and red siliciclastics of the overlying Brachina Formation. Although geomagnetic reversal frequency throughout the last 200 Ma—our only guide for interpreting these patterns—has varied by at least two orders of magnitude, we may offer tentative suggestions for sediment accumulation rates through the glacial-postglacial interval. First, the stratigraphic juxtaposition of three reversals in Nuccaleena dolostone with at least six in the underlying, glaciogenic Elatina Formation suggests that geomagnetic reversal rate may have been generally high in the late Precambrian. Second, deposition of most of the Elatina Formation need not have lasted substantially longer than deposition of its cap carbonate.

Two more magnetic polarity reversals may exist, constrained in each of the three sections extremely close to the base of the cap carbonate by only one or two specimens as of May 2005. Centimetre-resolution magneto-stratigraphy at Elatina Creek and Enorama Creek, sections separated by ca. 5 km, will test the veracity of these "cryptochrons." Cenozoic geomagnetic reversal records indicate a transitional duration of ca. 2-8 thousand years, with briefer transitions in tropical latitudes, regardless of bounding chron duration. Adoption of transitional-direction interpretation for specimens of ambiguous polarity associated with these putative basal cap reversals would more confidently constrain both average sedimentation rate and total duration of sedimentation for the Nuccaleena dolostone. To first order in May 2005, the onset of cap carbonate deposition in the three Flinders Ranges sections would appear to be diachronous by ca. 3 thousand years, with peloidal carbonate deposition lasting ~110 thousand years.

Supercontinents and Earth Evolution Symposium 2005

METAMORPHIC PATTERNS IN ACCRETIONARY AND COLLISIONAL OROGENS - SECULAR VARIATION IN METAMORPHIC REGIMES AND PUNCTUATED TECTONIC EVOLUTION OF EARTH Michael Brown Laboratory for Crustal Petrology, Department of Geology, University of Maryland, College Park, MD 29742, USA (mbrown@geol.umd.edu)

INTRODUCTION There are two styles of orogenic system (OS) on modern Earth: accretionary orogenic systems (AOS), which form during ongoing plate convergence, that dominate during supercontinent break-up and dispersal; and, collisional orogenic systems (COS), which occur where an ocean closes, that dominate during assembly and formation of supercontinents. It follows that COS may be superimposed on AOS, although some AOS exist for of millions of years without terminal collision. AOS are of two types: extensional-contractional AOS, in which continental growth is related to mantle-derived magmatism in dominantly extensional arcs; and, terrane-dominated AOS, in which continental growth also occurs by accretion of allochthonous, commonly oceanic elements. I review metamorphic patterns in modern OS and compare them with ancient AOS and COS. Patterns of crustal metamorphism in the geological record indicate secular change. I argue for three distinct tectonic regimes in Earth history, but avoid the sterile semantics of what constitutes plate tectonics. METAMORPHIC PATTERNS For convenience, the metamorphic realm is divided into three types of metamorphism: LP-HT metamorphism reaching the granulite facies and UHTM conditions (T = >900X, P = -0.5-1.5 GPa); HP-HT metamorphism reaching the HP granulite/MT eclogite facies (T = -750->1000°C, P = -1.5->2.5 GPa); and, HP-LT metamorphism reaching blueschist/eclogite facies and UHPM conditions (T = -700->900X, P = ~2.5->6.0 GPa). These different types reflect decreasing thermal gradients; commonly HP granulite/MT eclogite facies rocks (and some UHPM rocks) exhibit overprinting by granulite facies conditions, including UHTM, recording HT decompression that reflects the process of exhumation. In modern extensional-contractional AOS, LPHT metamorphism is dominant, with looping or (ambiguous) CW or CCW P-T-t paths and peak metamorphic mineral growth late in relation to

tectonic fabrics. UHT and HP granulites are absent at outcrop; although rare, blueschists may occur early in the orogenic cycle, but UHPM is not recorded. Short-lived contractional phases of orogenesis probably relate to interruptions in the continuity of subduction caused by features on the ocean plate, particularly plateaus. Extensive granite (s.l.) magmatism accompanies metamorphism. In modern terrane-dominated AOS, 'paired' metamorphic belts are supposed to characterize the metamorphic pattern. However, this is a false construct that results from failure to recognize orogen-parallel terrane migration and the limitations of particular chronological datasets. Commonly a HP-LT (blueschist-eclogite) metamorphic belt occurs outboard and is separated from a LP-HT (andalusite-sillimanite type) metamorphic belt by a terrane boundary. In some systems an additional feature of the orogenic process is ridge subduction, which is reflected in the pattern of LP-HT metamorphism and the type of magmatism (adakitic). Granulites may occur at the highest grade of metamorphism in the LP-HT belt, where granite (s.l.) magmatism is common, but UHPM is rare in the outboard HP-LT belt. Although COS commonly are said to be characterized by syntectonic index minerals that record CW P-T-t paths and Barrovian metamorphic field gradients, generated by crustal thickening followed by exhumation, metamorphism to HP granulite/MT eclogite facies and extreme UHPM conditions also commonly occurs and may be more typical of COS on modern Earth. UHPM requires subduction and exhumation of continental crust, but it may be possible on the modern Earth for some continental crust metamorphosed under extreme UHPM conditions to be transported into the deep mantle rather than being returned to normal crustal depths. SECULAR CHANGE Data from ancient belts suggest secular change in thermal regimes and a punctuated tectonic

Supercontinents and Earth Evolution Symposium 2005

evolution of Earth. Archean high-grade gneisses record P-T conditions characteristic of granulite facies, whereas greenstone belts record lower T facies of the LP-HT type of metamorphisnn; blueschist facies rocks are not recorded, and although HT (type III) eclogite residues occur there is no widespread record of metamorphisnn of continental crust at mantle P. MT (type II) eclogites occur in the Neoarchean Belomorian Belt, Fennoscandian Shield (Earth's oldest crustal Omp-Grt rocks, ca. 2.72 Ga), in the Usagaran Orogen, Tanzania (ca. 2.0 Ga) and in the Grenvillian belts; they record initiation of oceanic lithosphere subduction. HP granulite/MT eclogite metamorphism is dominantly a Proterozoic-Phanerozoic phenomenon related to COS, and UHT granulite metamorphism is dominantly a Proterozoic phenomenon associated with crustal aggregation and supercontinent formation. Blueschists (HP-LT) first appear in the Neoproterozoic Era and become common through the Phanerozoic Eon,

reflecting the colder style of subduction typical of modern plate tectonics. UHP metamorphism also is dominantly a Phanerozoic phenomenon, related to deep continental subduction within COS. TECTONIC REGIMES The change in tectonometamorphic style through the Archean-to-Proterozoic transition (3.1-2.6 Ga) reflects mantle cooling and registers introduction of a Proterozoic plate tectonics regime, as evidenced by occurrence of AOS and COS, and MT eclogite and/or HP granulite in the Neoarchean-Proterozoic crustal record. The Neoproterozoic transition to the modern plate tectonics regime registers a change to deep subduction of oceanic lithosphere (to the CMB) and whole mantle convection as oceanic lithosphere became thicker with decreased thermal gradients, evidenced by occurrence of blueschist (HP-LT) belts and UHPM of continental crust.

Siipercontinents and Earth Evolution Symposium 2005

PALAEOPROTEROZOIC ECLOGUE EXHUMATION AND IMPLICATIONS FOR ANCIENT TECTONIC PROCESSSES Steven M Reddv\ Alan S Collins^ ^ Craig Buchan^ ^Tectonics Special Research Centre, Department of Applied Geology, Curtin University of Technology, PO Box U1987, WA 6845, Australia (S.Reddy@curtin.edu.au) ^Now at: School of Earth and Environmental Sciences, University of Adelaide, SA 5005, Australia

Subduction-related Palaeoproterozoic eclogites provide an opportunity to study the nature of tectonic processes in this important period of Earth history. One of the Earth's oldest examples of subduction-related eclogite facies rocks is preserved in the Palaeoproterozoic Usagaran orogenic belt of Tanzania. Gneiss exposed in the high-grade, eclogite-bearing part of this orogen (the Isimani Suite) records a complex deformation and thermal history that can be broadly subdivided into five events. The first of these (Di), associated with formation of eclogite facies metamorphism (-18 kbar & 750X), is strongly overprinted by a pervasive, amphibolitefacies ( - 6 kbar & 600X), high-strain deformation (D2). REE and U-Pb SHRIMP zircon data from the Usagaran eclogites indicate high pressure metamorphism at 1999 ± 1 Ma. A post-D2 pegmatite that cuts high strain D2 fabrics yields an intrusive crystallisation age of 1991 ± 2 Ma. These data require Di eclogites to be exhumed to 6kb and cooled by ca. 150°C over ca. 8 million years at integrated exhumation and cooling rates of -1.5 kbar/million years and -25X/million years. These rates are comparable to estimates derived for exhumation of subduction-related Tertiary eclogites. Temporally, eclogite exhumation must take place during D2 deformation. The geometry of foliations and lineations developed during D2 deformation are variable and are associated with variable kinematic shear directions. Electron backscatter diffraction studies of quartz within polymineralic samples from five kinematicallydefined D2 structural domains show the development of both orthorhombic and complex monoclinic CPOs (crystallographic preferred orientations). These data indicate partitioning of

pure and simple shear components, both between the documented structural domains and within individual samples. Rotation of c-axis patterns within domains demonstrating monoclinic symmetry confirms the shear direction interpreted from macro-scale kinematic indicators. Macro and mesoscopic analysis of the five D2 domains therefore indicates strain and kinematic partitioning linked to sinistral-reverse transpressional shearing during eclogite exhumation. Additional complexity within the orogen is evidenced by structural and geochronological data. Hornblende "^^Ar/^^Ar data record ages that vary from 1860 Ma in the west to 1630 Ma in the east. These ages are interpreted to represent Neoproterozoic telescoping of rocks recording post-eclogite cooling rates of <1°C/million years. Evidence for telescoping comes from greenschist facies structures (D4 & D5). These are localised as thrust and extensional shear zones, which separate D2 domains and indicate post-D2 reactivation of the Isimani Suite and thrusting of the Isimani Suite over volcano-sedimentary units of the Konse Group. The timing of syn-D4/5 deformation is constrained by Neoproterozoic data that indicate greenschist facies metamorphism during the Neoproterozoic East African Orogen. The data outlined above indicate exhumation and cooling rates comparable to those of modern eclogites. Detailed analysis of fabrics indicates complex flow kinematics also consistent with modern plate tectonic environments. We conclude that the Usagaran eclogites of Tanzania appear to provide support for plate tectonic processes, similar to those in the modern Earth, acting 2 billion years ago.

Siipercontinents and Earth Evolution Symposium 2005

LINKING SUPERCONTINENT ASSEMBLY, SUBDUCTION INITIATION AND ACCRETIONARY OROGENESIS Craig Buchan\ Peter A Cawood^ ^Tectonics Special Research Centre, Department of Applied Geology, Curtin University of Technology, PO Box U1987, WA 6845, Australia (c.buchan@curtin.edu.au) ^Tectonics Special Research Centre, University of Western Australia, WA 6009, Australia

Age relations for assembly of Gondwana and Pangea indicate that the timing of collisional orogenesis between amalgamating continental bodies was synchronous with subduction initiation and contractional orogenesis within accretionary orogens located along the margins of these supercontinents. In the case of Gondwana, final assembly occurred between ca. 570-510 Ma, amalgamating the various components of East and West Gondwana. This was coeval with a switch from passive margin sedimentation to convergent margin activity along the Pacific margin of the supercontinent. Timing of subduction initiation along the Pacific margin ranges from 580-550 Ma evidenced by the first appearance of arc derived detrital zircons in the upper Byrd group sediments and the oldest supra-subduction zone plutons along the Antarctic segment of the margin. A phase of extension in eastern Australia began around 580 Ma, resulting in eruption of the rift-related Mount Wright volcanics, and was punctuated by suprasubduction zone ophiolite generation between 535-500 Ma preserved in greenstone successions. This extensional episode immediately precedes and overlaps RossDelamerian contractional orogenesis between 520-490 Ma, inboard of the plate margin, coinciding with the cessation of collisional orogenesis between the amalgamating blocks of Gondwana. Supra-subduction zone igneous activity was continuous throughout this period indicating that subduction was on-going.

The final stages of assembly of the Pangean supercontinent occurred between ca. 320-250 Ma. Major plate boundary reorganization during this time was accompanied by regional orogenesis along the Pacific margin. The East Gondwanan margin segment experienced extension and strike-slip activity from ca. 310 Ma until ca. 270 Ma after which convergence along the plate margin was re-established, marked in eastern Australia by the migration of arc magmatism indicating a migration of the plate margin. Synchronous with this phase of plate readjustment was the Gondwanide Orogeny (305230 Ma) affecting the entire Pacific margin of Pangea. Temporal relations between interior collisional orogeny, associated with assembly of continental blocks, and marginal accretionary orogens during supercontinent amalgamation, suggest a linked history between interior and exterior processes related to global plate kinematic adjustments. Orogenesis in accretionary orogens occurs in the absence of colliding bodies during ongoing subduction and plate convergence and must therefore be driven by a transitory coupling across the plate boundary. Correspondence of coupling with, or immediately following, subduction initiation and plate boundary reorganization, suggest it may reflect plate readjustments involving a temporary phase of increased relative convergence across the plate boundary.

Supercontinents and Earth Evolution Symposium 2005

ACCRETIONARY GROWTH IN THE CENTRAL ASIAN OROGENIC BELT OF MONGOLIA AND KASAKHSTAN AND COMPARISON WITH THE ARABIAN-NUBIAN SHIELD Alfred K r o n e r ^ B r i a n F Windley^ G Badarch^ O Tomurtogoo^ E H e g n e r ^ S G r u s c h k a \ A D e m o u x \ DY L i u ^ MTD Wingate^ Înstitut fur Geowissenschaften Universitat Mainz, Germany (kroener@mail.uni-nnainz.de) ^Tectonics Special Research Centre, University of Western Australia, WA 6009, Australia ^Dept. of Geology, University of Leicester, UK "Înstitute of Geology Mineral Resources, Mongolian Academy of Sciences, Ulaanbaatar, Mongolia Înstitut fur Geo- und Umweltwissenschaften, Universitat Munchen, Germany ^Chinese Academy of Geological Sciences, Beijing 100037, China

The Central Asian Orogenic Belt (CAOB) records a ca. 800 Ma history of arc and microcontinent accretion, from S to N, during evolution and closure of the SW Pacific-type Palaeo-Asian ocean in the period -1000 to --300 Ma. We contest previous evolutionary models in terms of a single, large island arc. The earliest history of ocean opening is recorded by -1020-1050 Ma ophiolites in southern Siberia. The next younger events farther S are exemplified by the evolution of -850-570 Ma arc-ophiolite terrains, followed by earliest Palaeozoic subduction/accretion and suturing during which some of the arc terrains were metamorphosed to granulite-facies. Precambrian microcontinental fragments are documented by Archaean to Mesoproterozoic ages, and we also found Archaean to Neoproterozoic detrital zircons in Mongolian arcderived sediments, suggesting derivation from continental sources. Voluminous intermediate to felsic island-arc magmatic rocks in northern and central Mongolia were generated at 460-417 Ma within a broad belt extending from northern Mongolia to the Gobi Desert. Zircon xenocrysts and Nd model ages for some of these rocks are between 600 and >1300 Ma, suggesting that older material was involved in their generation. SHRIMP detrital zircon ages for arc-derived sandstones document continental input dating back to the late Archaean. By the end of the Ordovician, the northern part of the CAOB had amalgamated to create a new continental margin along the Main Mongolian Lineament. This probably is a major boundary separating two crustal provinces with different isotopic characteristics. Several microcontinental fragments have been identified

in southern Kazakhstan with zircon ages between 2.0 and 2.7 Ga. Extensive rifting with formation of bimodal volcanic sequences occurred at -775-820 Ma. Precambrian detrital zircons from clastic sediments resting on ocean floor supports the view that the oceanic domains between the microcontinents and early Palaeozoic arc complexes were narrow, probably marginal basins. A metadacite and granodiorite from an arc complex near Lake Balkash provided identical zircon ages of 478-480 Ma and contain zircon xenocrysts as old as 2290 Ma. One 480 Ma diorite sample from central Kazakhstan contains a 3.88 Ga zircon xenocryst. These rocks are unlikely to have formed in an intraoceanic environment, and we favour an Andean- or Japan-type setting. Overall the CAOB records the formation of SW-Pacific style small forearc and backarc ocean basins that probably evolved between island arcs and microcontinents and were closed during continuous accretion between the Neoproterozoic and Palaeozoic. Final closure of the Palaeo-Asian ocean probably occurred in the late Permian when the North China craton was attached to the CAOB. The presence of large volumes of felsic volcanic rocks in the Mongolian arcs, together with ubiquitous Precambrian xenocrystic and detrital zircons, argue for significant involvement of older material in the production of CAOB crust, and previous crustal growth models require substantial revision. Similar inheritance has recently been reported from seemingly juvenile Neoproterozoic felsic rocks of the ArabianNubian shield where a purely intra-oceanic development is now also in doubt.

Supercontiuents and Earth Evolution Symposium 2005

EMPLACEMENT OF THE WESTERN TETHYAN OPHIOLITES AND ATLANTIC OCEAN-FLOOR SPREADING Alan G Smith Department of Earth Sciences, University of Cambridge, Downing St, Cambridge, CB2 3EQ, UK (ags1 @esc.cam.ac.uk)

TETHYAN OPHIOLITES Broadly speaking, there are three separate "Tethyan" ophiolite suites in the Mediterranean region, sensu lato: one, the Apennine-LigurianAlpine ophiolites in Corsica, Italy and Switzerland; two, the Hellenic-Dinaric ophiolites in Greece, Albania and former Yugoslavia; and three, the ophiolites of Turkey, Cyprus, Syria, Iran and Oman. The Apennine-Ligurian-Alpine ophiolites are MORB-like, range in age from --169 to 148 Ma, and were emplaced in post-Jurassic time. The Hellenic-Dinaric examples are mostly suprasubduction (SSZ) ophiolites with some MORB, ranging in age from 173 to 168 Ma. They were emplaced, eroded and covered by younger sediments by -140 Ma. The ophiolites of Turkey and their SE continuation are also mostly SSZ, but are of midCretaceous age. TECTONIC SETTING OF THE HELLENICDINARIC OPHIOLITES Mid-Jurassic spreading on the newly formed central Atlantic ridge coincides with the creation of the Hellenic-Dinaric ophiolites. Metamorphic soles are 1-2 million years younger. The creation of both the "Alpine" and the Hellenic-Dinaric ophiolitic suites is attributed to the motion of Adria, which formed a promontory as the central Atlantic opened. Extension to the W of Adria created the Ligurian Sea, generating MORB crust. Simultaneous convergence to the E caused a pre-existing Triassic ocean to be subducted, accompanied by roll-back that created SSZ ophiolites and the metamorphic soles. The ophiolites were emplaced onto adjacent continental margins in the later stages of convergence at -140 Ma, or possibly earlier. EMPLACEMENT BY ROLL-BACK In general, the geometry of roll-back leads to emplacement of a SSZ ophiolite onto the margin that is being subducted. Because large-scale subduction of a continent appears to be precluded by the low density of continental crust, the subduction zone eventually seizes up.

TECTONIC SETTING OF THE MIDCRETACEOUS OPHIOLITES The initial setting of the mid-Cretaceous ophiolites is clearest in Oman. The Oman ophiolite appears to have formed during roll-back of Triassic (or possibly Permian) ocean-floor from the Iranian margin at about 95 Ma, and to have produced metamorphic soles that are also -1-2 million years younger than the ophiolites. Shallow marine limestones of early Maastrichtian to early Tertiary age lie conformably on the youngest foredeep sediments, but onlap unconformably onto the thrust sheets. AFRICA TO EUROPE MOTIONS AND OPHIOLITE EMPLACEMENT Despite the complexities and uncertainties in the Atlantic spreading data, the relative motion of Africa to stable Europe calculated from these data is remarkably simple. A major slowing of the relative motion from MO-50 mm yr'^ to less than 5 mm yr"^ takes place from --140-125 Ma, broadly synchronous with the cessation of Hellenic-Dinaric ophiolite emplacement. The motion then speeds up to - 2 0 mm yr'\ and slows again to - 4 mm yr"\ from 75 Ma to 55 Ma, broadly synchronous with the cessation of emplacement of the mid-Cretaceous ophiolites. The slowdown is attributed to the attempted subduction of the continental margin that formed the Africa-western Eurasia plate margin at the time. The speeding up of the motion that follows emplacement is attributed to the action of a new Africa-western Eurasia plate margin (or margins among which the convergence is distributed) elsewhere. Hence, it is conjectured that ophiolite emplacement, or its obverse, continental margin subduction, exerted a significant control on the spreading pattern of the N and central Atlantic. How these ideas can be applied to the emplacement of ophiolites where the relative velocities of the colliding fragments are unknown, is unclear.

Supercontinents and Earth Evolution Symposium 2005

THE TEMPORAL DISTRIBUTION OF MINERAL DEPOSITS: A STRONG REFLECTION OF TECTONIC AND LITHOSPHERIC EVOLUTION David I Groves\ RM Vielreicher\ RJ Goldfarb^ KC Condie^ JMA Hronsky"^ ^Centre for Global Metallogeny & Tectonics Special Research Centre, The University of Western Australia, Crawley, Australia (dgroves@cyllene.uwa.edu.au) ^United States Geological Survey, Denver, USA ^New Mexico Institute of Mining and Technology, Socorro, USA % M C , Belnnont, Australia

Mineral deposits exhibit heterogeneous distributions, with each major deposit type showing distinctive, commonly unique, temporal patterns. These reflect a complex interplay between formational and preservational forces that, in turn, largely reflect changes in tectonic processes and environmental conditions in an evolving Earth. The major drivers were the supercontinent cycle and evolution from plumedominated to modern-style plate tectonics in a cooling Earth. Consequent decrease in the growth rate of continental crust, and change from thick, buoyant sub-continental lithospheric mantle (SCLM) in the Precambrian to thinner, negatively buoyant SCLM in the Phanerozoic, led to progressive decoupling of formational and preservational processes through time. This affected the temporal patterns of gold-bearing deposit types including orogenic gold, porphyry and epithermal deposits, volcanic-hosted massive sulphide (VHMS), palaeoplacer gold, iron oxide Cu-Au (lOCG), and intrusion-related gold deposits. Orogenic gold deposits, which formed over at least 3.4 billion years, had the highest preservation potential of any gold deposit type. The pattern of formation and preservation, from episodic to more cyclic, broadly mirrors that of crustal growth. Early Precambrian (mostly ca. 2.7 and 2.0 to 1.8 Ga) deposits, protected from uplift and erosion in the centres of buoyant cratons, are rare between ca. 1.7 Ga and 0.6 Ga, owing to the change to more modern-style platetectonic processes, with non-preservation of deposits of this age due to uplift and erosion of more vulnerable orogenic belts. VHMS deposits were accreted into convergent margin terranes in which orogenic gold deposits were forming. Their temporal distribution, from strongly episodic to more cyclic peaks, also supports a model of selective preservation. The first appearance of lOCG deposits at -2.55 Ga closely follows development of early Precambrian SCLM. Their genesis involved

melting of metasomatised SCLM, so they could not form until such metasomatised SCLM evolved below cratons with buoyant lithosphere. Giant Precambrian palaeoplacer gold deposits probably formed by effective fluvial sorting under extreme climatic conditions but were largely preserved due to early buoyant SCLM below hosting foreland basins. Unequivocal intrusionrelated gold deposits are related to complex felsic intrusions with a mixed mantle-crustal signature, that intruded deformed shelf sedimentary sequences close to, but outside, craton margins. Given that postPalaeoproterozoic uplift and erosion is likely in vulnerable orogenic belts with negatively buoyant lithosphere, this deposit type is likely to be rare in Palaeozoic and older terranes. Yet other types of mineral deposits appear to require the prior existence of thick, stable, early Precambrian SCLM, with low geothermal gradients and incompatible element-rich metasomatism of marginal SCLM, for formation. Hence, these make their first appearance in the late Archaean or Palaeoproterozoic. These include primary diamond deposits and PGE deposits in layered intrusions. They show drastically different temporal patterns, however, with giant PGE deposits mainly restricted to the Precambrian (c.f. lOCG deposits) due to their relatively deep crustal level of emplacement, or protection by cover rocks. In contrast, economic diamond deposits became progressively more abundant towards the late Phanerozoic because of their high crustal level of emplacement and susceptibility to weathering of hosting kimberlites and lamproites. First-order controls on sedimentary rockhosted lead-zinc deposits are more problematic. The SEDEX deposits first formed in the early Mesoproterozoic, following the assembly of the first giant supercontinent and related generation of large rift basins with appropriate structural and sedimentary architecture, and again during the

Siipercontinents and Earth Evolution Symposium 2005

assembly of Pangea; precise processes are equivocal. Sedimentary mineral deposits mined for redox-sensitive metals show highly anomalous temporal patterns in which specific deposit types are restricted to particular times in Earth history. In particular, palaeoplacer uranium, BIF and BIFassociated manganese carbonates that formed

in the early Precambrian do not reappear in younger basins. The most obvious driver is progressive oxidation of the atmosphere, with consequent long term changes in the hydrosphere and biosphere, the latter influencing the temporal distribution and peak development of deposits such as MVTs, hosted in biogenic sedimentary rocks.

Supercontinents and Earth Evolution Symposium 2005

WHAT CONTROLS OROGENY AT A CONVERGENT MARGIN - THE ANDEAN CASE Onno Oncken GeoForschungsZentrum Potsdam, Telegrafenberg, 14473 Potsdam, Germany (mailto:oncken@gfzpotsdam.de)

Uplift of the Central Andean Cordillera commenced in early Tertiary times and accelerated since the Early Miocene. Plateau formation and the contribution of different processes (crustal shortening, magmatic addition, mantle delamination and hydration) are a matter of debate. Deep geophysical data across the Central Andes between 20°S and 24°S (ANCORP'96 and associated geophysical studies) indicate the widespread presence of partial melts or metamorphic fluids at mid-crustal level under the plateau between the Cordilleras bounding the latter. From structural balancing studies, these fluids or melts are associated with decoupling of upper crustal shortening and lower crustal thickening, hinting at the key role of thermal processes. Isotopic dating of syn-tectonic sediments of the Cenozoic intramontane basin system building the Altiplano, as well as seismic-sequence analysis, demonstrate that the Southern Altiplano crust was deformed with a complex partitioning of deformation between various subunits that were partly synchronized. The general acceleration of shortening rate shows only shows a weak link to plate convergence rates in the

early stages until the Middle Miocene. In contrast, our results show that the difference between upper plate velocity and oceanic plate slab rollback velocity is crucial in determining amount and rate of shortening as well as their lateral variability at the leading edge of the upper plate. This first order control is tuned by factors affecting the strength balance between the upper plate lithosphere and the plate interface of the Nazca and South American plates. These factors particularly include a stage of reduced slab dip accelerating shortening (33 and 20 Ma) and an earlier phase of higher trenchward sediment flux reducing plate interface coupling with slowed shortening and enhanced slab rollback (45 and 33 Ma). The combination of these parameters (in particular differential trench-upper plate velocity evolution, high plate interface coupling from low trench infill, and the lateral distribution of weak zones in the upper plate leading edge) was highly uncommon during the Phanerozoic, leading to very few plateau-style orogens at convergent margins like the Cenozoic Central Andes in South America or the Laramide North American Cordillera.

Supercontinents and Earth Evolution Symposium 2005

URSEIS'95: A MULTI-SEISMIC STUDY ACROSS THE SOUTHERN URAL OROGEN Ramon Carbonell, D Brown Institute de Ciencias de la Tierra "Jaume Almera", CSIC, Barcelona, Spain (rcarbo@ija.csic.es) The URSEIS '95 multiseismic experiment was an international initiative to constrain the crustal structure and physical properties of the Urals, a Variscan orogenic belt. The multiseisnnic experiment included normal incidence Vibroseis and explosive seismic reflection data to constrain the structure of the crust and upper mantle. Additionally, a wide-angle/refraction dataset was acquired to provide constraints on physical properties and seismic P and S wave velocities. Seismic images reveal a bivergent collisional orogen with a preserved crustal root reaching to --52 km. Three tectonic terranes that form the orogen can be distinguished: 1) the East European Craton (EEC), with a -40 km thick crust; 2) the core of the orogen with a thickness of 52-53 km, and 3) Siberian crustal terranes to the east with a thickness of - 4 3 km. Wide-angle/refraction seismic data provided the velocity structure, which is characterized by velocities of 6.3 ± 0.1 km/s for the upper crust for the P-waves and 3.9 ± 0.2 km/s for the S-waves. The middle and lower crust feature Vp's from 6.5 to 6.8 km/s, and a little over 7.0 km/s above the Moho in the central and eastern part. Vs ranges from 3.7 to 3.9 km/s, increasing to 3.9 to 4.0 km/s at the Moho. The crust-mantle transition is marked by an increase in Vp to values higher than 8.0 km/s and in Vs to values higher than 4.6 km/s. Stacking of the wide-angle shots provided images of the Moho across the orogen revealing its bivergent geometry. The Moho beneath the root zone does not provide a normal incidence PmP reflection whereas the wide angle shots, including the fan-recordings, reveal arcuate events in the deep crust in this area and a PmP with a 1 to 2 s long coda. Additionally, a relatively high frequency Pn phase can be identified at offsets beyond 225 km offsets. This feature rules out gradient models for the Moho boundary. While a layered model for the Moho provides a long coda it does not explain the lack of Moho in the normal incidence recordings. A 6 km thick, laterally and vertically heterogeneous crustmantle transition properly simulates the seismic signature at all offsets. The heterogeneous

structure can be achieved by laterally discontinuous layers (boudins) with a horizontal correlation length of 600 to 1000 m or by a Moho with an irregular topography on the scale of 5 km. These models can represent different crustal re-equilibration mechanisms: one characterized by magmatism and a second which features metamorphic processes. Finally, one of the most densely sampled wide-angle seismic reflection images of the lithosphere was obtained by this multi-seismic experiment. This unique dataset, recorded by an 18 km long spread, revealed that at wide angles the shallow subcrustal mantle features high amplitude reflectivity, which contrasts with a lack of reflectivity at later travel times. This change in the seismic signature is located at approximately 120-150 km depth, which correlates with the depth estimates of the lithosphereasthenosphere boundary (LAB) of previous Deep Seismic Sounding (DSS) studies. This seismic signature can be simulated by two layer mantle model. Both layers with similar average velocities differ in their degree of heterogeneity. The shallow heterogeneous layer and the deeper and more homogeneous one correlate with the lithosphere and the asthenosphere, respectively. Studies involving surface outcrops of ultramafic massifs and mantle xenoliths infer that the upper mantle is a heterogeneous mixture of ultramafic rocks (Iherzolites, harzburgites, pyroxenites, perid-otites, dunites, and small amounts of eclogites). Laboratory measurements of physical properties of these mantle rocks indicate that compositional variations alone can account for the wide-angle reflectivity. A temperature increase would homogenize the mixture, decreasing the seismic reflection properties due to melting processes. It is proposed that this would take place below 120-150 km (1200°C, the LAB). The URSEIS data has provided constraints on the construction of a petrophysical model of the crust and upper mantle across the Variscan Urals orogenic belt.

Siipercontinents and Earth Evolution Symposium 2005

MAGNETOTELLURIC IMAGING OF THE CENTRAL AUSTRALIAN LITHOSPHERE Kate Selwav\ Martin Hand\ Graham Heinson\ Ian Scrimgeour^ ^Continental Evolution Research Group, School of Earth and Environmental Sciences, University of Adelaide, SA 5005, Australia (katherine.selway@adelaide.edu.au) ^Northern Territory Geological Survey, GPO Box 3000, Darwin NT 0801, Australia

The central Australian lithosphere is currently the focus of significant effort aimed at refining accretionary models for the evolution of the Australian Proterozoic. The region is also the location of remarkable late Neoproterozoic to mid Palaeozoic intracratonic reworking. With the aim of providing three-dimensional spatial constraints to geological reconstructions, three magnetotelluric (MT) surveys have been carried out in Central Australia: in the southern Arunta margin, the eastern Alice Springs Orogen and the northern Gawler Craton. These surveys image proposed accretionary sutures and the crustal expression of intracratonic orogeny. MT is a passive electromagnetic method that produces resistivity cross-sections by analysing the earth's electric and magnetic fields. It is particularly useful for imaging and understanding orogenic processes because the resistivity of the crust is strongly affected by fluid flow, bulk composition and the production and destruction of mineral assemblages. Depths of investigation in orogenic studies are typically of the order of 50-100 km. MT is therefore an excellent addition or alternative to seismic reflection studies. SOUTHERN ARUNTA MARGIN The Redbank Survey is a 140 km long, northsouth survey located in the southern Arunta Region. The line crosses the Redbank Thrust Zone (RTZ), a -45° north-dipping structure which offsets the Moho by - 1 5 km whose most recent reactivation was during the intracratonic 450-300 Ma Alice Springs Orogeny (ASO). The line also crosses the boundary between the Proterozoic Warumpi and Aileron Provinces, which is inferred to be a major late Palaeoproterozoic accretionary boundary. The MT data indicate the existence of a sharp resistivity boundary (maximum gradient of 12000 Qm over 10 km) with a surface location coinciding with both the mapped RTZ and the boundary between the Warumpi and Aileron Provinces. This boundary separates resistive rocks of the Warumpi Province from conductive

rocks of the Aileron Province to the north. Its dip follows that of the RTZ to a depth of - 5 km but then becomes sub-vertical to the modelled depth of 40 km. We interpret this near-vertical boundary as reflecting juxtaposition of the juvenile Warumpi Province against the older Aileron Province of the north Australia Craton. Available geochronological data suggest this suturing occurred at -1630 Ma. EASTERN ALICE SPRINGS OROGEN The Eastern Arunta Survey is a 150 km long, north-south survey extending from the Amadeus Basin in the south to the Georgina Basin in the north, and as such is the first geophysical survey to cross the whole extent of the intracratonic ASO. Data show a bivergent orogenic architecture. Crustal-scale resistive zones correspond to outcrop locations of regions that underwent significant fluid-rock interaction during the Alice Springs Orogeny while conductive regions correspond to relatively fluid-unaffected rocks. The MT data suggest that much of the crust involved in the ASO underwent fluid infiltration during intracratonic reworking. NORTHERN GAWLER CRATON The --350 km long Northern Gawler Craton Survey extends from the Archaean Christie Domain north through the Palaeoproterozoic Nawa Domain and into the Mesoproterozoic Musgrave Block. Preliminary results show a sharp contrast in geoelectric strike between the Christie and Nawa Domains and a major resistivity contrast between the Gawler Craton and Musgrave Block. MT results, in combination with continuing geochemical and geochronological investigations, suggest that the Nawa Domain may have been exotic to the Gawler Craton at 1650 Ma. CONCLUSION MT surveys in central Australia have imaged proposed Proterozoic sutures and the crustal expression of intracratonic orogeny.

Supercontinents and Earth Evolution Symposium 2005

OROGENIC PROCESSES OF THE NORTHEAST TIBETAN PLATEAU, CHINA, FROM DEEP CRUSTAL SEISMIC PROFILING DATA Walter D Moonev\ Youxue Wang^, Xuecheng Yuan^, NIhal Okaya^ ^US Geological Survey, 345 Middlefield Rd., Menio Park, CA 94025, USA (mooney@usgs.gov) ^China University of Geoscience, 29 Xueyuan Lu, Beijing 100081, China ^China Geological Survey, 31 Xueyuan Lu, Beijing 100081, China

We discuss orogenic process across the northeastern Tibetan plateau based on activesource seismic data recorded along a 1600 km long profile crossing the southern Tarim basin, the western flank of the South-Qilian Shan, the northeastern margin of Qaidam basin, EastKunlun Shan, Songpan-Ganzi terrane, and Sichuan basin. The crustal P- and S-wave velocity structure and Poisson's ratio outline characteristics of the crustal structure and provide constraints on crustal composition. The derived crustal cross section shows several significant features. (1) The crustal thickness varies considerably. North of the Kunlun fault variations in crustal thickness and topography correlate well. The crust thickens from 48 km below the Tarim basin to 70 km beneath the northeastern margin of the Qaidam basin, and then thins again to about 56 km depth beneath the eastern flank of the Qaidam basin. The crust thickens again to 70 km beneath the East-Kunlun Shan. Across the Songpan-Ganzi terrane, the crust steadily thins from 70 km just south of the Kunlun fault to 48 km beneath the Sichuan basin, despite the fact the topography remains constant across the Songpan-Ganzi terrane and then abruptly drops by 10% in the Sichuan basin. (2) North of the Kunlun fault, variation in crustal

thickness is mainly caused by variations in lower crustal thickness, whereas south of the Kunlun fault they are caused by variations in upper-, middle- and lower-crustal thickness. (3) North of the Kunlun fault we detect a mid-crustal lowvelocity zone, which is not apparent south of the fault. Across the plateau, Poisson's ratio is nearly constant, with a value of 0.24-0.25 in the upper and middle crust indicating a felsic bulk composition. In the lower crust the Kunlun fault seems to act as a boundary with a Poisson's ration of 0.29 north of the fault (Kunlun-Qaidam) and 0.26 south of the fault (Songpan-Ganzi). Poisson's ratio and P-wave velocity values suggest that the lower crust throughout the Tibetan plateau (South-Qilian Shan, margins of the Qaidam Basin, East-Kunlun Shan, SongpanGanzi terrane) is of intermediate composition. Thus the NE Tibetan plateau along our profile is missing a mafic crustal layer at the base of the crust. The Tarim basin, which borders the Tibetan plateau to the north, shows a typical platform-like crustal structure with a felsic upper and middle crust, and a mafic lower crust. The Sichuan basin, which borders the Tibetan plateau to the east, also has a felsic upper and middle crust, and an intermediate or mafic lower crust.

Supercontinents and Earth Evolution Symposium 2005

ASSEMBLY OF WEST GONDWANA- PERSPECTIVE FROM THE DAMARA OROGEN COLLISIONAL TRIPLE JUNCTION David R Grav\ Ben Goscombe^, Richard A Armstrong^, David A Foster"*, Cees Passchier^, Rudolph Trouw® ^School of Earth Sciences, University of Melbourne, VIC 3010, Australia (email) ^Northern Territory Geological Survey, Alice Springs, NT 0871, Australia ^PRISE, Research School of Earth Sciences, Australian National University, ACT 0200, Australia "^School of Geological Sciences, University of Florida, Gainesville, FL 32611-2120, USA ^Institut fur Geoswissenschaften, JohannesGutenburg University, 55099 Mainz, Germany ^Institute de Geociencias, Universidade Federal do Rio de Janeiro, 21910-900, Brazil

New kinematic, geochronological and thermochronological data from the Damara Orogen of Namibia supports previously published temporally distinct suturing and amalgamation of South America (Rio de la Plata craton) with the Congo and Kalahari cratonic nuclei of southern Africa. This three-pronged orogenic system is essentially a collisional triple junction (after Hoffman et al., 1994) made up of a coastal arm, with north and south expressions as the Kaoko and Gariep Belts respectively, and an inland Damara Belt extending through the Lufilian Arc and Zambezi Belt into the Mozambique Belt of eastern Africa. Oblique convergence between the Rio de la Plata craton (South America) and the Congo and Kalahari cratons (Africa) involved sequential closure of the Adamastor Ocean, where the main ocean basin and/or sub-basins were closed first in the north and subsequently southwards from 550 Ma. Transpressive convergence in the Kaoko and Gariep coastal arms produced major westdipping listric shear zones that caused oblique crustal overriding of external or more outboard parts over and towards the African cratonic nuclei. Transpressional orogenesis in the Kaoko Belt occurred from 580-550 Ma and had clearly ceased by •^535 Ma, with cratonisation marked by intrusion of post-kinematic granite and pegmatite between 535-505 Ma. The Kaoko Belt is a thermally softened margin in transpression, dominated by major sinistral strike-slip shear zones defining regional scale shear lozenges within high-grade amphibolite fades Damara sequence turbidites, incorporating basement slivers and sheared Pan-African age granitoids. The Gariep Belt underwent transpressional convergence between --550 and 540 Ma (Frimmel & Frank, 1998) with erosion into the Nama foreland basin commencing at -540 Ma

(Gresse & Germs, 1993; Gresse, 1994; Frimmel, 2000) and intrusion of post-tectonic granites at -530-520 Ma (Allsopp et al., 1988). E-directed thrusting continued inboard within the Nama foreland basin through 496 Ma (Gresse et al., 1988). The Gariep Belt is mostly of low metamorphic grade, consisting of an arcuate belt of SW-vergent, stacked oceanic thrust-sheets including melange with blueschist blocks, metagreywacke turbidites and metabasalts, thrust over the passive continental margin of the Kalahari craton. The Damara Belt (Inland Branch) is a classic doubly vergent orogen with craton-directed thrust systems on both margins (e.g. Hakos and Naukluft nappes in the Southern Margin Zone and Southern Foreland). It records closure of the Khomas Ocean basin by high angle convergence between the Congo and Kalahari cratons, with a marked tectonothermal magmatic history extending from 570 Ma through -490 Ma with the intrusion of A-type granites (McDermott et al., 2000; Jung & Mezger, 2003). Ar-Ar data suggests Naukluft thrust-nappe emplacement on the southern foreland occurred at -570-550 Ma. Overall, the Damara and Gariep Belts appear to show both younger deformation and metamorphism than in the Kaoko Belt, although deformation may have been occurring in the Damara Belt by -570 Ma. Continued highangle convergence through 530 Ma in the Damara Belt coincides with the Shortening Phase deformation in the Kaoko Belt, reflected by large-scale open E-W trending folds and by major NE-trending kink-like flexures, local subvertical NE-trending crenulation cleavage and penetrative foliation in contact aureoles in the Ugab Domain (Southern Kaoko Belt). Data from the Damara Orogen suggest that the assembly of this part of West Gondwana had occurred by 480 Ma.

Supercontinents and Earth Evolution Symposium 2005

CRUSTAL STRUCTURE OF THE EAST-EUROPEAN CRATON: IMAGING EARLY PRECAMBRIAN ACCRETIONARY AND COLLISIONAL OROGENS Michael Mints\ Arsen Suleimanov^, Robert Berzin^ Nadezhda Zamozhniaya^, Vladimir Stupak^, Pavel Babayants^, Yury Blokh^, Alexey Trusov^, Irina Philippova^ ^Geological Institute of the RAS, Moscow, Russia (michael.nnints@mtu-net.ru) ^Spetsgeofisica, Povarovka, Moscow region, Russia ^Aerogeophysica, Moscow, Russia

New models of the deep crustal structure of the East European Craton (EEC) from the Barents Sea in the north and southward to the Caspian Sea have been created based on recently performed reflection seismic imaging of the EEC lithosphere down to 80 km depth (20 s) along the CDP geotransect 1-EU (3500 km long) and transverse profile 4B (270 km). A geological map of the platform basement beneath sedimentary cover, crossed by the 1-EU geotransect, was created on the basis of deep drilling data and maps of effective magnetization and density values at the basement surface computed using modern algorithms in the "Aerogeophysica". Crustal thicknesses within the Archean Karelian Craton, Volgo-Uralian and Sarmatian crustal segments and Paleoproterozoic Svecofennian accretionary orogen are 40-50, 6568, 48, and 45-50 km, respectively. The 5-15 km thick suture zones built by high-density and highmagnetic basalts, amphibolites, and sediments divide the Karelian Craton and Kola Province from the Paleoproterozoic Lapland - White-Sea - Mid-Russia - South-Baltica collisional orogen with crustal thickness of 50-60 km and Sarmatia from Volgo-Uralia. Length of these zones from the basement surface down to the Moho can reach 350 km.

Contrary to previous ideas that early Precambrian crust of the EEC is formed by multiple blocks with near-vertical boundaries and subhorizontal internal layering, it appears that crust as a whole is built mainly by successions of inclined tectonic sheets, which can be traced in some cases from the present day or basement surface down to the Moho. The "crocodile-type" relations between separate tectonic units are rather usual. Mantle areas adjacent to places where lower-crustal assemblages and sheets interpreted as suture zones reach the Moho discontinuity are characterized by irregularly distributed and partially oriented reflection events, which continue dimly the main features of crustal structure to the mantle. These areas may consist of "crust-mantle mixture" resulting from plunging of crustal sheets into the mantle. Final structure of the EEC was formed as a result of accretion and collision events until the end of the Paleoproterozoic. Their beginning at -1.9-1.85 Ga was marked by initiation of the Svecofennian accretionary orogen, and terminated with EEC creation at -1.7 Ga after closure of an ocean between Sarmatia and a composite Fennoscandia - Volgo-Uralia continent.

Supercontinents and Earth Evolution Symposium 2005

DEEP REFLECTIONS, LITHOSPHERIC STRUCTURE AND SUPERCONTINENT EVOLUTION Larry D Brown Institute for the Study of the Continents, 3120 Snee Hall, Cornell University, Ithaca, NY 14853, USA (brown@geology.cornell.edu)

Deep seismic reflection profiling over the past 30 years has accunnulated a substantial inventory of lithospheric innagery, much of it focused on orogenic systems of various ages and provenance. These results have both answered and generated important questions regarding the evolution of these orogens and the supercontinents they have stitched. Here I would like to highlight the following issues related to deep reflection imagery that transcend local interest. CRUSTAL HETEROGENEITY Although it is now taken for granted, at least among the seismic reflection community, the pervasive heterogeneity implied by most deep seismic reflection profiles is an important constraint on composition as well as structure. Any average composition derived from bulk seismic velocity measurements must be consistent with the preservation of the observed reflection contrasts at depth. In effect, deep reflectivity is a measure of lithological mixing on geological time scales. DEEP CRUSTAL FLOW A popular tectonic model in modern orogens is ductile flow in channels within the lower crust. Such flow would seem to be consistent with the subhorizontal lamination in the lower crust observed in some areas, particularly those representing post-orogenic thermal extremes. Broad zones of dipping reflectivity may represent comparable flow associated with major shear zones which now reach the surface. Cross cutting reflectivity in other cases, however, either post-dates flow or suggests alternative origins for layered reflectivity (e.g. intrusion). MAGMA ENTRAINMENT Seismic "bright spots" on a number of deep reflection profiles are indicative of fluids, most likely magma, in several orogenic systems. The apparent clustering of such features at 15-20 km depth (sometimes at the Moho), suggests rheological control. Correlation of such features in modern collisional orogens (e.g. Tibet) with granitic systems in older orogens (e.g. the Himalayas) provides fresh perspective on the role of partial melting in collisional orogeny.

MOHO TECTONICS "Flat" Moho morphology beneath extended terranes is often cited in support of post-orogenic collapse with attendant lower crustal flow. It is important to recognize, however, that a flat Moho requires a weak mantle as well as a mobile crust. Preservation of Moho "offsets" in some orogenic systems implies that mechanical re-equilibration is not always complete. Examples of angular unconformity between crustal and Moho reflections implies that the crust-mantle boundary is a likely candidate for intralithospheric detachment. Conversely, examples of continuity in crustal reflectivity "across" the Moho has been cited as fresh support for the phase change explanation for the seismic Moho locally. SUBDUCTION SCARS AND PLATE TECTONICS Among the most dramatic of deep reflections are the pronounced dipping events which emerge out of the lowermost crust and dip well into the underlying lithospheric mantle. These events not only contrast with the more homogeneous nature of the mantle in general, they represent fossils of orogenic assembly that serve as markers for post-orogenic preservation of mantle lithosphere. The most common interpretation of these events is that they are traces of subducted oceanic lithosphere. Their existence in Archean systems is strong evidence for plate tectonics in early Earth history. Not all mantle reflections have the geometry of subduction scars. However, they too serve as key markers in evaluating tectonic or magmatic transfer of material to or from the lithosphere. DELAMINATION (OR NOT) It is hard to envision delamination of lithospheric crust which is consistent with the preservation of subduction scars. However, incipient delamination involving the lower crust is the compelling interpretation of deep seismic profiles from at least some active collisional systems. Best known of these are the Alpine transects; another is a deep seismic survey across the Hidaka collisional belt on Hokkaido. The degree to which such systems continue to excavate lower crust and remove it from the lithosphere is a key consideration in the

Supercontinents and Earth Evolution Symposium 2005

refinement of bulk continental connposition. The character of the subduction scars would seem to argue that such tectonic delamination is relatively inefficient; however, they say little about foundering mechanisms. DISCUSSION The interpretation of reflection imagery on individual deep seismic profiles tends to be biased toward the latest tectonic events evident at the surface. The complex histories implicit in the amalgamation and dispersal of lithospheric fragments to form and reform the various supercontinents should serve as a reminder that reflectivity records the accumulation of tectonic offenses to the body lithospheric. Rather than discourage interpretation, this recognition places greater value on synthesis of individual results

along and between orogenic systems. Important gaps in coverage remain, but recurrent similarities suggest that such syntheses can be both tractable and robust. CONCLUSION Reflections mapped by deep seismic surveys form a basis of markers for tracing the evolution of continental lithosphere though supercontinent cycles. The heterogeneity of reflection patterns is commensurate with the diversity of tectonic process that imprints upon the lithosphere. However, the patterns are far from chaotic. Key features are often preserved as testimony to the relative importance of mobilization and preservation of lithospheric components through time.

