Stratigraphy, Sedimentation and Age of The

ResearchOnline@JCU

This file is part of the following reference:

Tucker, Ryan Thomas (2014) Stratigraphy, sedimentation and age of the upper cretaceous Winton formation, central-western Queensland, Australia: implications for regional palaeogeography, palaeoenvironments and Gondwanan palaeontology. PhD thesis, James Cook University.

Access to this file is available from:

http://researchonline.jcu.edu.au/34439/

The author has certified to JCU that they have made a reasonable effort to gain permission and acknowledge the owner of any third party copyright material included in this document. If you believe that this is not the case, please contact [email protected] and quote http://researchonline.jcu.edu.au/34439/

Stratigraphy, Sedimentation and Age of the Upper Cretaceous Winton Formation, central-western Queensland, Australia: Implications for regional palaeogeography, palaeoenvironments and Gondwanan palaeontology

Thesis Submitted by

Ryan Thomas Tucker March 2014

For the degree of: Doctorate of Philosophy

In the: SCHOOL OF EARTH AND ENVIRONMENTAL SCIENCES FALCULTY OF SCIENCE AND ENGINEERING JAMES COOK UNIVERSITY

Statement of Access

I, the undersigned author of this thesis, understand that James Cook University will make this thesis available for use within the university library and allow access in other approved libraries after its submission. All users consulting this thesis will have to sign the following statement: In consulting this thesis I agree not to copy or closely paraphrase it in whole or in part without the written consent of the author; and to make proper public written acknowledgement for any assistance which I have obtained from it.

Beyond this, I do not wish to place any restrictions on access to this thesis.

Ryan Thomas Tucker January 2014

Declaration

I declare that this thesis is my own work and has not been submitted in any form for another degree or diploma at any university or other institution or tertiary education. Information derived from the published or unpublished work of others has been acknowledged in the text and a list of those references is given.

Ryan Thomas Tucker January 2014

Statement of Contribution by Others

Nature of assistance Contribution Coauthors and Affiliations Academic Support Grant Writing Dr. Eric Roberts Dr. Paul Dirks Dr. Steven Salisbury Data Analysis Dr. Eric Roberts Dr. Yi Hu Dr. Richard Wormwald Dr. Tony Kemp Processing Dr. Eric Roberts Dr. Yi Hu Dr. Richard Wormwald Editing Dr. Eric Roberts Dr. Tony Kemp Dr. Steven Salisbury Vikie Darlington Mark Munro Owen Li Field-Analytical Field Research Dr. Eric Roberts Support Dr. Steven Salisbury Dr. Marc Hendrix Dr. Leif Tapanila *BNP field crew

Data Analysis Dr. Eric Roberts Dr. Yi Hu Dr. Tony Kemp Dr. Richard Wormwald

Financial Support Field & Analytical SEES & GRS ($13,300.00) Maggie Smith ($2000.00)

Conference & Travel EGRU ($1,000.00) SVP-JSGSMTG ($600.00)

III

Chapter Details of publication(s) Nature and extent of or input by thesis author Appendix 2 Tucker, R.T., Roberts, E.M., Hu, I wrote most of the Y., Kemp, A.I., Salisbury, S.W., paper to and 2013. Detrital zircon age development most of constraints for the Winton concepts. Kemp and Formation, Queensland: Hu set up the LA- Contextualizing Australia's Late ICPMS system at JCU Cretaceous dinosaur faunas. and developed the Gondwana Research, 24(2), 767- tuning method for the 779. instrument. I conducted all of the analyses, but benefited greatly from the advice and editing provided by all supervisors and contributors. 3 Tucker R.T., Roberts, E.M., I wrote most of the Henderson, R.A., Kemp, A.I.S. paper and developed the majority of the concepts. Kemp in particular, assisted me with Hf-isotope analysis and interpretation. Henderson provided key insights about the geology of N. QLD. 4 Tucker, R.T., Roberts, E.M., I contributed to the Darlington V. writing and development of concepts. Darlington played an important role on the GIS model and contributed to the writing of the methods for the basin model construction.

Appendix Romilio, A., Tucker, R.T., & I contributed to the 1 Salisbury, S.W., 2013. writing and Reevaluation of the Lark Quarry development of the dinosaur Tracksite (late Albian– paper as a whole, but Cenomanian Winton Formation, my specific focus central-western Queensland, pertained to the Australia): no longer a stampede?. revised Journal of Vertebrate Paleontology, sedimentological 33(1), 102-120. interpretations of Lark Quarry Track Site and the interpretations of the depositional environments Appendix I contributed to the 2 writing and development of concepts I confirm the candidate’s contribution to this paper and consent to the inclusion of the paper in this thesis.

Primary Advisor: Eric Roberts

Signature:

Date:

Acknowledgements This work, in its entirety, would not have occurred if it were not for all those involved in the last three and a half years. Although not everyone can be thanked in text, to everyone I pass on a heartfelt thank you. To my Ph.D. committee, I owe the greatest debt of gratitude. I thank Dr. Steven Salisbury for that SVP dinner which has brought about a wonderful and enriching academic adventure. I thank Dr. Paul Dirks, without whom I know this project and my arrival at James Cook University would have never happened. However, I owe my greatest debt of gratitude to my principle advisor, Dr. Eric Roberts. From the first day in Bladensburg to my first steps as a Lecturer, it is your guidance, patience, and fortitude which made this possible. I hope that in the years to come, I can work with the same passion, integrity, and drive which you devote to everything you set out to accomplish. I also greatly look forward to the buttes needing stratigraphic sections, to the big walls needing climbing, and to the bottles of wine waiting to be opened. I graciously thank all of the lecturer staff at James Cook University; including but not limited to Carl Spandler, Tom Blenkinsop, Bob Henderson, Tony Kemp, Jan Marten Huizenga, and Nigel Chung. I would also like to thank all of the support staff within the School of Earth and Environmental. I would like to also thank all of the staff at the Advanced Analytical Centre at James Cook University. This is especially true for Dr. Yi Hu, for all efforts taken in acquiring the LA-ICPMS geochronological data, a true Doctor ‘WHO’. For continued financial support, I thank the Graduate Research Scheme and the School of Earth and Environmental Sciences, along with the Jackson School Travel Grant, The Economic Geological Research Unit, and the generous funds provided by Maggie Smith. I would also like to thank my fellow PhD and Honours students over the past several years. In particular I would like to thank Owen, Johannes, Vikie, Rob, Mark, Babo, George, Cassey, Taka, Matt, and Hannah. I wish to thank you all for the questions, comments, and many laughs which made this journey a memorable one. Most of all, I thank my family and friends for their unwavering support throughout my life concerning these crazy rock based endeavors. I thank my parents for their continued encouragement and the energy they put forth to further my passions in life. I thank Richard Thompson and Garrett Schmitz, both of whom have kept me somewhat sane these many years. Lastly but certainly not in the least, I would like to thank my partner in life, Astrid Vachette. Your laughter and love in our time together made this period of my life the best it could have ever been.

Abstract The mid-Cretaceous Winton Formation is one of Australia’s most important sources of Mesozoic terrestrial fossils. In recent years it has produced some of the continent’s most significant body and trace fossil records of mid-Cretaceous dinosaurian and crocodilian assemblages. Additionally, the Winton Formation preserves a diverse assemblage of lungfish and primitive ray-finned fish, aquatic lizards, turtles, numerous invertebrates along with a high diversity of plant macrofossils, including some of the world’s earliest flowering plants. The Winton Formation is exposed over large portions of remote central-western

Queensland, in addition to northern New South Wales, north-western South Australia and the south-western corner of the Northern Territory. Despite its importance for our understanding of Australian terrestrial environments during the latter part of the Mesozoic, very little in the way of detailed geological work has been carried out on the Winton

Formation. As a consequence, palaeoenvironmental and palaeoecological conditions associated with many of Australia’s key dinosaur faunas are poorly understood. A greater understanding of the stratigraphy, sedimentology, age, and taphonomy of these sites will provide critical context for evaluating Australia’s late-Mesozoic vertebrate taxa to other well-known Gondwanan faunas.

This study utilizes detrital zircon geochronology and Lu-Hf isotope analysis to constrain the depositional age, tectonic setting, basin evolution, and stratigraphic context of the Winton Formation, northeastern Australia. A number of geological studies in the past several decades have focused on U-Pb detrital zircon geochronology for maximum depositional age; however, this approach is still underutilized in palaeontology. Thus, these ten samples, which were composed of ~100 grain detrital zircon samples from different

VII stratigraphic levels and key fossil locations throughout the Winton and underlying units

(basin wide) were analyzed. Detrital zircon ages were obtained via U-Pb LA-ICPMS geochronology and the resulting U-Pb grain ages were subjected to various metrics to interpret maximum depositional age and sedimentary provenance. The results of this work considerably improve upon existing palynological age constraints, suggesting that there are two distinctly different aged faunas: one that is likely ~ 100-98 Ma (Isisford Fauna); and one that is no older than earliest Turonian or latest Cenomanian (92-94 Ma), which includes most other Winton vertebrates (Lark/Bladensburg Fauna).

The most abundant detrital zircon age population clusters between 92-115 Ma, suggesting that much of the volcanic-rich sediment that characterizes the Winton Formation was eroded syndepositionally or shortly after emplacement of an active continental volcanic arc system located along the eastern margin of Australia (presumably the

Whitsunday Volcanic Province). This volcanic arc activity was not a singular event; rather I have identified near-continuous detrital zircon grain ages between 92-330 Ma, indicating intermittent arc volcanism along that eastern margin due to continuous slab subduction of the Phoenix/Pacific Plate under the eastward migrating Gondwana margin. A particularly interesting result is the identification of Jurassic-age grain populations that represent a period not previously associated with significant arc magmatism in northeastern

Queensland. The identification of fairly continuous detrital zircon ages is compared with established terranes including the New England Province via Lu-Hf isotopes and found to be of similar isotopic signatures. By combining U-Pb detrital zircon ages and ƐHf values, I interpret that these sediments were derived from a mixed juvenile magma source associated with a fairly long-lived tectonic system (300-92 Ma) on the east coast of northern Australia.

VIII

In addition to this, multiple populations have been identified from other, older easterly sources (330 -900 Ma) including but not limited to the Macrossan, Anakie, Cape River,

Greenvale and Georgetown Provinces. Small populations of apparently recycled

Proterozoic and Archean grains are also variably present in each of the samples. By coupling the above results with the Kolmogorov-Smirnov Test and paleocurrent data, this study provides compelling evidence for sediment input into the Winton Formation from a long-lived continuous subduction margin along the eastern margin of Australia. Eroded materials from this region were transported via transverse fluvial systems westward into the

Eromanga Basin and form the principal provenance source for all Winton sandstones investigated in this study.

In addition to understanding provenance patterns and refining the age of the Winton

Formation, the other primary objective of this thesis was to place the important Winton flora and fauna into a refined palaeoenvironmental context. In order to achieve this, detailed facies and architectural element analysis of both the Winton and the top of the underling Mackunda Formation was conducted and a basin model was developed. Based on a combination of field and core investigation, 23 distinct lithofacies were identified, and these were used to construct nine distinct facies assemblages. The Winton Formation can be informally separated into an upper and lower unit respective of gross changes in depositional patterns, alluvial architecture, and fauna, and confirmed by distinct maximum depositional age constraining zircon populations and a newly constructed basin model.

Environmentally, the lower Winton formation preserves the last vestiges of the shallow marine conditions in the Eromanga Basin, and an up section transition from coastal and tidally influenced deltaic strata to alluvial strata. The upper Winton Formation preserves

IX mature floodplain strata, dominantd by a series of stacked channel sandstones with west- to southwest- oriented paleocurrent indicators.

Table of Contents

Statement of Access ...... I Declaration ...... II Statement of Contribution by Others ...... III Acknowledgements ...... VI Abstract ...... VII Table of Contents ...... XI List of Figures ...... XV List of Tables ...... XVI List of Digital Data ...... XVI Chapter 1 Introduction ...... 1 Structure of this Thesis ...... 1 Project Summary ...... 2 Chapter 2 Applying LA-ICP-MS U-Pb geochronology of detrital zircons to constrain the depositional age of the Winton Formation, Queensland: contextualizing Australia’s Late Cretaceous dinosaur faunas ...... 7 ABSTRACT ...... 8 1. Introduction ...... 9 2. Geological background ...... 12 2.1 Stratigraphy and sedimentology of the Winton Formation ...... 15 3. Detrital Zircon Geochronology ...... 18 3.1 Location ...... 18 3.2 Zircon Separation ...... 19 3.3 LA-ICP-MS U-Pb dating of zircon ...... 20 3.3.1 Experimental Approach ...... 20 3.3.2 U-Pb dating of the Winton Formation ...... 21 3.4 Discussion of detrital zircon geochronology ...... 22 3.4.1 Youngest detrital zircon age determination ...... 22 3.4.2 Potential sources of bias in detrital zircon studies ...... 24 4. Results ...... 24 XI

4.1 Lark Quarry sample ...... 24 4.2 Bladensburg National Park sample ...... 25 4.3 Isisford sample ...... 26 4.4 GSQ Eromanga 1 core sample ...... 26 4.5 GSQ Longreach 1 core sample ...... 27 4.6 Youngest Grains ...... 33 5. Discussion ...... 36 5.1 Maximum depositional ages ...... 36 5.2 Age of Winton Formation vertebrate fossil localities ...... 38 5.3 Recalibrating Palynomorph zones ...... 41 5.4 Implications for palaeobiogeography ...... 45 6. Conclusions ...... 46 Chapter 3 U-Pb and Lu-Hf detrital zircon provenance analysis of the Upper Cretaceous Winton Formation, Eromanga Basin, Queensland: implications for the tectonic evolution 48 ABSTRACT ...... 49 1. Introduction ...... 50 2. Geological background ...... 54 2.1 Eromanga Basin ...... 54 2.2 The Winton Formation ...... 56 2.3 Study area and sample locations ...... 57 3. Methods ...... 59 3.1 Field and core sampling ...... 59 3.2 Detrital zircon separation ...... 60 3.3 U-Pb Analysis ...... 61 3.4 Lu-Hf Analysis ...... 68 3.5 Kolmogorov-Smirnov Test ...... 69 3.6 Potential sources of bias in detrital zircon studies ...... 69 4. Results ...... 70 4.1 Winton & Mackunda Formation Temporal Placement ...... 70 4.2 Detrital Zircon Populations ...... 71 4.2.1 Mackunda Formation Grain Ages ...... 76

XII

4.2.2 Winton Formation Grain Ages ...... 78 4.3 Kolmogorov-Smirnov Test) ...... 79 4.4 Palaeocurrent Analysis ...... 83 4.5 Lu-Hf Results ...... 85 5. Discussion ...... 87 5.1 Interpreted source areas and fluvial drainage patterns for the Winton Formation ..... 87 5.2 Source & Tectonic History of Mesozoic Detrital Zircons ...... 90 5.2.1 Model-1: Long-lived Phoenix-Pacific Plate Subduction ...... 90 5.2.2 Model-2: Northeastern Gondwana S- LIP ...... 91 5.2.3 Regional Tectonic Relationships and Implications ...... 93 5.3 Lu-HF Isotope Analysis ...... 94 6. Conclusion ...... 97 Chapter 4 Depositional environments, stratigraphy and correlation of the Winton Formation: implications for Australia’s most important Mid-Cretaceous terrestrial ecosystem ...... 99 ABSTRACT ...... 100 1. Introduction ...... 101 2. Geological Background ...... 103 2.1 Basin History ...... 103 2.2 Basin Stratigraphy ...... 105 2.3 The Winton Formation ...... 109 2.4 Structural Geology of the Northern Eromanga Basin ...... 111 3. Study Area ...... 113 3.1 Field Locations ...... 113 3.1.1 Bladensburg National Park ...... 114 3.1.2 Lark Quarry Conservation Park ...... 114 3.2 Core Logging ...... 115 4. Methods ...... 116 4.1 Stratigraphy and Sedimentation ...... 116 4.2 Arc GIS Reconstruction Models and Fence Diagram ...... 117 5. Results ...... 119

XIII

5.1 Sedimentology ...... 119 5.1.1 Major Sandstones (FA1) ...... 125 5.1.2 Minor sandstones (FA2) ...... 129 5.1.3 Major Siltstones (FA3) ...... 131 5.1.4. Minor Siltstones (FA4) ...... 134 5.1.5. Major Mudstones (FA5) ...... 136 5.1.6. Minor Mudstones (FA6) ...... 138 5.1.7. Interlaminated sandy siltstones to silty mudstones (FA7) ...... 139 5.1.8. Plant-rich or Carbonaceous silty-Mudstone to Coal (FA8) ...... 141 5.1.9. Intraformational Conglomerate (FA9) ...... 142 5.2 Fossil preservation and distribution in the Winton Formation ...... 144 5.2.1 Trace Fossils ...... 144 5.2.2 Vertebrate Fossils ...... 145 5.3 Stratigraphy ...... 146 5.3.1 Outcrop Stratigraphy ...... 148 5.3.2 Core Stratigraphy ...... 152 6. Discussion ...... 156 6.1 Depositional environments ...... 156 6.1.1 Middle to Upper Mackunda; Distal Delta Front to Prodelta (subtidal platform) depositional environments ...... 157 6.1.2 Lower Winton Formation; Tidally influenced depositional environments ...... 158 6.1.3 Upper Winton Formation; Alluvial Depositional Environments ...... 159 6.2 Revised stratigraphy and correlation of the Winton Formation and its floras and faunas ...... 161 7. Conclusion ...... 164 Chapter 5 Thesis Summary ...... 168 Reference ...... 174 Appendix 1 ...... 204

XIV

List of Figures Figure 2- 1 Map of Queensland ...... 13 Figure 2- 2 Cross section of the Central Eromanga Basin, ...... 15 Figure 2- 3 Youngest graphical (histogram based) detrital zircon age peak ...... 34 Figure 2- 4 Temporal relationship of the seven metrics ...... 36 Figure 2- 5 (right) Original and revised chronostratigraphy ...... 44 Figure 2- 6 Interpreted temporal relationships ...... 44 Figure 3- 1 Map of Queensland, northeastern Australia displaying surface exposures of the Winton Formation ...... 52 Figure 3-2 Plausable source terranes ...... 72 Figure 3- 3 Youngest graphical (histogram based) detrital zircon age peak (YPP) results for all individual samples ...... 73 Figure 3- 4 Youngest graphical (histogram based) detrital zircon age peak (YPP) results based on stratigraphic placement ...... 77 Figure 3- 5 K-S test results displayed in terms of CDF’s ...... 82 Figure 3- 6 Simplified map of eastern Australia displaying the plausible source terranes .. 84 Figure 3- 7 Lu-Hf isotope results presented in ƐHf(T) ...... 86 Figure 3- 8 Simplified paleogeographic reconstruction maps ...... 88 Figure 3- 9 Tectonic reconstructions of the Eastern Australian Margin ...... 92 Figure 4- 1 Map of eastern Australia displaying surface exposures of the Winton and Mackunda formations ...... 102 Figure 4- 2 Time scale displaying Middle Jurassic to Quaternary ...... 107 Figure 4- 3 Map of exposed faults and structures within the exposed Winton and Mackunda formations ...... 112 Figure 4- 4 Arc GIS (10.1) Geo-referenced based map of exposed Winton and Mackunda Formation ...... 113 Figure 4- 5a Typical outcrop patterns of Facies Associations ...... 122 Figure 4- 5b Typical patterns of Facies Associations ...... 123 Figure 4- 6a Observed hierarchy of bounding surfaces in outcrop ...... 124 Figure 4- 6b Observed hierarchy of bounding surfaces in core ...... 124 Figure 4- 7 Outcrop Facies associations ...... 128 Figure 4- 8 Observed facies associations in the Lower Winton Formation ...... 131 Figure 4- 9 A) Displays Units A B and C ...... 147 Figure 4- 10 ArcGIS based cross section from Lark Quarry Conservation Park to Bladensburg Nation Park ...... 148 Figure 4- 11 Facies associations observed in outcrop of the upper Winton Formation ..... 151 Figure 4- 12 Facies associations observed in cored sections of the Winton and Mackunda Formation ...... 153 Figure4- 13 Simplified paleogeographic maps ...... 157 Figure 4- 11 Artistic reconstruction of the uppermost Winton Formation ...... 160

List of Tables Table 2- 1 Detrital zircon via LA-ICPMS ...... 32 Table 2- 1 A compilation of all metric's utilized within the study ...... 35 Table 3- 1 (A-E) Detrital zircon via LA-ICPMS results for all samples including isotope and concentration data ...... 67 Table 3- 2 K-S tests results ...... 81 Table 4- 1 Lithofacies codes ...... 120 Table 4- 2 Facies associations ...... 121 Table 4-3 Architectural elements ...... 127

List of Digital Data Digital Appendix 1: Publication during Ph.D

Digital Appendix 2: Detrital Zircon CL Images

Digital Appendix 3: Hylogger Data

Digital Appendix 4: All Detrital Zircon records

XVI

Chapter 1

Introduction

Structure of this Thesis This thesis is structured so that each individual chapter is a stand-alone manuscript, and ready to be published in leading international journals. The thesis is broken into three principal chapters, although I have also included a discrete additional research product in Appendix 1. Appendix 1 represents a published manuscript in which I contributed a significant component of the original research conducted during the course of my own research in the Winton Formation, but of which I am not the senior author. Each section is cross-referenced as a separate paper accordingly. Prior to the submission of this thesis, Chapter 1 and Appendix 1 both went through peer-review and were published. Chapter one; which focuses on using U-Pb detrital zircon geochronology to refine maximum depositional age constraints for key vertebrate fossils localities in the Winton Formation, is published in Gondwana Research. Appendix 1, which presents and reinterprets the famous Lark Quarry “dinosaur stampede” trackway in the Winton Formation and its depositional environments, was published in the Journal of Vertebrate Paleontology. Chapters three and four are nearly ready for journal submission, which will take place immediately following thesis examination and final thesis submission. Chapter three will be submitted to GSA Bulletin and Chapter four is going to be submitted to Sedimentary Geology for peer- review and publication. Additionally during the thesis period, one manuscript was fully published within Geobios and one is in review with Plos One. This format was agreed upon by the committee and the candidate in question during the initial stages of this project. It should be understood that while each chapter builds on the previous, each chapter is an individual body of work and is presented in similar format to the journal to which it will be submitted to. Subsequently, minor repetition of overriding goals or methodologies is inevitable, however, each chapter has been written as clearly and succinctly as possible. It should also be noted that all references used within this dissertation are grouped into a single section at the end, which follows Chapter 4.

Project Summary Recently within vertebrate palaeontology, a major focal point has been to better understand the abiotic and biotic responses to Mesozoic floras and faunas during the separation of and the subsequent isolation of many once cosmopolitan Mesozoic taxa. The current understanding of this topic is uneven and varies significantly from one Gondwanan landmass to the next. The situation is particularly grim for Australia, which is sharply constrained by deep weathering and a general lack of good late Mesozoic exposures and fossils from this time period. Australian terrestrial vertebrate assemblages for the later part of the Mesozoic are restricted to four basins, with the best exposures and fossils associated with the northeastern Eromanga Basin. Although fossil material was discovered by

Longman in central-western Queensland as early as 1933, detailed investigation of

Cretaceous vertebrate fossil from Australia has been critically hampered by a general lack of stratigraphic and palaeoenvironmental context for many specimens that were collected sporadically over the intervening 80 years. Due to this lack of local/regional sedimentologic and taphonomic context, much debate exists concerning Cretaceous vertebrates from Australia and their relationships to each other and other Gondwanan faunas. However, a suite of newly discovered Winton vertebrate and plant fossils over the last decade demonstrates potential for the Australian Cretaceous continental fossil record and promises to provide crucial information about early terrestrial floras and faunas, particularly in the Eromanga Basin. Recently discovered fossils include dinosaurs, crocodyliforms, aquatic squamates, turtles, lungfish and teleost fishes. Preliminary, taxonomic identifications are underway and already leading to a broad range of questions concerning Cretaceous terrestrial ecosystems of Australia following the fragmentation of

Gondwana (Sereno, 1997; Cracraft, 2001; Sereno, 2003; Sereno et al., 2004; Coria, et al.,

2007; Brusatte and Sereno; 2008; Ksepka and Reisz, 2008; Currie et al., 2009; D’Emic et al., 2009; Forster et al., 2009; Agnolin, 2010; Voice et al., 2011).

Thulborn et al. (1994), Hocknull et al. (2009), and Molnar (2010) represent some of the few published taphonomic and sedimentologic context reports for Cretaceous terrestrial fossil sites in the basin. Despite the importance of the fauna, poor age control has, in particular, been a major impediment to regional and global correlations, limiting the certainty of evolutionary and palaeobiogeographic assessments. As a result, it is presently impossible to place the middle to Late Cretaceous terrestrial vertebrates and plants of the

Winton Formation into a reliable context for addressing these and other pertinent questions.

Consequently, the goal of this Ph.D. dissertation was to develop and refine the geological context of the Winton Formation, particularly as it pertains to both newly discovered and historic floral and faunal fossil localities.

The first facet of this study applies U–Pb laser ablation ICPMS-MS geochronology to refine the depositional age of selected horizons within the Winton Formation (across the basin). Because my PhD project coincided, fortuitously, with the recent, but protracted set up of the LA-ICPMS facility at the Advanced Analytical Centre at JCU, my project represented one of the first detailed U-Pb geochronology projects at JCU. Considerable work to get the lab established by Dr. Yi Hu and Dr. Tony Kemp resulted in new instrumental tuning protocols, and my samples happened to be among the first samples tested using these novel procedures, hence the lab method for instrument tuning and zircon dating is included in my first chapter and stands as an important reference for all future users of the lab. The first chapter systematically investigates detrital zircon grain ages for five Winton samples, from different stratigraphic levels, many directly associated with

3 vertebrate-bearing fossil locations in the Winton Formation. Seven different metrics were compared to interpret the maximum depositional age of each detrital zircon samples.

Results indicate that sedimentation of the Winton Formation commenced no earlier than latest Albian (~103.0–100.5 Ma) and that deposition of most of the upper vertebrate fossil- rich portion of the section began no earlier than Cenomanian–Turonian boundary (~93.9

Ma). These results imply that the formation and its entombed flora and fauna were deposited primarily during the Late Cretaceous. These conclusions provide a significant advancement in understanding the age of the Winton Formation's flora and fauna, and will help to contextualize Australia's Late Cretaceous terrestrial biota within a broader

Gondwanan framework.

The second facet of my project was focused on developing better tectonic and palaeogeographic context for Winton Formation and the late Mesozoic of eastern Australia in general. Thirteen detrital zircon samples were analysed for sedimentary provenance by

U-Pb LA ICPMS geochronology and a subset of grains from these analyses were investigated further through the use of Lu-Hf isotopes. Small populations of apparently recycled Proterozoic and Archean grains were also variably present in each of the samples.

Multiple populations have been identified, which are interpreted to be sourced from other easterly sources (330-900 Ma), including, but not limited, to the Macrossan, Anakie, Cape

River, Greenvale and Georgetown Provinces. A particularly interesting result was the identification of Jurassic grain populations; a period of time not previously associated with significant arc magmatism along the eastern Australian Margin. The most abundant detrital zircon populations cluster between 92-115 Ma. This strongly indicates that much of the volcanic-rich sediment that characterizes the Winton Formation was syndepositionally or

4 near-syndepositionally eroded from an active continental arc system located along the eastern margin of Australia. The source of this volcanic material is interpreted to be from the Whitsunday Volcanic Province. This volcanic arc activity was not a singular event, but more likely a long-lived volcanic system based on the near-continuous age-spectrum of detrital zircons identified between 92-330 Ma. This most parsimoniously suggests that volcanic activity continued between the termination of the New England Orogeny (~230-

200 Ma) directly up to rifting associated with the rifting (‘unzipping’) of the Tasman Line

(~89-84 Ma). Transport of volcaniclastic sediment derived sediment rich in zircon is interpreted to have been dominantly west to south westward away from easterly to northeasterly sources. This is consistent with transverse fluvial systems that flowed across the

Eromanga Basin as the Eromanga Sea retreated to the west and then eventually to the south as the sea retreated completely. Data supports recent assertions that the thick sediment package that formed the Ceduna Delta in South Australia has a genetic link to the Winton

Formation. This is supported by not only in-field paleocurrent information but also by detailed sedimentologic data (Chapter 3) and Kolmogorov-Smirnov test statistical analyses.

The third facet of this project focussed on the interpretation of both local and regional patterns of deposition within the Winton Formation. Detailed facies and architectural analysis of the Winton and underlying Mackunda Formation were conducted through detailed field studies, extensive core logging, and the development of a basin model. Twenty-two lithofacies were identified and utilized to construct and interpret nine facies associations and their depositional environments. Ultimately, these results were coupled with palaeontological information and other data sets on the Winton Formation to reconstruct the palaeoenvironments associated with what is Australia’s most important

Cretaceous continental vertebrate record. These data also provide a more clear understanding of the palaeogeographic evolution of the Eromanga Basin during and following the retreat of the Eromanga Sea. Deposition of the Mackunda Formation is dominantd by shallow marine (middle to Lower Mackunda) to prodelta and shoreline depositional environments (Upper Mackunda). The Lower Winton Formation records a transition from marine to delta plain and fluvial systems. This work demonstrates that the initial stages of Eromanga Seaway regression began no earlier than the earliest Cenomanian

(102 Ma), based on new maximum depositional age constrain for the Mackunda Formation presented in Chapter 1. The Upper Winton Formation is dominantd by an association of alluvial floodplain deposits that hosted sinuous channel systems within a low-lying, heavily vegetated floodplain with abundant ponds and wetlands capable of supporting a significant flora and fauna. These assemblages are significantly younger than previously estimated

(Cenomanian, with most key fossil localities Turonian or younger).

Chapter 2

Applying LA-ICP-MS U-Pb geochronology of detrital zircons to constrain the depositional age of the Winton Formation, Queensland: contextualizing Australia’s Late Cretaceous dinosaur faunas

ABSTRACT The Winton Formation provides an important snapshot of Australia's late Mesozoic terrestrial biota, boasting a vertebrate fauna that includes dinosaurs, crocodyliforms, aquatic squamates, turtles, lungfish and teleost fishes, and a flora that has previously been considered to include some of the world's earliest known flowering plants. Despite its significance, poor age control has thus far prevented precise regional and global correlations, limiting the depth of palaeobiogeographic assessments. The goal of this study was to use U–Pb isotope dating of detrital zircons by laser ablation to refine the depositional age range of selected horizons within the Winton Formation. We applied this technique, with refined instrumental tuning protocols, to systematically investigate detrital zircon grain ages for five samples from different stratigraphic levels and vertebrate-bearing fossil locations throughout the Winton Formation. Seven different metrics for interpreting the maximum depositional age of each of the detrital zircon samples were compared and our results suggest that sedimentation of the Winton Formation commenced no earlier than latest Albian (~103.0–100.5 Ma) and that deposition of the upper vertebrate fossil-rich portion of the section began roughly near or after the Cenomanian–Turonian boundary (93.9 Ma), demonstrating that the formation and its important flora and fauna were deposited primarily during the Late Cretaceous. These results provide a significant advancement in understanding the age of the Winton Formation's flora and fauna, and will help to contextualize Australia's Late Cretaceous terrestrial biota within a broader Gondwanan framework.

1. Introduction Detrital zircon geochronology has traditionally been used as a provenance tool for reconstructing landscape evolution and tectonics by tracing known age populations of zircons back to their metamorphic or igneous (and in some cases recycled sedimentary) points of origin (e.g., Roback and Walker, 1995; Ireland et al., 1998; Hoskin and Ireland,

2000; Fedo et al., 2003; Adams et al., 2007). Over the past decade sedimentary provenance analysis has evolved significantly due to numerous advances in U–Pb geochronology that have made it possible to rapidly and economically date large populations of detrital minerals, in particular zircon (e.g., Ireland et al., 1998; Kowallis et al., 1998; Hoskin and

Ireland, 2000; Fedo et al., 2003; Jackson et al., 2004; Andersen, 2005; Link et al., 2005;

Gerdes and Zeh, 2006; Surpless et al., 2006; Gehrels, 2008, 2011; Barbeau et al., 2009;

Dickinson and Gehrels, 2009; Frei and Gerdes, 2009; Carrapa, 2010). This has been particularly beneficial in provenance studies focused on tectonic and paleogeographical reconstruction and landscape evolution (Hallsworth et al., 2000; Dickinson and Gehrels,

2003; Kusuhashi et al., 2006; Gonzalez-Leon et al., 2009; Roberts et al., 2012; Babinski et al., 2012; Herve et al., 2013).

Although a number of studies have recently focused on the application of detrital zircons and other minerals for maximum depositional age constraint (Dickinson and

Gehrels, 2009; Lawton et al., 2010), it is noteworthy that in paleontology, a field for which this approach has great potential, there has been little application of this technique to date

(e.g. Jinnah et al., 2009; Chure et al., 2010; Irmis et al., 2011; Varela et al., 2012). Within the field of paleontology, particularly when dealing with continental ecosystems, detrital zircon geochronology has major potential for refining the age of terrestrial floras and faunas, which are often notorious for their poor temporal and stratigraphic controls due to 9 ambiguous biostratigraphy. Detrital zircons are nearly ubiquitous in continental clastic sediments, commonly sourced from syndepositional or closely contemporaneous volcanic rocks (located within or even outside of the basin), which can often provide more precise temporal constraints than through biostratigraphy alone. Australia's Winton Formation is a perfect example of a situation where an important floral and faunal assemblage is only grossly constrained biostratigraphically, and ashbeds have yet to be identified.

Furthermore, the fact that this important succession likely straddles the middle of the

‘Cretaceous quiet zone’ makes magnetostratigraphy an unlikely option for improving age constraint. The Winton Formation is an ideal target for this application because of its relative proximity to coeval volcanic and plutonic source rocks along Australia's east to northeastern coast (New Caledonia and the volcanic rocks of the Whitsunday Volcanic

Province) that have the potential to be maximum depositional age constraining provenance sources (dated between 95 and 115 Ma; Bryan et al., 1997, 2012; Cluzel et al., 2011). Both are sparsely preserved, though currently these two sources have been strongly suggested as a potential provenance source for this middle Cretaceous zircon population (Bryan et al.,

1997, 2012; Cluzel et al., 2011).

The vertebrate fauna of the Winton Formation includes titanosauriform sauropods

(Coombs and Molnar, 1981; Molnar, 2001, 2010, 2011; Molnar and Salisbury, 2005;

Salisbury et al., 2006b; Hocknull et al., 2009), megaraptoran theropods (Salisbury, 2003,

2005; Hocknull et al., 2009; Benson et al., 2010; Agnolin et al., 2010; Salisbury et al.,

2011; White et al., 2012), thyreophoran and ornithopodan ornithischians (Salisbury, 2005;

Hocknull and Cook, 2008), basal eusuchian crocodyliforms (Molnar and Willis, 1996;

Salisbury, 2005; Salisbury et al., 2006a), dolicosaurian squamates (Scanlon and Hocknull,

2008), turtles (Molnar, 1991; Salisbury, 2005), dipnoan and teleostean fishes (Dettmann et al., 1992; Kemp, 1997; Faggotter et al., 2007; Berrell et al., 2008, 2011), possible cynodonts and basal mammaliaforms (Salisbury, 2005; Musser et al., 2009), along with likely plesiosaurs (Salisbury, 2005; Salisbury et al., 2006b). Occasional invertebrate fossils have also been recorded (Hocknull, 1997, 2000; Jell, 2004; Cook, 2005). Trace fossils from the Winton Formation include dinosaur tracks at Lark Quarry Conservation Park, indicative of two types of ornithopods and ‘possibly’ a theropod (Thulborn and Wade, 1984; Romilio and Salisbury, 2011; Romilio et al., 2013).

Upwards of 50 plant macrofossil taxa representing 10 different orders are known in the Winton Formation, dominantd by conifers and angiosperms, with rarer bennettitaleans, cycadophytes, ferns, ginkgoales (ginkgophytes) and pentoxyloleans (Bose, 1955;

Whitehouse, 1955; Peters and Christophel, 1978; Burger and Senior, 1979; Peters, 1985;

Burger, 1990; Dettmann et al., 1992; McLoughlin et al., 1995; Pole, 1998; Pole and

Douglas, 1999; Dettmann and Clifford, 2000; Pole, 2000a,b; Clifford and Dettmann, 2005;

Dettmann et al., 2009; McLoughlin et al., 2010). Cretaceous angiosperm pollen has been well studied in the Eromanga Basin by Bose (1955), Dettmann and Playford (1969), Burger and Senior (1979), Burger (1980, 1986, 1989, and 1990) Dettmann et al. (1992),

McLoughlin et al. (1995), Pole and Douglas (1999), Dettmann and Clifford (2000), and

Dettmann et al. (2009; and see references therein). The Winton Formation has been placed into Burger's (1990) Suit III biozone, characterized by Coptospora paradoxa and

Phimopollenites pannosus, and on this basis interpreted as being late Albian to early

Cenomanian in age. This pollen zone has also been assigned to samples from

11 stratigraphically lower units, including the Toolebuc, Allaru, and Mackunda formations, thereby providing solid, but broad stratigraphic control for this important interval.

In addition, there has not been any attempt to determine the stratigraphic relationship between the numerous fossil-bearing localities. As a consequence, the temporal and palaeoenvironmental context of one of Australia's most important Cretaceous terrestrial biotas and its relationship to those from other Gondwanan landmasses is poorly understood.

In lieu of unambiguous ash beds and the volcanoclastic nature of the Winton

Formation strata, the goal of our study was to utilize U–Pb detrital zircon geochronology to obtain maximum depositional age estimates for a range of fossil-bearing localities. We present the results from five samples collected from throughout the stratigraphic and geographic ranges of the formation, which significantly refine existing palynologically- based age constraints. This work has significant implications for intrabasinal correlations and understanding the relationship of the Winton vertebrate fauna and flora within a

Gondwanan, and indeed, global palaeobiogeographic and evolutionary framework.

2. Geological background

Figure 2- 1 Map of Queensland, northeastern Australia; displaying the Winton Formation within the Eromanga Basin. Symbols indicate detrital zircon sample locations derived from at or near recently discovered (Bladensburgh National Park and Isisford, Queensland) and historical fossil localities (Lark Quarry Conservation Park) along with samples derived from Geological Survey of Queensland stratigraphic core logs (GSQ Eromanga 1 and GSQ Longreach 1). Also displayed are a number of historical archosaurian fossil assemblage localities throughout the Winton Formation (A–D) based on described findings from but not limited to: Thulborn and Wade (1979), Coombs and Molnar (1981), Thulborn and Wade (1984, 1979, 1989), Dettmann et al. (1992), Molnar (2001), Molnar and Salisbury (2005), Salisbury et al. (2006a,b), Agnolin et al. (2010), Hocknull et al. (2009), and Molnar (2010).

The Eromanga Basin forms a major portion of the Great Artesian Basin (GAB; Fig. 2-1),

the tectonic history of which is still a matter of debate. Studies such as those by Gallagher

(1990) have proposed that the GAB occupied a foreland basin setting during the

Cretaceous, whereas others, such as Gray et al. (2002), suggest a more complex,

13 intracratonic basin setting. Either way, there is consensus that either a major arc- or rift- related volcanic system was located on the eastern margin of the basin, providing a synorogenic source for the abundant feldspathic and volcanolithic petrofacies that characterize the Winton Formation (Bryan et al., 1997). Draper (2002) and others also demonstrated that a large portion of the basin was inundated by an empiric seaway during the Cretaceous as a result of globally elevated sea levels. There is evidence for two major transgressive–regressive cycles in the basin, though direct correlation to global events is problematic (Gallagher and Lambeck, 1989). Gallagher and Lambeck (1989) suggest that the voluminous input of volcanic detritus into the basin is directly related to the transgression and regression events rather than global effects, although this is debated

(Draper, 2002). The richly fossiliferous, alluvial to coastal plain strata of the Winton

Formation were deposited following the final regression of the interior sea from central

Australia (Draper, 2002).

The Winton Formation is exposed over large portions of Queensland and portions of northern New South Wales, north-western South Australia and the south-western corner of the Northern Territory (Fig. 2-1). Originally designated the “Winton Series”, the unit was initially described by Dunstan (1916), as a succession of terrestrial (alluvial) sandstones, shales and minor coal seams. The Winton Formation overlying the dominantly marine Rolling Downs Group, originally defined by Jack (1886) and Dunstan (1916) and unconformably overlain by modern fluvial deposits in the form of ‘black soil’. However, the first mention of the Winton Formation and regional synthesis of the Cretaceous nomenclature was not until Whitehouse (1954, 1955), Exon (1966), Casey (1970) and also

Senior and Mabbutt (1979), all of whom defined the formation as the upper portion of the

Rolling Downs Group and designated it as an Upper Cretaceous continental unit conformably overlying the marine Mackunda Formation and provided the first detailed sedimentologic characterization of the horizons comprising it. More recently, Gray et al.

(2002) grouped the Winton together with the Mackunda Formation to form the Manuka

Subgroup (Fig. 2-2).

Figure 2- 2 Cross section of the Central Eromanga Basin, Queensland, with stratigraphic positions of analyzed zircon samples, modified from Draper (2002). Stratigraphic cross-sections A–A′ was based on GSQ McKinlay 1 south to Isisford Water Bore Log, and includes stratigraphic position of detrital zircon samples derived from Lark Quarry Conservation Park, Bladensburg National Park and Isisford, Queensland. Stratigraphic cross-sections B–B′ is a cross-section form GSQ Longreach 1, Isisford Water Bore Log and GSQ Eromanga 1, and includes stratigraphic position of detrital zircon samples derived from GSQ Longreach 1 and GSQ Eromanga 1. (1*) denotes that above the in places (Lark Quarry and Bladensburgh National Park) that the Winton Formation is exposed in ‘jump up's’ which is capped by laterite and Quaternary alluvium, whereas in other locations (Isisford, Queensland) the upper Winton Formation is very poorly exposed and blanketed by 'black soil'.

2.1 Stratigraphy and sedimentology of the Winton Formation This study focuses on vertebrate fossil-bearing localities within the Winton

Formation near Winton (Bladensburg National Park and Lark Quarry Conservation Park) and Isisford (Fig. 2-2). Additional investigations of GSQ cores (GSQ Eromanga 1 and GSQ

Longreach 1) were conducted to place these sites into a regional framework. Exposures of the Winton Formation between Lark Quarry and Bladensburg National Park are continuous and dominantd by fine-grained sandstones, siltstones and mudstones. Much of the exposure at both localities is heavily weathered, characterized by kaolinized feldspars and volcanic detritus. Sandstones are typically feldspathic to feldspatholithic, fine to medium grained, well sorted, and exhibit sub-angular to subrounded grains. Sedimentary structures commonly noted in outcrop. Sedimentary structures commonly noted in outcrop at both localities include, but are not limited to: asymmetrical and symmetrical current ripple cross- lamination, trough and tabular cross-stratification, scour and fill structures, and lenticular channel macroform elements. Mudstones are commonly gray, purple or olive green, and in some areas exposures reveal ‘shrink and swell’ smectitic characteristics, including haystack mounds and popcorn texturing. Fossil material is commonly exposed at or near the base of

(locally known as 'jump-ups'). Well-preserved associated to poorly articulated body fossils, along with mollusk shells, are often only found 1.0–3.0 m below the surface. Commonly surface material is typically limited to weathered fossil bone shards.

In contrast, the sedimentology at Isisford presents a dramatic departure from the suite of facies seen at the Bladensburg National Park and Lark Quarry Conservation Park sites, with much of the fossil material preserved in medium-to-coarse grained Fe-oxide and calcite cemented sandstone nodules, some of which are still in situ. The Winton Formation in this area is poorly exposed and mostly covered by ‘black soil’, a recent alluvial deposit that blankets much of the Eromanga Basin (Twidale, 1966). The large fossil-bearing nodules (many exceeding 3–4 m in diameter) are located at or near the weathering surface, but their precise stratigraphic position within the Winton Formation is difficult to constrain

16 precisely. Only some of the nodules are in situ, including the dinosaur bearing nodule used in this study. Those which are not may represent surface lag, concentrated as erosion of softer, poorly cemented led to stratigraphic condensation of the nodules. Based on correlation with nearby well and core logs (ESSO Australia Isis Downs-1, GSQ Longreach

1, and GSQ Maneroo 1) the stratigraphic position of the Winton Formation surface exposures (and nodule) in the Isisford area is towards the top of the unit in this part of the

Eromanga Basin.

The overall character of strata at Bladensburg National Park and Lark Quarry

Conservation Park suggests a fluvial origin, with evidence of channels, floodplain mudrocks, crevasse splays, abandoned channels, and oxbow lakes, representative of a broad and expansive fluvial floodplain system. Although some previous studies suggest a dominantly lacustrine setting for parts of the Winton Formation, such as at the famous dinosaur ‘stampede’ tracksite at Lark Quarry (Thulborn and Wade, 1979, 1984, 1989), we regard this as a relatively less abundant facies within the broader Winton depositional system. Much of the Winton Formation lacks significant surface exposure due to overlying alluvium and deep weathering; however, trenches excavated near fossil localities indicate that the subsurface facies are similar to those at Lark Quarry and Bladensburg National

Park.

3. Detrital Zircon Geochronology

3.1 Location Three samples for detrital zircon geochronology were acquired at, or near, key vertebrate fossil localities within the Winton Formation, while two other samples come from Geological Survey Queensland (GSQ) cores (Fig. 2-2). Sampling was conducted to provide spatial coverage across the Winton Formation field areas and range stratigraphically from just below the base of the Winton Formation (i.e., at the top of the

Mackunda Formation; GSQ Longreach 1) to the upper-most exposed Winton Formation outcrop in western Queensland (Lark Quarry Conservation Park) (Fig. 2-2). In total five samples for detrital zircon analysis were collected. Bulk samples (~5–10 kg each) were extracted from freshly exposed, unweathered fine- to medium-grained sandstones, interpreted as low-energy fluvial channel deposits except for Bladensburg National Park, which we interpret as a proximal floodplain deposit.

The first sample was collected from Lark Quarry Conservation Park (Lat:

23°00′58.62″S; Long: 142°24′42.73″E) and was taken from ~1.0 m above the horizon that preserves the famous Lark Quarry dinosaur tracksite (Thulborn and Wade, 1979, 1984,

1989). This was stratigraphically, the highest sample collected from the Winton Formation, near the top of the section. The second highest Winton Formation sample comes from

Bladensburg National Park (Lat: 22°30′53.35″S; Long: 143°02′19.89″E), ~2–3 m above the main fossil-bearing horizon in the park, slightly lower than the Lark Quarry trackway horizon. The third sample was collected from a locality at Isisford (Lat: 24°15′33.51″S;

Long: 144°26′29.35″E). It was taken directly from a sandstone nodule that preserves the remains of a presently undescribed dinosaur that is in close proximity to other key fossil- bearing localities in the area (see Salisbury et al., 2006a). The nodule was discovered in- 18 situ in the uppermost portion of the bedrock, but most of it was initially buried by the

‘black soil’, which directly overlies most of the upper-most unexposed Winton Formation within the Isisford region. The final two samples came from cores drilled in central-western

Queensland by Geosciences Queensland (GSQ), including one sample from the GSQ

Eromanga 1 core (ERO1-11-001). GSQ Eromanga 1 core was drilled east of Eromanga,

Queensland (Lat: 26°36′52.93″S; Long: 143°52′48.41″E) and this sample (ERO1-11-001) was collected from the core interval between 16.5 and 26.7 m (boxes 2–5) below surface

(roughly 146.0 m above base of the Winton Formation). This sample is taken from the upper-most subsurface section of the recovered core. A second sample was taken from the

GSQ Longreach 1 (LO1-11-001) core, which was drilled northeast of Longreach,

Queensland (Lat: 23°06′42.93″S; Long: 144°35′58.41″E) and (LO1-11-001) was collected from the core interval between 20.8 and 35.8 m (boxes 1–6) below surface of retrieved core. This sample is collected from the upper most portions of the Mackunda Formation

(Fig. 2-2).

3.2 Zircon Separation All samples were crushed and milled in a tungsten carbide disc mill and then sieved using both 250 and 500 μm meshes to enhance the possibility of sampling potential distal volcanic ash derived zircons. The material was washed and decanted numerous times to remove the clay-sized fraction. Heavy minerals were separated using lithium polytungstate adjusted to a specific gravity of 2.85–2.87. Mineral separates were then washed, dried, and a hand magnet was used to remove strongly magnetic minerals, followed by a Frantz magnetic separator at progressively higher magnetic currents of 0.5, 0.8, 1.3, and 1.5, set at a constant 10° side slope. The non-magnetic heavy mineral separates were then selected via hybrid selection protocols that involved handpicking zircons as randomly as possible from

19 the greater population within a defined field of view. Following this, the remainder of each sample was handpicked a second time with the intention of selecting the clearest, most euhedral grains remaining in each sample; with the goal of increasing the likelihood of analyzing maximum age constraining young zircons. For each sample, ~100–150 grains were mounted in a 25 mm epoxy resin puck, polished to expose their mid-sections and imaged using a Jeol JSM5410LV scanning electron microscope with attached cathodoluminescence detector in order to document microstructures, cracks, inclusions and other complexities.

3.3 LA-ICP-MS U-Pb dating of zircon

3.3.1 Experimental Approach All work was done at the Advanced Analytical Centre of James Cook University, using a Coherent GeolasPro 193 nm ArF Excimer laser ablation system connected to a

Bruker 820-MS(formerly Varian 820-MS). The ablation cell was connected to the Bruker

820-MS via Tygon tubing and a 3-way mixing bulb (volume ~5 cm3). The standard cylindrical sample cell was used throughout the study, but with a custom-designed polycarbonate insert to reduce the effective volume to 4 cm3 (see Supplemental materials).

This insert combined with the mixing bulb provides both a very stable time-resolved signal and rapid signal washout.

The Bruker 820-MS employs an ion mirror design, which reflects the ion beam exiting the skimmer cone by 90° and focusses this into the mass analyzer. Non-ionized large particles and neutrals, as well as partially ionized particles, are not reflected and extracted by a pump located behind the mirror. In this way, the electrostatic mirror acts as a particle size filter to admit only fully atomized and ionized particles into the quardupole mass filter and detector. The advantage of this unique configuration is that it facilitates

20 tuning of the ICP-MS to minimise instrumental mass fractionation focusing on the key ratio of Pb/U, as described below. The instrumental parameters and operating conditions are provided in the Supplemental materials.

All instrument tuning was performed using a 5 Hz repetition rate, 44 μm beam aperture and 6 J/cm2 energy density, as determined by energy meter at the ablation site.

Under these conditions, the ablation rate for NIST 610 and zircon was about 0.1 μm per laser pulse and 0.06 μm per laser pulse respectively. Tuning was achieved by iteratively adjusting the He carrier gas, Ar sampling gas, sheath gas flow rate, RF Power, 1st, 2nd, 3rd

Extraction lens and corner lens voltage to achieve 238U/232Th ratio ~1, ThO/Thb1% typically 0.5% and 206Pb/238U ~0.22 in NIST610. Tuning the instrument towards the

‘true’ 206Pb/238U ratio thus minimises the magnitude of the total Pb/U fractionation correction applied zircon analyses, reducing the inherent uncertainties in this correction procedure where there are large age differences between standard and sample zircons.

Using this technique improved the accuracy and reproducibility of zircon U–Pb isotope analysis in our laboratory (See Supplementary Table 2-1A–E). For sample analysis, the total measurement time was set at 65 s. The first 30 s was for gas blank measurement (laser firing but with the shutter closed), with the shutter opened to allow sample ablation for the final 35 s, standard bracketing was used throughout the study to correct for remaining elemental fractionation and mass bias.

3.3.2 U-Pb dating of the Winton Formation Approximately 100 detrital zircons for each of the five samples were analyzed using the optimized LA-ICP-MS tuning method outlined above. A 32 μm beam diameter was used for the following samples: Lark Quarry Conservation Park, GSQ Eromanga 1, and

GSQ Longreach 1. However, due to the small grain size of samples from Bladensburg

National Park and Isisford, a 24 μm beam diameter (spot size) was used for these samples.

As different aspect ratios (e.g. laser pit diameter/depth ratio) will have different elemental fractionation characteristics, when one beam size is used (either 24 μm or 32 μm beam), this beam size is uniformly applied to all zircon to be dated including calibration zircon standard. Over the duration of ablation, groups of 10–12 zircon grains were analyzed, followed by at least two analyses each of a primary (GJ-1, 609 Ma, Jackson et al., 2004) and secondary in-house zircon standard (Temora-2 [TEM-2] 416.8 Ma, Black et al., 2003).

If grains exhibited a greater discordance of 30%, those grains were omitted from the populations and the study as a whole. All standard analyses were within 2% of the expected ages, and most were within 1% of the expected age. NIST 610 or 612 was analyzed at the beginning and end of each session, and at least once in between, for the purpose of calibrating Th and U concentrations.

3.4 Discussion of detrital zircon geochronology

3.4.1 Youngest detrital zircon age determination When trying to constrain the maximum depositional age of strata or fossil localities, it is a common practice to identify the youngest detrital zircon grains in a sample in order to determine the youngest possible age of deposition (Rainbird et al., 2001). Recent studies, including a robust investigation by Dickinson and Gehrels (2009), have reappraised the various methodologies for ascertaining the maximum depositional age and highlighted the importance of utilizing several different approaches to obtain the most accurate representation of the maximum depositional age of a sample. This study compares seven of the most common metrics used to determine the maximum depositional age of a detrital zircon datasets (e.g., see Dickinson and Gehrels, 2009; Johnston et al., 2009; Lawton and

Bradford, 2011; Robinson et al., 2012), including:1) youngest single grain age (YSG); 2)

22 youngest graphical detrital zircon age (YPP); 3) youngest detrital zircon age (YDZ); 4) the weighted mean average age (YC1σ) (+3); 5) Weighted Average; 6) weighted mean average age (YC2σ) (+3); and 7) TuffZirc (Zircon Age Extractor) (+6).

YSG involves singling out the youngest detrital zircon grain within the population.

However, this approach is highly questionable and the least rigorous method because it represents only a single data point (Dickinson and Gehrels, 2009). As such, this study employs six other methods to provide a more reliable youngest maximum depositional age approximation. YPP is the youngest graphical detrital zircon age peak recorded on the histogram; and is obtained by identifying the first maximum age peak (several grains or grain cluster (DZ≥3)) along an age-probability plot or age distribution curve, calculated within ISOPLOT (Ludwig, 2009; Dickinson and Gehrels, 2009; Lawton and Bradford,

2011). YDZ is calculated using an algorithm within ISOPLOT, which extracts the youngest subset of DZ ages within the whole DZ population using a Monte Carlo analysis

(Dickinson and Gehrels, 2009; Ludwig, 2009). YC1σ (+3) and YC2σ (+3) or the weighted mean averages at 1 and 2 σ, respectively, incorporate both internal analytical error and external error of the youngest population within the tested group of grains (Dickinson and

Gehrels, 2009; Jones et al., 2009). Dickinson and Gehrels (2009) and others apply a minimum of two grains (n≥2) to achieve a maximum depositional age result. However in this study, we used a minimum three grain cluster limit. YC1σ (+3) and YC2σ (+3) are derived from the AGE PICK program, generated by the University of Arizona LaserChron

Center. Weighted Average is also an algorithm within Isoplot that utilizes age values and errors to generate an inverse variance-weighted average that helps to deal with excessive scatter within a batch of grains (Ludwig, 2009). The last metric applied in this study is the

TuffZirc age extractor (+6), an Isoplot algorithm (originally based on the TuffZirc algorithm by Ludwig and Mundil (2002)) that implements a mathematically based approach on the loss and inheritance of Pb and its error (Ludwig, 2009). It is suggested by

Ludwig (2009) to include 10 or more grains within the process. However, as this study seeks the youngest maximum depositional age, the youngest six grains were utilized from each population as it is the minimum number of grains required for the program to run.

Together, seven methodologies are utilized and evaluated for each sample in order to the most reliable maximum depositional age for the Winton Formation.

3.4.2 Potential sources of bias in detrital zircon studies A number of recent studies have highlighted the potential for error and bias in detrital zircon studies. Most of these studies focus on but are not limited to hydrological sorting, transport variability and affects to deposition by climate (Hietpas et al., 2011; Lawrence et al., 2011). These biases mostly affect studies dealing with provenance-based detrital zircons studies, and have little bearing on studies focused on maximum depositional age reconstruction. The obvious exception is that techniques, such as multiple sampling of grains and high resolution CL imaging to identify grain irregularities or cores, are in fact pertinent to maximum depositional age-based studies like this one, and were applied to this study.

4. Results

4.1 Lark Quarry sample Zircon grain morphology ranges from euhedral to abraded and well rounded.

Overall zircon grains derived from these rocks are smaller (75 μm≥n≤100 μm) than those in other samples. Ninety grains were analyzed (76 reported), yielding multiple age

24 populations, mostly composed of various Mesozoic populations (70%), with the remaining grains dominantly from Paleozoic sources (24%), and a few Pre-Cambrian grains (6%). Of the Mesozoic grains, the dominant population is within the Cretaceous (68%), of which

33% are Late Cretaceous and 67% are Early Cretaceous. The remaining populations are divided between the Jurassic (13%) and Triassic (19%). Analytically, this sample yields the youngest maximum detrital zircon grain ages of all five samples. Results of the seven methodologies are as follows: YSG is 92.5 (±1.2) Ma (Table 2-1a), YPP is 93.0 Ma (Fig. 2-

3,Table 2-2), YDZ is 94.5 (+2.1/−2.3) Ma (Table 2-2), YC1σ (+3) is 94.5 (±1.8) Ma (Table

2-2), Weighted Average is 94.5 (±5.3) Ma (Table 2-2), YC2σ (+3) is 94.5 (±3.1) Ma (Table

2-2), and TuffZirc is 97.5 (+0.3/−3.0) Ma (Table 2-2).

4.2 Bladensburg National Park sample Zircon grain morphology ranged from euhedral to abraded and well rounded, though they were distinctly larger than those collected at Lark Quarry (100 μm≥n≤200 μm).

Ninety grains were analysed (82 reported), yielding multiple age populations, mostly composed of various Mesozoic populations (65%), with the remaining grains of Paleozoic

(26%) and a few Pre-Cambrian grains (9%). Of the Mesozoic grains, the dominant population is Cretaceous (68%) of which 25% are Late Cretaceous and 75% are Early

Cretaceous. The remaining populations are divided between Jurassic (13%) and Triassic

(19%). Analytically, this sample produced the second youngest detrital signature and the seven methods’ results were as follows: YSG is 93.3 (±1.2) (Table 2-1b), YPP is 96.0 Ma

(Fig. 2-3, Table 2-2), YDZ is 93.8 (+1.9/−1.8) Ma (Table 2-2), and YC1σ (+3) is 94.0

(±1.7) Ma (Table 2-2), Weighted Average is 94.0 (±1.4) Ma (Table 2-2), YC2σ (+3) is 94.5

(±2.9) Ma (Table 2-2), and TuffZirc is 95.6 (+2.2/−2.3) Ma (Table 2-2).

4.3 Isisford sample Zircon grains are typically large with well-preserved euhedral crystals, although older grains are commonly abraded and well rounded, and average grain size ranged between 100 μm≥n≤150 μm. Seventy two grains were analyzed (62 reported), yielding multiple age populations, mostly composed of various Mesozoic populations (65%), and lesser Paleozoic (22%) and a few Pre-Cambrian grains (13%). Of the Mesozoic grains, the dominant population is Cretaceous (63%) of that Cretaceous population 100% are Early

Cretaceous. The remaining populations are divided between Jurassic (7%) and Triassic

(27%). Analytically, this sample produced the second oldest maximum depositional ages of all five samples. The seven age metrics are as follows: YSG is100.5 (±1.1) Ma (Table 2-

1c), YPP is 103 Ma (Fig. 2-3, Table 2-2), YDZ is 101.3 (+1.8/−1.8) Ma (Table 2-2), YC1σ

(+3) Isisford is 101.6 (±1.8) (Table 2-2), Weighted Average is 101.6 (±1.3) Ma (Table 2-2),

YC2σ (+3) is 101.6 (±2.9) Ma (Table 2-2), and TuffZirc is 102.2 (+1.9/−1.9) Ma (Table 2-

2).

4.4 GSQ Eromanga 1 core sample Zircon grain morphology ranges from euhedral to abraded, well rounded grains with average grain size between 90 μm≥n≤200 μm. Also, these zircons were the most varied in colour of all the samples taken. Ninety-nine grains were analyzed (99 reported), yielding multiple age populations, mostly composed of various Mesozoic populations (80%) and much fewer Paleozoic grains (15%) and Pre-Cambrian grains (5%). Cretaceous grains

(84%) dominantd the Mesozoic population, of which 6% are Late Cretaceous and 94% are

Early Cretaceous. The Jurassic (6%) and Triassic (10%) grains represent the other

Mesozoic populations. Maximum depositional ages are as follows: YSG is 93.0 (±1.1) Ma

(Table 2-1d), YPP is 96.0 Ma (Fig. 2-3, Table 2-1), YDZ is 95.6 (+1.8/−1.9) Ma (Table 2-

1), YC1σ (+3) is 95.2 (±1.6) Ma (Table 2-1), Weighted Average is 101.6 (±1.3) Ma (Table

2-1), YC2σ (+3) is 95.2 (±2.7) Ma (Table 2-1), and TuffZirc is 101.1 (+1.3/−1.4) Ma

(Table 2-1).

4.5 GSQ Longreach 1 core sample Zircon grain morphology is characterized by notably fewer euhedral crystals. Many of the grains within the sample are abraded and well-rounded with average grain size between 90 μm≥n≤150 μm. Seventy-three grains were analyzed (69 reported), yielding multiple age populations, represented by Mesozoic (57%), Paleozoic (29%) and a few Pre-

Cambrian (14%) grains. Early Cretaceous grains (57%) dominant the Mesozoic populations with fewer Jurassic (25%) and Triassic (18%) grains. Analytically this sample contains the oldest maximum depositional ages; YSG is 102.5 (±1.3) Ma (Table 2-1e), YPP is 104 Ma

(Fig. 2-3, Table 2-2), YDZ is 102.8 (+1.7/−2.1) Ma (Table 2-2), YC1σ (+3) is 103.1 (±2.0)

Ma (Table 2-2),Weighted Average is 103.2 (±1.5)Ma (Table 2-2), YC2σ (+3) is 103.1

(±3.5)Ma (Table 2-2), and TuffZirc is 104.3 (+1.6/−1.8) Ma (Table 2-2).

Table 2- 1 Detrital zircon via LA-ICPMS results for all samples including isotope and concentration data: Th (in PPm) is Thorium concentrations measured in parts per million; U (in PPM) is Uranium concentrations in parts per million; Th/U is the divided ratio of Uranium concentrations; Both Pb206/U235 and error calculated at ±1σ and Pb206/U238 and error calculated at ±1σ were used to calculate and Error Correction Value (Error Corr.). The ratio values for Pb206/U238, Pb207/Pb206, and Pb207/ Pb206 with all error calculated at ±1σ with the preferred age (Pb206/U238) with error ±1σ interpreted as YSG. Age is calculated by error propagation and included errors and U and U/Th Concentration are calculated against NIST 612 and are within ~30% error.

4.6 Youngest Grains Of the five detrital zircon analyses, three samples (Lark Quarry, Bladensburg

National Park, and GSQ Eromanga 1) had three or more grains at or younger than the

Albian-Cenomanian boundary (n≤100.5 Ma). To test the robustness of the youngest maximum depositional ages, the twenty-four youngest grains in this study (effectively all

Late Cretaceous grains) were grouped and analyzed as a separate sample. Using this approach, the maximum depositional age metrics for the entire study (n=388 grains) produces the following ages: YSG is 92.5 (±1.2) Ma (Table 2-2, Supplementary Table 2-

1a,b,d), YPP is 98 Ma (Table 2-1), YDZ is 91.8 (+1.6/−2.3) Ma (Table 2-2), YC1σ (+3) is

92.9 (±1.7) Ma (Table 2-2),Weighted Average is 92.9 (±1.3) Ma (Table 2-2), YC2σ (+3) is

92.9 (±2.9) Ma (Table 2-2), and TuffZirc is 93.5 (+2.1/−1.0) Ma (Table 2-2) (see

Supplementary Fig. 1 for Concordia Plots). Using this approach, one can make a much stronger argument for a younger (post-Cenomanian/Turonian boundary) depositional age for at least the upper fossiliferous portion of the Winton Formation.

Figure 2- 3 Youngest graphical (histogram based) detrital zircon age peak (YPP) results for all samples (Lark Quarry Conservation Park at 93 Ma; Bladensburg National Park at 94 Ma; 103 Ma; fossil entombing nodules from Isisfor Queensland at 103 Ma; GSQ Eromanga 1core at 96 Ma; and GSQ Longreach 1 core at 104 Ma), based on the first maximum age peak (several grains or grain cluster (DZ≥3)) along an age-probability plot or age distribution curve presented and calculated within ISOPLOT (Dickinson and Gehrels, 2009; Ludwig, 2009).

Table 2- 1 A compilation of all metric's utilized within the study. YSG of all six populations presented with a ±1σ error. YPP or the graphical age based on a histogram derived from Isoplot (see Fig. 3). YDZ of all six populations is presented with the Age, range (+ or −) and the confidence of that metric. YC1σ (+3) of all six populations is presented with the Final Age, Weighted Mean Average, Systematical error and MSWD at a ±1σ error. Weighted Average of all six populations is presented with the Age, Confidence, the amount of grains rejected, MSWD and the overall probability. YC2σ (+3) of all six populations is presented with the Final Age, Weighted Mean Average, Systematic Error, and MSDW at a ±2σ error. TuffZirc of all six populations is presented with the Final Age, Confidence, and the Group Size (the number excluded against the grains used by the metric).

5. Discussion

5.1 Maximum depositional ages The aim of this study was to determine the youngest maximum deposition age of the Winton Formation and to temporally constrain its globally significant fauna and flora

(Fig. 1-4). The youngest single grain ages (YSG) for all five samples range from 92.5 Ma at

Lark Quarry, at the top of the section, to 102.5 Ma in the GSQ Longreach 1 core, from the uppermost sections the Mackunda Formation. If the single grain age is taken as a proxy for the youngest maximum age of the lower to middle portions of the Winton Formation, it strongly supports continued deposition of the formation into the Turonian or earlier

(Grandstein and Ogg, 2004; Walker and Geissman, 2009; Ogg and Hinnov, 2012) (Fig. 1-

4).

Figure 2- 4 Temporal relationship of the seven metrics (YSG, YPP, YDZ, YC1σ (+3), Weighted Average, YC2σ (+3), and TuffZirc (+6)) utilized with in this study, of the six populations (Lark Quarry Conservation Park, Bladensburg National Park, Isisford, GSQ Eromanga 1, GSQ Longreach 1 and the youngest combined population). Temporal constraint is suggested for the faunal assemblages at each of the three fossil localities (Lark Quarry Conservation Park, Bladensburg National Park, and Isisford, Queensland).

However, these ages are based on single grains and may be erroneous; hence they should be used with caution. The value of identifying YSG ages is that they may actually indicate a small, but true population of younger grain ages that can be verified through additional zircon analyses. On the other hand, the application of the TuffZirc and YPP metrics demonstrably produce the oldest ages, and is considered much less sensitive for the interpretation of the maximum youngest depositional age. Given this insensitivity, we recommend against using these two metrics for this application. Thus, excluding the insensitive TuffZirc and YPP ages, along with the potentially inaccurate YSG ages, we find that each of the other metrics yields relatively similar and robust maximum depositional age constraints for the upper fossiliferous portion of the Winton Formation (i.e., Lark

Quarry, Bladensburg National Park, GSQ Eromanga-1), between 92.5 Ma and 95.6 Ma.

However, when all samples are compiled (n=388 grains), there is a solid population, and the application of the four preferred maximum depositional age metrics strongly implies a post-Cenomanian age (≤93.6 Ma) for the depositional age of the upper fossil-bearing portion of the Winton Formation. Based on this study, the base of the Winton Formation appears to be at or near the Albian/Cenomanian boundary, whereas the upper Winton

Formation very likely continued into or past the early Turonian. It should also be highlighted that in the review by Dickinson and Gehrels (2009) the following variances within the methodologies are as follows; YSG and YC1σ (+3) on average are within 5 Ma of deposition, YDZ is within 2 Ma, YPP only 5% of the time younger than that of depositional age and YC1σ (+3) is older by 10 Ma and is not a more reserved measure.

Also it should be noted that these derived zircon ages only represent an age for the youngest sediment sources and are not a true depositional age of the formation.

Additionally, given that the uppermost 10–20 m of the Winton Formation was not sampled, 37 it may well be that these horizons preserve detrital zircons that are younger still. Minimally, the results of this study suggest that the uppermost portions of the Winton strata in the northern part of the basin were deposited near to or after the Cenomanian/Turonian boundary. Our findings significantly improve upon continental biostratigraphic age control for the Winton Formation and support assertions by Helby et al. (1987) that the deposition of the Winton Formation may extend into the early Turonian.

5.2 Age of Winton Formation vertebrate fossil localities The stratigraphic position of the various samples that we have considered and their respective ages indicate that the age of the basal-most Winton Formation and its overlying thickness vary throughout the basin. Fielding (1992) has shown that in marginal regions of the Eromanga Basin, the Winton Formation is less than 100 m thick, but that the unit thickens to more than 1200 m in central areas about the Queensland-South Australian border. The Isisford sample, which sits approximately 200 m above the inferred contact with the Mackunda Formation, is here dated as late Albian. This age is closer to that of the geographically proximate GSQ Longreach 1 sample than any of the other samples that we considered.

Our results indicate that vertebrate fossil-bearing localities in the Winton area (Lark

Quarry and Bladensburg National Park) are of a similar age (statistically close to overlapping), deposited near to or after the early Turonian–late Cenomanian, respectively.

This result is consistent with the very shallow dip (~8°) in the broader Winton area, and the apparent continuity of strata between the two localities. Other fossil sites in the Winton area, such as those on Alni Station, 50 km NW of Winton (Lat: 22°11′00″S; Long:

142°28′00″ E; Coombs and Molnar, 1981; Molnar, 2001, 2010, 2011; Molnar and

Salisbury, 2005), Belmont Station, 60 km NE Winton (22°5′S, 143°30′E; Clifford and

Dettmann, 2005; Salisbury, 2003, 2005; Salisbury et al., 2006a,b, 2007; Scanlon and

Hocknull, 2008; Hocknull and Cook, 2008) and Lovelle Downs Station/Elderslie Station,

48 km WNW of Winton (Lat: 22°11′59″S; Long: 142°31′43″E; Dettmann et al., 2009;

Hocknull et al., 2009; White et al., 2012) are therefore likely to be of similar depositional history, and would be no older than the Cenomanian/Turonian boundary. The latter sites traverse a low ridge between Lovelle Downs Station and Elderslie Station, and are all within a few kilometers of each other (Dettmann et al., 2009; Hocknull et al., 2009; White et al., 2012). This area is 30 km northwest of Bladensburg National Park and 100 km northwest of Lark Quarry Conservation Park. Previously, Dettmann et al. (2009; whose age assessment was followed by Hocknull et al., 2009) proposed that the floral assemblage from the Lovelle Downs/Elderslie locality was from the late Albian. This assessment was based primarily on the presence of Cicatricosisporites and Crybelosporites pollen in association with Clavatipollenites and Phimopollenites, indicating that the sediments were within the C. paradoxa or P. pannosus spore–pollen zones (of Helby et al., 1987) and therefore could not be older than middle Albian. A late Albian age was proposed by

Dettmann et al. (2009) because the sediments that the fossils were found in were within

~265 m of the contact with the underlying Mackunda Formation (Watson, 1973). However, based on the relatively close proximity of the Lovelle Downs/Elderslie locality to

Bladensburg National Park (~30 km), and the likely stratigraphic continuity between these two areas and Lark Quarry, we suspect that an early Turonian–middle Cenomanian age is more likely.

There are notable differences between the vertebrate assemblages of Isisford and those from the Winton area (Bladensburg, Lark Quarry, Elderslie, Lovelle Downs, Alni,

Belmont). The remains of titanosauriform sauropods are the most abundant vertebrate fossils encountered at sites around Winton (Coombs and Molnar, 1981; Molnar, 2001,

2010, 2011; Molnar and Salisbury, 2005; Salisbury et al., 2006b; Hocknull et al., 2009).

Those that have been described to date are usually preserved as either isolated elements or as partial, disarticulated and associated skeletons, often in multi-individual bone beds and associated with isolated and frequently reworked micro-vertebrate remains (Salisbury,

2005; Salisbury et al., 2006b; Hocknull et al., 2009; Molnar, 2010). Disarticulated or semi- articulated and associated skeletons, such as the holotype of Diamantinasaurus matildae

(Hocknull et al., 2009), have proven to be much less common. The latter specimen was found in association with a partial, disarticulated theropod skeleton, Australovenator wintonensis (Hocknull et al., 2009; White et al., 2012). Thus far, other smaller-bodied taxa from Winton sites have only been described on the basis of isolated elements (Dettmann et al., 1992; Kemp, 1997; Salisbury, 2003, 2005; Hocknull and Cook, 2008; Scanlon and

Hocknull, 2008; Musser et al., 2009; Salisbury et al., 2011). The Isisford assemblage, on the other hand, is thus far devoid of sauropods. The most commonly encountered vertebrates are teleost fishes (Berrell et al., 2008, 2011) and small-bodied crocodyliforms

(Salisbury et al., 2006a). Dinosaurs are rare, but those that have been discovered and prepared to date are all small (cow-sized or smaller; Fletcher et al., 2009). In nearly all instances, the vertebrate fossils from the Isisford localities are either near-complete or partial skeletons, with elements preserved in full articulation, semi-articulation or close association. With the exception of the crocodyliform remains, none of the other micro-

40 vertebrates recovered from the Winton sites has turned up at Isisford (including lungfishes and turtles).

While many of the differences between the faunal and floral assemblages of the

Winton sites and Isisford may relate to taphonomy and depositional setting, it is also likely that the overall biotic composition of each area was distinct in its own right. This study indicates that vertebrate fossil-bearing sites in the two areas have distinctly different maximum depositional age profiles, with the Winton sites being middle at least mid

Cenomanian, but likely post-Cenomanian, and the Isisford sites yielding a very early

Cenomanian (100.5–102.2 Ma) maximum depositional age. Interestingly, the stratigraphic position of the fossil-bearing horizon at Isisford (as represented by the nodule sampled herein) appears to be well above the Winton-Mackunda boundary in this part of the

Eromanga Basin. Combined with an older age, this strongly suggests local variation in deposition across the basin, such that the age and depth of boundary between the Winton and Mackunda Formation may vary geographically. With more precise temporal and stratigraphic constraints, it should also be possible to more closely examine the variations in depositional history of the Winton Formation.

5.3 Recalibrating Palynomorph zones As far back as the 1950s, problems were recognized with the utilization of pollen as a temporal constraint, due to spores being highly susceptible to reworking in the same potential manner as the entombing host-rock. In ideal conditions, all constituents of a rock may be structurally preserved through reworking, but in reality pollen grains are commonly altered, damaged, or annihilated from the record (Muir, 1966). It is also now acknowledged that a greater stratigraphic disturbance of pollen is due to marine transgressive and

41 regressive cycles. With three of the four formations within the Rolling Downs Group represented by P. pannosus (Toolebuc, Allaru Mudstone, Mackunda Formation and the

Winton Formation), the first three of which span marine phases, many problems could arise. However when the detrital zircon results are coupled with palynological data, a more reliable interpretation can be formulated.

Historically, palynology and plant macrofossils have been utilized for bracketing a temporal constraint for mid-Jurassic to mid-Cretaceous sediments within the Eromanga

Basin strata (Hutton Sandstone to the Winton Formation) since the mid-sixties (Vine and

Day, 1965; Evans, 1966). In previous literature (Burger, 1990 and references therein, Sec.

2.2; and McLoughlin et al., 1995) the relative age of the Winton Formation relative to other units was been based on the occurrence of P. pannosus and A. distocarinatus, (Dettmann and Playford, 1969; Burger, 1973; Dettmann, 1973, 1994; Morgan, 1980; Burger, 1993).

However, Helby et al. (1987) placed the Winton Formation between the Albian and the

Cenomanian, noting the possibility that deposition could have extended into the early

Turonian. Recently, McLoughlin et al. (2010; pg. 4) noted that plant remains and palynological data have yet to obtain a definitive age determination of the upper Winton

Formation. Detrital zircons obtained just below the contact between the upper-most

Mackunda Formation date around 102 Ma to 103 Ma (latest Albian; see sec 3.5.5), thus supporting McLoughlin et al. (2010) age position of samples recovered from the base of the

Winton Formation. However Dettmann et al. (2009) described Lovellea wintonensis from surface deposits as late-Albian in age based on reports and site descriptions from Peters and

Christophel (1978; pg. 3119–3120) for Lovelle Downs, 48 km WNW of Winton

Queensland. An earlier site description of Lovelle Downs by the Peters and Christophel

(1978) study originally described the floral assemblage being entombed within an exposed 42 lenticular body containing silicified plant remains “just above” the supposed Winton

Formation but failed to described the body in detail (Peters and Christophel (1978; pg.

3119)). It is worth noting here that a plethora of plant remains occur within in a distinct horizon that occurs both upon exposed surfaces and also within lenticular bodies at the base of ‘jump ups’ (buttes). These common localities occur in the upper exposed and preserved

Winton Formation throughout the Winton area, including multiple sites just to the south of

Lovelle Downs at Bladensburg National Park and Lark Quarry Conservation Park. The stratigraphic position of surface and entombed remains is interpreted as portions of the lower–upper to upper sections of the preserved and exposed Winton Formation, which is in agreement with the observations noted by Peters and Christophel (1978, pg. 3120),

McLoughlin et al. (1995; pg. 274) and Pole and Douglas (1999, p. 542). These lenticular beds occur 1–3 m vertically below Lark Quarry's DZ samples described in this study, which are estimated at 92 (±1.2) Ma to 94.5 (±1.3) Ma in the early to mid-Turonian to very Late

Cenomanian, and just 2–4 m vertically above the dinosaur bone bed quarries at

Bladensburg National Park, which are estimated to be between 93.3 (±1.2) and 95.6 (±1.4)

Ma in the early Turonian to Late Cenomanian (if error is taken into account, then statistically both sites temporally overlap) (Figs. 1-5 and 1-6). Thus, adjustment of the spore–pollen zones originally described by Helby et al. (1987), Partridge (2006) and other authors is needed (Figs. 1-5 and 1-6). This ‘new’ temporal placement of the upper sections of the Winton Formation partly supports the original observations of Helby et al. (1987)

(Fig. 1-6).