Supercontinents and Earth Evolution Symposium 2005

THURSDAY 29 SEPTEMBER 2005

Supercontinents and Earth Evolution Symposium 2005

PALEOMAGNETISM AND SUPERCONTINENTS: RESOLVING THE RODINIA PUZZLE Rob Van der Voo Department of Geological Sciences, University of Michigan, Ann Arbor, Ml 48109-1005, USA (voo@umich.edu)

The supercontinent Rodinia is generally accepted to be premised on the following concepts: (1) the roughly contemporaneous Grenville, Kibaran and Sveconorwegian orogenic belts mark the assembly phase of the supercontinent, (2) the Neoproterozoic passive margins of the Laurentian continent indicate that it occupied a central position in Rodinia, and that Rodinia break-up began at about 800 Ma, and (3) the cratonic elements that later constituted Gondwana were dispersed in various periLaurentian positions. The actual configuration of Rodinia, however, remains poorly constrained by these premises and paleomagnetic data are

therefore essential in order to make progress. After it seemed, at first, that a Rodinia reconstruction roughly similar to that of P. Hoffman was supported by the available paleopoles, subsequently published new data have generally created more confusion than coherence in creating an acceptable Rodinia reconstruction. In this presentation, the available paleomagnetic data for the cratonic blocks in Rodinia will be critically reviewed; however, the general conclusion must be that several tectonic elements remained separated from other blocks by oceanic realms within Rodinia during its 1100 - 800 Ma history.

Supercontinents and Earth Evolution Symposium 2005

PROGRESSIVE PROTEROZOIC DEVELOPMENT OF SOUTHERN LAURENTIA: NEW MAP AND LINKED DATABASES Karl E Karistrom\ Steven J Whitmeyei^ ^Dept. of Earth & Planetary Sciences, Univ. of New Mexico, Albuquerque, NM 87131, USA (kek1@unnn.edu) Dept. of Geology & Environmental Science, James Madison University, Harrisonburg, VA 22807, USA IGCP 440 has resulted in a new map compilation with sequential time slices and linked databases. Laurentia offers one of the most complete geological, geophysical and isotopic data sets of any continent and hence provides a template to test piercing points. We attempt to show all pre750 Ma tectonic elements that may be useful for reconstructions. Rifted margins, because they are thinned and weakened during extension are invariably overprinted by younger orogens, hence the importance of combining Nd and Pb isotopic data with detailed geology (e.g. for the Mojave province and pre-Appalachian rifted margin). Detrital zircon and monazite studies, including "fingerprinting" of monazite zoning patterns, locally provides information about distinctive spacing of tectonic events and provenance outside Laurentia. The goal is an integrated understanding of the geological history of the Laurentian craton that will ultimately provide the best confidence in linking once-adjacent blocks. The southwestern U.S. preserves a rich record of Proterozoic accretion of juvenile terranes and processes related to their stabilization into continental lithosphere. A zone >1000 km in width was added to Laurentia between 1.8 and 1.0 Ga by successive additions of dominantly juvenile oceanic terranes and magmatic arcs to a long-lived compressive/ transpressive plate margin. Early juvenile terrane accretion during the -1.88-1.83 Ga Penokean orogeny was followed by reworking of Archean (?) and Paleoproterozoic (1.84 Ga) basement within the 1.78-1.68 Ga Mojave province. Accretion of juvenile bimodal volcanic/plutonic successions continued with the 1.8-1.7 Ga

Yavapai and 1.67-1.60 Ga Mazatzal orogenies. Subsequent addition of -1.5-1.3 Ga juvenile crust (i.e. present-day mid-continental U.S.) suggests accretion of outboard magmatic arc(s) linked temporally to -1.45-1.35 Ga A-type backarc magmatism that intruded and stabilized older Proterozoic provinces. Accretion of juvenile crust to southern Laurentia culminated with the 1.3-1.0 Ga Grenville orogeny: the final stage in the assembly of Rodinia prior to collapse and breakup between -0.8 and 0.55 Ga. Assembly of southern Laurentia probably took place by subduction-accretion processes as recorded in subduction scars imaged by combined seismic and geologic studies across several major shear zones and province boundaries. Suture geometries include interwedged crustal terranes, deep crustal bivergent root zones, and non-uniform mantle interfaces. Shortening and thickening of crust due to progressive accretionary events and later reactivation has transposed shallow boundaries into predominantly subvertical fabrics. Syn- and post-accretion lithospheric differentiation produced blooms of granitoids that likely were accompanied by progressive depletion of lithospheric mantle and development of a mafic lower crust, further stabilizing the lithosphere. Subduction-related assembly of Proterozoic juvenile terranes in southern Laurentia resulted in anomalously hydrous, fertile, and thick (> 200 km) chemical lithosphere (compared to Archean lithosphere). Heterogeneity preserved in today's mantle at the 10 km scale reflects differing responses of Proterozoic compositional domains to later hydration and heating events, including modern asthenospheric upwelling.

Supercontinents and Earth Evolution Symposium 2005

A REVISED POLE FOR THE -1070 Ma (?) ALCURRA (KULGERA) INTRUSIONS Phillip W Schmidt\ George E Williams^, Alfredo Camacho^ ^CSIRO Exploration & Mining, PO Box 136, North Ryde, NSW 1670, Australia (Phil.Schnnidt@csiro.au) Discipline of Geology and Geophysics, University of Adelaide, SA 5005, Australia ^Department of Geological Sciences, Queens University, Kingston, Canada

The Alcurra Dykes (AD), formerly known as the Kulgera Dykes, intrude the Musgrave Craton, which is a mid- to late Proterozoic mobile zone of metamorphic and intrusive rocks covering 120 000 km^ in the centre of the Australian continent. The AD form an east-west arcuate belt 90 km long by 10 km wide, with dykes commonly 500 m long, 2 m wide, sub-parallel or anastomosing and shallowly dipping (5° to 30°). They tend to dip ESE in the east and S to SSE in the west, where their greatest concentration occurs. The AD has yielded Sm-Nd and Rb-Sr isochron ages (2a) of 1090 ± 32 Ma and 1054 ± 13 Ma, respectively. The dykes are intruded by granites having a zircon U-Pb SHRIMP age of 1071 ± 5 Ma, which therefore provides a lower constraint for the age of the AD. It has been suggested that the AD form part of the Warakurna large igneous province (LIP) which is thought to be -1070 Ma. In this case the 'granite' may represent a highly fractionated late-stage intrusion, and the age of 1071 Ma may also apply to the AD. Palaeomagnetic directions from the AD have previously been found to fall into two groups. The less stable magnetizations appear to be overprints associated with the Carboniferous Alice Springs Orogeny, while the more stable magnetizations are similar to those from the related Stuart Dykes (SD). We have added new data from two more dykes whose orientations are different from those previously studied. This has allowed the primary nature of palaeomagnetic direction to be demonstrated through applying a fold test. The directions pass the fold test with 99 percent confidence and the revised mean direction is declination = 291.2°, inclination = 50.8° and ags = 8.0°, yielding a pole position at latitude = 7.7°S, longitude = 75.0°E (dp = 4.4°, dm = 8.0°). The revised pole position is very similar to that of the SD. While not enough information is available to allow a rigorous comparison of the SD pole with the revised AD, if it is assumed that at least ten Stuart dykes were sampled then it is possible to estimate the relevant Fisher statistics

to allow a tentative McFadden-Lowes test to be applied. Using N = 10 and K = 24 for the SD yields a test statistic of 3.156 which is less than F(2,34) = 3.274. This indicates that the hypothesis that the SD and AD have a common mean direction cannot be rejected. The high degree of overlap of the 95 percent confidence ovals about the poles also indicates that the SD and the AD poles are very similar. Although the ages of the SD and AD are not precisely determined, their pole positions would seem to be reliable and represent poles for --1070 Ma for the Arunta Craton and the Musgrave Craton respectively. Mafic sills from the Bangemall Superbasin, which are apparently of similar age to the SD/AD and are also thought to be part of the Warakurna LIP, have yielded directions that are quite different (mean declination = 339.9°, inclination = 46.5° and ags = 8.4°). We note that the 'Gawler B' (GB) dykes from the Gawler Craton possess directions broadly similar to the AD/SD with a declination = 272.5°, inclination = 61.1° and ags = 8.5°. These may be feeders to the Beda Volcanics, which have also been suggested to form part of the Warakurna LIP. These palaeomagnetic results therefore represent an area covering the Gawler Craton, the Musgrave and Arunta Cratons and the edge of the Pilbara Craton. If these four sets of dykes are indeed contemporaneous then their directions should be similar, unless they have undergone different rotations about vertical axes since the time of intrusion. Such rotations have been demonstrated within the Superior Craton of North America, where directions and strikes of the Matachewan Dykes can be aligned by applying a simple rotation scheme. Whether the rotations suggested here were distributed evenly across the cratons similar to the deformation of the Superior Craton, or occurred in narrow belts between cratons, remains to be determined but some rotation appears to be unavoidable if the Stuart Dykes/Alcurra Dykes (±GB Dykes) are contemporaneous with the Bangemall sills.

Supercontinents and Earth Evolution Symposium 2005

NEW PALEOMAGNETIC DATA FOR THE 1.1 Ga DIABASES OF ARIZONA - IMPLICATIONS FOR RODINIA ASSEMBLY Lauri J Pesonen\ Kimmo Korhonen^ Alex Deutsch^, Fablo Donadini\ Stephen Harlan"*, Heikki Nevanlinna^ ^Division of Geophysics, PO Box 64, FI-00014 University of Helsinki, Helsinki, Finland (lauri.pesonen@helsinki.fi) Geological Survey of Finland, PO Box 96, FI-02151 Espoo, Finland ^Institute of Planetology, University of Munster, D-48149 Germany "^George Mason University, Fairfax, VA, USA ^Finnish Meteorological Institute, Space Research, PO Box 503, FI-00101 Helsinki, Finland

Large Igneous Provinces (LIPs) represent times when huge volumes of mafic magma were emplaced in a relatively short period of time. Links to tectonic events, such as rifting, assembly or break-up of supercontinents, or to mantle plumes have been suggested. One of the most dramatic LIP events is the 1.11 - 1.08 Ga Keweenawan magmatism in North America, which can be linked to the Mid-Continent Rift. This event can be correlated with the diabase province of the SW USA and with the Moore's Lake intrusive event in Saskatchewan. A presumably coeval event is the Umkondo magmatic activity in South Africa. These events coincide with the Grenvillian (North America) and Natal-Namaqua (southern Africa) collisional events and in this way link them together in the assembly of Rodinia. THE PROBLEM It is a major challenge to prove paleomagnetically that the 1.11 - 1.08 Ga LIPs are related to the amalgamation of Rodinia. However, paleomagnetic data for North America show significant inclination asymmetries between the normal and reversed polarities, leading to uncertainties in estimating the position of Laurentia at 1.11 Ga. The asymmetries, so far identified with certainty only in data from the Lake Superior region, are explained by four

models: (i) continental drift during the reversal crossing (from R to N), (ii) unremoved secondary components, (iii) non-averaging of secular variation in paleomagnetic data and (iv) a significant contribution by non-dipole field components at 1.11 Ga. To test which of these models is most appropriate, additional paleomagnetic data are needed for -1.1 Ga rocks from North America and other continents. NEW SAMPLING To shed further light on the problem, we initiated a project which includes a global analysis of 1.1 Ga paleomagnetic, paleointensity, and paleosecular variation data. For this reason we carried out new sampling of the sills and dykes of the Gila County and Sierra Ancha Mountains of southwestern Arizona. The new samples are studied petrographically and the least altered ones will be dated by isotopic techniques. Preliminary paleomagnetic results for the Arizona samples will be presented and compared with those from the Lake Superior region of North America, from the Umkondo Province of South Africa, and from probably coeval sediments of Siberia. Possible non-dipole field contamination will be identified with novel spherical harmonic techniques applied to the 1.1 Ga paleomagnetic data.

Supercontinents and Earth Evolution Symposium 2005

PALEOMAGNETISM OF MIDDLE RIPHEAN DYKES FROM THE LADOGA LAKE REGION OF NORTHERN KARELIA Natalia Lubnina\ Satu Mertanen^, Tatiana Vasilieva^ ^Moscow State University, Geological Department, Moscow, Russia (lubnina@rannbler.ru) ^Geological Survey of Finland, PO Box 96, FI-02151 Espoo, Finland ^Geological Institute, RAS, Russia Following the Svecofennian orogeny at ca. 1.91.8 Ga, rifting processes took place in the Fennoscandian Shield during the Middle Riphean. Most of this activity was concentrated in the western Fennoscandian Shield, where magmatic complexes of rapakivi granites and associated dykes, and rift structures with basic dykes were formed. In Russia, Middle Riphean complexes occur in the Lake Ladoga region. A dyke swarm of Fe-rich olivine dolerites was sampled for paleomagnetic investigations. Chemical composition of the dykes corresponds to the Salmi basalts dated at 1457 ± 2 Ma which is therefore regarded as the age of the dykes (Vasilieva & Frank-Kamenetsky, 2002). Two types of dykes are present. Type 1 dykes are thin (up to 1.5-2 m) aphanitic dolerites, which have narrow contact zones. These dykes were collected near the village of Helyla, at SuurHaapasaari, and at the northwestern and southeastern parts of the Tamkhanka islands. Type 2 dykes also consists of Fe-rich olivine dolerites, but are up to 30 m wide with contact zones of ca. 10-15 cm. About 30 to 50 oriented samples were collected from each dyke and about 5-10 samples from the host gneisses and granites at each dyke. Remanence measurements were performed using a 2G cryogenic magnetometer (Geological Survey of Finland) and JR-5A spinnermagnetometer (Paleomagnetic Laboratory of VSEGEI, St. Petersburg). Both alternating field and thermal demagnetizations were conducted. The aphanitic Type 1 dolerites carry two remanence components. The intermediate T coercivity component has Tub up to 380-440°C and coercivities <40mT and the high 'H' coercivity component has Tub up to 590°C and coercivity » 9 0 m T . The T component is of single polarity and is directed to the northeast with shallow negative (upwards) inclination. A similar remanence component pointing northeast and slightly upwards was identified also in the granites, taken from baked contact zones to the dykes and in the gneisses, taken about 2-5 meters from the contact zone. Based on this negative baked contact test, we therefore argue

in favour of a secondary origin for the T component. The 'H' component is directed to the northeast with positive (downwards) inclination. It is demagnetized between 380°(40mT) and 590°C (lOOmT). Granite samples in the baked contact zone have a similar remanence direction as the dykes, but samples ca. 2-5 meters from the contact carry a northwest-directed downwards direction. We argue that the 'H' component in type 1 dykes is primary, based on the positive contact test. Type 2 dykes are subdivided into weakly (NRM 0.7-1.5 A/m) and strongly (NRM 2.5-4.3 A/m) magnetized samples. Samples carrying mainly single-component NRM were collected from endo- and exocontact zones of the dykes. This remanence component was unblocked gradually up to 590°C. The mean direction of this component is close to the T direction. However, samples from the central part of the dykes have a multicomponent NRM. A low-stability remanence component was removed between 200° - 440°C (<20-40 mT) and is directed to the northeast with negative inclination. Its mean direction is close to the T direction. According to thermomagnetic analysis the principal carrier of the 'H' component is low-titanium titanomagnetite. Most of these specimens show dual polarity, which indirectly supports a primary origin for this component. Paleomagnetic poles were compared with Precambrian 'key poles' for the Fennoscandian Shield (Buchan et al., 2000) and with the Paleozoic apparent polar wander path (APWP) for Baltica (Smethurst et al., 1998). The paleomagnetic pole corresponding to T component directions lies on the Late Paleozoic part of the Baltica APWP, implying that it may be related to late sedimentary processes after the Caledonian orogeny. The paleomagnetic pole corresponding to 'H' component directions (Lat. = 16.2°, Long. = 177.0°, dp = 2.7°, dm = 5.3°) falls between the 1.27 Ga and 1.58 Ga 'key poles' for the Fennoscandian Shield (Buchan et al., 2000) and places the shield in a near-equatorial position at that time.

Super continents and Earth Evolution Symposium 2005

MESO- AND NEOPROTEROZOIC TECTONIC AND MAGMATIC EVENTS ALONG THE WESTERN MARGIN OF SIBERIA: ASPECTS RELATED AND NOT RELATED TO RODINIA ASSEMBLY Valerv A Vernikovskv, DV Metelkin United Institute of Geology, Geophysics and Mineralogy, Siberian Branch of the R A S , Novosibirsk, Russia

(tainnyr@uiggnn.nsc.ru)

The configuration of the Meso- to Neoproterozoic Rodinia supercontinent and the position of Siberia in Rodinia rennain subjects of nnuch debate (Pavlov et al., 2002; Meert & Torsvik, 2003; Pesonen et al., 2003; Pisarevsky & Natapov, 2003; Vernikovsky & Metelkin, 2004). Fornnation of the Rodinia supercontinent was a result of Grenvillian (late Meso- to early Neoproterozoic) collisional events and its breakup was connected with Neoproterozoic rifting. The analysis of geological structures of Mesoand Neoproterozoic passive and active continental nnargins of Siberia and their evolution, together with paleonnagnetic studies, can provide innportant constraints on the existence of Siberia in the Rodinia supercontinent and its position relative to other continents. There is much evidence that the western and northwestern margins of Siberia were passive continental margins in the late Meso- and early Neoproterozoic. The transformation of these into active margins occurred in the early Neoproterozoic. New data from the Central Taimyr accretionary belt provides evidence that an island arc system existed near the Siberian margin by 960 Ma. Paleomagnetic poles have

been determined for the 960 Ma Three Sisters island arc felsic volcanic rocks. These poles are very close to similar-age poles for Siberia (Pavlov et al., 2002), but demonstrate about 30° difference in declination and less than 9° difference in latitude. This is in good agreement with the view that the Three Sisters island arc developed at similar latitudes as the Taimyr margin of Siberia and was moved towards the continent with counter-clockwise rotation prior to amalgamation. We have no evidence about Grenvillian age (1100-1000 Ma) orogenic belts along the (present) western margin of Siberia. However, there are abundant 940-850 Ma granites in the Yenisey Ridge and Taimyr fold belts (MamontShrenk, Faddey, and Central Angara terranes). These granites have a very similar composition and age. Therefore they could have formed in any Grenvillian - Post-Grenvillian age orogenic belt outside of Siberia prior to their collision with Siberia. This paper presents new paleomagnetic data for the northwestern and western margins of Siberia, and discusses the Mesoand Neoproterozoic Siberian APWP and implications for the relative positions of Siberia and Laurentia.

Supercontinents and Earth Evolution Symposium 2005

STRATIGRAPHIC TESTS OF THE PROTEROZOIC SOUTHEAST SIBERIA - SOUTHWEST LAURENTIA CONNECTION James W Sears\ Andrei K Khudoley^, Andrei V Prokopiev^ ^University of Montana, Missoula, MT, 59812 USA (james.sears@umontana.edu) ^St. Petersburg State University, St. Petersburg, 199034, Russia ^Diamond and Precious Metal Geology Institute, 39 Lenin Ave, Yakutsk, 677891, Russia

New stratigraphic data from southeastern Siberia and Death Valley, USA, test a proposed Mesoproterozoic connection of the eastern margin of the Siberian craton against the southwestern margin of Laurentia that was mainly based on basement piercing-points.

The reference Mesoproterozoic section of the Siberian craton crops out along the Belaya River gorge, which crosses the Gornostakh anticlinorium in the Sette-Daban range of the Verkhoyansk Mountains. In the proposed restoration, this section plots within 200 km of correlative Mesoproterozoic strata exposed in the Death Valley region. The Aimchan, Kerpyl, and Lakhanda groups of southeastern Siberia compare with the lower two formations of the Pahrump Group of the Death Valley region (Crystal Spring and Beck Spring formations). Both sections begin with a profound angular unconformity, overlain by trough-crossbedded, pebbly sandstone (Talyn Formation and basal Crystal Spring sandstone). The sandstone grades into red shale that passes upward into cherty, stromatolitic, pink and orange dolomite (Svetiyi Formation and Crystal Spring dolomite). The dolomite in both regions is capped by a distinctive, decimeter-thick quartz-and-dolomite pebble conglomerate, and succeeding red shale and coarse, trough cross-bedded red and white sandstone (Bik Formation and upper clastic member of Crystal Spring Formation). Both sections next contain a distinctive brown dolomite and yellow shale (Muskel Formation and uppermost Crystal Spring Formation). Next comes a black, petroliferous dolomite containing massive breccias, slump folds, microbial mats, and 2-3 m tall cyanobacterial domes (Malgin and Tsipanda Formations and Beck Spring Dolomite). Depositional cycles from red and yellow shale to black stromatolitic dolomite with pisolites (Lakhanda Group and western facies of

Beck Spring Formation) completes the sequence of similar platformal facies. For both cratons these platformal facies are younger than ca.1.38 Ga intrusives in the unconformably underlying rocks, and are older than sets of diabase sills dated to ca. 1.0 Ga. The Siberian section is systematically thicker than the Death Valley section. Both sections grade into more basinal facies; Siberia from west to east, and Death Valley from east to west, potentially forming two sides of an intracratonic basin. Both platformal sections are abruptly overlain by deep-water turbidite facies (Uy Group and basal Kingston Peak Formation), indicating sudden basin collapse. There are associated mafic intrusions and flows in both turbidite sections. One significant difference between the two regions is that the Siberian section lacks the diamictites for which the Kingston Peak Formation is well-known. The basal Uy Group contains sills dated to 930-1000 Ma, but the upper Uy Group is not dated. The Vendian-Early Cambrian sections of the Sette-Daban region match those of the WhiteInyo Range of eastern California better than those of Death Valley, suggesting separation along a rift-transform system may have begun. Both sections contain thick cream-colored dolomite and interbedded sandstone that unconformably overlie older beds (Yudoma Group and Reed dolomite). The uppermost part of the dolomite in both sections contains distinctive cloudiniid fossils. The massive dolomite is overlain by interbedded stromatolitic dolomite and quartz arenite (Sytyga and Deep Springs formations); both contain passage beds from Vendian into Early Cambrian. Both sections have an abrupt appearance of black siliceous shales (Akra and Campito formations) and contain similar early trilobite faunas. Middle Cambrian limestone-shale sections are also very similar.

Supercontinents and Earth Evolution Symposium 2005

RODINIA: THE AFRICAN PERSPECTIVE Richard E Hanson Department of Geology, Texas Christian University, Fort Worth, TX 76129, USA (r.hanson@tcu.edu)

Africa contains four main cratons (West African, Saharan, Congo, and Kalahari) separated by Neoproterozoic-Early Paleozoic (Pan-African) orogenic belts, and the original positions of these cratons within (or removed from) Rodinia remain controversial. To date, only the Congo and Kalahari cratons in central and southern Africa have yielded a significant record of Mesoproterozoic tectonism that may bear on models of Rodinia evolution. Within the Congo craton, the Angola-Kasai and Bangweulu-Tanzania blocks are separated by the Kibaran belt, which records convergence (of uncertain magnitude) at 1.38 Ga, well before the main phase of Rodinia assembly. Other tectonothermal events occurred in the belt at 1.21.0 Ga, the younger of which may link with orogenesis along the SE edge of the Bangweulu block. There, within the Irumide belt, a 1.9-1.6 Ga terrigenous sequence deposited on 2.7-1.9 Ga crust was deformed and intruded by voluminous granitoids at 1.05-0.95 Ga. Farther SE, a complex, poorly understood region extending to the coast of Mozambique contains 1.1-1.0 Ga continental-margin and oceanic arc terranes, associated locally with a 1.4 Ga ophiolite and elsewhere with possible microcontinental fragments containing older crust. This region appears to represent a longlived convergent margin active within the time frame of Rodinia assembly, but Pan-African overprinting makes it difficult to establish original relations between terranes. To the south, 1.1-1.0 Ga orthogneisses also occur within the Zambezi and Mozambique belts along the north and east margins of the Kalahari craton, but the extent to which these orthogneiss terranes were displaced during Pan-African orogenesis is unclear. Another major Mesoproterozoic convergent plate boundary is defined by the Namaqua-NatalMaud orogen along the southern margin of the Kalahari craton. This orogen records arc magmatism, terrane accretion, and syn- to latetectonic granite plutonism at 1.38-1.02 Ga. The Sinclair terrane and Rehoboth inlier in the western part of the orogen wrap the SW corner of the Kalahari craton but then disappear farther NE beneath extensive cover in the Kalahari Desert. These terranes are inferred to connect with the 1.35-1.2 Ga Choma-Kalomo block

exposed in the NW part of the Kalahari craton in southern Zambia; the NW Botswana rift is superimposed on this largely buried belt. Bimodal magmatism in the NW Botswana rift represents the western portion of the Umkondo large igneous province, which developed over much of the Kalahari craton at 1112-1106 Ma and may have an important bearing on Rodinia models. Close similarity in timing of this igneous event with voluminous intraplate magmatism in Laurentia suggests that the two events were linked, although paleomagnetic data require the two cratons to have been separated by a significant distance at this time. Younger intraplate magmatism in the Congo and Kalahari cratons and the surrounding PanAfrican belts was associated with development of major Neoproterozoic basins that were later destroyed during Pan-African orogenesis. Bimodal assemblages, A-type felsic rocks, alkaline complexes, and/or carbonatites were emplaced at 1000-910 Ma in the West Congo belt, at 850-710 Ma in parts of the eastern Congo craton, at 880-740 Ma in the Damara-LufilianZambezi orogen between the Congo and Kalahari cratons, and at 830-740 Ma in the Gariep belt and adjacent parts of the SW Kalahari craton. This magmatism overlaps in time with similar igneous activity that occurred on a number of other cratons worldwide and is inferred to record initial stages of Rodinia breakup. These data provide circumstantial evidence that both Congo and Kalahari formed parts of Rodinia and were involved in its disaggregation. Original relations between the Congo and Kalahari cratons across the younger DamaraLufilian-Zambezi orogen are controversial. Neoproterozoic eclogites with MORB affinities in the Zambezi belt suggest that an ocean basin was destroyed during collision between the two cratons. However, the original extent of this ocean remains unclear, because of the lack of well-defined, long-lived Neoproterozoic volcanicarc assemblages in most parts of the orogen. Resolution of this problem awaits acquisition of well-dated paleomagnetic poles from rock units of appropriate age from both the Congo and Kalahari cratons.

Supercontinents and Earth Evolution Symposium 2005

FROM PROTO-KALAHARI TO GREATER KALAHARI: SIGNIFICANCE FOR THE RECONSTRUCTION OF RODINIA Joachim Jacobs\ Sergei A Pisarevsky^, RJ Thomas^, T Becker"* ^Universitat Bremen, FB Geowissenschaften, PF 330440, 28334 Bremen, Germany lpojacobs@uni-bremen.de) Tectonics Special Research Centre, Univ. of Western Australia, Crawley, WA 6009, Australia ^British Geological Survey, Keyworth, UK "^Geological Survey of Namibia, Windhoek, Namibia

Kalahari might form an important part of Rodinia. It represents a major plate that was involved in repeated phases of plate assembly and dispersal from Paleoproterozoic to Late Neoproterozoic times. We review the components of Kalahari, its geometry and its possible position within a Rodinia supercontinent. Clifford (1970) defined Kalahari as the southern African craton that remained stable since the end of the Kibaran orogeny in which he included the Late Mesoproterozoic rocks of the Namaqua-Natal Belt. Thus, Kalahari consists of a Paleoproterozoic nucleus, plus a rim of Mesoproterozoic rocks of various plate-tectonic settings. Although Clifford did not include parts into Kalahari that rifted away during Gondwana breakup, for considerations within Rodinia it makes sense to include any parts that formed part of this Greater Kalahari plate by the end of the Mesoproterozoic. Today, Greater Kalahari is strongly fragmented with differently sized fragments exposed in southern Africa, South America (Falkland Islands), different parts of Antarctica (Dronning Maud Land, Haag Nunataks) and possible further, yet unidentified fragments elsewhere. The major part of Kalahari is exposed in southern Africa and it is often only this part that is shown in most Rodinia reconstructions. The margin of the Kalahari plate is completely surrounded and to various degrees overprinted by late Neoproterozoic to Early

Paleozoic "Pan-African" mobile belts. Therefore, obtaining a good estimate of the size and geometry of pre-Pan-African Kalahari is a difficult task. However, it appears that Kalahari was significantly larger than shown in most Rodinia reconstructions, with significance for Rodinia reconstructions. The paleomagnetic history from ProtoKalahari to Greater Kalahari is moderately well constrained. It appears that Kalahari oscillated between 60° and the equator during the entire time without large rotational components. Besides paleomagnetic constraints, there are a few boundary conditions that any reconstruction of Kalahari in Rodinia should fulfil. The highgrade rocks of the Namaqua-Natal-Maud Belt are interpreted as representing an indenter and the timing of the high-grade metamorphism along the length of this extensive orogen is taken as the time of continent-continent collision. Therefore, the southern margin of Kalahari should face another continent. The NW margin of Kalahari is probably a long-lasting continental arc, the detailed paleogeographic evolution of which needs incorporation into any model. The NE margin underwent intense overprinting during orogenesis along the East African-Antarctic Orogen. Increasing new work in Northern Mozambique sheds light on the complex history of this area.

Supercontinents and Earth Evolution Symposium 2005

RELATIONS OF THE KALAHARI CRATON TO RODINIA: PALEOMAGNETIC EVIDENCE Wulf A Gose\ Richard E Hanson^ ^Department Geological Sciences, University of Texas, Austin, TX 78712, USA ifwulf@mail.utexas.edu) ^Department of Geology, Texas Christian University, Fort Worth, TX 76129, USA

Rigorous reconstructions of the Rodinia supercontinent are hampered by the scarcity of paleomagnetic data for which the ages of magnetization are well determined. Here we focus on poles in the 1100-1000 Ma age range, within the time frame of final Rodinia assembly. For the Kalahari craton, there are paleomagnetic results from 75 sites in the Umkondo LIP in Botswana, South Africa, and Zimbabwe, with

shallow southerly directions and 9 sites of opposite polarity. To this, one can add the results from 33 sites in the Grunehogna Province of Antarctica after restoring East Antarctica to its position next to southern Africa. Grouping the sites geographically yields 10 poles with a site mean pole of 63.7"N, 38.6°E, A95=3.7^ Umkondo dolerites have yielded U-Pb singlecrystal baddeleyite or zircon crystallization ages of 1112 ±0.5 to 1108 ±0.9 Ma. From the Namaqua-Natal belt along the southern margin of the Kalahari craton there are several pole positions from intrusive rocks. Two sampling areas are in the Namaqua segment of the belt and two areas are in the eastern part in Natal. Their site mean pole position lies at 9.4°N, 329.5°E, A95=17.8°. None of these rocks carries a primary magnetization; their magnetizations were set when the Namaqua-Natal belt cooled below the 500°C isotherm at -1000 Ma. The Kalkpunt Formation at the top of the Koras Group, which rests unconformably on Namaqua gneisses, has yielded a pole that falls between the Namaqua and Umkondo poles and has been widely cited in recent compilations. However, the cited age for this pole (1065 Ma) is based on discordant, multigrain U-Pb zircon analyses from felsic rocks that predate the Kalkpunt Formation; the exact age of the pole is unknown. No other paleomagnetic data from rock units of appropriate age in the Kalahari

craton have their ages of magnetization sufficiently well constrained to be of use for Rodinia reconstructions. The Umkondo pole can be correlated with excellent poles from Laurentia. Because both sets of poles are precisely of the same age as well as of one polarity, the relative orientation of the two cratons is fixed, implying that the Namaqua-Natal belt faced away from Laurentia. The Namaqua-Natal pole has no well-dated counterpart in Laurentia, but a pole for 1000 Ma can be reasonably well estimated. Assuming the simplest plate motion for Kalahari between 1100 Ma and 1000 Ma keeps the Kalahari craton well south of Laurentia and permits Kalahari to occupy a position SW of Laurentia at 1000 Ma (but separated from it by another craton), as suggested in some recent Rodinia models. No paleomagnetic data exist from the Congo craton in Africa for this time period. There are, however, three pole positions constrained by Ar/Ar dates of 1078 ± 18 to 1012 ± 24 Ma for the Sao Francisco craton, which represents a continuation of the Congo craton into Brazil. These poles allow close proximity of the Kalahari and Congo-Sao Francisco cratons at -1000 Ma and suggest that they were well separated from Laurentia. Based on these and other data, some workers have argued that neither Congo nor Kalahari were part of Rodinia. However, the Congo and Sao Francisco blocks are partly separated by a deep embayment occupied by the Pan-African West Congo-Aracuai orogen, which records a vise-like closure between the two blocks. Significant vertical-axis rotation of one or both blocks is likely to have occurred during this event, adding uncertainty in using the Sao Francisco paleomagnetic data to constrain the exact positions of the Congo craton relative to Kalahari or other possible parts of Rodinia.

Supercontinents and Earth Evolution Symposium 2005

TIMING OF COLLISIONS AND THERMAL EVENTS IN THE -'I.O Ga NAMAQUA-NATAL PROVINCE OF SOUTHERN AFRICA David H Cornell Asa Pettersson Geology Dept, Earth Sciences Centre, Goteborg University, Box 460 Goteborg SE40530 Sweden (cornell@gvc.gu.se)

When attempting reconstruction of Rodinia by reassennbling continental fragments according to their ancient Grenville-age collisions, it is important to know the exact timing of collision and other orogenic events in the different continents. The Namaqua-Natal Province of southern Africa has long suffered from a lack of precise dating, so that palaeomagnetic data and timing of orogenic events have been poorly constrained. The province comprises at least eight distinct terranes, some of which are clearly juvenile arcrelated, with pre-collision ages in the 1300 to 1200 Ma range. Others are much older, between 1900 and 1600 from Nd model ages, and one Archean component has been identified. New ion probe and Sm-Nd data from localities 200 km apart along the western Namaqua Front zone shows that the arc-continent collisions which established the Namaqua Front took place close to 1210 Ma, equivalent to the Elseverian of Canada, and not strictly Grenvillian. In Natal, the collision between mafic arcs and the Kaapvaal Craton has traditionally been thought of as significantly younger, but recent dating shows that it took place between 1209 and 1180 Ma. Considering the regional picture, all the highgrade terranes contain predominant granitoids in the 1200 to 1160 Ma interval, which testifies to crustal thickening followed by thermal relaxation and melting of deep crust. Thus although at least five collision zones should lie within the province, all the different terranes must have been assembled in approximately their present configuration before 1160 Ma. The entire Namaqua-Natal Province exhibits a two-stage orogenic history, the second event being in the range 1100 to 1020 Ma, about 100 Ma after the collision events, although varying somewhat in age from one terrane to the other.

The second event is expressed in all terranes as a thermal peak, accompanied by mainly granitic magmatism and high-T, low-P metamorphism. Deformation at that time was largely concentrated in regional faults and shear zones, related to a north-directed stress field. Although this could conceivably reflect a collision event far to the south, there is no direct evidence for this in the exposed terranes. The major Umkondo mafic magmatic event, documented in the Kaapvaal Craton (and Laurentia) at 1100 Ma, corresponds in time to 1090 to 1080 Ma intrusive and extrusive magmatism and regional-contact low-P metamorphism in the Namaqua-Natal Province bordering the craton. Terranes further from the craton seem to have experienced the same thermal peak somewhat longer or later, with intrusive and metamorphic events recorded from 1065 to 1020 Ma. The second event seems unrelated to collisions, but rather to a regional thermal event which originated in the mantle. This may have taken the form of either a mantle plume, or a mantle delamination event. As this mantle process affected the Kaapvaal Craton as well as the Namaqua-Natal Province, it is difficult to understand how the craton retained its deep lithospheric keel and low geothermal gradient necessary for the preservation of diamonds. The two-stage history of the Namaqua-Natal Province is now sufficiently well-established to enable comparisons with other continental fragments such as Laurentia, with which it shares many features, although lacking anorthosite massifs. Palaeomagnetic fits between the Kalahari Craton (Kaapvaal Craton plus Namaqua-Natal Province) and other parts of Rodinia should be made at or after the time of its assembly at 1200 Ma.

Siipercontinents and Earth Evolution Symposium 2005

RODINIA DESCENDANTS IN SOUTH AMERICA Reinhardt A Fuck\ Benjamin Bley de Brito Neves^, Carlos Schobbenhaus^ ^Universidade de Brasilia, 70910-000, Brasilia, Brazil (rfuck@unb.br) ^Universidade de Sao Paulo, 05508-080, Sao Paulo, Brazil ^Servigo Geologico do Brasil, 70830-030, Brasilia, Brazil

Several reconstructions of Rodinia have been presented since the 1991 Hoffnnan senninal paper. In these proposals, insufficiency of data from South America have led workers to consider only the main cratonic blocks (Amazonia, Sao Francisco, Rio de La Plata) to have been derived from Rodinia fission. We try to take into consideration all possible descendants of Rodinia, including continental blocks from shield areas (pre-Brasiliano and Brasiliano domains) and from the basement of Phanerozoic basins of the South American Platform, as well as basement inliers of the Andean Cordillera. The main areas recording late Mesoproterozoic orogenic events in South America comprise the basement of the Andean Chain and adjacent areas, the southwestern part of the Amazon Craton, and several deeply reworked occurrences in the Brasiliano Borborema, Mantiqueira, and Tocantins provinces. Within the northern Andes mountain belt granulite, orthogneiss and related rocks have been described in the Guajira, Sierra Nevada de Santa Marta, Santander, and Garzon blocks. High-grade metamorphism was dated at ca. 1.21.3 Ga and correlation with the Oaxacan Complex in S Mexico was suggested. In the central Andes, Proterozoic basement rocks appear in the Arequipa-Antofalla, Pampia, and Cuyania crustal blocks. Although rocks of Paleoproterozoic age are present, as in the Arequipa massif, available data suggest that these blocks were deformed and metamorphosed during the Mesoproterozoic. Cuyania is correlated with Grenville basement of eastern Laurentia. Several Mesoproterozoic events were recognized in the SW Amazon Craton. Magmatic arc rocks are preserved in the Cachoeirinha rocks (1.52-1.55 Ga) and in the Santa Helena arc (1.42-1.45 Ga). Ca. 1.35 high-grade metamorphism is recorded in the Rondonian and San Ignacio domains, followed by ca. 1.1 orogenic activity in the Nova Brasilandia and Sunsas-Aguapei belts. Metavolcanic and metasedimentary rocks, orthogneiss and associated granites (1.0-0.95

Ga, Cariris Velhos Orogeny) form the Pajeu terrane within central Borborema Province. Similar ages were reported from the Riacho do Pontal Belt farther south. The Punta del Este terrane, within the Dom Feliciano Belt, Mantiqueira Province, Uruguay, bears gneiss and migmatite formed ca. 1.0 Ga, thought to be a small piece of the Namaqua province of the Kalahari Craton. The Falklands/Malvinas islands are thought to be a piece of the Natal province, attached to South America after Gondwana breakup. Mesoproterozoic rift-related rocks are reported from the southern Ribeira Belt, Mantiqueira Province. Oceanic tholeiite and related metasedimentary rocks ca. 1.4 Ga compose the Serra do Itaberaba Group, Sao Paulo. Mafic volcanics (ca. 1,45 Ga) and metasedimentary units appear also in the Perau, Aguas Claras, and Votuverava groups, Parana. Bimodal volcanics, metasedimentary rocks, layered complexes, and granites (1.25-1.3 Ga) are part of the Goias Massif, Brasilia Belt, Tocantins Province. All these units were reworked during the Brasiliano orogeny. Two additional groups of potential descendants of Rodinia have not been included in previous reconstructions. One includes cratonic blocks hidden below the major Phanerozoic sedimentary basins of Parnaiba (Parnaiba block, NE Brazil) and Parana (Paranapanema block, SE Brazil). The second group comprises a number of minor blocks that have worked out as microplates and microcontinents during Brasiliano collage. Such basement inliers of the Neoproterozoic orogenic systems were deeply reworked. They include Central Ceara, Rio Grande do Norte, Moxoto, and Pernambuco-Alagoas (Borborema Province), Goias, Rio Apa (Tocantins Province), Juiz de Fora, Cabo Frio, Curitiba, Luis Alves (Mantiqueira Province), and several others that were descended from the same ancestor of the major cratons. Brasiliano overprinting and granite plutonism have hampered recognition of their close geological relationships with the major cratons.

Supercontinents and Earth Evolution Symposium 2005

MESOPROTEROZOIC SUPRA-SUBDUCTION MAGMATISM IN THE SOUTHERN IRUMIDE BELT, CENTRAL SOUTHERN AFRICA: IMPLICATIONS FOR THE CONGO CRATON IN RODINIA RECONSTRUCTIONS Simon P Johnson\ Bert De Waele^, K Tani\ Francis Tembo^ ^Institute for Research on Earth Evolution, JAMSTEC, 2-15 Natsushinna-cho, Yokosuka, 237-0061, Japan (sjohnson@jamstec.go.jp) ^Tectonics Special Research Centre, University of Western Australia, Crawley, WA 6009, Australia ^School of Mines, University of Zambia, PO Box 32379, Lusaka, Zambia

The Irumide Belt is a terrain of deformed basement, supracrustal rocks and late Mesoproterozoic magmatic rocks that occurs along the southern margin of the Congo Craton of central Africa. Because the belt was the locus of magmatism and convergent tectonics during the time frame normally assigned to the assembly of the Rodinia supercontinent, it forms a critical feature, the understanding of which may help elucidate whether or not the Congo Craton participated in the amalgamation of the supercontinent. The southeastward extension of the Irumide Belt into eastern Zambia, northeastern Zimbabwe and northwestern Mozambique, a region we term the "Southern Irumide Belt" (SIB), remains tenuous mainly because of a large exposure gap across the Mesozoic Luangwa graben and a lack of modern age data. In recent years, a few reliable data have become available for some key areas in the SIB. The Chewore Inliers of NE Zimbabwe, and the Chongwe and Chakwenga regions of Zambia have recently been the focus of detailed and extensive geological investigations. These regions are characterised by a similar tectonostratigraphic sequences of felsic to mafic gneisses, abundant mafic amphibolites and metasomatic whiteschists. The Chewore region is comprised of four tectonically stacked terranes; the Ophiolite Terrane (OT) comprising the 1.39 Ga Chewore Ophiolite and the 1.08 Ga Kaourera Arc; the South Zambezi Terrane (SZT), comprised of Kfsmegacrystic felsic to intermediate orthogneiss; the Granulite (GT), Quartzite (QT) and North Zambezi (NZT) Terranes are predominantly of meta-sedimentary origin but are intruded by a variety of felsic to intermediate orthogneisses, that in the GT and NZT have been dated at 1.07 and 1.08 Ga, respectively. Geochemistry of these orthogneisses, including those from the SZT, reveal that they formed in a suprasubduction zone setting. In the GT, QT, and NZT

high-temperature (>800°C), low-pressure (<4.4 kbar) granulite facies metamorphism of host sediments occurred contemporaneously with intrusion of the orthogneisses. The Chongwe and Chakwenga areas comprise Kfs-megacrystic felsic orthogneisses, structurally overlain by folded and highly deformed mafic to felsic gneisses and whiteschists. Geochemical and whole rock isotopic (Sm-Nd) analyses of the various lithologies indicate formation in a suprasubduction setting comparable to that of the Kaourera Arc. Arc magmatism occurred in a similar time frame to that in the Chewore Inliers, between 1.09 and 1.04 Ga. Imbrication tectonics, similar to that described for the Chewore Inliers, is apparent in the Chongwe and Chakwenga areas. The span in magmatic activity, covering some 50 million years, suggests that both regions represent a succession of juvenile arcs that were imbricated together during either Meso- or Neoproterozoic tectonism. It is interesting to note that the age of magmatism in the SIB overlaps with the ages of the Mpande and Munali Hills granitic gneisses (1.09-1.04 Ga) that form the basement to the Neoproterozoic Katangan sediments of the Lufilian Belt, south of the Mwembeshi Shear Zone. Geochemistry of the Mpande and Munali gneisses also display arc-like signatures, and it is possible that these gneisses also formed due to subduction processes and are part of the SIB, rather than part of the Congo Craton basement as is traditionally considered. Our data suggest that at the time of Rodinia assembly, the SIB was dominated by 1.09-1.04 Ga supra-subduction oceanic-arc magmatism and thus the southern margin of the Congo Craton faced an open ocean. Mesoproterozoic metamorphism and deformation at 1.02 Ga in the Irumide Belt itself, and at 1.05 Ga in the eastern parts of the SIB (i.e. around Chipata in eastern Zambia), may indicate oceanic arc formation

Supercontinents and Earth Evolution Symposium 2005

outboard of the continental nnargin, followed by accretionary or collisional tectonics at ca. 1.051.02 Ga during amalgannation of Rodinia. However, it cannot be ruled out that the current

configuration of stacked terranes is the result of Neoproterozoic tectonics, during the formation of Gondwana.

Supercontinents and Earth Evolution Symposium 2005

EPISODIC CRUSTAL RECYCLING IN THE METACRATONIC IRUMIDE BELT: GEOCHEMICAL AND ISOTOPIC EVIDENCE Bert De Waele\ Jean-Paul Liegeois^ ^Tectonics Special Research Centre, University of Western Australia, Crawley, WA 6009, Australia (bdewaele@tsrc.uwa.edu.au) Departement de Geologie, Musee Royal de I'Afrique Centrale, B-3080 Tervuren, Belgiunn

The Irunnide Belt (IB) occurs along the southern nnargin of the Bangweulu Block (BB), and comprises the following units: (a) a complex of Palaeoproterozoic crystalline basement dated between 2.05 and 1.93 Ga (G1a), with a minor Archaean component dated at 2.73 Ga (GO); (b) the Muva Supergroup psammo-pelites, unconformably /structurally overlying the basement, and including rare thin mafic and felsic volcanic units dated between 1.88 and 1.85 Ga (G1b); (c) a volumetrically minor suite of anorogenic plutons dated between 1.66 and 1.55 Ga (G2); and (d) voluminous K-feldspar porphyritic granitoids intruded between 1.05 and 0.95 Ga (G4), coincident with the climax of Irumide tectonism dated at 1.02 Ga (MP-HT metamorphism). The most voluminous groups ( G l a and G4) despite their significant age difference, have similar whole-rock geochemistry (high-K calcalkaline) and Nd TDM model ages (3.1-3.3 Ga), indicating the same Archaean source at depth. G i b dacites-rhyolites in the IB have very similar geochemistry to contemporaneous granites and dacites present in the BB. The G i b Group, both on the BB and in the IB, display a range of Nd TDM between 2.9 and 2.3 Ga, indicating mixing between Archaean crust and a mantle component. The anorogenic G2 plutons are more enriched in most incompatible elements than the former groups and correspond geochemically to A-type granitoids. Their Nd TDM model ages range from 3.2 to 2.8 Ga indicating a mainly crustal source; their particular compositions can be ascribed to a mantle component but more probably to a lower degree of partial melting of Archaean crust.

Nd TDM model ages clearly indicate that Archaean crust is present beneath the IB. The major Usagaran { - 2 Ga) and Irumide (^1 Ga) events do not appear to have added new material to the crust in the region. The volumetrically minor G i b (Ubendian - -1.85 Ga) and G2 magmatic pulses added little new material, being also mainly crustal in origin. Geochemical characteristics of all groups are dominated by the nature of the Archaean crust, rendering deductions on geotectonic environments tenuous. We consider the IB as the southern boundary of the BB, reworked during Usagaran (2 Ga) and Irumide (1 Ga) orogenesis. At 1 Ga, this reworking generated amphibolite to granulite fades rocks but also preserved 1.88-1.85 Ga supracrustal and volcanic rocks. This suggests important vertical tectonism. No active margin rocks have been observed, whereas just to the south, on the other side of the Mwembeshi mega-shear zone, ca. 1 Ga juvenile terranes are known. Irumide reactivation could have resulted in continental subduction of the southern margin of the BB, inducing mainly fracturing of the cratonic boundary and intrusion of crustal granitoids in response to asthenospheric upwelling along subvertical shear zones, and leaving intact the overall structure of the cratonic boundary. A similar scenario can be suggested for the Usagaran ( G l a ) phase along the same margin, whereas G i b and G2 magmatism would have occurred in response to more distant collisions, exploiting the rheologically weakened structure of the proto-lrumide Belt. A far-field setting would favour the limited mantle contribution seen in these magmas.

Rb-Sr isotopic data for all the magmatic units in the IB record significant disturbance, ascribed to mobility during Irumide and possibly PanAfrican tectonism. This effect renders previously reported whole rock Rb-Sr dates on deformed lithologies within the IB suspect.

The IB shows all the hallmarks of a cratonic continental passive margin, subjected through time to several convergent periods. These events were able to transform this margin of the BB into a metacratonic margin, susceptible to reactivation.

Siipercontinents and Earth Evolution Symposium 2005

FORMATION OF THE SOUTH CHINA BLOCK: EVIDENCE FROM SIBAOAN OROGENIC MAGMATISM Xian-hua Li\ Wu-Xian Li\ Zheng-Xiang Li^, Jian Wang^ ^Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, China (lixh@gig.ac.cn) ^Tectonic Special Research Centre, University of Western Australia, Australia ^Chengdu Institute of Geology and Mineral Resources, China Geological Survey, China

The South China Block (SCB) consists of the Yangtze craton to the northwest and the Cathaysia block to the southeast. There are increasing evidences suggesting that the SCB was formed through the amalgamation of the Yangtze and Cathaysia blocks during the Proterozoic Sibaoan orogenesis (Li et al., 2003), but the timing of this amalgamation is still a matter of debate. Mesoproterozoic meta-sedimentary rocks are widespread within the Sibao Orogen. They are tightly folded and are typically overlain, across an angular unconformity, by Neoproterozoic (<820 Ma) volcaniclastic rift sequences. However, their ages and tectonic significance are poorly understood. On the other hand, the orogenic magmatic rocks occur sporadically at the southeastern section of the orogen, providing important constraints on the timing and evolution of the Sibao Orogeny. The Shuangxiwu Group volcano-sedimentary sequence occurs in northern Zhejiang, consisting mainly of intermediate to felsic volcanic rocks and volcaniclastics. New SHRIMP U-Pb zircon analyses yield ages of 924 ± 14 and 887 ± 12 Ma for volcanic rock samples from the Beiwu and Zhangcun Formations, respectively, providing lower and upper age limits for the Shuangxiwu volcanic rocks. These volcanic rocks constitute a complete calc-alkaline series, and display characteristic arc-like trace element patterns. An ophiolite complex occurs in the NE Jiangxi, extending for ca. 100 km. A Sm-Nd mineral isochron age of 1034 ± 24 Ma was reported for the Zhuangshuduan ophiolitic block (Chen et al., 1991). Volcanic rocks within the ophiolite are mostly tholeiitic basalts and basaltic andesites. They show a general geochemical affinity to back-arc basin basalts and N-MORB. These rocks have uniform, high £Nd(T) values of

+5 to +7 (Li et al., 1997), reflecting derivation from a homogeneous and long-term depleted mantle source without crustal contamination. In general, the NE Jiangxi Ophiolite is typical of supra-subduction zone ophiolites. A number of granitic lenses are enclosed within the Xiwan serpentinised ultramafic block, one of the largest ophiolitic blocks occurring within the NE Jiangxi Ophiolite. Two types of granites i.e., adakitic granite and obduction-type granite, have been identified through petrological and geochemical investigations. An albite granite sample was dated at 968 ± 23 Ma by SHRIMP U-Pb zircon analyses (Li et al., 1994), which was previously interpreted as the timing of oceanic crustal formation. However, recent geochemical studies indicate that these albite granites are typical of adakites. They are thus re-interpreted as products generated by low degrees of partial melting of subducted oceanic crust at pressures high enough to stabilize garnet and amphibole (Li and Li, 2003). Consequently, the U-Pb zircon age of 968 ± 23 Ma was re-interpreted as the timing of subduction of oceanic crust. A few biotite granites have also been identified within the Xiwan ultramafic block. Preliminary SHRIMP U-Pb zircon analyses yield a crystallization age of -880 and a metamorphic age of -230 Ma. These biotite granites are peraluminous, and have negative £Nd(T) values indicative of crustal origin. They were generated by melting of sedimentary rocks beneath ophiolite thrust sheets during obduction of the ophiolites onto continental crust. The U-Pb age of -880 Ma for the "obduction-related" granites provides a constraint on the timing of final amalgamation between the Yangtze and Cathaysia blocks.

Supercontinents and Earth Evolution Symposium 2005

LATE MESOPROTEROZOIC - EARLY NEOPROTEROZOIC BASIN RECORD OF THE SIBAO OROGENESIS IN WESTERN SOUTH CHINA BLOCK AND ITS RELATIONSHIP TO THE ASSEMBLY OF RODINIA Matthew Greentree\ Zheng-Xiang Li\ Xian-hua Lp ^Tectonics Special Research Centre, University of Western Australia, Crawley, WA 6009, Australia (mgreentree@tsrc.uwa.edu.au) Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, China

Along the western margin of the South China block (SCB) thick Mesoproterozoic sedimentary successions are exposed. Locally these sedimentary rocks are termed the Kunyang Group and were thought to be a continuous early Mesoproterozoic succession that overlies Paleoproterozoic basement. However, SHRIMP dating during this study suggests that the Kunyang Group is the product of two late Mesoproterozoic depositional episodes separated by a 140 Ma hiatus. The lower Kunyang Group consists of alkali basalt, tuff and siltstone. SHRIMP dating of zircons from a tuff bed indicates a Pb age of 1147 ± 15 Ma for these volcanic rocks. Detrital zircons from a shale unit indicate a predominantly late Archean (2700-2600 Ma) to early Paleoproterozoic (2390-2000 Ma) provenance. Several younger Mesoproterozoic grains are present (1400 Ma and 1190 Ma). The alkali basalts have pronounced enrichments in Th, Ta, Nb, LREE, Sr, P, Zr, Hof, Ti, smooth LREE enriched patterns and humped trace element spider diagrams. Such a geochemical signature is similar to modern alkali basalts found in the western branch of the East African Rift. The alkaline nature of these basalts, the lack of MOR basalt and the relatively thin sedimentary sequence suggest limited crustal extension occurred during the period of rifting. Mesoproterozoic to earliest Neoproterozoic intracontinental magmatism has been found along the southern margins of Laurentia and on the Kalahari Craton. The presence of late Mesoproterozoic intracontinental magmatism in

South China suggests a possible spatial link to Laurentia and Kalahari in Rodinia. The late Mesoproterozoic volcanic and sedimentary rocks are overlain by upper Kunyang Group sedimentary rocks, including siltstone, sandstone and limestone deposited in marine, lagoonal and fluviatile environments. The depositional age of the upper Kunyang Group is constrained directly by a 996 ± 15 Ma SHRIMP zircon age from a tuff bed near the base of the succession (the Heishantou Formation). The distribution of detrital zircon ages from the succession indicates changing provenance with time. Sediments at the base and middle of the succession contain Archean (ca. 3570-2660 Ma) zircons and a large Paleoproterozoic (ca. 1950-1800 Ma) component. At the top of the Kunyang Group (the Liubatang Formation) the depositional environment changes to predominately fluvial. Significant changes occurred in the provenance, with a largely Paleoproterozoic (2140-1730 Ma) and Mesoproterozoic (ca. 1500-1200 Ma) populations. Several younger grains at ca. 1093-960 Ma provide a maximum age of deposition. These new data indicate two periods of basin formation along the southwestern margin of the South China Block, separated by a hiatus of -140 million years. During the late Mesoproterozoic, intracontinental magmatism, possibly within a failed rift, was overlain by more extensive early Neoproterozoic foreland basin sediments. Changes in the provenance of the sediments suggest that during the earliest Neoproterozoic more sediment was derived from Cathaysia.

Siipercontinents and Earth Evolution Symposium 2005

Ar-Ar DATING OF 1 Ga SIBAO OROGENIC EVENTS IN SOUTH CHINA: THE ^MISSING LINK' IN THE ASSEMBLY OF RODINIA? Jo-Anne Wartho\ Zheng-Xiang Li^, Sandra A Occhipinti^ ^John de Laeter Centre of Mass Spectrometry, Department of Applied Geology, Curtin University, GPO Box U1987, Perth, W A 6845, Australia (J-A.Wartho@curtin.edu.au)

^Tectonics Special Research Centre, University of Western Australia, Crawley, WA 6009, Australia Worldwide Grenvillian orogenic events at ca. 13001000 M a are believed to be responsible for assembly of the Neoproterozoic supercontinent Rodinia. However, there is still much uncertainty over the make-up of the supercontinent fragments and the fit of these puzzle pieces.

classic S W E A T fit, because the Sibao orogenic belt forms a suture between Australia, east Antarctica, and Laurentia. The Mesoproterozoic Tianii muscovite schists were metamorphosed during the Sibao Orogeny, and are unconformably overlain by the Neoproterozoic Wengjialing Formation, which contains abundant fragments of the Tianii schists. The Wengjialing Formation contains unmetamorphosed volcaniclastic conglomerates and sandstones, and is interpreted as a succession deposited during the early Neoproterozoic rifting following the Sibao Orogeny. The history of the Tianii schists makes them excellent targets for A r - A r laser dating.

In recent years, it has been suggested that the South China Cathaysia Block was part of southwestern Laurentia, and the Sibao belt was the suture zone between Laurentia-Cathaysia and the Yangtze craton, which was in turn connected to eastern Australia by a similar-aged suture zone. This theory is consistent with palaeomagnetic data obtained from South China, and provides an explanation for the geological mismatches in the

Recent ultra-violet laser Ar-Ar dating of muscovite in the Tianii schists has yielded ca. 1 Ga ages, indicating Grenville-aged metamorphism along the Yangtze side of the Sibao orogenic belt, and confirms preliminary data reported by Chinese researchers. This event dates the amalgamation of the Cathaysia and Yangtze Blocks, possibly as part of the assembly of the Rodinia supercontinent.

A large deformation belt is found throughout much of South China, termed the Sibao orogenic belt, which separates the Yangtze and Cathaysia cratonic blocks. The Sibao belt has been interpreted as representing at least one orogenic event in South China, however the timing and kinematics of deformation in the belt are poorly constrained.

Supercontinents and Earth Evolution Symposium 2005

RELICTS OF A CA. 1100 - 1000 Ma MOBILE BELT WITHIN NEOPROTEROZOIC PASSIVE MARGINS ATTRIBUTED TO THE BREAK-UP OF RODINIA: EVIDENCE FROM THE RUKER PROVINCE, EAST ANTARCTIC SHIELD Glen Phmips\ SD Boger\ CJL Wilson\ IH Campbell, CM Allen^ ^School of Earth Sciences, The University of Melbourne, Parkville, Vic. 3010, Australia (g.phillips2@pgrad.unimelb.edu.au) Research School of Earth Sciences, Australian National University, Canberra, ACT 0200, Australia

Laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS) of detrital zircon from the Neoproterozoic Sodruzhestvo Series of the southern Prince Charles Mountains, indicates a significant input of grains from a ca. 1100-1000 Ma source. Although the identity of this source remains ambiguous, palaeocurrent directions and the dominance of this age population could indicate that a significant mobile belt of this age lies beneath the ice cover within the interior the East Antarctic Shield. Similar aged detritus on the opposite side of East Antarctica (Skelton and Beardmore Groups) supports a central and common source for these grains. We speculate that the continuation of the Pinjarra Orogen, which is exposed in the Leeuwin Complex and Northampton inlier, into the East Antarctic Shield, was the source of these detrital zircons. If correct, this suggests that the East Antarctic craton is dissected by an orogenic belt of broadly 'Grenville' age. The depositional age of the Sodruzhestvo Series has been constrained within a window of ca. 950 - 650 Ma. This is a significant period of extension in Rodinia reconstructions, as rifting land masses, namely East Antarctica-Australia

and Laurentia, separated. Similar aged sediment and detrital zircon along both margins of the East Antarctic Shield indicate sediment deposition was not only evident along the margin of the future Palaeopacific, but also within the southern Prince Charles Mountains. The lack of overlying Palaeozoic sediment (cf. Transantarctic Mountains, Lachlan Fold Belt) in the southern Prince Charles Mountains supports the closing of this margin and a switch to orogenesis during the Early Palaeozoic. RESULTS We present new U-Pb LA-ICPMS data from detrital zircon extracted from Neoproterozoic cover sequences of the Ruker Province. Detrital zircon from this sequence indicate prominent age populations between ca. 1100-1000 Ma, and minor constituents of ca. 2.1-2.0 Ga and 2.7-2.5 Ga zircon. CONCLUSION U-Pb provenance data support the existence of a ca. 1100-1000 Ma mobile belt within the East Antarctic Shield. Similar ages observed from the Pinjarra Orogen support the continuation of this mobile belt into the East Antarctic Shield.

Siipercontinents and Earth Evolution Symposium 2005

ASSEMBLY AND DISPERSAL OF RODINIA: EVIDENCE FROM THE NORTHWESTERN INDIAN SHIELD Kamal K Sharma Government College, Sirohi, 307 001, Rajasthan, India (sharmasirohi@yahoo.com)

ASSEMBLY OF RODINIA The northwestern Indian shield holds a unique position to explain supercontinent building processes during Precambrian time. The oldest reported continental element from this region is the Udaisagar granite, dated at 3.3 Ga. This proto-continent was presumably part of the oldest continent "Ur". Several Proterozoic orogenic belts (i.e. Aravalli, Delhi and Sirohi) have been described from the Aravalli mountain range of the northwestern Indian shield. The tectonothermal reconstitution of the southern Delhi fold belt and development of a short-lived basin between 1100 and 850 Ma marks amalgamation of the supercontinent Rodinia in the region. This beginning of Rodinia assembly is evidenced by intrusion of gabbro-diorite at Ranakpur and the Abu Road region at 1000 Ma. This magmatic event is part of the development of a narrow linear basin from Ras to Sirohi, named the Sirohi Basin. The Sirohi Group contains rocks of shalecarbonate fades. There is a total absence of rocks with larger clasts (conglomerate, arkose, and quartz arenite) as well as volcanic rocks. The Sirohi orogenic cycle ended with the intrusion of granites at 850 Ma in the Pali, Abu, and Sirohi regions. The final assembly of Rodinia is marked by closure of orogenic cycles in the Aravalli mountain range. DISPERSAL OF RODINIA The Malani magmatic province (-750 Ma) of northwestern India is an example of a Silicic Large Igneous Province developed at the time of Rodinia break-up. The Malani volcano-plutonic province is spread over approximately 50,000 km^ in western Rajasthan. In addition, Malani magmatic activity has been reported from the Sind Province of Pakistan, and from Kutch, Madagascar, and the Seychelles. Continental fragmentation caused the dispersal of the large Malani province into detached outcrops on

different landmasses. The largest is the Malani province of India. Malani magmatism is generally terrestrial in origin. Aqueous conditions are observed only locally or in the initial stage of volcanism. The volcanism included ignimbrite eruptions, rhyolite flows, hot avalanches, and ash fall eruptions through multiple fissure/rift systems that developed in an intraplate tectonic setting. This was followed by A-type anorogenic granite plutonism and terminal felsic/silicic dykes. Rodinia dispersal resulted in widespread Neoproterozoic anorogenic, commonly bimodal, magmatism on most of the continents under extensional tectonic regimes. However, plume-hotspot models are invoked to explain the Malani activity. The present study indicates that: a) the Malani magmatism resulted from melting of silicic crust signifying shallow magma generation, b) there is an absence of large-scale mafic and ultramafic components along with the felsic magma, c) the Malani activity shows rift settings in the field, and d) Malani magmatism shows an erosional unconformity with basement rocks. Thus, the felsic Malani magmatism, though intraplate, anorogenic, and closely associated with Rodinia break-up, is not consistent with a deep mantle plume-hotspot model. The crust became thick and remained thermally insulated for a long time after the assembly of Rodinia. Prolonged heat build-up in the lithosphere caused thermal expansion and a geoid bulge in the silicic crust resulting in extensional tectonics and intraplate anorogenic magmatism along rift zones. Dispersal of the Rodinia supercontinent at 750 Ma led to development of intraplate anorogenic rift magmatism in East Gondwana continents, comprising India, Seychelles, Madagascar, Antarctica, and Australia, parts of the early Neoproterozoic Rodinia supercontinent.

Supercontinents and Earth Evolution Symposium 2005

A COMPREHENSIVE PALEOMAGNETIC STUDY OF THE VINDHYAN SUPERGROUP, CENTRAL INDIA: PRELIMINARY RESULTS Joseph G Meert\ Manoi K Pandit^, Endale Tamrat\ Linda Sohl^, Dhiraj M Banerjee"*, Vimal R Pradhan\ Shawn Malone\ Laura Gregory^ ^Department of Geological Sciences, University of Florida, Gainesville, Fl 32611, USA ^Department of Geology, University of Rajasthan, Jaipur, 302004, India (mpanditjp1@sancharnet.in) ^Lamont-Doherty Earth Observatory of Columbia University, Palisades, NY 10964, USA "^Department of Geology, University of Delhi, Delhi, 110 007, India

The Meso- to Neoproterozoic Vindhyan Supergroup that crops out in the Son Valley region and Rajasthan State in the Vindhyachal basin is a broad arcuate synclinorium extending east-west across the northern portion of peninsular India. The Vindhyan Supergroup is subdivided by an angular unconformity into the Lower Vindhyan (Semri Group) and Upper Vindhyan (Kaimur, Bhander and Rewa Groups). The reported age of the Lower Vindhyan is --1630 Ma for the Deonar Porcellainite and Rampur shale ash beds of Son Valley. The age of the Upper Vindhyan (Bhander-Rewa) is controversial, but recent discoveries of Ediacaran-type fauna in the Bhander limestone suggest an age of -560 Ma for at least part of the uppermost Vindhyan. Previous paleomagnetic work on the Vindhyachal Basin rocks concentrated on sections from Bhander-Rewa (Upper Vindhyan) and Kaimur Groups, although no detailed demagnetization was reported. We have collected more than 1100 core samples from 105 sites within the Vindhyan Supergroup representing all major sedimentary/stratigraphic units. In addition, samples from the intrusive Majhgawan kimberlite were collected for a more robust "^^Ar/^^Ar age determination on phlogopite. Our preliminary data suggest that this part of peninsular India was located at intermediate (-45 degrees) latitude during lowermost Vindhyan times (<1800 Ma) and drifted to low latitudes during Middle Semri time (-1630 Ma) and then back to higher latitudes (-60-70 degrees) during upper Semri time (-1200-1400 Ma). This

suggests either a period of rapid motion during the interval from the end of Kheinjua deposition into Rohtas-Khorip time or that a hiatus is present between these Groups. The hiatus would be consistent with the recent ^^Sr/^^Sr interpretation. Data from the Majhgawan kimberlite and the Upper Vindhyan sequence largely confirm the early findings; however, we do note the following key information regarding these units. The Bhander and Rewa Groups show a series of stratabound reversals suggesting a primary magnetization. A fold test is also possible for these units, but we currently have not demagnetized sufficient samples to report a conclusive result. Our results on the Majhgawan kimberlite are identical to those previously reported. Thus, the exact age of the Majhgawan kimberlite is important for a number of reasons. First, it will provide some constraints on the age of the Vindhyan sequence that it intrudes. The Majhgawan has definite intrusive relationships into the Kaimur sequence that lies stratigraphically below the Rewa and Bhander Groups. The age of the Bhander-Rewa sequence is generally taken as Neoproterozoic to Cambrian. It is interesting to note that the Majhgawan paleomagnetic directions obtained by other workers and also a suite of samples collected by us from Majhgawan are identical to the directions observed in the Bhander-Rewa sequence (NE-SW shallow). We hope to have the "^^Ar/^^Ar geochronological control on the Majhgawan kimberlite ready for presentation at this meeting.

Super continents and Earth Evolution Symposium 2005

MESOPROTEROZOIC NALLAMALAI SEDIMENTATION IN THE CUDDAPAH BASIN: A LINK BETWEEN INDIA AND EAST ANTARCTICA IN THE RODINIA SUPERCONTINENT Ananya Biswas Bengal Engineering and Science University, Shibpur, India {annie_geol@rediffnnail.com)

The southern extension of the Eastern Ghats Granulite Belt (EGGB) occurs to the east of the Nallamalai Fold Belt (NFB) with an intervening - 5 0 knn wide tract of the Nellore Schist Belt. The latter is thought to be equivalent to the Dharwar schists of the Karnataka craton. Model melting reactions and geochemical studies have already correlated high-Mg metapelites of the Eastern Ghats Granulite Belt (EGGB) with those in the Rayner Complex, East Antarctica. The Rodinia supercontinent was assembled by late Mesoproterozoic (1300-1000 Ma) orogenies that produced two major mobile belts in India: the EGGB and the Central Indian Tectonic Zone in eastern and central India, respectively. A connection between the Dharwar Craton of Peninsular India and Archaean rocks of the Napier Complex of Antarctica was established possibly earlier than 1.0 Ga. The Proterozoic Rayner Complex surrounds the Archean Napier Complex and contains reworked Napier Complex crust. The EGGB and southern granulite terrane of Tamilnadu and Kerala are mainly composed of Palaeoproterozoic to Archean protoliths affected by phases of high-grade metamorphism. The suture linking India with Antarctica is represented by the EGGB-Circum East Antarctic Mobile Belt. Mesoproterozoic suturing of the Circum East Antarctica Mobile Belt and the EGGB provide a geodynamic and tectonostratigraphic framework to study relations between Antarctica, Australia, India, and Sri Lanka within Rodinia. The deformed Mesoproterozoic Nallamalai Group of mainly metasedimentary rocks is truncated in the east by the high-grade Nellore thrust-fold belt of the EGGB. The Nallamalai Basin can therefore be considered as a foreland

basin that formed as a result of west-directed thrusting in the EGGB. Diverse lithotectonic ensembles of the belt include huge volcaniclastic-rich flysch deposits that now include greenschist facies rocks, metapelites, marbles, and calcsilicates. The deep-water Nallamalai flysch basin was also associated with episodic intrusive phases of kimberlites and lamproites (most about 1 . 2 - 1 . 1 Ga) and also generation of felsic plutons of the Ishwarakuppam Complex. Nepheline-bearing gneisses may be remnants of fossil plate boundaries, reflecting destruction of an ocean basin during plate collision. The Proterozoic cratonic basins situated along the eastern margin of the eastern Dharwar and Bastar cratons in India occur immediately west of the high-grade metamorphic belt of the Eastern Ghats. The spatial association has led to the plausible explanation of the NFB as an evolved back-arc basin or one in a retro-arc setting behind the collisional belt, i.e. the Eastern Ghats. In this scenario, deformation in the arcuate NFB is related to E-W directed compression during the (latest Mesoproterozoicearly Neoproterozoic) Eastern Ghats orogeny. Similarity in lithologic associations (charnockites, etc.), structural characteristics, and ultra-high temperature (UHT) metamorphism of rocks of the EGGB and those from the Eastern Antarctica has long been pointed out. More recently, there have been contributions bearing on UHT metamorphism, geochronology, and the structure of the EGGB. These form the basis of a more refined East Gondwana reconstruction that tries to fit the eastern coast of India against eastern Antarctica and corroborates the inclusion of EGGB contiguous with the Rayner Complex in East Gondwana.

Siipercontinents and Earth Evolution Symposium 2005

FRONTAL TO LATERAL TECTONISM AT A NEWLY RECOGNIZED SYNTAXIS IN THE EASTERNMOST GRENVILLE PROVINCE, CANADA Charles F Gower Department of Natural Resources, Government of Newfoundland and Labrador, PO Box 8700, St. John's, Newfoundland and Labrador, A1B 4J6, Canada (cgower@gov.nl.ca)

The Grenville Province in easternmost Canada does not continue eastward into the Atlantic Ocean as it has been traditionally depicted, but terminates in a lateral ramp having a southeast trend. The Namche Barwa syntaxis in the eastern Himalayas might be a modern-day analogue. This radical re-assessment of the tectonic configuration at the eastern end of the Grenville Province is the outcome of 25 years of regional geological mapping and geochronological investigations that now permit discrimination of Grenvillian (1080-985 Ma) from earlier Labradorian (1710-1600 Ma) and Pinwarian (1520-1460 Ma) tectonism in the same region. In contrast to the remainder of the Grenville Province, which has an overall northeast structural trend and is characterized by northwest-verging thrusts, the eastern end is a zone of southeast-trending, right-lateral strikeslip faults. In its simplest terms, this zone is interpreted as the northeastern lateral ramp and dextral transposition zone to Grenvillian allochthonous terranes that were transported on frontal ramps located farther northwest. In the interior allochthonous terranes, Grenvillian tectonism, deformation and granitoid pluton emplacement have had a major impact, whereas in the exterior parauthochthonous region, Grenvillian effects are moderate or minor. Discrimination between Grenvillian and preGrenvillian orogenic effects has been facilitated by high-precision U-Pb geochronological data, coupled with re-evaluation of older K-Ar and ArAr geochronological information. Northeast of the lateral ramp, Labradorian titanite ages are preserved, thus demonstrating that the highgrade metamorphic assemblages in that region cannot be post-Labradorian. In contrast, within the southeast-trending ramp zone, zircon, monazite and titanite U-Pb data all point to orogenic activity culminating between 1080 and 1040 Ma. Perhaps surprisingly, it is the earlierobtained K-Ar and Ar-Ar data, when depicted as thermochrons, that most clearly demonstrate the Grenvillian syntaxis configuration. Thermochron

maps published in the original articles offer strong hints for the syntaxis, but without the present wealth of geological mapping and extensive geochronological database, the original authors were poorly positioned to make the interpretations now possible. The distribution pattern of late- to postGrenvillian granitoid plutons provides another persuasive line of evidence for the syntaxis. Whereas these are prevalent throughout much of the interior allochthonous region, the northern and eastern limits to their extent are sharp and mimic the dog-leg change in thermochron pattern from northwest- to southeast-trending. Rather than a simple frontal- versus lateralramp configuration, it is possible that recent lower-crustal channel-flow models may have applicability in the region. Such models envisage a melt-weakened zone in the lower to middle crust, coupled with ductile extrusion, driven by surface denudation, of high-grade metamorphic rocks between coeval normal- and thrust-sense shear zones. Such models have been invoked to explain very high-pressure metamorphic assemblages found within frontal ramp systems in the central part of the Grenville Province. In the easternmost Grenville Province, however, the structural configuration dictates that interpretation should be directed at what happens at the side of the orogen, rather than its leading edge. It is suggested here that the flank is a roughly 50-km-wide zone that, during Grenvillian orogenesis, was characterized by higher heat flow, low viscosity, subhorizontal transport, and structural complexity at the frontflank syntaxis; the flank only need be characterized by high-pressure Grenvillian assemblages where it merges with the front. This model has dramatic implications regarding inter-continental reconstructions, as these currently assume continuation of the Grenville Province eastward and linkage with the Sveconorwegian Province in Scandinavia. Such models w i l l require re-evaluation.

Supercontinents and Earth Evolution Symposium 2005

SYNTHESIS OF THE ASSEMBLY AND BREAK-UP HISTORY OF RODINIA: RESULTS FROM IGCP 440 Zheng-Xiang Li\ Svetlana V Bogdanova^, Sergei A Pisarevsky\ Bert De Waele\ and all members of the Rodinia Map Committee ^Tectonics Special Research Centre, University of Western Australia, Crawley, WA 6009, Australia (zli@tsrc.uwa.edu.au) Department of. Geology, Lund University, Solvegatan 12, 223 62 Lund, Sweden

This paper gives a brief overview of the construction of the Geodynamic Map of Rodinia, a final product of UNESCO/IGCP 440: Rodinia Assembly and Breakup (1999-2004). Contributors to this work include past leaders C.McA. Powell, R. Unrug, A.B. Kampunzu, and map compilers A. Collins, A. Davidson, R.E. Ernst, D.A.D. Evans, I.C.W. Fitzsimons, R.A. Fuck, D.P. Gladkochub, J. Jacobs, K.E. Karlstrom, S. Lu, J-P Milesi, J.S. Myers, L.M. Natapov, M.K. Pandit, V.L. Pease, K. Thrane, and V. Vernikovsky. RATIONALE OF THE MAP Although a huge amount of data have accumulated since the idea of a late Precambrian supercontinent was proposed, there is no consensus on the configuration of Rodinia. The aim of the Rodinia map, which includes maps for all major Precambrian cratonic blocks compiled using GIS databases, is to illustrate major tectonic elements relevant to the formation, configuration, and breakup of Rodinia, in a transient configuration. This will enable researchers to see the constraints available and make their own reconstructions using the compiled maps and new information as it becomes available. Features in the 1600 to 700 Ma range are given prominence in the map. WHEN DID THE CORE OF RODINIA COME INTO EXISTENCE? At the beginning of the Rodinia debate, the predominant view was that Laurentia, Australia and East Antarctica, forming the core of Rodinia, were together since 1900-1600 Ma. However, this view was challenged by the truncations of crustal provinces between Australia-Antarctica and Laurentia, the existence of late Mesoproterozoic orogenic events in Queensland, the King Island and NW margin of Laurentia, and palaeomagnetic data which indicate that none of the currently suggested configurations between these continents could have existed at -1200 Ma. Recent work suggests that assembly of the core of Rodinia may have lasted until -900 Ma.

WHICH RODINIA? Both geological and palaeomagnetic data suggest that central and western Rodinia broke apart at -750 Ma. The sparsity of palaeomagnetic data between 1000 and 750 Ma precludes unique solutions, but in combination with geological constraints, the field of choices is narrowing. In the Rodinia map we adapted a revised SWEAT fit with South China between southwestern Laurentia and eastern Australia. This configuration is consistent with current palaeomagnetic data, strong similarities between Hainan Island of South China and southwestern Laurentia, and similar Neoproterozoic magmatic and rifting histories in these blocks, although we realise that other possibilities exist. We reconstructed the relative positions of India, Tarim, North China, Siberia, Baltica, Congo-Sao Francisco and Amazonia, using palaeomagnetic data and geological constraints. The positions of West Africa and most small continental blocks remain poorly constrained. TIMING AND MECHANISM OF BREAKUP Rodinia breakup started as early as ca. 850 Ma, but large-scale rifting in Australia, South China, India, and Kalahari did not start until ca. 820 Ma. This global rifting event was accompanied by emplacement of LIPs, including radiating dyke swarms and rapid continent-scale doming, indicating the presence of mantle plumes (or a superplume?). Recent palaeomagnetic data suggests that Rodinia could have stretched from the equator to polar regions at 820-800 Ma, and underwent a rapid 90° rotation before 750 Ma which brought the entire supercontinent to a lowlatitude position. It has been proposed that a 830800 Ma superplume beneath the polar end of Rodinia caused an IITPW event, and thus the rapid 90° rotation. Enhanced carbon burial in and around an equatorial supercontinent, and higher albedo, may have led to the Sturtian glaciation. Global rifting and associated magmatism lasted until at least 750-740 Ma when Rodinia started to break apart (rift-drift transition). Breakup along eastern Laurentia did not occur until after 600 Ma, when Gondwanaland started to assemble.

Supercontinents and Earth Evolution Symposium 2005

FRIDAY 30 SEPTEMBER 2005

Supercontinents and Earth Evolution Symposium 2005

LATE NEOPROTEROZOIC PALAEOGEOGRAPHY: ALTERNATIVE MODELS AND PROBLEMS Sergei A Pisarevskv, Michael TD Wingate, Peter A Cawood Tectonics Special Research Centre, School of Earth and Geographical Science, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia (spisarevsky@tsrc.uwa.edu.au)

Understanding of past continental configurations has advanced by extending our knowledge backwards through tinne fronn the Phanerozoic supercontinents of Pangaea and Gondwanaland, to the late Mesoproterozoic to midNeoproterozoic supercontinent of Rodinia. Although the evolution of Gondwanaland is relatively well established, the configuration of the Rodinia, and its assennbly and breakup histories, are still widely debated. Our understanding of Rodinia and earlier times is severely hampered by our poor knowledge of palaeogeography in the late Neoproterozoic, the interval during which Rodinia broke apart and Gondwanaland was formed. This critical gap in our knowledge results from ambiguities and contradictions in late Neoproterozoic palaeogeographic models. Many high-quality palaeopoles are used to construct Phanerozoic APWPs for the majority of continents, and there is general agreement about Phanerozoic tectonic history (e.g. the evolution of Gondwanaland and Pangaea). In contrast, late Neoproterozoic palaeomagnetic data are scarce and controversial and it is impossible at this stage to apply the traditional APWP method owing to time gaps of more than 100 million years between some data points. Such gaps lead to palaeomagnetic problems of longitudinal uncertainty and polarity ambiguity. The latter is of particular importance, because even high-quality results can be used to support remarkably different palaeogeographic models by choosing alternative polarity options. Thus, for late Mesoproterozoic and Neoproterozoic time, there are four alternative Australia-Laurentia reconstructions, at least six proposed positions of Siberia, and several for Baltica, Kalahari, Amazonia, and Congo. The proliferation of alternative models will continue until the problems of late Neoproterozoic palaeomagnetic data are resolved. The latest Neoproterozoic to Early Cambrian interval is marked by at least two major tectonic reconfigurations of the Earth: the final breakup of

the remnants of the Rodinia supercontinent and the assembly of Gondwanaland. This was also one of the greatest orogenic epochs resulting in the Pan-African, Baikalian, Cadomian, and Timanian orogenies. Latest Neoproterozoic palaeomagnetic data from Laurentia are relatively abundant, but also controversial. They form two groups of roughly similar reliability; one that supports a low-latitude position of Laurentia, and the other that supports a high-latitude position. New preliminary results however, confirm the apparent co-existence of both low- and high-latitude positions. The data, together with recent geochronological information, suggest that Laurentia underwent several 'jumps' between near-equatorial and near-polar positions between 615 and 550 Ma. Recent palaeomagnetic data from northern Russia and western Ukraine are in favour of the "conservative" model in which Baltica rifted off Laurentia-Greenland-Amazonia around 600 Ma with the opening of East lapetus and Tornquist Sea along its Scandinavian and TeisseyreTornquist margins. This was followed by separation of Amazonia and Laurentia and opening of West lapetus. At the same time, a complicated process of collision between several continental blocks (Australia-Mawson, India-Rayner, KalahariDronning Maud Land, Congo-Sao Francisco, Amazonia-West Africa, Rio de La Plata, and a number of microcontinents) on the other side of the globe caused closure of oceanic basins (Adamaster, Brasiliano, Mozambique and others) and the assembly of Gondwanaland. New palaeomagnetic and geochronological data suggest that Siberia rifted away from northern Laurentia and drifted independently until closure of the Ural Ocean and reassembly of Laurasia in the Permian. This was accompanied by major accretionary events along (in presentday coordinates) the northern Gondwana (Cadomian orogeny), eastern Baltica (Timanian orogeny) and southern Siberia (Baikalian orogeny) margins.

Supercontinents and Earth Evolution Symposium 2005

THE SAO FRANCISCO BASIN, EASTERN BRAZIL: THE RECORD OF RODINIA(?) BREAKUP AND GONDWANA ASSEMBLY Marcelo A Martins-Neto^'^, Fernando Flecha de Alkmim^ ^Geology Department, Federal University of Ouro Preto, Caixa Postal 173, 35400-000, Ouro Preto/MG, Brazil (mamneto@uai.conn.br) ^NUPETRO - Nucleos of Petroleum Geology, Gorceix Foundation, Caixa Postal 173, 35400-000, Ouro Preto/MG, Brazil

The late Paleoproterozoic to Mesozoic sedimentary cover of the Sao Francisco craton, as well as of the adjacent Brasilia and Araguai orogenic belts, comprises the record of successive first-order basin fill cycles (megasequences), which have been combined in order to define the Sao Francisco basin. Integrated seismic, well, and outcrop studies attest that the record of the late Mesoproterozoic to early Cambrian Rodinia(?)-Gondwana succession in the Sao Francisco basin started with the development of a full passive margin basin setting and ended with a convergent cycle related to the assembly of West Gondwana (backarc and foreland basins). The fragmentation process of a paleocontinent, coeval with Rodinia, composed of an Archean to Paleoproterozoic basement crosscut by a network of late Paleoproterozoic to early Mesoproterozoic failed rifts, started at about 1300 Ma. A initial rift system evolved to a classic passive margin, defining the western margin of the Sao Francisco-Congo paleocontinent. This stage is represented by the clastic-carbonate successions of the Canastra/ Paranoa Megasequence. These deposits crop out in the Brasilia orogenic belt and, according to seismic data, cover large areas of the Sao Francisco craton. The onset of subduction processes and the development of magmatic arcs at ca. 950 Ma mark the end of the passive-margin setting on the western border of the Sao Francisco paleocontinent. The Araxa/lbia Megasequence, consisting of a volcano-sedimentary package deposited in a backarc setting, is representative of this development stage. Simultaneously (at ca. 950 Ma), rifting processes took place on the eastern Sao Francisco paleocontinent, leading to breakup and establishment of a gulf-like passive

margin. The Macaubas Megasequence represents the record of the rift to passivemargin basin. At ca. 790 Ma, the accretionary processes that led to development of the Brasilia fold belt started at the western border of the Sao Francisco paleocontinent, culminating with a continent-continent collision (Amazonia and Sao Francisco cratons) between 630 and 600 Ma. During this interval (790-600 Ma), the cratonic portion of the Sao Francisco paleocontinent behaved as a flexural basin in response to the tectonic load of the in-building Brasilia fold belt, where the Bambui Megasequence was deposited. Seismic data indicates that Bambui foreland deposits have a classic wedge shape with thicknesses of ca. 4000 m to the west at the depocentre and ca. 800 m to the east, close to the flexural border. Subduction and magmatic arc construction was initiated at the eastern Sao Francisco paleocontinent between 650-620 Ma, and the Salinas Megasequence deposited in a backarc setting. This scenario evolved until ca. 600 Ma, when collisional processes closed the ocean and erected the Araguai A/Vest Congo orogen, deforming the eastern domains of the Bambui deposits. It is notable that there existed a phase shift between the evolution of the Brasilia and Araguai fold belts, the former being always one step ahead of the latter. Support for this statement is the age correlation between the ca. 900 Ma old glaciogenic Ibia diamictites, deposited in a backarc basin in the domains of the Brasilia fold belt to the west, and the Macaubas diamictites, deposited in a rift basin in the domains of the Araguaf fold belt to the east. Younger glaciogenic deposits (ca. 750 Ma) occur at the base of the Bambui foreland Megasequence.

Supercontinents and Earth Evolution Symposium 2005

NUTCRACKER TECTONICS DURING THE NEOPROTEROZOIC ASSEMBLY OF WEST GONDWANA: THE DEVELOPMENT OF THE CONFINED ARAgUAI-WEST CONGO OROGEN (BRAZIL/AFRICA) Fernando F Alkmim\ Stephen Marshak^, Antonio Carlos Pedrosa-Soares^, Simone Cruz\ Guilherme G Peres\ Allan Whittington"^ ^Departamento de Geologia, Escola de Minas, Universidade Federal de Ouro Preto, Ouro Preto, Minas Gerais, Brazil (alkmim@degeo.ufop.br) ^Department of Geology, University of Illinois, Urbana, Illinois, USA ^Instituto de Geociencias, CPMTC, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil "^Department of Geological Sciences, University of Missouri, Columbia, Missouri, USA

Paleogeographic reconstructions indicate that the Araguai orogen of eastern Brazil and the West Congo belt of western Africa formed a continuous lithospheric block that lay in the center of West Gondwana prior to opening of the South Atlantic. These reconstructions also reveal that the Araguai-West Congo orogen corresponds to a tongue-shaped enclave surrounded on three sides by stable crust of the Sao Francisco-Congo craton. How such a partially confined orogen formed has been a long-standing kinematic puzzle and a major problem in models postulated for the assembly of West Gondwana. Based on structural analysis conducted in eastern Brazil, we suggest that the Araguai-West Congo orogen resulted from the closure of a Red Sea-like basin (the Macaubas basin), which was partially floored by oceanic crust and separated the Sao Francisco peninsula from the Congo continent. The closure of the Macaubas basin started at around 630 Ma, when

the Sao Francisco peninsula rotated towards the Congo continent, so that the development of the orogen resembled the process of crushing of a nut between two arms of a nutcracker. As rotation progressed, the margins of the Macaubas basin converged and crustal thickening took place. This process led to the propagation of deformation into the foreland of both the Sao Francisco peninsula and Congo continent, generating the Araguai and WestCongo fold-thrust belts in the 585 to 565 Ma interval. The final stage of closure caused lateral escape of the orogen's internal domains, followed by a phase of extensional collapse and intrusion of voluminous post-tectonic granites. If only a narrow ocean closed during formation of the orogen, then subduction might not have been the only motivation for closure. We suggest that the nutcracker was forced to close by the collisions of Amazonia and Kalahari with the Sao Francisco-Congo plate.

Siipercontinents and Earth Evolution Symposium 2005

PAN-AFRICAN CRUSTAL THICKENING AND ECLOGUE FACIES METAMORPHISM IN CAMEROON D e n n y L o o s e \ V S c h e n k \ MH Bouyo^'^, J Penaye^, F Toteu^ ^Institute of Geosciences, University of Kiel, 24118, Kiel, Germany (dl@nriin.uni-kiel.de) ^IRGM, Garoua, PO Box 333, Cameroon

INTRODUCTION The transition zone between the northern border of the Congo craton and the E-W striking PanAfrican Central African Fold Belt (CAFB) is exposed in Cameroon. This offers a unique opportunity to study the orogenic processes and behaviour of the cratonic border during supercontinent formation in the central part of Gondwana. Moreover, the CAFB is suspected to be the locus of the suture zone that separates the Congo craton from its cratonic neighbours. However, ophiolites and eclogites representing remnants of a former ocean floor have so far not been identified. We have reconstructed metamorphic P-T paths and performed in-situ monazite dating to unravel the tectonic history of Palaeoproterozoic basement units and Neoproterozoic cover series in the CAFB. In addition, we analysed major and trace elements of metabasites to find possible remnants of the oceanic suture.

continuously, whereas Ca decreases stepwise from core to rim. Conventional geothermobarometry and differential thermodynamic calculations in metapelites, in combination with decompression textures in metabasites, point to a clockwise P-T path. It starts at high pressures (8-10 kbar) in the kyanite stability field and reaches 750-850X due to near-isobaric heating prior to uplift. RETROGRESSED ECLOGITES Some metabasic rocks of the Paleoproterozoic Nyong complex experienced eclogite facies conditions: omphacitic clinopyroxene (up to 23% jd) contains numerous plagioclase "exsolutions" pointing to a former significantly higher jadeite component during maximum subduction depth. Thermobarometry yields a minimum pressure of 16 kbar at 750-800X. HREE are 10-19 times chondritic, whereas LREE are depleted similar to those of MORB (low (La/Sm)N ratios (<1) at variable Nb/La (0.7-1.4)).

GEOLOGICAL SETTING The Congo craton in Cameroon consists of the Archaean Ntem Complex and the Paleoproterozoic Nyong Complex. It is assumed that neither was reworked during the Pan-African orogeny. In contrast, the Paleoproterozoic basement within the CAFB was overprinted in Pan-African times. These basement rocks locally display granulite facies mineral assemblages thought to have formed during a Paleoproterozoic metamorphic event. The main part of the CAFB consists of granites (mostly Pan-African) and Neoproterozoic cover series metamorphosed from lower greenschist to granulite facies.

GEOCHRONOLOGY Metapelitic monazites from the Paleoproterozoic and Neoproterozoic units exhibit a narrow range in chemical composition. Isochron ages for both units are in the range 602 ± 36 to 631 ± 31 Ma, i.e. all metamorphic monazite grew during the Pan-African orogeny. So far, no evidence for Paleoproterozoic metamorphism has been found, indicating that the two published zircon ages of 2.1 and 2 Ga for the Paleoproterozoic unit may reflect magmatic formation ages and cannot be correlated with granulite facies metamorphism.

GRANULITES Granulite facies metapelites (Grt-Bt-Ky/Sil) and metabasites (Grt-Cpx-Qtz±Opx) have been studied from three areas in Paleoproterozoic basement and one area in the Neoproterozoic unit. Both units are supposed to have experienced different metamorphic histories. Surprisingly, metapelitic garnet from both Paleoproterozoic and Neoproterozoic units display the same type of well preserved prograde zoning, indicating a common prograde growth history: Mg/(Mg+Fe) ratio increases

CONCLUSION The studied areas of granulite facies Paleoproterozoic basement and Neoproterozoic cover series in the CAFB experienced only one metamorphic event at about 602-631 Ma, due to substantial crustal thickening. The wellpreserved stepwise growth zoning of metapelitic garnet indicates a short duration for this metamorphism. Retrogressed eclogites with MORB type chemistry occurring near the NW-edge of the Congo craton may indicate the so far unknown site of the suture zone in the CAFB.

Supercontinents and Earth Evolution Symposium 2005

PAN-AFRICAN REWORKING OF THE NORTHEASTERN CORNER OF THE CONGO CRATON IN UGANDA Volker Schenk\ Peter Appel\ Niels J6ns\ Denny Loose\ Andreas Schumann^ Heike Wegner^ ^Institutfur Geowissenschaften, Universitat Kiel, Germany (vs@min.uni-kiel.de) ^Department of Geology, Makerere University, PO Box 7062, Kampala, Uganda

INTRODUCTION The Basement Complex of Uganda (BC) is regarded as the NE corner of the Archean Congo craton bordering the Pan-African Central African Fold Belt (CAFB) in the north and the East African Orogen (EAO) in the east. Seismic tomography revealed that the BC, like the Tanzania craton, is distinct from the rest of the craton because it is missing a thick (>170 km) lithospheric root. In search of the decratonisation process, we studied the tectonic history of the Archean basement (West Nile area and Labwor Hills of NW and NE Uganda) and of its Neoproterozoic cover series by means of metamorphic P-T paths and metamorphic ages (U-Th-total Pb chemical dating of monazites). NORTHERN BORDER A clockwise P-T evolution during a first metamorphism (M1) in the West Nile area is constrained by prograde sillimanite and later formed kyanite included in garnet, peakmetamorphic Grt-Sil-Bt-bearing pelitic assemblages, and the late-stage formation of cordierite rims around garnet. Peak P-T conditions were estimated at 9-10 kbar and 940°C (GASP barometry and feldspar thermometry). Decompression to about 8 kbar and subsequent near-isobaric cooling is deduced from late-stage Grt-Sil-Crd-Qtz assemblages and from Grt-Cpx±Qtz coronas around Opx in metabasic and charnockitic rocks. A second metamorphism (ca. 8 0 0 X at 7 kbar) is evident from late-stage metapelitic garnet overgrowing Grt-Bt-Sil/Ky-Crd symplectites formed during retrogression after M l . Monazite dating revealed ca. 2.42 Ga for the first metamorphism and ca. 570 Ma for Pan-African reworking. In summary, the Archean rocks experienced granulite metamorphism after crustal thickening at 2.4 Ga. For about 1.8 billion years, they remained in the deep cratonic crust before they were affected by Pan-African reworking and subsequent uplift.

100

EASTERN BORDER Aluminous Opx (up to 11.5 wt% AI2O3 when coexisting with Grt) and assemblages with SplQtz-Grt-Osumilite (pseudomorph) or Opx-Sil-Grt point to extreme metamorphic temperatures (>1000X) at 8-10 kbar in a very large area in eastern Uganda (Labwor Hills). In contrast, latestage Grt-Cpx-Qtz coronas in charnockitic and metabasic rocks where formed at 6-7.6 kbar only after near-isobaric cooling to temperatures of 650-680°C. Uplift of the UHT-granulites is not documented petrologically. Monazite growth has been dated at 633 ± 11 Ma. NEOPROTEROZOIC COVER SERIES Metasediments that occur at the Murchison Falls of the Nile River (NW Uganda) have not experienced granulite facies metamorphism, unlike the basement rocks on which they are resting, and have been interpreted to be postArchean in age. They are Grt-Ky-Ms-Bt-PI-Qtzbearing mica schists metamorphosed at 8 kbar and 650-680°C during a clockwise P-T path. Monazites show a single-stage growth history and give ages of 621 ± 26 and 633 ± 27 Ma for the metamorphic event caused by crustal thickening. CONCLUSION The NE corner of the Congo craton exposed in the Basement Complex of Uganda experienced substantial reworking in Pan-African times. In the NW, crustal thickening is indicated by burial of Neoproterozoic cover series to depths of 25-30 km and by uplift of 2.4 Ga old granulites subsequent to a Pan-African granulite facies overprint. Towards the EAO in the east, the cratonic crust experienced strong Pan-African reheating resulting in UHT-metamorphism (>1000°C) over a large area (>2000 km^) that was followed by near-isobaric cooling. We propose that in the course of Pan-African reworking, the cratonic crust lost its lithospheric root, which led to the unusually high temperatures in the crust.