Figure 2- 5 (right) Original and revised chronostratigraphy; Original chronostratigraphy based on Appendicisporites distocarinatus and Phimopollenites pannosus spore–pollen zones originally described by Helby et al. (1987), and (left) revised to be based on A. distocarinatus (Hoegisporis uniforma) and P. pannosus by Clifford and Dettmann (2005), Partridge (2006), Dettmann et al. (2009), Hocknull et al. (2009) and McLoughlin et al. (2010). Revised spore–pollen zonation was based on detrital zircon geochronological information interpreted within this study.

Figure 2- 6 Interpreted temporal relationships (based on the temporal relationship of the seven metrics (YSG, YPP, YDZ, YC1σ (+3), Weighted Average, YC2σ (+3), and TuffZirc (+6)) utilized with in this study, seen on the right) for detrital zircon samples derived from the three faunal assemblage localities in the upper Winton Formation (Lark Quarry Conservation Park, Bladensburg National Park, and Isisford, Queensland), along with the temporal placement of detrital samples derived from GSQ core logs (GSQ Eromanga 1, GSQ Longreach 1) and the interpreted position of the conformable contact between the underlying Mackunda Formation and the Winton Formation.

5.4 Implications for palaeobiogeography The rapid separation of Gondwana, principally during the Cretaceous, represented one of the most important palaeobiogeographic dispersal and subsequent vicariance events in Earth history. Archosaurian evolutionary lineages, post-separation of Gondwana, have been widely described as evolutionarily unique to their Laurasia counterparts (Bonaparte,

1986; Agnolin et al., 2010). However, the relationship of fossils from the eastern-most portion of Gondwana, in particular Australia, is still poorly understood due to a sparse and poorly dated fossil record. The many recent discoveries of new archosaurian taxa in the

Winton Formation are beginning to shed light on the radiation and dispersal of dinosaurian clades between South America and Australia via Antarctica. However, prior to this study had been undertaken taxa could not be reliably placed into a rigorous temporal framework.

The second temporal constraint used by many studies including Molnar (1980), Coombs and Molnar (1981), Dettmann et al. (1992), Molnar and Willis (1996), Molnar (2001),

Molnar and Salisbury (2005), Salisbury et al. (2006a,b), Hocknull et al. (2009), Fletcher and Salisbury (2010), and Molnar (2010) has been an association with the P. pannosus palynomorph zone (Suite III; Burger, 1990 and therein). However, problems arise with the utilization of this method when it is not used in association with other temporal constraining methods. P. pannosus spans four stratigraphic units (the Toolebuc Formation, the Allaru Mudstone, the Mackunda Formation, and the Winton Formation), over a time span of about 15 My.

With each new discovery, along with the re-description of previously-identified taxa, it is becoming increasingly apparent that Australian taxa have strong evolutionary ties to similarly aged forms from South America. Agnolin et al. (2010) revised much of the

45 non-avian dinosaur taxa from the Winton Formation, and found that the majority had affinities to other Gondwanan taxa, particularly from South America, rather than Laurasia.

With the implementation of the chronostratigraphic framework generated by this study, for the first time, paleontologists can better place Australia's early Late Cretaceous biota

(including recently discovered undescribed taxa) into a meaningful global context (Sereno,

1997, page 474; Sereno et al., 2004). By continuing to place each fossil horizon into a temporal framework based on geochronology we can better compare Australian taxa with coeval forms from other parts of the world (Upchurch et al., 2007; Upchurch, 2008).

6. Conclusions To accurately contextualize the Winton Formation's terrestrial Cretaceous vertebrate biota, improved temporal and stratigraphic resolution is paramount. We attempted to address this problem by testing the potential of utilizing U–Pb LA-ICP-MS detrital zircon geochronology to refine the age of the formation by systematically investigating the youngest maximum depositional age of zircons at different stratigraphic levels and locations throughout the unit. We analyzed five samples (six populations; see sec 4.6) and compared different approaches (seven) to interpret the youngest maximum depositional age in order to both refine the age of the formation and to improve regional stratigraphic relationships between widely separated fossil assemblages. Of the seven approaches (YSG,

YPP, YDZ, YC1σ (+3), Weighted Average, YC2σ (+3), and TuffZirc (+6)), five metric's are recommended for deriving the youngest maximum depositional age (YSG (when coupled with other metrics), YDZ, YC1σ (+3), Weighted Average, YC2σ (+3)). Our results demonstrate that the formation was largely deposited during or shortly after a long period of intense volcanic activity along the eastern margin of Australia, with erosion of the

46 resulting topography being the main source of sediment. We identified multiple temporally arrayed Cretaceous volcanic sources that fed the formation, and that the youngest of these populations was likely deposited syndepositionally, providing greatly improved age constraints on deposition. Using various metrics for interpreting maximum depositional age, we find that the youngest samples from Lark Quarry Conservation Park Bladensburg

National Park and drill-core GSQ Eromanga 1 ranged between 92.5 (±1.2) and 93.3 (±1.2)

Ma, and indicate that deposition of the Winton Formation and associated fossil material occurred within the early-Turonian to Turonian–Cenomanian boundary (Fig. 1-6). As a result, the previously interpreted earliest known flowering plants occurred at or younger than the early-Turonian to Turonian–Cenomanian boundary rather than previously indicated. Fossil material collected from the uppermost underlying Mackunda Formation is interpreted to be around 102.5 to 104 Ma.

Chapter 3

U-Pb and Lu-Hf detrital zircon provenance analysis of the Upper Cretaceous Winton Formation, Eromanga Basin, Queensland: implications for the tectonic evolution of eastern Australia

ABSTRACT In order to address a variety of questions that have recently been posed around the Winton Formation and the tectonics of eastern Australia, 10 detrital zircon samples were collected from various stratigraphic levels and locations across the basin from both the Winton Formation and the underlying Mackunda Formation. U-Pb LA-ICPMS geochronology was conducted on 719 individual detrital zircons, which revealed dominant population clusters between 200 -92 Ma. Very minor populations of Archean, Paleoproterozoic and Mesoproterozoic grains were also observed and interpreted to be recycled into the Winton Formation from younger sedimentary units. These grains were likely originally derived from western Australian cratons. A number of minor Neoproterozoic to Paleozoic grain populations (ranging from 1000-330Ma) were also identified and are interpreted to be from nearby sources in northeastern Queensland including but not limited to the Macrossan, Anakie, Cape River, Greenvale and Georgetown Provinces. Significant Late Paleozoic to Triassic detrital zircon populations (330-200Ma) are recorded in this study and these are linked to well-documented local sources in the New England and Kennedy Provinces. One of the most significant findings from this study is the documentation of near-continuous volcanic activity from Triassic through ~92 Ma. The youngest grain ages in this range are interpreted as syndepositional. Grains of this age, particularly Cretaceous, the most voluminous observed in this study and are interpreted to be derived from volcanic arc-derived sediments that were transported westward into the basin from the eastern margin of the continent. This indicates significant and previously undocumented evidence of continuous arc magmatism along the eastern margin of Australia during the Jurassic and Cretaceous. The Whitsundays Volcanic Province has previously been suggested as a source of sediment for the Winton Formation, and these results support that. The Jurassic, and Cretaceous detrital zircon populations, along with other selected grain populations were further investigated through Lu-Hf isotope analysis. The Jurassic-Cretaceous zircons are derived from juvenile mixed crustal and shallow mantle sources, consistent a typical recycled arc provenance, rather than a Large Igneous Province, which has recently been suggested. On the other hand, several historical and recent studies have suggested continuous slab subduction of the Phoenix/Pacific Plate under the eastward migrating Gondwana margin has been proposed as the mechanism for the long-lived volcanism that apparently characterized this margin. This model would have provided a significant volume of volcanic-derived sediment that characterizes the Winton Formation, with much of this material being transported westward into the basin syndepositionally or near-sydepositionally by large river systems.

1. Introduction Clastic sediments are important archives of past climate, tectonics and palaeogeography. Zircon is an ideal detrital mineral for provenance studies because of its durability, both structurally and chemically, and its abundance in a wide array of rock types

(Köppen and Carter, 2000; Andersen, 2005, Roberts et al., 2012). Advances in detrital zircon geochronology over the last decade have greatly expanded research potential, and in many ways, led to a renaissance in sedimentary petrology (Dickinson, 1974; Cawood et al.,

2007; Gehrels, 2011). This is due to numerous advances in U-Pb geochronology that have made it possible to rapidly and economically date and analyze large populations of detrital minerals, in particular zircon (e.g., Fedo et al., 2003; Surpless et al., 2006; Gehrels, 2008;

Dickinson & Gehrels, 2009; Frei & Gerdes, 2009; Gehrels, 2011).

The application of detrital zircons recently played a significant role in helping to unravel the late Mesozoic tectonic evolution of New Zealand, New Caledonia, eastern

Australia and the South Pacific islands (Adams et al., 1998; Cluzel et al., 2011, 2012;

Bryan et al., 2012, Seton et al., 2012). In particular, the eastern Australia margin has been the subject of considerable debate with multiple hypotheses arising for the timing and mode of volcanism and plate boundary interaction (Ewart et al., 1992; Adams et al., 1998;

Veevers et al., 2000; Cluzel et al., 2011, 2012; Bryan et al., 2012, Seton et al., 2012).

Although an active volcanic system was located on the eastern margin of the Australian

Craton during the middle Cretaceous, questions still remain concerning the mode and origin of volcanism at that time. In particular, it has been debated whether volcanism during this period was related to either long-lived subduction related arc volcanism or to extension and rift related Siliceous-Large Igneous Province (S-LIP) volcanism (~130-95 Ma) (Veevers et al., 1997; Bryan et al., 1997; Veevers, 2000 a,b,c; Bryan et al., 2012). 50

In a recent study, Bryan et al. (2000, 2012) investigated the fragmentary remnants of the Cretaceous volcanic bodies in the Whitsunday Volcanic Province (WVP), which extend from Maryborough, Queensland to just north of Mackay, Queensland. Their study found evidence of both intrusive and extrusive igneous rocks in the WVP dominantd by rhyolitic and dacitic compositions. Based on geochemistry and isotopic dating, they interpreted the WVP as evidence of a previously unrecognized episode of S-LIP volcanism that predated later Cretaceous rifting (~85 Ma), which led to the northern unzipping of the

Tasman Sea Basin (Coral Sea). Bryan et al. (2012) also concluded that the fragmentary remnants of the WVP represent only a small portion of a much larger original volcanic region. Their study also used sedimentary provenance data from adjacent westerly lying basins to constrain the timing and cessation of volcanism, which they concluded was between ~130-95 Ma (Bryan et al., 2012).

In contrast to the S-LIP model, the long held tectonic theory for the northeastern

Australian Margin is a long-lived subduction margin and continuously associated arc magmatism (Kemp, et al., 2009; Li et al., 2014). Much of eastern Australia is composed of the Tasmanides, an easterly younging series of orogenic terranes (Li et al., 2014). The

Tasmanides are estimated to originate in the Neoproterozoic and preserve a tectonic record which extends into the Triassic (Glen, 2005). This accreted series of terranes dominantly records continuous convergent-transform tectonism coupled with episodic subduction zone retreat along the developing eastern Gondwana margin until at or around the Late Triassic

(Fergusson, 1999; Glen, 2005; Gray et al., 2006; Kemp et al., 2009). Extensive isotopic evidence also reliably correlates these pre-Mesozoic terranes to distinct Australian terranes or established tectonic events post-dating the unzipping of the Tasman Line (Luyendyk,

1995; Veevers, 2000a,b; Collot et al., 2009; Waschbusch et al., 2009; Champion &

Bultitude, 2013).

Figure 3- 1a Map of Queensland, northeastern Australia displaying surface exposures of the Winton Formation within the Eromanga Basin. Symbols indicate detrital zircon sample locations derived from at or near recently discovered and historical fossil localities (Bladensburg National Park, Lark Quarry Conservation Park, and Isisford, Queensland) along with detrital samples derived from Geological Survey of Queensland stratigraphic core logs (GSQ Eromanga 1, GSQ Longreach 1-1B, GSQ Blackall 2, GSQ McKinlay1, and GSQ Maneroo 1). Figure 3-1b. Stratigraphic cross-section A-A’ includes stratigraphic position of all detrital zircon samples derived from the Eromanga Basin for both the Mackunda and Winton Formation. Australian image modified from GeoScience Australia 2013.

Although many workers have hypothesized continuous arc volcanism and subduction along eastern Australia into the Late Cretaceous, the lack of well-preserved arc rocks of this age severely limit the testability of this idea or the more recent S-LIP model.

Because very little of the Mesozoic volcanic Arc or rift-related volcanics remains in eastern

Queensland as a result of rifting of the Coral Sea, sedimentary provenance analysis may provide the best opportunity to understand the late Mesozoic tectono-volcanic history of this part of Australia. The uppermost strata preserved in the Eromanga Basin are mid- to

Late- Cretaceous in age and represent the best starting point for investigating this question.

In particular, the Winton Formation, a unit best known for its prolific terrestrial dinosaur fauna, has been indicated to have strongly volcanolithic petrofacies (Grey, 2002, and references there in). However, relatively little sedimentological or provenance analysis of the Winton Formation has actually been conducted to demonstrate this or its potential for revealing information about late Mesozoic volcanic source areas of Eastern Australia.

Bryan et al. (1997) specifically hypothesized a source in the Whitsunday Volcanic Province to account for the putatively volcanic-rich sediments of the Winton Formation.

Additionally, only a handful of studies (with very little agreement) have tried to extrapolate sediment input and drainage patterns within the Eromanga Basin during the middle

Cretaceous. Most recently, Veevers (2000), King and Mee (2004), and MacDonald et al.

(2013), dispute a link between Cretaceous-age detrital zircons observed in Cretaceous-

Eocene age deposits of the Ceduna Delta in South Australia and erosion of volcanic detritus from the Whitsundays Volcanic Province or the Eastern Australian Highlands. However, this and other assertions must first be tested through careful provenance analysis of the

Winton Formation.

Hence, the goal of this project is to comprehensively investigate sedimentary provenance of the mid-Cretaceous Winton and Mackunda formations within the Eromanga

Basin via U-Pb detrital zircon geochronology and Lu-Hf isotopes (Fig. 3-1). By analyzing ten detrital zircon samples (~850 U-Pb ages & 83 Lu-Hf analyses) from key stratigraphic levels within the Winton and underlying Mackunda formations, we are able to interpret the major sources of sediment for these units, and reconstruct middle Cretaceous continental drainage patterns. For instance, these data are crucial to testing the recent hypothesis that

Upper Cretaceous sediments of the Ceduna Delta System are genetically linked to the

Winton Formation and ultimately the Whitsunday Volcanic Province. Most importantly, these results shed new light on the tectono-magmatic history of eastern Australia over the latter half of the Mesozoic, and also aid in reconstruction of the palaeogeography and palaeoenvironments associated with one of Australia’s most import dinosaur faunas.

2. Geological background

2.1 Eromanga Basin The expansive Eromanga Basin is one of the largest sedimentary basins in Australia

(Fig. 3-1), covering much of central and southern Queensland, easternmost Northern

Territory, and the very northern portions of South Australia and New South Wales. The

Eromanga Basin also forms the central portion of the Great Artesian Basin, and according to Draper (2002b), the Eromanga Basin developed above portions of the earlier Cooper,

Galilee and, Georgina Basins, and the Mt. Isa Province and the Maneroo Platform (Draper,

2002b; Cook et al., 2013). The Eromanga Basin was originally interpreted as intracratonic sag (Grey, 2002), however other workers have interpreted it to be a foreland basin

(Middleton, 1989; Gallagher, 1990; Cartwright & Dewhurst, 1998; Draper, 2002c; Cook et

54 al., 2013). At present there seems to be little consensus on this debate (e.g., see Fielding

1996; Isern et al., 2002, Bryan et al. 2012).

Regardless of its tectonic origins, the Eromanga Basin is widely considered to have initiated during the late Triassic, with the Triassic Cuddapan Formation representing the earliest depositional unit in the basin (Draper, 2002). However, deposition during the earliest phase of basin development was restricted to low-lying depressions at the southern end of the Bowen Syncline (Whitehouse, 1954). Subsequent Jurassic sediments were deposited in dominantly terrestrial –fluvial environments, which are interbedded with argillaceous rocks (Senior et al., 1978; Exon and Burger, 1981; Draper, 2002). The Lower

Jurassic succession is limited to the interior of the basin, whereas the Upper Jurassic units extend beyond the outer fringes of the Eromanga Basin (Gallagher & Lambeck, 1989).

During the Latest Jurassic to Earliest Cretaceous, the Eromanga Basin was slowly inundated by a marine flooding that culminated in an expansive interior seaway across much of the basin and existed for most of the Early Cretaceous. This event is recorded by the transition from basal fluvial and floodplain sandstones and siltstones of the Adori

Sandstone (late Oxfordian) to the Westbourne Formation (late Oxfordian – earliest

Tithonian) in the first recorded marine transgression. The overlying Westbourne Formation to Cadna-Owie Formation (earliest Tithonian-earliest Barremian) preserves the regression stage of the interior seaway, typified by near-shore to terrestrial fluvial and lacustrine environments (Senior et al., 1978; Gray et al., 2002). The overlying Wallumbilla and

Toolebuc formations, the Allaru Mudstone and the Mackunda Formation (earliest

Barremian-latest Albian-earliest Cenomanian) preserve sediments that are characteristic of a long-lived shallow-marine seaway (Exon and Senior, 1976).

2.2 The Winton Formation The Winton Formation is poorly but extensively exposed over large portions of

Queensland, as well as in a number of localities in the northern New South Wales, north- western South Australia and the south-western corner of the Northern Territory (Fig. 3-1a)

The Winton Formation was originally described by Dunstan (1916) as the “Winton Series”, consisting of sandstones, shales and minor coal seams that overlie the predominantly marine sediments of the Mackunda Formation. It was not until Whitehouse’s (1954) construction of the stratigraphic nomenclature of the Eromanga Basin that the Winton

Formation was formally defined. Vine and Day (1965) and Vine et al., (1967) redefined the Cretaceous stratigraphy in the basin as follows: Wilgunya Group at the base

(Wallumbilla, Toolebuc and Allaru formations), overlain by the Manuka Subgroup

(Mackunda and Winton formations). Vine et al. (1967) noted that the original divisions of the Roma and Tambo Series, as defined by Whitehouse (1955), were invalid stratigraphic rock units (valid only as biostratigraphical units), whereas the Winton Formation, first described as an interbedded freshwater sequence of mudstones, sandstones, and minor coal seams was supported as a mappable unit and considered a valid formation. More recently the Winton Formation was placed within the Rolling Downs Group (upper Manuka

Subgroup) and additional sedimentologic investigations refined the description of the unit as consisting of complex repetitive sequences of fine to medium-grained feldspatholithic or lithofeldspathic arenite, siltstones, mudstones and claystones with minor thinly bedded mud-clast conglomerates and rare limestones (Casey, 1970; Exon and Senior, 1976; Senior et al., 1978; Senior and Mabbutt, 1979; Coote, 1987; Gray et al., 2002).

2.3 Study area and sample locations Ten detrital zircon samples were collected from across the Winton Formation outcrop area and from Geoscience Queensland (GSQ) sedimentary cores drilled in central

Queensland (Fig 3-1a,b). Four of the samples were acquired at, or near, key vertebrate fossil localities within the Winton Formation (Isisford, Hades Hill, Eureka, and

Bladensburg). The remaining six core samples were chosen (GSQ Eromanga, Longreach,

Maneroo (2), McKinlay, and Blackall), based on stratigraphy and location, from the

Geological Survey Queensland (GSQ) core shed (Fig. 3-1a,b). Stratigraphically, samples were collected from the top of the Mackunda Formation, ~10 m below the contact with the base of the Winton Formation to the upper-most levels of the Winton Formation exposed in outcrop in western Queensland (Lark Quarry Conservation Park, and Bladensburg National

Park) (Fig. 3-1b). Outcrop samples collected from Lark Quarry Conservation Park include

Eureka (Lat: 23°00’58.62”S; Long: 142°24’42.73”E) and Hades Hill (Lat: 23°00’44.96”S;

Long: 142°24’44.95”E). Samples derived from Eureka were taken from ~1.0 m above the horizon that preserves the famous Lark Quarry dinosaur tracksite (Thulborn and Wade,

1979, 1984, 1989; Romilio et al., 2013). Eureka is stratigraphically the highest sample collected from the Winton Formation. The second sample from Lark Quarry Conservation

Park; Hades Hill was collected from several meters below the track horizon and is stratigraphically lower in section that the Eureka sample. This sample was located at the base of the local ‘jump-ups’ at a locality, which preserves abundant plant and vertebrate fossil remains. The third Winton outcrop sample comes from Bladensburg National Park

(Lat: 22°30’53.35”S; Long: 143°02’19.89”E), ~2–3 m above the main fossil-bearing horizon in the park, slightly below the level of the Lark Quarry trackway horizon. The fourth outcrop sample was collected near Isisford, Queensland (Lat: 24°15’33.51”S; Long:

144°26’29.35”E). This sample was taken directly from a large carbonate-cemented sandstone concretion containing a new, undescribed dinosaur. This sample site is also in close proximity to other key fossil-bearing localities in the area (see Salisbury et al.,

2006a).

Six additional detrital zircon samples were derived from GSQ cores, including

Eromanga 1, McKinlay 1, Blackall 2, Longreach 1-1B, and Maneroo 1 (Fig 3-1a,b). GSQ

Eromanga 1 core was drilled east of Eromanga, Queensland (Lat: 26°36'52.93"S; Long:

143°52'48.41"E) and this sample (ERO1-11-001) was collected from the core interval between 16.5-26.7 meters (boxes 2-5) below surface (roughly 146.0 meters above base of the Winton Formation). Two samples were taken from GSQ McKinlay 1, drilled near

McKinlay, Qld (Lat: 21°35'33.00"S; Long: 142°15'35.00"E). The uppermost sample

(McKinlay_02) was collected from near the top of the core at a depth of 9.16-9.54m (Box

2), which corresponds to the upper Winton Formation. The lowermost sample

(McKinlay_16) was collected at a depth of 53.82-54.30m, and is identified as a middle

Winton Formation sample.

A samples was collected from GSQ Blackall 2 core (Lat: 24°09'45.33"S; Long:

144°13'25.00"E), and was collected because of its proximity to the Isisford fossil site. This sample was collected from a depth of 20.8-35.8m (Box 1-6). Another sample was collected from GSQ Longreach 1-1B core (sample LO1-11-001), which was drilled northeast of Longreach, Queensland (Lat: 23°06'42.93"S; Long: 144°35'58.41"E). This sample was collected from the interval between 20.8-35.8 meters below surface (boxes 1-

6), and represents the uppermost portion of the Mackunda Formation (Fig. 3-1). The last of the sampled cores was collected from GSQ Maneroo 1, which was drilled northeast of

Maneroo, Queensland (Lat: 23°23’09.10"S; Long: 144°28'04.41"E). The Maneroo sample was collected from box 331, between 342.3-343.3 m below surface, and is the stratigraphically lowest sample analyzed in this study, from the middle Mackunda

Formation.

3. Methods

3.1 Field and core sampling Detailed sedimentologic and stratigraphic analysis of the Winton and Mackunda formations conducted as part of this study. In particular, outcrop investigations were focused on Bladensburg National Park and Lark Quarry Conservation Park. To get a more holistic understanding of the Winton Formation, cores were also investigated and sampled as part of this study and detailed descriptions on each of the GSQ cores were performed.

Decimeter-scale, measured sections were constructed at the four outcrop sample locations using a Jacob’s Staff and Brunton compass.

To ensure consistent sampling, detailed facies analysis was also performed at each sample area to characterize the depositional environments of the Winton and reconstruct regional palaeogeography of the Winton depositional system (Tucker et al., 2014). Bulk samples weighing 5-10 Kg each were collected from each field locality from freshly exposed, unweathered fine- to medium-grained sandstones. In addition, palaeocurrent measurements on fluvial sandstones were collected at Bladensburg National Park and Lark

Quarry where possible; however intense surface weathering has greatly limited the potential for identifying and recording accurate palaeocurrent measurements.

Sampling of the six core samples was conducted at the Geoscience Queensland core facility in Zillmere, Queensland. This was performed as part of a larger core logging study of the Winton Formation focused on geochronology and reconstruction of palaeoenvironments (Tucker et al., 2013, 2014). Because of the precious nature of these cores, all core samples were chosen with care and collected from half-cores collected over several meters from each box, with each sample totaling ~5kg of fine- to medium-grained sandstone from dominantly fluvial or shallow marine sandstone facies. Detailed core logs were constructed and photographed through each of the sampled intervals.

3.2 Detrital zircon separation All samples were crushed and milled in a tungsten carbide disc mill. Samples were washed to remove the clay-sized fraction, but all other grains below <500 µm were utilized to enhance the possibility of sampling potential distal volcanic ash derived zircons. Heavy minerals were separated using lithium Polytungstate adjusted to a specific gravity of 2.85-

2.87. Mineral separates were then washed, dried and a hand magnet was used to remove strongly magnetic minerals, followed by a Frantz magnetic separator at progressively higher magnetic currents of 0.5, 0.8, 1.3, and 1.5 set at a constant 10° side slope. Non- magnetic heavy fractions (>1.3) were then selected via a hybrid selection protocol that involved handpicking zircons as randomly as possible from the greater population within a defined field of view. Zircon populations derived from fossil localities at LQCP, BNP, and

Isisford Queensland, along with GSQ Eromanga 1 and GSQ Longreach 1-1B were utilized in a prior study that focused on refining the age of the Winton Formation through the systematic identification of the maximum depositional age at each sample locality (Tucker et al., 2013). Provenance data from these samples are investigated here for the first time, in combination with new data from the additional samples Hades Hill. For each sample,

~100-150 grains were mounted in a 25 mm epoxy resin puck, polished to expose their mid- sections and imaged using a Jeol JSM5410LV scanning electron microscope with attached cathodoluminescence detector in order to document cores, microstructures, cracks, inclusions and other complexities.

3.3 U-Pb Analysis All work was done at the Advanced Analytical Centre of James Cook University using a

Coherent GeolasPro 193nm ArF Excimer laser ablation system connected to a Bruker 820-

MS(formerly Varian 820-MS). Instrumental tuning, sample protocal and set up followed the methodologies outlined in Tucker et al. (2013). All analyses were performed using a 5

Hz repetition rate, 44 µm beam aperture and 6 J/cm2 energy density, as determined by energy meter at the ablation site. Under these conditions, the ablation rate for NIST 612 and zircon was about 0.1 µm per laser pulse and 0.06 µm per laser pulse respectively. For all sample analyses (which averaged 100 grains per sample) the total measurement time was set to 65 seconds. Standard bracketing was used throughout the study to correct for remaining elemental fractionation and mass bias (Tucker et al., 2013).

A 32 µm beam diameter was used for the following samples: Eureka, Hades Hill,

Eromanga 1, McKinlay 1, Blackall 2, Longreach 1-1B, and Maneroo 1 (2). However, due to the small grain size of samples from Bladensburg and Isisford, a 24 µm beam diameter

(spot size) had to be used for these samples. The primary zircon standard used in all analyses was GJ-1 (609 Ma, Jackson et al., 2004) and the secondary zircon standard used was Temora- 2 (TEM-2; 416.8 Ma, Black et al., 2003). After NIST glass and standards were used to tune the instrument at the beginning of each analytical session, groups of 10–

12 unknown zircons were analyzed, followed by at least two analyses of the primary standard and 2 analyses of the secondary standard. During the course of this study, all

61 standard analyses fell within 2% of the expected ages and most were within 1% of the expected age. NIST 612 were analyzed at the beginning and end of each session and at least once between for the purpose of calibrating Th and U concentrations (Tables 1a-e).

Table 3- 1 (A-E) Detrital zircon via LA-ICPMS results for all samples including isotope and concentration data: Th (in PPm) is Thorium concentrations measured in parts per million; U (in PPM) is Uranium concentrations in parts per million; Th/U is the divided ratio of Uranium concentrations; Both Pb206/U235 and error calculated at ±1σ and Pb206/U238 and error calculated at ±1σ were used to calculate an Error Correction Value (Error Corr.). The ratio values for Pb206/U238, Pb207/Pb206, and Pb207/ Pb206 with all error calculated at ±1σ with the preferred age (Pb206/U238) with error ±1σ interpreted as YSG. Age is calculated by error propagation and included errors and U and U/Th Concentration are calculated against NIST 612 and are within ~30% error. 67

3.4 Lu-Hf Analysis Following U-Pb analysis on each of the ten detrital zircon samples, a subset of grains representing each of primary age peaks selected for Lu-Hf isotope analysis. This work was conducted via Thermo Scientific Neptune MC-ICPMS and GeoLas Pro 193nm

ArF laser system in the Advanced Analytical Centre, James Cook University. Analytical protocols and methods follow those established by for the lab by Kemp et al. (2009) and

Næraa et al. (2012). A total of 108 detrital zircons selected from ten primary detrital populations based U-Pb age concordance (<20%) and age.

The largest beam size possible was used in these analyses, which was typically 32

μm or 42 μm spots, and on rare occasions, a 60 μm. A 4 Hz laser pulse repetition rate over a 60 second ablation period applied at a constant energy level at 5-6 J/cm2. A He carrier gas was used during ablation, which was combined with Ar and a small (~0.005 l/min) N2 flow prior to transport into the ICP-MS. Ablation was conducted as close as practical to the U-

Pb spots. The isobaric interference of Lu and Yb on 176Hf was corrected by monitoring the interference-free 171Yb and 175Lu intensities during the analysis and then deriving 176Yb and

176Lu using 176Yb/171Yb = 0.897145 (Segal et al., 2003) and 176Lu/175Lu = 0.02655

(Vervoort et al., 2004). Yb isotope ratios were normalized to 173Yb/171Yb = 1.130172

(Segal et al. 2003) and Hf isotope ratios to 179Hf/177Hf = 0.7325 using an exponential law.

The mass bias of Lu was assumed to follow that of Yb. Data was processed offline using a customized Microsoft Excel spreadsheet; within-run outlier rejection was set at 3 standard errors of the mean. Reference zircons analyzed concurrently with the unknowns yielded mean 176Hf/177Hf values (± 2 sd) of 0.282185 ± 24 (n = 5) for FC1 and 0.281611 ± 16 for

QGNG (n = 11), identical to the solution 176Hf/177Hf values of 0.282184 ± 16 and 0.281612

± 4, respectively, reported by Woodhead and Hergt (2005). All Hf isotope data were 68 normalized to the solution 176Hf/177Hf value of Mud Tank zircon (0.282507 ± 6, Woodhead and Hergt, 2005, reported relative to JMC 475 176Hf/177Hf = 0.282160) using laser ablation data generated from this zircon during the same analytical session (0.282497 ± 15, n = 7, 2

SD). Quoted Hf uncertainties in sample zircons combine within-run analytical errors plus the reproducibility of Mud Tank zircon.

3.5 Kolmogorov-Smirnov Test The K-S test is a non-parametric method for comparing cumulative probability distributions (DeGraaff-Surpless et al., 2003; Barbeau et al., 2009; Guynn & Gehrels,

2010). This method is a very practical comparative tool due to the lack of assumptions of data distributions (Amidon et al., 2005).The K-S test essentially determines, at the 95% confidence level (2 sigma error), if two populations cannot be derived from the same source. For P values less than 0.05, the two sets of grain-age data could not have been derived from the same original source, or population of ages. However, if the P value is greater than 5%, then the differences between the two samples are simply due to random chance and they are considered to have the same source. Additionally, the K-S test calculates the maximum probability distance between the Cumulative Distribution

Functions (CDF’s) of two sets of age distributions (Barbeau et al., 2009). Thus, by setting a chosen confidence level, the compared sets of data will either pass or fail. This study utilizes CDF’s based on ages and uncertainties, and excludes those that fail this test. Even though a 95% confidence interval was applied, all results are reported.

3.6 Potential sources of bias in detrital zircon studies A number of recent studies have highlighted the potential for error and bias in detrital zircon studies. Most of these studies focus on, but are not limited to, hydrological sorting, transport variability and climate effects on deposition (Hietpas et al., 2011; 69

Lawrence et al., 2011). Biases which directly affect the interpretation of provenance source, age, and sampling errors have been addressed by various workers, including

Moecher and Samson (2006) and Cawood et al. (1999, 2007b). Similarly, all possible precaution was taken during this study to eliminate potential biases associated with this detrital zircons study. We recognize that there are inherent biases associated with a study of this kind and attempted to minimize possible pitfalls through careful sampling strategies,

CL imaging and grain selection.

4. Results

4.1 Winton & Mackunda Formation Temporal Placement By coupling detrital zircon data from Tucker et al. (2013) with the newly acquired samples this study can more confidently place constraints on the deposition of the Winton and Mackunda formations. The Middle Mackunda Formation (Maneroo_331) contains grains that are younger than 128 ±1 Ma. Thus, the maximum depositional age for the

Middle Mackunda is older than 134 ± 1 Ma to 128 ±1 Ma. Tucker et al. (2013) interpreted the youngest maximum depositional age for the uppermost Mackunda Formation

(Longreach 1-1B) to be no older than 102 ±1 to 104 ±1 Ma and the lowermost Winton

Formation (Isisford, Qld) no older than 100 ±1 to 102 ±1 Ma. Furthermore, Tucker et al.

(2013) interpreted the Upper Winton Formation to be no older than that of 92-94 Ma based on samples from Bladensburg National Park, Lark Quarry Conservation Park (Eureka), and

Eromanga_1. Additional results reported here further support these interpretations, and indeed suggest that the top of the formation may be even younger than suggested. Three new Upper Winton Formation samples (McKinlay 1, Lark Quarry Conservation Park

(Hades Hill), and Blackall 2) all contain similar maximum depositional age-constraining

70 populations between 92 and 94 Ma. However, a stratigraphically lower sample from the middle Winton Formation sample (McKinlay 16) was also dated and found to preserve similar age grains between 92-94 Ma.

4.2 Detrital Zircon Populations 719 grains were dated from the ten different samples collected from the Winton and

Mackunda formations. General observations are first discussed based on the distribution of grains (n=719 grains) by age (Tables 3-1). The most prominent population observed is the high proportion of Mesozoic (particularly Cretaceous) grains, moderate to minor populations of Paleozoic grains and the very low proportion of Precambrian grains. About

20% of grains are Precambrian: Neoproterozoic populations (10.6%), Mesoproterozoic

(5%), Paleoproterozoic (1.3%), and Archean (1.9 %). Paleozoic age grains also contribute

~20% of the total proportion of grains analyzed. However, Mesozoic grains clearly dominant the sedimentary provenance, representing over 60% of all grains analyzed. A full review of the detrital zircon provenance patterns observed in this study is presented below by sample.

Figure 3-2 Plausable source terranes for detrital zircons recovered from the Winton and Mackunda formations. (Modified from Jell, 2013).

Figure 3- 3 Youngest graphical (histogram based) detrital zircon age peak (YPP) results for all individual samples along an age-probability plot or age distribution curve presented and calculated within ISOPLOT (Ludwig, 2009).

When all samples are collectively reviewed, very minor populations of Archean grains (middle Paleoarchean-Neoarchean) are recorded. The oldest Paleoarchean (3.3-3.2

GA) grain clusters present are composed of 2-4 grains each. Similarly, the Mesoarchean

(3.2-2.8 Ga) is represented by three clusters (n<3 each), and the Neoarchean (2.8-2.5 Ga) is represented by only two distinctive clusters (n<4 each). Currently, there are no known

Archean terranes in northeastern Australia, indicating these grains to be not locally derived.

Thus, it is more likely that these detrital grains are from western and southwestern terranes.

Plausible terranes include the Gawler Craton, along with the Musgrave, Arunta, and

Tennant Creek inliers (Fig. 3-2A). The Paleoproterozoic (2.5-1.6 Ga) is represented by eight clusters with many minor populations composed of two to three grains; however there is a population around 1.7 Ga represented by five grains. The Mesoproterozoic (1.6-1.0

Ga) is also represented by minor populations represented by five to eight grains each, with a significant population of 18 zircons around 1.0 GA. Contrary to these minor clusters in the early and middle Proterozoic, the Neoproterozoic contains roughly five distinctive age populations containing from eight to 31 grains each. The number of grains in each of these populations is consistently more as the populations get younger (Fig. 3-2b,c,d).

Regionally there are several potential sources for the Proterozoic grains, including the Musgrave Province, the Georgetown and Mt. Isa inliers of Queensland, Australia

(Blake & Stewart, 1992; Murgulov et al., 2007; Foster & Austin, 2008; Withnall & Hutton,

2013). Other plausible local source terranes include Greenvale and the Cape River

Metamorphics, which lie within the Charters Towers Province of Queensland (Fig. 3-2c,d)

(Meyers et al., 1996; Draper, 1997; Fergusson et al., 2005; Fergusson et al., 2007b).

Paleozoic zircons (n=145) are significantly more abundant than Archean or

Proterozoic grains (Fig 3-3). Generally, grain population density increases from the

Paleozoic to Permian. When all samples are collectively reviewed in terms of all grains ablated, Paleozoic clusters fluctuate. Cambrian and Ordovician (541-443 Ma) grains form six distinct populations grains, and generally populations decrease into the Ordovician.

The Siliurian to Devonian (443-358 Ma) is represented by eight distinct clusters which fluctuate (3Silurian and a positive general increase into the Late Devonian. During this period (Late Cambrian into the Devonian),

Eastern Australia was dominantd by the Tasman Fold Belt System (Scheibner & Veevers,

2000; Fergusson & Henderson, 2013). Plausible local source terranes in Queensland include the Anakie province, and both the Macrossan and Pama Igneous Associations (Fig.

3-2e) (Murray & Kirkegaard 1978; Day et al., 1983; Withnall et al., 1996; Hutton et al.,

1997; Green et al., 1998; Fergusson & Henderson, 2013).

The number of distinct populations and the amount of grains within each distinct population both increase between the Middle to Late Paleozoic (Carboniferous-Permian).

Carboniferous (358-298 Ma) grains form five distinctive peaks (4

3-2e) (Scheibner & Veevers, 2000; Champion and Bultitude, 2013; Donchak et al., 2013;

Fergusson & Henderson, 2013). This collection of volcanic terranes was the result of a westward subduction zone between the Gondwana/Australian margin and the subducting

Phoenix-Pacific Plate which continued the Late Devonian to Permian (380-252 Ma)

(Murray et al., 1987; Little et al., 1992; Withnall et al., 1997; Betts et al., 2002; Champion

& Bultitude, 2013).