Supercontinents and Earth Evolution Symposium 2005

COARSE CLASTIC MARKERS OF RODINIA BREAKUP AND GONDWANA ASSEMBLY IN THE LUFILIAN BELT, PANAFRICAN OROGEN OF CENTRAL AFRICA Marek Wendorff University of Botswana, Private Bag 0022, Gaborone, Botswana (wendorff@mopipi.ub.bw)

INTRODUCTION The Lufilian arc, a segment of the Neoproterozoic to Early Palaeozoic Pan-African orogenic network within southern and central Africa, deforms the sedimentary succession of the Katanga Supergroup (880-500 Ma) and contains large bodies of fragmental rocks considered by previous workers as tectonic friction breccias marking two regional decollement zones related to thrusting during Lufilian orogenesis. However, the present author has recently shown that that these rocks are syntectonic conglomeratic complexes deposited in response to a pronounced uplift of source regions composed of Katangan rocks. The major lines of evidence for sedimentary genesis of the Katangan 'megabreccias' are: (1) lack of pervasive shearing or other textures that would point to tectonic fragmentation of brittle rocks; (2) derivation of clasts from identifiable Katangan lithostratigraphic units and vertical changes in clast composition recording progressive unroofing of the source region; (3) vertical successions of lithologies and sedimentary features diagnostic of deposition from sediment gravity flows; (4) lateral facies gradients from proximal in the south to distal in the north, away from the uplifted sources in the south; (5) unconformable lower boundaries of 'breccia' bodies, which represent stratigraphic, not tectonic contacts; (6) soft-sediment injections of conglomerate projecting into open fractures in allochthonous blocks, recording loading of unconsolidated sediments beneath older, rigid Katangan rocks. Sedimentary genesis revealed two previously unrecognized sedimentary basins and tectonic stages of the Lufilian belt evolution. These are reflected in the revised stratigraphy, according to which the Katanga Supergroup is subdivided into four groups: two syn-rift successions - Roan and Guba (or N'guba), and synorogenic foreland suites - Kundelungu, Fungurume and Plateau.

NEWLY RECOGNISED BASINS Major uplift (>765 Ma) in the southern part of the Roan rift basin terminated deposition of the Roan Group infilling the first rift basin of the Lufilian belt. Thick, coarse conglomerates/ olistostromes of the Mufulira Formation derived from the uplifted zone in the south record subsequent opening of the Guba rift. As the Guba rift expanded northward, the conglomerates prograded beyond the northern margin of the older Roan rift and were deposited nonconformably upon pre-Katangan basement in what is now the northern part of the Lufilian belt. Thus, the Roan and Guba Groups record two distinct rifting stages and a northward shift of the rifted zone. The Kundelungu Group rests unconformably upon folded rift sequences, fills the first synorogenic foreland basin and consists of a proximal conglomerate wedge in the south of the Lufilian belt, grading northwards to distal finegrained facies. During the second contractional deformation, in response to the load of elevated nappes composed of Katangan strata, the Fungurume foreland basin was formed in the northern part of the belt and filled with conglomerates and allochthonous blocks derived from the orogenic source and deposited ahead of the northward-advancing orogenic front. The youngest foreland unit, late syn- to post-orogenic Plateau Group continental molasse (<575 Ma), was deposited partly synchronously with the Fungurume Group in the basin extending north of the Fungurume foreland. CONCLUSION Recognition of sedimentary genesis of the Katangan megabreccias reveals two major rifting stages due to Rodinia break-up and two major convergence phases between the Kalahari Craton in the south and Congo Craton in the north, which led to Gondwana assembly in the central African sector of the Pan-African orogen.

101

Supercontinents and Earth Evolution Symposium 2005

DETRITAL ZIRCON PROVENANCE OF CENTRAL MADAGASCAR AND NEOPROTEROZOIC TERRANE TRANSFER ACROSS THE MOZAMBIQUE OCEAN Ian CW Fitzsimons\ Bregje Hulscher^ ^Tectonics Special Research Centre, Department of Applied Geology, Curtin University of Technology, GPO Box U1987, Perth, WA 6845, Australia 0.Fitzsimons@curtin.edu.au) ^Tectonics Special Research Centre, The University of Western Australia, WA 6009, Australia

The Neoproterozoic East African Orogen records closure of the Mozambique Ocean and the resultant assembly of Gondwana, but details of its evolution remain uncertain. In the north, the Arabian-Nubian Shield comprises continental fragments and Neoproterozoic arc terranes assembled within the Mozambique Ocean between 850 and 650 Ma. These were accreted to the Saharan Metacraton along the Keraf Suture at 650-600 Ma. Sutures are difficult to identify further south, where gneissic rocks reflect late Neoproterozoic collision of the Congo and Dharwar cratons of Africa and India. Recent attempts to define a suture in the south have focused on 820-720 M a continental arc plutons in Madagascar. These were emplaced at an ocean margin, but it is unclear whether this ocean lay west or east of the arc. Critical to these arguments is whether the Malagasy Arc was emplaced into rocks of African or Indian provenance, and whether these rocks remained connected to their parent craton during the lifetime of the Mozambique Ocean.

Madagascar has two Archaean basement domains. The Antongil Block in the east is a fragment of the Dharwar Craton comprising 3190-2510 Ma gneisses and granite. The Antananarivo Block in central Madagascar comprises 2550-2500 Ma granite gneiss and is overlain to the southwest by greenschist- to granulite-facies quartzite, pelite and marble of the Itremo Group. Kyanite schist and ultramafic rocks of the Ambodiriana Group crop out between these two basement blocks and were metamorphosed at 520 Ma. The Antananarivo Block and Itremo Group were intruded by 820720 Ma plutons of the Malagasy Arc, and deformed and metamorphosed at 640-510 Ma during renewed granitic magmatism associated with final Gondwana assembly. Detrital zircon ages from the Itremo Group are identical to those in the Muva Supergroup of Zambia, implying that both units were deposited on the proto-Congo Craton at 1800 Ma. The Itremo Group and Antongil Block were therefore on opposing sides of the Mozambique Ocean before juxtaposition along an intervening suture, 102

most likely in the ultramafic-bearing Ambodiriana Group. However, the Itremo Group and Congo Craton are separated by 640-540 Ma granulitefacies gneiss in Gondwana reconstructions. This 400-km wide domain of Neoproterozoic tectonism contains juvenile crust and is associated with 850-650 Ma calc-alkaline plutons in Tanzania. It is along strike from ophiolitic sutures in Kenya and the Arabian-Nubian Shield, has a reversal of structural vergence, and is widely regarded as the best candidate for a collisional suture in the southern East African Orogen. These contradictions are resolved if the Itremo Group and its basement were rifted from the Congo Craton and accreted to the Dharwar Craton before final ocean closure. We correlate the Malagasy Arc with subduction of oceanic crust between the Antananarivo and Antongil blocks, and argue that they amalgamated when magmatism ceased at 720 Ma. Another strand of the Mozambique Ocean lay west of the Itremo Group, subduction of which beneath the Congo Craton was responsible for 850-650 Ma calcalkaline magmatism in Tanzania followed by final assembly of the Congo and Dharwar cratons at 650-600 Ma. Resultant 640-500 Ma orogenesis was most intense between the Congo Craton and Itremo Group, but also reactivated structures further east, including the Ambodiriana suture. This pattern of westward-younging sutures matches that in the Arabian-Nubian Shield, and we correlate the suture between central Madagascar and the Congo Craton with the Keraf Suture that terminated terrane accretion further north. We conclude that final closure of the Mozambique Ocean was preceded by terrane transfer from the African to the Indian margin. This model is analogous to the transfer of Gondwanan continental fragments across the lapetus, Rheic, and Tethys oceans, and, extending this analogy, we interpret multiple sutures in the East African Orogen in terms of Palaeo- and Neomozambique ocean basins.

Supercontinents and Earth Evolution Symposium 2005

P-T EVOLUTION OF THE BEMARIVO BELT (NORTHERN MADAGASCAR): THE FINAL ASSEMBLY OF GONDWANA Niels J6ns\ Volker Schenk\ Peter Appel\ Theodore Razakamanana^ Vstitut fur Geowissenschaften, Universitat Kiel, Germany (nj@nnin.uni-kiel.de) ^Departennent des Sciences de la Terre, Universite de Toliara, Madagascar

INTRODUCTION Madagascar holds a central position in the East African Orogen (EAO), which was formed during amalgamation of the supercontinent Gondwana. The island largely consists of Achaean and Proterozoic rocks that have been affected by Pan-African (ca. 650-500 Ma) metamorphic overprinting. However, the timing of collision and the sequence of different metamorphic events is still under debate. This is mainly due to the fact that most rocks record polymetamorphic histories, resulting in ambiguous geochronological data. The key to understanding the formation history of this part of Gondwana lies in detailed observation of metamorphic reaction textures, in combination with in-situ techniques of geochronology. The WNW-ESE-striking Bemarivo Belt of northern Madagascar truncates all other tectonic units. Its northernmost part consists of low-grade metamorphic epicontinental series, whereas the southern part is dominated by granulite-facies metapelitic rocks. Large volumes of granitoids and charnockites occur between both distinct areas. We used metapelites of the southern part of the belt to determine the pressure-temperature (P-T) conditions of metamorphism. To get an idea of the timing and the duration of metamorphism, we performed texturallycontrolled in-situ U-Th-total-Pb dating of monazites. P-T CONDITIONS A clockwise P-T evolution is constrained by prograde inclusions of kyanite in garnet, peakmetamorphic Grt-Opx-Sil-bearing politic assemblages, and the late-stage formation of cordierite rims around garnet. The prograde stage is confirmed by geobarometry (GASP equilibria) and prominent sillimanite pseudomorphs formed after kyanite. Peakmetamorphic conditions of 970°C and 8-10 kbar are inferred from the alumina content of orthopyroxene (8 wt.%) coexisting with garnet and sillimanite, feldspar thermometry, as well as GASP equilibria. Near-isothermal decompression

to pressures of 5-7 kbar is deduced from latestage Grt-Sil-Crd-Qtz assemblages. The subsequent cooling followed a near-isobaric path. GEOCHRONOLOGY Monazites generally consist of a homogeneous core (Mi) and a narrow overgrown rim (M2). In rare cases an additional magmatically-zoned core (Mo) is preserved. From differences in monazite chemistry and from inclusion relationships we conclude a two-stage growth history. The Mi monazites are dated at ca. 531 Ma, correlated with prograde garnet growth. The M2 rims formed at ca. 504 Ma, likely during the peak of metamorphism and decompression. MQ monazites are dated at ca. 717 Ma, and interpreted as detrital grains giving the maximum age of sedimentation of the protolith. Although the Mi and M2 ages overlap within error, textural control and differences in chemistry indicate that it is likely that they represent true metamorphic ages, pointing to a duration of ca. 30 million years from prograde garnet growth to heating and decompression. CONCLUSION The 717 Ma age of the detrital grains is interpreted as a maximum age of sedimentation. From the northern part of the Bemarivo Belt and from the Seychelles, which may be the source regions for the sediments, similar magmatic ages are known. At ca. 531 Ma, the Bemarivo Belt collided with the already amalgamated parts of Gondwana, leading to burial of the sediments to depth of at least 30 km. Heating to temperatures of nearly 1000°C is explained with heat supply by magmatic underplating just ca. 30 million years later. The decompression points to rapid extensional orogenic collapse followed by cooling at mid-crustal levels. The hightemperature metamorphism is one of the youngest known from the EAO and points to a very late accretion of the Bemarivo Belt to the Gondwana supercontinent.

103

Supercontinents and Earth Evolution Symposium 2005

THE CAUVERY SHEAR ZONE, SOUTHERN GRANULITE TERRAIN, SOUTHERN INDIA: A CRUSTAL-SCALE FLOWER STRUCTURE Talari RK Chettv National Geophysical Research Institute, Hyderabad-7, India (chettytrk@yahoo.co.in)

Structural interpretation of satellite data, followed by extensive field traverses and critical outcrop nnapping, suggest that the Southern Granulite Terrain (SGT), Southern India, can be divided into discrete tectonic blocks separated by a network of Proterozoic shear zones. Most prominent among them is an east-west tract (350 X 65 km) constituting a network of shear belts, collectively known as the Cauvery Shear Zone system (CSZ), which separates the Archaean Dharwar Carton to the north and Proterozoic granulites to the south. Multi-scale structural observations across the CSZ, along a north-south, 100 km wide corridor reveals a set of major shear belts: MoyarBhavani, Chennimalai-Noyil, Dharmapuri, Devattur-Kallimandeum and KarurOddanchatram shear zones. The charnockitic rocks have protolith ages (Nd model ages) between 3.6 to 2.6 Ga, whereas their formation and metamorphism could be at 2.5 Ga. Development of shear zones and associated intrusive granitoids and alkaline rocks can be related to a Neoproterozoic event (750-550Ma). Analysis of mesoscopic structures, such as deflection of foliation, S-C fabrics, duplex structures, extensional crenulation cleavages, development of ultramylonites and associated ashaped K-feldspar porphyroclasts and other

104

displacement patterns, clearly suggest dextral movements along all the shear zones. The behaviour and the attitude of stretching lineations along the CSZ are complex. Whereas steeply plunging lineations are dominant in some areas, many other areas are characterized by shallow to gently plunging lineations. This complex pattern of lineations is in some cases recorded even within a single outcrop. The regional disposition and geometry of shear zones, remarkable variations in foliation fabrics, persistent dextral kinematic indicators on all scales of observation, and coeval development of mylonitic fabrics (^550 Ma, based on Rb-Sr ages), suggest that the CSZ could represent a crustal-scale flower structure, typical of transpressive tectonics related to collisional processes. The shear zones may reflect inherited features from pre-existing thrust planes that acted as zones of weakness repeatedly reactivated during successive strains. Seismic refraction and wide-angle reflection studies also indicate the presence of disturbed heterogeneous crust associated with nearvertical fault patterns supporting the geometry of a flower structure at depth. Observed extrusive tectonics and heterogeneous strain patterns favor correlation of the CSZ with transpressive tectonic models of triclinic symmetry.

Supercontinents and Earth Evolution Symposium 2005

METAMORPHIC AND MAGMATIC PROCESSES IN DRONNING MAUD LAND DURING THE PAN-AFRICAN EVENT, EAST ANTARCTICA Masaaki Owada\ Sotaro Baba^, Kazuyuki Shiraishi^ and Yasuhito Osanai"^ ^Department of Earth Sciences, Yamaguchi University, 1677-1 Yoshida, Yamaguchi, 753-8512 Japan (owada@sci.yamaguchi-u.ac.jp) ^Department of Natural Environment, University of the Ryukyus, 1 Senbaru, Nishihara, Okinawa, 9030213, Japan ^ National Institute of Polar Research, 1-9-10 Kaga, Itabashi, 173-8515, Japan "^Division of Evolution of Earth Environments, Graduate School of Social and Cultural Studies, Kyushu University Ropponmatsu 4-2-1, Fukuoka, 810-8560, Japan

The Dronning Maud Land (DML) mountain range is inferred to represent the southeastern continuation of the East African Orogen. The Orogen formed as a result of collision of East and West Gondwana during the Pan-African event in late Neoproterozoic to Early Palaeozoic time. The older, Mesoproterozoic (ca. 1.1 Ga) basement rocks in DML were differentially reworked during the Pan-African event. Recent geochronological studies revealed that the age of high-grade metamorphism and voluminous igneous activity occurred at approximately 530 to 490 Ma in DML. In this presentation, we describe the metamorphic and igneous history of central to eastern Dronning Maud Land and discuss the tectonic evolution with respect to Gondwana amalgamation during the Pan-African event. GENERAL GEOLOGY OF DRONNING MAUD LAND Metamorphic rocks in central and eastern DML underwent granulite fades metamorphism. The Pan-African tectonothermal overprint was less intense in the western DML. Voluminous granitic rocks are exposed over a large area, from H.U Sverdrupfjella (2°E) in the west to Sor-Rondane (28°E) in the east. METAMORPHISM A metamorphic history for Filchnerfjella (central DML) is summarized below. 1) Symplectic orthopyroxene grew at the expense of garnet and hornblende porphyroblasts in the garnet-orthopyroxene mafic gneiss. 2) In the garnet-sillimanite gneiss, cordierite includes sillimanite, hercynite, and ilmenite, and the garnet locally contains hercynite inclusions. The following reactions can be inferred from the early stage to the later stage.

Grt + Qtz = Opx + PI Hbl + Qtz = Opx + PI + H2O Grt + Sil + Qtz = Crd These reactions are promoted by decreasing pressure and/or increasing temperature in Filchnerfjella. Similar metamorphic stages with clockwise PT-t paths can be recognized in Sor-Rondane. The peak metamorphic conditions reach approximately 800 MPa and 800°C. TECTONIC SIGNIFICANCE OF INTRUSIVE ROCKS Emplacement of the granite-syenite-charnockite suite in DML must postdate the main metamorphism and deformation event, because the syenite has not undergone any pervasive deformation or metamorphism. The ages of post-tectonic plutonic rocks are 520 to 490 Ma in central DML and 530 to 500 Ma in eastern DML. The suite is petrographically and geochemically similar throughout the entire area of DML. 525 to 500 Ma lamprophyre dykes are also present in DML. It is noteworthy that the magmatic rocks intruded the deformed high-grade rocks 30 to 40 million years after the main deformation period. Geochemistry of the igneous rocks, combined with the deformation and metamorphic history, indicates they were formed by melting of lower crust after crustal thickening due to continental collision. We infer that a significantly elevated geotherm was required to produce DML granitic magmatism and related high-grade metamorphism. This geotherm can be attributed to removal of thickened lithospheric mantle due to continental collision and a subsequent upwelling mantle plume beneath DML during the Pan-African event.

105

Supercontinents and Earth Evolution Symposium 2005

IN-SITU SHRIMP DATING OF ZIRCON IN MIGMATITE: RESOLVING THE PAN-AFRICAN DEVELOPMENT OF THE LUTZOW-HOLM COMPLEX, EAST ANTARCTICA Dan J Dunklev\ Y Motoyoshf, CJ Carson^ T Hokada\ K Shiraishi^ ^National Institute of Polar Research, 1-9-10 Kaga, Itabashi-ku, 173-8515, Tokyo, Japan (danield@nipr.ac.jp) Northern Territory Geological Survey, PO Box 3000, Darwin, NT 0800, Australia

Supracrustal lithologies at Skallevikshalsen, a coastal bluff in Lutzow-Holm Bay, East Antarctica, underwent high-strain deformation during metamorphism with a clockwise P-T-t path and decoupled peak pressures of >10 kbar and peak tennperatures of 780-960°C. Anatexis began before the thernnal peak, producing abundant felsic nnagnnas that crystallized during and after extensional shearing related to orogenic collapse and deconnpression. Zircon nnorphology and composition in migmatized metapelite differs between leucosome and melanosome hosts. Zircon grains in restitic assemblages (pl-sill-crn-ilm-grt-apmon) are equant and either unzoned or sectorzoned. Inherited cores are minute or not present. Moderately high U and low Th contents are typical. In leucosome and garnet-sillimanite-felsic gneiss, which are interpreted to have crystallized from anatectic melts, zircon grains have complex morphologies, with varying proportions of inherited cores, weakly-zoned moderate- to highU inner zones, moderate- to low-U zones with oscillatory zoning, moderate-U, weakly zoned outer zones, and thin high-U rims. Oscillatoryzoned zircon dominates most grains, and is attributed to kinetically-retarded growth in felsic melt. Zircon growth in the presence of melt is

106

indicated also by zircon grains with inclusions of sillimanite in the rims of peritectic garnet porphyroblasts. Zircon also grew during the release of fluids by melt crystallization, as recorded by U-rich overgrowths on monazite and ilmenite and thin rims on zircon grains. Isotopically concordant SHRIMP ages are dispersed between ca. 600 and 500 Ma, consistent with zircon growth at various stages of a progressive metamorphic 'cycle'. Oscillatoryzoned zircon ages cluster between 540 and 560 Ma, whereas high-U weakly zoned zircon ages are more widespread, with modal ages of ca. 530Ma. The former age group is associated with melt-present zircon growth, whereas the latter is associated with fluid activity, especially before and after anatexis. Zircon grew in several stages during PanAfrican orogenesis, with compositions and morphologies that reflect fluid vs. melt-dominated modes of transport. Isotopic age data can be associated with specific metamorphic processes, rather than with generic 'peak metamorphic' conditions. No isotopic ages were obtained that would indicate igneous or metamorphic activity in the Lutzow-Holm Complex during the formation of Rodinia.

Supercontinents and Earth Evolution Symposium 2005

THE EAST AFRICAN OROGEN AND THE AMALGAMATION OF GONDWANA-A PALAEOGEOGRAPHIC PERSPECTIVE Alan S Collins^

Sergei A Pisarevsky^

^Continental Evolution Research Group, Geology and Geophysics, School of Earth and Environnnental Sciences, University of Adelaide, Adelaide, SA 5005, Australia (Alan.Collins@adelaide.edu.au) ^Tectonics Special Research Centre, School of Earth and Geographical Sciences, The University of Western Australia, Crawley, WA 6907, Australia

The Neoproterozoic global reorganisation that saw the demise of Rodinia and the amalgamation of Gondwana took place during an incredibly dynamic period of Earth evolution. To better understand the palaeogeography of these times, and hence help quantify the interrelations between tectonics and other Earth systems, we here integrate Neoproterozoic palaeomagnetic solutions from the major blocks that made up eastern Gondwana, with the large amount of recent geological data available from the orogenic belts that formed as eastern Gondwana amalgamated. From this study, we have: 1) identified large regions of pre-Neoproterozoic crust within late Neoproterozoic-Cambrian orogenic belts that significantly modify the geometry and number of continental blocks present in the Neoproterozoic world; 2) suggested that one of these blocks, Azania, which consisted of Archaean and Palaeoproterozoic crust within the East African Orogen of Madagascar, Somalia, Ethiopia and Arabia, collided with the Congo-Tanzania-

Bangweulu Block at -650-630 Ma to form the East African Orogeny; 3) postulated that India did not amalgamate with any of the Gondwana blocks until the latest Neoproterozoic-Cambrian forming the Kuunga Orogeny between it and Australia-Mawson and coeval orogenesis between India and the previously amalgamated Congo-Tanzania-Bangweulu - Azania Block (we suggest the name 'Malagasy Orogeny' for this event); and 4) produced a palaeomagnetically and geologically permissive model for Neoproterozoic palaeogeography between 750 and 530 Ma, from the detritus of Rodinia to an amalgamated Gondwana. The East African Orogen is one of the major orogens that formed as Gondwana amalgamated and preserves much evidence within it to constrain the tectonic evolution of this region. In this presentation we will integrate new and existing data from Madagascar, East Africa and southern India and place this in a global palaeogeographic construct.

107

Siipercontinents and Earth Evolution Symposium 2005

THE PAN-AFRICAN OROGENY IN NORTHWESTERN HIMALAYA: THE MISSING LINK IN THE TRANSITION FROM RODINIA TO GONDWANA Pradip Kumar Dasgupta Department of Geology, University of Calcutta, India (pradip_geol@rediffmail.com)

East Gondwana has been considered to have formed during the late Neoproterozoic circum east Antarctica/Rodinia Orogeny and reactivated by the late Neoproterozoic to early Palaeozoic Pan-African/Brasiliano Orogeny. The basic structure and tectonic framework of the Indian subcontinent is a mosaic of four distinct Archaean to Palaeoproterozoic crustal blocks/terrains, one of which, the Bundelkhand protocontinent, extends below the Himalaya in the north, while Trans-Aravalli Fold Belt continues westward and merges with the Arabian-Nubian Shield. The Neoproterozoic Rodinia supercontinent, of which India formed a part, amalgamated during the interval 1300-900 Ma. However, the Rodinia supercontinent was short-lived and breakup started at about 900 Ma and continued until 600 Ma. The signature of Rodinia is discernible in the basement of Tethyan sedimentary sequences, intruded by Palaeozoic granites. The Central Indian Tectonic Zone (CITZ) with Bundelkhand Trans-Aravalli - Central Arabian Craton, decisively represent a part of this supercontinent. Short-lived Rodinia evolved into the Gondwana supercontinent (600-500 Ma) in response to PanAfrican collisions. East Asian allochthonous crustal fragments were successively rifted and detached from the margin of East Gondwana as three elongate continental slivers during Devonian, early Permian, and late Triassic-late Jurassic times, causing successive opening of Palaeo-, Meso- and Neo-Tethys which occupied the spaces between the slivers and Gondwana. The northwestern Himalayan terrain recorded coupling of right- and left lateral shear over long distances and progressive suturing of these micro-continental slivers. The Upper Lahaul, Spiti, and Zanskar regions in the northwestern Himalaya documented sedimentation and tectonics related to the northern passive margin of Indian plate. The Higher Himalayan crystalline sequence (HHCS) of Zanskar includes metasedimentary rocks. These are metamorphic equivalents of the Phe Formation shed over the Zanskar Shear by the

108

end of the Precambrian. The HHCS was intruded by orthogneiss, protoliths of which are associated with late Pan-African tectonothermal events (constrained by a range of Rb-Sr isochron and zircon U-Pb ages between 463 ± 13 Ma and 549 ± 17 Ma), greatly affecting the Zanskar area. Further evidence of pre-Himalayan deformation can be observed in the stratigraphic record of the Zanskar region. The Permian Panjal Traps rest directly on, and truncate folds within, the deformed Precambrian to Cambrian Phe and Karsha Formations. Sediments younger than the Permian, therefore, are not affected by this pre-Himalayan deformation. The Cambrian Kurgiakh Formation follows these undulations of the underlying sediments, but is unconformably overlain by the Ordovician Thaple Formation. Thus, in Tanze and in Spiti, the presence of an unconformity at the base of Ordovician points towards a distinct tectonic event at the Cambro-Ordovician boundary. Despite the absence of schistosity related to deformation, uplift and erosion of a cratonic basement, related to a late Pan-African orogenic cycle, might have occurred during CambroOrdovician time. A Pan-African orogeny has been postulated on the basis of Ordovician granites and Cambrian alluvial conglomerates. Furthermore, the Cambro-Ordovician unconformity is consistent with the Pan-African orogeny. It can be added that the transition from Rodinia to Gondwana is manifested in different parts of the Indian subcontinent with the stamp of Pan-African events. In conclusion, it is observed that there are dozens of felsic igneous plutons related to the pre-Himalayan Orogeny, intruded during Early Palaeozoic time (-500 Ma) in different tectonic settings all along the Himalayan range. Signatures of Pan-African orogeny in the CITZ, Trans-Aravalli Fold Belt, Bundelkhand protocontinent, Arabian Craton, and Himalayan Orogen serve as a link for the transition from Rodinia to Gondwanaland in the northwestern Himalaya, where a record of Pan-African events is consistently preserved.

Supercontinents and Earth Evolution Symposium 2005

EVIDENCE FROM BREAKUP OF RODINIA TO ASSEMBLY OF GONDWANA IN CHINA CONTINENT Lu Songnian, Li Huaikun, Chen Zhihong, Xiang Zhenqun Tianjin Institute of Geology and Mineral Resources, CGS (tjlsongnian@cgs.gov.cn)

Neoproterozoic thermo-tectonic events are well developed in the Yangtze Platform (YTP), Tarim Platfornn (TRP), and the Central China Orogen (CCO). Apart from the early Neoproterozoic events related to the assembly of Rodinia, major thermo-tectonic events reflecting the dispersal of Rodinia and assembly of Gondwana are also evident. Key evidence includes mafic and ultramafic dyke swarms, layered intrusions, within-plate flood basalts, and A-type granites from 820 to 800 Ma, and eclogites, granulite metamorphism, arc volcanic rocks, ophiolites, and syn-orogenic granites with ages concentrated between about 520 and 400 Ma. The rifting records of western Rodinia have been recognized by Australian and Chinese geologists. Isotopic ages of 827 ± 6 Ma and 824 ± 4 Ma have been reported for the Gairdner Dyke Swarm of South Australia and for dykes in South China, respectively. Recent geological and geochronological data show that coeval dyke swarms and layered intrusions are not only found in South Australia and South China, but also in the TRP and CCO. For example, ultramafic intrusion in the Jinchang area of the Longshoushan Mountains is dated at 827 ± 8 Ma. A diabase dyke swarm dated by SHRIMP UPb at 821 ± 21 Ma intruded marbles of the Mesoproterozoic Wandonggou Group in the Quanji area. Mafic-ultramafic intrusions are recognized from the north margin of Tarim Basin. These intrusions are mainly composed of pyroxenite, olivine gabbro, and gabbro. Zircons from olivine gabbro yielded a SHRIMP U-Pb age o f 8 1 7 ± 11 Ma. In addition, several SHRIMP U-Pb zircon and baddeleyite ages between 810 and 820 Ma have

been obtained in the Hannan Complex, situated in the northwestern margin of YZP. For example, we have measured SHRIMP U-Pb baddeleyite and zircon ages of 822 ± 25 Ma and 811 ± 8 Ma for gabbro, and a SHRIMP U-Pb age of 820 ± 20 Ma was obtained for granite in the same area. It is noteworthy that the Early Paleozoic thermo-tectonic events related to an orogeny process (the Pan Huaxia Orogeny of 520-400 Ma) are distinguished from the CCO, which caused assembly between the Tarim and North China Platform, as well as some massifs in the CCO. Metamorphism of eclogites and granulites occurred at about 500 Ma in the north margin of the Qaidam Massif and central Altyn Tagh. The formation of ophiolites in the West and East Kunlun and North Qilian Mountains occurred mainly between 500 and 400 Ma. Arc volcanics in the North Qilian Mountain, Quanji area, and northern Qinling Range are 520 to 430 Ma in age. Most syn-orogenic granites are dated at 500 to 400 Ma. An important unconformity between the upper Devonian and pre-Devonian strata is commonly developed in CCO. There is no doubt that the above events are responses to the early Paleozoic Pan-Huaxia orogeny. It is clear that the time range of the orogen is younger than that of Pan-African orogenic episode by about 50 to 100 million years. We consider that CCO to be different from the East African Orogen, and was situated in peri-Gondwanaland. This project was supported by China Geological Survey grants (No. 200113900070 and 200313000060) and by a National Science Foundation of China grant (No. 40032010).

109

Supercontinents and Earth Evolution Symposium 2005

DETRITAL ZIRCON AND INTRUSIVE GEOCHRONOLOGY OF THE KALAK NAPPE COMPLEX, ARCTIC NORWAY: A NEOPROTEROZOIC ACCRETIONARY COMPLEX Chris L Kirkland\ JS Daly\ MJ Whitehouse^ ^Geology Department, University College Dublin, Belfield, Dublin, Ireland (Chris.Kirkland@ucd.ie) ^Swedish Museum of Natural History, Stockholm, Sweden

The Kalak Nappe Complex (KNC) in Finnmark, Arctic Norway, is a classic example of a thrustassembled tectonostratigraphy within the Caledonian orogenic belt. Early studies from this region played a major role in defining the orogenic events that underpin our understanding of the Scandinavian Caledonides. A preCaledonian, late Neoproterozoic history is now well established through dating of granites that cut early structures. Zircon rims from two granites yield an age of 840 ± 5 Ma, validating the Porsanger Orogeny. This is supported by 828 ± 5 Ma ages for two intrusive pegmatites. Discordant structural relationships are only preserved in rare cases of low overprinting Caledonian strain. Granites dated at 976 ± 3 Ma (n=3) are found at lower structural levels within the KNC. Deformation of the enclosing psammite is constrained between ca. 980 Ma and ca. 1025 Ma, suggesting a link with the Sveconorwegian (Grenville) orogeny. Synkinematic migmatitic leucosomes within the uppermost nappe of the KNC, dated at 709 ± 4 Ma, are identical in age to metamorphic zircon overgrowths in the underlying nappe. Hence juxtaposition of these nappes occurred at ca. 710 Ma, clearly predating Caledonian events. These spatial and temporal patterns of granite intrusion and deformation point to episodic terrane accretion from Early Neoproterozoic times up to the Cryogenian. Dating of detrital minerals is a powerful tool to test the model of episodic terrane accretion for the KNC. The KNC has been assumed to contain a single conformable stratigraphic package, the Soroy Succession. However, the uppermost unit is a Llandovery deposit of Laurentian affinity, much younger than the others. The lowermost member of the Soroy Succession, the Klubben Psammite, must represent at least two distinct units. Within the nappes affected by the Grenville event, detrital zircons have age peaks at ca.

110

1720, 1640, 1500, and 1300 Ma, along with a minor peak at 1100-1000 Ma. These psammites were deposited between ca. 980-1030 Ma, constrained by intrusive granites and detrital components. Within the overlying nappe to the west, affected by the Porsanger Orogeny, there are no ca. 1720 or 1300 Ma peaks and a new broad peak at ca. 1180-920 Ma is found along with a few Archaean grains. Deposition of this psammite unit occurred between ca. 840 and 920 Ma. Within the uppermost nappe the youngest recorded component is 910 ± 15 Ma. The latter two sets of psammites may belong to the same stratigraphic package. In all populations there are few Palaeoproterozoic grains older than 1800 Ma. The age peaks at around 1650 and 1500 Ma correspond to the Labradorian and Pinwarian magmatic events in NE Laurentia and Baltica whereas the younger age peaks at 1100 Ma correspond to early phases of the Grenville orogen in Laurentia and the Sveconorwegian region in Baltica. The detrital age populations are similar to those from the Scottish Moine, which have a suggested Laurentian affinity. However, the detrital population is also compatible with deposition of the KNC psammites close to Baltica with some exotic influence from Laurentia. With Baltica inverted in relation to Laurentia during most of Neoproterozoic, the Timanian/Baikalian, Avalonian and Cadomian arcs can be linked to Baltica and Laurentia respectively in a Pacific-rim scenario. This palaeogeography allows a similar depositional setting for the Moines and the terranes of the KNC. The ca. 980, 840, and 710 Ma orogenic events within the KNC likely reflect accretion of terrane fragments from the Peri-Gondwanan region. Intriguingly, their times and modes of docking with Baltica are as yet poorly constrained.

Supercontinents and Earth Evolution Symposium 2005

A BALTICA PROVENANCE FOR DETRITAL ZIRCONS FROM EARLY PALEOZOIC SEDIMENTS OF SOUTHERN NOVAYA ZEMLYA, ARCTIC RUSSIA V i c k y P e a s e \ R S c o t t ^ K Eliaeson^ ^Dept. of Geology & Geochemistry, Stockholm University, 106 91 Stockholm, Sweden (vicky.pease@geo.su.se) CASP, Dept of Earth Sciences, University of Cambridge, West Building, 181A Huntingdon Road, Cambridge, CBS ODH, UK

The Novaya Zemlya archipelago lies between the Barents Sea and Kara Sea on the Russian Arctic Shelf. The archipelago is generally considered to be part of the Uralian Orogen, which stretches from the Caspian in the south to the Arctic Ocean in the north, and records the Late Paleozoic collision of Baltica with Siberia. In the northern part of the orogen, an additional independent crustal block, the North Kara Terrane, has been invoked by some authors to have been involved in the collisional events. Understanding the nature and extent of these basement provinces in the high Arctic is, however, hampered by lack of data. Precambrian through Early Paleozoic basement elements are exposed at a number of localities along the length of the Novaya Zemlya archipelago. In the southernmost part of the archipelago, Precambrian through Early Paleozoic strata are exposed in the core of an Early Mesozoic antiformal structure. An angular unconformity beneath Early Ordovician coastal and shallow marine sediments is well exposed at a number of localities. Previous interpretations suggest that rocks below the unconformity are Neoproterozoic in age, implying that the unconformity was generated in response to endPrecambrian Timanide orogeny well documented to the southwest in Timan-Pechora and adjacent regions. To understand the affinities and ages of strata above and below the angular unconformity in

southernmost Novaya Zemlya, samples of sandstone were collected at regular stratigraphic intervals for detrital zircon provenance investigations. Several sites were sampled and the results of the first site analyzed indicate that: (1) strata below the unconformity are not Neoproterozoic as mapped but Cambrian or younger (<530 Ma); (2) the maximum age of strata overlying the unconformity is Late Cambrian/Early Ordovician (<490 Ma), in good agreement with Tremadocian paleontological data; (3) peaks in the cumulative probability curve can be derived 'locally' from known Baltica sources; (4) samples above the unconformity record a cratonic sediment component reflecting erosion of the Baltic shield. These results imply that the unconformity does not document the typical late Precambrian Timanide event, but rather a relatively short-lived tectonic event of end-Cambrian age. This suggests affinities with an unconformity of identical style, age, and duration recognized in Severnaya Zemlya to the northeast, a key component of the so-called North Kara Terrane. However, because the cumulative probability curves indicate a clear Baltica signature for southern Novaya Zemlya, we regard this as further evidence that the North Kara Terrane was not a separate entity, but a northeast continuation of Baltica.

Ill

Supercontinents and Earth Evolution Symposium 2005

IS THERE A LINK BETWEEN AMALGAMATION OF GONDWANA AND THE SPAWNING OF COMPLEX LIFE? Richard J Squire\ Ian H Campbell^ Charlotte M Allen^ Chris JL Wilson^ ^School of Earth Sciences, University of Melbourne, Victoria, 3010, Australia (frsquire@unimelb.edu.au) Research School of Earth Sciences, Australian National University, Canberra, ACT, 0200

The influx of quartz-rich sediment displaying the prominent Pan African and Grenvillian-aged peaks on what is now six continents (Africa, Arabia, India, Australia, New Zealand and Antarctica) during the Early Cambrian, is the most striking feature of the early Palaeozoic detrital-zircon age spectra from Gondwana. The timing of this sediment influx from the same source rocks broadly coincides with the convergence of East and West Gondwana. Therefore, we suggest that the collision of these two large palaeocontinents produced an enormous mountain chain along the eastern margin of Africa, the presently preserved root zone of which is referred to as the East AfricanAntarctic Orogen; here, we refer to the surface expression of this collisional zone as the Transgondwanan Supermountains to emphasise its palaeogeographic significance. The river systems that drained this enormous mountain chain to either side transported huge sediment volumes and then deposited them in a series of giant (turbidite) fans, named here the Gondwana Super-fan System. The eventual shutdown of the Super-fan System is recorded by a reduction in areal extent of the quartz-rich sediments during the Late Ordovician, but this means that the Gondwana Super-fan System was active for at least 100 million years. We propose that the diachronous convergence of East and West Gondwana from -650 Ma fluxed increasing volumes of continentally derived nutrient-rich sediments into equatorial oceans during the Ediacaran period (e.g., increasing seawater ®^Sr/®®Sr ratios) and must have dramatically modified the circulation patterns of Earth's oceans. Although establishment of the Transgondwanan

112

Supermountains also overlaps with the catastrophic Marinoan and Gaskiers glaclation events (i.e., snowball Earth), its impact on these global climate-change events is presently uncertain. Despite this, we believe that the radically altered ocean-atmosphere conditions created by establishment of the Transgondwanan Supermountains generated the biologically explosive environment that spawned complex life during the Ediacaran period. The rapid decline of Ediacara fauna towards the end of the Ediacaran period (-543 Ma) was followed by the sudden appearance and highly explosive diversification of skeletonised marine phyla, the diversification rates of which were markedly higher than at any other period in Earth's history. This catastrophic biological event coincided with rapid growth of the Gondwana Super-fan System as the Transgondwanan Supermountains rotated and drifted away from the equator and into the adjacent monsoonal belt. Therefore, we believe that establishment of the monsoonal conditions that rapidly increased erosion rates and development of the Gondwana Super-fan System may have also produced a rapid increase in sequestration of CO2 into the oceans. The sudden appearance and explosive diversification of skeletonised marine phyla may thus reflect the dramatic change in the biological carbon cycle from -543 Ma following establishment of the intense monsoonal conditions associated with the Transgondwanan Supermountains. The demise of the supermountains, as recorded by the rapid reduction in areal extent of the Gondwana Super-fan System, occurred during the middle to late Ordovician.

Siipercontinents and Earth Evolution Symposium 2005

DID SLIVERS OF PRECAMBRIAN CONTINENTAL CRUST BREAK OFF RODINIA TO FORM BASEMENT TO THE TASMANIDES OF EASTERN AUSTRALIA? Richard A Glen^ and Anthony J Crawford^ ^Geological Survey of New South Wales, Department of Primary Industries, Hunter Region Mail Centre, PO Box 344, NSW 2310, Australia ^Centre for Ore Deposit Research, University of Tasmania, GPO Box 252-79, TAS 7001, Australia

The Tasmanides of eastern Australia represent a series of orogenic belts that record the break-up of Rodinia followed by formation of continental crust along the margin between Gondwana and the proto-Pacific plate. Several workers have suggested that the rifting of the Rodinia supercontinent between 830 and 520 Ma resulted in the calving of slivers of Precambrian continental crust, some of which occur as basement to parts of the Tasmanides of eastern Australia. The best place to assess this probability is in Victoria, which contains the Delamerian Orogen in the west and the Lachlan Orogen in the east. Western Victoria is occupied by the outboard part of the Delamerian Orogen that marks the Pacific edge of Gondwana that formed as a result of the multiphase Delamerian Orogeny. The first phase there, at -505 Ma, consisted of accretion of 520-510 Ma Cambrian boninitictholeiitic forearc crust that lay west of a westfacing intraoceanic arc. This was followed by extension, with the formation of rift basins and post-collisional Late Cambrian (500 Ma) age andesitic volcanics, followed by a second phase of the Delamerian Orogen. Data from western Tasmania to the south show identical relationships but also show that the forearc rocks were accreted onto Precambrian crust overlain by 600-580 Ma rift basins. In Victoria, the Lachlan Orogen lies east of this Delamerian margin and consists mainly of Ordovician to Carboniferous rocks. However, several hundred kilometres east of this Delamerian margin are several major fault systems, the hanging walls of which contain deformed Cambrian igneous rocks and overlying sediments that pass up into Ordovician cratonderived turbidites. The Cambrian volcanics are either andesitic, matching identically postcollisional volcanics in western Victoria and also western Tasmania, or boninitic-tholeiitic volcanics that match the forearc rocks of the Delamerian Orogen. Other Cambrian mafic to intermediate igneous rocks occur within the Lachlan Orogen along the south coast of Victoria

and have been linked by aeromagnetic data across the Cretaceous-Tertiary Bass Basin into Cambrian volcanics in western Tasmania. These aeromagnetic data have been used to suggest that central Victoria is underlain by the Selwyn Block, consisting of Precambrian continental crust plus accreted mafic to andesitic volcanics that have been through a mid - Late Cambrian Delamerian deformation. Our alternative interpretation of the basement geology of central Victoria is based on the view that the Delamerian Orogeny was followed by major and rapid roll-back of the proto-Pacific plate of approximately 1000 km. As a result, parts of the Delamerian margin and outboard Cambrian volcanics that lay east of the Delamerian deformation front were rifted away from the margin in the earliest Ordovician to form local basement to the Ordovician turbidites that represent submarine fan deposits shed off the older craton. The key to this interpretation comes from stratigraphy. Accreted Cambrian volcanics in the Delamerian Orogen pass rapidly up into latest Cambrian turbidites. In contrast, outboard volcanics pass up from mass flow deposits involving limestone detritus, through shales, into deep sea cherts, abruptly overlain, with no unconformity, by basal Ordovician turbidites. To us, these indicate that the outboard volcanics underwent rapid submergence over 13-15 million years, consistent with rifting. We also apply this interpretation to the geochemically similar boninitic rocks that lie along the Peel-Manning Fault System of the New England Orogen that formed outboard (?), and to the north, of the Lachlan Orogen. In this interpretation, there are no slivers of Precambrian continental crust under the Lachlan Orogen and no Selwyn Block. On a wider scale, the generation of the North Queensland Orogen above Precambrian continental crust highlights its formation adjacent to the Delamerian margin. This contrasts with the evolution of the southern Tasmanides that was marked by major accretion events that reflect the importance of back-arc spreading in the

IJ3

Supercontinents and Earth Evolution Symposium 2005

Ordovician and again in the Silurian. Whether there was a sharp transform fault between the northern and southern Tasmanides across the Delamerian margin inherited from Rodinia breakup is unknown.

114

Glen publishes with the permission of the Deputy Director-General NSW Department of Primary Industries-Minerals

Supercontinents and Earth Evolution Symposium 2005

THE TIMING AND DURATION OF THE DELAMERIAN OROGENY: CORRELATION WITH THE ROSS OROGEN AND IMPLICATIONS FOR GONDWANA ASSEMBLY J o h n F o d e n \ M a r l i n a A Elburg^'^ ^Geology and Geophysics, University of Adelaide, Adelaide, SA 5005, Australia Oohn.foden@adelaide.edu.au) Department of Geology and Soil Sciences, Ghent University, Krijgslaan 281 S8 9000, Ghent, Belgium

The Antarctic Ross and the Australian Delamerian orogenies are the consequence of stress transfer to the outboard trailing edge of the newly assembled Gondwana supercontinent. This tectonic reorganisation occurred in the Early to mid-Cambrian on completion of Pan-African deformation and subduction along the sutures between eastern and western Gondwanan continental fragments. Prior to this, Neoproterozoic to Early Cambrian rocks in eastern Australia were formed in a passive margin and record dispersion of Rodinia with consequent opening of the proto-Pacific. Our new U-Pb and Rb-Sr geochronology shows that in the South Australian (Adelaide Fold Belt) domain of the Delamerian Orogen, contractional orogenesis commenced at 514 ± 3 Ma and persisted for -2A million years until 490 ± 3 Ma, terminated by rapid uplift, cooling, and extension, in association with post-tectonic magmatism. Integration of new and published UPb and "^^Ar-^^Ar geochronology from the entire Ross-Delamerian belt shows that although both the Delamerian and Ross Orogens have synchronous late magmatic and terminal cooling histories, the Ross Orogen commenced its convergent orogenic history at about 540 Ma. This was 25 million years before Delamerian deformation began. During the Early Cambrian, eastern Australia was still in a state of extension (or transtension), with opening of the Kanmantoo Basin and associated anorogenic, largely mafic magmatism. This basin received sediment from the already exposed Ross Orogen to the south.

The simultaneous first occurrence of strain fabrics and subduction-related magmatism (including boninite, granite and andesite lavas) at -514 Ma in New Zealand, Victoria, South Australia, New South Wales, and Tasmania implies that the Delamerian orogeny was driven by ridge-push forces transmitted on the initiation of westward-dipping subduction. Subsequent eastward slab rollback at 490 Ma may have occurred when the new slab had reached the transition zone at 650 km depth, resulting in upper-plate extension and anorogenic Basin and Range-style magmatism in South Australia and Tasmania (Mt Read belt). The delayed onset of subduction in the Australian sector of the margin implies that westward motion of the Australian portion of eastern Gondwana continued to be accommodated during the late Early Cambrian by subduction or deformation along either the Mozambique Suture or at the northern end of the South Prince Charles Mountains - Prydz Bay suture. The footprint of the Ross and Delamerian Orogens subsequently became the locus of the Ferrar and Tasmanian Jurassic low-Ti tholeiite magmatic provinces. Based on highly anomalous Nd, Sr, Pb, and O isotopic compositions, these appear to reflect the melting of anomalous crustally contaminated lithospheric mantle. We link this mantle contamination to terminal RossDelamerian exhumation and uplift, possibly associated with mantle lithosphere delamination.

115

Supercontinents and Earth Evolution Symposium 2005

POSTERS

116

Supercontinents and Earth Evolution Symposium 2005

A FIRST TENTATIVE CLASSIFICATION OF RODINIA S DESCENDANTS IN SOUTH AMERICA Beniamim Blev de Brito Neves\ Reinhardt A Fuck^ Carlos Schobbenhaus^ ^Universidade de Sao Paulo, 05508-080. Sao Paulo - SP, Brazil (bbleybn@usp.br) ^Universidade de Brasilia, 70910-000, Brasilia - DF, Brazil ^CPRM - Brazilian Geological Survey, Brasilia - DF, Brazil

Research of Rodinia supercontinent descendants in South America is still under development. Starting points for this goal are first the analysis of these considered as paradigm (the Amazonia Block) and second, to look for some genetic expected characteristics for these fragments, which are today masked and modified by the many superposed Neoproterozoic (Brasiliano) and Phanerozoic (Fammatinian, Hercynian and Andean) tectonic, thermal and accretionary events. Some progress could be performed on this subject along the lUGS-IGCP 440. The main recognizable characteristics are: (a) the presence of some Archean seed-nuclei, bounded and arranged inside (b) the framework of the Paleoproterozoic collage, mainly Rhyacian and Orosirian orogeneses; (c) the tectonicmagmatic marks of the Statherian taphrogenesis (rifting, dike-swarms, volcanic traps, intracratonic granitic plutonism etc); (d) Mesoproterozoic intracratonic volcanic and sedimentary traps; (e) Orogeneses or other deformational processes of late Mesoproterozoic - early Neoproterozoic age. Tectonothermal Brasiliano reworking is widespread throughout the Precambrian provinces of the continent, privileging the Mesoproterozoic belts (then, younger thermal ages sites) and only the Amazonia Block could be discriminated as a relatively preserved domain. Reworking by Phanerozoic orogeneses was intense along the westernmost Mesoproterozoic belts and blocks (some minor Paleoproterozoic areas) which now occur as scattered basement windows and inliers (in the Andean Chain), of rather difficult discrimination. Following the above mentioned characteristics and according to the present geological conditions (reworking rates) it is possible to recognize different categories of candidates to be descendants of Rodinia. Most of them have been absent in Rodinia

reconstructions of the last 14 years. Actually, only three blocks, Amazonia, Sao Francisco, and Rio de la Plata, have been included in most previous reconstructions. Based on both the degree of knowledge and intensity of tectonicthermal reworking, there are different categories that may be disclosed. The first one is reserved to these well-preserved blocks, cases of Amazonia and Sao Francisco blocks (a part of Congo-Kasai-Angola). The same is true for the (Archean) Luis Alves cratonic segment. It is necessary to point out that some large descendants/candidates do not crop out, for they occur beneath the Gondwanan sedimentary basins of Parnaiba (Parnaiba block) and Parana (Paranapanema block) which identification is recent, based upon geophysical surveys, isotopic studies, and data from several boreholes. The conditions of tectonothermal reworking of this (second) group are not yet understood. The third group is assembling partially reworked segments, but with some significant territorial losses in their original extents and shapes (like Rio de La Plata, Rio Apa) and segments presenting extensive reworking, but which exerted an important role as structural highs in the Brasiliano framework (Granja, Central Ceara, Rio Grande do Norte, Eastern Pernambuco-Alagoas, Goias Central/Tocantins, Curitiba, Nico Perez). The fourth group includes segments and lithostructural portions that were completely reworked and that were incorporated to the framework of the Brasiliano provinces (Alto Moxoto, Alto Pajeu, Riacho Gravata, Guanhaes, Cabo Frio, Punta del Este etc.). The fifth group was reserved to include those windows of Paleoproterozoic (e.g. Arequipa) and (mostly) upper Mesoproteroic terranes (such as Garzon-Santa Marta, Cuyania, Pampia) along the basement of the Andean Chain.