Mesozoic grain populations contain the majority of all grain ages obtained by this study (436/719 grains, Fig. 3-3). Generally, Cretaceous grains (n=286) are more abundant than those in the Jurassic (n=75) or Triassic (n=75). Within the Triassic, seven distinct clusters are readily identified (5

Cretaceous grains (up to 92 Ma) represent the largest of cluster of grain ages obtained in this study (286/719 grains). Cretaceous grains are represented by eight distinct clusters with peaks between 110-100 Ma (85 grains and 83 grains respectively). Whitsunday

Volcanic Province has previously been postulated as the source for Cretaceous grains in the

Winton Formation (Bryan et al., 2012) (Fig. 3-2f 7 3-3).

4.2.1 Mackunda Formation Grain Ages The Middle Mackunda Formation (Maneroo_331, 89 grains) contains 15 clusters

(Longreach 1-1B, 69 grains), pre-Ordovician clusters are very minor (12 clusters with 2-4

76 grains in each) comparatively to the Middle Mackunda Formation. A distinct peak does not occur until ~490Ma (1 cluster with 6 grains).

Figure 3- 4 Youngest graphical (histogram based) detrital zircon age peak (YPP) results based on stratigraphic placement within the Winton of underlying Mackunda Formation, along an age-probability plot or age distribution curve presented and calculated within ISOPLOT (Dickinson and Gehrels, 2009; Ludwig, 2009).

These minor populations are more likely regionally derived from eastern to northeastern source terranes rather than multimodal sources. The upper Mackunda

Formation contains the first occurrence of middle Cretaceous grain clusters (102-110 Ma, 2 clusters containing 15 and 30 grains respectively). These clusters represent the first influx of eroded sediment from the eastern lying Whitsunday Volcanic Province.

4.2.2 Winton Formation Grain Ages The lower Winton Formation (Isisford, Queensland, 62 grains), contains few pre-

Permian clusters. Pre Neoproterozoic grain clusters are represented by four peaks (n<2).

Neoproterozoic populations are represented by four discrete clusters N<3 each) (Fig. 3-4).

These older clusters are likely reworked polyphase grains from northern and northeastern lying terranes (Mt. Isa, Greenvale and the Cape River Metamorphics). Very late

Neoproterozoic to late Carboniferous nine discrete grains clusters are recognized (2

Pama Igneous Associations). Mesozoic populations range from the Triassic, which average about four grains in each cluster to the middle Cretaceous with several clusters containing

27-30 grains each. Grains between the 300 and 100 Ma are interpreted to be derived from the New England and Kennedy Provinces along with the Whitsunday Volcanic Province.

The Middle Winton (McKinlay_16, 63 grains) contains single cluster prior to the

Neoproterozoic at ~1.55Ga (2 grains) (Fig. 3-4). These grains are likely from north to northeastern sources (Greenvale and the Cape River Metamorphics, and Mt. Isa

Provenance). Very minor Paleozoic grain clusters fluctuate between 2-4 grains within each. Distinct clusters only start to occur at 200 Ma, containing between 6-8 grains in each.

The first several distinctive clusters occur in the middle Cretaceous (~34 grains), and also

78 contains the first occurrence of the youngest subset of grains between 92-94 Ma. These younger clusters are interpreted to be solely sourced from the Whitsunday Volcanic

Province which was deposited in lowland swamps, deltaic and coastal deposition environments.

The Upper Winton Formation contains a broader suite of grain clusters

(McKinlay_02 (63 grains): Bladensburg National Park (82 grains), Lark Quarry

Conservation Park (Eureka (76 Grains) & Hades Hill (42 grains), Blackall_2 (73 grains),

Eromanga_1 (99 grains), 435 grains total); however, pre Mesoproterozoic grains remain very minor (2-3 grains each) (Fig. 3-4). Nine minor grain clusters occur within the

Paleozoic, and greatly fluctuate between each cluster (10

Cretaceous grain clusters greatly increase in the number of grains within each cluster. At

300 Ma the main clusters includes ~75 grains, yet the 100-92 Ma cluster includes 220 grains. These major subsets of grains are interpreted to have been sourced from that later

New England and Kennedy Provinces and the Whitsunday Volcanic Province and were deposited in fluvial and associated floodplain environments.

4.3 Kolmogorov-Smirnov Test) This study seeks to address whether or not detrital zircons collected from throughout the Winton Formation are derived from similar source terranes. This study also seeks to better contextualize these patterns over time, in particular the transitions from the

Mackunda Formation and the Winton Formation. If detrital zircon age populations are similar, then this would support the hypothesis for a common source area, whereas, if they do not compare well, then this suggests changing source areas through time or space.

Additionally, late Mesozoic detrital populations are largely unstudied from northeastern

Australian and these zircon populations could potentially provide important new data for understanding the tectonic evolution of eastern Australia. To address these questions, the null hypothesis that different distributions of detrital zircons (i.e., samples) are the same or are derived from the same age distribution, was tested using Kolmogorov-Smirnov goodness-of-fit tests. All samples were tested against each other and results are briefly discussed below and presented in Figure 3-5 and Table 2.

Despite general similarities in detrital zircon ages, the middle Mackunda Formation

(Pre-Triassic) failed all K-S tests between all other samples (Maneroo-331). This strongly indicates a fundamental change in source area or sediment delivery between the middle to

Upper Mackunda. This is the only sample that fails all K-S tests and shows no genetic relationship with other samples (Fig. 3-5a; Table 3-2). The upper Mackunda (Longreach 1-

1B) with the lower Winton Formation (Isisford), and moderately passed the K-S with the upper Winton Formation (McKinlay-2; Bladensburg; LQCP Eureka; Blackall-2; LQCP

Hades Hill; but failed K-S tests between Eromanga-1 and McKinlay-16) (Fig. 3-5a,b; Table

3-2). The lower Winton Formation (Isisford) strongly passed the K-S with the upper

Mackunda Formation and moderately to strongly passed the K-S with the upper Winton

Formation (Fig. 3-5a,b; Table 3-2). The middle Winton Formation (McKinlay-16) failed the K-S with the upper Mackunda and also the lower Winton Formation. The middle

Winton Formation did however, strongly to moderately pass with the upper Winton

Formation

Table 3- 2 K-S tests results for middle Mackunda Formation (Maneroo-331), the upper Mackunda Formation (Longreach 1-1B), the lower Winton Formation (Isisford, Queensland), the middle Winton Formation (MacKinlay-16) and the upper Winton Formation (Bladensburg National Park, Lark Quarry Conservation Park (Hades Hills and Eureka), GSQ Eromanga 1, Blackall 2, McKinlay-1). Tables display “P” values with confirmed similarity in grain sourced highlighted in yellow (P>5%) and negative results in white (P<5%). Table A displays similarity of all samples (Mackunda and Winton Formation samples), whereas Table B is a restricted comparison between Winton formation samples 81

Figure 3- 5 K-S test results displayed in terms of CDF’s, with all samples displaying weak to very strong (overlapping) similarity of source terranes except for that of Maneroo-331.

(Fig. 3-5a,b; Table 3-2). All upper Winton Formation samples barely pass to strongly pass the K-S when compared to other stratigraphic sections, as stated earlier. The upper Winton

Formation samples also display strong relationship with those that are regional neighbors and weak relationship with samples from further distances within the basin (Fig. 3-5; Table

3-2). For example, both Eromanga and Blackall display more similarity to Winton based samples then those of the Mackunda; however any similarity is weak. This could indicate similar headwaters and different drainage patterenes within the basin (north vrs south)

(Fig. 3-5).

4.4 Palaeocurrent Analysis Where possible, palaeocurrent orientations were collected from trough-cross stratified fluvial sandstones in the Winton Formation to help determine the orientation of flow in these ancient channel systems during the Cretaceous. The goal of this part of the study was to support the detrital zircon provenance reconstruction. However, deep surface weathering of the formation led to extremely poor three-dimensional preservation of cross bedding in the formation, which severely limited the impact of this approach. A thorough analysis of palaeocurrents in the Winton Formation resulted in only 35 unambiguous measurements (Fig. 6). However, a general trend in palaeocurrent orientations was observed, with most palaeoflow measurements ranging between northwest and southwest.

Paleocurrents observed in Bladensburg National Park (N=24) commonly ranged between

270º and 290º with some minor deviations at 135º, 215-218º (Fig. 3-6). One major deviation comes from a set of measurements from a single site that is oriented between 60º-

64º, which is interpreted as an isolated channel meander. Fewer orientations could be collected in Lark Quarry Conservation Park due to a lack of preservations (N=11).

Generally these flow orientations trended west by south/northwest (Fig. 3-6). This is in

83 agreement with palaeocurrent data presented in Hawlader (1990) and Bryan et al. (1997,

2012). Overall, these data indicate that fluvial channel flow was dominantly towards the west-southwestern channel flow was dominantly oriented to the west- southwest.

Furthermore, these findings corroborate the southwest transport of sediment via large–scale river system suggested by Vevers (2000) and King and Mee (2004).

Figure 3- 6 Simplified map of eastern Australia displaying the plausible source terranes and orientation of likely emplacement. Also displayed are paleocurrent data derived from Bladensburg National Park (n=24) and Lark Quarry Conservation Park (n=11) where outcrop Quality permitted.

4.5 Lu-Hf Results To further elucidate the provenance interpretations and tectonic implications of the detrital zircons from the Winton Formation, ~100 grains (83 of which yielded good results) of different ages were selected for Lu-Hf isotope analysis (Fig. 3-7). In particular, grains were selected from the pre-Mesozoic (New England & Kennedy Provinces), Triassic and

Jurassic (terranes undocumented), and the Early and Middle Cretaceous (Whitsunday

Volcanic Province) populations, and from all stratgraphic levels across the basin. The majority of grains yielded ƐHf(T) between +5 and +10 (a few select grains had just below a

+12 value) (Fig. 3-7). There is no discernable change in the isotopic pattern transitioning from the Mackunda Formation into the overlying Winton Formation. A handful of grains yielded negative Hf (T) values. This is interpreted to reflect incorporation of older crustal components into the magma.

The Hf isotope data indicates that parent magmas for many of the zircons were derived from a mixed source or from a range of isotopically heterogeneous sources, rather than a single homogeneous crustal or mantle source. This is because at any single age, in both individual samples and samples grouped by formation, the data scatter over several Ɛ

Hf units, indicating the involvement of isotopically-contrasting source reservoirs. None of these grains, however, exhibit isotopic signatures that suggest derivation from a depleted

MORB source mantle (Fig. 3-7) (Kemp et al., 2009). Additionally, in each sample analyzed

(individual or by formation), a consistent increase in radiogenic Hf is identified with a

85 decrease in age.

Figure 3- 7 Lu-Hf isotope results presented in ƐHf(T) for both detrital zircon recovered from the Winton and Mackunda formations from the Eromanga Basin (A: N=83)) and those derived from the New England Province (N=258). C) Compared results, which display a common trend in ƐHf(T) (+5≤n≤+10) with no discernible change between 350-92 Ma. (Modeled from Vervoort et al., 1999)

5. Discussion

5.1 Interpreted source areas and fluvial drainage patterns for the Winton Formation Similar detrital zircon ages from various stratigraphic levels in the Mackunda and

Winton formations suggest common source terranes, most notably for pre-Mesozoic

Populations. However, the amount of grains in each pre-Mesozoic population greatly decreases up-section by sample in the Winton Formation. Moreover, when compared to the

Winton Formation, the middle Mackunda (Maneroo-1) based samples lack the youngest detrital record (n>120 Ma). Furthermore, Mackunda samples preserve a much broader zircon-age spectra, including grains that are most likely sourced from Proterozoic terranes to the north and northwest. In addition, many of the Mackunda Grains preserve a significantly higher proportion of grains that clearly represent reworked grains that have been transported into the region at earlier stages and deposited in a variety of Proterozoic basins, prior to being recycled during the Cretaceous. These older grain ages quite likely originated in Central and Western Australia. Thus, we interpret this diverse range of zircon populations to represent a broad spectrum of potential source populations that were located to the north and east of the sample site. Given the fully marine nature of the Mackunda

Formation, the most likely mechanism for this diverse array of provenance signature reflects long shore current detrital mixing (e.g., Sircombe, 1999) (Fig. 8a,b). This is further supported by K-S test results, which indicates that the middle Mackunda Formation shows no genetic relations to any stratigraphically higher samples (Winton Formation in particular).

The Upper Mackunda is characterized by the retreat of the Eromanga Shallow

Seaway from offshore depocentres to nearshore coastal and deltaic environments.

Continuing up section, the lowermost Winton Formation transitions from tidal swamps and

Figure 3- 8 Simplified paleogeographic reconstruction maps with symbols showing the retreat of the Eromanga Shallow Interior Seaway and the development of the west to south-westward basin drainage which is similar to conditions seen today. A) coastal deposits to shallow marine conditions were emplaced no older than 104-102 Ma, B) distal deltaic to coastal deposited no older than 100-102Ma, C) fluvial floodplain dominantd deposits no older than 92-94Ma, D) terrestrial associations of sediments trace and body fossil assemblages (modified from Veevers, 2000; and references therein).

88 delta plain depocentres to distal fluvial-floodplain environments with meandering channels

(Tucker et al., 2014, see Ch. 4; Fig. 8a,b). Upper Mackunda to the lower Winton

Formation samples yield a considerably less diverse spectrum of detrital zircon populations dominantd by Mesozoic sources; however, minor input or recycling of older strata sourced from Proterozoic terranes to the north and northwest still persists. By utilizing these younger populations, we can reliably interpret that the upper Mackunda Formation was deposited no more than 102-104 Ma and the lowermost Winton Formation was deposited no more than 100-102 Ma (Tucker et al., 2013). These younger grain populations are interpreted to be derived from eastern volcanic sources, in particular, the Whitsunday

Volcanic Province. The close match in ages found in the Whitsundays with detrital zircon grain ages reported here provide strong support for this hypothesis, which was originally put forward by Bryan et al. (1997). These young grains within the uppermost Winton

Formation are of similar age to the Whitsunday Volcanic Province, and hence interpreted to be sourced from these volcanics. Furthermore, K-S tests between different upper Winton

Formation samples uniformly pass and indicate a similar provenance pattern during both the lower and upper Winton Formation throughout the basin. K-S tests between all middle and upper Winton Formation samples also indicate similar provenance sources, which are confirmed by similar paleocurrent measurements (Fig. 8c,d). Hence, the source and mechanism for sediment delivery into the basin from areas along the eastern arc remained fairly constant through the early to Middle Cretaceous.

The most significant observation between the lower Winton provenance and that of the middle to upper Winton provenance is a major younging in maximum depositional age.

Whereas the youngest grains observed in the Upper Mackunda and Lower Winton intervals

89 range from 104-102 Ma and 102-100 Ma, respectively; the youngest grains observed in the middle to upper Winton Formation are uniformly in the 92-95 Ma range (Fig. 8c,d). The

K-S tests’ results confirm that source areas have remained similar between the upper

Mackunda Formation and the upper Winton Formation. Lu-Hf isotope results for younger populations distinctly display similar values to zircons derived from older arc rocks of the

New England and Kennedy Province. Furthermore, Lu-Hf isotope signatures also remain similar between the ~200 Ma and the younger 95-92 Ma population, indicating that the change in grain ages does not reflect a shift in source areas for these young grains (Fig 8d).

This implies that the Whitsunday volcanics were active for considerably longer than previously thought (Bryan et al., 1997)

5.2 Source & Tectonic History of Mesozoic Detrital Zircons

5.2.1 Model-1: Long-lived Phoenix-Pacific Plate Subduction The classic interpretation of late Mesozoic volcanism along the eastern margin of

Australia is one of long-lived subduction of the Phoenix/Pacific Plate associated with the breakup of Gondwana (Henderson & Stephenson, 1980; Coney et al., 1990; Holcombe et al., 1997b; Veevers, 2000a) (Fig. 3-9). Collot et al. (2009; and others) have applied magnetic, geochemical and zircon data to interpret moderately active volcanism from the

Permian to the Cretaceous along the eastern margin of Australia. Volcanic activity during this time should have been present as a result of low-angle subduction of the Phoenix-

Pacific Plate beneath the continental crust (Veevers, 1991a,b, 2000; Collot et al., 2009). A proposed modern analogue is the present day Cordilleran orogenic belt along the western margin of South America (Veevers, 1991a,b; Collot et al., 2009) (Fig. 3-9).

5.2.2 Model-2: Northeastern Gondwana S- LIP To address unresolved issues of irregular tectonic and sedimentary histories within the Surat and Eromanga Basin Ewart et al., (1992), and more recently Bryan et al., (1997,

2000, 2003, 2008, 2011, 2012) challenged the established model of long-lived subduction

(Fig. 3-9). Utilizing the well-documented volcanic rich sediments, Bryan et al. (1997) hypothesised a link between the apparently volcano-lithic sedimentary provenance of the mid-Cretaceous strata in the Eromanga Basin and coeval Cretaceous volcanics in the

Whitsunday Volcanic Region. Over time, Bryan et al. (2000, 2008, 2012) interpreted these detrital grains to have been emplaced by a very large, long-lived S-LIP. They attributed the initiation of this S-LIP to a shift from subduction-related to rift-related extensional tectonism in the Late Jurassic to Early Cretaceous, and associated underplating along the eastern margin of Australia (Bryan et al., 2012). In this model, the Whitsunday S-LIP formed an intracratonic intercontinental volcanic belt (around ~2,500km long and 300 km wide), which originally extended along much of the northeastern coastline of Queensland

(Fig. 3-9). This intraplate rift preceded the development of the main rifting and eastern- most fragmentation of Gondwana and progressive rifting of the Tasman Sea Basin (~95-85

Ma) (Bryan et al., 2000). The volcanics and their intrusive equivalents are interpreted to have been emplaced as mostly siliceous-rich magmas, dominantd by rhyolitic and dacitic compositions (Ewart et al., 1992). Both Bryan et al. (1997, 2000, 2012) and Ewart et al.

(1992) suggest that volcanic activity initiated around ~132 Ma, with the main period of activity between 120 to 105 Ma. They argue that intermittent volcanism continued until 95

Ma, marking the termination of the Whitsunday S-LIP (Bryan et al., 2012) (Fig. 3-9).

The initiation of the Whitsunday S-LIP is suggested to be like any other hotspot with asthenospheric-derived mafic volcanism (Bryan et al., 2012). Bryan et al. (2000, 2012)

91 interpret that volcanic activity in the Whitsunday S-LIP was dominantd by multiple low- relief volcanic vents with large calderas. Older volcanics are suggested to have been

Figure 3- 9 Tectonic reconstructions of the Eastern Australian Margin: Scenario One - Continued subduction of the Phoenix/Pacific plate under the northeasterly migrating Australian continental crust Scenario Two - S-LIP+ Rifting between 135 and 95 Ma Veevers (2000).

92 dominanted by explosive and non-explosive dacitic pyroclastics, indicating a bimodal suite of lavas and ignimbrites (Bryan et al., 2000).

5.2.3 Regional Tectonic Relationships and Implications Several studies have confidently reassembled many small Zealandian fragments including Lord Howe Rise, Challenger Plateau, New Caledonia and the North Island of

New Zealand, to form a rifted ribbon of what was once attached to the eastern Australian

Margin (Sutherland, 1999; Pickard et al., 2000; 2008; Adams et al., 2009; Collot et al.,

2009; Cluzel et al., 2011). Furthermore, there is overwhelming petrological, geochemical, mineralogical, faunal and detrital evidence for the tectonic relationships of the pre-Jurassic

Zealandia and eastern Australia (Pickard et al., 2000; Schellart et al., 2006; Adams et al.,

2007, Waschbusch et al., 2009; Tulloch et al., 2009; Cluzel et al., 2010, 2011, 2012).

Similar relationships can be correlated to surveys of the Lord Howe Rise by Higgens et al.

(2011) with striking similar populations at 97 Ma. However, there is a distinct lack of similar detrital populations between 90 -74 Ma. This can be readily interpreted to show pre and post-stages of rifting between the Australian Craton and Zealandia. With the incorporation of the recovered detrital record from the Eromanga basin, it is interpreted that

Mesozoic grain populations are derived from similar sourced terranes as those identified in

Zealandia. This similar detrital record is interpreted to be derived from volcanic activity along the Eastern to Northeastern Australian margin, including but not limited to the New

England Orogeny, Kennedy Province, and the later Whitsunday Volcanic Province.

Significantly, each of these former fragments of the Australian continental margin also preserve a distinctly younger population of detrital zircons (~89-85 Ma) not observed in this study or by other workers in any detrital zircon studies or geochronologic

93 investigations of the Whitsundays Volcanic Province (Cluzel et al., 2011). Coincidently, several early to middle Mesozoic southern Zealandian terranes, especially those preserved in New Zealand (ergo. Nambucca Black) contain Triassic and Jurassic terranes, now missing (rifted away) from the current eastern Australian margin (Campbell & Coombs,

1966; Adams et a., 2007; Tulloch et al., 2009; Cluzel et al., 2010, 2011; Adams et al., 2012,

2013a,b). Other parsimonious detrital zircon records have recently been described from

Northland Allochthon and Mt. Camel Terranes, New Zealand (125, 120, 116, 108, 104 Ma)

(Adams et al., 2009; Cluzel et al., 2011). Furthermore, studies by Cluzel et al. (2011) advocated for a localized New Caledonia source terrane for grains younger than 89 Ma.

This would then imply that rifting initiated just prior to 90 Ma, and the 100-92 Ma zircon population in the Eromanga represents the terminal stage of subduction related volcanism along the rifted eastern margin.

5.3 Lu-HF Isotope Analysis If, during the Mesozoic (~250-92 Ma) there was a dramatic shift in the mechanism for volcanic activity in northeastern Australia (i.e slab subduction to S-LIP), this should be reflected in the isotopic record of magmatic rocks. However, a significant portion of the

Eastern Australian Margin (Triassic-Cretaceous rocks) rifted away or was recycled. As a result, the detrital zircon record represents the best archive to reconstruct the last vestige of tectonic events in northeastern Australia prior to rifting.

Currently, a LIP or large igneous province is defined as a large volume intraplate magmatic event consisting of flood basalts, layered intrusions (ultramafic), and many crisscrossing networks of dyke swarms (Ernst & Buchan, 2001; Ernst et al., 2013). LIPs are characteristically short-lived (1-5 Myr) and are the result of very deep-seated mantle

94 up-wellings (Ernst & Buchan, 2001; Ernst et al., 2013). Silicic LIP’s can occur in places where underplating has occurred simultaneously with the emplacement of a LIP (Bryan &

Ernst, 2008; Ernst et al., 2013). In the first instance, the duration of S-LIP activity in the

Whitsunday Volcanic Province (upwards of 40 Myrs) seems to contradict the defined LIP parameters. To accommodate this, Bryan et al. (2012) suggested that this S-LIP experienced short intervals of magmatism between 130-95 Ma; pulses ranging between several million years to ~15 Ma (Ewart et al., 1992; Bryan et al., 1997). If this were accurate a strongly negative Ɛ Hf(T) would be reflected in the Hf isotopic records relative to age. Secondly, as LIP’s are classically defined as the result of injected mantle plumes, even if there was underplating, it would stand to reason that over time this would result in dramatic shifts in radiogenic isotope compositions and likely reflect strongly depleted and enriched isotopic signatures (- and + Ɛ Hf(T)) (Söderlund et al., 2006). However, this is not the case for Cretaceous detrital zircon populations analyzed in this study. Moreover, a relatively continuous spectrum of Cretaceous (and indeed late Mesozoic) grain ages documented in this study are also inconsistent with the pulsed volcanism model required to develop an S-LIP. Together, these data call the S-LIP model into question.

For instance, this study has dated numerous grains between 300- 130 Ma via LA-

ICPMS (Tables 2-1a-e & 3-1a-e). ƐHf(T) values for those grains indicate Mesozoic magmatic activity akin to that described for northeastern Australia between the Ordovician through to the Permian. For the most part, a majority of the grains are above ƐHf(T) +5 indicating that much of the source material was derived from young crust or mantle.

Isotopically, this study recognizes a fairly continuous pattern for zircons300 and 92 Ma in all samples (ƐHf(T) values between +5 and +10), indicating a dominance of such juvenile

95 sources. A minor caveat to this observation is a cluster of grains (~4) display slightly lower

ƐHf(T) values between 180-140 Ma. This may be an anomalous isotope signature, given the small number of data; however, this may also represent a period in which a sizable amount of older crust was incorporated into the magmas. If this is the case, it could represent a period of contraction and crustal thickening, which is isotopically and tectonically well documented throughout the tectonic history (pre and post-Mesozoic) of the Eastern

Australian Margin (Sdrolias et al., Schellart et al., 2006; Kemp et al., 2009; Li et al., 2014), involving migration of arc magmatism inboard of the continent. To properly address this paradigm, a greater number of Hf isotope data are needed for grains between 130-180 Ma.

If this event occurred, its effect was short-lived as Ɛ Hf(T) between 135-85 Ma immediately returns to between +5 and +11. This indicates a return to juvenile magma sources and away from old continental crustal influences.

Regardless, minor ƐHf(T) fluctuations between 300-92 Ma are interpreted to be a reflection of heterogeneity in the magma source material (crustal). This mixing is due to variable incorporation of young and old crustal material; potentially including portions of

‘missing’ northeastern Australian Triassic-Jurassic sources terranes and possible

Precambrian terranes, along with fractions derived from the subducted Phoenix-Pacific

Plate. These materials were available to be incorporated into arc magmas. These results are also consistent with ƐHf(T) signatures and patterns from the New England Orogen (e.g., see Kemp et al., 2009, and Fig. 3-6), which greatly strengthens the correlation of similar tectonic conditions from the Ordovician to the mid-Mesozoic in northern Australia.

Additionally, between 300 Ma (New England and Kennedy Province derived) and

92 Ma (Whitsunday Volcanic Province) there is a broad, continuous increase in zircon ƐHf.

This increase in epsilon Hf values that continues until 92 Ma might have tectonic connections to the eventual outboard slab retreat and unzipping of the northern Lord Howe

Rise (Zealandia) and the shift to an extensional tectonic environment. This pattern has been recognized by Kemp et al (2009), who showed a similar rise in ƐHf and ƐNd in the

Lachlan and New England Orogeny, and interpreted this to represent slab retreat, such that magmas were progressively isolated from old crustal sources. If this is a continuation of the same slab subduction into the Late Mesozoic, then this may indicate a repeated tectonic pattern during the life of this east-northeastern migrating subduction margin. This pilot study has only begun to shed light on what is a very complex tectonic regime, and requires a great deal more study.

6. Conclusion The overriding goal of this project was to comprehensively investigate the sedimentary provenance of the mid-Cretaceous Winton and Mackunda formations within the Eromanga Basin via U-Pb detrital zircon geochronology and Lu-Hf isotopes. By utilizing ten detrital zircon samples (~719 U-Pb ages & 83 Lu-Hf analyses) from key stratigraphic levels within the Winton and underlying Mackunda formations, we are able to confidently interpret both syndepositional and polyphase and/or reworked source terranes.

Moreover, these data yield significant new insight into the tectono-magmatic history of eastern Australia over the Mesozoic. The identification of Triassic to Early Cretaceous grain populations is a period of time not previously associated with significant arc magmatism in Eastern Australia. This fairly continuous detrital zircon record is tested via

Lu-Hf against more well established terranes including the New England Province and found to be of similar isotopic signatures. By combining U-Pb detrital Zircon ages and ƐHf

97 values, we can interpret that these detrital grains were derived from a mixed juvenile magma source from a fairly long-lived tectonic system (300-92 Ma). By coupling these data with k-s and paleocurrent, this study can confidently conclude that grains were derived from a long-lived continuous subduction margin along the eastern margin of Australia, and then transported via transverse fluvial systems into westerly lying continental basins.

Finally, these data can also provide crucial context, which support a recent hypothesis discussed by Veevers (2000) and more recently MacDonald et al. (2013) which suggests long-distance transport and sedimentary provenance of the Cretaceous Ceduna Delta

System is connected to the Whitsunday Volcanic Province via long-lived and current-day southwestward drainage through the Eromanga Basin.

Chapter 4

Depositional environments, stratigraphy and correlation of the Winton Formation: implications for Australia’s most important Mid-Cretaceous terrestrial ecosystem

ABSTRACT The Winton Formation of northeastern Australia is recognized as one of the most important source of middle Cretaceous terrestrial floras and faunas in Australia. Yet, detailed sedimentologic investigations have never been comprehensively attempted. Therefore, this study focused on conducting a detailed facies analysis at key fossil localities in outcrop, along with local to regional scale stratigraphic correlation of these key fossil assemblages. Early outcomes included the identification of 23 distinct facies, which colbinate into nine distinct assemblages. Moreover, this study advocates for the informal speration of the Winton Formation into an upper and lower unit. This separation is based on gross changes in facies, alluvial architecture, maximum depositional ages of detrital zircons, and trends in the faunal composition. This study has determined that the lower Winton Formation predominatly preserves marginal-marine to coastal palaeoenvironments, which is interpreted as the initial stages of seaway regression, and occurred no later than the very latest Albian to very early Cenomanian. On the other hand, the upper Winton Formation is fully alluvial and preserves a broad range of floodplain depositional environments dominanted by multiple westward meandering river systems, which occurred no later than the Cenomanian-Turonian boundary. Thus, this study finds that the upper Winton Formation’s multitaxic vertabrate assemblages of Lark Quarry Conservation Park and Bladensburg National Park are near-contemporaneous. Whereas, Isisford, Queensland presveres temporally dissimilar multitaxic vertabrate assemblages. These results suggest that upper Winton fossil ammaeblages are likely correlative to the Rio Limay Subgroup of the Neuquén Basin of Argintina, and the Santo Anastácio and Adamantina Formations of the Bauru Basin, of Brazil. This provided significant context for reconstructing the important mid Cretaceous continental flora and fauna of the Winton Formation.

100

1. Introduction The Cretaceous represents one of the most dramatic periods of supercontinent breakup and landscape fragmentation in Earth history (Stampfli & Borel, 2002; Skelton,

2003; van Hisbergen et al., 2011). Plate reorganization had profound effects on terrestrial ecosystems and landscapes, leading to palaeobiogeographic dispersal and development of evolutionary patterns that ultimately influenced the modern world (Skelton, 2003; Mannion et al., 2012, 2014). After the separation of Pangea, archosaur lineages from Gondwana are widely described as being evolutionarily distinct from their Laurasian counterparts

(Bonaparte 1986; Sereno, 1997; Agnolin et al., 2010; Benton et al., 2014). However, many eastern Gondwana assemblages still remain poorly understood (Agnolin et al., 2010). This is especially true for Cretaceous fossil assemblages and ecosystems from Australia. The best record of post-separation, Cretaceous dinosaur faunas and continental ecosystems in

Australia comes from the expansive continental deposits of the Winton Formation, in central Queensland. Recent fossil discoveries in the Winton Formation include a diverse range of dinosaurs, along with crocodilians, lungfish and primitive ray-finned fish, aquatic lizards, turtles, invertebrates and a high diversity of plant macrofossils, including some of the world’s earliest flowering plants (Dettmann et al., 1992; Molnar, 2001; Salisbury

2006a,b; Hocknull et al., 2009; Agnolin et al., 2010; Fletcher and Salisbury, 2010; Molnar,

2010; Leahey & Salisbury, 2013). Despite a renaissance of paleontological research in the

Winton Formation over the last decade, the stratigraphy, age, and depositional environments of this important stratigraphic unit remain problematic and poorly understood; both at a local and regional-scale.

Historically, the Winton Formation has been interpreted as the uppermost depositional phase of the Eromanga Basin, represented by a westward-thickening 101 continental wedge deposited during the final regression of the Cretaceous Eromanga

Shallow Seaway in Queensland (Fig. 4-1) (Veevers, 2000a,b; MacDonald et al., 2013).

Recent detrital zircon investigations of the Winton Formation have significantly improved the temporal constraints on deposition of the formation and identified the Basin’s major sources of sediment (Tucker et al., 2013; 2014), making it possible to place this important flora and fauna into a more precise global context.

Figure 4- 1 Map of eastern Australia displaying surface exposures of the Winton and Mackunda formations in the Eromanga Basin of Queensland Australia. Symbology identifies locations of fossil localities and GSQ Core localities, which were utilized for both stratigraphic and temporal correlation. Map is modified from GeoScience Australia 2013.

102

This paper presents the results of a detailed sedimentologic investigation of the

Winton Formation, to improve the understanding of the depositional environments of the

Winton Formation and to correlate Winton deposits and fossil localities across the basin.

Good exposure of the Winton Formation is relatively limited and most stratigraphic sections are isolated. This study combines both field investigation and detailed core logging, from sites across the basin. Detrital zircon samples were collected at most of the key field sections, as well as from a number of cores. These samples have previously been investigated for sedimentary provenance and maximum depositional age constraints

(Tucker et al., 2013; Tucker et al., in review), however, in this study we re-examine these data and demonstrate their potential for stratigraphic correlation (Fig. 4-1). Here, we combine these datasets to create a basin model to identify new structural and stratigraphic trends that have important implications for correlating and interpreting the Winton flora and fauna.

2. Geological Background

2.1 Basin History The Eromanga Basin is interpreted to have initiated in the Triassic, which continued into the late Mesozoic (Draper, 2002; Gray et al., 2002 Cook et al., 2013).

The Eromanga Basin forms the central sub-Basin of the Great Artesian Basin, with the

Carpentaria Basin to the north and Surat Basin to the southeast. The Eromanga is bounded to the west by the Georgina, Amadeus, and Officer Basins, the Northern

Territory and the Western Plateau. Within the study area from east to west, the

Eromanga Basin overlies the Galilee, Adavale, Cooper, Warburton, Simpson, and

Pedirka Basins along with the centrally located Warrabin Trough. Presently, within the

103

Eromanga, drainage patterns generally have a SW-oriented flow into the Lake Eyre region of South Australia.

The tectonic origin and developmental history of the Eromanga Basin have been debated for many decades. Because of its location away from active plate boundaries, the Eromanga Basin has traditionally been interpreted as an intracratonic basin

(Angevine et al., 1990; Draper, 2002). However, some authors have argued that the

Eromanga Basin is a foreland basin (e.g., Gallagher, 1990). Classically, intracratonic basins are interpreted to be associated with failed rift systems and linked to regional subsidence and slow, steady generation of accommodation space; resulting in a broad concave up morphology (Kingston et al., 1983; Mitchell and Reading, 1986; Draper,

2002). Instead, the Eromanga Basin displays a concave down shape, more typical of a foreland Basin (Gallagher, 1990; Draper, 2002). Moreover, foreland basins are typified by pulsed subsidence and generation of accommodation space (e.g., rapid then slow), which compares more favourably with the patterns observed in the Eromanga Basin

(Gallagher and Lambeck, 1989). The association of a remnant volcanic arc and hypothesized late Mesozoic subduction margin along the east coast of Australia also emphasizes a probable foreland or retroforeland basin origin for the Eromanga Basin

(Gallagher & Lambeck, 1989; Gallagher et al., 1994; Russell & Gurnis; 1994; Veevers,

2000a,b). Furthermore, the western portions of the basin apparently experienced higher rates of subsidence, due to asymmetric loading as would be expected with thrust loading on the western side of the basin (Gallagher, 1990; Draper, 2002).

Several alternative mechanisms for subsidence have been suggested, including:1) thermal contraction (Gallagher & Lambeck, 1989); 2) passive thermal subsidence (Fielding, 1996); and 3) dynamic topography (loading) (Russell & Gurnis, 104

1994). Among the varied explanations to explain the observed basin geometry, extensional, collisional or plate rotation, mantle heat flux, and detached slab subduction at the Eastern Australian Margin have all been proposed; however, little agreement exists and the question of basin type and subsidence remains largely unresolved

(Kuang, 1985; Gallagher & Lambeck, 1989; Krieg et al., 1995; Gurnis et al., 1998;

Draper, 2002; Mavromatidis & Hillis, 2005; Heine et al., 2008;

Matthews et al., 2011; Cook, 2012).

2.2 Basin Stratigraphy Greater consensus exists concerning the Mesozoic stratigraphy of the Eromanga

Basin. During the late Mesozoic, but particularly during the Early Cretaceous, the

Eromanga Basin was influenced by marine transgressions (Exon & Burger, 1981; Burger,

1986; Draper, 2002; Gray et al., 2002). However, correlation of sea-level cycles recorded in the basin to eustatic sea level curves has proven difficult (Draper, 2002). Sedimentation within the Eromanga Basin is interpreted to have initiated in the Late Triassic (Carnian to

Rhaetian), with the Cuddapan Formation (Draper, 2002). The Cuddapan is between 20-50 m thick, and commonly composed of quartzose sandstone, carbonaceous siltstone, mudstone, and coal (Alexander et al., 1998; Gray et al., 2002). This pattern of deposition continued into the unconformably overlying Poolowanna Formation (Moore, 1986). The

Early Jurassic (Sinemurian-Toarcian) Poolowanna has a maximum thickness of 165 m and is commonly composed of granular to fine grained sandstone in lower sections, and siltstone, mudstone, and coals in upper sections (Almond, 1983; Moore, 1986). The

Poolowanna Formation depositionally represents repeated shifts from fluvial to lacustrine facies characteristics. Uppermost Poolowanna Sedimentary deposits record sea level rise can be directly correlated to the Toarcian global sea-level rise. The Middle Jurassic

105

(Aalenian-Callovian) is dominantd by the deposition of the Hutton Sandstone and the

Birkhead Formation (Exon, 1966; Gray et al., 2002). The Hutton ranges in thickness between 90-210 m and is commonly composed of fluvial to aeolian sandstone with very minor siltstone and mudstone (Almond, 1983). On the other hand, the Birkhead Formation can range in thickness between 40 to 100 m, and is composed of interbedded sandstone and siltstone. The Birkhead Formation is interpreted to have dominantly been emplaced by fluvio-lacustrine process with variable influx of volcaniclastic sediments (Gray et al.,

2002).

The Late Jurassic (Oxfordian-Tithonian) sedimentary record is preserved in the

Adori Sandstone and the conformably overlying Westbourne Formation (Fig. 4-2) (Gray et al., 2002). The Adori sandstone in its maximum thickness is estimated to be 55 m, though ranges between 15-55 m. Commonly, the Adori is composed fluvial coarse sandstone to siltstone (Almond, 1983, 1986). Deposition continues into the overlying Westbourne

Formation which is usually between 70 to 130 m thick and commonly composed of near shore to lacustrine sandstone, siltstone and shale Almond, 1983, 1986; Gray et al., 2002).

The Late Jurassic- Early Cretaceous (Tithonian to very early Valanginian) is preserved in the Hooray Sandstone. The Hooray can range between 90-165 m in total thickness. The

Hooray is typified by fluvial derived sandstone with minor siltstone and mudstone.

106

Figure 4- 2 Time scale displaying Middle Jurassic to Quaternary sedimentary units within the Eromanga Basin including stratigraphic location of archosaurian assemblages in the Winton Formation. Modified from Draper et al., 2002 & Tucker et al., 2013.