117

Supercontinents and Earth Evolution Symposium 2005

VOLGO-URALIA: SHRIMP EVIDENCE OF STRONG PALAEOPROTEROZOIC REWORKING OF ARCHAEAN CRUST Svetlana Bogdanova\ Bert De Waele^, Elena Biblkova^, Alexander Postnikov^ Lubov' Popova"* ''Dept. of Geology, Lund University, Solvegatan 12, 22362 Lund, Sweden (Svetlana.Bogdanova@geol.lu.se) T S R C , University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia Vernadsky Institute of Geochemistry and Analytical Chemistry, RAS, Kosygin street 19, Russia "^Gubkin State University of Oil and Gas, Leninsky pr. 65, 117296 Moscow, Russia

Volgo-Uralia (VU) is an enigmatic crustal segment that occupies the eastern third of the East European Craton (EEC). Its crystalline crust is buried completely beneath thick Proterozoic to Phanerozoic sedimentary cover. Information on the crystalline basement is therefore solely based on geophysical data and a great number of drill cores, particularly numerous because of the high oil and gas potential of the region. VU crust is dominated by many separate belts of -2.9-2.7 Ga metasedimentary and metaigneous granulites. There are also greenstone sequences with komatiitic volcanics. Associated with these belts are NE-SW to E-W trending zones of strong shearing and mylonitisation. Superimposed on the Archaean structural pattern are large domal structures featuring circular concentric magnetic and gravity anomalies, and intense retrograde reworking of the granulites. This is in marked contrast to tectonic style of Fennoscandia and Sarmatia, two other crustal segments of the EEC. Palaeoproterozoic metasedimentary rocks with rare metavolcanic rocks and numerous granitoids have been intersected in the interior parts of the domes. Apart from that, large areas of Palaeoproterozoic turbiditic metapelitic mica schists, silts, sandstones and carbonaceous shales occur along some margins of VU. Whether reworking of the Archaean granulites and formation of domal uplifts was Archaean or Palaeoproterozoic is unresolved, mainly due to the absence of reliable isotopic ages for the overprinting metamorphism. To settle the issue, zircons from twelve samples of the most representative VU rocks were subjected to SHRIMP analysis. Among these rocks were meta-igneous and metasedimentary granulites as well as granitoids both from granulite facies areas and from zones of strong retrograde reworking of the Archaean granulites. The newly obtained SHRIMP and available TIMS zircon data confirm Meso and Neoarchaean ages for the protoliths and define the following

118

tectonothermal events: (1) Regional granulite and amphibolite facies metamorphism, emplacement of granitoids, and deformation at 2.72-2.71 Ga; (2) Granitoid magmatism of uncertain tectonic setting at 2.67-2.65 Ga; (3) Intrusion of post-collisional monzonitic granitoids, gabbro-anorthosites and anorthosites at --2.60 Ga which form large elongated intrusions associated with major thrust zones; (4) sedimentation between 2.30 and 2.10 Ga, followed by high-grade metamorphism and anatexis at 2.08 Ga; and (5) metamorphism between 1.90 and 1.80 Ga. Thus, Archaean crust in VU was strongly reworked during multiple events between 2.1 and 1.8 Ga, i.e. analogous to similar events in Sarmatia. This suggests that after 2.1 Ga the Sarmatia and VU segments shared a common evolution. The doming, which occurs only in the Archaean crust of VU, is interpreted to have occurred prior to reworking and assembling with Sarmatia between 2.1 and 1.8 Ga. Widespread continental sedimentation recognised in VU between 2.3 and 2.1 Ga may be a result of doming, which is potentially linked to convergent processes that eventually led to collision with Sarmatia around 2.1 Ga. Mantle plume tectonics may also have played a role in the large-scale doming between 2.6 and 2.1 Ga. Geochronology of zircon from small drill core samples of VU highlight the complexity of zircon growth in multiply reworked granulite facies terranes. All thermal events in VU were accompanied by new zircon growth, complicating traditional zircon geochronology. Our U-Pb SHRIMP analyses clearly demonstrate the presence of several generations of zircon growth in many samples, and widespread zircon inheritance, making unequivocal determination of protolith ages troublesome. This complexity, combined with lack of exposure, which limits any study to small core samples, renders geochronological work on the VU Shield difficult.

Siipercontinents and Earth Evolution Symposium 2005

CRUSTAL STRUCTURE OF THE URALIDE OROGEN D Brown\ R Carbonell\ C Juhlin^ ^Institute de Ciencias de la Tierra "Jaume Almera", CSIC, Barcelona, Spain (dbrown@ija.csic.es) ^Departnnent of Earth Sciences, Uppsala University, Uppsala, Sweden

Geology and geophysical data show that the Uralides preserves a bivergent collisional architecture and confirms the existence of a crustal root along the orogen. Reflection and refraction seismic data show that the crust thickens eastward from - 4 0 to - 5 2 km in the East European Craton (EEC) to 50-55 km across the volcanic arcs, before thinning to 4 0 45 km in the easternmost part of the orogen. In the South (URSEIS) and Middle (ESRU) Urals, EEC crust is imaged as subhorizontal to eastdipping reflectivity that can be related to its Uralide and older orogenic events. In both data sets, the EEC extends eastward beneath the Magnitogorsk-Tagil zone. The Magnitogorsk (URSEIS)-Tagil (ESRU) zone displays moderate to weak upper crustal reflectivity, but the middle and lower crust reflectivity is diffuse for the Magnitogorsk arc, but quite strong for Tagil. The Moho is not imaged beneath the Magnitogorsk zone, but is fairly sharp beneath the Tagil zone. East of the arc complexes, upper and middle crust is imaged as clouds of diffuse reflectivity interspersed with, or cut by sharp, mainly westdipping reflections that extend from middle into lower crust where it appears to merge with the Moho. In the ESRU data it is characterised by abundant lower crustal reflectivity. In both the URSEIS and ESRU data the Trans Uralian zone dips westward beneath the East Uralian zone. The velocity structure of the Uralide crust is determined along the URSEIS transect. Upper crustal Vp reaches 6.3 kms'\ and Vs reaches 3.9 kms"\ with the higher values in the Magnitogorsk arc. In the middle and lower crust, Vp ranges from 6.5 to 6.8 kms"\ reaching 7.1 kms"^ above the Moho in the central and eastern part. Vs is 3.7 to 3.9 kms"\ increasing to 3.9-4.0 kms'^ at the Moho. The Moho is marked by an increase in Vp to >8.0 kms'"" and Vs >4.6 kms ''. Uralide heat flow density (HFD) is characterised by a minimum along the central part of the orogen, with values as low as 10 mW m"^, reaching up to 60 mW m"^ on either side. The short wavelength of the HFD anomaly suggests a shallow origin for the minimum, although heat production data determined from surface samples is too high to allow it to be simulated. This implies more ultramafic material at depth beneath the volcanic arcs than below

the EEC crust. Alternatively, the HFD minimum is due to propagation of ground surface temperature changes to depth as a result of palaeoclimatic changes. Modeling assuming the first alternative suggests that the Uralides thermal structure is characterised by relatively flat geotherms, with a Moho temperature of ~600°C. This does not suggest a very cold root. The Bouguer anomaly in the South and Middle Urals is characterised by a low of -60 to 45 mGal across the EEC, indicating upper and middle crustal densities of -2.80 and 2.90 g/cm^ and lower crustal densities of 2.98 and 3.02 g/cm^. There is an abrupt increase in the Bouguer anomaly to - 0 to -40 mGal in the Magnitogorsk-Tagil zone, falling to about -70 and -40 mGal in the East Uralian zone, and about -30 to -10 mGal across the East Uralian zone. This indicates an upper crustal density between 2.71 and 2.80 g/cm , a middle crustal density between 2.92 and 2.95 g/cm^ and a lower crustal density between 2.98 and 3.07 g/cml The upper mantle has a density of 3.34 g/cm^. The magnetic signature of the South and Middle Urals is characterised by short wavelength features, with a long wavelength low that reflects the magnetic character of the EEC. In the URSEIS transect, magnetic crystalline basement is truncated - 5 0 km west of the Main Uralian fault and magnetic susceptibilities of the rocks are 0 and 1.5 A/m to the east and west of this truncation, respectively. Magnetic susceptibilities of the upper crust are 0.25 to 0.5 A/m, with areas reaching 2 and even 2.5 A/m locally. The middle and lower crust has magnetic susceptibilities of 0.5 and 1.5 A/m. Petrophysical modeling along the URSEIS transect shows clear differences between the composition of the old continental crustal nucleus of the EEC and newly-added crust of the accreted arc terranes to the east. The EEC crust is more felsic than that of the Magnitogorsk and East Uralian zones, and the latter two have a lowermost crust whose characteristics indicate a high garnet content (mafic garnet granulite) and/or the presence of hornblendite. The overall composition of the arc terranes is basaltic. Physical properties data suggest that eclogite is not present in the lower crust, or if present it exists in such small amounts that it is below the resolution of the data set.

119

Supercontinents and Earth Evolution Symposium 2005

TRENDS IN SEDIMENTARY PROVENANCE ALONG EAST LAURENTIA, BALTICA, AND AMAZONIA DURING ASSEMBLY AND BREAKUP OF RODINIA Peter A Cawood\ Alexander A Nemchin^, Rob Strachan^, Tony Prave"*, J Orestes S Santos^, Maarten Krabbendam® ^TSRC, School of Earth and Geographical Sciences, University of Western Australia, 35 Stirling Hwy, Crawley, WA 6009, Australia (pcawood@tsrc.uwa.edu.au) ^WASM, Curtin University, Locked Bag 22, Kalgoorlie, 6433, WA, Australia ^School of Earth & Environmental Sciences, University of Portsmouth, Portsmouth, P01 3QL, UK "^School of Geography and Geosciences, University of St Andrews, St Andrews, KY16 9AL, Scotland ^Brazilian Geological Survey, Rua Banco Provincia 105 Santa Teresa, 90840-030 Porto Alegre, Brazil ^British Geological Survey, Murchison House, West Mains Road, Edinburgh, EH9 3LA, Scotland

Mesoproterozoic to Phanerozoic siliciclastic units and their detrital zircon age signatures from East Laurentia, Baltica and Amazonia provide a record of Rodinia assembly and dispersal. Four lithotectonic assemblages, based on age relations with respect to Rodinia are recognized: 1) basins which predate Rodinian assembly; 2) basins formed synchronously with collisional orogeny and Rodinian assembly; 3) intracratonic basins developed within an assembled Rodinia; and 4) late to post-Rodinian basins related to supercontinent breakup. Assemblage 1 - consists of Mesoproterozoic sedimentary basins developed along the margin of East Laurentia-Baltica and also independently along the western Amazonia margin. In East Laurentia-Baltica, most of these basins are associated with an active convergent margin. In addition, inboard of the Grenville-Sveconorwegian Orogen is the Mesoproterozoic Stoer Group (NW Scotland), interpreted as part of a rift system inboard of the active margin. Detrital zircons from the Stoer Group are dominated by Archaean detritus with minor Palaeoproterozoic and no Mesoproterozoic detritus, whereas those from the active margin show a range of ages from latest Archean to latest Mesoproterozoic. Assemblage 2 - accretionary margin tectonism along East Laurentia-Baltica terminated in the late Mesoproterozoic through continent-continent collision with the inferred Amazonia craton, resulting in GrenvilleSveconorwegian-Sunsas orogenesis. Basins developed during collisional orogenesis reveal an early history between around 1190 to 1100 Ma, typically associated with bimodal igneous activity, followed by deposition of a cover sequence between 1120 and 1010 Ma. In addition, sedimentary successions accumulated on the Laurentian foreland synchronously with Grenville orogenesis and include the Middle Run 120

Formation and the Torridon and Sleat Groups, as well as the Midcontinent Rift System. Detrital zircons from these units range from Archean to late Mesoproterozoic and indicate derivation from the orogenic hinterland. Assemblage 3 - intracratonic extension and subsidence within Rodinia between about 1000700 Ma is preserved in the Krummedal supracrustal succession of East Greenland, the Moine Supergroup of Scotland, Brennevinsfjorden Group of Svalbard, and Mount Rogers Formation of the US Appalachians. In the British and Greenland Caledonides, pulses of extension alternate with crustal thickening and orogenesis. These units are characterized by late Paleo- and Mesoproterozoic zircon detritus with Archean age zircons absent or rare. Assemblage 4 - consists of late Neoproterozoic to early Paleozoic successions that accumulated during Rodinia breakup and include Neoproterozoic to Cambrian mixed siliciclasticcarbonate sequences that extend from the southeast US to Greenland The detrital zircon record of the group is characterized by late Archean, late Paleoproterozoic, and Mesoproterozoic detritus, except for the lower Dalradian (Grampian Group) and Eleonore Bay successions, which are similar to Assemblage 3, and lack Archean detritus. These latter units may have recycled detritus from Assemblage 3. The significant feature of the detrital zircon record with respect to the paleogeographic evolution of the region is the incoming of late Mesoproterozoic detritus in Assemblage 2. The persistence of this age component in younger assemblages indicates the longevity of the collisional orogenic welt. The absence of Archean detritus from Assemblage 3, but their location adjacent to Archean basement indicates that this welt dominated palaeogeography and sediment supply.

Siipercontinents and Earth Evolution Symposium 2005

DEPOSITIONAL AGE AND PROVENANCE RECORD OF ULTRA-HIGH TEMPERATURE METASEDIMENT PROTOLITHS OF SOUTHERN MADAGASCAR AND INDIA A l a n S Collins^ ^ Pete D K i n n y ^ M S a n t o s h ^ T R a z a k a m a n a n a ^ Continental Evolution Research Group, Geology and Geophysics, School of Earth and Environnnental Sciences, University of Adelaide, Adelaide, SA 5005 Australia (alan.collins@adelaide.edu.au) ^Tectonics Special Research Centre, University of Western Australia, Crawley, WA 6009, Australia ^Tectonics Special Research Centre, Department of Applied Geology, Curtin University of Technology, PO Box U1987, Perth, WA 6845, Australia "^Faculty of Science, Kochi University, Akebono-cho 2-5-1, Kochi 780-8520, Japan ^Departement des Sciences de la Terre, Universite de Toliara, Toliara, Madagascar

INTRODUCTION South India and southern Madagascar are dominantly made up of ultra-high temperature metasedimentary rocks that were metamorphosed in the Ediacaran-Cambrian orogenies that formed as Gondwana amalgamated. These metasediments lie in a key part of the Gondwana coalition, close to three of the main Gondwana-forming suture zones, the eastern Malagasy Betsimisaraka suture, a probable suture to the west of Madagascar that may crop out in southern Kenya, NW Tanzania, and the Kuunga suture to the south. As such, the information contained in the U-Pb ages of detrital zircons have considerable importance in both determining the age of deposition (and therefore basin evolution) and delineating possible source regions. These data can help to unravel the Neoproterozoic prior to Gondwana amalgamation. MALAGASY SAMPLES Southern Madagascar is divided into three main lithofacies units: the Vohibory Series, the Graphite Series and the Androyen Series. Sample M03-06 is from the Vohibory Series that lies in the far southwest of Madagascar and is associated with many meta-basaltic rocks. Detrital zircon cores from Vohibory samples have 207pj^/206p^ ages that cluster at -850 Ma suggesting that this rock was deposited after -850 Ma and was dominantly sourced from rocks of this age. We suggest that these rocks may have formed during rifting of Azania (a microcontinent involving central Madagascar parts of Somalia, Ethiopia and Yemen) from eastern Africa.

Rocks from within the Androyen and Graphite Series preserve considerably different detrital zircon age spectra with many discordant analyses giving Palaeoproterozoic ages with maxima between -1.8 and 2.2 Ga. These are similar to detrital profiles reported from the probable Palaeoproterozoic Itremo Group in central Madagascar. INDIAN SAMPLES U-Pb data from zircon cores in the Southern Granulite Terrane of India are commonly discordant and reflect the fact that the host rocks have experienced extremely high metamorphic temperatures-close to current estimates for Pb diffusion in zircon. ages of zircon cores that are within 10% of concordance again show probability density distribution maxima between 1.8 and 2.2 Ga. There are also a few analyses that preserve near-concordant Mesoand Neoproterozoic ages that suggest that at least some of these metasediment protoliths were deposited in the Neoproterozoic. DISCUSSION Sensitive High Resolution Ion MicroProbe (SHRIMP) analyses of detrital zircons from southern Madagascar and India show that large regions of the East African Orogen are composed of Neoproterozoic sedimentary rocks that were deformed and metamorphosed in the Neoproterozoic amalgamation of Gondwana. These data also show that many sedimentary rocks of possible Palaeoproterozoic age in both southern Madagascar and India are sourced from a region rich in 1.8-2.2 Ga rocks.

121

Supercontinents and Earth Evolution Symposium 2005

IN SITU U-Th-Pb MICROPROBE DATING OF EARLY DIAGENETIC MONAZITE IN NEOPROTEROZOIC BLACK SHALES OF THE WESTERN UNITED STATES Laura J Crossev\ Michael L Williams^ Karl E Karlstrom\ Michael Jercinovic^ John D Bloch\ Carol M Dehler^ ^Dept. of Earth & Planetary Sciences, Univ. of New Mexico, Albuquerque, NM 87131, USA (lcrossey@unm.edu) Dept. of Geosciences, University of Massachusetts, Amherst, MA 01003, USA ^Dept. of Geology, Utah State University, Logan, UT 84332, USA

U-Th-Pb in situ dating of monazite via the electron microprobe offers a rich potential for directly dating sedimentary sequences, as well as understanding provenance (detrital monazite), resolving sedimentation rate, calibrating chemostratigraphic oscillations (e.g., C-isotope records), and understanding linkages between biologic systems and other geologic processes. This paper describes the occurrence of monazite, and develops chemical criteria for distinguishing authigenic from detrital monazite. We report new dates from Neoproterozoic shales of the western U.S. Black shales contain two populations of monazite. One population consists of rounded grains exhibiting zoning and yielding dates of 1.74-1.65 Ga, consistent with derivation of detrital grains from the adjacent, well-dated basement. A second population consists of tiny (1-5 micron) grains showing delicate textures that we interpret as early

122

diagenetic authigenic products. In the Chuar Group, for example, these yield ages of -743 ± 13, compatible with a known 742 ± 6 Ma age from an ash bed higher (Walcott Mbr.) in the Chuar section, and with regional correlations. These results provide constraints on sedimentation rate, and the timing of three large magnitude C-isotope shifts (-10 permil) during the depositional interval that just predates the first (Sturtian) low-latitude global (?) glaciation. Textural and chemical characterization of these minute monazites, and their association with well-preserved microfossils, shows the operation of a P cycle not unlike present processes in anoxic mudstones. This new ability to directly date sedimentary rocks, improved dramatically with smaller electron beam diameters in the current generation of monazite-capable electron microprobes, will advance our understanding of sedimentary systems of all ages.

Siipercontinents and Earth Evolution Symposium 2005

TECTONICS OF EASTERN AFRICA: THE MOZAMBIQUE BELT AND ITS REGIONAL SETTING Huntiv NC Cutten\ Simon P Johnson^ ^Tectonics Special Research Centre, School of Earth and Geographical Sciences, University of Western Australia, 35 Stirling Highway, Crawley, WA 6009 Australia (hcutten@tsrc.uwa.edu.au) ^Institute for Research on Earth Evolution, Japan Agency for Marine-Earth Science and Technology, 2-15 Natsushima-cho, Yokosuka, Kanagawa 237-0061, Japan

Developnnent of the East Africa Database, using ArcView GIS and Microsoft Access linked by software Av2Ax, has provided a research tool to archive new and published tectonically significant data from the Mozambique Belt in eastern Africa. The database provides an objective tool to test tectonic models against a large and diverse dataset, and is a new approach in tectonic synthesis. Application to the Mozambique Belt and formulation of tectonic models for the collision leading to the formation of Gondwana has been a test case and the database could be applied to more extensive regions with much larger datasets. The Mozambique Belt is part of the East African Orogen (EAO), which extends from the Arabian Nubian Shield in the north to as far south as east Antarctica. From Kenya to Tanzania an exhumation gradient is observed with supracrustal rocks in the north, and deep crustal levels of EAO in the south, exposed by erosion following continent-continent collision and isostatic rebound from crustal thickening. Data used in this synthesis include: lithology, metamorphic grade, igneous, metamorphic and detrital geochronology, geochemistry, P-T conditions, P-T-t path and structural geology. This data delineates the Mozambique Belt as two terranes. The Western Mozambique Belt (WMB), comprising upper amphibolite-grade gneisses of the tectonised Tanzania Craton, with emplacement ages of 2970 Ma to 2648 Ma, and the Eastern Granulites (EG) with high grade, arcderived lithologies, and emplacement ages of 841 Ma to 632 Ma. Both terranes include intercalated Neoproterozoic metasediments of the West and East Mozambique Basin respectively. Nd isotopics of WMB show Archaean model ages and sNd(t) similar to the Tanzania Craton. The EG show model ages generally ranging from 1300 Ma to 900 Ma and sNd(t) from 5.36 to 0.38 indicating mixing of juvenile arc material with some older crustal

melts. Peak metamorphism in both WMB and EG centres on two age clusters of -640 Ma and -550 Ma. P-T-t paths were reported to be clockwise in WMB with peak metamorphic conditions of -12-13 kbar and 750-800°C, dated at -640 Ma, and attributed to continent-continent collision. From the EG an anti-clockwise P-T-t path was reported with -640 Ma peak metamorphic conditions of 9.5-11 kbar and 810 ± 40°C. This anti-clockwise P-T-t path was attributed to magmatic underplating and upper crustal magma loading in these arc-derived rocks, preceding continent-continent collision. It is suggested here that the early part of the anticlockwise P-T-t path reflects conditions in the base of the arc and that peak metamorphism at -640 Ma in both WMB and EG reflect their juxtaposition by continent-continent collision. The contact between WMB and EG appears to be a low-angle east-dipping thrust and EG has been further subdivided into three thrust sheets: the Anorthosite thrust sheet, and Typical Eastern Granulite thrust sheet, each representing different levels of a Neoproterozoic continental arc, and the Accretionary Wedge thrust sheet, representing continental shelf deposition outboard of the arc. The EG have previously been identified as a continental arc formed above a west-dipping subduction margin on the flank of the Tanzania Craton. It is suggested here that available data are more consistent with an east-dipping subduction margin on the flank of a continental block (likely Madagascar) well away from the African margin and that the arc rocks were subsequently thrust over the African margin with closure of the 'western' Mozambique Ocean and continent-continent collision. A later -550 Ma peak metamorphic event is also recorded throughout the Mozambique Belt and reflects subsequent collision of additional continental blocks.

123

Supercontinents and Earth Evolution Symposium 2005

PALEOMAGNETISM OF BOREHOLE GKP-01 OF THE AGOURON-GRIQUALAND PALEOPROTEROZOIC DRILLING PROJECT, SOUTH AFRICA M O d e K o c k \ NJ B e u k e s \ DAD E v a n s ^ I H i l b u r n ^ J L K i r s c h v i n k ^ E Rose^ ^Department of Geology, University of Johannesburg (Aucklandpark Kingsway Campus), PO Box 524, Aucklandpark, 2006, South Africa (glgy5@na.rau.ac.za) ^Department of Geology and Geophysics, Yale University, PO Box 208109, New Haven, CT 065208109, USA ^California Institute of Technology, 1200 E. California Blvd., Pasadena, CA 91125, USA

INTRODUCTION The Agouron-Griqualand Paleoproterozoic Drilling Project was initiated to study the interval from the Archean/Proterozoic boundary interval at 2.6-2.5 Ga to the critical oxygenation and snowball episodes around 2.2 Ga as recorded in the western part of the Kaapvaal craton. One goal of this project is to establish a highresolution chronostratigraphic framework for the Campbellrand carbonate platform and the Kuruman Iron Formation by paleomagnetic techniques and other methods. RESULTS After removal (at relatively low levels of demagnetization) of a drilling induced magnetization (DIM) and present-field viscous remanent magnetization (VRM) that presumably reside in multi-domain magnetite, a further three magnetic components were revealed. The first of these components (named INT) was recorded by all lithologies sampled and is carried in an iron sulfide, presumably pyrrhotite. A few samples from all the lithologies, except for the tuffs recorded a second component (HI), which was revealed by either great circle arcs away from the INT position towards a very shallow NNW direction or as poorly defined nonzero seeking, upwards directed shallow lines. The third component (H2) was revealed at high temperature demagnetization steps either as great circle arcs away from the INT or HI components or as stable end points of demagnetization trajectories. The H2 component is directed steeply downwards towards the southwest.

124

DISCUSSION The INT direction is similar to those of ca. 2050Ma rocks of the Kaapvaal craton and is interpreted to represent fluid-activated chemical remanent magnetization (ChRM) possibly associated with the intrusion of the Bushveld Complex or other events of that age. The HI pole, although showing some similarities to Paleoproterozoic poles, display remarkable overlap with the 1240 ± 20 Ma Namaqua Eastern Zone pole from the Namaqua metamorphic belt (extension of the Gascoyne belt of Western Australia). The H2 pole does not bear any direct resemblance to other 2220 to ca. 1750 Ma Kaapvaal craton poles, but it does plot close to the younger of these poles. The Kaapvaal craton is bordered to the west by the older Kheis terrane and the ca. 1100 Ma Namaqua belt as mentioned above. Deformation in the Kheis terrane can be largely be explained by two episodes at 2100 and 1100 Ma. It appears as if the Campbellrand carbonate escaped these events but possibly suffered pervasive remagnetization at ca. 2050 Ma as discussed above. The Paleo- to Mesoproterozoic age for the H2 component might be representative of the initiation of subduction to the west of the Kaapvaal craton that preceded the collisional event responsible for the Namaqua belt. CONCLUSION None of the revealed magnetic components can be shown to be primary, but the paleomagnetic work can constrain the nature and timing of hydrothermal fluids potentially responsible for resetting other geochemical and isotopic records. So far the results implicate the likelihood of a substantial alteration event at ca. 2050 Ma.

Siipercontinents and Earth Evolution Symposium 2005

ORIGIN OF GONDWANA MARGIN PERMO-JURASSIC GRANITE-RHYOLITE PROVINCES: THE SOUTH AMERICAN CHOIYOI EXAMPLE Gabriela V Depine, Suzanne M Kay, Jason Phipps Morgan EAS and INSTOC, Cornell University, Snee Hall, Ithaca, NY 14853, USA (gvd2@cornell.edu)

A broad range of special geologic processes occurs during a supercontinent stage on the Earth. Some are characteristic of continental accretion, others of dispersion. In particular, large igneous provinces have been cited as indicators of the beginning of supercontinent break-up. Typically, these provinces are of basaltic composition, erupt over a short period, and are thought to be mantle-related. There are also cases of granite-rhyolite provinces through Earth's history, with Permo-Jurassic graniterhyolite terranes being the youngest, in which rhyolitic ignimbrite and granite dominate over andesite and basalt. They are called provinces because their voluminous outputs are on the order of flood basalts (>350,000 km^ for the Choiyoi province discussed here). They differ from mafic provinces in having a longer duration. There is increasing agreement in assigning the generation of these rhyolites to lower crustal melting, but fundamental thermodynamic and geologic questions exist as to how such gigantic amounts of lower crustal melt can be generated. Extensive outcrops of Permian to Jurassic granite and rhyolite occur along the Gondwana margin in Peru, Bolivia, Chile, Argentina, Antarctica and eastern Australia where they form a belt that coincides with regions of Paleozoic terrane accretion. Different regions or provinces are called by different names. Their formation was diachronous, beginning in Peru in the Permo-Triassic and ending in eastern Australia in the Late Jurassic-Early Cretaceous. Although the provinces are bimodal, the predominant rock types are dacite to high-Si rhyolite, with basalt and mafic andesite occurring only locally and commonly postdating large silicic eruptions. Compositions of the silicic rocks range from S- to I- to A-type. These differences are better explained by variations in the underlying crust rather than by fundamental differences in melt generation processes. The South American Choiyoi province extends from 20°S to 42°S. Paleozoic subduction and terrane collision along this part of the western Gondwana margin ended in the early Permian with a transpressional event called the San Rafael orogeny. Voluminous high-Si Choiyoi

magmas were emplaced into these deformed rocks as a mildly extensional tectonic regime developed in response to mechanical relaxation. This regime could also have been affected by counterclockwise rotation of western Gondwana relative to North America. Ignimbritic sequences >2 km thick and shallow granitic intrusive rocks dominate the Choiyoi province. These mostly highly silicic rocks (>73 wt% Si02) have low LaA'b and large negative Eu anomalies consistent with eruption through a thin crust in an extensional regime. Scarce basalts occur among the younger rocks in a true rift setting (e.g. Cuyo rift). The basalts analyzed have intraplate-like chemistry (e.g., Triassic Uspallata basalts in Argentina). The silicic rocks can have an arc-like chemistry, but this does not necessarily indicate a subductionrelated origin as they can be melts of crust formed and modified by subduction processes. Processes proposed to explain the genesis of granite-rhyolite provinces like the Choiyoi all call for basaltic underplating and crustal melting. The problem is to explain the tectonic setting of the mafic magmas and the heat to produce the large volume silicic magmas. Processes proposed are: 1) Slab break-off and/or delamination of thickened crust caused asthenospheric upwelling that produced basalts that melted immature and hydrated crust. These events would follow the San Rafael orogenic event attributed to relative overriding of Gondwana over the oceanic plate or the possible collision of a terrane now largely removed by forearc subduction erosion. A problem is why large quantities of similar rhyolites did not erupt after earlier Paleozoic terrane collisions. 2) Steepening of a formerly shallow subduction zone lead to extensive melting of hydrated continental lithosphere (ignimbrite flare-up). Problems come from comparisons with slab-steepening models for the central Andean plateau in which largely andesitic to dacitic (not rhyolitic) melts are generated and erupt in a contractional (not extensional) setting in a thickening crust. 3) A high geotherm produced by a stationary position of Gondwana over the mantle and cessation of subduction leading to plume generation.

125

Supercontinents and Earth Evolution Symposium 2005

CONTINENTAL FREEBOARD AND GEOLOGICAL EVOLUTION: A PRECAMBRIAN PERSPECTIVE Patrick G Eriksson\ Raiat Mazumder^, Octavian Catuneanu^ ^Dept. of Geology, University of Pretoria, Pretoria 0002, South Africa (perikssn@nsnper1.up.ac.za) ^Dept. of Geology, Asutosh College, Kolkata 700 026, India ^Dept. of Earth and Atmospheric Sciences, University of Alberta, Edmonton, Alberta T6G 2E3, Canada

Continental crustal growth rates, crustal volumes/thicknesses, and continental freeboard are considered to be intimately related; however none are easy to quantify, least of all in the Precambrian. Additionally, crustal volumes/ thicknesses and isostasy are related to changing mantle heat and the concomitant variation in thickness of ocean crust. Another set of variables would be the interaction of plate tectonics (and the onset of plate tectonics, a contentious issue in itself) and mantle plumes/superplumes. Further complication is provided by the ranges of freeboard elevations and eustatic sea level changes overlapping in scale, and sharing common genetic causes. The "constant freeboard model" suggests that the net results of the interaction of this complex set of variables has been that continental elevation relative to sea level, has remained essentially similar, on a

126

global average basis, over much of geological time, assuming an essentially constant ocean volume since c. 3.8 Ga. Application of this model to the Precambrian geological record suggests that sigmoidal crustal growth rate models are probably the most reJiable. A conundrum of the constant freeboard model would be that freeboard would tend to become lowered towards mean sea level for any specific craton over time; this "freeboard-equilibrium" would be interrupted for a particular craton by active geodynamic processes resulting from the interplay of plate tectonics and mantle thermal instabilities. In this paper we test the validity of this thesis of alternating approximate freeboardequilibrium and active geodynamics, by examining the volcano-sedimentary record of a number of cratons: Kaapvaal, Singhbhum and Pilbara.

Supercontinents and Earth Evolution Symposium 2005

RODINIA RADICALLY REVISED? David Evans Department of Geology and Geophysics, Yale University, PO Box 208109, New Haven, CT, 06520, USA (dai.evans@yale.edu)

A new configuration of the Rodinia supercontinent differs dramatically from all previous reconstructions. Based primarily on palaeomagnetic data with important constraints from global tectonostratigraphy, the most notable cratonic juxtapositions are: (1) Arctic North America adjacent to Conqo-Sao Francisco at its Damaride margin; ANACONDA; (2) Baltica's Norwegian margin adjacent to SE Greenland and Newfoundland; (3) Australia/Mawsonland inverted such that the Queensland margin joins Baltica's Caspian-South Uralian margin; (4) eastern India adjacent to NW Australia; (5) South China's Sichuan margin joining NW India; (6) Kalahari's NW margin aligned with Mawsonland's Vostok margin; (7) Siberia as either a separated block or possibly connected along its Igarka margin to W India. Amazonia and West Africa are reconstructed according to two options. The first is a standard location of Amazonia with its Sunsas margin colliding against Llano with subsequent sinistral displacement along the Grenville orogen; West Africa occupies a position such that its Pharuside margin faces the Cordilleran margin of Laurentia. The second option for these blocks is a rotated placement of Amazonia (about the 1200 Ma Nova Floresta pole) so that the Sunsas belt is a continuation, through various Central- and South American terranes, of the Grenville-Llano belt. In this alternative, the Cordilleran margin of Laurentia links with Brasiliano and Pharuside belts of the West Gondwanaland cratons (COBRA), and all of these blocks are presumed to have been joined as early as Palaeoproterozoic time. This novel Rodinia requires two non-standard interpretations of the global palaeomagnetic database. First, dyke swarms in Bahia, Brazil, must be slightly older than their existing Ar/Ar

ages would suggest: ca. 1100 Ma rather than 1080-1010 Ma. Second, a large APW loop at ca. 800 Ma, currently identified among highquality data from India, South China, and Congo, must be included as a previously unrecognised excursion in the APWPs of other blocks, including Laurentia. The magnitude and rate of this loop would suggest true polar wander as its underlying cause. There are many regional tectonic features that must be viewed from this global perspective if this radically revised Rodinia is to be validated. The most intriguing new connections include: (1) identification of the "other side" of the giant Mackenzie radiating dyke swarm, as the precisely coeval Kabanga-Musongati and Niquelandia complexes respectively in the eastern Congo and western Sao Francisco blocks; (2) the Roan and Amundsen evaporites at ca. 800 Ma as part of a single lithospheric sag basin; (3) recognition of a Baltica-Australia collision as recorded in the Beloretzk and Cape River terranes; (4) an extensive Rodinia-forming orogen linking the Eastern Ghats to the Namaqua belt across a broad internal zone now covered by Antarctic ice; (5) continuation of the Mojavia province into Transamazonian terranes in the COBRA option; and (6) identification of the Sveconorwegian-Grenville-Llano-Sunsas belt as a Pacific-rim type of subduction-accretion orogen rather than two-continent collision, also as part of the COBRA scenario. Perhaps the most important contribution of this alternative Rodinia, rather than simply testing its validity, is the way in which it requires us to ask new questions of the world's Proterozoic geology, in many places (e.g.. West Africa) traditionally ignored as marginal and irrelevant in previous Rodinia reconstructions.

127

Supercontinents and Earth Evolution Symposium 2005

CONTINENTAL ACCRETION IN THE NW SIERRAS PAMPEANAS OF ARGENTINA L A D Fernandes\ CC Porcher\ GI Vujovich^, CJ Chernicoff^ ^CNPq & Centre de Petrologia e Geoquimica/UFRGS PO Box 15065- 91501-970-POA-RS, Brazil ^CONICET & Laboratorio de Tectonica Andina/UBA, Argentina

Basement rocks comprising orthoand paragneisses whose tectono-metamorphic evolution is poorly known, are exposed in the Sierras de Umango, Maz-Espinal and Las Ramaditas, in the northwest of the La Rioja Province, Argentina. These units were included in the Maz, El Taco, El Zaino complexes, as well as the Tambillos Metamorphites and would (i) belong to the northern end of the Cuyania Terrane, a microcontinent derived from Laurentia that was amalgamated to the margin of Gondwana during the Ordovician, or (ii) be part of the western margin of the supercontinent. Geochemical studies, trace and REE elements and Nd-Sm model ages of these rocks allowed the recognition of several episodes of crustal accretion. The oldest one occurred at ca. 2.2 Ga in a arc/backarc environment and is represented by rocks exposed along the eastern segment of the Sierra de Maz. The following episode of crustal accretion that formed rocks in this region was at ca. 1.4 Ga and is registered by tonalites emplaced in an extensional environment that crop out in the western flank of the Sierra del Espinal. In the Sierra de Umango, an arc/backarc sequence marks an episode of crustal accretion, probably during the Grenvillian cycle (ca. 1.1 Ga). The last episode of crustal accretion detected is represented by 800 Ma alkaline volcanism in the Sierra de Umango that could represent the first stage of break-up of the Rodinia supercontinent. The metamorphic grade

128

reached by these rocks is mostly represented by fabrics with mineral assemblages of intermediate to high pressures and high temperatures, typical of collisional environments. The oldest rockforming fabrics tectono-metamorphic episode recognised is of Mesoproterozoic age (ca. 1.04 to 0.969 Ga, garnet-WR Sm/Nd) being registered by metapelites from Maz Complex that attained temperatures of 650°C at 6.3 kbar. A younger metamorphic event, (463 Ma, garnet-WR Sm/Nd) is verified in metatonalites intrusive into these metapelites. The final event was recognised in metasediments from El Taco Complex (301 Ma, garnet-WR Sm/Nd). Peak metamorphic conditions of this event, probably registering the last major tectonic episode that affected rocks of this area were 868°C and 9.8 kbar. Geophysical evidence indicates the presence of extensive WNW-oriented lineaments that separate basements blocks of different magnetic and gravimetric signatures that might represent ancient Grenville-age suture zones. The Valle Fertil Lineament that runs between sierras de Umango and Maz-Espinal, is interpreted as marking the eastern boundary of the Cuyania Terrane. On the other hand, isotopic data as well as the contrasting history of tectonometamorphic events, as determined for both of these segments of the NW Sierras Pampeanas, suggests that Sierras de Maz-Espinal and Las Ramaditas could belong to the western margin of Gondwana.

Supercontinents and Earth Evolution Symposium 2005

THREE STAGES OF MESOPROTEROZOIC METAMORPHISM DURING CRUSTAL SHORTENING IN THE ALBANY-FRASER OROGEN OF SW AUSTRALIA Ian CW Fitzsimons, Pete D Kinny Tectonics SRC, Department of Applied Geology, Curtin University of Technology, GPO Box U1987, Perth, WA 6845, Australia (l.Fitzsimons@curtin.edu.au )

The Mesoproterozoic Albany-Fraser Orogen of southwestern Australia has been divided into several tectonic units. The Northern Foreland is the southern margin of the Archaean Yilgarn Craton that was reworked during the development of the Albany-Fraser Orogen. It is bounded to the south by the allochthonous Biranup, Fraser and Nornalup Complexes dominated by granulite-facies orthogneiss and late-tectonic granite plutons. Lower-grade metasedimentary rocks overlie this gneissic basement, the best studied of which are the Mount Barren Group schists. Numerous U-Pb zircon dating studies of early-, syn-, and posttectonic granitoids in the basement complexes have revealed two phases of metamorphism and magmatism at 1345-1260 Ma and 1215-1140 Ma, each associated with a discrete episode of regional NW-SE directed shortening (Stages I and II of the Albany-Fraser tectonism). Most workers have regarded crustal thickening during regional contraction as the cause of both Stage I and II magmatism and metamorphism, but this has been challenged by recent 1200 Ma xenotime and monazite ages from amphibolite-facies schists of the Mount Barren Group. These ages have been interpreted as dating the final stages of peak metamorphism in the Albany-Fraser Orogen. This implies that metamorphism pre-dated major Stage II crustal thickening in the basement dated at 1190-1160 Ma, but immediately post-dated a 1210 Ma mafic dyke swarm emplaced across much of southwestern Australia. This has been used to argue that peak metamorphism in the Albany-Fraser Orogen was caused by mafic magmatism and not by compressional tectonics. Our detailed petrographic study of Mount Barren schists indicates that the 1200 Ma xenotime and monazite grew under greenschistfacies conditions sometime before peak metamorphism. We have also identified a 1030 Ma monazite population in a limited range of bulk compositions that is likely to reflect peak metamorphism in the Mount Barren Group during a third phase of regional shortening (Stage III of the Albany-Fraser tectonism).

No isotopic record of Stage I tectonism has been identified in the Mount Barren Group, suggesting that the Mount Barren Group was located some distance away from the basement complexes at 1300 Ma, or that it remained at a shallow crustal level. The Mount Barren Group does preserve evidence of metamorphism during the beginning of Stage II tectonism in the basement, but at lower grade conditions. The timing of this early metamorphism in the Mount Barren Group is consistent with it resulting from the thermal influence of mafic intrusions, and a mafic sill does constitute up to 30% of rock exposure in the Mount Barren Group. It is unlikely, however, that mafic intrusions caused high-grade Stage II metamorphism in the basement, given that mafic dykes in the basement were themselves metamorphosed during this event. Granulite-facies Stage II assemblages in the basement are associated with 1190-1160 Ma ductile shear zones developed during crustal thickening. Stage III tectonism appears only to have affected the schists of the Mount Barren Group and not the basement, presumably because the latter was recrystallized and dehydrated at high metamorphic grades during Stages I and II. The regional significance of 1030 Ma Stage III tectonism is unknown, although it involved NWdirected thrusting and the growth of kyanite-zone Barrovian mineral assemblages, consistent with a period of crustal thickening. It is apparent that the Albany-Fraser Orogen had a protracted 1350-1000 Ma tectonic history involving at least three discrete episodes of metamorphism associated with N to NW-directed thrusting and crustal thickening. The first of these is assumed to represent initial assembly of the orogen, but the causes of contractional reactivation at 1190-1160 Ma and 1030 Ma are unknown. The similar style of all three events, and the fact that they did not affect all units to the same degree, means that tectonic correlations in the Albany-Fraser Orogen cannot be based solely on structure or metamorphic grade and require careful isotope geochronology coupled with detailed petrographic study of dated phases.

129

Supercontinents and Earth Evolution Symposium 2005

SYN-AND POST-EXTENSIONAL OROGENESIS IN THE PALAEOPROTEROZOIC RIFT SEQUENCES OF BROKEN HILL AND MT ISA AND IMPLICATIONS FOR RECONSTRUCTIONS OF RODINIA G e o r g e Gibson^'^, Narelle N e u m a n n ^ Peter S o u t h g a t e ^ L a u r i e Hutton^'^ VmdCRC and ^Minerals Division, Geoscience Australia, GPO Box 378, Canberra, ACT 2601, Australia (George.Gibson@ga.gov.au) ^Geological Survey of Queensland, Dept. of Natural Resources & Mines, Indooroopilly, QLD 4068, Australia

Palaeoproterozoic rift sequences bordering the eastern margins of the Curnamona and North Australian cratons share a comnnon history of early Mesoproterozoic (1600-1590 Ma) orogenesis, culnninating in basin inversion, crustal thickening and LPHT metamorphism up to amphibolite or granulite fades (Clary and Mt Isa orogenies, respectively). Intense deformation accompanying these events has masked much of the original basin architecture and pre-1600 Ma crustal history although there is increasing evidence that basin formation in both the Broken Hill and Mt Isa regions was linked at depth to ductile deformation, bimodal magmatic intrusion, and an earlier phase of LPHT metamorphism. These relations are particularly evident in the Mt Isa Western Succession where LPHT amphibolite facies metamorphism related to syntectonic intrusion of the 1670 Ma Sybella Batholith overlapped in time with half-graben formation and deposition of the 1690-1670 Ma Prize Supersequence (including Surprise Creek Formation). Coarse clastic sedimentation, together with U-Pb zircon ages from bimodal volcanic rocks and granite immediately underlying the Surprise Creek Formation indicate that this phase of basin formation and rifting commenced no later than 1710 Ma. A detachment surface separates the 1670 Ma granites and their LPHT country rocks from the overlying and more weakly metamorphosed sedimentary basins, including post-rift sediments of the 1670-1590 Ma Isa Superbasin. A similar pattern of LPHT metamorphism related to 1670-1690 Ma bimodal magmatism is evident in basin sequences of comparable age (1710-1690 Ma) at Broken Hill (lower Willyama Supergroup), indicating formation in a common (although not necessarily contiguous) tectonic setting. Taken together, the Broken Hill and Mt Isa regions have a crustal structure and history that is strongly reminiscent of the Eritrea-Red

130

Sea area or Basin and Range Province in North America. An appropriate tectonic analogue for the Broken Hill and Mt Isa regions might therefore be an extensional environment in which initial LPHT metamorphism was linked to the formation and unroofing of metamorphic core complexes, and driven by increased heat flow and magmatic intrusion accompanying asthenospheric upwelling and thinning of the mantle lithosphere. Such a model has already been proposed for Broken Hill where early metamorphism in the inferred lower plate was associated with bimodal magmatism, anatexis, and formation of decompression textures in sillimanite- to granulite-grade mineral assemblages. As in the Mt Isa terrane, an extensional detachment (now folded) separates the lower plate rocks from an initially less intensely metamorphosed upper plate in which metamorphism followed an anticlockwise P-Ttime path and peaked during the 1600 Ma Clary orogeny. Matching of 1.7-1.60 Ga orogenic belts in eastern Australia and SW Laurentia is a critical component of several published reconstructions of Rodinia. However, while both Broken Hill and Mt Isa preserve a common record of 1.72-1.67 Ga intracontinental rifting and extension followed by major orogenesis at 1590-1600 Ma, crustal evolution in SW Laurentia throughout much of the Late Palaeoproterozoic followed a very different path involving plate convergence, terrane assembly and arc magmatism (Yavapai and Mazatzal orogenies). Such differences raise doubts about the validity of existing reconstructions of Rodinia based on simple matching of Palaeoproterozoic terranes (e.g. AUSWUS) but do not preclude a much younger age for the assembly of Rodinia based on matching of late Neoproterozoic events and sequences in eastern Australia and SW Laurentia.

Supercontinents and Earth Evolution Symposium 2005

UNRAVELLING THE NEW ENGLAND OROCLINE AND IMPLICATIONS FOR END-PALAEOZOIC TO EARLY MESOZOIC OROGENESIS ALONG THE PACIFIC MARGIN OF GONDWANA Sarah C Goss, Peter A Cawood Tectonics Special Research Centre, University of Western Australia, 35 Stirling Highway, Crawley, 6009, WA, Australia (sgoss@tsrc.uwa.edu.au)

Oroclines are map-view bends of originally quasi-linear lithospheric elements. The New England orocline lies within the Eastern Australian segment of the Terra Australis Orogen and developed during Late Palaeozoic to Early Mesozoic orogenesis that extended along the Pacific margin of the Gondwana supercontinent. The orocline is doubly vergent with the southern, Manning-Hastings segment having undergone up to 230° of counter-clockwise rotation and the northern Texas-Coffs Harbour segment showing clockwise rotation of at least 120°. The orocline deformed an arc assemblage consisting of a western magmatic arc, an adjoining forearc basin and an eastern subduction complex. Assuming an originally linear trend for the arc system, oroclinal bending resulted in a minimum of 50% shortening in New England from an original length of at least 1200 km. This has in part been accommodated by some 300 km of lateral displacement of the arc onto the Gondwana foreland. The doubly vergent nature of the orocline, with the N and S segments related to dextral and sinistral convergent regimes, respectively, has led to contrasting models of formation. We resolve these conflicting kinematic settings with a model involving buckling of the arc system about a vertical axis due to progressive northward translation of the New South Wales (New England) segment of the arc system against the Queensland segment which is pinned relative to cratonic Gondwana. Northward motion is driven by coupling between the Gondwana and Pacific plates. Specific features of the model are: Termination of magmatism at around 310 Ma followed by regional extension in New England and deposition of a marine succession unconformable on the subduction complex and eastern fore-arc sequences. Mafic magmatism associated with extension resulted in melting of the deeper parts of the subduction complex and emplacement of S-type granites. This was followed by regional deformation of the subduction complex which, at least in the

southwest, involved regional asymmetric sinistral-verging folds with an axial planar subvertical cleavage. Continued northward motion of the arc system in New England resulted in oroclinal bending. Paleomagnetic data from the Rouchel, Gresford and Myall blocks, that lay progressively further south in the southern segment of the forearc, show anticlockwise rotations of 80°, 80° and 120°, respectively. The Hastings Block, which represents the inferred most southerly segment of the forearc, was displaced northward and outboard of the subduction complex and shows evidence for up to 230° counterclockwise rotation. In contrast, the 120° clockwise rotation of the forearc and subduction complex elements within the northern segment of the orocline produced a regional Z-fold with a wavelength of around 250 km and amplitude of almost 200 km. Oroclinal bending was complete by 270 Ma when arc magmatism was re-established. The new magmatic arc formed a linear belt lying largely within the pre-existing subduction complex in New England, cutting across the orocline and also extending into the forearc in Queensland. The apparent eastward displacement of the magmatic arc between its pre-310 Ma and its post-270 Ma positions probably reflects foreshortening and westward displacement of the upper crustal lithosphere of the arc system rather than a dramatic shift in the position of the subduction zone within an asthenospheric reference frame. If correct this requires a detachment within the lithosphere above which oroclinal motion and foreland displacement took place. The spatial correspondence of oroclinal bending with earlier extension and crustal melting suggests that thermal weakening of the crust may have facilitated development of a crustal decollement above which the orocline formed. Westward motion of the orocline onto the foreland during buckling and vertical axis rotation may have been controlled by the rigid oceanic lithosphere of the Pacific plate restricting eastward motion of the arc system.