107

The Cadna-Owie Formation is the lowest Cretaceous unit deposited in the Basin

(Valanginian-Early Barremian). It commonly ranges in thickness from 60-90m and is composed of interbedded sandstone, siltstone and mudstone (Dettmann, 1987b; Gray et al.,

2002). The Cadna-Owie is interpreted to have been deposited within a marginal shallow marine to deltaic setting (Senior et al., 1975; Gray et al., 2002). Conformably overlying this unit is the Wallumbilla Formation, which marks the onset of the last major marine transgression of the Cretaceous into the basin during the Barremian to Albian stages (Fig.

4-2) (Williams, 1983b; Dettmann, 1984a, 1988). The Wallumbilla Formation is much thicker, ranging from 200-350 m. It is characterised by carbonaceous grey to black mudstone and siltstone that yield plant fragments and invertebrate shells. Deposition is interpreted to have been deposited under shallow-marine conditions with the upper units being deposited during marine regression (Gray et al., 2002). The very thinly bedded (20-

45 m) and conformably overlying Toolebuc Formation is dominated by poorly sorted, weakly-laminated mudstone, minor siltstone, conglomerate, and thin coquina lenses (Fig.

4-2) (Gray, 2002). The Toolebuc Formation is also interpreted to have been deposited in shallow marine conditions, which extended into the Albian stage (Moore et al., 1986). The conformably overlying 200-300 m-thick late Albian to early Cenomanian Allaru Mudstone continued the deposition of shallow marine units (Exon and Senior, 1976). This unit is dominated by blue-grey mudstone that is partially pyritic with interbedded calcareous siltstone, cone-in-cone limestone and minor sandstone (Exon and Senior, 1976; Price et al.,

1985; Helby et al., 1987; Draper, 2002; Gray et al., 2002).

The final regression of the Eromanga Seaway regression is recorded in the

Mackunda Formation. The Mackunda Formation commonly ranges in thickness 75-100m,

108 and is characterized by interbedded shallow-marine sandstone, siltstone, and mudstone, and is late Albian to early Cenomanian (Gray et al., 2002). The Mackunda is conformably overlain by the dominantly continental Winton Formation.

The final regression of the Eromanga Seaway regression is recorded in the

Mackunda Formation. The Mackunda Formation commonly ranges in thickness from 75-

100m, and is characterized by interbedded shallow-marine sandstone, siltstone, and mudstone, and is late Albian to early Cenomanian (Gray et al., 2002). The Mackunda is conformably overlain by the dominantly continental Winton Formation.

2.3 The Winton Formation The Winton Formation is exposed over large portions of Queensland along with portions of northern New South Wales, north-western South Australia and the southwestern corner of the Northern Territory (Fig. 4-1). The Winton Formation was originally described by Dunstan (1916) as the “Winton Series”, consisting of sandstones, shales, and minor coal seams that stratigraphically overlie predominantly marine sediments of the Mackunda Formation. Most recently, the uppermost Rolling

Downs Group consists of the Manuka Subgroup, which is subdivided into the Winton and Mackunda formations.(Fig. 4-2).

The Winton Formation conformably overlies the Mackunda Formation and ranges in thickness between 400-1000 m; thickness generally increases westward (Gray et al., 2002). The Winton Formation is composed of interbedded sandstones, sandy siltstones, siltstones, and mudstones with minor coal seams and intraformational conglomerates (Clark, 1965; Vine et al., 1967; Casey, 1970; Senior and Mabbutt, 1979;

Coote, 1987; Draper, 2002). The Winton Formation is well-known for its entombed

109 fossils, which includes plants, palynomorphs, invertebrates and vertebrates. The plant diversity of the Winton Formation is particularly significant, with a range of taxa including members of the Bennettitales, Conifers, Cycadophytes, Ferns, Ginkgoales

(Ginkgophyte), Pentoxylolean, and most significantly early angiosperms (Burger and

Senior, 1979; Burger, 1990; McLoughlin et al., 1995; Pole and Douglas, 1999;

Mcloughlin et al., 2010). Angiosperm palynomorphs in the formation have been well- documented and reviewed by Dettmann and Playford (1969) and Burger (1980, 1981,

1986, 1989, 1990). Prior to more precise age constraints provided by detrital zircon geochronology (Tucker et al., 2013), the Winton Formation was fairly broadly dated between Albian to Cenomanian based on palynology. Maximum depositional age constraints from detrital zircons now indicate that the lower sections of the Winton

Formation are no older than ~100-102 Ma, while the upper Winton Formation has been shown to be no older than the latest Cenomanian, with evidence for the uppermost strata being Turonian (~93-92 Ma) or younger (Tucker et al., 2013; 2014).

The Winton Formation is one of the only stratigraphic units in Australia that preserves a middle Cretaceous terrestrial vertebrate body-fossil assemblage (Agnolin et al., 2010). New fossil vertebrate discoveries in the Winton Formation are beginning to confirm that close similarities exist between this fauna and those of comparable age units in South America (Sereno, 1997; Agnolin et al., 2010; Leahey & Salisbury 2013;

Mannion et al., 2013). Important basal archosaur fossils have recently been described from the Winton Formation, including the Eusuchian crocodyliform Isisfordia duncani, found near Isisford, Queensland. This important taxon represents a transitional form between Neosuchian and Crocodylian representatives (Salisbury et al., 2006). Recently discovered vertebrate fossils in the Winton Formation range from isolated bones and 110 bone fragments to nearly complete to associated skeletons (Coombs and Molnar 1981;

Salgado et al., 1997; Wilson and Sereno, 1998; Molnar, 2001; Salisbury and Molnar,

2005). Among the recent discoveries are two new, fragmentary sauropod skeletons,

Diamantinasaurus matildae and Wintonotitan wattsi, and a fragmentary theropod,

Australovenator wintonensis. This also includes the first isolated teeth from an ankylosaur from the Winton Formation (Hocknull et al., 2009; Leahey & Salisbury,

2013).

2.4 Structural Geology of the Northern Eromanga Basin Several major faults have been identified within the Eromanga Basin, including the

Canaway Fault, the Weatherby fault system, and Cork fault system, along with numerous synclines and anticlines across the basin (Fig. 4-3) (Casey, 1967a,b; Mathur, 1983; Moss and Wake-Dyster, 1983; Moore & Pitt, 1984; Wopfner, 1985; Finlayson et al., 1988;

Murray et al., 1989; Green et al., 1991; Cartwright & Lonergan, 1997). These residual features are interpreted to be thermally and/or tectonically related: tectonic subsidence related to plate motions and underlying mantal conditions (Gallagher & Lambeck, 1989;

Gurnis et al., 1998; Watterson et al., 2000; Veevers, 2006; Heine et al., 2008). It is generally assumed that the flat lying nature of most stratigraphic units in the basin underwent minimal tectonic deformation during deposition, and that most of the units are relatively conformable. However, recent studies focused on various economic aspects of the basin have resulted in the discovery of additional and more extensive faulting then previously recognised (Esso 1981, Carne & Alexander, 1997; Hower, 2011; Sentry

Petroleum, 2011a,b). This work has demonstrated that many individual coal seams are offset in places and are laterally discontinuous over tens of kilometres (Casey, 1967a,b;

Hower, 2011; Sentry Petroleum, 2011a,b).

111

Figure 4- 3 Map of exposed faults and structures within the exposed Winton and Mackunda formations within the study area. This figure highlights historical (A. Hoffmann et al., 1991) and recent (B. Howler et al., 2011) depicting a great deal more faulting than what was originally described. Map modified from GeoScience Australia 2013.

112

3. Study Area

3.1 Field Locations Three primary field localities were investigated during this project, including: expansive exposures at Bladensburg National Park (BNP) and Lark Quarry Conservation

Park (LQCP) near the town of Winton, Queensland, and an important fossil bearing locality with poor surface exposure near IQ (Fig. 4-4). Nine stratigraphic sections were measured at LQCP and an additional six sections were measured at BNP. Due to a lack of exposure, no sections were measured at IQ. Instead, the stratigraphy of IQ was investigated and correlated through investigation of regional well logs and cores.

Figure 4- 4 Arc GIS (10.1) Geo-referenced based map of exposed Winton and Mackunda Formation, along with the expansive Quaternary Alluvium & Black Soil across the study area. Map also displays location of all field sights and core localities across the study area. Cross section A-A’ displays the fairly constantly gentile dip across the basin and the general thickening of the Winton Formation in the west.

113

3.1.1 Bladensburg National Park Bladensburg National Park (BNP) covers more than 48,900 hectares of Mitchell grassland and channel country (Fig. 4-4). Much of the park is covered with “Jump Ups” or buttes that offer exposure of Winton Formation. Fossil exploration in the area has been minimal and yet to be documented.

3.1.2 Lark Quarry Conservation Park South of Winton Queensland, Lark Quarry Conservation Park (LQCP) is the site of one of the southern hemisphere’s best known dinosaur track sites (Fig. 4-4). The Lark

Quarry dinosaur tracksite (Australian Natural Heritage List, Place ID 105664) preserves between 3,000-4,000 dinosaur footprints in an area little more than 200 m2. The original interpretation of this site (Thulborn and Wade 1979, 1984, and 1989) suggested that a large theropod chased a small heard of ornithopodian dinosaurs away from a lake shore.

However, a more recent investigation of the trackway, in combination with morphometric analysis of the tracks and, detailed sedimentologic investigation of this site offers a different interpretation. Romilio and Salisbury (2010) identified the large tracks as those of a large, ornithiscian dinosaur, rather than a carnivore. And most recently, Romilio et al.

(2013) provide several lines of evidence to suggest that track site was produced over a much longer duration of time in a fluvial setting; not the result of a ‘stampede’.

3.1.3 Isisford, Queensland

Isisford, Queensland (IQ), in the heart of the Queensland Outback, is world renowned for the Isis Downs Shearing Shed, which is the largest in Australia (Fig. 4-4).

Fossil diversity is particularly rich in the region, but fossils are almost exclusively recovered from large sandstone concretions. The most important discovery in this area is

114 that of Isisfordia duncani, the earliest known Cretaceous crocodiliformes to exhibit modern crocodilian characters. Exposure of the Winton Formation in this area is exceedingly poor; however several locally derived cores were recovered in the area by Geoscience

Queensland and utilized for subsurface correlation.

3.2 Core Logging Drill core analysis of the Winton and Mackunda formations were conducted at the

Geological Survey of Queensland core storage house in Zillmere, Queensland. In total, eight cores were logged (although only five cores are utilized in this study). An additional seven cores were digitized using the Hylogger system and the images and spectral analyses for these cores were investigated. Finally, historic drill records/notes for eight additional cores were also used in this study, and together, data from each of these core records was used to construct the GIS basin model shown in Figure 4-4.

Eromanga 1 core was drilled east of Eromanga, Queensland (S26.6147/ E143.8801) and 250 m of Winton strata were logged, starting at surface. The Winton Formation in

GSQ McKinlay 1, near McKinlay, Qld (S21.5925/ E142.2597) was logged from surface to a depth of 240 m. GSQ Blackall 2 core (S 24.1626 /E144.2236), drilled near Blackall,

Queensland was logged between the Winton Formation at the surface exposure to a depth of 300 m. GSQ Longreach 1-1B was drilled northeast of Longreach, Queensland

(S23.1119/E144.5996) and was logged between the Mackunda Formation surface exposures and a depth of 350 m; however, no Winton Formation was encountered in this core. GSQ Maneroo 1, which was drilled northeast of Maneroo, Queensland

(S23.3859/E144.4679) and was logged between the Winton Formation at the surface a depth of 420 m. For additional stratigraphic correlation and visualization, published core

115 logs for a number of other GSQ core logs were also utilized, including those for: Manuka 1,

Muttaburra 1, Machattie 1, Connemara 1, Augathella 1, Quilpie 1, and Thargomindah 1-1a

(Fig. 4-4).

4. Methods

4.1 Stratigraphy and Sedimentation Sedimentologic analysis of the Winton Formation was conducted between 2010 and

2013. Detailed facies and architectural element analysis were performed following the conceptual framework established by Miall (1977, 1985) and modifications to this approach by Eberth and Miall (1991), Fielding (2001, 2006), Roberts (2007), and others.

For consistency, a uniform set of facies codes used to describe both the outcrop sections and cores. In outcrop, detailed measured sections were constructed at the decimetre scale in each of the three study areas using a Jacob’s Staff, Brunton Compass, and GPS.

Particular emphasis was placed on understanding and correlating important vertebrate fossil localities in each study area. This included construction of bonebed and trackway maps for sites at both BNP and LQCP based on historic records and photographs (Romillio et al.,

2013). Many horizons were walked out in order to define the horizontal and lateral continuity of beds and facies.

The following types of field data were collected for each area and section: 1) lithology; 2) the nature of the upper and lower bounding surfaces; 3) external unit geometry and lateral extent; 4) scale and thickness of units; and 5) sedimentary and biogenic structures (Miall, 1985). A series of distinct marker horizons were identified in the field based on physical and sedimentologic characteristics and these were used to trace outcrops along strike and lateral facies changes. In order to correlate the subsurface geology, core

116 logs and drill records were used to construct a detailed GIS basin model (see below; Fig. 4-

4). Weathered and unweathered colour is based on the GSA Munsell Colour Chart (2009) and the GSA Rock Colour Chart Committee (Goddard et al., 1980).

4.2 Arc GIS Reconstruction Models and Fence Diagram

A 3D visualization of the subsurface within the investigation area of Eromanga

Basin was created combining a suite of digital resources. Principally, this study utilized

ArcGIS 10.1 for digital reconstructions of the basin wide stratigraphic sections and correlations. This study also utilized 3D GIS within Arc GIS 10.1 to better visualize subsurface geology. All digital work was carried out in accordance with the “Working with 3D GIS, Using ArcGIS” (V10.1, 2010). GDA94 is the official geodetic datum adopted nationally across Australia on 1 January 2000. This was the datum used to gain consistency across all raster and vector data displayed. Zone 55 was identified as the most relevant for this portion of the Eromanga Basin. Regional geological maps of the northern and central Eromanga Basin (Rock Unit & Structures) were imported and floated on the digitised topographical surface for the region. The rock units of interest were displayed. Only the fault structures were selected for display. The faults were extruded to provide a pseudo 3D visualization. Both magnetic and gravitational survey raster data were acquired. Three magnetic survey images were considered. This included a 1st vertical derivative regional survey (magmap04) that covered the entire region, a TMI and a higher resolution 1st vertical derivative (GSQP795tmivd1 image) providing a partial coverage. The gravitational material covered a similar expanse as the

TMI. Seismic survey data was selected using the 2007 21 December 2D seismic collector sheet. These results were not displayed within the visualization.

117

Logged drill information from 15 sites (including cores logged as part of this study and additional original GSQ drill logs from surrounding wells) was used to enable stratigraphic surfaces to be interpolated for the Winton and Mackunda formations. To ensure consistent surface interpolations it was necessary to include both overlying and underlying stratigraphic units. The core length applied relative to RL supplied the vertical component (z), with the positions (x,y) input using the latitude and longitude converted to decimal degrees. Top surfaces of each unit were generated using the

Ordinary kriging function. The semivariogram properties used a spherical model to create the surface. The base surface of the Toolebuc Formation was also generated. The function used 12 outer drill core details to create the interpolated surface. The interpolated surface was then used to create a Triangulated Network (TiN).

A traditional block diagram, encompassing all fifteen drill locations was generated using the extrude function. Extrude joins the nominated TiN surfaces and a polygon shape that defines the area of interest. The polygon shape files used to provide the outline for the extrude function were drawn to allow adjustment for faulting. These polygon shapes had to be restricted to slightly inside the outer edge of the TiN to ensure the extruded block created a “closed” shape. The blocks affected by dip-strike faults were adjusted vertically to ensure the fence diagrams would display the faulting appropriately. To create a fence diagram it was necessary to intersect a multipatch (3D) shape with the block diagram. The intersecting multipatch was first drawn as a 2D surface then extruded and converted to a multipatch. The intersect tool generates a multipatch file of the intersection between the block unit and fence multipatch.

118

5. Results

5.1 Sedimentology Outcrop and core based facies analysis of the Winton Formation and the top of the underlying Mackunda Formation resulted in identification of 22 lithofacies (Table 4-1; in text all faces codes are listed in descending order of abundance). The repeated occurrence of lithofacies in distinct combinations, together with the presence of diagnostic architectural elements, was used to identify and interpret nine facies associations (FA) (Fig.

4-5a, b; Table 4-2). A detailed description of all nine facies associations is presented below, and these are interpreted in terms of their depositional environment. Following the approach and codes of Miall (1985, 1996), architectural elements, or lithosomes were identified within different facies associations based on their geometry, associated facies, and scale (Table 4-3). Bounding surfaces were commonly 0th to 4th order; however some

5th order may be inferred, but not strongly supported (Fig. 4-6a,b). Within the study areas, two distinctive sedimentary patterns are observed. First, facies are overall coarser in BNP than in LQCP, and second, in outcrop FAs gradually coarsen up-section in both areas.

Outcrop exposure in IQ is too limited to assess any trends. General coarsening upward trends are also identified in most cores studied and published core logs.

119

Table 4- 1 Lithofacies codes identified in the Winton and Mackunda formations. Modified from Miall, 2010.

120

Table 4- 2 Facies associations identified in in the Winton and Mackunda formations. Modified from Miall, 2010.

121

Figure 4- 5a Typical outcrop patterns of Facies Associations 1-9 in outcrop at both Lark Quarry Conservation Park and Bladensburg National Park: (FA1) massive-bedded sandstone, (FA2) thinly-bedded sandstone, (FA3) massive-bedded siltstone, (FA4) thinly bedded siltstone, (FA5) tuffaceous mudstone, (FA6) thinly bedded protosol, (FA7) interlaminated finely grained sandstone and siltstone, and (FA9) paraconglomerate.

122

Figure 4- 5b Typical patterns of Facies Associations 1-8 in core.

123

Figure 4- 6a Observed hierarchy of bounding surfaces in outcrop of the Winton Formation at Lark Quarry Conservation Park and Bladensburg National Park.

Figure 4- 6b Observed hierarchy of bounding surfaces in core from both the Winton and Mackunda Formation from around the study area.

124

5.1.1 Major Sandstones (FA1)

Description Major sandstones are a moderately common FA in both outcrop and core within the

Winton Formation. FA1 is composed of the following lithofacies: Sp, Sh, St, Se, Si, Sd,

Sm, Fl, Gmm, Gmx, and Sr (Sr is rare/poorly preserved) (Table 4-1). Individual sand bodies range from massive (Sm) with only weak evidence of bedding (primarily due to weathering) to well-bedded with excellent preservation of internal sedimentary structures

(St, Sr, Si, and Sp in particular). In outcrops, FA1 is most common in the middle and upper portions of the Winton Formation, but only moderately abundant in the lower Winton

Formation. In core, FA1 is identified in the middle to upper Winton within Eromanga 1,

Blackall 2, McKinley 1, and Maneroo 1. Interestingly, FA1 is restricted to the very uppermost part of the Mackunda Formation in the Longreach 1-1B core and absent from the Mackunda Formation in Eromanga 1, Blackall 2, McKinley 1, and Maneroo 1.

FA1 units are characterised by sub-angular to sub-rounded, moderately to well- sorted, medium to finely-grained sandstone that commonly fine upward. Less commonly,

FA1 units are characterized by coarse grained sandstones that exhibit less degrees of textural maturity. In rare occurrences, FA1 grades to a very finely-grained siltstone at the uppermost contacts (St, Sr). FA1 represents thick sandstone bodies exceeding ~0.5 m thick; however, most beds are restricted to 0.5- 3.0 m in thickness. FA1 units can extend continuously for a kilometre or more. FA1 is typified by basal 4th- to maybe 5th order surfaces, which commonly form sharp to erosive contacts (more commonly, 4th order surfaces are observed). Where basal 4th order surfaces are present, intraclast conglomerates

(Gmx) are almost always observed in the basal 0.10- 0.30 m of FA1. Internally, 0th to 3rd order, bounding surfaces are common within FA1 sand bodies, with 3rd order surfaces

125 characterized by thin lenses of Gmx. Individual beds within FA1 vary in thickness and typically have multiple erosional to gradational internal bounding surfaces (1st-3rd order).

Individual cross-sets (St, Sp) are commonly between 0.3- 1.3 m thick. Lower bounding surfaces are commonly sharply truncated or erosional.

Within FA1, five different architectural elements are recognized (sensu Miall, 2010,

2014), including channel elements (CH), lateral accretion elements (LA), sandy bed forms

(SB), scour hollows (HO), and laminated sands (LS) (Table 4-3). CH, LA, and SB are the most prevalent and well-preserved at both LQCP and BNP. CH elements range in thickness from 0.8- 2.0 m and extend laterally for 10.0- 40.0 m (sheets can laterally extend for several kilometres or more). LA elements are relatively common and their morphology is typically that of wedges or sheets, though rare lobes and trough-sets are present. LA toe to crest height ranges from 0.5- 2.5 m and their widths generally exceed 5.0 m; however, lack of continuous exposures and poor preservation inhibits precise measurements in most cases. SB is also very common in outcrop, and ranges in thickness between 0.5- ~1.5 m, and typically extends 10’s of meters in lateral extent. SB most commonly occurs in sheets; however, rare lenses and wedges have been identified. Other architectural elements, such as

HO and LS, are present, but rare.

Interpretation FA1 is interpreted to represent fluvial channel deposits. Large-scale macroform elements identified within outcrop, including channel, LA, SB, LS, and CH are all consistent with this interpretation (Fig. 7a,b,c,d) (Allen 1963; Nadon, 1994; Miall, 1996,

Fielding, 2006; Gibling, 2006; Nichols & Fisher, 2007; Miall, 2014). LA elements are interpreted as ancient point bars and scroll bars, whereas the association of SB and CH in

126 certain areas provide some evidence of high-sinuosity fluvial channels with well-defined margins. The complexity and abundance internal bounding surfaces throughout individual

FA1 sand bodies suggest that flow waxed and waned over the course of deposition, as would be expected in fluvial channel system. FA1 units commonly fine upwards, suggesting that much of the bed is composed of migrating point bars, channel lag and

thalweg sediments (Fi elding et al., 1999; Miall, 2014).

Table 4-3 Architectural elements (lithosomes) identified in the Winton Formation outcrops of Lark Quarry Conservation Park and Bladensburg National Park. Modified from Miall, 2010, 2014.

127

Figure 4- 7 Outcrop Facies associations (white), bounding surfaces (black), and fluvial architecture (green) in outcrop at both Lark Quarry Conversation Park and Bladensburg National Park: A) Single in-filled abandoned channel (CH(FF)) bounded by a basal 4th order erosive surface, B) Multistorey FA2 scroll bars, each bounded by erosive 3rd order surface, C) Moderately developed paleosol, D) Basal sheet FA4 overlain by angular bedded course grained FA1, E) Basal tuffaceous mudstone overlain by multistorey FA2.

128

5.1.2 Minor sandstones (FA2)

Description Minor sandstones are a moderately common FA in both outcrop and core within the

Winton Formation. FA2 is composed of the following lithofacies: Sh, Sfv, Sr, Sd, Sm, Si,

St, Si, Fl, Gmm, Gmx and IvB. Generally, individual sand bodies range from massive (Sm) to thinly-bedded with well-preserved sedimentary structures (St, Sr, and Si). In outcrop,

FA2 is most common in the upper Winton Formation; though, only moderately common in the lower Winton Formation. FA2 is identified in all logged cores throughout the Winton

Formation, but this FA is considerably more abundant in the upper Winton Formation than in the lower Winton or Mackunda formations.

FA2 units are characterized by moderate to well sorted, moderately rounded, fine grained sandstones. Coarse-grained sandstones do occur, though not as common and typically exhibit less textural maturity. FA2 represents thinly-bedded sandstone bodies that are less than 0.5 m thick (most typically ~0.02- 0.3 m). Lateral extent of these units is variable, but units can extend for 100’s of meters to several kilometres. Individual lithofacies within FA2 vary in thickness and are characterized by both erosional to gradational internal bounding surfaces (1st-3rd order). Individual trough and planar-cross stratification is generally small scale (<0.20 m sets) and ripple cross lamination is particularly abundant. All FA2 units generally fine-upward and are typified by basal 3rd-4th order surfaces that commonly form sharp to erosive contacts. Where basal 4th order surfaces are present (very rare possible 5th order), intraclast conglomerates (Gmx) are observed in the lower 0.02-0.10 m of these units. Very similar to FA1, internal bounding

129 surfaces are common, ranging from 0th to 3rd order. Internal 3rd order surfaces tend to be defined by minor basal intraclast conglomerates (Gmx).

Interpretation FA2 is interpreted to be associated with low-energy fluvial channel infill, crevasse splays, sheet floods, levees, and waning flow with minor bed- load to suspension settling

(Allen 1963; Miall, 1996; Fielding, 2006, Nichols & Fisher, 2007). Strong lateral correlation of these units can in certain cases be used to identify lateral relationships between minor sandstone bodies (FA2) hosted within floodplain fines that are adjacent to major sandstone bodies (FA1), confirming their precise origin as crevasse splay deposits.

The presence of upward fining and waning flow features, in combination with CH elements in a handful of FA2 units is interpreted to represent a waning flow stage and channel filling of secondary or tertiary tributary or auxiliary fluvial channels. They may potentially be records of inactive, infilled channels within larger scale fluvial channel belts (Fig. 4-7b, c, d; 8) (Nichols & Fisher, 2007; Miall, 2014). In some cases, these units also likely represent isolated sheet flood deposits.

130

Figure 4- 8 Observed facies associations in the Lower Winton Formation with outlined stratigraphic section of Hades Hill at Lark Quarry Conservation Park: A) Uppermost section of the lower exposed Winton Formation with interbedding of FA4 and multistory FA2, B) Middle lower exposures with basal interbedded siltstone, C) Middle lower exposed massive bedded FA5 tuffaceous mudstone.

5.1.3 Major Siltstones (FA3) Description

131

Major siltstones are a common FA in both outcrop and core within the Winton

Formation. FA3 is commonly composed of the following lithofacies: Sh, Sr, Sm, Sd, Se,

Sfv, Ss, Ssc, Fr, Fl, Flf, Fsc, C, IvB, P, Pb and Gmx. FA3 commonly forms minor cliffs where exposures are preserved and commonly exhibits spheroidal in altered strata. In core,

FA3 is also identified throughout the upper and Lower Winton Formation in Eromanga 1,

Blackall 2, McKinlay 1, and Maneroo 1. This facies is limited to the very uppermost

Mackunda Formation in Maneroo 1 and McKinlay 1, and common in Longreach 1-1B.

Most FA3 units are characterised by moderate to well-rounded, well-sorted sandy siltstones. Siltstones at both BNP and LQCP are commonly moderately– to well-cemented.

FA3 units are generally coarsely-grained to finely-grained siltstone and occasional mudstone; altered units do not preserve grading characteristics in the uppermost Winton

Formation. FA3 bodies can exceed 4.0 m; however, most beds are restricted to 0.5- 3.0 m.

FA 3 is typified by basal 2nd to 3rd order surfaces which commonly range from gradational, sharp, to irregular-eroded contacts. Third-order surfaces are present in two distinct morphologies, immature paleosols and channelized structures. Third order surfaces are also commonly characterised by thin intraformational conglomerate drapes; though these may also be granular to coarse sand (Gmx). Individual facies within FA3 range in thickness with sharp-erosional to gradational internal bedding surfaces (1st-3rd order).

Individual cross sets (St, Sp, Sr) are commonly between 0.04- 2.0 m thick. Lower

132 bounding surfaces are commonly sharply truncated to erosive with rip-up clasts, dominantly clay-mud rich. Bioturbation structures are commonly preserved (P, IvB, Pb).

Soft-sediment deformation and dewatering structures are also present, though preservation is poor.

Within FA3, five different architectural elements are recognised including channel elements (CH), lateral-accretion macroforms (LA), laminated sand sheets (LS), levee (LV), and floodplain fines (FF). CH, LS, and FF are the most prevalent and well preserved at both LQCP and BNP. CH presents in sheets that range in thickness from 0.5-4.0 m in thickness and extend laterally for at most a kilometre; though, units generally extend for

200- 300 m. LA morphology is typically sheet-like, with minor variation to thickness over many meters (50- 100 m); however, variable internal preservation inhibits complete description. LA toe to crest height ranges from 0.5- 1.5 m and widths are generally smaller, between 3.0-6.0 m. LS are present, though preservation quality over lateral distance is variable and commonly discontinuous. Thickness ranges between 0.03-0.5 m, and can range laterally between 100-150 m. LV is identified in upper and lower sections of the exposed Winton Formation and commonly ranges in thickness between 2.0- 4.0 m in thickness and between 200- 300 m (variable to outcrop quality).

Interpretation

133

FA3 is interpreted to represent very low energy fluvial channel and proximal floodplain deposits based on their physical and structural characteristics (Fielding, 2006;

Miall, 2014). In several cases, lower bounding surfaces present erosive contacts and rip up clasts interpreted as low energy sheet floods of active channel flow (Miall, 1996, 2014).

FA3 is commonly overlain by finer sediments which commonly exhibit well preserved plant remains, paleosol and bioturbation structures, indicating channel migration and shortly thereafter bank- levee development (Allen 1963; Miall, 1996). Lateral correlation of these beds suggests that some of these events were long-lived in broad sweeping fluvial dominantd flood plains.

5.1.4. Minor Siltstones (FA4)

Description Minor siltstones are a common FA in both outcrop and core throughout the Winton

Formation. FA4 is commonly composed of the following lithofacies: Sm, Sh, Fm, St, Sd,

Sr, Fl, Fm, Fr, Tlcf, P, Pb, Gmx, and IvB (Table 4-1). Individual silt bodies range from massive and void of internal structure (or poorly preserved) to well-developed and well- preserved internal bedding and sedimentary structures (St, Sr, IvB). In outcrop at all study areas, FA4 is identified in both upper and lower exposures of the Winton Formation. In cores Eromanga 1, Blackall 2, McKinley 1, and Maneroo 1, FA4 is identified in both upper and lower stratigraphic sections. This facies is also abundant in the Mackunda Formation in cores McKinley 1, Blackall 2, Eromanga 1, Longreach 1-1B and Maneroo 1.

Most FA4 unites are characterised by moderate to well-rounded grain that commonly show an overall fining upward pattern into very finely-grained siltstone to very silty mudstones (St, Sr). Less commonly, FA4 units can be characterised by coarse grained

134 silts which are commonly grain supported, and poorly consolidated. Overall, FA4 represents thin siltstone bodies which do not exceed a thickness of 0.5 m, with most units ranging between ~0.01- 0.3 m. Lateral extent in FA4 is highly variable upon outcrop quality. In some cases FA4 can be traced for half a kilometre; however, in most cases FA4 extends between 100-300 m. FA4 is typified by basal 0th to 3rd order surfaces which commonly exhibits both sharp and erosive contacts. Basal 3rd order surfaces exhibit both erosive surfaces including coarse sand to granular intraclasts (Gmx). Upper surfaces are commonly 1st to 3rd order, and many are gradational in nature. Internally, 0th to 3rd order surfaces are common within FA4 sand bodies. Commonly, internally sharp contacts of 1st to 2nd order are presented in thinly bedded lamina (0.02- 0.10 m). Individual cross-sets (St,

Sp, Sr) are commonly between 0.4 and 1.5 m thick. Lower bounding surfaces are commonly sharply truncated and erosional. Lowermost inclined beds are commonly courser with sand, mud or clay-clasts inclusions and generally fine-upwards. Thicker units containing higher percentages of fines (silts/muds), and more often contain moderately to well-preserved silty soil (protosol) and bioturbation structures (Sm, P, Pb, and IvB). In

FA4, also preserved vertical and bifurcating roots are within moderate to well-preserved soil horizons.

Interpretation FA4 is interpreted to be associated with fluvial channels and overbank fine deposits based on their physical and structural characteristics (Allen 1963; Miall, 1996, 2014).

Smaller scale to isolated singular events includes minor sheet floods, abandoned channel infill, levee development, and suspension settling along with weakly to moderately- developed rooted paleosol horizons (Fig. 4-7c). Moderate lateral continuity of these beds

135 suggests that some of these events were well to moderate developed channel sets with associated levee development.

5.1.5. Major Mudstones (FA5)

Description Major mudstones are a moderately common FA in both outcrop and core within the

Winton Formation. FA5 is commonly composed of the following lithofacies: Ssc, Fm, Fl,

Flf, Fr, Fsc, VFc, IvB, C, PB and P (PB and P are infrequent). Individual mud bodies range from massive (Fm) with only weak bedding to well-bedded with excellent preservation of sedimentary structures to units that are moderate (Ssc, Fl, Fsc, VFc) to highly kaolinized- lateritized. In outcrop at all study areas, FA1 is rarely present in the upper sections of the

Winton Formation and moderately common in the lower Winton Formation. In core, FA5 is identified in all sections of the Winton Formation in most cores. This facies is very common in the upper and middle Mackunda Formation cores too.

FA5 is common and most beds restricted to between 0.8-2.0 m in thickness, but with some units exceeding 3 m thick. Characterisation of lateral extent in FA5 is variable upon the bedding morphology and outcrop quality. Lateral continuity can extend for several hundred meters to several kilometres. FA5 is typified by basal 0th to 3rd order surfaces that can have gradational, sharp, or irregular contacts. Lower contacts are commonly expressed by 2nd or 3rd order surfaces. Internally, many units exhibit 0th-1st order bounding surfaces with upper contacts displaying erosions to irregular. FA5 occurs in three distinctive lithotypes: 1) moderate to highly kaolinized-lateritized units that commonly erode into small cubic ped structures; 2) bentonitic, very clay rich units that exhibit haystack to popcorn weathering; and 3) highly bioturbated units with evidence of

136 colour mottling and development of colour horizons. In the lithtype, blocky ped structures and small slickensides are also common, and the trace fossils are sometimes in-filled with coarser silts and sands and secondary iron cements and small concretions.

Architectural Elements identified within FA5 include, in descending order of abundance: FF, LV, and CH(FF). FF is typically observed in thick to massive bedded units, which range 0.7-4.0 m, and laterally extended for kilometres. LV is commonly observed in medium to thick beds which taper to a pinch out. Thickness of LV can range from 0.8-3.2 m, and laterally extend for 200- 500 m. CH(FF) macroform elements are a very distinctive feature associated with FA5 in outcrop. FA5 units with CH(FF) marcro form elements have silty to sandy bases above erosional contacts and are most significant in that preserve large quantities of well-preserved plant macrofossils (typically leaves). The

CH (FF) elements are small, typically only 1.5- 2.0 m thick and taper to a distinct pinch- out, ranging about 30-80 m wide.

Interpretation FA5 is interpreted to reflect deposition within low-energy proximal to distal floodplain settings (Allen 1963; Miall, 1996; Ghosh et al., 2006). A range of depositional environments are recognized for FA5 units, including ancient channel levee deposits and small secondary channel fill deposits in the case of FA5 units with CH(FF) macroform elements. Many FA5 units can be interpreted as pedogenically modified flood plain deposits, or paleosols, based on the presence of such features as ped structures, slickensides, mottling, colour banding, small Fe-oxide staining and concretions, and the common presence of bioturbation, including clear root traces. The combination of these and physical features indicate that most paleosols are only weakly developed, fitting the

137 classification of ‘protosols’ (Brown and Kraus, 1987; Retallack, 1997; Demko et al., 2004;

Ghosh et al., 2006). FA5 units with distinct haystack and popcorn weathering features are interpreted as swelling bentonitic mudstones that imply the presence of devitrified volcanic ashes (Roberts and Hendrix, 2000). However, several of these units were collected and separated to look for datable volcanic phenocrysts. In both cases, the coarse grained separates included abundant detrital sediments and dating of the zircon separates indicated that these are not primary ashes, but more likely developed as a result of weathering of volcanic-lithic rich sandstones and siltstones. No evidence of primary volcanic ash beds could be identified, although considerable effort was made to determine if such beds exist.

5.1.6. Minor Mudstones (FA6)

Description Minor mudstones are a common occurring FA in all exposures and cores of the

Winton Formation. FA6 is commonly composed of the following lithofacies: Fm, Fl, Flf,

Fr, Fsc, VFc, IvB, PB, C, P and Ssc. Individual thinly bedded mudstone units range from medium bedded (Fm) to very well thinly laminated. In outcrop at all study areas, FA6 is exposed in both lower and upper sections in varying frequency. In core, FA6 is identified in the middle to lower sections of the Winton Formation. FA6 is infrequent in the upper most units. This facies is very common in the cores that also preserve the upper and middle

Mackunda Formation.

FA6 represents thinly-bedded mudstone bodies that do not exceed 0.5 m thick, with most beds ranging between 0.05-0.3 m in thickness. Thicker units (0.2-0.5 m) tend to partially preserve fine to medium scale lamina (Sm, Fl); however, many units are void of internal sedimentary structure. Laterally continuous units can extend for many kilometres;

138 however, most commonly extend for 200 to 400 m. This FA is better preserved in the lower to middle exposures, primarily because upper exposures are deeply weathered and highly altered. FA6 is typified by basal 0th to 3rd order surfaces with gradational to sharp contacts. Lower bounding surfaces are commonly sharply truncated, flat lying and uniform.

Bioturbation in FA6 is common (IvB, PB, and P), and facies with rare mud cracks (Fm, Fl) are present.

Interpretation FA6 is interpreted to reflect deposition within low-energy floodplain settings.

Some FA6 units with evidence of mudcracks and bioturbation features, are interpreted be very weakly developed paleosols. In several outcrop sections of preserved FA6, indications of various levels of pedogenesis and stability, with some more intensely developed than others (Fig. 4-7c).

5.1.7. Interlaminated sandy siltstones to silty mudstones (FA7)

Description Interlaminated sandy siltstones to silty mudstones are moderately common throughout the outcropping Winton Formation. FA7 is commonly composed of the following lithofacies: Ssc, Sr, Fm, Fr, Fl, IvB, Pb, C, and P. FA7 tends to preserve fine to medium scale, flat-lying lamina (Sm, Fl, Fr, Fm, Ssc). Individual couplets are tightly packed with fairly continuous upper and lower bounding surfaces. In several instances however, planar and ripple cross laminations are observed, along with abundant, small- scale dewatering features. Bioturbation is also common. In outcrops, at all study areas,

FA7 is most commonly exposed in middle and upper sections of the Winton Formation. In core, FA5 is identified in lower sections of the Winton Formation. This facies is very common in cores that preserve the upper and middle Mackunda Formation.

139

FA7 units commonly range in outcrop from 0.1- 3.0 m in thickness and are laterally continuous for up to a kilometre at most; however, most only extend for a few hundred meters. Coarser laminae commonly exhibit sharp to erosive basal surfaces (2nd-3rd order surfaces; Gmx). Internal bounding surfaces are 1st to 2nd order and range from gradational to sharp boundariesIn core, two distinctive morphologies are identified: 1) tightly packed rythmite (lenticular bedding with 0.0005-0.01 m sets) beds that are 0.10-5.0 m-thick, restricted to the Mackunda Formation and 2) Thicker, flaser- bedded, interlaminated siltstones and mudstone couplets (that reach up to 8 m thick in total) and are restricted to the Mackunda and lowermost Winton Formation. Very little evidence of FA7 is found in the middle or upper Winton Formation.

Interpretation FA7 and associated sedimentary structures indicate minor cycles of increased and decreased flow velocities within tidal flats and tidally influenced fluvial channels that are most likely associated with middle to upper delta plain settings. This FA is almost exclusively restricted to the Mackunda and Lower Winton Formation; however, rare examples of FA7 are found alluvial upper Winton Formation outcrops. In these rare cases,

FA7 is most likely to have been generated by a secondary or tertiary fluvial channel with waxing and waning flow velocities. Association of ripples, vertebrate tracks, and invertebrate bioturbation markers indicate periods of repeated ponding in inactive fluvial channels, closely followed by minor stream flow (Thulborn and Wade, 1984; Buatois and

Mangano, 2011; Romilio et al., 2013). In core, nearly all FA7 units are restricted to the

Makunda and lower Winton Formation and are interpreted to have been produced in marginal marine, coastal, deltaic, and very distal tidally-influenced fluvial channels (Pontén

& Björklund, 2007). These tightly packaged units are interpreted as cyclical tidal

140 rhythmites and non-cyclical couplets of flaser, wavy, and lenticular bedding. An upsection progression from mudstone-dominantd lenticular bedding to sandstone and siltstone dominantd flaser bedding is interpreted to represent a transition from tidal flat and lower delta plain to middle and upper delta plain settings with a tidal marsh to deltaic setting.

5.1.8. Plant-rich or Carbonaceous silty-Mudstone to Coal (FA8)

Description

Description Carbonaceous siltstones, mudstones, and thin coals comprise this FA in both outcrop and core within the Winton Formation. FA8 is commonly composed of the following lithofacies: Ssc, Fl, Fsc, C, PB, and P. FA9 is ranges from thin- to thick-bedded siltstone and mudstone units that have moderate to very high amounts of carbonized or whole plant material and coalified fragments (Fl, Ssc, Fsc, Fcf). In core, coal grade commonly ranges from peat to lignite, although in rare occasions subbituminous coal is present. In outcrop, plant material is very well-preserved and readily identified.

Individual horizons in outcrop range from 0.05 to 4.0 m in thickness; however, with most beds range from 0.10 to 2.0 m. Lateral extent of FA8 units is variable, extending upto 100-150 m. In core, FA8 does not 8.0 m in overall thickness and averages between

1.5 -5.0 m. FA8 is typified by upper and lower 0th to 2nd order bounding surfaces, which commonly exhibit sharp contacts. Internally, weakly developed 0th to 1st order laminae are common, defined by plant hashes and well-developed coalified fragments along boundaries. In core, carbonaceous fragments range from isolated fragments (upper

Winton) to highly disseminated (lower Winton Formation).

141

In outcrop, FA8 is only exposed in the lowermost portion of the Upper Winton

Formation. Ch(FF) macroform elements are an important feature associated with the rare

FA9 units observed in outcrop. In core, FA8 is identified in both the middle and lower

Winton Formation This facies is absent in all but the uppermost Mackunda Formation in core, where it is very common.

Interpretation

FA8 is interpreted to represent very low energy anoxic to disoxic accumulations of plant material within abandoned channels, oxbow’s, billabongs and low-lying ponds and swamps (Allen 1963; Fielding, 1985, 1987; Davies-Vollum and Wing, 1998). This FA is most common in the uppermost Mackunda and lower Winton formations and most likely associated with coastal mires and oxbow lakes, and backswamp settings to upper delta plain settings. Thicker sections of peat to sub-grade coal are interpreted to have been generated in swamp accumulations adjacent to back swamps to very distal fluvial environments (Kraus & Aslan, 1993, Wing 1998). Many FA8 units occur in FA8 units with CH(FF) macroform elements that are interpreted to represent oxbow swamps that formed inabandoned channels (Fig. 4-7a). However, isolated non-lenticular horizons are likely to have been ponds, minor swamps, or even plant and mud rich flood debris.

5.1.9. Intraformational Conglomerate (FA9)

Description Intraformational conglomerate is a rare FA in both outcrop and core within the

Winton Formation. FA9 dominantly matrix supported paraconglomerate composed of the following lithofacies: Gmm, Gmx, Sh, and Ss. In outcrop, at all study areas, FA9 is restricted to the uppermost Winton Formation and in many weathered outcrops FA9 forms

142 slump structures in the uppermost sections of many “Jump Ups”. In core, FA9 is isolated to a handful or examples, none of which exceed a thickness of 0.30 m. This FA is not identified in the Mackunda Formation.

FA9 represents an unstratified unit that is dominantd by matrix supported pebble to cobble sized clasts commonly composed of Intraformational mud and silt. Matrix varies from very coarse sand to very fine silt, which forms a support matrix for various sized clast inclusions (0.01-0.30 m). In outcrop FA9 units are commonly about 1.0m in thickness, and some units can be traced laterally for quite some distance in LQCP and BNP. Beds are moderately continuous but thickness varies laterally to form discontinuous lense-like to sheet-like geometries. Basal bounding surfaces are erosionally incised 4th order surfaces, which are very irregular in nature. Upper surfaces are commonly weakly gradational into overlying FA and can exhibit sharp upper surfaces. FA 9 units are void of any internal structure.

Interpretation The matrix supported nature of these conglomerates, composed of intraformation mudstone, siltstone and sandstone clasts, suggests that these deposits represent minor, in channel bank collapse events in some cases and small, localized debris flows in other cases.

The lateral continuity of some of these beds indicate dilute debris flows that maintained matrix support over fairly long run-out distances prior to frictionally freezing. Several distinct horizons observed at both LQCP and BNP may represent minor flooding events.

Upper bounding surfaces, which are gradations in nature, indicate a return to normal stream flow conditions. It is perhaps possible that these units are associated with periods of intense volcanic eruptions to the east in the Whitsundays Volcanic Province, and represent

143 the distal expression of in channel lahars. However, it is perhaps, more likely that these are simply the result of particularly intense flooding events.

5.2 Fossil preservation and distribution in the Winton Formation

5.2.1 Trace Fossils In outcrop, facies assemblages FA2-FA9 preserve a broad suite of trace fossils of varying diversity, quality, and intensity. Bann et al. (2008) constructed a hierarchy

(bioturbation index [BI]) for classifying the intensity of bioturbation in a given sedimentary horizon from BI 0 (none) to BI 6 (intense). Following the criteria established by Bann et al.

(2008), a general characterization of the BI for different FAs of the Winton Formation was performed. FA1 units typically show a BI of 0-1, though 1 is very rare and restricted to finely-rained units. FA2 units typically record BI values of 0-2. FA3 typically records BI values of 2; however, a several horizons have been identified with BI up to 3. FA4 exhibit the broad range of bioturbation intensity (0 to 4). FA5 and FA6 either present bioturbation and commonly a BI of ~3 (and commonly silty infill) or do not preserve at all. FA7, 8, and

9 commonly range around a BI of 2, however some fluctuation is noted and though only roughly this may be classified into a BI of 3.

Many of the FAs that preserve diverse trace fossil assemblages are interpreted to have formed within stable long-lived and well-developed horizons. The primary ichnotaxa indentified throughout the Winton Formation are Skolithos, Scoyenia, and Planolities. The association of these particular trace fossils together is characteristic of the Scoyenia ichnofacies (Buatois and Mangano, 1995). The Scoyenia ichnofacies has been interpreted to represent a continental environment with low-energy flow and periodic subaerial conditions

(MacEachern et al., 2010). This is consistent with sedimentological interpretations

144 presented above for the Winton Formation, which indicate fluvial systems with well- defined channel banks and point bars (FA1-4), along with incipient paleosol development in many units (FA3-6, 8-9).

5.2.2 Vertebrate Fossils Preservation styles of vertebrate fossil remains described from Isisford, Queensland are unique to all other described from the Winton Formation. The Isisford assemblage is interpreted to be entombed within the lower Winton Formation (Tucker et al., 2013; 2014).

Skeletal remains are well-preserved and commonly skeletons are associated and articulated.

High quality of preservation is a direct result of skeletal material being preserved in iron cemented-volcanilithic-rich sandstone nodules. These nodules are commonly preserved within FA1 and FA3. Fossil remains include Isisfordia duncani, a Eusuchia

Crocodyliform, along with a newly identified Ichthyodectiform fish and a newly discovered dinosaurian (archosaur) taxa (Salisbury et al., 2006a). A direct result of these assemblage patterns and lack of any outcrop has greatly inhibited their description. For the first time, this locality can be interpreted based on the temporal placement of its entombed fossil assemblage (Chapter 2), and coupled with a better understanding not only the source of the entombing sediment (Chapter 3) and the facies which directly affected the emplacement on preservation of said fossil accumulations (this chapter).

On the other hand, many of the more advanced dinosaurian forms are described from the upper Winton Formation at either LQCP or BNP (Salisbury et al., 2006b). These associated to isolated remains are of poorer quality when directly compared to those described from Isisford. Elements found at the surface are commonly weathered and fragmentary; though preservation quality greatly increases with burial. This is especially

145 true for a minor multi-individual bonebed at BNP. Commonly, these isolated-associated elements are preserved in coarsely grained FA3-FA4, though some do occur in finely grained FA1 and FA2. Presently, dinosaurian taxa associated with these upper horizons are the very fragmentary remains of the theropod (Saurischia, Dinosauria), Australovenator wintonensis, and the fragmentary remains of several sauropods (saurischia, dinosauria),

Diamantinasaurus matildae and Wintonotitan wattsi (Hocknull et al., 2009), along with the newly discovered yet undescribed material by Salisbury and crew.

5.3 Stratigraphy In both BNP and LQCP surface exposures are spotty at best. Many possible surface exposures are covered by ‘black-soil’ and Quaternary Alluvium with scattered exposures in low-lying depressions, drainage channels, and dry modern fluvial channels (Fig. 4-4, 4-8,

4-9, 4-10, 4-11). ‘jump-up’ exposure ranges ~14- 36 m at LQCP, and similarly at BNP.

Consistently, each ‘jump-up’ utilised could be separated into three vertical subgroups based on levels of alteration (Fig. 9). Unit A, the base unit has little to no alterations, and contained the best preserved sedimentary structures in any of these unites. Typically, Unit

A was about 12-16 m in thickness and laterally continuous beds can stretch for several kilometres. Individual units within Unit A are overall fine grained then overlying units.

Bedding and architectural notes are readily identifiable, and facies are easily diagnosed.

Unit B preserves many structural and sedimentological features though is characterised by moderate levels of physical and chemical alteration. Depending on the level of surficial weathering, Unit B can range between 3-8 m. Preservation of individual and associated sedimentary features or architectural elements is moderate to spotty at best. Large scale architectural elements are rare and often poorly preserved or isolated. Moderate to higher degrees of weathering is identified by the intensity of associated spheroidal weathering,

146 which can be very pervasive. Unit C, the uppermost unit is characterized by being greatly altered and most often highly weathered. Unit C generally forms 1-4 m thick cliff-like cap rock to all of the observed ‘jump ups’. Unit C is only laterally continuous though out the study area; however, correlation of individual beds can be limited to 500 m to a maximum of 1-2 kms. Many of these units have been altered to a laterite, and many compositional characteristics are hard to discern. However, many of these units preserve medium to fine scale sedimentary features and associated trace fossils.

Figure 4- 9 A) Displays Units A B and C in vertical section of Monkey Head Hill in Lark Quarry Conservation Park. B) Simplified cross section across Lark Quarry Conservation Park displaying Units A, B and C in vertical schematic.

Stratigraphic relationships are somewhat troublesome between LQCP and BNP.

Based solely on average strike and dip (n<1º NE; in agreement with Senior et al., 1978),

LQCP would be slightly higher in stratigraphic relationship compared to BNP (Fig. 4-10).

Stratigraphically, key marker beds contain floral, faunal and dated detrital minerals. It is also interpreted that stratigraphic sections are slightly higher and younger at LQCP then those at BNP (Tucker et al., 2013; Tucker et al., 2014). However, many studies are indicating this basin is highly faulted, indicating that lateral correlation should be rigorously tested. Also, Tucker et al. (2014) and this study find a great deal of evidence to illustrate a depositional hierarchy between BNP and LQCP.

147

Figure 4- 10 ArcGIS based cross section from Lark Quarry Conservation Park to Bladensburg Nation Park of key correlative bed of the last occurring tuffaceous FA5 mudstone with extrapolated regional dip of ~1º.

5.3.1 Outcrop Stratigraphy In LQCP, the first and lowermost exposed sections of the Winton Formation are best exposed and the most unaltered in Hades Hill (~ 13.5 m; Unit A). Features identified in Hades Hill are observed all other Jump Ups or surface exposures in LQCP (Fig. 8a-c,

11). In BNP, units are inherently very similar if not identical; however, quality of outcrop is a great deal less than LQCP. It should be noted, these lower units contain overall higher percentages of muds and clays as support matrix than any other overlying unit. Commonly, basal units are composed of repetitive and interbedded packages of FA2-FA6, with minor occurrences of FA1, FA5, GA6 or FA8. FA2 occurs in eight distinctive successive step- like beds which are commonly 0.4-1.3 m in thickness. In-between each FA2 is groupings of interbedded and repeating succession of FA3-FA6 with very infrequent beds of FA8

(FA5 and FA6 contains high percentages of silt progressing up-section).

Several major, but isolated lenticular bodies preserve FA8; at its thickest part ranges between 0.8-1.0 m and commonly pinches-out. FA8 is interpreted to be in-filled

148 abandoned channels with preserved plant remains and trace fossils preserved therein. FA2,

FA4 and FA6 average between 0.02-0.4 m in thickness. The common association of FA1,

FA2, and FA4 is interpreted to be in-channel to directly adjacent to actively migrating channels. Stacked sections of FA5 and FA6 (sometimes interbedded with FA4), which exhibit both paleosol and bioturbation characteristics likely represents minor bank and levee development to well-developed and long-lived bank (Fig. 8b,c). Both cross bedding

(FA2 & FA4) and very small-scale ripple lamination reflects in-channel activity is common in FA1-FA4, and is interpreted to represent channel fill. The ratio between channel/overbank percentages is estimated between 80%/20%; however, this may be erroneous considering local preservation quality. Palaeocurrent data based on in-channel features and averaged in a westerly orientation (335º-334º).

In both LQCP and BNP, the exposed middle sections (Unit B) are moderately altered, though preserved both sedimentary and architectural elements (Fig. 9a-h). The amount of alteration can readily be addressed; a higher presence of spheroidal weathering structures is directly indicative of higher amounts of weathered/altered feldspar. The contact of units A and B in outcrop is quite distinct; the uppermost stratum in Unit A is either a fine grained FA4 or a silty FA6 with either is noted as a gradational or an erosive contact with the very lowermost stratum of Unit B. As many of these upper units seem to have a great deal less clays and muds within FA1-FA4 its contact is likely a result of digenetic iron saturation and porosity levels (Fig. 9a). Unit B, throughout LQCP is similar in many ways to that of unit A; however, FA1 and FA3 have a great deal more presence.

FA1-FA9 is present in various forms and this unit is partially describable in each ‘jump-up’ at LQCP and also BNP.

149

There is a distinct lack of muds FA5-6 in the upper section of Unit B at both LQCP and BNP, where a distinct thickening of FA1-4 is observed. FA5 and FA6 are rare, but if present likely represent waning channel floods or short-term levee development. Overall, better quality of outcrop is at LQCP; however, large-scale architectural notes are better preserved in BNP. This includes many channelized elements (point bars and scroll bars) in

FA1-FA4. Interlaminated finely grains sandstones and siltstone (FA7) are interpreted to represents saturated sediment deposited during waxing and waning flow periods (Fig. 9d,e).

Uppermost sections of Unit B, thicker units of FA1 occur along with fine grained and FA3 and FA4 which commonly preserve numerous palaeosol indicators, bioturbation, and root traces (Fig. 4-9). This is interpreted to represent that upper units in the northwest likely experienced channel migration and therein developed into a moderately to poorly drained floodplain directly adjacent to the channels. On the other hand, other upper sections of Unit B are dominantd by FA1, FA2, FA7, and capped by FA9. FA1 is a great deal coarser and occasionally grain supported. There is a distinctive loss of supporting clays and muds, though some very fine sands and silts are present. The increase of overall grain size strongly indicates a rise in fluvial energy from lower channelized structures identified in the underlying Unit A. FA2 is present in much thinner horizons and likely to represent minor sheet floods within individual channel(s). FA7 is best preserved in these sections, indicating a repetitive waxing and waning of flow within these channels. The upper most section of Unit B in many sections is a thick-bedded paraconglomerate (FA9) with a highly erosive basal surface. Spheroidal weathering is present in varying degrees in

FA1-FA4 in upper portions of Unit 2. However, secondary digenetic iron has somewhat aided in the preservation of biogenic features within these altered units.

150

Figure 4- 11 Facies associations observed in outcrop of the upper Winton Formation: Section type 1 dominantly contains poorly to moderately developed palaeosols to protosols (9a,b,c). These stable and developing horizons preserve a broad suite of soil, floral and bioturbation indicators (9a,b), incised banks (9c), which commonly form short cliff-like structures (9d). Interlaminated FA7’s occur regularly and are interpreted to be erosive waxing and waning periods (9e), On the other hand, Section type 2 is typified by lower bank and levee deposits which are then incised by a very active group of channel coarse grained cross bedded sandstones (9f,g,h).

151

In LQCP, the uppermost section of Unit C are altered, though preserves sedimentary and architectural elements (1.0-4.0 m). At BNP, the uppermost sections are highly alerted, with most commonly forming lateritized to heavily kaolinized caprocks.

The contact of unit B and C is gradational to erosive, though this is highly variable to preservation quality. Much of Unit C could be described as lateritic to extremely kaolinized. Many are weathered into a series of heavily cubic fracturing beds averaging about 0.5-1.0m. If beds are stacked, the lowermost is still fairly competent, however progressing up-section, beds quickly become heavily fractured and disjointed. As with

Unit 2, the same east and west (east-fluvial channel and west-floodplain) separation of environmental indicators and sediment patterns are recognised in Unit 3 and interpreted as such. Units at BNP preserve less sedimentary features and structures then those at LQCP.

In sections at Hidden Hill and Dingo Den, horizons are craggy and commonly form short cliff-like structures with no discernable features.

5.3.2 Core Stratigraphy By utilizing preserved stratigraphic core logs from around that basin, a much better understanding of the Winton as a whole was achieved (Fig. 4-10a-d). The first occurrence of the Winton Formation is at surface in all cores except for that of GSQ Longreach 1-1B.

All cores preserving various portions of the Winton Formation that were re-reviewed by this study include GSQ cores; Eromanga 1, Blackall 2, McKinley 1, Maneroo 1, and

Longreach 1-1B. All GSQ Cores preserve the underlying Mackunda Formation.

152

Figure 4- 12 Facies associations observed in cored sections of the Winton and Mackunda Formation: A) upper Winton Formation sections dominantd by massively bedded FA1 with 2nd to 3rd, B) The lower Winton contains a great deal more heterolithic lamina and bedding dominantly silty in nature, C) The conformable contact between the Winton and Mackunda Formation is identified by the first occurring coal measure, D) The uppermost Mackunda is distinctly silt and mud dominantd with very thick beds of tidal rhythmites and deformed lenticular lamina.

In the middle and lower Mackunda Formation, a distinct lack of all sands is noted

(Fig. 4-10d). Bedding becomes dominantd by well-preserved lenticular and tidal rhythmite sequences. Numerous normal and reverse grading was identified throughout these sections of interbedded clay rich muds and very fine siltstones. It is likely that many of these 153 deposits in the upper to middle Mackunda indicate a transition from the delta front to the pro delta (Bhattacharya, 2010). This section of core is also the lowermost occurrence of calcite and other evaporites minerals within the strata. These minerals can occur in discrete lamina (during deposition) or as crack and fissure infill (post-deposition). In core there was no obvious identified marine biogenic trace, however there were many horizons which displayed deformation as a result of dewatering in multiple sequences of lamina. It is interpreted that the upper Mackunda preserves a period of deposition which was dominantly shallow marine, coastal to the mouth of a fluvial dominantd delta (lower-upper delta plain) (Bhattacharya and Walker, 1992; Helland-Hansen and Gjellberg, 1994;

Bhattacharya, 2010). The upper Mackunda is much the same; however, laminae become a great deal more heterolithic, shifting from flaser to wavy or wavy to lenticular mixed laminae. This greatly signifies mixed energies as these areas transitioned from delta-front- coastal to upper delta plain to the lower delta plain in the very lowermost Winton

Formation.

The contact between the lowermost Winton Formation and uppermost Mackunda

Formation is diagnosed by the lowermost major coal seam in the Mackunda (or last occurring coal in the Winton Formation) (Fig. 4-12c), which this study agrees with (Draper,

2002). In many cores, it was observed that this first occurrence of coal was often accompanied by the last occurrence of marine oyster shells and shell hash. A third characteristic of this contact, though very gradual, is a general fining of sediment down section and an increase of muds. High degrees of coalified plant debris likely indicate the local to regional development of swamps, wetlands or coastal woodlands. Disseminated to isolated and fragmentary coalified flakes indicate minor incorporation of floral remains but

154 more likely indicates deposition within adjacent to swamps and mangroves within the upper to lower delta plain.

Within the the lower Winton Formation, the first distinctive occurrence of FA5,

FA7, FA8 and loss of FA 1 and a dramatic thinning of FA2 to be no thicker than 0.8m (Fig.

4-12a,b). Interpreted facies transition from dominantly terrestrial environments to very tidally influenced distal channels to tidal, backswamp, and coastal environments. Major bounding surfaces transition from 2nd- 3rd order surfaces. Internally these units dominantly contain 1st to 2nd order surfaces. However, prevalent sedimentary structures still indicate fluvial in-channel structures (ripple cross-laminations, low-angle cross stratification, and trough cross cutting). These sedimentary structures are commonly accompanied by mud drapes and settling fines which are interbedded. The abundance of these minor mud lenses quickly increases progressing down hole. The occurrence of these minor lenses likely occurred in association with fluvial environments events such as floods and levee bank development or in adjacent-to fluvial environments.

The uppermost beds of the Winton Formation (n<30.0m-40.0m) in core are commonly very coarse- to finely-grained sandstones and interbedded coarse siltstones,

FA1-FA4 (Fig. 4-10a). The upper Winton Formation is herein defined by the greater amount of terrestrial facies. Both macro and micro-scale sedimentary features were identified in core; however, the lateral extent of these features is at best inferred.

Architecture notes preserved in core include large and very small-scale ripple lamination reflects which is interpreted to be in-channel activity. Very minor interlaminated, very fine silts and very fine sands are interpreted as thinly bedded (n<2.0m) waning channel floods, along with very minor levee development. The ratio between channel/overbank

155 percentages is estimated at 90%/10%. This depositional pattern is moderately to fairly continuous for the first 20.0-30.0m. In underlying sections, there is a loss of any sandy units, signifying a complete loss of FA1 and FA2.

6. Discussion

6.1 Depositional environments Previous studies have interpreted the mid-Cretaceous portion of the Eromanga

Basin to have undergone a region wide seaway regression (Fig. 4-13). This study is in agreement that marine deposition dominantd much of the lower to middle sections of the

Mackunda Formation and that a major mechanism of environmental change within the basin is regression. However, the progression into overlying strata is a bit dubious. This study finds that the transition from the Upper Mackunda Formation to the Lower Winton

Formation, depositional environments indicate deltaic to coastal and shallow-marine environmental influences. The Lower Winton Formation is more associated with distal floodplain to delta associated environments with a great deal more tidal influences. A majority of Upper Winton Formation depositional environments are dominantly alluvial in nature within a broad sweeping floodplain. This study strived to provide more localized environmental interpretations that accompany recent fossil discoveries. The results of this study greatly support the historical broad-scale interpretation of the Eromanga Basin.

However, this study has successfully provided key context concerning the environments and deposition of sediments therein. To understand the evolution of environments within the northern Eromanga Basin, stratigraphic, sedimentological, paleontological, and paleobiological (derived from both field sites and regional core) are herein synthesized.

156

Figure4- 13 Simplified paleogeographic maps modified from Veevers (2000; and references therein) with symbols showing the retreat of the Eromanga Shallow Interior Seaway: A) No older than 104-102 Ma, the first occurrence of coastal deposits in the far east of the basin, as the central and western cores preserve dominantly shallow marine conditions, B) No older than 100-102Ma, the lower Winton Formation preserve distal deltaic to coastal environments; however far western cores preserve shallow marine sediments, C) No older than 92-94Ma, the middle Winton Formation deposited preserves fully terrestrial environments except the farthest western core which preserved coastal-rhythmites, D) The transition into the upper Winton Formation which is estimated to be deposited no older than 92 Ma, in all cores preserves terrestrial associations of sediments and both trace and body fossil assemblages. Temporal rates, paleocurrent and sediment source terranes discussed in detail within Tucker et al. (2014).

6.1.1 Middle to Upper Mackunda; Distal Delta Front to Prodelta (subtidal platform) depositional environments Much of the Mackunda Formation and underlying strata are interpreted to represent a transgressive/regressive shallow marine seaway (Fig. 4-10d;4-13; Draper, 2002; Veevers,

2000a,b,c). This study finds that shallow seaway deposition shifted in the Middle

Mackunda Formation which is interpreted to be part of an evolving prodelta system. The transition between the proximal delta fronts and distal fronts are interpreted to occur in the uppermost Mackunda Formation. This is interpreted as the early stages of the shallow

157 seaway retreat and the initiation of deltaic depositional conditions. Upper sections of the

Mackunda Formation exhibit a sedimentary record which is in transition from a lower delta plain to an upper delta plain with tidal influence, which continue into the lowermost

Winton Formation. This change in deposition co-occurs with the first appearance of coal within the middle-upper delta plain (wetlands/swamps). This also accompanies a shift from dominantly marine to marginal-estuarine faunal assemblages. This transition is not recorded in all cores. Centrally located cores indicate transition environments; western cores are dominantly shallow marine.

6.1.2 Lower Winton Formation; Tidally influenced depositional environments The Lower Winton Formation is interpreted from both core and field observations.

Key sedimentary features indicate that the Upper Mackunda Formation and Lower Winton

Formation occurred in marginal to deltaic associated environments (Fig. 4-10b,c;4-13).

This section of strata is interpreted to reflect the transition in deposition from prodelta/coastline to upper deltaic environments. Many associated sedimentary features indicate that this was likely a channel dominantd delta with periodic floods. Many associated environments likely represent very broad lower delta plain wetlands, swamps, and marshes rich in floral accumulation in anoxic environments (Fig. 4-10b,c). Vertebrate fossils also agree with the interpreted association of marginal environments (Salisbury et al., 2006a). Transitioning into the upper sections of the Lower Winton Formation also indicates another environmental shift. This progression likely indicates the delta prograding westward and this eastern portion of the Eromanga Basin is likely an evolving distal floodplain (marginally still tidally influenced). A caveat to this is deposition in the cores of the western portion of the basin which in the Middle Winton Formation is still shallow marine to prodelta (Fig. 4-10c,d).

158

6.1.3 Upper Winton Formation; Alluvial Depositional Environments Key sedimentary elements identified in the Upper Winton Formation from both outcrop and core aided in providing insights into its varied depositional history (Fig. 4-13).

In a regional scale, deposition within the Upper Winton Formation strongly indicates that at this point in basin evolution, alluvial environments were the main mechanism for the emplacement of sediment. This broad area is interpreted to have been a long-lived freshwater floodplain with multiple meandering channels. On a local scale, channel deposits indicate active migration and channel development throughout the floodplain.

Associated structures including overbank and abandoned channel deposits are identified throughout the stratigraphy. Development of both levees and banks ranged from short- lived and isolated to long-lived and very stable. Biogenic activity ranged from poor to intensely bioturbated protosol horizons, along with deep-reaching bifurcating roots.

Accompanying this, are the preserved remains of secondary and tertiary plant communities, which ranges from leaves, branches, and bark to stumps and moderately complete tree sections. With the association of these factors, it is likely that this flood plain represented wet conditions with minor seasonality. Emerging, yet to be described, microvertebrate and invertebrate assemblages from both LQCP and BNP ranges from freshwater bivalves to freshwater fish also support a long-lived wet and stable environment.

159

Figure 4- 11 Artistic reconstruction of the uppermost Winton Formation based on sedimentological, trace and body fossil assemblages at Lark Quarry Conservation Park. Reconstruction done by Owen Li (Petrified Pencils, 2013).

160

6.2 Revised stratigraphy and correlation of the Winton Formation and its floras and faunas A key objective in this investigation was to follow-up on the considerably refined age of the Mackunda and Winton formation reported in Tucker et al. (2013 [Chapter 1]) by refining the stratigraphic correlations of key fossil sites, and indeed the entire stratigraphy.

This objective is a particularly daunting task because of the extremely flat, low relief nature of the exposed Winton and Mackunda stratigraphy across the Eromanga Basin. There is actually very little stratigraphy at surface to work with in the first place and those few areas where relatively thick sections are exposed (at ‘jump ups’) are also characterized by intense surficial weathering that further complicates long distance correlations between sites.

Traditionally, because the dip on these beds is relatively gentle, facies tend to look relatively similar, and there are few signs of clear structures exposed at surface, most of the

Winton Formation fossil localities have been assumed to be roughly correlative and lumped into the Upper Winton Formation.

However, as noted above, significant variations and taphonomic characteristics do characterize different fossil localities in the Winton Formation. For instance, the important

Isisford vertebrate fauna from Isisford, Queensland has traditionally been assumed to be part of the Upper Winton Formation based on regional stratigraphic dip trends and location within the basin. However, when compared to other putatively upper Winton localities like

Bladensburg, Lark Quarry and Australian age of Dinosaurs sites, there are clear differences, not only in the composition of the fauna but also in its palaeoenvironmental reconstruction and depositional setting (Salisbury et al., 2006). For instance the Isisford fauna is dominantd by aquatic forms, some of which commonly have marine affinities (Salisbury, pers com. 2014). Much greater similarity seems to exist between the other three localities.

161

Hence, a key component of this study involved investigating new approaches for refining stratigraphic correlations between these sites and within the stratigraphy. To this end, detrital zircon datasets presented in Tucker et al. (2013 [Chapter 2], and 2014 [Chapter

3]) were reinvestigated for their stratigraphic utility. In addition, detailed stratigraphic sections and cores logged in this study were combined with other published core logs and a number of scanned core records provided by GSQ (via their Hylogging system). These logs were correlated and used to develop a basin model in ArcGIS. In addition, a thorough review of the available geophysical and structural data and literature on the Eromanga

Basin were compiled and used to identify key structural complications in the basin.

A number of important stratigraphic patterns are revealed in the detrital zircon datasets. Firstly, near continuous spectra of detrital zircon ages are observed between ~200

-92 Ma (especially between ~110-92) Ma when all zircon ages (n=719) are observed. This suggests that arc volcanism was relatively continuous throughout the Mesozoic, and that syndepositional, or near syndepositional age zircon grains were continuously shed into the basin. Although it should be noted that these grain ages represent maximum deposition ages only, and that their absence does not necessarily mean that a stratum is only as old as the youngest zircon. For instance, the Middle Mackunda detrital zircon sample yields a maximum depositional age at or around 128Ma (see Chapters 2-3), yet we have identified that this unit is no older than ~105 Ma (Tucker et al., 2013; InPrep). Directly above this unit, the youngest grains in the very Uppermost Mackunda Formation are no older than 102

Ma.

Detrital zircon samples recovered from the conformably overlying lower Winton

Formation indicate a maximum depositional age at or around 100 Ma. Progressing up

162 section into the middle Winton Formation, recovered grain ages indicate the first occurrence of grains younger than 100 M. For the middle Winton Formation, youngest maximum depositional age would be between 93-94 Ma (two individual grains are younger; however these do not constitute true clusters or populations). This is conformable with overlying of the uppermost Winton Formation, which contains youngest maximum depositional grain ages that are between 92-93 Ma. This significant maximum depositional grain age pattern is a repeated occurrence in six individual samples from across the basin.

However, recovered grain ages from Isisford, which were originally interpreted to be surface exposures of the uppermost Winton Formation, are not in agreement with this conclusion. Therefore, this study questioned the validity of Isisford’s stratigraphical positioning when compared to other fossil bearing localities. This problem was further exacerbated by the lack of similarity between associated fossil assemblages found in

Isisford compared to other upper Winton Formation localities. This was originally thought to represent minor variation in depositional regimes and sedimentary input. However, when this study reconstructed basic basin wide facies models via key fossil localities and core logs, Isisford directly correlated with facies associations identified in the lower

Winton. Thus, Isisford’s faunal composition and palaeoenvironmental setting reflects subsurface structural variability rather than a reflection of unique depositional conditions.

Upon review of modern and historical records, the upper units of the Eromanga Basin are highly faulted over very short distances.

This study sought to construct a multifaceted palaeontological and sedimentological based framework to address the placement of key fossil localities of the Winton Formation.

As an unexpected result, this study has found compelling evidence for the first

163 identification of two, temporally separate vertebrate assemblages within the Winton

Formation. When combining near identical fossil assemblages (floral and faunal), with congruent facies associations and a very robust detrital record, this study confidently places the deposition of the upper Winton Formation at an age no older than 92-94 Ma. This is also true for the lower Winton Formation. By coupling fossil and facies information form

Isisford with regional cores, this study confidently places the deposition of the lower

Winton Formation no older than 100 Ma. Excitingly, this has several key impacts on vertebrate assemblages: 1) that fossil assemblages within Isisford are not contemporaneous with Lark Quarry Conservation Park or Bladensburg, 2) that by coupling fossil, detrital and facies associations, this study was able to identify and resolve cryptic stratigraphic qualms, and 3) for the first time, confidently display that the Winton Formation can be informally separated into a lower and upper section. The informal split of the Winton Formation strongly suggests a considerable temporal unconformity and could be the result of a sequence boundary associated with the final retreat of the Eromanga Sea from the basin.

7. Conclusion Detailed facies and architectural analysis of the Winton Formation and underlying

Mackunda Formation, has resulted in a much better understanding of the palaeoenviroments and stratigraphy of this important vertebrate-bearing succession.

Twenty-two reoccurring facies codes were identified and utilized to construct nine distinct facies associations. Each facies association was coupled with available palaeontologic and near-syndepositional geochronological data, and as a result, the progression of environmental change could be reliably interpreted within the Mackunda and Winton formations. Depositional conditions no older than 104 Ma indicates that much of the

164

Eromanga Basins was inundated with the last stages of a shallow marine seaway.

However, by an age no older than 100 Ma, sedimetological evidence indicates receding shallow marine to coastal environments in the eastern-most reaches of the basin. On the other hand, the overlying Winton Formation is interpreted to preserve evidence of an evolving terrestrial depositional regime. Based on the association of sedimentary and biological indicators, the lower Winton Formation is interpreted to have deposited sediment within very distal flood plains to lower delta plains. The Upper Winton Formation is dominantd by an association of alluvial floodplain deposits which hosted multiple migrating channels within a low-lying, flood plain. This west- to southwest retreat of the shallow marine seaway continued for much of the Cenomanian, 102-92 Ma, or younger.

This study has found that for much of the Cenomanian to Turonian, deposition of sediments was dominantd by the developments of a westward expanding progradational river system dominanted delta front under normal regression influences.

The utilization of these methodologies has not only facilitated stratigraphic correlation between key fossil localities but also aids in the identification of two temporally distinct faunal assemblages in the Winton Formation. Based on these results, faunal assemblages and their palaeoecosystems currently excavated at Lark Quarry Conservation

Park and Bladensburg National Park are near-contemporaneous. On the other hand, fossil assemblages and their palaeoecosystems near Isisford, Queensland are unique to those in the central portions of the Eromanga Basin and interpreted to preserve lower stratigraphic units of the Winton Formation. It should be stressed that there is likely numerous shallow fault systems inbetween these and similar fossil assemblage sites between Isisford and

Winton, Queensland. Thus, any temporal placement of this fossil material should be

165 rigorously tested and coupled with other means of stratigraphic or temporal placement, ergo radiometric dating.

Furthermore, this multifaceted approach has identified of cryptic structural complexities not readily noticed in the field, along with the informal separation of the lower and upper Winton Formation with an associated temporal unconformity. Detrital zircon data were critical to identifying all of these factors and emphasize the utility of detrital zircon geochronology for better constraining other terrestrial faunal assemblages. For the first time, this study has confidently provided a robust geological and palaeontological based framework for relating various Winton based assemblages. Furthermore, for the first time, this framework will allow much more accurate comparisons between Australian

Winton taxa to similar forms in other locations around Gondwana (i.e Rio Limay Subgroup of the Neuquén Basin of Argintina, and the Santo Anastácio and Adamantina Formations of the Bauru Basin, of Brazil).

166

167

Chapter 5

Thesis Summary

168

The rapid separation of Gondwana, principally during the Cretaceous, represented one of the most important palaeobiogeographic dispersal and subsequent vicariance events in Earth history. The relationship of fossils from the eastern-most portions of Gondwana, in particular Australia; remain poorly understood due to poor temporal and palaeoenvironmental constraints on this important fossil record (Bonaparte, 1986; Agnolin et al., 2010). In light of many recent discoveries of new fossils in the Winton Formation that are beginning to shed light on the radiation and dispersal of dinosaurian clades between

South America and Australia via Antarctica (Agnolin et al., 2010), the issue of poor geologic context has become acute. Hence, the overriding goal of this study was to provide key temporal, stratigraphic and palaeoenvironmental information for the Winton Formation and its entombed flora and fauna.

In order to improve upon previous palynological-based temporal constraints, this thesis first investigated the potential for utilizing maximum depositional age constraints derived from detrital zircons via U-Pb LA-ICPMS geochronology. This investigation proved to be far more successful than expected, demonstrating that much of the formation formed syndepositionally with the final stages of arc volcanism associated with subduction along the east coast of northeastern Australia. By utilizing various statistical approaches for determining maximum depositional age from detrital zircon populations, this study concluded in Chapter 1 that the most richly fossiliferous portion of the Upper Winton

Formation, was deposited no earlier than Cenomanian-Turonian Boundary (92-94 Ma), and indeed, these fossils are most likely wholly Turonian or younger. Additional detrital zircon samples were investigated for sedimentary provenance analysis in Chapter 2 and this work both confirmed the age of the Upper Winton Formation and extended the post-

169

Cenomanian-Turonian boundary age interpretation for the formation for strata of the middle Winton Formation too. More specifically the combined results of Chapters 1 and 2 now indicate that nearly all known Winton Fossil are Turonian or younger. However, one very interesting stratigraphic trend that was also determined first through detrital zircon patterns, and since supported by seismic and structural data, is that a considerable unconformity likely exists between the lower and upper Winton Formation. This is recognized by a distinct gap in the detrital zircon age spectra between all Mackunda and lower Winton samples and all middle to upper Winton Samples. Specifically, the youngest detrital zircons observed in the middle to upper Mackunda formation are ~104-102 Ma, and

102-100 Ma in all lower Winton samples, whereas all other middle and upper Winton

Formation samples preserve detrital zircons that encompass the entire age spectra between

100 to 94 Ma, with several key samples preserving 93-92 Ma grains. This suggests that the lack of grains younger than 100 Ma in any lower Winton Formation samples is a true stratigraphic signal. This finding is particularly important in that one of the most important fossil localities in all of the Winton Formation (Isisford), which was originally thought to be part of the Upper Winton Formation, yielded a zircon age spectra similar to that of the

Upper Mackunda and lower Winton Formation. Absolutely no young (post 100 Ma) grains were observed. This led to a closer inspection of the regional geology and geophysics and the development of a GIS basin model (Chapter 3). As a result of this investigation, a significant fault zone was identified between Isisford other the key fossil localities at

Bladensburg and Lark Quarry in the upper Winton Formation and confirms that Isisford is indeed part of the Lower Winton Formation. A brief inspection of the Isisford fauna indicates a distinctly different faunal composition than the rest of the Winton Formation.