131

Supercontinents and Earth Evolution Symposium 2005

ASSEMBLY OF WEST GONDWANA - A PERSPECTIVE FROM THE DAMARA OROGEN COLLISIONAL TRIPLE JUNCTION David R Grav\ Ben Goscombe^, Richard A Armstrong^, David A Foster"*, Cees Passchier^, Rudolph Trouw® ^School of Earth Sciences, University of Melbourne, Melbourne, VIC 3010, Australia ^Northern Territory Geological Survey, Alice Springs, NT 0871, Australia ^ R I S E , School of Earth Sciences, Australian National University, Canberra, ACT 0200, Australia ^School of Geological Sciences, University of Florida, Gainesville, FL, 32611-2120, USA ^Institut fur Geoswissenschaften, JohannesGutenburg University, 55099 Mainz, Germany ^Institute de Geociencias, Universidade Federal do Rio de Janeiro, 21910-900 Rio de Janeiro, Brazil

New kinematic, geochronological and thermochronological data from the Damara Orogen of Namibia supports previously published temporally distinct suturing and amalgamation of South America (Rio del Plata craton) with the Congo and Kalahari cratonic nuclei of southern Africa. This three-pronged orogenic system is essentially a collisional triple junction (after Hoffman et al., 1994) made up of a coastal arm, with north and south expressions as the Kaoko and Gariep Belts respectively, and an inland Damara Belt extending through the Lufilian Arc and Zambezi Belt into the Mozambique Belt of eastern Africa. Oblique convergence between the Rio del Plata craton (South America) and the Congo and Kalahari cratons (Africa) involved sequential closure of the Adamastor Ocean, where the main ocean basin and/or sub-basins were closed first in the north and subsequently southwards from 550 Ma. Transpressive convergence in the Kaoko and Gariep coastal arms produced major westdipping listric shear zones that caused oblique crustal overriding of external or more outboard parts over and towards the African cratonic nuclei. Transpressional orogenesis in the Kaoko Belt occurred from 580-550 Ma and had clearly ceased by -535 Ma, with cratonisation marked by intrusion of post-kinematic granite and pegmatite between 535-505 Ma. The Kaoko Belt is a thermally softened margin in transpression, dominated by major, sinistral strike-slip shear zones defining regional scale shear lozenges within high-grade amphibolite facies Damara sequence turbidites, incorporating basement slivers and sheared Pan-African age granitoids. The Gariep Belt underwent transpressional convergence between -550-540 Ma (Frimmel and Frank, 1998) with erosion into the Nama foreland basin commencing at -540 Ma (Gresse

132

and Germs, 1993; Gresse, 1994; Frimmel, 2000) and intrusion of post-tectonic granites at -530520 Ma (Allsopp and others, 1988). East-directed thrusting continued inboard within the Nama foreland basin through 496 Ma (Gresse and others, 1988). The Gariep Belt is mostly of low metamorphic grade, consisting of an arcuate belt of SW-vergent, stacked oceanic thrust-sheets including melange with blueschist blocks, metagreywacke turbidites and metabasalts, thrust over the passive continental margin of the Kalahari craton. The Damara Belt (Inland Branch) is a classic doubly vergent orogen with craton-directed thrust systems on both margins (e.g. Hakos and Naukluft nappes in the Southern Margin Zone and Southern Foreland). It records closure of the Khomas Ocean basin by high angle convergence between the Congo and Kalahari cratons, with a marked tectono-thermal magmatic history extending from 570 Ma through -490 Ma with the intrusion of A-type granites (McDermott et al., 2000; Jung & Mezger, 2003). Ar-Ar data suggest Naukluft thrust-nappe emplacement on the southern foreland was at -570-550 Ma. Overall, the Damara and Gariep Belts appear to show both younger deformation and metamorphism than the Kaoko Belt, although deformation may have been occurring in the Damara Belt by -570 Ma. Continued highangle convergence through 530 Ma in the Damara Belt coincides with the Shortening Phase deformation in the Kaoko Belt, reflected by large-scale open E-W trending folds and by major NE-trending kink-like flexures, local subvertical NE-trending crenulation cleavage and penetrative foliation in contact aureoles in the Ugab Domain (Southern Kaoko Belt). Data from the Damara Orogen suggest that the assembly of this part of West Gondwana had occurred by 480 Ma

Supercontinents and Earth Evolution Symposium 2005

RELICS OF THE MOZAMBIQUE OCEAN: GEOCHEMISTRY OF THE VOHIBORY BLOCK (MADAGASCAR) Niels J6ns\ Volker Schenk\ Timm John\ Theodore Razakamanana^ ^Institut fur Geowissenschaften, Universitat Kiel, Germany (nj@nnin.uni-kiel.de) ^Departement des Sciences de la Terre, Universite de Toliara, Madagascar

INTRODUCTION The assembly of Gondwana and the associated formation of the East African Orogen (EAO) in Pan-African times (ca. 650-500 Ma) has been a subject of interest for many years. From petrological, geochronological, and palaeomagnetic data, a conclusive reconstruction of the Gondwana supercontinent has been developed. However, no occurrences of Pan-African eclogites, which would give direct evidence for the positions of Neoproterozoic subduction zones, have been reported anywhere in the EAO. Locating the relics of former oceans is therefore of importance for confirming the position of suture zones and to contribute to the question of the missing eclogites. Our work is focused on the Mozambique Ocean, the former existence of which was postulated by indirect evidence. This ocean separated Eastern and Western Gondwana and its closure is thought to have led to the final formation of Gondwana during the Pan-African orogeny. Metabasites and biotite-hornblende gneisses dominate the Vohibory Block of southwestern Madagascar. Furthermore, there are intercalated bands of marbles, amphibolites, serpentinites, and rare metapelites. From its lithologies, this tectonic unit has been interpreted as a shelf sequence tectonically interleaved with orthogneisses and slices of ophiolitic complexes, which may represent relics of the Mozambique Ocean floor. To test this hypothesis, we analysed major and trace elements of metamorphosed ultramafic to intermediate rocks and deduced the P-T conditions of metamorphism. GEOCHEMISTRY Based on trace element chemistry, the analysed rocks are subdivided into two groups: a first group that resembles MORB-type rocks ("MORB group") and a second group that can be interpreted as having formed in an oceanic arc or continental arc setting ("arc group"). In chondritenormalized REE patterns the MORB group rocks

show a slight depletion of the LREE ([La/Sm]N=0.58-0.66). Furthermore, they have Ce/Zr ratios of 0.11-0.14 and Nb/La ratios of 0.48-0.55, pointing to formation in a mid-ocean ridge or backarc spreading environment. The arc group is highly enriched in LREE (up to 180 times chondritic values) compared to HREE (3-17 times chondrite). In MORB-normalized patterns, they show negative Nb, Ta, and Ti anomalies. In combination with elevated Ce/Zr ratios of 0.23-0.49 and (La/Sm)N of 1.09-3.27, this points to an arc-related setting. The Nb/Zr ratios of samples from both MORB and arc groups are generally below 0.6, indicating a depleted mantle source for the whole suite of rocks. P-T CONDITIONS Previous workers have determined pressures and temperatures of 9.0-11.5 kbar and 750800°C, deduced from phase petrological considerations in metatroctolites. These estimates are confirmed by our conventional thermobarometry applied to garnet-hornblendebiotite gneisses (Grt-PI-Hbl-Qtz, Grt-Hbl, Grt-Bt), metabasites (Grt-Cpx, Grt-PI-Cpx-Qtz), and rare metapelitic assemblages (Grt-AS-Qtz-PI). CONCLUSION The lithologies of the Vohibory Block represent a tectonic melange consisting of basaltic rocks, orthogneisses, marbles, and metapelites. From trace element analyses, metabasites are interpreted as metamorphosed ocean floor basalts and rocks related to a magmatic arc setting. Therefore they are likely to have originated during Pan-African closure of the Mozambique Ocean. The absence of highpressure rocks and the amphibolite- to granulitefacies metamorphic overprint may be due to collision of an island arc with a continental margin after closure of a back-arc basin.

J33

Supercontinents and Earth Evolution Symposium 2005

DEEP CRUSTAL STRUCTURE OF THE INDIAN LITHOSPHERE: SEISMIC IMAGING JR Kaval Geological Survey of India, 27, JL Nehru Road, Kolkata, 700 016, India (jY_kayal@hotmail.com)

Peninsular India constitutes one of the largest Precambrian shield areas of the world. The Indo-Gangetic Alluvium Plain (IGAP) separates the Himalaya to the north and the peninsular shield to the south. The Shillong Plateau in the northeast constitutes an outpost separated from the main shield by the Bengal basin and from the Himalaya by the Brahmaputra river. The Peninsular shield is made up of three main cratonic regions: the Aravalli, Dharwar, and Singhbhum cratons, which are separated by Proterozoic rifts and mobile belts. The major rifts that separate the southern and northern blocks of the shield are the Narmada Son and Tapti lineaments, together called Son-Narmada Tapti lineament (SONATA). The other rift basins are the Kutch, Cambay, Godavari, Cuddapah etc. A National Working Group (NWG) was formed by the Geological Survey of India (GSI), under project IGCP 474, to evaluate the crustal structure of the Indian lithosphere. This group, in addition to the GSI, incorporates earth scientists of the National Geophysical Research Institute (NGRI), Indian Institute of Geomagnetism (IIG), India Meteorological Department (IMD), Indian School of Mines (ISM), and the Indian Institute of Technology, Bombay (IITB), etc. A large amount of work was conducted using Deep Seismic Sounding (DSS) by the NGRI group of scientists. DSS results from refraction and wide angle reflection reveal much subsurface information: (1) Thickness of the Deccan traps varies from - 1 0 0 m in the northeast to --1500 m at the west coast, (2) Hidden Mesozoic basins are delineated below the Deccan traps in the SONATA zone, (3) The number of faults and their displacement pattern indicate block tectonics in peninsular India, (4) Crustal underplating is present in the SONATA zone. The presence of high velocity layers (7.07.3 km/s) from a depth of 8-12 km down to the Moho is indicative of mafic-ultramafic intrusives, (5) The Moho varies from 33 to 43 km in peninsular India region, but below the SONATA zone the Moho is deeper (40-43 km), (6) Forward modeling of refraction/reflection data yielded a five-layer velocity structure and (7) seismic reflection data identifies a suture in the Satpura mobile belt.

134

Receiver function studies have been conducted using teleseismic earthquake waveform data from 10 broadband stations spread over peninsular India and 5 stations over the northeast India. A thick normal (33-39 km) crust occurs below the south Indian Archaean shield; the Deccan basalts have not significantly affected the underlying crust. The predominant Proterozoic crust in the northern and eastern shield, on the other hand, exhibits a complex character. Moho conversions are considerably weaker compared to the Archaean terrains, and crustal thickness is >40 km. In northeast India, thinner (33 km) Archaean crust occurs below the Shillong Plateau; thickness increases to - 4 0 km below the Brahmaputra valley and to - 5 0 km below the northeastern Himalaya to the north. Detailed seismic images were obtained by Seismic tomography, using P and S phases of local earthquakes in northeast and peninsular India, the aftershock sequence of the 1993 Killari earthquake (M 6.3) in the southern Archaean shield, and the sequence of the 2001 Bhuj earthquake (Mw 7.7) in the Kutch rift basin. Seismic images of northeast India, including the Bengal basin and Indo-Burma ranges, clearly reveal low velocity, thick ( - 2 0 km) alluvium/ sediments in the Bengal basin and a high velocity dipping lithosphere below the IndoBurma ranges. Detailed imaging revealed high velocity mantle at 30-35 km depth beneath the Plateau and at 40 km depth below the Brahmaputra valley - results comparable with those of the receiver function study. Seismic imaging in the Killari area of the southern shield, however, was confined to shallow depths ( - 1 0 km); lateral heterogeneity in velocity structure is identified as the earthquake source area. In the Kutch rift basin, the source zone of the 2001 Bhuj earthquake sequence, the high velocity mantle at a depth of 38-40 km is found. A similar deeper mantle is reported below the SONATA. The NWG IGCP 474 in India is now preparing a seismological data base for more detailed seismic imaging of the Indian region, for both continental and oceanic lithosphere. The 2004 Sumatra-Andaman earthquake (Mw = 9.3) sequence provides a large dataset for this study.

Supercontinents and Earth Evolution Symposium 2005

3D MODEL OF CRUSTAL STRUCTURE OF THE SOUTHEASTERN FENNOSCANDIAN SHIELD: EVIDENCE FROM REFLECTION SIESMIC AND GEOLOGICAL DATA Michael Mints\ Arsen Suleimanov^, Robert Berzin^, Nadezhda Zamozhniaya^, Vladimir Stupak^ Alexander Konilov\ Valery Zlobin^ ^Geological Institute of the RAS, Moscow, Russia (nnichael.mints@nntu-net.ru) ^Spetsgeofisica, Povarovka, Moscow region, Russia

The southeastern Fennoscandian Shield includes the Neoarchean Karelian Craton, Archean-Paleoproterozoic Belonnorian innbricated thrust belt, and the southeastern edge of the Svecofennian accretionary orogen. The area is crossed by two CDP profiles, fronn north to south by the 1-EU geotransect (350-1350 km section) and in its central part by the latitudinal 4B profile (270 km length). Cross-sections provide reflection seismic images of the crust and upper mantle from the surface to -80 km depth (25 s). Archean granite-greenstone assemblages of the central and eastern Karelian Craton are transected by a series of Paleoproterozoic volcano-sedimentary belts that are usually understood as synclinal folds built by "protoplatform" assemblages or as deformed strata of riftogenic troughs and basins. In seismic images, these parts of the Karelian Craton appear as imbricated crust built by alternating slices of Archean granite-gneisses and Paleoproterozoic volcano-sedimentary assemblages. Tectonic slices dip steeply northeastward, then flatten and partially unite at 20 km depth. This level can be understood as the main detachment at the base of the Paleoproterozoic succession. Correspondingly, Paleoproterozoic linear belts are interpreted as outcropping frontal parts of these slices. In turn, the Onega

"depression" appears as an ensemble of gently dipping slices thrust northwestward. Tectonic slices of the Belomorian belt were thrust westward on top of the Karelian Craton. The most prominent tectonic zone observed along the 48 profile separates the Belomorian belt from the Karelian Craton and Paleoproterozoic Shombozero structure. The geometrical features of reflection images portray that zone as an overthrust-underthrust structure. Tectonic sheets several hundred kilometres long, formed of both Archean and Paleoproterozoic assemblages, can be traced from the present surface to the Moho discontinuity, where irregularly distributed and partially oriented reflection events continue the main lines of crustal structure down to the mantle. Similar features are characteristic of bending and plunging down to the mantle of high-reflective low-crustal fragments along the Karelia-Svecofennian boundary, which suggests they are remnants of Svecofennian oceanic lithosphere. Careful coordination of mapped geological structures and geological deciphering of seismic images along both cross-sections permit speculations on a 3D model of the crust and upper lithospheric mantle.

135

Supercontinents and Earth Evolution Symposium 2005

HOW THICK IS THE EARTH'S CRUST? Walter D Moonev\ Gary Chulick^'^, Shane Detweiler^ ^U.S. Geological Survey, 345 Middlefield Rd., MS 977, Menio Park, CA 94025, USA (nnooney@usgs.gov) University of Wisconsin, Whitewater, Wisconsin, USA

To understand continental evolution and other geological processes, it is first necessary to develop a thorough knowledge of the Earth's crustal structure. We present a recently-updated contour map of the thickness of the Earth's crust using a 10-km contour interval, with the 45-km contour also included. This contour map was created from about 8000 individual crustal data points that have been acquired during the past 65 years. The contour map honours all available seismic refraction measurements for features with a dimension greater than 2 degrees. Crustal thicknesses in Eurasia, North America, and Australia are well constrained by seismic refraction data, while new data has enhanced

136

resolution in Antarctica, South America, Africa, and Greenland. To a first approximation, the continents and their margins are outlined by the 30-km contour. The part of the continental interior enclosed by the 40-km contour and regions with crustal thickness of 45 to 50 km are found on all well-surveyed continents. Continental crust with thickness in excess of 50 km is exceedingly rare and accounts for less than 10% of surveyed continental crust. These observations, now available on a global basis, provide important information to be used for numerous scientific studies, including ongoing refinements to global crustal models.

Supercontinents and Earth Evolution Symposium 2005

PERMIAN CONTINENTAL COLLISION EVENT IN INDOCHINA REGION N Nakano\ Y O s a n a i \ M Owada^ S TaguchP, H Kaiden^ T Hokada^ I r a n N NarnVf^Cagami® ^Division of Evolution of Earth Environment, Kyushu University, 4-2-1 Ropponmatsu, Chuo-ku, Fukuoka 810-8560, Japan (cs303042@scs.kyushu-u.ac.jp) ^Department of Earth Sciences, Yamaguchi University, Yamaguchi 753-8511, Japan ^Department of Earth System Science, Fukuoka University, Fukuoka 814-0180, Japan "^National Institute of Polar Research, Tokyo 173-8515, Japan ^Department of Geosciences, Hue University, Hue, Vietnam ^Graduate school of Science and Technology, Niigata University, Niigata 950-2102, Japan

INTRODUCTION The Asian continent was formed by amalgamation of several microcontinents during Permo-Triassic time. The evidence of collision metamorphism is recorded in high-grade metamorphic rocks such as ultra high pressure (UHP) eclogites (e.g. Dabie-Sulu belt) and ultra high temperature (UHT) granulites (e.g. Higo terrane). The diamond-bearing UHP eclogites, occurring between the North and South China cratons, are identified as results of continental collision and deep continental subduction. The Kontum massif, distributed in central Vietnam, had long been considered to be Precambrian basement of the Indochina craton. However, recent chronological investigations suggest that part of the terrane was metamorphosed in the Permo-Triassic. UHT politic and mafic granulites were also reported, recently, which is considered to relate to Asian collision event resulted from their metamorphic ages and clock-wise P-T paths. In this presentation, we discuss details of P-T-t evolution of this UHT mafic granulite. OCCURRENCE AND PETROGRAPHY UHT mafic granulites occur as blocks within politic or felsic gneisses. The mafic granulite consists of coarse-grained garnet, Cpx, rutile and quartz as well as fine-grained Opx+PI±Mag±Spl symplectites, which are well-developed around garnet. The peak conditions of this granulite are 1050X and 1.3 GPa, based on Grt-Opx thermometry using garnet rims and symplectic Opx and thermodynamic data for the reaction of Grt+Cpx+Qtz=Opx+PI. Politic gneisses including mafic blocks also indicate granulite-facies mineral paragenesis of Grt+Opx±Sil±Spl.

UHP RELICTS UHP relicts in UHT mafic granulite are observed as quartz rods in Cpx and rutile needles in garnet. The recalculated Cpx and Grt show supersilicic and supertitanic compositions, respectively, which suggest the rock underwent UHP conditions preceding UHT conditions. GEOCHRONOLOGY SHRIMP ages of recrystallized or new grown zircon in host politic gneiss are ca. 300 to 240 Ma. Sm-Nd isochron ages of host politic gneiss and mafic block yield 247 ± 1 and 240 ± 2 Ma, respectively. Considering Sm-Nd closure temperatures for garnet, the Sm-Nd ages are identified as cooling ages after UHT conditions. The range of SHRIMP ages suggests that zircon grew during both prograde and retrograde stages. However, zircon containing biotite inclusions yields 250 ± 5 Ma, which is considered to be affected by retrogression at this time. The cooling age possibility corresponds with that from Sm-Nd system. Thus, we concluded that the metamorphism started at ca. 300 Ma and reached UHP stage at Permian time (280-270 Ma?). Subsequently, at the late Permian (ca. 250 Ma), the rock was already in cooling stage just after UHT conditions. CONCLUSIONS This study proved the presence of Permian UHP metamorphism in Vietnam, which would be due to collision between South China and Indochina cratons and deep crustal subduction. It is also the first possibility of UHT granulite preceded by UHP conditions, and could provide critical insights into continental collision tectonics elsewhere in the world. During the presentation, we will discuss above mentioned metamorphic evolution with the results of inclusions within zircon.

137

Supercontinents and Earth Evolution Symposium 2005

THE PALAEOPROTEROZOIC TO NEOPROTEROZOIC EVOLUTION OF THE CAPRICORN OROGEN, WESTERN AUSTRALIA: A SUMMARY IN SPACE AND TIME Sandra A Occhipinti\ Steven M Reddy\ Steve Sheppard^ ^Tectonics Special Research Centre, Curtin University of Technology, PO Box 1987, Perth, WA 6845, Australia (S.Occhipinti@curtin.edu.au) ^Geological Survey of Western Australia, 100 Plain St., East Perth, WA 6004, Australia The Palaeoproterozoic to Neoproterozoic tectonic evolution of the southern Capricorn Orogen can be explained in terms of various periods of rifting, basin formation, and orogenesis. A diagram that summarises all tectonic units and available data in space and time allows for rational correlations to be made different tectonic units in the region, as well as assessment of plausible tectonic models. Between 2300 and 2200 Ma, rifting in the southeastern Capricorn Orogen led to the formation of the Yerrida Basin, and the opening of an ocean on the northwestern margin of the Yilgarn Craton and the subsequent formation of the Bryah Basin. Along the southern margin of the Pilbara Craton a possible foreland basin developed at c. 2200 Ma, however, the method for initiation of this basin in a convergent setting is unknown. At 2010-1950 Ma, the Glenburgh Terrane developed in an arc-type setting above a northwest or west-dipping subduction zone. Subsequent accretion and collision of the Glenburgh Terrane (during the Glenburgh Orogeny) onto the northwestern Yilgarn Craton resulted in the formation of the Errabiddy Shear Zone, the closure of the Bryah Basin and formation of the Padbury foreland basin. Further east the lowermost units of the Earaheedy Group may have been deposited at this time. Between 1950 and 1830 Ma, there was an apparent hiatus in tectonism throughout most of the Capricorn Orogen, with the exception of deposition within the Mooloogool Group in the Yerrida Basin. Mafic intrusive rocks of the Mooloogool Group intruded into black shales at c. 1840Ma. The Capricorn Orogeny took place between 1830 and 1780 Ma, and was probably the result of collision between the Archaean Pilbara and Yilgarn Cratons. Felsic granitoid rocks preserved throughout most of the Capricorn Orogen were emplaced between 1830 and 1780 Ma. Intrusion of granitoid rocks was locally associated with high-grade metamorphism and deformation of

138

country rock. High-grade metamorphism of the Morrissey Metamorphic Suite metasedimentary rocks took place after -1840 Ma, and may have been coincident with the Capricorn Orogeny. In the south, flattening strains associated with shearing took place between shear zones that were deformed by dextral non-coaxial strain. At the same time in the northern part of the Capricorn Orogen, sedimentary rocks were deposited in north-verging fold-and-thrust belts between -1820 and -1780 Ma. In the southeastern Capricorn Orogen, the upper Earaheedy Group was probably deposited in a foreland basin setting as a result of the Capricorn Orogeny. Throughout much of the Capricorn Orogen sedimentary rocks of the Bresnahan Group and Mount James Formation were deposited sometime between 1780 and 1620 Ma. Peraluminous granitoid rocks of the Durlacher Supersuite were intruded in the northern and central Capricorn Orogen (Boora Boora, Mangaroon and Limejuice zones) at 1680-1620 Ma, and in the southern part of the Capricorn Orogen between 1640 and 1600 Ma during the Mangaroon Orogeny. Rifting, resulting in formation of the Bangemall Basin took place between 1600 and 1470 Ma. At -1465 Ma, dolerite sills intruded the Bangemall Basin, accompanying sediment deposition. At -1070 Ma, the Warakurna large igneous province formed along the central axis of the Capricorn Orogen. Deformation and lowgrade metamorphism between -1070 and 750 Ma, during the Edmundian Orogeny, resulted in exhumation of the southern to central western Capricorn Orogen and development of the lower parts of the Officer Basin to the east. The correlation diagram shows that between the Palaeo- and Mesoproterozoic, tectonism in the Capricorn Orogen likely changed from being dominated by interplate plate tectonic processes to intraplate processes, with far-field plate interactions influencing deformation and basin development well into the Neoproterozoic.

Supercontinents and Earth Evolution Symposium 2005

REGIONAL DIVISION OF METAMORPHIC PROCESSES IN THE NAPIER COMPLEX, EAST ANTARCTICA AS A FINAL EVENT OF CONTINENT COLLISION DURING ARCHEAN Yasuhito Osanai^ Tsuvoshi Toyoshima^, Masaaki Owada^, Tomokazu Hokada^ Toshiaki Tsunogae^ Warwick Crowe®, Samuel Mukasa^, Nobuhiko Nakano^ ^Division of Evolution of Earth Environments, Kyushu University, Ropponmatsu 4-2-1, Fukuoka, 8108560 Japan (osanai@scs.kyushu-u.ac.jp) ^Department of Geological Sciences, Niigata University Ikarashi 2-8050, Niigata, 950-2181 Japan ^Department of Earth Sciences, Yamaguchi University, Yoshida 1677-1, Yamaguchi, 753-8512 Japan "^National Institute of Polar Research, Kaga 1-9-10, Itabashi-ku, Tokyo, 173-8515 Japan ^Department of Earth Sciences, University of Tsukuba, Tenno-dai 1-1, Tsukuba, 305-8572 Japan ^Tectonics Special Research Centre, University of Western Australia, Crawley WA 6009, Australia ^Department of Geological Sciences, University of Michigan, Michigan 48109-1005, USA

The Napier Complex in eastern Enderby Land, East Antarctica is one of the most famous Archean ultrahigh-temperature (UHT) metamorphic terranes in the world. The dominant rock types of the Napier Complex are Opx- and garnet-bearing quartzofeldspathic gneisses of igneous origin (Archean TTG-type tonalitic orthogneiss) with subordinate ultramafic, mafic, politic, calcareous, siliceous and aluminous granulites. Characteristic UHT-type mineral assemblages of Spl+Qtz and Grt+Spr+Qtz, with or without Opx and osumilite (Osm), and Opx+Sil+Qtz are widespread in the Napier Complex. Recent geochronology reveals that the age of high-grade UHT metamorphism may have occurred at 2.5 to 2.8 Ga in the Napier Complex. We conducted geological and petrological investigations around Amundsen Bay (e.g. Tonagh Isl., Bunt Isl., Mt. Pardoe, Priestly Peak, Mt. Liiser-Larsen, etc.). In this presentation, we will discuss the tectono-metamorphic evolution of the entire Napier Complex using data from these areas with respect to primitive continental growth during Archean. METAMORPHIC GEOLOGY Metamorphic rocks in the entire Napier Complex undenA/ent UHT-granulite facies metamorphism. Characteristic low-pressure UHT-mineral assemblages of Spl+Qtz and Grt+Osm+Qtz are identified in the western Napier Complex (WNC), where Opx+Sil+Qtz is unstable. However, in the eastern Napier Complex (ENC), relatively highpressure UHT-assemblages of Spr+Opx±Grt+Qtz and Opx+Sil+Qtz without any Osm and Spl+Qtz are ubiquitous. Regional-scale analysis of geological structure using a strike-line map

indicates that there is a dominant structural gap between the ENC and WNC. METAMORPHIC EVOLUTION In the ENC (e.g. Bunt Isl.), clockwise P-T evolution under UHT-condition is indicated by Osm-forming reactions, but inclusions of Opx, sillimanite and quartz are also identified in garnet and Spr as an unstable previous high-pressure assemblage. This P-T path shows nearisothermal decompression, which would start from 1.2 GPa or more high-pressure at around lOOOX. Fine grains of srilankite also support the primary high-pressure condition. In the WNC (e.g. Tonagh Isl.), counterclockwise P-T evolution is indicated by changes in divariant mineral assemblages from Spl+Qtz to Opx+Sil via Spr+Qtz. The inferred P-T path indicates isothermal compression under UHT conditions, peaking at --1.0 GPa and -1100°C, followed by near-isobaric cooling. TECTONIC SIGNIFICANCE The age of peak metamorphism in the Napier Complex is uncertain. Detailed geochronological analyses using multi-isotope systems indicate that a -2.5 Ga event is recorded in almost all areas of the Napier Complex, whereas a slightly older event, at 2.7-2.8 Ga, is identified only in the ENC. We infer that micro-continent collision occurred during the late Archean. After formation of a crustal pile, the WNC was situated beneath the ENC, as indicated by different P-T paths for the two complexes. Metamorphic evolution of the WNC was completely overprinted by collision metamorphism at -2.5 Ga, although the precollision history of the overlying ENC is preserved.

139

Supercontinents and Earth Evolution Symposium 2005

TECTONICS OF THE NORTH WESTERN INDIAN PENINSULAR SHIELD AND ITS SIGNIFICANCE BS Paliwal, Alka Baghela Department of Geology, Jai Narain Vyas University, Jodhpur-342001, India (paliwalbhawani@yahoo.co.in)

The

Precambrian

crust

along

the

northwestern

Possibly,

the A r a v a l l i

Orogeny and the

Delhi

margin of the Indian Peninsular Shield evolved through cratonization of Archaean basement by the accumulation of several sedimentary cycles and their igneous counterparts at intervals. M o r e than 3.5

Orogeny represent two stages of a single orogeny rather than two separate orogenies. These orogenic movements were followed by widespread magmatism, accumulating the granitic plutons of

billion year old A r c h a e a n basement, popularly known as the Banded Gneissic Complex or the M e w a r Gneiss, intruded by a large number of Proterozoic granites like those of A h a r River, Gingala, Untala, Darwal, Udaisagar Granite,

Erinpura (850 Ma) and Sendra (790-840 Ma) etc. The 680 to 7 8 0 M a period included igneous activity along the western flank of the A r a v a l l i Mountain Range, and produced volcanic and plutonic rocks of the Neoproterozoic Malani Supergroup and

Erinpura, Anasagar, etc., suffered a tensional tectonic setting resulting in formation of a series of linear grabens and half grabens in which volcaniclastic and chemogenic sediments of the Palaeoproterozoic Aravalli Supergroup were

volcaniclastic sediments of the Sindreth and Punagarh Groups. It is possible that mantle plumes

deposited. The basins were filled in completely and evaporitic conditions marking the closing of a sedimentary cycle were evolved. Ultimately, a fluvial regime took over in these areas. A t this stage, the Palaeoproterozoic A r a v a l l i Orogeny took place in the region, and produced first generation folds in metasediments of the A r a v a l l i Supergroup. The Palaeoproterozoic A r a v a l l i Orogeny, which occurred along a very narrow zone, ---lOO km in width, was followed by another tensional tectonic event. This second major orogenic movement, the Delhi Orogeny ( 1 6 6 0 - 1 2 0 0 Ma), occurred along the same narrow zone where the A r a v a l l i Orogeny

developed at this time and fragmented the overlying crustal plate, resulting in outpouring of the M a l a n i Volcanics which cover more than 5 0 , 0 0 0 km^ Tectonism associated with development of faulted grabens produced three parallel basins, the Sindreth and Punagarh Basins on the eastern side of the A r a v a l l i Mountain Range in Rajasthan and the B i r m a n i a Basin on the western side, in which clastic sediments of the Randha Sandstone and chemogenic sediments of Birmania were deposited. This was followed by deposition of a huge quantity of sediments on either side of the A r a v a l l i Mountain Range. To the east, the Vindhyan Supergroup was deposited, whereas sediments of the Marwar Supergroup or the Trans-Aravalli Vindhyans were deposited on the western side.

( 2 5 0 0 - 1 9 0 0 Ma) occurred. The Delhi Orogeny produced an additional series of linear grabens and half grabens in the same area where the earlier A r a v a l l i extension produced the linear basins in which Palaeoproterozoic sediments of the A r a v a l l i

A l l these local events have a global significance and play an important role in understanding the assembly and break-up of Rodinia and growth of Asia. In particular, repeated development of tensional tectonic settings, formation of depositional

Supergroup were deposited. Siliciclastic and chemogenic sediments of the Mesoproterozoic Delhi Supergroup accumulated in these narrow linear basins. These orogenic movements that followed the development of tensional tectonics, occurred in the

basins, filling up of these sedimentary basins, and development of compressional forces that gave rise to the Palaeoproterozoic Aravalli and Mesoproterozoic Delhi Orogenies in a 100 km wide zone of the Earth's crust is quite interesting when

area presently occupied by the A r a v a l l i Range in northwestern India.

visualized in the global perspective.

140

Mountain

Supercontinents and Earth Evolution Symposium 2005

PRE-TERTIARY GLACIAL EVENTS IN NORTHWESTERN INDIA AND THEIR SIGNIFICANCE IN THE ASSEMBLY AND BREAKUP OF RODINIA AND GROWTH OF ASIA BS Paliwal, Alka Baghela Department of Geology, Jai Narain Vyas University, Jodhpur-342001, India (paliwalbhawani@yahoo.co.in)

In northwestern India, there is evidence of at least four glacial events of pre-Tertiary age. The first occurred during deposition of volcaniclastic and chemogenic sedinnents of the early Proterozoic Aravalli Supergroup in southern and central Rajasthan. A thick polymictic conglomerate bed, containing large, wellrounded boulders, occurs in the upper portion of the metasedimentary sequence of the Aravalli Supergroup, in the hills of the Aravalli Mountains. It is difficult to trace out glacial striations on the surfaces of the boulders because of the complex polyphase deformation which resulted from two major and several minor folding episodes that affected the region. However, the compositions of boulders, their fabric or texture, and the distribution of pebbles, cobbles, and boulders support a glacial origin. This boulder horizon occurs in four linear belts extending for several kilometres in the southern and central parts of the state. The best exposures are near the Kadia-Losing area north of Udaipur city in Rajasthan - the type locality of the Palaeoproterozoic Aravalli Supergroup. The very large boulders with high degrees of sphericity and roundness, together with finely laminated shales, indicates a glacial environment of deposition. The remaining three boulder beds occur to the west of the Aravalli Mountain Range in the arid and semi-arid regions of the great Thar Desert. The second glacial event, of Neoproterozoic age, immediately followed the widespread Malani volcanic activity (740 ± 10 Ma), and produced the Basal Conglomerate of Paliwal (1994, 1998, 1999), or the wrongly

identified as Pokaran Boulder Bed of Blanford (1877), Oldham (1886, 1888), and Pareek (1981, 1984) at the base of the Marwar Supergroup (the Trans-Aravalli Vindhyans of Heron, 1932, 1953) followed by the Sonia Shale Formation and Girbhakar Sandstone Formation of the Jodhpur Group. The third glacial event, representing the Eocambrian glaciation (cf. Dott and Botten 1976), resulted in deposition of the true Pokaran Boulder Bed just above the calcareous Bilara Group (Paliwal, 1994, 1998, 1999, 2001) overlain by arenaceous and evaporitic sediments of the Nagaur Group. The Neoproterozoic Basal Conglomerate and post- Bilara true Pokaran Boulder Bed are integral parts of the Marwar Supergroup (Neoproterozoic to Cambrian in age) and provide evidence of Neoproterozoic and Eocambrian glaciations, respectively, in this part of India. The fourth and perhaps the last pre-Tertiary glacial event in the region occurred during deposition of the Bap Boulder Bed of Talchir age (Gondwana Supergroup of the Permian to Carboniferous period) which can readily be correlated with its counterparts in the heart of Africa and other continents like Australia, South America, and Antarctica, etc. All these glacial events, in which glacial boulder beds were deposited, play significant roles in the geological history of the region, particularly in the context of the assembly and fragmentation of Rodinia and growth of Asia, because they were associated with global events and constrain the position of India relative to other continents at that time.

141

Supercontinents and Earth Evolution Symposium 2005

BALTICA IN THE CRYOGENIAN, 850-630 Ma V Pease\ JS Daly^, S-A Elming^, R Kumpulainen\ M Moczydlowska^ V Puchkov^ D Roberts®, A Saintot®'^, R Stephenson^ ^Dept. of Geology & Geochemistry, Stockholm University, SE-106 91, Sweden ifvicky.pease@geo.su.se) Dept. of Geology, University College Dublin, Belfield, Dublin 4, Ireland ^Dept. of Applied Chemistry & Geosciences, Lulea University of Technology, S-97187 Lulea, Sweden "^Dept. of Earth Sciences, Uppsala University, S-75236 Uppsala, Sweden ^Institute of Geology, Ufimian Science Centre RAS, KarlMarx St. 16/2, Ufa 450 000, Russia ^Geological Survey of Norway, N-7491 Trondheim, Norway ^Faculty of Life & Earth Sciences, Vrije Universiteit, De Boelelaan 1085, 1081 HV Amsterdam, Netherlands

During IGCP 440 - The Assembly and Break-up of Rodinia - a new tectonic synthesis for Baltica during the break-up of Rodinia, e.g. the Cryogenian, has been compiled. This synthesis provides a framework for understanding the dynamic evolution of Baltica and for constraining tectonic correlations within the context of Rodinia's break-up. Baltica in the Cryogenian easily divides into three principal domains: west-northwesterly, east-northeasterly, and southerly. The westnorthwesterly margin is characterized by Neoproterozoic to lower Cambrian sedimentary successions deposited on crystalline rocks of the Sveconorwegian (1200-900 Ma) mobile belt, reworked during Caledonian orogenesis. Attempted rifting of this margin (tholeiitic Blekinge-Dalarna dike swarm, ca. 800 Ma) was unsuccessful until ca. 620-550 Ma, when intense, partly alkaline dikes and mafic-ultramafic complexes (e.g. Seiland Igneous Province) were emplaced. The sediments, recording the change from an alluvial setting to a marine environment and eventually to a partially starved(?) turbidite basin, document the late rifting of Baltica from the supercontinent Rodinia. Baltica's east-northeasterly margin was a long-lived passive margin. Episodic rifting began in the early Mesoproterozoic (?) and was followed by passive margin deposition through the Cryogenian. A system of longitudinal border faults effectively separated its pericratonic or

142

platformal domain from its deeper-water, basinal domain. This margin was likely an ocean-facing part of the supercontinent Rodinia. It remained a passive margin during Rodinia break-up until the onset of Timanian orogenesis at ca. 616 Ma, when extension gave way to compression and ocean-continent collision with island arc accretion occurred. Baltica's southern margin remains enigmatic. Precambrian basement is generally hidden beneath 5-12 km thick Ediacaran and younger platform sediments and is poorly known. It probably represents the East European Craton, reworked by late Proterozoic and younger tectonism. Granites, diorites and gabbros from the central region yield Cryogenian and Ediacaran K-Ar ages (790 Ma and 640-620 Ma, respectively) interpreted to document Cryogenian intracratonic rifting; 700-600 Ma granitoids in the northern Caucasus mountains are thought to represent mantle/subduction-type magmatism; magmatism and high-grade metamorphism in the southern Caucasus are interpreted to document Neoproterozoic arc accretion. Thus the southern margin may i) represent a rifted margin developed during Rodinia break-up, ii) be contiguous with Baltica since at least the time of Baikalian orogenesis, or iii) have Cadomian affinities and defines a Paleozoic suture. At present, the nature of this margin is unknown and offers little to constrain the break-up of Rodinia

Supercontinents and Earth Evolution Symposium 2005

THE KORAS GROUP, SOUTH AFRICA: A TECTONOSTRATIGRAPHIC MARKER FOR THE NAMAQUA COLLISION EVENT? Asa Pettersson, David H Cornell Earth Sciences Centre, Goteborg University, Box 460, SE-405 30, Goteborg, Sweden (asap@gvc.gu.se)

In narrowing down the time constraint for the collision event(s) during the amalgamation of Rodinia at --1200-1000 Ma in southern Africa, one supracrustal sequence has been considered especially important. The Koras Group lies in the Kheis Terrane in the tectonic front of the Namaqua-Natal Province. It overlies highly deformed rocks related to the Namaqua collision, but is itself is considered to be undeformed. Thus it postdates deformation in the Kheis Terrane and possibly in the whole province, in which orogenic events have been dated between -1200 and 1020 Ma. Because of pervasive lowgrade metamorphism, the U-Pb system in zircon has been invoked to provide a reliable age for the Koras lavas. We have investigated the Koras Group in detail as well as a migmatite from the Areachap Terrane further to the west and at a higher grade of metamorphism. A recent SHRIMP study gave the Koras Group Pb-Pb ages of 1171 ± 7 Ma, confirmed by our 1162 ± 13 Ma U-Pb age, from the lower part of the Group. However, we discovered a major time

gap in the stratigraphy of the Koras Group, identified by Sanderson-Damstra in 1982 as an unconformity. Moreover, the lower part shows signs of deformation as slickenside striations and folding, whereas above the unconformity the rocks are completely post-deformation and now dated at 1090 ± 10 Ma. A new U-Pb zircon rim date at 1165 Ma in adjoining Areachap Terrane for the migmatization and deformation just barely postdates a major granitoid emplacement age in the whole region around 1180-1165 Ma. The 1165 Ma deformation event explains why the lower formations of the Koras Group are slightly deformed. This means that the collision event might not be constrained by the 1171 Ma age, as this probably formed during the peak of the deformation. Thus the role of the Koras Group as a tectonic marker is not as simple as previously assumed. Its position in the distal part of the Namaqua Front means that its deformational history does not fully represent that of the deeper seated terranes in the former Mountain Belt.

143

Supercontinents and Earth Evolution Symposium 2005

THE BASAL PALEOPROTEROZOIC COVER SEQUENCES OF LAURENTIA: FROM THE BREAK-UP OF KENORLAND TO THE ASSEMBLY OF NUNA Robert H Rainbird, William J Davis, Sally J Pehrsson Geological Survey of Canada, 601 Booth St., Ottawa, ON, K1A 0E8, Canada (rrainbir@nrcan.gc.ca)

Sequence stratigraphic studies in concert with UPb dating offer a powerful tool for evaluating the correlation and tectonic development of sedimentary basins that can contribute to paleocontinental reconstructions. Such studies have helped to further our understanding of the temporal and evolutionary relationships among Earth's oldest thick, mature, cratonic sedimentary successions and are critical to our understanding of the tectonic evolution and reworking of underlying crustal domains. In North America, areas of current research include the western Churchill province, west of Hudson Bay, and the Lake Superior region of the southern Superior province (Southern Province). A connection between these regions and the Wyoming craton was proposed over 30 years ago, based on lithostratigraphic correlation of a succession of glacial diamictites and overlying thick aluminous quartz arenites, and was used as evidence for the existence of a late Archean supercontinent (Kenorland).

from distal sources. The thickness and relative compositional maturity of the lower sequence records a protracted period of enhanced weathering on a stable supercontinent (Kenorland). This was followed by mafic volcanism and basin flooding, consistent with rifting and crustal extension attending supercontinent break-up. Overlying middle sequence strata record platformal sedimentation along the margins of rifted continental blocks. The upper sequence is interpreted to represent deposition in foreland basins during initial amalgamation of Laurentia (part of Nuna) at ca. 1.9 Ga. The sequence stratigraphy and detrital zircon geochronology of the Hurwitz Gp., which overlies the Hearne domain, is broadly similar to the cover sequence of the Rae domain, but differs through the presence of glacial diamictites and lack of 2.5 Ga detritus, a significant local source in the Rae. Our data suggest a RaeHearne connection by max. 1.91 Ga (i.e. max. age of deposition of the upper Hurwitz Gp.).

WESTERN CHURCHILL PROVINCE The basal Paleoproterozoic cover sequence of the western Churchill province comprises sedimentary basins overlying the Archean Rae and Hearne domains. The Rae domain cover (e.g. Amer, Ketyet River, Penrhyn, Piling and Lake Harbour groups) includes a shallow marinefluvial quartz arenite (lower) sequence with mainly Neoarchean detrital zircons of local provenance. The upper part of the quartz arenite sequence locally includes carbonates ± mafic volcanic flows and records an abrupt but conformable transition to deeper water deposits. Sandstones within this transition have a substantially different detrital zircon provenance signature dominated by early Paleoproterozoic zircons, with grains as young as 1.95 Ga. The deeper water (middle) sequence begins with black shales and platformal carbonates passing upward into siltstones and sandstones that record gradual emergence of the platform. An unconformably to conformably overlying (upper) sequence of immature marine sandstones, including turbidites, has a significant component of 2.0-1.9 Ga detritus that was probably derived

SOUTHERN PROVINCE The Southern province cover sequence comprises from west to east, Animikie Group, Marquette Range Supergroup, Huronian Supergroup-Whitewater Group and MistassiniOtish Mountains Group. Glacial diamictites occur at the base of all, although the Huronian preserves a more complete and older depositional record, including 3 successive glacial diamictite-quartzite sequences. The upper glacial sequence is capped by thick, mature, aluminous quartzites, similar in character and age to the lower western Churchill cover sequence. The Southern province sequence is unconformably overlain by relatively deep-water marine deposits ± Fe-formation ± mafic volcanic rocks above the quartz arenites, signifying continental flooding, perhaps associated with rifting and break-up of Kenorland. Above the deep-water deposits is a ca. 1.9-1.85 Ga coarsening upward succession interpreted to represent a foredeep, formed during collisional orogenesis (Penokean orogeny) attending amalgamation of Laurentia {Nuna).

144

Supercontinents and Earth Evolution Symposium 2005

CONCLUSION The cover sequences of the southern Superior and western Churchill provinces exhibit firstorder sequence stratigraphic similarities that indicate a shared history of stable continental sedinnentation followed by rifting, basin opening and basin closure. These sequences thus

represent a complete Wilson cycle of supercontinent break-up/ amalgamation. Inferences concerning their relative positions within these supercontinents must await geochronological and paleomagnetic studies of key stratigraphic horizons and mafic magmatic events.

145

Supercontinents and Earth Evolution Symposium 2005

DETRITAL ZIRCON GEOCHRONOLOGY OF THE AUSTRALIAN MARINOAN GLACIAL INTERVAL Timothy D Raub\ David AD Evans\ Michael TD Wingate^, Clive R Calver^, Catherine F Izard^ Vale University, PO Box 208109, New Haven, CT, 06520, USA (timothy.raub@yale.edu) ^Tectonics Special Research Centre, School of Earth and Geographical Sciences, University of Western Australia, Crawley, WA, 6009, Australia ^Mineral Resources Tasmania, PO Box 56, Rosny Park, Tasmania, 7018, Australia

Within the last year, suggested estimates for the age of the Marinoan glaciation have ranged from 635 to 580 Ma, based on correlations to the type sections in the Flinders Ranges of South Australia. A detrital zircon suite from the highest sandstone in a non-conglomeratic Elatina section at Burr Well, northern Flinders Ranges, yields an essentially identical provenance indication as that previously reported for Marino arkose, a gritstone unit defined lithologically and identified (or presumed by the first author) to lie beneath conglomeratic Elatina facies within the depocentre and near the top of the glaciogene succession on the Stuart shelf. This common provenance indicates pre- and post-glacial input of detritus from a "Grenville"-aged source (plausibly the Musgrave orogen or Warakurna large igneous province) and the adjacent Gawler craton, as well as a hitherto unrecognized -650 Ma source. Both Marino arkose and "Burr Well sandstone" signatures are distinct from the dominantly Gawler provenance inferred from reported detrital zircons in an erosive, flaser-

146

bedded, sandy unit intimately associated with the base of the Nuccaleena cap dolostone at the Ediacaran GSSP. A detrital zircon suite from a similar stratigraphic position, at the top of Cottons Breccia on King Island, yields an entirely different signature, dominated by Geon 6 zircons from a mysterious but presumably juvenile and proximal terrane. Altogether, and through May 2005 with support of only a limited paleomagnetic dataset, these geochronological diagnoses support, but do not require, a -635 Ma age for type-Marinoan glaciation. They permit, but do not particularly support, a -580 Ma age for the Marinoan succession as recently indicated for Tasmania based on stratigraphic correlations to the Flinders Ranges. Better dating of the type Marinoan sections, which contain the world's most robust paleomagnetic record of low-latitude glaciation, will be crucial for unravelling the nature of late Precambrian ice ages.