Furthermore, the characterization of the fauna’s entombing sediment also strongly indicates 170 a mixed depositional environment, including brackish to marine elements amongst the terrestrial components. This hypothesis requires greater faunal analysis and detailed taphonomic investigation; however, the stratigraphic implications are clear. Hence, the methodologies applied in each chapter successively build together to develop a robust approach to identifying cryptic structural complexities in the field and recognizing temporally distinct faunal assemblages in the Winton Formation. For the first time, we can reliably say that the Winton Formation preserves two distinct vertebrate assemblages. This includes: 1) the Lower Winton fauna (Isisford) that was deposited after 102 Ma and probably not later than 100-99 Ma; and 2) the upper Winton fauna, interpreted to be no older than the Cenomanian-Turonian boundary.

In addition to the significant temporal and stratigraphic advances to understanding the Winton Formation and its flora and fauna, this study has also led to a significantly improved understanding of the sedimentary provenance and drainage evolution of the final stages of the Eromanga Basin. Yet more significantly, this provenance investigation has helped to answer a number of long standing mysteries surrounding the tectonic history of the Eastern Margin of Australia during the mid-Cretaceous.

The sedimentary provenance of the Mackunda and Winton Formations is dominantd by late Mesozoic arc-derived volcanolithic sediment input from the east. Very minor polyphase reworked grain populations from the Archean to Proterozoic were identified in all samples. These grains are likely derived from western lying cratons and reworked into eastern lying metasediments that are re-eroded. Minor Palaeozoic (500 to 200 Ma) populations were also identified in all samples, and many of these grains can be readily associated with local terranes in northeastern Queensland. As a whole, all samples are

171 dominantd by Mesozoic grain ages spanning a near continuous age spectra between 250 to

92 Ma. Key detrital population between the Ordovician to Latest Triassic can be confidently associated with long-standing subduction of the Phoenix-Pacific Plate under that of the evolving Gondwana/Australian eastern margin (New England & Kennedy

Provinces). The presence of Triassic thru mid-Cretaceous age zircons is particularly interesting as this is a time period in which the tectonic activity of eastern Australia is not well understood. By utilizing the recovered zircon grain populations, we can readily interpret that there was volcanic activity throughout the Mesozoic, and that it continued nearly uninterrupted. Younger populations between 130 to 95 Ma are confidently associated with arc volcanics of this age known from Whitsunday Volcanic Province. A tail of younger grains between 94-92 Ma demonstrate that this volcanic system was active longer than previously thought and that rifting of the (Tasman Basin) took place only after

92 Ma. Thus, during the Mesozoic, near continuous production of arc-related magmatic activity along the eastern Australian margin supplied what would later be fossil entombing volcanalithic rich sediment.

The final contribution of this thesis to the field is a much better sedimentologic and palaeoenvironmental understanding of the uppermost stratigraphic units in the Eromanga

Basin, the Mackunda and Winton formation. This study has found that for much of the

Cenomanian and Turonian, as the Eromanga Sea retreated to the west, fully alluvial environments succeed deltaic and tidal systems. This progradational phase is a direct result of the retreat of the Eromanga Seaway which is interpreted to have initiated no older than

104 Ma and continued to drain well into the Turonian. In sum, the results of this thesis have led to a more rigorous and temporally constrained depositional history for the middle

172

Cretaceous in the Eromanga Basin, and a significantly expanded understanding the tectonic history of eastern Australia during this time. It is hoped that the results of this work will help to place the mid-Cretaceous flora and fauna of Australian into a constrained

Gondwana, and indeed global context.

173

Reference

Adams, C.J., Barley, M.E., Fletcher, I.R., Pickard, A.L., 1998. Evidence from U–Pb zircon and 40Ar/39Ar muscovite detrital mineral ages in metasandstones for movement of the Torlesse suspect terrane around the eastern margin of Gondwanaland. Terra Nova, 10(4), 183-189.

Adams, C.J., Campbell, H.J., Griffin, W.L., 2007. Provenance comparisons of Permian to Jurassic tectonstratigraphic terranes in New Zealand: perspectives from detrital zircon age patterns. Geological Magazine,144(4), 701-729.

Adams C.J., Cluzel D., Griffin W. L., 2009. Detrital-zircon ages and geochemistry of sedimentary rocks in basement Mesozoic terranes and their cover rocks in New Caledonia, and provenances at the Eastern Gondwanaland margin. Australian Journal of Earth Sciences, 56(8), 1023-1047.

Adams, C.J., Griffin, W.L., 2012. Rodinian detrital zircons in Late Cretaceous sandstones indicate a possible Precambrian basement under southern Zealandia. Precambrian Research, 212, 13-20.

Adams C.J., Mortimer N., Campbell H.J., Griffin W. L., 2013a. Detrital zircon geochronology and sandstone provenance of basement Waipapa Terrane (Triassic– Cretaceous) and Cretaceous cover rocks (Northland Allochthon and Houhora Complex) in northern North Island, New Zealand. Geol. Mag. 150(1), 89–109.

Adams C.J., Korsch R.J., Griffin W.L., 2013b. Provenance comparisons between the Nambucca Block, Eastern Australia and the Torlesse Composite Terrane, New Zealand: connections and implications from detrital zircon age patterns. Australian Journal of Earth Sciences, 60(2), 241-253.

Agnolin, F.L., Ezcurra, M.D., Pais, D.F., Salisbury, S.W., 2010. A reappraisal of the Cretaceous non-avian dinosaur faunas from Australia and New Zealnad: evidence for their Gondwanan affinities. Journal of Systematic Palaeontology, 8(2), 257–300.

Alexander, E.M., Gravestock, D.I., Cubitt, C. Chaney, A., 1998. Lithostratigraphy and environments of deposition. In: Gravestock, D.I.,Hibburt, J.E. & Drexel, J.F. (Eds): The petroleum geology of South Australia. Volume 4: Cooper Basin. Primary Industries and Resources SA, Report Book 98/9, 69–115.

Allen, J.R.L., 1963. The classification of cross-stratified units, with notes on their origin. Sedimentology, 2(2), 93–114.

Amidon, W.H., Burbank, D.W., Gehrels, G.E., 2005. U–Pb zircon ages as a sediment mixing tracer in the Nepal Himalaya. Earth and Planetary Science Letters, 235(1), 244-260.

174

Almond, C.S., 1983. Stratigraphic Drilling Report –GSQ Eromanga 1. Queensland Government Mining Journal, 84, 359–368.

Almond, C.S., 1986. Stratigraphic drilling report –GSQ Thargomindah 1-1A. Queensland Government Mining Journal, 87, 68–77.

Andersen, T., 2005. Detrital zircons as tracers of sedimentary provenance: limiting conditions from statistics and numerical simulation. Chemical Geology, 216(3), 249– 270.

Andrews P.B., Speden I.G., Bradshaw J.D., 1976. Lithological and paleontological content of the Carboniferous-Jurassic Canterbury Suite, South Island, New Zealand, New Zealand Journal of Geology and Geophysics, 19(6), 791-819.

Angevine, C.K., Heller, P.C., Paola, C., 1990. Quantitative sedimentary basin modeling. AAPG Continuing Education Course Notes Series, 32, 1–247.

Babinski, M., Boggiani, P.C., Trindade, R.I.F., Fanning, C.M., 2012. Detrital zircon ages and geochronological constraints on the Neoproterozoic Puga diamictites and associated BIFs in the southern Paraguay Belt, Brazil. Gondwana Research. ISSN 1342-937X. http://dx/doi.org/10.1016/j.gr.2012.06.011.

Bann, K.L., Fielding, C.R., 2004. An integrated ichnological and sedimentological comparison of non-deltaic shoreface and subaqueous delta deposits in Permian reservoir units of Australia. In: McIlroy, D. (Ed.), The Application of Ichnology to Palaeoenvironmental and Stratigraphic Analysis: Geological Society of London Special Publication, 228, 273–307.

Bann, K.L., Fielding, C.R., MacEachern, J.A., Tye, S.C., 2004. Differentiation of estuarine and offshore marine deposits using integrated ichnology and sedimentology: Permian Pebbley Beach Formation, Sydney Basin, Australia. In: McIlroy, D. (Ed.), The Application of Ichnology to Palaeoenvironmental and Stratigraphic Analysis: Geological Society of London Special Publication, 228, 179–211.

Bann, K.L., Tye, S.C., MacEachern, J.A., Fielding, C.R., Jones, B.G., 2008. Ichnological and sedimentologic signatures of mixed wave- and storm-dominantd deltaic deposits: Examples from the Early Permian Sydney Basin, Australia. In: Hampson, G., Steel, R., Burgess, P., Dalrymple, R., (Eds). Recent advances in models of siliciclastic shallow-marine stratigraphy. SSEPM Special Publications, 90, 293-332.

Barbeau, D.L. Jr., Gombosi, D.J., Zahid, K.M., Bizimis, M., Swanson-Hysell, N., Valencia, V., Gehrels, G.E., 2009. U-Pb zircon constraints on the age provenance of the Rocas Verdes basin fill, Tierra del Fuego, Argentina. Geochemistry, Geophysics, Geosystems, 10(12), 1-11.

175

Benson, R.B.J., Carrano, M.T., Brusatte, S.L., 2010. A new clade of archaic large-bodied predatory dinosaurs (Theropoda, Allosauroidea) that survived the latest Mesozoic. Naturwissenschaften, 97(1), 71–78.

Benton, M.J., Forth, J., Langer, M.C., 2014. Models for the Rise of the Dinosaurs. Current Biology, 24(2), 87-95. Berrell, R., Alvarado-Ortega, J., Yabumoto, Y., Salisbury, S.W., 2008. A new ichthyodectiform fish from the mid-Cretaceous of Australia. Journal of Vertebrate Paleontology, 28, 52A.

Berrell, R.W., Alvarado-Ortega, J., Yabumoto, Y., Salisbury, S.W., 2011. Ichthyodectiform fishes (Teleoste: Ichthyodectiformes) from the mid-Cretaceous Eromanga Basin of central-western Queensland, with comments on the validity of Cladocyclus sweeti Woodward, 1884. 13th Conference on Australasian Vertebrate Evolution, Palaeontology and Systematics. Geological Survey of Western Australia, Perth, p. 19.

Betts, P.G., Giles, G.S., Frick, L.R., 2002. Evolution of the Australian lithosphere. Australian Journal of Earth Sciences, 49(4), 661-695.

Bhattacharya, J.P., 2010. Deltas. In: James, N.P., Dalrymple, R.W. (Ed.). Facies models. Geological Association of Canada.

Bhattacharya, J.P., Walker, R.G., 1992. Deltas. In: Facies Models: Response to Sea-Level Change (Eds) Walker, R.G., James, N.P., 157–178. Geological Association of Canada, St Johns.

Bishop, D.G., Bradshaw, J.D., Landis, C.A., 1985. Provisional terrane map of South Island, New Zealand. In: Howell, D.G. (Ed.), Tectonostratigraphic Terranes, Circum-Pacific Council for Energy and Mineral Resources Earth Science Series, No. 1, Houston, Texas, 515–521.

Black, L. P., Kamo, S. L., Allen, C. M., Aleinikoff, J. N., Davis, D. W., Korsch, R. J., Foudoulis, C. 2003. TEMORA 1: a new zircon standard for Phanerozoic U–Pb geochronology. Chemical Geology, 200(1), 155-170.

Blake, D.H., Stewart, A.J., 1992. Stratigraphic and tectonic framework, Mount Isa. Bulletin of the Australian Geological Survey Organization, 243, 1-11.

Bonaparte, J.F., 1986. The early radiation and phylogenetic relationships of sauropod dinosaurs, based on vertebral anatomy. In: Padian, K. (Ed.), The Beginning of the Age of Dinosaurs. Cambridge University Press, Cambridge and New York, 247– 258.

Bose, M.N., 1955. Some Tertiary plant remains from Queensland, Australia. Botaniska Notiser, 108, 381–390.

176

Bown, T.M., Kraus, M.J., 1987. Integration of channel and floodplain suites, I. Developmental sequence and lateral relations of alluvial paleosols. Journal of Sedimentary Research, 57(4).

Bryan, S.E., Ernst, R.E., 2008. Revised definition of large igneous provinces (LIPs). Earth- Science Reviews, 86(1), 175-202.

Bryan, S.E., Constantine, A.E., Stephens, C.J., Ewart, A., Schön, R.W., Parianos, J., 1997. Early Cretaceous volcano-sedimentary successions along the eastern Australian continental margin: implications for the break-up of eastern Gondwana. Earth and Planetary Science Letters, 153(1), 85-102.

Bryan S.E., Ewart A., Stephens C.J., Parianos J., Downes P.J., 2000. The Whitsunday Volcanic Province, Central Queensland, Australia: lithological and stratigraphic investigations of a silicic-dominantd large igneous province. Journal of Volcanology and Geothermal Research, 99, 55–78.

Bryan, S.E., Fielding, C.R., Holcombe, R.J., Cook, A. Moffitt, C.A., 2003, Stratigraphy, facies architecture and tectonic implications of the Upper Devonian to Lower Carboniferous Campwyn Volcanics of the northern New England Fold Belt. Australian Journal of Earth Sciences, 50(3), 377–401.

Bryan, S.E., Cook, A.G., Allen, C.M., Siegal, C., Purdy, D.J., Greentree, J.S., I.T.U., 2012. Early-mid Cretaceous tectonic evolution of eastern Gondwana: From silicic LIP Magmatism to continental rupture. Episodes, 35(1), 142-152.

Bryan, S.E., Ferrari, L., (2013). Large igneous provinces and silicic large igneous provinces: progress in our understanding over the last 25 yr. Geological Society of America Bulletin, 125 (7/8), 1053-1078.

Buatois, L.A., Mángano, M.G., 2011. Ichnology: Organism-substrate interactions in space and time. Cambridge University Press.

Buatois, L.A., Mángano, M.G., 1995. The paleoenvironmental and paleoecological signiﬁcance of the lacustrine Mermia ichnofacies: an archetypical subaqueous nonmarine trace fossil assemblage. Ichnos, 4(2), 151–161.

Burger, D., 1973. Spore zonation and sedimentary history of the Neocomian, Great Artesian Basin. Queensland Society of Australian Special Publications, 4, 87–118.

Burger, D., 1980. Palynological studies in the lower Cretaceous of the Surat Basin. Australian Burro of Mineralogical Resources Bulletin, 189, 1–69.

Burger, D., 1986. Palynology, cyclic sedimentation and palaeoenvironments in the late Mesozoic of the Eromanga Basin. In: Gravestock, P.S., Moore, P.S., Mount, T.J., Pitt, G.M. (Eds.), Contributions to the Geology and Hydrocarbon Potential of the Eromanga 177

Basin. Special Publication of the Geological Society of Australia, 12, 53–70.

Burger, D., 1989. Stratigraphy, palynology, and palaeoenvironmentsof the Hooray Sandstone, eastern Eromanga Basin, Queensland and New South Whales. Department of Mines Report, 3, 1–28.

Burger, D., 1990. Early Cretaceous angiosperms from Queesnalnd, Australia. Review of Palaeobotany and Palynology, 65(1), 153–163.

Burger, D., 1993. Early and middle Cretaceous angiosperm pollen grains from Australia. Review of Palaeobotany and Palynology, 78(3), 183–234.

Burger, D., Senior, B.R., 1979. A revision of the sedimentary and palynological history of the northeastern Eromanga Basin, Queensland. Journal of the Geological Society of Australia, 26(3-4), 121–132.

Campbell, J.D., Coombs, D.S., 1966. Murihiku Supergroup (Triassic—Jurassic) of Southland and South Otago, New Zealand. Journal of Geology and Geophysics, 9(4), 393-398.

Casey, D.J., 1970. Northern Eromanga Basin. Geological Survey of Queensland, Report No. 41.

Carne, G.E., Alexander, E., 1997. Eromanga Basin Prospects Inventory Blocks ER97-A to C. Department of Mines and Energy Resources South Australia.

Carrapa, B., 2010. Resolving tectonic problems by dating detrital minerals. Geology, 38(2), 191–192.

Cartwright, J., Lonergan, L., 1997. Seismic expression of layer-bound fault systems of the Eromanga and North Sea Basins. Exploration Geophysics, 28(3), 323-331.

Cartwright, J.A., Dewhurst, D.N., 1998. Layer-bound compaction faults in fine-grained sediments. Geological Society of America Bulletin, 110(10), 1242-1257.

Casey, D.J., 1967a. Subsurface geology of the Longreach 1:250000 sheet area. Geological Survey of Queensland Records.

Casey, D.J., 1967b. Subsurface geology of the Maneroo and Winton 1:250000 sheet area. Geological Survey of Queensland Records.

Casey, D.J., 1970. Northern Eromanga Basin. Geological Survey of Queensland, Report No. 41.

Cawood, P.A., Nemchin, A.A., Leverenz, A., Saeed, A., Balance, P.F., 1999. U/Pb dating of detrital zircons: implications for the provenance record of Gondwana margin terranes. Geological Society of America Bulletin, 111(8), 1107-1119. 178

Cawood, P.A., Nemchin, A.A., Strachan, R., Prave, T., Krabbendam, M., 2007 sedimentary basin and detrital zircon record along east laurentia and baltica during assembly and breakup of rodinia. Journal of the geological society, 164(2), 257–275.

Champion, D.C., Bultitude, R.J., 2013. Kennedy Igneous Association. In: Jell, A.P., (Ed.) Geology of Queensland. Queensland Government, Brisbane, QLD, Australia, 225- 304.

Chure, D., Britt, B.B., Whitlock, J.A., Wilson, J.A., 2010. First complete sauropod dinosaur skull from the Cretaceous of the Americas and the evolution of sauropod dentition. Naturwissenschaften, 97(4), 379–391.

Clifford, H.T., Dettmann, M.E., 2005. First record from Australia of the Cretaceous fern genus Tempska and the description of a new species, T. judithae. Review of Palaeobotany and Palynology, 134, 71–84.

Cluzel, D., Adams, C.J., Meffre, S., Campbell, H., Maurizot, P., 2010. Discovery of early cretaceous rocks in New Caledonia: New Geochemical and U-Pb zircon age constraints on the transition from subduction to marginal breakup in the Southwest Pacific. The Journal of Geology, 118(4), 381-397.

Cluzel, D., Adams, C.J., Maurizot, P., Meffre, S., 2011. Detrital zircon records of Late Cretaceous syn-rift sedimentary sequences of New Caledonia: an Australian provenance questioned. Tectonophysics, 501(1), 17-27.

Cluzel, D., Maurizot, P., Collot, J., Sevin, B., 2012. An outline of the Geology of New Caledonia; from Permian-Mesozoic Southeast Gondwanaland active margin to Cenozoic obduction and supergene evolution. Episodes-News magazine of the International Union of Geological Sciences, 35(1), 72.

Collins, W. J., Richards, S. W., 2008. Geodynamic significance of S-type granites in circum-Pacific orogens. Geology, 36(7), 559-562.

Collot, J., Geli, L., Lafoy, Y., Vially, R., Cluzel, D., Klingelhoefer, F., Nouzé, H., 2008. Tectonic history of northern New Caledonia Basin from deep offshore seismic reflection: Relation to late Eocene obduction in New Caledonia, southwest Pacific. Tectonics, 27(6), TC6006.

Collot, J., Herzer, R., Lafoy, Y., Géli, L., 2009 Mesozoic history of the Fairway-Aotea Basin: Implications for the early stages of Gondwana Fragmentation. Geochemistry, Geophysics, Geosystems, 10(12).

Coney, P.J., Edwards, A., Hine, R., Morrison, F., Windram D., 1990. The regional tectonics of the Tasman orogenic system, eastern Australia. Journal of Structural Geology, 12, 519–543.

179

Cook, A.G., 2005. First record of fossil freshwater gastropods within the Winton Formation. Memoirs of the Queensland Museum, 51, 406.

Cook, A.G., Bryan, S.E., Draper, J.J., 2013. Post-orogenic Mesozoic basins and magmatism. In: Jell, A.P., (Ed.) Geology of Queensland. Queensland Government, Brisbane, QLD, Australia, 225-304.

Coombs, W.P., Molnar, R.E., 1981. Sauropoda (Reptilia, Saurischia) from the Cretaceous of Queensland. Memoirs of the Queensland Museum, 20, 351–373.

Coote, S.M., 1987, GSQ Blackall 2- Preliminary lithologic log and composit log. Geological Survey of Queensland, Record 1987/12.

Davies-Vollum, K.S., Wing, S.L., 1998. Sedimentological, taphonomic, and climatic aspects of Eocene swamp deposits (Willwood Formation Bighorn Basin, Wyoming). Palaios, 13(1), 28–40.

Day, R.W., Whitaker, W.G., Murray, C.G., Wilson, I.H., Grimes, K.G., 1983. Queensland Geology. Geological Survey of Queensland, 383, 1-194.

DeGraaff-Surpless, K., Mahoney, J.B., Wooden, J.L., McWilliams, M.O., 2003. Lithofacies control in detrital zircon provenance studies: Insights from the Cretaceous Methow basin, southern Canadian Cordillera. Geological Society of America Bulletin, 115(8), 899-915.

Demko, T.M., Currie, B.S., Nicoll, K.A. 2004. Regional paleoclimatic and stratigraphic implications of paleosols and fluvial/overbank architecture in the Morrison Formation (Upper Jurassic), Western Interior, USA. Sedimentary Geology, 167(3), 115-135.

Dettmann, M.E., 1973. Angiospermous pollen from Albian to Turonian sediments of eastern Australia. Special Publication of the Geological Society of Australia, 4, 3–34.

Dettmann, M.E., 1984a. Preliminary palynological report - Jackson No. 1. Mines Administration Pty Ltd Palynological Laboratory Report No. 13/141. Report held by the Department of Natural Resources and Mines Queensland as CR 12790.

Dettmann, M.E., 1984b. Palynology – Munkarie South No. 1. Mines Administration Pty Ltd Palynological Laboratory Report No. 13/198 as Appendix VI in DIO Munkarie South 1 Well Completion Report. Report held by the Department of Natural Resources and Mines Queensland as CR 13312.

Dettmann, M.E., 1987. Biostratigraphy, palaeoenvironments, and hydrocarbon source potential of Jurassic and Lower Cretaceous sediments in Monler No. 1, 3130ft – 4920ft. Appendix 9 in HEP Monler 1 Well Completion Report. Report held by the Department of Natural Resources and Mines Queensland as CR 16942.

180

Dettmann, M.E., 1988. Cretaceous sediments in GSQ Thargomindah 3, 842.57m – 1306.85m. Report held by the Department of Natural Resources and Mines Queensland as CR 16964.

Dettmann, M.E., 1994. Cretaceous vegetation: the microfossil record. In: Hill, R.S. (Ed.), History of the Australian Vegetation: Cretaceous to Recent. Cambridge University Press, Cambridge, 390–419.

Dettmann, M.E., Playford, G., 1969. Palynology of the Australian Cretaceous: a review. In: Campbell, K.S.W. (Ed.), Stratigraphy and Palaeontology - Essays in honor of Dorothy Hill. Australian National University Press, Canberra, 174–210.

Dettmann, M.E., Molnar, R.E., Douglas, J.G., Burges, D., Fielding, C., Clifford, H.T., Francis, J., Jell, P., Rich, T., Wade, M., Rich, P.V., Pledge, N., Kemp, A., Rozefelds, A., 1992. Australian Cretaceous terrestrial faunas and floras: biostratigraphic and biogeographic implications. Creteceous Research, 13, 207–262.

Dettmann, M.E., Clifford, H.T., 2000. Gemmae of the marchantiales from the Winton Formation (mid-Cretaceous), Eromanga Basin, Queensland. Memoirs of the Queensland Museum, 45, 285–292.

Dettmann, M.E., Clifford, H.T., Peters, M., 2009. Lovellea wintonensis gen. et. sp. nov. — Early Cretaceous (late Albian), anatomical preserved, angiospermous flowers and fruits from the Winton Formation, western Queensland, Australia. Cretaceous Research, 30(2), 339–355.

Dickinson, W.R., Gehrels, G.E., 2003. U–Pb ages of detrital zircons from Permian and Jurassic eolian sandstones of the Colorado Plateau, USA: paleogeographical implications. Sedimentary Geology, 163(1), 29–66.

Dickinson, W.R., Gehrels, G.E., 2009. Use of U-Pb ages of detrital zircons to infer maximum depositional ages of strata: A test against a Colorado Plateau Mesozoic database. Earth and Planetary Science Letters, 288(1), 115-125.

Donchak, P.J.T., Purdy, D.J., Withnall, I.W., Blake, P.R., Jell, P.A., 2013. New England Orogen. In: Jell, A.P., (Eds.) Geology of Queensland. Queensland Government, Brisbane, QLD, Australia, 225-304.

Draper, J.J., Fielding, C.R., 1997. Discussion. SHRIMP zircon dating of the Permian System of eastern Australia. Australian Journal of Earth Sciences, 44, 535–538.

Draper, J.J. (Ed.), 2002. Geology of the Cooper and Eromanga Basins, Queensland. Queensland Minerals and Energy Review Series, Queensland Department of Natural Resources and Mines.

181

Dunstan, B., 1916. Appendix B; Queensland Geological Formations, In: Harrap. G., A School geography of Queensland. Department of Public Instruction, Brisbane, 164- 176.

Eberth, D.A., Miall, A.D., 1991. Stratigraphy, sedimentology and evolution of a vertebrate- bearing, braided to anastomosed fluvial system, Cutler Formation (Permian- Pennsylvanian), north-central New Mexico. Sedimentary Geology, 72(3), 225-252.

Ernst, R.E., Buchan, K.L., 2001. Large mafic magmatic events through time and links to mantle-plume heads. In: Ernst, R.E., Buchan K.L., (Eds.), Mantle Plumes: Their Identification Through Time. Geological Society of America Special Paper, 352, 483–575.

Ernst, R.E., Bleeker, W., Söderlund, U., Kerr, A.C., 2013. Large Igneous Provinces and supercontinents: Toward completing the plate tectonic revolution. Lithos, 174, 1-14

ESSO Australia Pty Ltd., 1981. Isis Downs-1 well completion report, 268P, Queensland. (unpublished).

Evans, P.R., 1966. Mesozoic stratigraphic palynology in Australia. Australaian Oil and Gas Journal, 12(6), 58–63.

Ewart A., Schon, R.W., Chappell, B.W., 1992. The Cretaceous volcanic plutonic province of the central Queensland (Australia) coast—a rift related ‘calcalkaline’ province. Transactions of the Royal Society of Edinburgh. Earth Sciences, 83(1-2), 327-345.

Exon, N.F., 1966. Revised Jurassic to Lower Cretaceous stratigraphy in the south-east Eromanga Basin, Queensland. Queensland Government Mining Journal, 67, 232– 238.

Exon, N.F., Senior, B.R., 1976. The Cretaceous geology of the Eromanga and Surat Basins. BMR Journal of Australian Geology & Geophysics, 1, 33–50.

Exon, N.F., Burger, D., 1981. Sedimentary cycles in the Surat Basin and global changes and sea level. BMR Journal of Australian Geology & Geophysics, 6, 153-60.

Faggotter, S.J., Salisbury, S.W., Yabumoto, Y., 2007. A new possible haleocomorph fish from the mid-Cretaceous (Albian-Cenomanian) Winton Formation of Isisford, central-western Queensland, Australia. In: Warren, A. (Ed.), 2007 Conference on Australasian Vertebrate Evolution. Palaeontology and Systematics, Geological Society of Australia, p. 55.

Fedo, C.M, Sircombe, K.N., Rainbird, R.H., 2003. Detrital Zircon Analysis of Sedimentary Record, In: Hanchar, J.M., Hoskin, P.W.O. (Eds.), Zircon. Reviews in Mineralogy and Geochemistry. 53(1), 277-303.

182

Fergusson, C.L., 1999. Subduction-related assemblages in the Tasmanides of southeastern Australia. New England Orogen, 13.

Fergusson, C.L., Henderson, R.A., Lewthwaite, K.J., Phillips, D. Withnall, I.W., 2005. Structure of the Early Palaeozoic Cape River Metamorphics, Tasmanides of north Queensland: evaluation of the roles of convergent and extensional tectonics, Australian Journal of Earth Sciences. An International Geoscience Journal of the Geological Society of Australia, 52(2), 261-277.

Fergusson, C.L., Henderson, R.A., Withnall, I.W., Fanning, C.M., Phillips, D., Lewthwaite, K.J., 2007. Structural, metamorphic, and geochronological constraints on alternating compression and extension in the Early Paleozoic Gondwanan Pacific margin, northeastern Australia. Tectonics, 26, TC3008.

Fergusson, C.L., Henderson, R.A., 2013. Thomson Orogen. In: Jell, A.P., (Ed.) Geology of Queensland. Queensland Government, Brisbane, QLD, Australia, 225-304.

Fielding, C.R., 1985. Coal depositional models and the distinction between alluvial and delta plain environments. Sedimentary Geology, 42(1), 41-48.

Fielding, C.R., 1987. Coal depositional models for deltaic and alluvial plain sequences. Geology, 15(7), 661-664.

Fielding, C.R., 1996. Mesozoic sedimentary basins and resources in eastern Australia-a review of current understanding. Geological Society of Australia, Abstracts, 43,180- 185.

Fielding, C.R., 2006. Upper flow regime sheets, lenses and scour fills: Extending the range of architectural elements for fluvial sediment bodies. Sedimentary Geology, 190(1), 227-240.

Fielding, C.R., 1992. A review of Cretaceous coal-bearing sequences in Australia. In: McCabe, P.J., Parish, J.T. (Eds.), Controls on the Distribution and Quality of Cretaceous Coals: Special Publication. Geological Society of America, 267, 303–324.

Fielding, C.R., 1996. Mesozoic sedimentary basins and resources in eastern Australia-a review of current understanding. Geological Society of Australia, Abstracts, 43, 180- 5.

Fielding, C.R., Alexander, J., McDonald, R., 1999. Sedimentary facies from ground- penetrating radar surveys of the modern, upper Burdekin River of north Queensland, Australia: consequences of extreme discharge fluctuations. Fluvial Sedimentology VI, 347-362.

183

Finlayson, D.M., Leven J.H., Etheridge M.A., 1988. Structural styles and basin evolution in Eromanga region, eastern Australia. American Association of Petroleum Geologists Bulletin, 72, 33-48.

Fletcher, T., Berrell, R.W., Geddes, K., Salisbury, S.W., 2009. Acid preparation of vertebrate fossils (Teleosti and Dinosauria) from the mid-Cretaceous Winton Formation of Isisford, central-western Queensland, Australia. In: Travuillon, K.J., Worthy, T.H., Hand, S.J., Creaser, P. (Eds.), Conference on Australasian Vertebrate Evolution, Palaeontology and Systematics. Geological Society of Australia, The University of New South Wales, p. 74.

Fletcher, T.L., Salisbury, S.W., 2010. New Pterosaur fossils from the Early Cretaceous (Albian) of Queensland, Australia. Journal of Vertebrate Paleontology, 30, 1747– 1759.

Foster, D.R.W., Austin, J.R., 2008. The 1800-1610Ma Stratigraphic and magmatic history of the Eastern Succession, Mount Isa Inlier, and correlations with adacent Paleoproterzoic terranes. Precambrian Research, 163, 7-30.

Frei, D., Gerdes, A., 2009. Precise and accurate in situ U-Pb dating of zircon with high sample throughput by automated LA-SF-ICP-MS. Chemical Geology, 261(3), 261- 270.

Gallagher, K., 1990. Permian to Cretaceous Subsidence history along the Eromanga- Brisbane Geoscience transect, in: Finlayson, D.M., (Ed.), The Eromanga- Brisbane geosciences transect: a guide to basin development across Phanerozoic Australia in southern Queensland, 232, 133-51.

Gallagher, K., Lambeck, K., 1989. Subsidence, sedimentation and sea level changes in the Eromanga Basin Australia. Basin Research, 2(2), 115-131.

Gallagher, K., Dumitru, T.A., Gleadow, A.J., 1994. Constraints on the vertical motion of eastern Australia during the Mesozoic. Basin Research, 6(2‐3), 77-94.

Gehrels, G.E., 2008. Detrital zircon geochronology of Neoproterozoic to Permian miogeoclinal strata in British Columbia and Alberta. Canadian Journal of Earth Sciences, 35(2), 1380-1401.

Gehrels, G.E., 2011. Detrital Zircon U-Pb Geochronology: current methods and new opportunities, In: Busby, C., and Azor, A., (Eds.), Tectonics of Sedimentary Basins: Recent Advances. John Wiley & Sons, 47-62.

Gerdes, A., Zeh, A., 2006. Combined U–Pb and Hf isotope LA-(MC)-ICP-MS analyses of detrital zircons: comparison with SHRIMP and new constraints for the provenance and age of an Armorican metasediment in Central Germany. Earth and Planetary Science Letters, 249(1), 47–61.

184

Ghosh, P., Sarkar, S., Maulik, P., 2006. Sedimentology of a muddy alluvial deposit: Triassic Denwa Formation, India. Sedimentary geology, 191(1), 3-36.

Gibling, M.R., 2006. Width and thickness of fluvial channel bodies and valley fills in the geological record: a literature compilation and classification. Journal of sedimentary Research, 76(5), 731-770.

Glen, R.A., 2005. The Tasmanides of eastern Australia. In: Vaughan, A.P.M., Leat, P.M., Pankhurst, R.J. (Eds.), Terrane processes at the margins of Gondwana. Geological Society Special Publications, 246(1), 23–96.

Gonzalez-Leon, C.M., Valencia, V.A., Lawton, T.F., Amato, O.M., Gehrels, G.E., Leggett, W.J., Montijo-Contreras, O.M., Fernandez, M.A., 2009. The lower Mesozoic record of detrital zircon U–Pb geochronology of Sonora, Mexico, and its paleogeographic implications. Revista Mexicana de Ciencias Geologicas, 26(2), 301–314.

Grandstein, F.M., Ogg, J.G., 2004. Geological time scale 2004 — why, how, and where next. Lethaia, 37(2), 175–181.

Gray, A.R.G., McKillop, M., McKellar, J.L., 2002. Eromanga Basin stratigraphy, 30–56 + references and figures, in: Draper, J. J. (Ed.). Geology of the Cooper and Eromanga Basins, Queensland. Department of Natural Resources and Mines, Brisbane.

Gray, D. R., Foster, D. A., Korsch, R. J., Spaggiari, C. V., 2006. Structural style and crustal architecture of the Tasmanides of eastern Australia: Example of a composite accretionary orogen. Special Papers-Geological Society of America, 414, 119.

Green, P.M., Hawkins, P.J., Carmichael, D.C., Brian, T.J., Smith, R.J., Quarantotto, P., Genn, D.L.P., Hoffmann, K.L., McKellar, J.L., John, B.H., Shield, C.J., 1991. An assessment of hydrocarbon potential of the northern Eromanga Basin, Queensland. Queensland Government Mining Journal, 24-45. Green, T.J., Fergusson, C.L., Withnall, I.W., 1998. Refolding and strain in the Neoproterozoic‐ Early Palaeozoic Anakie Metamorphic Group, Central Queensland. Australian Journal of Earth Sciences: An International Geoscience Journal of the Geological Society of Australia, 45(6), 915-924.

Gurnis, M., Mueller, R.D., Moresi, L., 1998. Dynamics of Cretaceous vertical motion of Australia and the Australian-Antarctic discordance. Science, 279(5356), 1499–1504.

Guynn, J., Gehrels, G., 2010. Comparison of detrital zircon age distributions using the K–S test visualization and representation of age-distribution data histograms. University of Arizona

Hallsworth, C.R., Morton, A.C., Claoue-Long, J., Fanning, C.M., 2000. Carboniferous sand provenance in the Pennine Basin, UK: constraints from heavy mineral and detrital

185

zircon age data. Sedimentary Geology, 137(3), 147–185.

Hawlader, H. M., 1990. Diagenesis and reservoir potential of volcanogenic sandstones— Cretaceous of the Surat Basin, Australia. Sedimentary Geology, 66(3), 181-195.

Heine, C., Dietmar Müller, R., Steinberger, B., Torsvik, T.H., 2008. Subsidence in intracontinental basins due to dynamic topography. Physics of the Earth and Planetary Interiors, 171(1), 252-264.

Helby, R., Morgan, R., Partridge, A.D., 1987. A palynological zonation of the Australian Mesozoic. Memoir of the Association of Australasian Palaeontologists, 4, 1–94.

Helland-Hansen, W., Gjellberg, J.G., 1994. Conceptual basis and variability in sequence stratigraphy: a different perspective. Sedimentary Geology, 92(1), 31-52.

Henderson, R.A., Stephenson, P.J., 1980. The geology and geophysics of northeastern Australia. Geological Association of Australia, Queensland Division.

Herve, F., Calderon, M., Fanning, C.M., Pankhurst, R.J., Godoy, E., 2013. Provenance variations in the Late Paleozoic accretionary complex of central Chile as indicated by detrital zircons. Gondwana Research, 23 (3), 1122–1135.

Hietpas, J., Samson, S., Moecher, D., Chakraborty, S., 2011. Enhancing tectonic and provenance information from detrital zircon studies : assessing terrane-scale sampling and grain-scale characterization. Journal of the Geological Society, 168(2), 309-318.

Hieu, P.T., Chen, F., Me, L.T., Thuy, N. T.B., Siebel, W., Lan, T.G., 2012. Zircon U–Pb ages and Hf isotopic compositions from the Sin Quyen Formation: the Precambrian crustal evolution of northwest Vietnam. International Geology Review, 54(13), 1548- 1561.

Higgins, K., Hashimoto, T., Fraser, G., Rollet, N., Colwell, J., 2011. Ion microprobe (SHRIMP) U–Pb dating of Upper Cretaceous volcanics from the northern Lord Howe Rise, Tasman Sea. Australian Journal of Earth Sciences, 58(2), 195-207. Hocknull, S.A., 1997. Cretaceous freshwater bivalves from Queensland. Memoirs of the Queensland Museum, 42, 223–226.

Hocknull, S.A., 2000. Mesozoic freshwater and estuarine bivalves from Australia. Memoirs of the Queensland Museum, 45, 405–426.

Hocknull, S.A., Cook, A.G., 2008. Hypsilophodontid (Dinosauria: Ornithischia) from the latest Albian, Winton Formation, central Queensland. Memoirs of the Queensland Museum, 52, 212.

186

Hocknull, S.A., White, M.A., Tischler, T.R., Cook, A.G., Calleja, N.D., Sloan, T., Elliott, D.A., 2009. New mid-Cretaceous (Latest Albian) dinosaurs from Winton, Queensland, Australia. PLoS One, 4(7).

Holcombe, R.J., Stephens, C.J., Fielding, C.R., Gust, D., Little, T.A., Sliwa, R., McPhie, J., Ewart, A., 1997a. Tectonic evolution of the northern New England Fold Belt: Carboniferous to Early Permian transition from active accretion to extension. Tectonics and metallogenesis of the New England Orogen, 19, 66-79.

Holcombe, R. J., Stephens, C. J., Fielding, C. R., Gust, D., Little, T. A., Sliwa, R., Kassan, J., McPhie, J., Ewart, A., 1997b. Tectonic evolution of the northern New England Fold Belt: the Permian–Triassic Hunter–Bowen event. Tectonics and metallogenesis of the New England Orogen, 19, 52-65.

Hoskin, P.W.O., Ireland, T.R., 2000. Rare earth element chemistry of zircon and its use as a provenance indicator. Geology, 28(7), 627–630.

Hower, T.L., 2011. Evolution of the Sentry Petroleum Ltd Talundilly-1 Coal Seam Methane Project, NI 51-101 Evaluation. MHA Petroleum Consultants LCC, 1-28.

Hutton, L.J., Draper, J.J., Rienks, I.P., Withnall, I.W., Knutson, J., 1997. Charters Towers region. In: Bain, J.H.C. & Draper, J.J. (eds) North Queensland Geology. Australian Geological Survey Organisation Bulletin, 240,165–224.

Ireland, T.R., Flottmann, T., Fanning, C.M., Gibson, G.M., Preiss, W.V., 1998. Development of early Paleozoic Pacific Margin of Gondwana from detrital-zircon ages across the Delamerian orogeny. Geology, 26(3), 243–246.

Isern, A.R., Anselmetti, F.S., Blum, P., 2002. Leg 194 of the cruises of the Drilling Vessel, JOIDES Resolution Towsnville, Australia, to Apra Harbor, Guam, Sites 1192-1199, 3 January - 2 March 2001, proceedings of the Ocean Drilling Program. Initial Reports, 194.

Jack, R.L., 1886. Handbook of Queensland geology. Geological Survey of Queensland. Publication, 31.

Jackson, S.E., Pearson, N.J., Griffin, W.L., Belousova, E.A., 2004. The application of laser ablation-inductively coupled plasma-mass spectrometry to in situ U-Pb zircon geochronology. Chemical Geology, 211(1), 47-69.

Jell, P.A., 2004. The fossil insects of Australia. Memoirs of the Queensland Museum, 50, 1–124.

Jinnah, Z.A., Roberts, E.M., Deino, A.L., Larsen, J.S., Link, P.K., Fanning, C.M., 2009. New Ar–Ar and detrital zircon U–Pb ages for the Upper Cretaceous Wahweap and Kaiparowits formation on the Kaiparowits Plateau, Utah: implications for regional 187

correlation, provenance, and biostratigraphy. Cretaceous Research, 30(2), 287–299.

Johnston, S., Gehrels, G., Valencia, V., Ruiz, J., 2009. Small-volume U–Pb zircon geochronology by laser ablation-multicollector-ICP-MS. Chemical Geology, 259(3), 218–229.

Jones, J.V., Connelly, J.N., Karlstrom, K.E., Williams, M.L., Doe, M.F., 2009. Age, provenance, and tectonic setting of Paleoproterozoic quartzite successions in the southwestern United States. Geological Society of America Bulletin, 121(1–2), 247–264.

Kemp, A., 1997. Four species of Metaceratodus (Osteichthyes: Dipnoi, Family Cetatodontidae) from Australian Mesozoic and Cenozoic deposits. Journal of Vertebrate Paleontology, 17, 26–33.

Kemp, A. I. S., Hawkesworth, C. J., Collins, W. J., Gray, C. M., Blevin, P. L., 2009. Isotopic evidence for rapid continental growth in an extensional accretionary orogen: The Tasmanides, eastern Australia. Earth and Planetary Science Letters, 284(3), 455- 466.

King, S.J., Mee, B.C., 2004. The seismic stratigraphy and petroleum potential of the Late Cretaceous Ceduna Delta, Ceduna Sub-basin, Great Australian Bight. In Eastern Australasian Basins Symposium II, Petroleum Exploration Society of Australia, Special Publication, 63-73.

Kingston, D.R., Dishroon, C.P., Williams, P.A., 1983. Global basin classification system. AAPG bulletin, 67(12), 2175-2193.

Kinny, P.D., Maas, R., 2003. Lu–Hf and Sm–Nd isotope systems in zircon. Reviews in Mineralogy and Geochemistry, 53(1), 327-341.

Köppen, A., Carter, A., 2000. Constraints on provenance of the central European Triassic using detrital zircon fission track data. Palaeogeography, Palaeoclimatology, Palaeoecology, 161(1), 193-204.

Kowallis, B.J., Christiansen, E.H., Deino, A.L., Peterson, F., Turner, C.E., Kunk, M.J., Obradovich, J.D., 1998. The age of the Morrison Formation. Modern Geology, 22(1- 4), 235–260.

Kraus, M., Aslan, A., 1993. Eocene hydromorphic paleosols: significance for interpreting ancient floodplain processes. Journal of Sedimentary Petrology 63(3), 434–463.

Krieg, G.W., Rogers, P.A., 1995. Stratigraphy marine succession. In: Drexel, J.F. and Preiss, W.V., (Eds.) The geology of South Australia. Volume 2, The Phanerozoic. South Australia Geological Survey Bulletin, 54, 112-122.

188

Kuang, K.S., 1985. History and style of the Cooper–Eromanga Basin structures. Australian Society of Exploration Geophysicists Bulletin, 16(2-3), 245–248.

Kusuhashi, N., Matsumoto, A., Murakami, M., Tagami, T., Hirata, T., Iizuka, T., Handa, T., Matsuoka, H., 2006. Zircon U–Pb ages from tuff beds of the upper Mesozoic Tetori Group in the Shokawa district, Gifu prefecture, central Japan. Island Arc, 15, 378–390.

Lawrence, R.L., Cox, R., Mapes, R.W., Coleman, D.S., 2011. Hydrodynamic fractionation of zircon age population. GSA Bulletin, 123(1-2), 295-305.

Lawton, T.F., Bradford, B.A., 2011. Correlation and provenance of Upper Cretaceous (Campanian) fluvial strata, Utah, USA, from zircon U-Pb geochronology and petrography. Journal of Sedimentary Research, 81(7), 495–512.

Lawton, T.F., Hunt, G.J., Gehrels, G.E., 2010. Detrital zircon record of thrust belt unroofing in Lower Cretaceous synorogenic conglomerates, central Utah. Geology, 38(5), 436–466.

Leahey, L.G., Salisbury, S.W., 2013. First evidence of ankylosaurian dinosaurs (Ornithischia: Thyreophora) from the mid-Cretaceous (late Albian–Cenomanian) Winton Formation of Queensland, Australia. Alcheringa: An Australasian Journal of Palaeontology, 37(2), 249-257.

Li, P., Rosenbaum, G., Vasconcelos, P., 2014. Chronological constraints on the Permian geodynamic evolution of eastern Australia. Tectonophysics.

Link, P.K., Fanning, C.M., Beranek, L.P., 2005. Reliability and longitudinal change of detrital-zircon age spectra in the Snake River system, Idaho and Wyoming: an example of reproducing the bumpy barcode. Sedimentology Geology, 182(1), 101–142.

Little, T.A., Holcombe, R.J., Gibson, G.M., Offler, R., Gans, P.B., McWilliams, M.O., 1992. Exhumation of late Paleozoic blueschists in Queensland, Australia, by extensional faulting. Geology, 20(3), 231-234.

Ludwig, K.R., 2009. User's Manual for Isoplot 3.70. A Geochronological Toolkit for Microsoft Excel: Berkley Geochronology Centre Special Publication, No.4.

Ludwig, K.R., Mundil, R., 2002. Extracting reliable U–Pb ages and errors from complex populations of zircons from Phanerozoic tuffs. Journal Conference Abstract 12th Goldschmidt Conference.

Luyendyk, B.P., 1995. Hypothesis for Cretaceous rifting of east Gondwana caused by subducted slab capture. Geology, 23(4), 373-376 189

MacDonald, J.D., Holford, S.P., Green, P.F., Duddy, I.R., King, R.C., Backé, G., 2013. Detrital zircon data reveal the origin of Australia’s largest delta system. Journal of the Geological Society, 170(1), 3-6.

MacEachern, J.A., Pemberton, S.G., Gingras, M.K., Bann, K.L., 2010. Ichnology and facies models. In: James, N.P., Dalrymple, R.W. (Eds.), Facies Models, edition 4: Geological Association of Canada, St. Johns, Newfoundland, 19–58.

Maloney, K., Baptista, L., Vlok, N., Warboys, I., 2012. Geological Review, Longreach Project. Moultrie Group, Baru Resources Ltd.

Mannion, P.D., Benson, R.B., Upchurch, P., Butler, R.J., Carrano, M.T., Barrett, P.M., 2012. A temperate palaeodiversity peak in Mesozoic dinosaurs and evidence for Late Cretaceous geographical partitioning. Global Ecology and Biogeography, 21(9), 898- 908.

Mannion, P.D., Upchurch, P., Barnes, R.N., Mateus, O., 2013. Osteology of the Late Jurassic Portuguese sauropod dinosaur Lusotitan atalaiensis (Macronaria) and the evolutionary history of basal titanosauriforms. Zoological Journal of the Linnean Society, 168(1), 98-206.

Mannion, P.D., Upchurch, P., Benson, R.B., Goswami, A., 2014. The latitudinal biodiversity gradient through deep time. Trends in ecology & evolution, 29(1), 42-50.

Matthews, K.J., Hale, A.J., Gurnis, M., Müller, R.D., DiCaprio, L., 2011. Dynamic subsidence of Eastern Australia during the Cretaceous. Gondwana Research, 19(2), 372-383.

Mathur, S.P., 1983. Deep crustal reflection results from the central Eromanga Basin, Australia. In: Friedman, M., Toksoz, M.N., (Eds). Continental Tectonics: Structure, Kinematics and Dynamics. Tectonophysics, 100(1), 163-173.

Mavromatidis, A., Hillis, R., 2005. Quantification of exhumation in the Eromanga Basin and its implications for hydrocarbon exploration. Petroleum Geoscience, 11(1), 79- 92.

McLoughlin, S., Drinnan, A.N., Rozefelds, A.C., 1995. A Cenomanian flora from the Winton Formation, Eromanga Basin, Queensland, Australia. Memoirs of the Queensland Museum, 38, 273–331.

McLoughlin, S., Pott, C., Elliott, D., 2010. The Winton Formation flora (Albian– Cenomanian, Eromanga Basin): implication for vascular plant diversification and decline in the Australian Cretaceous. Alcheringa, 34(3), 303–323.

190

Miall, A.D., 1977. A review of the braided river depositional environment. Earth-Science Reviews, 13(1), 1-62.

Miall, A.D., 1985. Architectural-element analysis: a new method of facies analysis applied to fluvial deposits. Earth Science Reviews, 22(4), 261–308.

Miall, A.D., 1996. The geology of fluvial deposits (Vol. 575). Berlin: Springer.

Miall, A.D, 2014. The Facies and Architecture of Fluvial Systems. In Fluvial Depositional Systems, 9-68. Springer International Publishing.

Middleton, M. F., 1989. A model for the formation of intracratonic sag basins.Geophysical Journal International, 99(3), 665-676.

Mitchell, A.H.G., Reading H.G., 1986. Sedimentation and Tectonics. In: H.G Reading (Ed.), Sedimentary Environment, Blackwell, Oxford, 471–519.

Moecher, D.P., Samson, S.D., 2006. Differential zircon fertility of source terranes and natural bias in the detrital zircon record: Implications for sedimentary provenance analysis. Earth and Planetary Science Letters, 247(3), 252-266.

Molnar, R.E., 1980. Australian Late Mesozoic terrestrial tetrapods: some implications. Mémoires de la Société Géologique de France, 139, 131–143.

Molnar, R.E., 1991. Fossil reptiles in Australia. In: Vickers-Rich, P., Monaghan, J.M., Baird, R.F., Rich, T.H. (Eds.), Vertebrate Palaeontology of Australasia. Pioneer Design Studio, Melbourne, 606–702.

Molnar, R.E., 2001. A reassessment of phylogenetic position of Cretaceous sauropod dinosaurs from Queensland, Australia. Association Paleontologica Argentina, Publicacion Especial, 7, 139–144.

Molnar, R.E., 2010. Taphonomic observations on the eastern Australian Cretaceous sauropods. Alcheringa 34(3), 421–429.

Molnar, R.E., 2011. New morphological information about Cretaceous sauropod dinosaurs from the Eromanga Basin, Queensland, Australia. Alcheringa, 35(2), 329–339.

Molnar, R.E., Salisbury, S.W., 2005. Observations on Cretaceous sauropods from Australia. In: Carpenter, K., Tidwell, V. (Eds.), Thunder-Lizards: The Sauropodomorph Dinosaurs. Indiana University Press, Bloomington and Indianapolis, 454–465.

Molnar, R.E., Willis, P.M.A., 1996. A neosuchian crocodile from the Queensland Cretaceous. Journal of Vertebrate Paleontology, 16, 54A.

191

Morgan, R., 1980. Palynostratigraphy of the Australian Early and middle Cretaceous. Geographical Survey of New South Wales Paleontological Memoir, 18, 1–153.

Moore, P.S., Pitt, G.M., 1984. Cretaceous of the Eromanga Basin - implications for hydrocarbon exploration. The APEA Journal, 24(1), 358-376.

Moore, P.S., Pitt, G.M., Dettmann, M.E., 1986: The Early Cretaceous Coorikiana Sandstone and Toolebuc Formation: their recognition and stratigraphic relationship in the south-western Eromanga Basin. In: Gravestock, D.I., Moore, P.S. & Pitt, G.M. (Editors): Contributions to the geology and hydrocarbon potential of the Eromanga Basin. Geological Society of Australia, Special Publication, 12, 97–114.

Mortimer, N., Tulloch, A.J., Gans, P., Calvert, Walker, N., 1999. Geology and thermochronometry of the east edge of the Median Tectonic Batholith (Median Tectonic Zone): a new perspective on Permian to Cretaceous crustal growth of New Zealand. Island Arc, 8(3), 404–425.

Mortimer, N., Hauff, F., Calvert, A.T., 2008. Continuation of the New England Orogen, Australia, beneath the Queensland Plateau and Lord Howe Rise. Australian Journal of Earth Sciences, 55(2), 195-209.

Moss, F.J., Wake-Dyster, K.D., 1983. The Australian central Eromanga Basin project: an introduction. In: M. Friedman and M.N. Toksbz (Eds). Continental Tectonics: Structure, Kinematics and Dynamics. Tectonophysics, 100, 131-145.

Muir, M.D., 1966. Reworking in Jurassic and Cretaceous spore assemblages. Review of Palaeobotany and Palynology, 5(1), 145–154.

Müller, R.D., Gaina, C., Clark, S., 2000a. Seafloor spreading around Australia. In: J Veevers (ed.) Billion year earth history of Australia and neighbors’ in Gondwanaland, GEMOC Press, Departmenl of Earth and Planetary Sciences. Macquarie University, Sydney, 34-50.

Müller, R.D., Gaina, C., Tikku, A., Mihut, D., Cande, S.C., Stock, J., 2000b. Mesozoic/ Cainozoic tectonic events around Australia, Geophysical Monographs of the American Geophysico/ Union, 106, 161-85.

Murgulov, V., Beyer, E., Griffin, W.L., O’Reilly, S.Y., Walters, S.G., Stephens, D., 2007. Crustal evolution in the Georgetown Inlier, North Queensland, Australia: a detrital zircon grain study. Chemical Geology, 245(3),198–218.

Murray, C.G., Kirkegaard, A.G., 1978. The Thomson Orogen of the Tasman Orogenic Zone. Tectonophysics, 4(B), 299-325.

192

Murray, C.G., Fergusson, C.L., Flood, P.G., Whitaker, W.G., Korsch, R.J., 1987. Plate tectonic model for the Carboniferous evolution of the New England Fold Belt. Australian Journal of Earth Sciences, 34, 213-36.

Murray, C.G., Scheibner, E., Walker, R.N., 1989. Regional geological interpretation of a digital coloured residual Bouguer gravity image of eastern Australia with a wavelength cut‐off of 250 km, Australian Journal of Earth Sciences: An International Geoscience Journal of the Geological Society of Australia, 36(3), 423-449.

Musser, A.M., Luo, Z.-X., Martinelli, A.G., Lamanna, M.C., Weisbecker, V., Wroe, S., Salisbury, S.W., 2009. First Australian non-mammalian cynodont: new evidence For the unusual nature of Australia's Cretaceous Vertebrates. In: Travuillon, K.J., Worthy, T.H., Hand, S.J., Creaser, P. (Eds.), Conference on Australasian Vertebrate Evolution, Palaeontology and Systematics. Geological Society of Australia. The University of New South Wales, p. 47.

Næraa, T., Scherstén, A., Rosing, M.T., Kemp, A.I.S., Hoffmann, J.E., Kokfelt, T.F., Whitehouse, M.J., 2012. Hafnium isotope evidence for a transition in the dynamics of continental growth 3.2 [thinsp] Gyr ago. Nature, 485(7400), 627-630.

Nadon, G.C., 1994. The genesis and recognition of anastomosed fluvial deposits: data from the St. Mary River Formation, southwestern Alberta, Canada. Journal of Sedimentary Research, 64(4).

Nichols, G.J., Fisher, J.A., 2007. Processes, facies and architecture of fluvial distributary system deposits. Sedimentary Geology, 195(1), 75-90.

Ogg, J.G., Hinnov, L.A., 2012. Cretaceous. In: Grandstein, F.M., Ogg, J.G., Schmitz, M.D., Ogg, G.M. (Eds.), The Geological Time Scale 2012. Elsevier, Kidlington, Oxford, 794–814.

Partridge, A.D., 2006. Jurassic–Early Cretaceous spore–pollen and dinocyst zonations for Australia. Australian Mesozoic and Cenozoic Palynology Zonations—Updated to the 2004 Geologic Time Scale (coord E. Monteil): Geoscience Australia.

Peters, M.D., 1985. A taxonomic analysis of a mid-Cretaceous megaplant assemblage from Queensland. University of Adelade PhD thesis, Adelade, South Australia, Australia.

Peters, M.D., Christophel, D.C., 1978. Austroseqoia wintonensis, a new taxodiaceous cone from Queensland, Australia. Canadian Journal of Botany 56, 3119–3128.

Pickard, A.L., Adams, J., Barley, M.E., 2000. Australian provenance for Upper Permian to Cretaceous rocks forming accretionary complexes on the New Zealand sector of the Gondwanaland margin. Australian Journal of Earth Sciences, 47(6), 987–1007.

193

Pirajno, F., Hoatson, D.M., 2012. A review of Australia's Large Igneous Provinces and associated mineral systems: Implications for mantle dynamics through geological time. Ore Geology Reviews, 48, 2-54.

Pole, M.S., 1998. Latest Albian–earliest Cenomanian monocotyledonous leaves from Australia. Botanical Journal of the Linnean Society, 129, 177–186.

Pole, M.S., 2000a. Dicotyledonous leaf macrofossils from the latest Albian–earliest Cenomanian of the Eromanga Basin, Queensland, Australia. Paleontological Research, 4, 39–52.

Pole, M.S., 2000b. Mid-Cretaceous conifers from the Eromanga Basin, Australia. Australian Systematic Botany, 13, 153–197.

Pole, M.S., Douglas, J.G., 1999. Bennettitales, Cycadales and Ginkgoales from the mid- Cretaceous of the Eromanga Basin, Queensland Australia. Cretaceous Research, 20(5), 523–538.

Pontén, A., Plink‐Björklund, P.I.R.E.T., 2007. Depositional environments in an extensive tide‐influenced delta plain, Middle Devonian Gauja Formation, Devonian Baltic Basin. Sedimentology, 54(5), 969-1006.

Price, P.L., Filatoff, J., Williams, A.J., Pickering, S.A. Wood, G.R., 1985. Late Palaeozoic and Mesozoic palynostratigraphical units. CSR Oil & Gas Division, Palynology Facility, Palynological Laboratory Report 274/25. Report held by the Department of Natural Resources and Mines Queensland as CR 14012.

Rainbird, R.H., Hamilton, M.A., Young, G.M., 2001. Detrital zircon geochronology and proveance of the Torridonian, NW Scotland. Journal of Geology, 158(1), 15–27.

Retallack, G.J., 1997. Colour guide to paleosols. John Wiley & Sons Ltd.

Roback, R.C., Walker, N.W., 1995. Provenance, detrital zircon U–Pb geochronometry, and tectonic significance of Permian to Lower Triassic sandstone in southeastern Quesnellia, Brittish Columbia and Washington. Geological Society of America Bulletin, 107(6), 665–675.

Roberts, E.M., Hendrix, M.S., 2000. Taphonomy of a petrified forest in the Two Medicine formation (Campanian), northwest Montana: Implications for palinspastic restoration of the Boulder batholith and Elkhorn Mountains volcanics. Palaios, 15(5), 476-482.

Roberts, E.M., 2007. Facies architecture and depositional environments of the Upper Cretaceous Kaiparowits Formation, southern Utah. Sedimentary Geology, 197(3), 207-233.

194

Roberts, E.M., Stevens, N.J., O’Connor, P.M., Dirks, P.H.G.M., Gottfried, M.D., Clyde, W.C., Armstrong, R.A., Kemp, A.I.S., Hemming, S., 2012. Initiation of the western branch of the East African Rift coeval with the eastern branch. Nature Geoscience, 5(4), 289-294.

Robinson, A.C., Ducea, M., Lapen, T.J., 2012. Detrital zircon and isotopic constraints on the crustal architecture and tectonic evolution of the northern Pamir. Tectonics 31, TC2016.

Romilio, A., Salisbury, S.W., 2011. A reassessment of large theropod dinosaur tracks from the mid-Cretaceous (late Albian–Cenomanian) Winton Formation of Lark Quarry, central-western Queensland, Australia: a case for mistaken identity. Cretaceous Research 32(2), 135–142.

Romilio, A., Tucker, R.T., Salisbury, S.W., 2013. Reevaluation of the Lark Quarry dinosaur Tracksite (late Albian–Cenomanian Winton Formation, central-western Queensland, Australia): no longer a stampede?. Journal of Vertebrate Paleontology, 33(1), 102-120.

Russell, M., Gurnis, M., 1994. The planform of epeirogeny: vertical motions of Australia during the Cretaceous. Basin Research, 6(2‐3), 63-76.

Salgado, L., Coria, R.A., Calvo, J.O., 1997. Evolution of titanosaurid sauropods: Phytogenetic analysis based on the postcranial evidence. Ameghiniana, 34(1), 3-32.

Salisbury, S.W., 2003. Theropod teeth from the Lower Cretaceous Winton Formation, central-western Queensland, Australia. Conference on Australian Vertebrate Evolution, Palaeontology, and Systematics, Queensland Museum, Brisbane, July 7– 11 2003. Abstracts.

Salisbury, S.W., 2005. A new vertebrate assemblage from the mid-Cretaceous (Albian- Cenomanian) Winton Formation, central-western Queensland. In: Reed, L., Bourne, D., Megirian, D., Prideaux, G., Young, G., Wright, A. (Eds.), Proceedings of CAVEPS 2005. Alcheringa, Naracoorte Caves World Heritage Area, South Australia, p. 465.

Salisbury, S.W., Molnar, R.E., Frey, E., Willis, P.M.A., 2006a. The origin of modern crocodyliforms: new evidence from the Cretaceous of Australia. Proceedings of the Royal Society of London Series B 273, 2439–2448.

Salisbury, S.W., Molnar, R.E., Lamanna, M.C., 2006b. A new titanosauriform sauropod from the mid-Cretaceous (Albian-Cenomanian) Winton Formation of centralwestern Queensland, Australia. Journal of Vertebrate Paleontology, 26, 118A.

Salisbury, S.W., Molnar, R.E., Lamanna, M.C., 2007. Sauropods from the mid-Cretaceous (Albian-Cenomanian) Winton Formation of central-western Queensland, Australia.

195

In:Warren, A. (Ed.), Conference on Australiasian Vertebrate Evolution, Palaeontology and Systematics 2007. Geological Society of Australia Abstracts, 85, 26–27.

Salisbury, S.W., Rich, T.H.V., Vickers-Rich, P., Currie, P.J., 2011. Australian Cretaceous non-avian theropod dinosaur teeth. 13th Confernce on Australasian Vertebrate Evolution, Palaeontology and Systematics. Geological Survey of Western Australia, Perth, p. 73.

Scanlon, J.D., Hocknull, S.A., 2008. A dolichosaurid lizard from the latest Albian (mid- Cretaceous) Winton Formation, Queensland, Australia. Transactions of the Kansas Academy of Science (Fort Hays Studies Special Issue - Proceedings of the Second Mosasaur Meeting), 131–136.

Scheibner, E., Veevers, J.J., 2000. Tasman Fold Belt System. In J.J. Veevers (ed.) Billion- year history of Australia and neighbours in Gondwanaland, GEMOC Press, Department of Earth and Planetary Sciences, Macquarie University, Sydney, 154-234

Schellart, W.P., Lister, G.S., Toy, V.G., 2006. A Late Cretaceous and Cenozoic reconstruction of the Southwest Pacific region: tectonics controlled by subduction and slab rollback processes. Earth-Science Reviews, 76(3), 191-233.

Sdrolias, M., Müller, R. D., Gaina, C., 2001. Plate tectonic evolution of eastern Australian marginal ocean basins. In Eastern Australian Basins Symposium, 227-237.

Segal, I., Halicz, L., Platzner, I.T., 2003. Accurate isotope ration measurements of ytterbium by multi-collector inductively coupled plasma mass spectrometry applying erbium and hafnium in an improved double external normalization procedure. Journal of Analytical Atomic Spectrom, 18, 1217-1223.

Senior, B.R., Exon, N.F., Burger, D., 1975. The Cadna-owie and Toolebuc Formations in the Eromanga Basin, Queensland. Queensland Government Mining Journal, 76, 445– 455. Senior, B.R., Mond, A., Harrison, P.L., 1978. Geology of the Eromanga Basin. Australian Bureau of Mineral Resources. Geology and Geophysics Bulletin, 167, 1-102.

Senior, B.R., Mabbutt, J.A., 1979. A proposed method of defining deeply weathered rock units based on regional geological mapping in southwest Queensland. Journal of the Geological Society of Australia, 26(5-6), 237-254.

Sentry Petroleum Ltd., 2011a. February 11, 2011 Sentry Petroleum Uncovers Potential 2,000 Square Mile Coal Deposit in World Class Coal Seam Gas Region. Retrieved on the 16/08/2012, from http://www.sentrypetroleum.com/sentry-petroleum-uncovers- potential-2000-square-mile-coal-deposit-in-world-class-coal-seam-gas-region.

196

Sentry Petroleum Ltd., 2011b. Current Drilling. Retrieved on the 16/08/2012 from: http://www.sentrypetroleum.com/projects/unconventional/current-drilling

Sereno, P.C., 1997. The origin and evolution of dinosaurs. Annual Review of Earth and Planetary Sciences, 25(1), 435–489.

Sereno, P.C., Wilson, J.A., Conrad, J.K., 2004. New dinosaur link southern land masses in the Mid-Cretaceous. The Royal Society, 271, 1325–1330.

Seton M., Müller R.D., Zahirovic S., Gaina C., Torsvik T., Shephard G., Talsma A., Gurnis M., Turner M., Maus S., Chandler M., 2012. Global continental and ocean basin reconstructions since 200 Ma. Earth-Science Reviews, 113(3), 212–270.

Skelton, P.W., 2003. The cretaceous world. Cambridge university press. 361 pgs.

Söderlund, U., Elming, S. Å., Ernst, R. E., Schissel, D., 2006. The Central Scandinavian Dolerite Group—Protracted hotspot activity or back-arc magmatism?: Constraints from U–Pb baddeleyite geochronology and Hf isotopic data. Precambrian Research, 150(3), 136-152.

Stampfli, G.M., Borel, G.D., 2002. A plate tectonic model for the Paleozoic and Mesozoic constrained by dynamic plate boundaries and restored synthetic oceanic isochrons. Earth and Planetary Science Letters, 196(1), 17-33. Surpless, K.D., Graham, S.A., Covault, J.A., Wooden, J.L., 2006. Does the Great Valley Group contain Jurassic strata? Reevaluation of the age and early evolution of a classic foreland basin. Geology, 34, 21–24.

Sutherland, R., 1999. Basement geology and tectonic development of the greater New Zealand region: an interpretation from regional magnetic data. Tectonophysics, 308(3), 341–362.

Thulborn, R.A., Wade, M., 1979. Dinosaur stampede in the Cretaceous of Queensland. Lethaia, 12(3), 275–279.

Thulborn, R.A., Wade, M., 1984. Dinosaur trackways in the Winton Formation (Mid- Cretaceous) of Queensland. Memoirs Queensland Museum, 21, 413–517.

Thulborn, R.A., Wade, M., 1989. A footprint as a history of movement, In: Gillette, D.D. and Lockley M.G. (Eds.), Dinosaur tracks and traces. Cambridge University Press, 51-56.

Tucker, R.T., Roberts, E.M., Hu, Y., Kemp, A.I., Salisbury, S.W., 2013. Detrital zircon age constraints for the Winton Formation, Queensland: Contextualizing Australia's Late Cretaceous dinosaur faunas. Gondwana Research, 24(2), 767-779.

197

Tucker, R.T., 2014. Stratigraphy, Sedimentation and Age of the Upper Cretaceous Winton Formation, central-western Queensland, Australia: Implications for regional palaeogeography, palaeoenvironments and Gondwanan palaeontology. Ph.D thesis unpublished

Tulloch, A.J., Ramezani, J., Mortimer, N., Mortensen, J., van den Bogaard, P., Maas, R., 2009. Cretaceous felsic volcanism in New Zealand and Lord Howe Rise (Zealandia) as a precursor to final Gondwana break-up. Geological Society, London, Special Publications, 321(1), 89-118.

Twidale, C.R., 1966. Late Cainozoic activities of the Selwyn upwarp, northwest Queensland. Journal of the Geological Society of Australia, 13(2), 9-10.

Upchurch, P., 2008. Gondwana break-up: legacies of a lost world. Trends in Ecology & Evolution, 23(4), 229–236.

Upchurch, P., Barrett, P.M., Dodson, P., 2007. Sauropoda, In: Weishampel, D.B., Dodson, P., Osmolska, H. (Eds.), The Dinosauria, Second edition. University of California Press, Los Angeles, CA, 259–324. van Hinsbergen, D.J., Steinberger, B., Doubrovine, P.V., Gassmöller, R., 2011. Acceleration and deceleration of India‐Asia convergence since the Cretaceous: Roles of mantle plumes and continental collision. Journal of Geophysical Research: Solid Earth (1978–2012), 116(B6).

Varela, A.N., Poiré, D.G., Martin, T., Gerdes, A., Goin, F.J., Gelfo, J.N., Hoffmann, S., 2012. Zircon U–Pb age of the Cretaceous Mata Amarilla Formation, Southern Patagonia, Argentina: its relationship with the evolution of the Austral Basin. Journal of Andean Geology, 39(3), 359–379.

Veevers, J.J., 1991a. Phanerozoic Australia in the changing configuration of Proto-Pangea through Gondwanaland and Pangea to the present dispersed continents. Australian Systematic Botany, 4(1), 1-11.

Veevers, J.J., 1991b. Mid-Cretaceous tectonic climax, Late Cretaceous recovery, and Cainozoic relaxation in the Australian region. The Cainozoic in Australia: A Re- Appraisal of the Evidence: Geological Society of Australia, Special Publication, 18, 1-15.

Veevers, J.J., Walter, M. R., Scheibner, E., 1997. Neoproterozoic Tectonics of Australia‐ Antarctica and Laurentia and the 560 Ma Birth of the Pacific Ocean Reflect the 400 m.y. Pangean Supercycle. The Journal of Geology, 105(2), 25-242.

Veevers, J.J., 2000a. Change of tectono-stratigraphic regime in the Australian plate during the 99Ma (mid-Cretaceous) and 43 Ma (mid-Eocene) swerves of the Pacific. Geology, 28(1), 47-50.

198

Veevers, J.J. (Ed.), 2000b. Billion-year earth history of Australia and neighbours in Gondwanaland. GEMOC Press, Sydney. 400 pgs.

Veevers, J.J., 2000c. Mid-Cretaceous (99 Ma) and mid-Eocene (43 Ma) events in the Indo- Australian and Antarctic plates and coeval swerves of the Pacific Plate. Geology 28, 47– 50.

Veevers, J.J., 2004. Gondwanaland from 650– 500 Ma assembly through 320 Ma merger in Pangea to 185– 100 Ma breakup: supercontinental tectonics via stratigraphy and radiometric dating. Earth-Science Reviews, 68(1), 1– 132.

Veevers, J.J., 2006. Updated Gondwana (Permian–Cretaceous) earth history of Australia. Gondwana Research, 9(3), 231-260.

Veevers, J.J., 2012. Reconstructions before rifting and drifting reveal the geological connections between Antarctica and its conjugates in Gondwanaland. Earth-Science Reviews, 111(3), 249-318.

Vervoort J.D., Patchett P.J., Blichert-Toft J., Albarede F., 1999. Relationships between Lu–Hf and Sm–Nd isotopic systems in the global sedimentary system. Earth and Planetary Science Letters, 168, 79–99.

Vervoort, J. D., Patchett, P. J., Söderlund, U., Baker, M., 2004. Isotopic composition of Yb and the determination of Lu concentrations and Lu/Hf ratios by isotope dilution using MC-ICPMS. Geochemistry, Geophysics, Geosystems, 5(11), 1-15.

Vine, R.R., Day, R.W., 1965. Nomenclature of the Rolling Downs Groupe, Nothern Eromanga Basin, Queensland. Queensland Government Mining Journal, 65, 416-421.

Vine, R.R., Day, R.W., Casey, D.J., Milligan, E.N., Galloway, M.C., Exon, N.F., 1967. Revised nomenclature of the Rolling Downs Group, Eromanga and Surat Basins. Queensland Government Mining Journal, 68, 144-151.

Walker, J.D., Geissman, J.W., 2009. Geological Time Scale. Geological Society of America. Watson, P.J. 1973. Lovelle Downs No.1 Well Completion Report, A-P 166B, Queensland. Geological Survey of Queensland, Open File Report CR 4326 (unpublished).

Waschbusch P., Korsch R.J., Beaumont C., 2009. Geodynamic modelling of aspects of the Bowen, Gunnedah, Surat and Eromanga Basins from the perspective of convergent margin processes. Australian Journal of Earth Sciences: An International Geoscience Journal of the Geological Society of Australia, 56(3), 309-334

Watterson, J., Walsh, J., Nicol, A., Nell, P. A.R., Bretan, P.G., 2000. Geometry and origin of a polygonal fault system. Journal of the Geological Society, 157(1), 151-162.

199

Williams, A.J., 1983a. Palynology (preliminary report) Morney No. 1. Appendix V in DIO Morney 1 Well Completion Report. Report held by the Department of Natural Resources and Mines Queensland as CR 9551.

Williams, A.J., 1983b. Palynological report Chookoo No. 1. Delhi Petroleum Pty Ltd Palynological Report No. 13/182 as Appendix XII in DIO Chookoo 1 Well Completion Report. Report held by the Department of Natural Resources and Mines Queensland as CR 13118.

Williams, A.J., 1983c. Palynological report Bycoe No. 1. Delhi Petroleum Pty Ltd Palynological Report No. 13/185 as Appendix V in DIO Bycoe 1 Well Completion Report. Report held by the Department of Natural Resources and Mines Queensland as CR 13033.

White, M.A., Cook, A.G., Hocknull, S.A., Sloan, T., Sinapius, G.H.K., Elliott, D.A., 2012. New Forearm Elements Discovered of Holotype Specimen Australovenator wintonensis from Winton, Queensland, Australia. PLoS One, 7(6), 1–28.

Whitehouse, F.W., 1954. Cainozoic and later beds. Queensland Cordinator of Public Works, Report, 1-3.

Whitehouse, F.W., 1955. The geology of the Queensland Portion of the Great Artesian Basin. Artesian water supplies in Queensland. Queensland preliminary papers A. Appendix G, 1-20.

Willner , A.P., Gerdes, A., Massonne, H.-J., 2008. History of crustal growth and recycling at the Pacific convergent margin of South America at latitudes 29°–36° S revealed by a U–Pb and Lu–Hf isotope study of detrital zircon from late Paleozoic accretionary systems. Chemical Geology 253(3), 114–129.

Wilson, J.A., Sereno, P.C., 1998. Early evolution and higher-level phylogeny of sauropod dinosaurs. Journal of Vertebrate Paleontology, 18(S2), 1-79.

Wing, S.L., 1998. Late Paleocene–early Eocene floral and climatic change in the Bighorn Basin, Wyoming. Late Paleocene–Early Eocene Climatic and Biotic Events in the Marine and Terrestrial Records. Columbia University Press, New York, 380-400.

Withnall, I.W., Hutton, L.J., 2013. Noth Australian Craton. In: Jell, P.A., (Ed.) Geology of Queensland. Queensland Government, Brisbane, QLD, Australia, 23-111.

Withnall, I.W., Mackenzie, D.E., Denaro, T.I., Sain, J.H.C., Oversby, B.S., Knutson, J., Donchak, P.I.T., Champion, D.C., Wellman, P., Cruikshank, B.I., Sun, S.S., Pain, C.F., 1997. Georgetown Region. Bulletin of the Australian Geological Survey Organization, 240, 19-116.

200

Woodhead, J. D., Hergt, J.M., 2005. A preliminary appraisal of seven natural zircon reference materials for in situ Hf isotope determination. Geostandards and Geoanalytical Research, 29(2), 183-195.

Wopfner, H., 1985. Some thoughts on the post-orogenic development of northeastern South Australia and adjoining regions. In: Lindsay, J.M. (Ed.), Stratigraphy, palaeontology, malacology, papers in honour of Dr Nell Ludbrook. South Australia. Department of Mines and Energy. Special Publication, 5, 365-372.

Zeh, A., Gerdes, A., 2012. U–Pb and Hf isotope record of detrital zircons from gold- bearing sediments of the Pietersburg Greenstone Belt (South Africa)—Is there a common provenance with the Witwatersrand Basin?. Precambrian Research, 204, 46- 56.

Zirakparvar, N.A., Baldwin, S.L., Vervoort, J.D., 2013. The origin and geochemical evolution of the Woodlark Rift of Papua New Guinea. Gondwana Research, 23(3), 931-943.

201

202

203

Appendix 1

Romilio, A., Tucker, R. T., & Salisbury, S. W. (2013). Reevaluation of the Lark Quarry dinosaur Tracksite (late Albian–Cenomanian Winton Formation, central-western Queensland, Australia): no longer a stampede?. Journal of Vertebrate Paleontology, 33(1), 102-120.

204

205

206

207

208

209