Supercontinents and Earth Evolution Symposium 2005

GONDWANA FORMATION: A METAMORPHIC VIEW FROM THE MARGINS OF THE CONGO CRATON Volker Schenk, Peter Appel, Niels Jons, Timm John, Denny Loose Institutfur Geowissenschaften, Universitat Kiel, Germany (vs@min.uni-kiel.de)

INTRODUCTION The Congo craton has a central position in reconstructions of the Gondwana supercontinent and is surrounded by transcontinental orogenic belts thought to have formed during Pan-African orogenic events between 650 to 500 Ma: The Central African Fold Belt (CAFB) in the north, the East African Orogen (EAO) in the east, the Zambezi-Lufilian-Damara orogenic system (ZB) in the south and the Brasiliano Belt in the west. To decipher the timing of subduction processes and collisional events around the Congo craton, we studied the metamorphism of eclogites (ZB and CAFB), granulites (EAO, CAFB), and whiteschists (ZB, EAO). We consider the continental reworking of the margins of the craton as well as the burial of Neoproterozoic cover series, in addition to subduction metamorphism of oceanic crust. THE NORTHERN MARGIN (CAFB) Western end of CAFB Paleoproterozoic basement and Neoproterozoic cover series in Cameroon experienced a common granulite facies metamorphism at about 602-631 Ma, due to substantial crustal thickening. Retrogressed eclogites with MORBtype chemistry occurring near the NW-edge of the Congo craton may indicate the so far unknown site of a suture in the CAFB. Eastern end of CAFB The West Nile area forms part of the northern margin of the Congo craton in Uganda, and is overlain by Neoproterozoic cover series. Crustal thickening is evident by burial of these cover series to depths of 25-30 km (at 630 Ma) and by uplift of 2.4 Ga granulites subsequent to a PanAfrican granulite-facies overprint. THE EASTERN MARGIN (EAO) The Basement Complex of Uganda and the Tanzania craton can be seen as the eastern margin of the Congo craton along the EAO. The cratonic crust in NE Uganda experienced a strong Pan-African reheating (633 ± 11 Ma) resulting in UHT-metamorphism (>1000°C) over a large area (>2000 km^) that was followed by near-isobaric cooling prior to uplift.

Similarly, the Archean crust of the craton reworked in the EAO of Tanzania experienced an anticlockwise P-T path between 650-610 Ma pointing to a strong crustal heating prior to collision. In contrast, whiteschists near the western border of the EAO (Mautia Hill) close to the Tanzania craton were formed during crustal thickening at about 540 Ma, i.e. about 70 million years after the peak temperatures had been attained in most of the EAO of Tanzania. The Bemarivo belt of northern Madagascar experienced crustal thickening at about 530 Ma when it became attached to the Antananarivo block. So far, the only MORB-type chemistry of mafic rocks in the EAO of Madagascar was found in the Vohibory block. However, these rocks are of amphibolite grade and do not show traces of HP metamorphism. THE SOUTHERN MARGIN (ZB) Eclogites of the central Zambezi Belt display MORB-type geochemistry and were metamorphosed at -650-600 Ma. In contrast, the whiteschists within the Lufilian Arc-Zambezi Belt formed under high-pressure amphibolite facies conditions at -530 Ma. Their metamorphism was linked to the collision rather than to subduction zone processes. The crustal thickening affected the entire Lufilian Arc-Zambezi Belt almost simultaneously -70 million years after subduction of the oceanic basin. CONCLUSION The collision events along the northern margin of the Congo craton (CAFB in Cameroon and Uganda) happened nearly synchronously at 630 Ma and were preceded by eclogite formation of so far unknown age. Eclogite metamorphism at the southern margin (ZB) was at 650-600 Ma and pre-dated collision and whiteschist formation at 530 Ma for about 70 million years. The eastern margin of the Congo craton (EAO) in Uganda and Tanzania experienced a anticlockwise P-T evolution and very strong heating (650-610 Ma) prior to crustal thickening and whiteschist formation at 540 Ma. A slightly younger collision (530 Ma) is only known from the Bemarivo Belt in northern Madagascar.

147

Supercontinents and Earth Evolution Symposium 2005

CRUSTAL STRUCTURE AND PHANEROZOIC OROGENIC PROCESSES ON AND NEAR THE SOUTHERN MARGIN OF BALTICA Randell Stephenson Netherlands Research Centre for Integrated Solid Earth Sciences, Vrije Universiteit, Faculty of Earth and Life Sciences, De Boelelaan 1085, 1081 HV Amsterdam, Netherlands (randell.stephenson@falw.vu.nl)

The location and tectonic style of the southern margin of Neoproterozoic Baltica is poorly determined as a result of an extremely complex subsequent geological history involving several phases of Phanerozoic orogenesis, much of it involving extensional tectonic processes and the formation of thick sedimentary basins overlying crustal basement. From regional tectonic considerations, combined with what can be inferred about crustal structural affinity, it seems likely that Neoproterozoic Baltica included the crust of all of what is referred to as the Scythian Platform in southern Ukraine and Russia and very likely underlies most or all of what is now the Peri-Caspian Basin. To the (present-day) south-west it likely extended beyond North Dobrogea (south-east Romania-Ukraine) to include the northern part of the Moesian platform (Romania-Bulgaria) crust, on a margin that was subsequently developed during Rodinia breakup. The likely affinity of the crust of the Istanbul Zone (western Pontides) in western Turkey with that of the southern part of the Moesian Platform indicates that the eastern prolongation of this part of the Baltic margin lay somewhere within

148

the present day western Black Sea Basin or its northern shelf. Although the nature of the basement of the Crimean Foldbelt is unclear, the thinned crust of the Mid-Black Sea Rise may have been contiguous with Baltica since at least the Neoproterozoic, given its correlation with the Dzirula terrane of the western Georgian Transcaucasus. An important consideration in all of this is the evident lack of any significant crustal accretionary phases during the (Late) Palaeozoic (i.e., Variscan or Hercynian orogenesis) in this area, in contradiction to conventional thinking during the last decades. The dominant tectonic process affecting all or much of southern Baltica in the Late Palaeozoic was actually rifting and this was also the case during much of the Mesozoic. The main crustal shortening phase on the Baltica margin and, indeed, across the entire Alpine-Tethyan belt in this area occurred in the Late Cretaceous and Tertiary. This also resulted in thick-skinned, intracratonic, sedimentary basin inversion, as demonstrated by deep seismic images and new geological studies of the Donbas Foldbelt in southern Ukraine.

Supercontinents and Earth Evolution Symposium 2005

COOL CRATONS AND THERMAL BLANKETS: HOW CONTINENTS AFFECT MANTLE CONVECTION Valery P Trubitsyn^, Walter D Moonev^, Dallas H Abbott^ ^United States Geological Survey, 345 Middlefield Rd., MS 977, Menio Park, CA 94025, USA \amont-Doherty Earth Observatory, Palisades, NY 10964, USA

Mantle convection models with moving continents show that continents profoundly affect the form of mantle convection. If the continents are wider than the wavelength of convection cells (^3000 km), they cause neighbouring thermal upwelling zones to coalesce into a single focused upwelling. This focused upwelling zone will have a potential temperature anomaly of about 200 degrees C, much higher than the 100 degree C temperature anomaly of upwelling zones generated beneath typical oceanic lithosphere. This excess temperature anomaly

persists for about 100 Ma after supercontinent breakup. In contrast, small continental blocks (<3000 km diameter) do not induce focused upwelling zones. Instead small continental blocks are dragged to mantle downwelling zones. As a result of emplacement over relatively cold mantle (downwellings), small continental blocks develop thick, cold thermal boundary layers (cratonic roots), and are geologically favoured to keep these roots and to have low geothermal gradients.

149

Supercontinents and Earth Evolution Symposium 2005

LA-ICPMS U-Pb ZIRCON GEOCHRONOLOGY OF NEOPROTEROZOIC IGNEOUS ROCKS FROM NORTHERN GUANGXI, SOUTH CHINA: IMPLICATIONS FOR PETROGENESIS AND TECTONIC EVOLUTION Xiaolei Wang, JinchengZhou, Jiansheng Qiu Department of Earth Sciences, Nanjing University, Nanjing 210093, PR China (xlwangnju@yahoo.com.cn)

The western end of the Proterozoic Jiangnan orogen is located in northern Guangxi, South China. Neoproterozoic S-type granites are dominant (>90%) in the area, with ca. 8% being mafic-ultramafic rocks. The generation of these igneous rocks was previously considered to be related to a mantle plume (or superplume) event that led to breakup of the Rodinia supercontinent. In this work, we present new laser ablation-ICPMS U-Pb zircon data for the basic-acid rocks from Northern Guangxi. The new ages for the Zhaigun, Bendong, Dongma, Sanfang and Tianpeng granitic plutons are 834.6 ± 8.4, 822.7 ± 3.8, 824 ± 13, 804.3 ± 5.2 and 798 ± 12 Ma, respectively, and the Hejiawan layered diabases are 811.5 ± 4.8 Ma. These ages indicate a long duration of magmatic activity (ca. 35 million years), inconsistent with plume models that predict widespread magmatic eruption and emplacement within period of 1-5 million years. A total of 11 spot analyses of zircons gave inherited early Neoproterozoic ages ranging from

150

ca. 870 to 950 Ma. These ages might record subduction or collision-related magmatic events between 950-870 Ma in Northern Guangxi. Combined with previous geochronological and geochemical data, our new dating results favour post-collisional extension, rather than a mantle plume or superplume model, for the genesis of 835-800 Ma granites and basic rocks in Northern Guangxi. The upwelling of deep mantle due to detachment of a subducted slab and delamination of the lithosphere might cause partial melting of the continental crust to generate S-type granites. The continentcontinent collisional orogenic event along the Jiangnan orogen may have spanned ca. 70 million years from 870 to 800 Ma, and the early Neoproterozoic subduction might have lasted for ca. 130 million years. It is proposed that South China might alternatively have been located at the western margin of the Rodinia supercontinent during the ca. 870-800 Ma interval.

Supercontinents and Earth Evolution Symposium 2005

TWO LARGE IGNEOUS PROVINCES IN LATE MESOPROTEROZOIC AUSTRALIA Michael TP Wingate\ Paul A Morris^, Franco Pirajno^, Robert T Pidgeon^ ^Tectonics Special Research Centre, University of Western Australia, Crawley, WA 6009, Australia (mwingate@tsrc.uwa.edu.au) Geological Survey of Western Australia, 100 Plain St., East Perth, WA 6004, Australia ^Dept of Applied Geology, Curtin University of Technology, Perth, WA 6845, Australia

Large igneous provinces (LIPs) record rapid ennplacement (within a few million years) of enormous volumes (typically millions of km ) of intrusive and extrusive rocks of mainly mafic composition. We have employed high-precision U-Pb geochronology, geochemistry, and paleomagnetism to delineate two previously unrecognised LIPs of late Mesoproterozoic age that extend over large areas of western Australia. The Warakurna LIP consists of coeval mafic igneous rocks emplaced rapidly over .5 million km^ in western and central Australia. SHRIMP U-Pb dating of rocks separated by up to 1500 km indicates that magmatism occurred mainly between 1078 and --1070 Ma. The Warakurna LIP includes layered mafic-ultramafic intrusions and mafic to felsic volcanic rocks and dykes in central Australia, a 1000-km-long mafic sill province in Western Australia, and several swarms of mafic dykes. The large areal extent and short duration imply emplacement above a mantle-plume head. Despite their wide separation, the mafic rocks have similar MORBnormalized trace element patterns and rare earth element characteristics. West-directed paleocurrents, westward radiating dike swarms, and the occurrence of high-Mg rocks indicate that the center of the Warakurna plume head was located beneath central Australia. The newly-recognized Marnda Moorn LIP includes extensive swarms of mafic dykes concentrated around the margins of the Archean Yilgarn craton, and also some that extend well into the craton interior. 'Marnda Moorn' means 'black rocks' in the language of the Balladong people, who used these dyke rocks in trade as a form of currency. Recent SHRIMP U-Pb geochronology suggests that most of these dykes are similar in age at 1210 Ma. Most are dolerite or gabbro, although more felsic compositions, such as quartz diorite, are also present. In the southeastern Yilgarn, the NE-trending Fraser swarm is known only from aeromagnetic images and a single mine exposure of a dyke

with a baddeleyite age of 1212 ± 10 Ma. The Fraser swarm is likely continuous, beneath regolith, with the Gnowangerup swarm, which subparallels the southern Yilgarn margin, and within which two dykes furnished zircon ages of 1203 ± 15 Ma and -1238 Ma. A dyke in the Darling Scarp near Perth, and another 100 km inland, yielded zircon ages of 1214 ± 5 and 1204 ± 10 Ma, respectively. These two intrusions are part of the Boyagin swarm, which extends along the western Yilgarn margin, and includes strongly uralitized dykes in a wide variety of orientations. About 200 to 300 km NE of Perth, the Wheatbelt swarm contains poorly-exposed, roughly E-trending dykes with zircon ages of 1202 ± 13, 1209 ± 8, 1215 ± 11, and 1216 ± 11 Ma. Finally, in the far northwest corner of the Yilgarn craton, two ENE-trending dykes yielded preliminary zircon ages of 1207 and 1211 Ma. The dated intrusions have ages between 1202 and 1216 Ma, and an average age of 1210 Ma. They extend over at least 400,000 km^ in the western and southern Yilgarn craton; whether any occur in the northeastern Yilgarn is unknown. Interestingly, dykes of this age have not been reported from the Capricorn orogen or the Pilbara craton. Although the wide distribution and short duration of the Marnda Moorn event is consistent with a plume origin, the apparent concentration of dykes around the present craton margins implies that plate boundary stresses at 1210 Ma were a major influence on dyke emplacement. The Marnda Moorn event coincided with the second stage of tectonothermal activity in the Albany-FraserMusgrave orogen. Plate reorganisation at about 1.2 Ga is indicated also by a major bend in the apparent polar wander path for Australia. The 1075 Ma Warakurna LIP is significantly younger than the 1110 Ma Umkondo and Keweenawan LIPs, and the 1210 Ma Marnda Moorn LIP is not similar in age to any other major mafic event recognised so far. All four appear to be examples of LIP emplacement during assembly of the Rodinia supercontinent.

151

Supercontinents and Earth Evolution Symposium 2005

WORKSHOP: ISSUES WITH THE PRECAMBRIAN TIME SCALE 25 SEPTEMBER 2005

152

Supercontinents and Earth Evolution Symposium 2005

ISSUES WITH THE PRECAMBRIAN TIME SCALE: THE STATUS QUO AND BEYOND Wouter Bleeker Geological Survey of Canada, 601 Booth Street, Ottawa, Ontario, K1A 0E8, Canada (wbleeker@nrcan.gc.ca)

From the genesis of planet Earth (ca. 4.55 Ga) to the emergence of shelly fossils at the beginning of the Cambrian Period (ca. 542 Ma), the Precambrian spans 88% of Earth's history. Yet, there is no complete and coherent time scale and nomenclature for this larger part of the history of our planet. The present time scale (Fig. 1) is incomplete, partly informal, used inconsistently and, as some would argue, fundamentally flawed because of its reliance on boundaries defined by absolute ages [e.g., 1]. Due to small but non-trivial uncertainties in decay constants, even for the U-Pb decay system, it is strictly impossible to pin-point externally defined absolute age boundaries in extant stratigraphic sections with any certainty. For instance, at ca. 2500 Ma, typical fuzziness due to decay constants is +/- 6.5 Ma. In agreement with ICS^ objectives to define all major boundaries in terms of GSSPs^, the Precambrian time scale should be redefined in terms of the extant rock record and first-order events in Earth history. The final construct should be a comprehensive and "natural" time scale that 1) is pragmatic, 2) allows for easy and stable communication between scientist from different disciplines (e.g., early Earth researchers, planetary scientists, Precambrian stratigraphers), and 3) honours many of the firstorder events and transitions in the history of our planet. The early part of the time scale should be in agreement with the latest insights from planetary science. Such a natural time scale would be more intuitive and help convey the dynamic history of planet Earth, to scientists and non-scientists alike. A new international Subcommission on Precambrian Stratigraphy has been established, tasked with completing the time scale and advancing proposals for GSSP-based boundary definitions. General requirements for Precambrian GSSPs are that they reflect firstorder stratigraphic events. If possible, these

^ International Commission on Stratigraphy, a body of lUGS. ^ Global boundary Stratotype Section and Point, also known informally as "golden spikes". Workshop: Issues with the Precambrian Time Scale

events should be global in expression. And the selected boundaries and stratotype sections should lend themselves to precise dating, ideally by multiple methods. Boundaries will become fixed in the rock record, whereas their calibration in absolute time will involve an inherent uncertainty and may even change in detail if decay constants improve. Precisely dated boundaries may be correlated worldwide by using zircon geochronology or other methods, as long as similar methodology is employed in different locations (e.g. intercept ages). It is acknowledged that some potential boundaries may be largely symbolic, with little prospect for global correlation. An example would be a formal Hadean-Archean boundary defined by the first appearance (preservation) of supracrustal rocks in the geological record. Overall, the evolution of planet Earth divides itself into six natural eons: • •

•

An early eon of accretion and differentiation. A Hadean Eon, starting with the Moonforming giant impact, a tailing heavy bombardment, vigorous internal convection, and little remaining rock record. The Archean Eon, starting with the first appearance of a meaningful supracrustal record and culminating with a period of rapid crustal growth. A transition eon, starting with giant BIFs in regionally extensive and little deformed platform sequences and culminating with the progressive oxygenation of the atmosphere. A (shortened) Proterozoic Eon characterized by a geodynamically and geochemically almost modern Earth, its onset defined by the first appearance of unequivocal terrestrial red beds (--2.2 Ga). The Phanerozoic Eon, characterized according to its classical definition by the appearance of obvious metazoan life forms.

Replacing the present time scale with a GSSP-based time scale could be implemented without major change to terminology that has at least been partly accepted. An interesting question is whether the newly defined Ediacaran

153

Supercontinents and Earth Evolution Symposium 2005

Period should be part of a downward extended Phanerozoic Eon (i.e. "obvious life") or remain in the Proterozoic Eon.

Eon

REFERENCES [1] Bleeker, W., 2004. Towards a 'natural' tinne scale for the Precambrian—A proposal. Lethaia, vol.37, p. 219-222. [2] Plumb, K.A., 1991. New Precambrian time scale. Episodes, vol. 14 (2), p. 139-140.

Era (Bast^ of Cambrian)

Period 542

542

Nmproterozoic

HBOprotBrozoiC Ut

g^g

Cryogen/an

^^^

1000- Toman MesoprotBrozoic

\ Ecfas/ao Calymmian

Proterozoic

1600-; i

1400 1600 -1800

Orosifmn

Sca/e: 1200 m.y.

Stathenan

. 1200h

P^lmpfOtmozQic

2050 Hhyac'mn

2500

Siderian

2300 "2500"

Hmarchean

2800i fSBSoarthBan

3200

Archean Paleoarchean

No furthm siibdmsions irm pmods

3600 EoarchBan

• I Ages in Ma

Figure 1: The "status quo". The present chronometrically defined Precambrian time scale (after [2]). The Archean and its subdivisions remain informal, with no defined beginning. The commonly used Hadean eon has been added here for completeness but was not part of the 1991 proposal. The Neoproterozoic III has been redefined as the Ediacaran Period.

154

Workshop: Issues with the Precambrian Time Scale

Siipercontinents and Earth Evolution Symposium 2005

USING OXYGEN ISOTOPE "STRATIGRAPHY" TO DEFINE THE ONSET OF THE ARCHAEAN IN THE ABSENCE OF A ROCK RECORD Aaron J Cavosie\ Simon A Wilde^, John W Valley^ ^Department of Geology and Geophysics, University of Wisconsin, Madison Wl, USA ^Department of Applied Geology, Curtin University of Technology, Perth, Australia

There is growing opinion in favour of determining Precambrian boundaries using chronostratigraphic principles (placing them at events or transitions in preserved rock sequences of known geological age) rather than the generally accepted use of chronometric methods (defined in years based on geochronology and without specific reference to the rock record). This is despite the fact that applying the former would preclude the first 500 million years of Earth's history, for which there is no known rock record. This has fundamental implications for placing the onset of the Archaean, which is currently undefined. Proposals to define this boundary based on the extant rock record are flawed because: (i) they would force it to be located at a time younger or equal to the oldest known rocks (currently the 4.03 Ga Acasta Gneiss) and (ii) the rock record is merely a series of 'snapshot' occurrences through geological time, that becomes more fragmentary with age. Furthermore, whether or not individual rock outcrops record the beginning or ending of any fundamental process or event (i.e. boundary conditions) on the early Earth cannot be evaluated - but seems extremely unlikely. It is against this backdrop that we propose an innovative chemostratigraphic method for defining the beginning of the Archean. It is based on the condensation of the early steam atmosphere, an event of unquestionable importance and world-wide synchroneity, which resulted in a chemical signature recorded in the ancient zircon record. The oxygen isotope ratio of magmatic zircon is a robust recorder of the oxygen isotope composition of the parental magma. As no significant reservoir higher in than peridotite (e.g. whole rock = - 5 . 5 % o ) is known in the Earth's mantle, magmas in equilibrium with the mantle crystallize zircons that have a narrow range of = 5.3±0.3%o. Significant deviation from 'mantle zircon' values towards higher in magmatic zircon can only result if the parental magma incorporates higher material (e.g. supracrustal rocks) through melting or

Workshop: Issues with the Precambrian Time Scale

assimilation. Hence, a high value in zircon (e.g. above ~6.0%o) is a clear signature of melting crustal rocks that have experienced lowtemperature alteration by liquid water. Our data obtained from the oldest known crystals on Earth at Jack Hills, Western Australia, indicate that from 4400 to 4325 Ma, zircons record mantle values of 5.4±0.4%o (n=2), and from 4325 Ma to 4200 Ma zircons preserve mildly elevated values from 6.3 to 6.5±0.4%o (n=3). After 4200 Ma, many zircons with values as high as 7.3±0.3%o occur, identical to the maximum value of 7.5%o measured in zircons from >120 known Archaean igneous rocks. The oxygen isotope ratio of detrital igneous zircons from Jack Hills thus record a fundamental change in the chemistry of terrestrial magmas at M200 Ma. We therefore propose that the lower boundary of the Archean be provisionally set at 4200 Ma, based on the changes of igneous zircons, and hence magma compositions, as a consequence of interaction with liquid water on the Earth's surface. The dramatic rise in the range of to 7.3 %o for igneous zircons at 4200 Ma signifies the recycling of supracrustal rocks into magma, and the unambiguous presence of water on Earth by this time. A boundary at 4200 Ma not only reflects a change in igneous processes but may also signify stabilization of the Earth's continental crust, as evidenced by its weathering, alteration and erosion. The chemostratigraphy recorded in ancient igneous zircons is the only record of early surface cooling, and highlights the potential of using global chemostratigraphic changes to identify other significant Precambrian events for defining important time boundaries. Applying the features identified above, the Archaean would be defined as the first eon characterized by surface waters, supracrustal rocks, crustal recycling, and incorporation of crustal rocks in igneous magmas on a grand scale. This change also marks the earliest period when Earth was hospitable to life, and its onset is considered one of the most profound events in early Earth history.

155

Supercontinents and Earth Evolution Symposium 2005

ISUA, "GOLDEN SPIKES" AND THE LACK OF ZIRCONS: CONSTRAINING THE EARLIEST VOLCANO-SEDIMENTARY RECORD Allen Nutman Research School of Earth Sciences, The Australian National University, ACT 0200, Australia (allen.nutnnan@anu.edu.au)

There is an understandable desire and need to have the oldest parts (>3600 Ma) of Earth's volcano-sedimentary record dated and understood via a series of "golden spikes" to the same degree as for the Phanerozoic - for the same reasons of charting evolution of life, the hydrosphere and atmosphere. Indeed, the scope of modern zircon dating technologies means it should be possible to obtain accurate dates, and with the will and resources (=$$), to have precisions of only 1 million years or even less attached to them. Instead, the problem in defining the most ancient volcanic and sedimentary timescale comes from the rock record itself via (1) its tectonothermal complexity (all rocks are amphibolite-granulite facies tectonites), (2) the rarity of volcano-sedimentary rocks bearing zircons that give the actual time of deposition and (3) dispute over the protoliths (volcanic or intrusive) of some zircon-bearing rocks. In the 1980s mapping of the Isua belt suggested that it contained one or two sequences, disrupted by early tectonic breaks (Nutman et al., 1984), and there was a broad acceptance that the then available zircon dates of 3800 Ma (Michard Vitrac et al., 1977; Baadsgaard et al., 1984; Compston et al., 1986) on a single outcrop of a single unit interpreted to be a felsic volcanic could be taken as the age of the whole belt. Since then, this 1980s "golden spike" has been thrown into doubt by questioning whether the dated unit is a volcanic or (altered) intrusion, (e.g., Myers, 2001). New zircon dating in the 1990s showed that the belt contains ca. 3710 and >3790 Ma

volcano-sedimentary rocks (Nutman et al., 1996, 1997, 2002), and new field studies by different groups have all concluded that the belt is partitioned by early (3600 Ma?) tectonic breaks although there is yet to be consensus as to the details of the tectonic architecture (Nutman et al., 1997, 2002; Komiya, 1999; Appel et al., 1998). These new findings mean that there is a need for more zircon dating to define the age of the Isua rocks - yet suitable units are hard to come by given the predominance of amphibolites derived from pillow basalts. Presently the unit with the most robust, accurate date are some graded felsic volcanic rocks in the eastern end of the belt - which yields only 3710 Ma zircons (Nutman et al., 1996, 1997, 2002; B. Kamber pers comm., 2004). These are the present best candidate for "golden spike" status in the earliest timescale. Rocks with potential "golden spike" status are cherts and banded iron formations in the east of the belt which yield small amounts of 3700 Ma zircon that we interpret to be from small input of airborne volcanic material (Nutman et al., 2002). This still leaves most of the earliest volcano-sedimentary record dominated by amphibolites from Isua and elsewhere without accurate dates. Instead, tonalite dykes that intrude them provide a minimum age. Thus amphibolites in the southern part of the belt are intruded by ca. 3790 Ma tonalities (Nutman et al., 1996, 1997, 2002; Crowley et al., 2002) and must be at least 80 million years older than the felsic volcanic 3710 Ma "golden spike" candidates in the east of the belt.

156

Workshop: Issues with the Precambrian Time Scale

Supercontinents and Earth Evolution Symposium 2005

ISSUES WITH THE PRECAMBRIAN TIME SCALE: WESTERN AUSTRALIA GEOLOGY AND GSSPs Martin Van Kranendonk Geological Survey of Western Australia, 100 Plain Street, East Perth, WA 6004, Australia (nnartin.vankranendonk@doir.wa.gov.au)

To GSSP, or not to GSSP7 That is the question regarding division of the Precambrian timescale. Can it be done? Some would argue that without a reliable biostratigraphic record, Phanerozoic timescale concepts cannot be applied to the Precambrian. Are there sufficient, datable Precambrian GSSPs that truly reflect global events? What are the global events? Are they truly global? If GSSPs are impractical, what combination of GEONS, chrono-stratigraphy, orogenic cyclicity, or a combination of all of the above, and possibly more, will need to be applied? One of the best areas in the world to test the GSSP concept for Precambrian time is Western Australia, where a nearly continuous stratigraphic record is preserved from 3515- 490 Ma. In the Pilbara Craton, the East Pilbara Granite-Greenstone Terrane consists of four unconformity-bound groups (Pilbara Supergroup: PS) deposited from 3515-3426, 3350-3310 Ma, 3270-3235 Ma, and 3235-3000 Ma, the last being undated. Are these depositional (and hiatal) periods reflected on other cratons and what exactly do they represent? Controversy exists, but it has been proposed that the craton records a change in tectonic style from vertical, plume-dominated tectonics to horizontal subduction-accretion tectonics at ca. 3200 Ma [1, 2] and could be used to define the early Archaean-mid Archaean boundary. But does this represent a global change in tectonic style? More importantly, this change is not reflected in a single stratigraphic section, but preserved in disparate tectonic elements separated by major tectonic boundaries. Following late tectonic clastic sedimentation (2970-2930 Ma De Grey Supergroup) and emplacement of 2890-2830 Ma "post-tectonic" granites, the Pilbara Craton was unconformably overlain by the 2775-2445 Ma Mount Bruce Supergroup of the Hamersley Basin [3]. This comprises three groups: three unconformitybounded flood basalt units and interbedded

^ Global boundary Stratotype Section and Point, also known informally as "golden spikes".

Workshop: Issues with the Precambrian Time Scale

sedimentary rocks of the 2775-2630 Ma Fortescue Group; banded iron-formation, shale, dolomite and volcanic rocks of the 2630-2445 Ma Hamersley Group; and undated, unconformably overlying clastic rocks of the <2445 Ma, >2200 Ma Turee Creek Group. The Fortescue and Hamersley Groups record continuous deposition over 330 million years and span the transition to an oxygenated atmosphere. One could argue that the unconformable base of the Fortescue Group can be used as a GSSP for the start of the Late Archaean, as the 2775 Ma age reflects the beginning of magmatic events on most cratons. However, unconformities are inconsistent with GSSP practice. The onset of BIF deposition in the Hamersley Group, at ca. 2600 Ma (Marra Mamba Iron Formation), is too old, however, to represent the end of the Archaean, as several cratons record events to ca. 2500 Ma. However, a younger period of BIF deposition, from ca. 2500-2450 Ma (Brockman Iron Formation), might be used to this end, as much of this succession is well exposed in outcrop or accessible in drill core. A series of Proterozoic basins unconformably overlie the Pilbara and Yilgarn Craton including the c. 2200 Ma Yerrida Basin, the c. 1900 BryahPadbury Basins, the 1840 Ma Earaheedy Basin, the 1650-1070 Ma Bangemall Basin, and the 840-490 Ma Officer Basin. Imprecise dating leaves it unclear as to whether these basins contain continuous sections through major global events, or, if so, where GSSPs may be placed. Although new techniques [4] may help refine the ages of depositional events, it probably will not aid the current project. My personal view is that the GSSP concept will not work for the Precambrian and that a division based on geochronology of global tectonic events, in combination with lithostratigraphy, is more appropriate. REFERENCES [1] Van Kranendonk eta!., 2002. Econ. Geo!. 97, 695732. [2] Smithies et a!., 2005. Earth & Planet. Sci. Lett. 231, 221-237. [3] Trendall, A.F. et a!., 2004. Aust J Earth Sci, 51, 621-644. [4] McNaughton et a!., 1999. Science 285, 78-80.

157

Siipercontinents and Earth Evolution Symposium 2005

SOME POTENTIAL PROTEROZOIC GSSPS David AD Evans Department of Geology & Geophysics, Yale University, PO Box 208109, New Haven, CT 06520, USA (dai.evans@yale.edu)

Fornnal adoption of a Precambrian chronostratigraphic tinnescale will require placements of Global Stratotype Sections and Points (GSSPs) at well chosen locations in the geological record (Bleeker, 2004a, b; Robb et al., 2004). Formalization of GSSPs in a chronostratigraphic framework does not require complete abandonment of chronometric scales; for example, the geon scale (Hofmann, 1990, 1999) may be used in parallel with any stratigraphically based scheme according to the convenience of scientific communication. Given wide usage of prior chronometric subdivisions in Precambrian time, GSSP selection should be guided by these subdivisions, at least at the levels of eons and eras (see Robb et al., 2004). Thus an Archean-Proterozoic boundary should be placed in a rock succession at approximately 2500 Ma age, and eras within the Proterozoic Eon should conform broadly to a three-part subdivision of Paleo-, Meso-, and Neo-, that correspond approximately with their current chronometric boundaries at 1600 and 1000 Ma (Plumb, 1991). The names of these eons and eras need not change, for we have ample precedent of consistent terminology usage both before and after GSSP formalization. The symbolic meanings of three Proterozoic eras can remain unchanged as long as our present chronostratigraphic definitions adhere to the features of Earth history that governed selection of the currently formalized chronometric tripartite subdivision (most thoroughly described by Plumb and James, 1986). If the present endeavour results in Proterozoic eras defined by different geological features and events, then new names may need to be sought for those conceptually new intervals of time. The Archean-Proterozoic transition encompassed several important irreversible changes to the Earth system, most notably the progressive cratonization of extant continental fragments (extensive dyke swarms, broad epicratonic cover successions, long and linear orogenic belts) and the onset of atmospheric oxygenation (earliest Superior-type banded ironformations, disappearance of mass-independent fractionation of sulfur isotopes). Other transitions, such as growth of the geomagnetic field and increases in biodiversity, may have occurred

158

during this interval but are presently too poorly known to constrain GSSP selection. As one of the most fundamental boundaries of geological time, with two of the most commonly used terms of the Precambrian interval, chronostratigraphic placement of the Archean-Proterozoic transition deserves the most respect for prior usage and definitions. A "transitional" eon inserted between Archean and Proterozoic (Bleeker, 2004b) is not favoured here, for it abandons this widespread terminology and would occupy an interval of time (about 300 Myr) that is grossly discordant to other intervals of eon status (all > 500 Myr). The best preserved and described stratified successions spanning this age are the Hamersley and Transvaal systems, respectively in Australia and South Africa, and the Huronian and Karelian successions, respectively in Canada and Russia. Each pair represents a different interval of time: the southern hemisphere deposits (in their more continuously exposed lower parts) span ca.2650-2450 Ma, whereas the northern hemisphere basins span ca.2450-2200 Ma (and younger in Karelia). If a GSSP is intended to approximate the existing Archean-Proterozoic chronometric boundary, then Australia or South Africa would be the most appropriate locales. One candidate GSSP would be the transition between underlying Gamohaan Formation carbonate into overlying Kuruman banded iron-formation (BIF) in the Kuruman Hills of South Africa; this section is well dated by U-Pb on interbedded volcanic ashes, constrained by carbon-isotope stratigraphy, and is part of an expansive regional exposure that allows litho-, sequence-, and chemo-stratigraphic correlation and can be well placed within a paleoenvironmental framework of deposition (e.g., Beukes, 1980, 1987; Klein and Beukes, 1989; Beukes et al., 1990; Halbich et al., 1992; Sumner and Bowring, 1996; Altermann and Nelson, 1998; Sumner and Grotzinger, 2004). Placing the GSSP at the base of the Kuruman iron-formation would also confine most of the world's Superior-type BIFs to the Proterozoic Eon, and, depending on the PaleoMesoproterozoic GSSP placement, probably also to the Paleoproterozoic Era. These Superior-type BIFs, which are most extensive in the geological record between 2470 and 1850

Workshop: Issues with the Precambrian Time Scale

Supercontinents and Earth Evolution Symposium 2005

Ma (Klein and Beukes, 1992), require stable continental platfornns for their extensive depositional areas, and at least indirectly manifest changes in atmospheric oxidation and/or biospheric development. Their widespread appearance in the rock record thus represents an appropriate choice for defining the Archean-Proterozoic transition. Atmospheric oxygenation was a protracted process that is indicated by different proxies at different ages (see Bekker et al., 2004), but all currently known examples would be confined to the Paleoproterozoic given the definition as recommended above. Defining boundaries between eras of the Proterozoic is made more difficult by the luxury of a greater pool for selection among well preserved stratified successions. In addition, the current chronometric definition of the PaleoMesoproterozoic boundary, at 1600 Ma, is meant to emphasize the rather subtle distinction between stabilization of cratons following supercontinent assembly (Statherian) and tectonic subsidence with epicratonic sedimentation (Calymmian). In the intervening years following formal definition, the gross distinction between Paleoproterozoic and Mesoproterozoic has come to be characterized as one separating a world of dramatic and exciting global events in the former era, from one of relative tectonic and climatic quiescence and subtle evolutionary changes, especially toward the beginning of the latter interval (see Brasier and Lindsay, 1998). If a Paleo-Mesoproterozoic GSSP is to conform to the concepts used in its sanctioned chronometric boundary, then it should be placed within a post-orogenic succession that represents the global aftermath of supercontinent assembly (with proposed names including Hudsonland, Nuna, Columbia, Capricornia, and perhaps others) at ca.1800 Ma. To my current knowledge, the best preserved rifted to passivemarginal successions deposited during and shortly after that age, in temporal order and with ages of initiation, are the Olifantshoek of South Africa (1930 Ma), Jixian of North China (1770 Ma), Espinhago of Brazil (1750 Ma), Coppermine of northern Canada (1740 Ma), McArthur-Lawn Hill of northern Australia (1730 Ma), circumSiberian margins (ca.1650 Ma), Bangemall of Western Australia (1640 Ma), Vindhyan of India (1630 Ma), and Riphean of Baltica (ca.1600 Ma). The most intensively studied among these is the McArthur-Lawn Hill superbasin, with diverse sedimentary fades that are amenable to study by numerous analytical methods. Within this expansive superbasin, a general evolution between synrift and postrift sedimentation is

Workshop: Issues with the Precamhrian Time Scale

found within the Gun Supersequence of Southgate et al. (2000). In the McArthur River area, this transition is recorded within the Masterton-Mallapunyah lithostratigraphic interval, well exposed in the gorge of the Kilgour River. To the southeast, in the Lawn Hill / Mt Isa region, a compromise between low metamorphic grade and stratigraphic completeness exists around the Fiery Creek dome, where the Gunpowder and Paradise Creek Formations are separated by the Mt Oxide Chert marker horizon (a potential GSSP level, ca.1660 Ma, Page et al., 2000). This level is advantageous because it can be traced almost continuously across a marine paleobathymetric slope toward the Mt Isa region (Southgate et al., 2000), allowing fuller characterization of oceanic geochemistry and paleobiology through the Paleo-Mesoproterozoic boundary interval (e.g., Shen et al., 2003). The Meso-Neoproterozoic chronometric boundary at 1000 Ma is most closely linked in Earth-system events to the culmination of Rodinia supercontinent assembly. Generally, the Neoproterozoic Era is characterized by more dramatic climatic and biological events than the preceding Mesoproterozoic interval, and GSSP selection should also reflect this aspect of surficial planetary evolution. Many of the world's best preserved volcano-sedimentary successions have ages either predominantly or completely on either side of 1000 Ma, with a uniquely notable exception of the Turukhansk in western Siberia. That composite section, comprising primarily carbonate rocks, has a well defined chemostratigraphy (Bartley et al., 2001) that permits correlations with other successions globally. As the currently defined Neoproterozoic Era contains an apparently monotonic rise (at present resolution) in marine 87Sr/86Sr carbonate values, as well as a general increase in the amplitude of 13C/12C carbonate oscillations (Shields and Veizer, 2002), a suitable level for the basal Neoproterozoic GSSP would be at the nadir of 87Sr/86Sr values in the Derevnya-Burovaya succession of the Turukhansk, with current age estimates of ca.1020 Ma (Bartley et al., 2001). Further subdivision of the Proterozoic Eon into chronostratigraphically defined Periods (i.e., with GSSPs) need not adhere to the chronometric subdivisions previously approved by ICS (Plumb, 1991), for some important updates to the ages of key rock successions and tectonic events may suggest alternative classifications of the various time intervals. For example, the formally defined Stenian Period at the end of the Mesoproterozoic Era has etymological basis in the "narrowing" of oceans that closed in concert with Rodinia assembly.

159

Supercontinents and Earth Evolution Symposium 2005

This aspect alone is appropriate for the 12001000 Ma interval, but increasing recognition of widespread large igneous provinces during the same period (e.g., Keweenawan, Umkondo, Warakurna) would suggest a more generally dynamic term stemming from the Greek "kinesis" as originally proposed by Plumb and James (1986). The following chronometrically defined period, Tonian (1000-850 Ma; Plumb, 1991), is named for "stretching" of the newly assembled supercontinent; however, the earliest widespread manifestations of Rodinia's fragmentation are now dated at ca.830 Ma (Li et al., 2003), thus most post-Rodinia rifting is ironically excluded from the Tonian as presently defined. Other chronometrically defined periods, such as Siderian, Orosirian, and Cryogenian, remain pertinent to current dating of representative rock successions and could be formalized chronostratigraphically by GSSP selection (in the case of Siderian this coincides neatly with my above-recommended basal-Proterozoic GSSP). REFERENCES Altermann, W., and Nelson, D.R., 1998. Sedimentation rates, basin analysis and regional correlations of three Neoarchaean and Palaeoproterozoic sub-basins of the Kaapvaal craton as inferred from precise UPb zircon ages from volcaniclastic sediments. Sedimentary Geology, v.120, p.225-256. Bartley, J.K., Semikhatov, M.A., Kaufman, A.J., Knoll, A.H., Pope, M.C., and Jacobsen, S.B., 2001. Global events across the Mesoproterozoic-Neoproterozoic boundary: C and Sr isotopic evidence from Siberia. Precambrian Research, v.111, p.165-201. Bekker, A., Holland, H.D., Wang, P.-L., Rumble III, D., Stein, H.J., Hannah, J.L., Coetzee, L.L., and Beukes, N.J., 2004. Dating the rise of atmospheric oxygen. Nature, v.427, p.117120. Beukes, N.J., 1980. Stratigrafie en litofasies van die Campbellrand-Subgroep van die Proterofitiese Ghaap-Groep, Noord-Kaapland. Transactions of the Geological Society of South Africa, v.83, p.141-170. Beukes, N.J., 1987. Fades relations, depositional environments and diagenesis in a major Early Proterozoic stromatolitic carbonate platform to basinal sequence, Campbellrand Subgroup, Transvaal Supergroup, southern Africa. Sedimentary Geology, v.54, p.1-46. Beukes, N.J., Klein, C., Kaufman, A.J., and Hayes, J.M., 1990. Carbonate petrography, kerogen distribution, and carbon and oxygen isotope variations in an Early Proterozoic transition from limestone to iron-formation deposition, Transvaal Supergroup, South

160

Africa. Economic Geology, v.85, p.663-690. Bleeker, W., 2004a. Towards a 'natural' time scale for the Precambrian~A proposal. Lethaia, v.37, p.219-222. Bleeker, W., 2004b. Toward a "natural" Precambrian time scale. In: Gradstein, P.M., Ogg, J.G., and Smith, A.G., eds., A Geologic Time Scale 2004 (Cambridge, Cambridge University Press), p.141-146. Brasier, M.D., and Lindsay, J.F., 1998. A billion years of environmental stability and the emergence of eukaryotes: New data from Northern Australia. Geology, v.26, p.555-558. Halbich, I.W., Lamprecht, D.F., Altermann, W., and Horstmann, U.E., 1992. A carbonatebanded iron formation transition in the early Proterozoikum of South Africa. Journal of African Earth Sciences, v. 15, p.217-236. Hofmann, H.J., 1990. Precambrian time units and nomenclature~The geon concept. Geology, v.18, p.340-341. Hofmann, H.J., 1999. Geons and geons. Geology, v.27, p.855-856. Klein, C., and Beukes, N.J., 1989. Geochemistry and sedimentology of a facies transition from limestone to iron-formation deposition in the Early Proterozoic Transvaal Supergroup, South Africa. Economic Geology, v.84, p. 1733-1774. Klein, C., and Beukes, N.J., 1992. Time distribution, stratigraphy, sedimentologic setting, and geochemistry of Precambrian iron-formations. In: Schopf, J.W., and Klein, C., eds.. The Proterozoic Biosphere (Cambridge, Cambridge University Press), p.139-146. Li, Z.X., Li, X.H., Kinny, P.D., Wang, J., Zhang, S., and Zhou, H., 2003. Geochronology of Neoproterozoic syn-rift magmatism in the Yangtze Craton, South China and correlations with other continents: evidence for a mantle superplume that broke up Rodinia. Precambrian Research, v.122, p.85-109. Page, R.W., Jackson, M.J., and Krassay, A.A., 2000. Constraining sequence stratigraphy in north Australian basins: SHRIMP U-Pb zircon geochronology between Mt Isa and McArthur River. Australian Journal of Earth Sciences, V.47, p.431-459. Plumb, K.A., 1991. New Precambrian time scale. Episodes, v.14, p.139-140. Plumb, K.A., and James, H.L., 1986. Subdivision of Precambrian time: Recommendations and suggestions by the Subcommission on Precambrian Stratigraphy. Precambrian Research, v.32, p.65-92. Robb, L.J., Knoll, A.H., Plumb, K.A., Shields, G.A., Strauss, H., and Veizer, J., 2004. The Precambrian: the Archean and Proterozoic

Workshop: Issues with the Precambrian Time Scale

Supercontinents and Earth Evolution Symposium 2005

Eons. In: Gradstein, F.M., Ogg, J.G., and Smith, A.G., eds., A Geologic Time Scale 2004 (Cambridge, Cambridge University Press), p.129-140. Shen, Y., Knoll, A.H., and Walter, M.R., 2003. Evidence for low sulphate and anoxia in a mid-Proterozoic marine basin. Nature, v.423, p.632-635. Shields, G., and Veizer, J., 2002. Precambrian marine carbonate isotope database: Version 1.1. Geochemistry Geophysics Geosystems, V.3, no.6, doi: 10.1029/2001GC000266. Southgate, P.N., Scott, D.L., Sami, T.T., Domagala, J., Jackson, M.J., James, N.P., and Kyser, T.K., 2000. Basin shape and

Workshop: Issues with the Precambrian Time Scale

sediment architecture in the Gun Supersequence: a strike-slip model for Pb-ZnAg ore genesis at Mt Isa. Australian Journal of Earth Sciences, v.47, p.509-531. Sumner, D.Y., and Bowring, S.A., 1996. U-Pb geochronologic constraints on deposition of the Campbellrand Subgroup, Transvaal Supergroup, South Africa. Precambrian Research, v.79, p.25-35. Sumner, D.Y., and Grotzinger, J.P., 2004. Implications for Neoarchaean ocean chemistry from primary carbonate mineralogy of the Campbellrand-Malmani Platform, South Africa. Sedimentology, v.51, p.1273-1299.

161

Siipercontinents and Earth Evolution Symposium 2005

PROGRESS TOWARDS THE FURTHER SUBDIVISION OF NEOPROTEROZOIC TIME Graham Shields School of Earth Sciences, James Cook University, Townsville, Queensland 4811, Australia (grahann.shields@jcu.edu.au)

Subdivision of Precambrian time based on globally recognisable geological events began with the banging in of the basal Ediacaran golden spike in South Australia earlier this year. The Nuccaleena Formation, hosting the golden spike, is a thin (<5 metre) dolostone unit that drapes glaciogenic deposits of the Elatina Formation laid down during the Marinoan phase of Neoproterozoic glaciation. It is the only GSSP that does not rely on biostratigraphy for its placement, and could act as a benchmark for future Precambrian subdivision. The choice of this level presumes that Nuccaleena-type "cap carbonates" can be correlated globally using their characteristic lithology and isotope stratigraphy. Although reports exist of carbonate units above Neoproterozoic diamictites at other levels, the central tenet of one unique "cap carbonate" event is consistent with existing biostratigraphic, geochronological and isotopic information. Firm age constraints based on U-Pb zircon dating of tuffs have been reported from Namibia and China, and indicate that the Ediacaran Period began around 632 Ma, which is consistent with more ambiguous age constraints from other locations, including Australia, Oman and Burkina Faso. These new age constraints have immediate implications for the further subdivision of the Ediacaran Period and definition of the immediately underlying "Cryogenian". Most importantly, glaciogenic strata that lack typical cap carbonate facies were clearly deposited both before and after this circa 635 Ma Marinoan event at 730 Ma (Sturtian) and 580 Ma (Gaskiers). It is not known whether these others were single events or of the areal extent and magnitude of the "Snowball Earth" Marinoan glacial episode but they will likely form important marker horizons for global stratigraphic correlation. Acritarch biostratigraphy holds great promise for subdividing the Ediacaran period (cf. K. Grey's abstract). A characteristic, large spiny acritarch assemblage first appears in South China just above the basal Ediacaran cap dolostone there, while similar microfloral

162

assemblages are known from post-630 Ma, but pre-550 Ma strata elsewhere in the world. Ediacaran-type soft-bodied fossils may also be of use as these, e.g. Charniodiscus, are known only from post-580 Ma strata in Newfoundland, Namibia, China and Australia, making that approximate level a favourite for subdividing the quite long (90 My) Ediacaran period. However, Precambrian biostratigraphy is still in its infancy. Given the ambitious time-frame set by the ICS, imminent subdivision of the Neoproterozoic cannot realistically be based on biostratigraphy. Defining the base of the "Cryogenian" Period represents a thorny problem. Absolute age constraints on the oldest of the Neoproterozoic ice ages are relatively poor, while the correlation of mid-Neoproterozoic glacial deposits is uncertain at best. Biostratigraphic information from Sturtian - Marinoan strata is rare and often stratigraphically ambiguous, so isotope stratigraphy will likely play a decisive role. 730 Ma - 640 Ma carbonates in Australia, Canada and Namibia exhibit unusually elevated •13C values, up to >10%o. Such high • 1 3 C values are central to existing correlation schemes, and with 87Sr/86Sr isotopes, are used to identify Cryogenian-equivalent strata in Mongolia, Scotland, Greenland, Svalbard, China, Brazil and USA. The "Tonian" "period", the first named interval of the Neoproterozoic, seems to lack any evidence for glaciation. This long (>250 Ma) chunk of Precambrian time is remarkable for its large drowned cratons covered in stromatolites, abundant molar-tooth structure and a characteristic assemblage of large organicwalled fossils, such as Chuaria and Tawuia. This interval appears to be typified by moderately, but not uniquely positive • 1 3 C values, while 87Sr/86Sr values remained below 0.7065 with minimal variation. Despite the potential of acritarchs such as Trachyhystrichosphaera for biostratigraphy, there does not yet seem to be sufficient distinction between Neoproterozoic and Mesoproterozoic strata to allow for any biostratigraphic base to the Neoproterozoic. One important aspect for the future will undoubtedly involve the study of Neoproterozoic

Workshop: Issues with the Precambrian Time Scale

Supercontinents and Earth Evolution Symposium 2005

glacial deposits worldwide. If the widely presunned hypothesis that these deposits are globally synchronous turns out to be correct, then this knowledge, along with improved geochronological constraints, will be crucial in

Workshop: Issues with the Precambrian Time Scale

establishing correlatable horizons that integrate sequence-, isotope, litho-, magneto- and biostratigraphy. This is a major goal of a recently established IGCP 512 project "Neoproterozoic Ice Ages", and of the Ediacaran Subcommission.

7 63

Supercontinents and Earth Evolution Symposium 2005

CORRELATING AND SUBDIVIDING THE EDIACARAN IN AUSTRALIA Kath Grey Geological Survey of Western Australia, 100 Plain Street, East Perth, WA 6004, Australia (kath.grey@doir.wa.gov.au) For nnany years, the lack of geochronology severely limited Australian Neoproterozoic correlations and stratigraphic interpretations had to be based mainly on lithostratigraphy. Although lithostratigraphic units showed a surprising uniformity across Australia, they were generally viewed with scepticism. In the 1990s, mineral and hydrocarbon exploration in the Centralian Superbasin and Adelaide Rift Complex, coupled with attempts to define a terminal Neoproterozoic system and period, prompted an integrated approach that tackled the problem from several perspectives. Current Ediacaran correlations use lithostratigraphy, event stratigraphy (recognition of key markers like glaciations and the Acraman impact ejecta layer), sequence stratigraphy, seismic interpretation, isotope chemostratigraphy, and biostratigraphy. Through these techniques, the Flinders Ranges succession can be correlated with successions across much of Australia. Lithostratigraphic correlations have been largely confirmed and further subdivisions appear feasible. The efforts of the working group on the terminal Proterozoic culminated in the ratification of a Global Stratotype Section and Point (GSSP) defining the base of the new Ediacaran System and Period. The GSSP is a lithostratigraphic marker backed by isotope chemostratigraphy at the base of the cap carbonate, (Nuccaleena Formation) overlying the diamictitic Elatina Formation in Enorama Creek. The successful defining of the Ediacaran System highlights important lessons for Proterozoic correlation and subdivision. For example, it is best not depend on a single approach, a principal advocated by Ken Plumb when the Precambrian Timescale was proposed in 1991. The consistencies demonstrated by the various lines of evidence strengthen the case for an integrated approach. Current correlations comprise a mosaic constructed from independent, but mutually supportive, data. Even if one method later shows flaws, the framework remains stable. Consequently, Australia-wide correlations can be viewed with considerable confidence.

164

Biostratigraphy has been largely disregarded as a Proterozoic correlation method, despite its extensive use throughout the Phanerozoic. Reasons cited for ignoring biostratigraphy include low species diversity, taxonomic uncertainty, morphological simplicity, sporadic distribution, and conservative evolution rates. However, it seems likely that Precambrian correlation schemes are lacking because palaeontologists have failed to adopt a systematic approach and have rarely tried to compile that standard biostratigraphic tool: the range chart. The Neoproterozoic palynological record is currently based on more than 2000 samples from at least 30 drillholes in the Adelaide Rift Complex, and Officer, Amadeus, and Georgina Basins. About half the samples are Ediacaran in age. To test the validity of a biostratigraphic approach, palynology and stable isotope analyses were conducted on the same samples. It is early days. Proterozoic palynology is at a similar stage to Phanerozoic palynology about forty years ago. However, the basic framework is there, and there is scope for refinement. Acritarchs (organic-walled, acid-insoluble, phytoplanktonic, polyphyletic microfossils) appear to be the most important fossils for correlation. They are ideal for biostratigraphy because they are abundant, have complex morphology, short stratigraphic ranges, and wide geographic and lithofacies distributions. Five zones have been identified to date. Samples from the diamictite, cap dolomite and immediately overlying mudstone succession are barren. Above this, simple spheroidal acritarchs (leiospheres) increase rapidly in actual numbers but not in species diversity (contrary to Snowball Earth expectations). A marked change takes place during the second post-glacial sea-level rise, with the first appearance of >50 species of large, acanthomorph (spiny) acritarchs. Diversification was rapid, but ended around the first appearance of bilaterians in the Ediacara fauna, c.565 Ma. The carbon isotope curve supports Ediacaran palynological interpretations across Australia (also the case for the Cryogenian).

Workshop: Issues with the Precambrian Time Scale

Supercontinents and Earth Evolution Symposium 2005

Global extension of the zonation scheme is not yet possible because ranges of key species have not been determined outside Australia. Nevertheless, there are species in common with Svalbard, Norway, Siberia, and China, indicating a potential for global correlation. However, attempts to match successions from the Australian Ediacaran and the Chinese

Workshop: Issues with the Precambrian Time Scale

Doushantuo Formation produces major inconsistencies. The anomalies apply to correlations based on both the biostratigraphy and carbon isotope curves and raise doubts about correlations of Neoproterozoic glaciations. More detailed studies are required to resolve these contradictions.

165

Supercontinents and Earth Evolution Symposium 2005

WHOLLY NUMERICAL DEFINITIONS OF MAJOR STRATIGRAPHIC BOUNDARIES? Alan G Smith Department of Earth Sciences, University of Cannbridge, Downing St, Cambridge, CB2 3EQ, UK (ags1 @esc.cam.ac.uk)

The comments in this presentation spring from involvement with preparation of the new Geologic Time-Scale produced in 2004 as a result of a major collaborative international effort involving over 40 contributors, supported by the International Commission on Stratigraphy. In particular, the volume highlights the major differences between methods currently used to define stratigraphic boundaries in the Phanerozoic and in the Precambrian. GLOBAL BOUNDARY STRATOTYPE SECTIONS AND POINTS ("GOLDEN SPIKES") The major divisions of the Phanerozoic timescale are defined by Global boundary Stratotype Sections and Points, i.e. GSSPs or "golden spikes", located in specific stratigraphic sections. The purpose of the golden spike method for defining stratigraphic boundaries is primarily to set a global standard for the definition in rock of the boundary between two stratigraphic units. Once such a boundary has been defined, its position in all other sections must be found by correlation. Thus golden spike sections are ideally defined in a section that has several features useful for precise correlation to other sections, such as rapidly evolving fossils, a geomagnetic polarity transition; dateable ash beds; geochemical changes, and so on. CONSEQUENCES Each GSSP uniquely defines a stratigraphic boundary in rock. It is the only place in the world where such a boundary is defined: elsewhere the position of the boundary has to be determined by correlation, which always involves uncertainty. The uncertainty is only negligible in special cases, such as a GSSP defined in a Milankovitch sequence and correlated to another Milankovitch sequence elsewhere, or perhaps by a GSSP defined at a polarity transition. In the case of a boundary defined exclusively by fossils, the uncertainty in correlation is generally of the order of 0.5 m.y., and in some cases may be as much as a few million years. NUMERICAL BOUNDARY DEFINITIONS FOR THE PRECAMBRIAN. The current integers assigned to the numerical boundaries of Precambrian subdivisions are not

166

uniformly distributed in time, but are chosen at what are considered to be the boundaries between major cycles of sedimentation and tectonics. Thus the boundaries are chronometric, rather than stratigraphic. It has been cogently argued that the boundaries between subdivisions should be moved to reflect "natural" subdivisions in the evolution of the planet, particularly for the early Earth. Such subdivisions would probably be applicable to the stratigraphic evolution of other terrestrial planets, including the Moon. By contrast, it has also been proposed that these numerical boundaries should be replaced by Precambrian GSSPs defined in rock sequences. CONSEQUENCES Numerical boundaries are abstract concepts that are nowhere located in rocks. For example, the current 2500 Ma Archean/Proterozoic boundary can never be located precisely in a stratigraphic sequence. A high precision date of 2500 Ma will always have an experimental error associated with its determination. Moreover the date - and the stratigraphic position - will change when decay constants are redetermined, or the apparent homogeneity of standards used for calibration is shown to be incorrect by the application of new techniques. The alternative - to define boundaries by GSSPs - is likely to be difficult to implement because of the great difficulties in correlating Precambrian sequences in the absence of rapidly evolving fossils. A SUGGESTION It may be worth considering the consequences of the somewhat radical suggestion that all stratigraphic boundaries eventually be defined mostly by integers related to their stratigraphic evolution, rather than by arbitrarily chosen values. How one could reach this goal while at the same time preserving the amazing detail in the Phanerozoic biostratigraphic stratigraphic record will be discussed. A major advantage of such an approach is that it might make stratigraphy more comprehensible to other scientists.

Workshop: Issues with the Precambrian Time Scale

Supercontinents and Earth Evolution Symposium 2005

AN UPDATED PRECAMBRIAN TIME SCALE OF CHINA Lu Sonqnian\ Wan Yushen^ Li Huaikun\ Wang Huichu\ Guo Jinghui^ ^Tianjin Institute of Geology and Mineral Resources, CGS, China (tjlsongnian@cgs.gov.cn) ^Geological Institute of China Geological Academy, China ^Geological and Geophysical Institute of China Science Academy, China

A new concept on the "natural" Precambrian time scale has been proposed by Gradstein et al. in 2004. They suggested that boundaries in the Precambrian should be placed at key events or transitions in the stratigraphic record, to highlight important milestones in the evolution of the Earth. Based on the new special reference to Precambrian time scale proposed by Gradstein et al., and the Precambrian time scale accepted by the China Stratigraphic Commission (CSC, 2001), we suggest an updated Precambrian time scale of China. The time range of the Sinian in the Precambrian time scale of China in 2001 is between 543 Ma and 680 Ma. Now according to an age of 629 ± 6.7 Ma (Yin et al., 2005) for the base of the Sinian, the time range of the Sinian now extends from 542 Ma to 630 Ma. The Sinian Period was previously placed in the latest Neoproterozoic, however, now it is as the earliest period of Paleozoic Era of Phanerozoic Eon, because of occurrence of metazoan. The subdivision of Pre-Sinian in China is a three fold classification, i.e., the Pre-Sinian is subdivided into the Proterozoic, Archean and Hadean Eons with boundaries 3850 Ma, 2500 Ma and 630 Ma, respectively. PROTEROZOIC EON The Proterozoic Eon contains Paleoproterozoic, Mesoproterozoic and Neoproterozoic Eras based on strata-type sections in Jixian Country and Yangtze Gorge in China. Their time boundaries are placed at 1800 Ma, 1000 Ma and 630 Ma, respectively.

Workshop: Issues with the Precambrian Time Scale

The Paleoproterozoic Era can be informally subdivided into three periods, namely Paleoproterozoic I, Paleoproterozoic II, and Paleoproterozoic III, with boundaries at 2300 Ma and 2100 Ma, respectively. The Mesoproterozoic Era includes, in ascending order, the Changcheng Period from 1800 Ma to 1600 Ma, the Gaoyuzhuang Period between 1600 Ma and 1400 Ma, and the Jixian Period from 1400 Ma to 1000 Ma. The Neoproterozoic Era is made up of the earlier Qingbaokou Period and later Nanhua Period with the time boundary at 820 Ma. The Nanhua Period is characterized by glaciation. ARCHEAN EON We agree with the subdivision of the Archean proposed by CGC, which is a four-fold classification of Archean into Eoarchean, Paleoarchean, Mesoarchean and Neoarchean Eras with boundaries at 3600 Ma, 3200 Ma, and 2800 Ma, respectively. HADEAN EON Nearly no geological records from the Hadean Eon have been found on Earth. The geological evolution of this eon of the Earth refers to the Lunar stratigraphy suggested by Gradstein et al. (2004). It is quite clearly understood that the local record may not always neatly fit into the globally accepted time limits. This project was supported by China Geological Survey grants (No. 200113900070 and 200313000060) and by the National Science Foundation of China (grant No.40032010).

167

Supercontinents and Earth Evolution Symposium 2005

THE HADEAN EON ON THE MOON Harrison H Schmitt University of Wisconsin-Madison, USA (schmitt@engr.wisc.edu)

The Moon incorporates the record of asteroid and cometary innpact activity in the inner solar system during the Hadean Eon on Earth. Any definition of the details of the Hadean should recognize this record and provide for correlation of related impact activity on Earth. The Moon, Earth and other terrestrial planets originated through "impact accretion" of material in the inner portion of the solar nebula within a few tens of millions of years of the 4.567 billion years birth of that nebula (To) [1]. This model continues to be supported by observations of young stars with observable accretionary disks surrounding them [2]. Consensus also exists that about 30 million of years after To the Moon belatedly came into existence as a result of a giant impact between the very young Earth and a chondritic asteroid [3] that was eleven to fourteen percent the mass of the Earth [4]. This consensus holds that at the time of that collision, both the Earth and the impactor would have been at least partially differentiated from their chondritic parent material by the separation of core-forming iron-rich liquid. Some additional chemical differentiation probably also had taken place by partial crystallization of magma oceans created during the earlier accretion of the two bodies [5]. This hypothetical lunar origin by giant impact primarily offers an explanation for the unusually high angular momentum of the EarthMoon system [6], and allows many of the lunar geochemical characteristics to be attributed to those of the original, at least partially differentiated, impactor. For example, the fifty percent higher iron content of the Moon relative to the Earth's mantle requires that at least eighty to ninety percent of the Moon be derived from the impactor rather than the core-depleted mantle of the Earth [7]. Major consequences of such a Moon-forming giant impact are presumed to be as follows: Most core-forming material in both bodies remained with or was quickly re-acquired by the Earth. The Moon formed through the rapid orbital reaggregation of some of the material ejected by the giant impact. According to Canup's recent computer simulations, that material would have reached temperatures between 3000 and 5000 K. Thus, according to this proposed origin, the bulk of material that formed the Moon vaporized during ejection and passed through a molten

168

stage during condensation and re-aggregation into a coherent body, erasing many primordial isotopic and volatile element signatures of the chondritic parents. Soon after or during lunar aggregation, the lunar core formed [8] and a magma ocean existed above this core [9]. Strong arguments exist, however, against the "giant impact hypothesis" for the origin of the Moon [10]. The geophysics and geochemistry of the upper mantle of the Moon are inconsistent with the implications of such a hypothesis. Geophysics strongly suggests that the lower mantle of the Moon, that is the portion below about 550 km [11], has remained largely unmelted, a conclusion that appears incompatible with giant impact [12]. Similarly, the non-glass component of the Apollo 15 and 17 orange and green pyroclastic glasses indicates that this lower mantle is volatile-rich relative to the upper mantle [13]. Further, chondritic isotopic and elemental systems exist in this non-glass component for tungsten [14], lead [15], samarium-neodymium [16], rhenium-osmium [17], and the siderophile and chalcophile elements [18]. These geochemical characteristics also appear incompatible with a giant impact origin for the Moon. Unfortunately, "giant impact" has permeated the thinking of the planetary geology community and currently is accepted widely as fact. The alternative of the capture of an independently accreted Moon [19], possibly at one of the two the co-orbiting Sun-Earth libration points [20], does not appear to have been seriously considered by the modeling community. This is in spite of the consensus that no more than about 10% of the Moon can be from the coredepleted mantle of the Earth thus making the giant impact hypothesis an "impact assisted capture" hypothesis. Can near impacts also result in capture? It does not appear that this question has been addressed in a comprehensive way. THE HADEAN Whatever was the origin of the Moon, an original solid crust was apparently present on the Moon by 4.52 billion years as the extinct isotope system (decays to indicates the lunar magma ocean was largely crystallized and fractionated within 40-50 million years of the beginning of the evolution of the solar nebula

Workshop: Issues with the Precambrian Time Scale

Supercontinents and Earth Evolution Symposium 2005

[21]. The '"'Hf-'^'W isotopic system also indicates that the separation of core fornning materials from the magma oceans of the Earth and Moon occurred at about 25-30 million years and for Mars at about 13 million years [22]. (Most workers still assume that "separation" is equivalent to core formation.) Actual core formation, however, appears to have been delayed in the Moon by 700 million years[23] and may have been delayed on Mars by as much as 400 million years and on Earth by as much as 100 million years [24]. As crust and core formation would be dependent on the size of the planets, such events cannot be correlated in time between planets. Fractional crystallization of the magma ocean initially left the ultramafic upper mantle extending down to -550 km, an iron-rich anorthosite crust --60 km thick and, in between, a zone of residual silicate liquid relatively rich in iron and potassium, rare earth elements, phosphorous (urKREEP), and thorium [25]. Interesting and complex as it is, the history [26] of solidification of the lunar mantle, and its future partial melting, does not relate to Hadean events on Earth. After the lunar crust had solidified sufficiently to record continued impacts (about 40 million years after To), the lunar surface was saturated with craters 60-70 km in diameter [27]. This created a megaregolith at least 25 km deep. The saturation cratering recorded in the lunar crust probably represents the sweep up of the last of pre-planetary accretion debris in the inner solar system and/or material ejected from the asteroid belt by its earliest interactions with Jupiter. On Earth, the effects of this intense cratering period probably dominated the first 300 million years of the Hadean. The development of a terrestrial mega-regolith, rich in glass and pulverized mineral debris, would have made clays the dominant mineral species in the waterrich environments at the surfaces of Earth [28]. This may have been the ideal environment for the development of complex organic molecules as precursors to replicating life. Such life, as discussed below, would have to wait to at least 3.8 billion years ago and the end of catastrophic cratering before such life could be assured of survival. Another significant process was unfolding simultaneously with the formation of the lunar and terrestrial mega-regoliths, the latter now recorded as "cratered highlands" relative to the lowlands plains of the lunar maria. During that same period, a few very large basins formed on the Moon and certainly also on the other terrestrial planets. The youngest and most obvious of these, the largely far-side basin South Pole-Aitken (-2500 km in diameter) [29], is

Workshop: Issues with the Precambrian Time Scale

recognized by the entire planetary geology community [30]; however, evidence exists for at least four other older basins of comparable or greater size [31]. The largest of these, Procellarum, is -3200 km in diameter. The age of South Pole-Aitken is estimated to be about 4.2 billion years because its rim remains clearly definable and has not been destroyed by the saturation cratering that sculpted the cratered highlands [32]. Procellarum would have formed earlier than South Pole-Aitken, possibly about 4.3 billion years ago, as the on-going saturation cratering and subsequent large basin impacts have significantly modified its rim. The overlap of the South Pole-Aitken and Procellarum ejecta blankets produced the thickest crust on the Moon (about 100 km thick versus an average of about 60 km and minima of about 30 km) [33]. The 4.2 billion year mark identified above is not well pinned down and may be off by ±0.1 billion years. It deserves, however, to be a placeholder for the separation of two, inner solar system eras of great importance and contrast. Prior to that time, in addition to saturation cratering and the clay dominated sedimentary environment, the seeds of the first terrestrial continents could have been formed by the fractional crystallization of thick, water-rich impact melt sheets formed in the very large, continental scale impact basins comparable to South Pole-Aitken and Procellarum [34]. Crystallization dates for detrital zircon from ancient basin sediments in Australia at 4.4 billion years of age, and evidence of their formation in the presence of water, strongly support this conclusion [35]. Subsequent to -4.2 billion years, saturation cratering and the formation of very large basins had ceased and the period of large basin forming impacts had begun or their basins began to be preserved [36]. The definable, -50 large basins (300-1000 km in diameter) formed before the main period of eruption of lunar mare basalts that began about 3.85 billion years ago. They may represent the appearance of a new source of impactors for the inner solar system [37]. This source probably was the proto-Kuiper belt of icy bodies, many of which were injected into the inner solar system when Neptune and other giant planets interacted as they apparently migrated outward early in solar system evolution [38]. On the Moon, old large basins are distinct from young large basins [39], the younger being more crisply defined in shape and having both mass concentrations (mascons) in their interiors and mass deficiencies beneath their rims [40]. The lunar crustal strengthening "transition event" that separates young mascon basins from older non-mascon basins probably occurred gradually

169

Supercontinents and Earth Evolution Symposium 2005

between 4.2 and 3.9 billion years, but this event is lunar specific in time. The transition represents the nnovennent of potassium, rare earth element, phosphorus (KREEP), and thorium-rich residual magma ocean liquid into the crust and its gradual solidification. A similar upward movement and solidification of residual magma ocean liquid probably occurred on every planet, but its composition, timing, and effects would be largely unique to the size of the planet and the thickness of each crust. The 3.8 billion year date for the end of the Hadean is solidly based on the well-defined end of large basin (300-1000 km diameter) formation on the Moon and therefore on all the inner solar system planets [41]. Subsequent to 3.8 billion years, the formation of impact craters larger than -100 km in diameter has been very rare. In lunar time-stratigraphic nomenclature this would be the end of the combined "Nectarian and Imbrium Systems". In addition, the planetary geology community has adopted this date as the approximate time of the "cataclysm" or "late bombardment" of impacts in the inner solar system [42]. Like the "giant impact" hypothesis, we should question the planetary geology community's consensus that there was a "cataclysm" of impacts in the inner solar system between 3.8 and 3.9 billion years that accounted for the vast majority of lunar basins and highland craters. Alternatively, the lunar wide distribution of ejecta and thus the sample bias created by the last 1015 (out of about 50) large (300-1000 km) basinforming events during the -100 million year period around 3.8 billion years probably accounts for the concentration of argon ages around 3.85 ± 0.05 billion years [43]. These "mascon basins" make up the combined "Nectarian and Imbrium Systems" of the USGS time-stratigraphic system for the Moon [44]. In spite of the apparent consensus for a cataclysm, strong arguments can be made for a more prolonged impact history. Various lunar samples and several lines of reasoning suggest that this period of large basin-forming events, the latter part of the "Pre-Nectarian System", began about 4.2 billion years ago [45]. This, therefore, suggests a -400 million year "cataclysm" which, in turn, suggests that adjustments in the orbits of the outer planets took longer than implied by current computer models. This prolonged cataclysm would involve only large basin formation, rather than only -100 million years that including almost all cratering recorded on the Moon in an -100 million year interval. For the latter to be the case, the following major episodes in lunar history would need to occur in that relatively short period [46]:

170

1. Saturated impacts covered the lunar crust with craters 60-70 km in diameter. 2. A number (1-5) of very large craters formed. 3. Old large basins (-35) formed. 4. Magma ocean residual liquid migrated into crust and cryptomaria/KREEP basalt lavas erupted, probably represented by lunar samples dated as — billion years old [47]. 5. Young large basins (-15) formed, and 6. Cryptomaria surfaces covered by young large basin ejecta. Clearly, 100 million years is a long time and all of the above could have occurred in such a period. To this observer, however, that would seem unlikely and worthy of much further investigation. SUMMARY Currently known lunar history suggests that two major boundaries can be defined for the Hadean. The first is at about 4.2 billion years, the end of saturation cratering and very large basin formation in the inner solar system. The second boundary and the potentially definable end of the Hadean is at about 3.8 billion years, the end of large basin formation in the inner solar system. Although not yet fully confirmed, the 3.8 billion year boundary also may be the point at which the first isotopic evidence of life on earth appears [48]. REFERENCES: [1] (Carlson and Lugmair 2000, Alexander et al. 2001, Taylor 2001, Jacobsen 2003, Amelin et al. 2004), Carlson, R. W., and G. W. Lugmair, 2000, Timescales of planetesimal formation and differentiation based on extinct and extant radioisotopes, in R. M., Canup and K. Righter, Eds., Origin of the Earth and Moon, University of Arizona Press, Tucson, and Lunar and Planetary Institute, Houston, pp. 25-44; Alexander, C. M. O'D, Boss, A. P., and Carlson, R. W., 2001, The early evolution of the inner solar system: A meteoritic perspective, Science, 293, pp. 64-68; Taylor, S. R., 2001, Solar System Evolution, Cambridge University Press, Cambridge, p. 392; Jacobsen, S. B., 2003, How old is planet Earth?, Science, 300, pp. 1513-1514; Amelin, Y, Krot, A. N., and Twelker, E., 2004, Pb isotopic age of the CB chondrite Gujba, and the duration of the chondrule formation interval, Geochimica et Cosmochimica Acta, 68, Abstract E958. [2] Wilner, D., et al, 2005, Astophysical Journal Letters, June 20; Cowen, R. 2005, Panning distant dust. Science News, 168, pp. 10-12. [3] Spudis, P. D., 1996, The Once and Future Moon, p. 164; Canup R. M., and Righter, K., 2000, eds. Origin of the Earth and Moon, Workshop: Issues with the Precambrian Time Scale

Supercontinents and Earth Evolution Symposium 2005

University of Arizona Press, 555 pp; Taylor, S. R., 2001, Solar System Evolution, Cambridge University Press, Cambridge, p. 392; Warren, P. H., 2003, The moon, in A. Davis, ed.. Treatise on Geochemistry, 1, Elsevier, Amsterdam, pp. 559-599; Palme, H., 2004, The giant impact formation of the Moon, Science, 304, pp. 977-979. [4] (Hartmann and Davis 1975, Cameron and Ward 1976, Hartmann 1986, Cameron 2002, Canup 2004, Palme 2004) Hartmann, W. K., and D. R. Davis, 1975, Satellite-sized planetesimals and lunar origin: Icarus, v. 24, p. 504-515; Cameron, A. G. W., and W. R. Ward, 1976, The origin of the Moon, Lunar Science Conference 7, abstract, pp. 120-122, Lunar Science Institute, Houston; Hartmann, W. K., 1986, Moon origin: The impact trigger hypothesis, in W. K. Hartmann, R. J. Phillips, and G. J. Taylor, eds.. Origin of the Moon: Lunar and Planetary Institute, Houston, p. 579-608; Canup, R. M., 2004, Simulations of a late lunar-forming impact, Icarus, 168, p. 433; Palme, H., 2004, The giant impact formation of the Moon, Science, 304, pp. 977979. [5] Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [6] Taylor, S. R., 2001, Solar System Evolution, Cambridge University Press, Cambridge, p. 392. [7] Taylor, S. R., 2001, Solar System Evolution, Cambridge University Press, Cambridge, p. 392; Canup, R. M., 2004, Simulations of a late lunar-forming impact, Icarus, 168, p. 433. [8] Agee, C. B., 1991, High-pressure melting of carbonaceous chondrites, in C. B. Agee and J. Longhi, Eds., Workshop on the Physics and Chemistry of Magma Oceans from 1 Bar to 4 Mbar, Technical Report Number 92-03, Lunar and Planetary Institute, Houston, pp. 11-12; Neal, C. R., Ryan, J., Jain, J. C., and Chazey, W., 2000, The nature of the lunar mantle: Generally chondritic of the Mare basalt sources, but with Garnet in the source of the volcanic glasses. Lunar and Planetary Science Conference 31, Abstract #1944; Neal, C. R., and Ely, J. C., 2002, Sulfide immiscibility in the lunar magma ocean: evidence for a primitive lunar lower mantle and the origin of high-fj mare basalts. Lunar and Planetary Science Conference 31, Abstract #1821. [9] (Wood et al. 1970, Smith et al. 1970, Warren 1985) Wood, J.A., Dickey, J. S, Marvin, U. B, and Powell, B. N., 1970, Lunar anorthosites and a geophysical model of the Moon, Proceedings Apollo 11 Lunar Science

Workshop: Issues with the Precambrian Time Scale

Conference, pp. 965-988; Smith, J. V., Anderson, A. T., Newton, R. C., Olsen, E. J., Wyllie, P. J., Crewe, A. V., Isaacson, M. S., and Johnson, D., 1970, Petrologic history of the moon inferred from petrography, mineralogy, and petrogenesis of Apollo 11 rocks. Proceedings Lunar Science Conference 1, pp. 897-925; Warren. P. H., 1985, The magma ocean concept and lunar evolution. Annual Reviews in Earth and Planetary Science, 13, pp. 201-240. [10] Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [11] Goins, N. R., Dainty, A. M., and Toksoz, M. N., 1981, Lunar seismology: The internal structure of the Moon: Journal of Geophysical Research, v. 86, p. 5061-5074; Kahn, A., Mosegaard, K, and Rasmussen, K. L., 2000, Lunar models obtained from a Monte Carlo inversion of the Apollo seismic P and S. waves (Abstract #1341): Lunar and Planetary Science Conference 31. [12] Kahn, A., Mosegaard, K, and Rasmussen, K. L., 2000, Lunar models obtained from a Monte Carlo inversion of the Apollo seismic P and S. waves (Abstract #1341): Lunar and Planetary Science Conference 31; Neal, C. R., 2001, Interior of the Moon: The presence of garnet in the primitive deep lunar mantle: Journal of Geophysical Research, v. 106, E l l , p. 27865-27885. [13] (Wasson et al. 1976, Krahenbuhl 1980, Meyer 1989). Wasson, J. T, Boynton, W. V., Kallemeyhn, G.W., Sundberg, L. L., and Wai, C. M., 1976, Volatile compounds released during lunar lava fountaining, Proceedings Lunar Science Conference 7, pp. 1583-1595; Krahenbuhl, U., 1980, Distribution of volatile and non volatile elements in grain-size fractions of Apollo 17 drive tube 74001/2, Proceedings Lunar and Planetary Science Conference 11, pp. 1551-1564; Meyer, C., 1989, A brief literature review of observations pertaining to condensed volatile coatings on lunar volcanic glasses, in J. W. Delano and G. H. Heiken, Eds., Workshop on Lunar Volcanic Glasses: Scientific and Resource Potential, Technical Report #90-02, Lunar and Planetary Institute, pp. 50-51. [14] Lee, D., Halliday, A. N., Snyder, G. A., and Taylor, L. A., 2000, Age and origin of the Moon: Science, v. 278, 1997, p 1098-1103. [15] Nunes, P. D., Tatsumoto, M., and Unruh, D. M., 1974, U-Th-Pb systematics of some Apollo 17 lunar samples and implications for a lunar basin excavation chronology: Lunar Science Conference 5, p. 1487 1514.

171

Supercontinents and Earth Evolution Symposium 2005

[16] Snyder, G. A. Borg, L. E. , Nyquist, L. E., and Taylor, L. A., 2000, Chronology and isotopic constraints on lunar evolution, in R. M. Canup and K. Righter, Origin of the Earth and Moon, Part III, University of Arizona Press, p. 381. [17] Walker, R. J., Horan, M. F., Shearer, C. K., and Papike, J. J., 2004, Depletion of highly siderophile elements in the lunar mantle: evidence for prolonged late accretion. Earth and Planetary Science Letters, 224, pp. 399413. [18] Neal, C. R., 2001, Interior of the Moon: The presence of garnet in the primitive deep lunar mantle: Journal of Geophysical Research, v. 106, E l l , p. 27865-27885; Walker, R. J., Horan, M. F., Shearer, C. K., and Papike, J. J., 2004, Depletion of highly siderophile elements in the lunar mantle: evidence for prolonged late accretion, Earth and Planetary Science Letters, 224, pp. 399-413. [19] Alfven, H. and Arrhenius, G., 1972, Origin and evolution of the Earth Moon system. The Moon, V. 5, p. 216; Schmitt, H. H., 1991, Evolution of the Moon: The Apollo Model: American Mineralogist, v. 76, p 775-776; Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed., Encyclopedia of Space: Wiley, New York, Chapter 1. [20] Belbruno, E., and Gott, J. R., Ill, 2004, Where did the Moon come from?. Astronomical Journal, 129, p. 1724. [21] Shearer, C. K.., and Newsom, H. E., 2000, W-Hf isotope abundances and the early origin and evolution of the Earth-Moon System, Geochimica et Cosmochimica Acta, 64, pp. 3599-3613. [22] Kleine, T., Munker, C., Jezger, K., and Palme, H., 2002, Rapid accretion and early core formation on asteroids and the terrestrial planets from Hf-W chronometry: Nature, v. 418, p. 952-955. [23] Lin, R. P., Mitchell, D. L., Harrison, L., et al, 1999, Miniature magnetospheres on the Moon and their relation to albedo swirls, Lunar and Planetary Science Conference 30, Abstract #1930; Lin, R. P., Anderson, K. A., and Hood, L., 1988, Lunar surface magnetic field concentrations antipodal to young large impact basins: Icarus, v. 74, p. 529-541; Mitchell, D. L., Lin, R. P., Harison, L., et al, 2000, Solar wind interaction with lunar crustal magnetic fields: relation to Albedo Swirles, Lunar and Planetary Science Conference 31, Abstract #2088. [24] Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed., Encyclopedia of Space: Wiley, New York, Chapter 1.

172

[25] Warren, P. H., and J. T. Watson, 1978, Compositional-petrographic investigation of pristine nonmare rocks. Lunar and Planetary Science Conference 9, pp. 185-217. [26] Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed., Encyclopedia of Space: Wiley, New York, Chapter 1. [27] Wilhelms, D. E., 1987, The Geologic History of the Moon: U.S. Geological Survey Professional Paper 1348, U.S. Government Printing Office, Washington, 302 p.; Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [28] Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [29] Wilhelms, D. E., 1987, The Geologic History of the Moon: U.S. Geological Survey Professional Paper 1348, U.S. Government Printing Office, Washington, 302 p. [30] Wilhelms, D. E., 1987, The Geologic History of the Moon: U.S. Geological Survey Professional Paper 1348, U.S. Government Printing Office, Washington, 302 p.; Spudis, P. D., 1993, Geology of Multi-Ring Impact Basins, Cambridge University Press, New York, 263 pp.; Spudis, P. D., Reisse, R. A., and Gillis, J. J., 1994, Ancient multiring basins on the Moon revealed by Clementine laser altimetry. Science, 266, pp. 1848-8151; Spudis, P. D., 1996, The Once and Future Moon, p. 164. [31] Wilhelms, D. E., 1987, The Geologic History of the Moon: U.S. Geological Survey Professional Paper 1348, U.S. Government Printing Office, Washington, 302 p.; Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [32] Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [33] Wieczorek, M. A., and R. J. Phillips, Journal of Geophysics, 103, 1998, Plate 2. [34] Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [35] Wilde, S. A., Valley, J. W., Peck, W. H., and Graham, C. M., 2001, Evidence from detrital zircons for the existence of continental crust and oceans on the Earth 4.4 Gyr ago: Nature, V. 409, p. 175-178; Turner, G., Harrison, T. M., Holland, G., Mojzsis, S. J., and Gilmour, J., 2004, Extinct 244Pu in ancient zircons. Science, 306, pp. 89-91; Mojzsis, S. J., Harrison, T. M., and Pidgeon, R. T., 2001, Oxygen-isotope evidence from ancient zircons for liquid water at the Earth's surface

Workshop: Issues with the Precambrian Time Scale

Supercontinents and Earth Evolution Symposium 2005

4,300 Myr ago: Nature, v. 409, p. 178-181; Watson, E. B., and Harrison, T. M., 2005, Zircon thermometer reveals minimum melting conditions on earliest Earth, Science, 308, pp. 841-844. [36] Wilhelms, D. E., 1987, The Geologic History of the Moon: U.S. Geological Survey Professional Paper 1348, U.S. Government Printing Office, Washington, 302 p.; Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [37] Schmitt, H. H., 1999, Early lunar impact events: terrestrial and solar system implication: Geological Society of America Annual Meeting Abstract #50440; Schmitt, H. H., 2001, Lunar cataclysm? Depends on what "cataclysm' means: Lunar and Planetary Science Conference, Abstract #1133; Dones, L., 2002, Dynamics of possible late heavy bombardment impactor population: Lunar and Planetary Science Conference 33, Abstract #1662; Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [38] Levinson, H. F., Dones, L., Chapman, C. R., Stern, S. A., Duncan, M. J., and Zahnle, K, 2001, Could the lunar "late heavy bombardment" have been triggered by the formation of Uranus and Neptune?, Icarus, 151, pp. 286-306; Morbidelli, A., 2004, How Neptune pushed the boundaries of our solar system. Science, 306, pp. 1302-1304; Kerr, R. A., 2004, Did Jupiter and Saturn team up to pummel the inner solar system?. News Focus, Science, 306, p. 1676. [39] Schmitt, H. H., 1989, Lunar crustal strength and the large basin-KREEP connection; in G. J. Taylor and P. H. Warren, eds., Workshop on Moon in Transition: Apollo 14, KREEP, and Evolved Lunar Rocks: Technical Report #89-03, Lunar and Planetary Institute, Houston, p. 111-112. [40] Muller, P. M., and Sjogren, W. L., 1968, Mascons: Lunar mass concentations: Science, 161, p. 680-684. [41] Wilhelms, D. E., 1987, The Geologic History of the Moon: U.S. Geological Survey Professional Paper 1348, U.S. Government Printing Office, Washington, 302 p. [42] Tera, F., Papanstassiou, D. A., and Wasserburg, G. J., 1974, Isotopic evidence for a terminal lunar cataclysm. Earth and Planetary Science Letters, 22, pp. 1-21; Ryder, G., 1990, Lunar samples, lunar accretion, and the early bombardment history of the Moon, EOS, AGU, 71, pp. 313 and 322-323; Ryder, G., Koeberl, C., and Mojzsis, S. J., 2000, Heavy bombardment of the Earth

Workshop: Issues with the Precambrian Time Scale

at -3.85 Ga: The search for petrographic and geochemical evidence, in R. M. Canup and K Richter, eds.. Origin of the Moon and the Earth: University of Arizona Press, Tucson, pp. 475-492; Hartmann, W. K., Ryder, G., Dones, L, and Grinspoon, D, 2000, The timedependent intense bombardment of the primordial Earth/Moon System, in R. M. Canup and K. Righter, Origin of the Earth and Moon, Part III, University of Arizona Press, pp. 493-512; Cohen, B. A., Swindle, T. D. and Kring, D. A., 2000, Support for the lunar cataclysm hypothesis from lunar meteorite impact melt ages. Science, 290, pp. 17541756; Ryder, G., 2001, Mass flux during the ancient lunar bombardment: The cataclysm, Lunar and Planetary Science Conference 32, Abstract #1326; Cohen, B. A., James, O. B., Taylor, L. A., Nazarov, M. A., and Baruskova, L. D., 2004, Lunar highland meteorite Dhofar 026 and Apollo sample 15418: Two strongly shocked, partially melted, granulitic breccias, Meteoritics and Planetary Science, 39, 9, pp. 1419-1447. [43] Schmitt, H. H., 2001, Lunar cataclysm? Depends on what "cataclysm* means: Lunar and Planetary Science Conference, Abstract #1133; Chapman, C. R., Cohen, B. A., and Grinspoon, D. H., 2002, What are the real constraints on commencement of the late heavy bombardment?: Lunar and Planetary Science Conference 33, Abstract #1627; Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [44] Wilhelms, D. E., 1987, The Geologic History of the Moon: U.S. Geological Survey Professional Paper 1348, U.S. Government Printing Office, Washington, 302 p. [45] Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [46] Schmitt, H. H., 2003, Apollo 17 and the Moon, in H. Mark, ed.. Encyclopedia of Space: Wiley, New York, Chapter 1. [47] Taylor, L. A., and co-workers, Earth and Planetary Science Letters, 66, 1983, pp 3347; Heiken, G. H., and co-workers. Lunar Sourcebook, 1991, p 209; Heiken, G. H., and co-workers. Lunar Sourcebook, 1991, p 218219. [48] Mojzsis, S. J., Arrhenius, G., McKeegan, K. D., Harrison, T. M., Nutman, A. P., and Friend, C. R. L., 1996, Evidence for life on Earth before 3800 million years age: Nature, V. 384, p. 55-59; Moorbath, S., 2005, Dating earliest life, Nature, 434, p. 155; Westfall, F., 2005, Life on the early Earth: A sedimentary view. Science, 308, pp. 366-367.

173

Siipercontinents and Earth Evolution Symposium 2005

174

Siipercontinents and Earth Evolution Symposium 2005

INDEX

(bold entry indicates presenter) Abbott, DH Alkmim, FF Allen, CM Appel, P Armstrong, RA Baba, S Babayants, P Badarch, G Baghela, A Balasubramonian, G Banerjee, DM Barley, M Barovich, K Becker, T Bernnan, RG Berzin, R Betts, PG Beukes, NJ Bhattacharya, S Bibikova, E Biswas, A Bleeker, W Bloch, JD Blokh, Y Bodorkos, S Bogdanova, SV Boger, SG Borel, GD Bouyo, MH Brown, D Brown, LD Brown, M Buchan,C Buchan, KL Bunge, H-P Calver, CR Camacho, A Campbell, IH Carbonell, R Carson, CJ Catuneanu, 0 Cavosie, AJ Cawood, PA Chernicoff, CJ Chetty, TRK Chulick, G Close, D Collins, AS Collins, WJ Condie, KC Corkeron, M Cornell, DH Crawford, AJ Crossey, LJ Crowe, W

149 97, 98 89, 112 100, 103, 147 66, 132 105 67 58 140, 141 18 91 27 38, 39, 40, 42, 43 79 13 67, 135 41 24,124 16 118 92 3, 153 122 67 33, 34, 44 14, 94, 118 89 4 99 63, 119 68 54 56, 57 9 7 51,146 73 89, 112 63, 119 106 126 155 57, 96, 120, 131 128 104 136 36 56, 107, 121 12 60 52 81, 143 113 122 139

Cruz, S Cutten, HNC Daly, JS Dalziel, IWD Dasgupta, PK Davis, WJ de Brito Neves, BB de Kock, MO De Waele, B Dehler, CM Demoux, A Depine, GV Detweiler, S Deutsch, A Donadini, F Dorland, HC Dunkley, DJ Edgoose, C Elburg, MA Eliaeson, K Elming, S-A Eriksson, PG Ernst, RE Evans, DAD Evins, LZ Evins, PM Fernandes, LAD Ferris, G Fitzsimons, ICW Foden,J Forbes, CJ Foster, DA Fuck, RA Gibson, G Giles, D Glen, RA Goldfarb, RJ Goscombe, B Gose, WA Goss, SC Gower, CF Gray, DR Greentree, M Gregory, L Grey, K Groves, Dl Gruschka, S Hand, M Hanson, RE Harlan, S Hatton, C Hegner, E Heinson, G Henderson, RA Hilburn, 1 Hoffman, PF Hokada, T Hou, G

98 123 110,142 2 108 13, 144 82, 117 24, 124 83, 85,94,118 122 58 125 136 74 74 24 106 36 115 111 142 126 9 11,22, 23, 24, 53, 124, 127, 146, 158 50 12 128 42 102, 129 115 41 66, 132 82, 117 37, 130 8,41 113 60 66, 132 80 131 93 66, 132 87 91 51, 164 60 58 38, 39, 40, 42, 43, 64 78, 80 74 10 58 64 50 124 47 106, 137, 139 15

175

Supercontinents and Earth Evolution Symposium 2005

Hough, M Howard, HM Hronsky, JMA Huaikun, L Huichu, W Hulscher, B Hutton, L Izard, CF Jacobs, J Jercinovic, M Jia, Y Jinghui, G John, T Johnson, SP Johnston, ST Jons, N Juhlin, C Kagami, H Kaiden, H Karlstrom, KE Kay, SM Kayal, JR Khudoley, AK Kinny, PD Kirkland, CL Kirschvink, JL Konilov, A Korhonen, K Korsch, R Kositcin, N Krabbendam, M Krapez, B Krishnendu, NR Kroner, A Kumpulainen, R Li, J Li, S Li, WH Li, X Li,ZX Liegeois, J-P Liu, DY Loose, D Lubnina, N Mackenzie, S IVIalone, S IVlarshak, S IVlartins-Neto, MA Maruyama, S Masago, H Mazumder, R Meert, JG Mertanen, S Metelkin, DV Mints, M Moczydlowska, M Mohanty, H Mooney, WD Morgan, JP

176

50 44 60 109, 167 167 102 130 146 79 122 49 167 133, 147 83, 123 4 100, 103, 133, 147 119 137 137 72, 122 125 134 77 41,121,129 110 124 135 74 26 26 120 27 18 58 142 15 20 86 86, 87 9, 86, 87, 88, 94 85 58 99, 100, 147 75 50 91 98 97 30 30 126 91 75 76 67, 135 142 19 65, 136, 149 125

Morris, PA Motoyoshi, J Mukasa, S Mukherjee, R Nakano, N Nam,TN Nemchin, AA Neumann, N Nevanlinna, H Nutman, H Occhipinti, SA Okaya, N Oncken, 0 Osanai, Y Owada, M Page, RW Paliwal, BS Pandit, MK Passchier, C Payne, JL Pease, V Pedrosa-Soares, AC Pehrsson, SJ Penaye, J Peres, GG Pesonen, LJ Pettersson, A Philippova, 1 Phillips, BR Phillips, G Pidgeon, RT Pirajno, F Pisarevsky, SA Popova, L Porcher, CC Postnikov, A Pradhan, VP Prasad, AK Prave, T Prokopiev, AV Puchkov,V Qian, X Qiu, J Radhakrishna, T Rainbird, RH Raub, TD Raub, TMD Razakamanana, T Reddy, SM Roberts, D Rose, E Sahu, AK Saintot, A Salminen, J Santos, OS Santosh, M Schenk, V Schmidt, PW Schmitt, HH

151 106 139 17 137, 139 137 120 37, 38 130 74 156 31,88,138 65 62 105, 137, 139 105, 137, 139 33 140, 141 91 66, 132 39 111,142 98 13, 144 99 98 25, 74 81, 143 67 7 89 151 44, 151 79, 94, 96, 107 118 128 118 91 19 120 77 142 15 150 18 13, 23, 144 53, 146 23 103, 121, 133 31, 34, 56, 138 142 124 19 142 25 120 121 99, 100, 103, 133, 147 73 168

Siipercontinents and Earth Evolution Symposium 2005

Schobbenhaus, C Scholl, DW Schumann, A Scott, R Scrimgeour, 1 Sears, JW Selway, K Sharma, KK Sheppard, S Shields, GA Shiraishi, K Sircombe, K Smirnov, A Smith, AG Smithies, RH Sohl, L Songnian, L Southgate, P Spaggiari, CV Squire, RJ Stephenson, R Strachan, R Strauss, H Stupak, V Suleimanov, A Sun, M Swain, G Szpunar, M Taguchi, S Tamrat, E Tani, K Tembo, F Thomas, RJ Tomurtogoo, 0 Toteu, F Toyoshima, T Trouw, R Trubitsyn, VP Trusov, A Tsunogae, T Tyler, IM Valley, JW Van der Voo, R Van Kranendonk, M Vasilieva, T Vernikovsky, VA Vielreicher, RM von Huene, R Vujovich, Gl Wade, B Wang, C Wang, J Wang, X Wang, Y Wartho, J-A Wegner, H Weinberg, R Wendorff, M Whitehouse, MJ

82, 117 5 100 111 35, 36, 64 77 36, 64 90 29, 33, 138 50, 162 105, 106 32 22 59, 166 44 91 109, 167 37, 130 28 112 142, 148 120 50 67, 135 67, 135 20 27,42 40 137 91 83 83 79 58 99 139 66, 132 149 67 139 33,44 155 71 157 75 76 60 5 128 43 15 86 150 65 88 100 41 101 110

Whitmeyer, SJ Whittington, A Wilde, SA Williams, GE Williams, ML Wilson, CJL Windley, BF Wingate, MTD Yuan, X Yuen, DA Yushen,W Zamozhniaya, N Zhao, G Zhenqun, X Zhihong, C Zhou,J ZIobin, V

72 98 20, 155 73 122 89, 112 58 9, 36, 58, 96, 146, 151 65 6 167 67, 135 20 109 109 150 135

177

Siipercontinents and Earth Evolution Symposium 2005

178

Supercontinents and Earth Evolution Symposium 2005

Notes

179

Supercontinents and Earth Evolution Symposium 2005

Notes

180

Supercontinents and Earth Evolution Symposium 2005

Notes

181

Siipercontinents and Earth Evolution Symposium 2005

Notes

182

ILLUSTRATION: ASTER satellite image of intensely folded rock units of central Madagascar. Individual Itremo Group quartzite layers (light colours) of just 100 m thickness outline 10 km wide fold interference structures (top-centre). Bands 3-2-1. Image: Bregje Hulscher