Living near permanent water in the upper Murray-Darling Basin Implications from the micromorphology of buried soils near artesian springs Malcolm Connolly Department of Archaeology University of Cambridge This dissertation is submitted for the degree of Doctor of Philosophy Darwin College March 2022 Declaration I hereby declare that except where specific reference is made to the work of others, the contents of this dissertation are original and have not been submitted in whole or in part for consideration for any other degree or qualification in this, or any other university. This dissertation is my own work and contains nothing which is the outcome of work done in collaboration with others, except as specified in the text and Acknowledgements. This dissertation contains fewer than 65,000 words including appendices, bibliography, footnotes, tables and equations and has fewer than 150 figures. Malcolm Connolly March 2022 Living near permanent water in the upper Murray-Darling Basin Implications from the micromorphology of buried soils near artesian springs Malcolm Connolly Abstract Understanding links between landscape change and early peoples that lived along Eulo Ridge in the upper Murray Darling Basin, Australia are hampered by poor environmental data and chronological frameworks. To address such issues, steep sided gullies (arroyos), soil micromorphology and pedology, and optically stimulated luminescence (OSL) dating provide a framework for the palaeo-human setting. It considers several questions, including: When did landscapes change and why? How did dry and humid conditions effect the circumstances and behaviours of early peoples living in Australia’s drylands? It is hypothesised that early peoples were able to adjust their behaviours to cope with episodic erosional events and successfully occupy drylands for millennia. This thesis implements a multi proxy approach to reveal evidence of oscillating humid and dry conditions from ~55 Ka to ~0.7 Ka. From 55.9 Ka ± 5.9 Ka BP, humid conditions dominated the region which slowly transitioned into cold frosty dry conditions toward the commencement of the last glacial maximum (LGM). Extreme dry cold conditions characterised the LGM (~24 Ka to ~20 Ka) with people living near groundwater fed pools and groups of springs. From ~20 Ka to ~15 Ka, the region was characterised by intermittent rainfall, fires, and people living near springs. From ~15Ka to ~12 Ka, activations of the monsoons brought wet conditions, mass movement of sediments downslope, and intermittent fires across the region. From ~5 Ka to ~0.7 Ka, conditions were humid with significant landscape changes with dune formation, landslides, and burial of the former land surfaces. This was followed by dry conditions and the familiar boom and bust periods with short humid conditions and long dry phases. The last 1,000 years is seen as a period of significant cultural and behavioural change with larger populations and technological innovations. It is argued that early peoples coped with landscape change by adjusting the extent of their home range to confront the opposing impacts of extreme humid and dry conditions, and implementing new technologies such as a greater use of fire to cope with the continual boom and bust phases. Finally, this study demonstrates that micromorphology is a valuable tool for geoarchaeologists to reconstruct both palaeoenvironments and to decipher the behaviours of early peoples living near springs for long periods of time. Dedication I would like to dedicate this thesis to my mother Daphne Joan Connolly (nee Adams). Sadly, Mum passed away on the 2nd of October 2019 while I was away in Cambridge. This dedication is because of her loving nature and support throughout my life. Mum encouraged me to improve myself in all circumstances no matter how big or small the situation, and it through her energy and support that I found myself undertaking an academic career and this Doctor of Philosophy at the University of Cambridge. My mother came from humble beginnings at Goodooga Mission, New South Wales. It is this start to her life that encouraged her children and grandchildren (and now great grandchildren) to rise above these difficult but happy times. I am sure that I have done her proud and that I am able to take-on Mum’s role and encourage other Aboriginal and Torres Strait Islander people to embark on an academic career of their choice. Thanks to my loving mother. Acknowledgements First and foremost I am extremely grateful to my supervisor Professor Charles French for his invaluable advice, continuous support, and patience during my PhD. His immense knowledge and experience encouraged me to follow this study through to the end. I am indebted to his generosity, kindness, and continued support in all circumstances. I am very grateful to Richard Potok, founder and former Chief Executive Officer of the Aurora Project and Aurora Education Foundation, and funding and support received as part of the Charles Perkins, Cambridge Trust, and Chevening scholarships to undertake this PhD. The support of Aurora employees Sharon Kumar, Lola Alexander, and Kris Wilson for their support and encouragement. I thank Dr Richard Robins, Tim and Evelyn Robins at Everick Heritage Consultants for their financial support to obtain essential optically stimulated luminescence dates. I acknowledge and thank the Budjiti People for allowing me to undertake this study on their lands especially Phillip Eulo, Lorna and Elizabeth McNiven, Michael McNiven, and Dinny McKellar, Margaret Hearn, Carolyn Hooper and Melissa Bryan. I also thanks Daniel McKellar for introducing me to the archaeology of the Eulo region and to regularly receive his feedback and generosity for his personal knowledge about country. This has allowed me to formulate research questions and, to undertake this project to better understand our ancestors. I am indebted to Emeritus Professor Ian Lilley, Dr Trudi Tate, Dr Delyna Baxter, Dr Richard Robins and Mr Matthew Connolly who provided comments on drafts and assistance throughout the thesis. I thank my brother Michael Connolly who provided permission for me to use his paintings throughout the thesis. I thank Sam Rando and Megan Howie for their encouragement and support to take the step of leaving full-time work to undertake this PhD. I thank Peter Madden for helping me with the early fieldwork and Pilot Study on Eastern Arrernte lands. I thank people of the Artitjere Community (Harts Range, NT) and staff of the Central Land Council for allowing me to speak to them about my intended PhD research that unfortunately was cut short due to a change in research focus to the Murray Darling Basin and artesian springs. It was unfortunate that I was unable to continue my research in central Australia. I would also like to thank these people for their technical support: Tonko Rajkovaca, David Redhouse, Laura Healy, Ningsheng Wang, Sean Taylor, Tamsin O’Connell, Philip Nigst, Ian 8 | Moffat, Luc Vrydaghs, Yannick Devos, Petros Chatzimpaloglou, Simon Stoddart, Federika Sulas, Philipe De Smedt, and Ana Polo-Diaz. And to my fellow PhD students Jeremy Bennett, Dylan Gaffney, Huiru Lian, David Kay, Mike Lewis, Joanna Walker, and Camila Alday for their support in the Charles McBurney laboratory and chats over coffee. Thanks to Michael Westaway for regular catch ups and chats during his visits to Cambridge. I particularly thank my partner Delyna Baxter who was always there for me to check drafts and to listen to my banter about geoarchaeology and my PhD. I also want thank my son Matthew Connolly for his continued support, patience and encouragement especially at the end when I arrived back in Queensland during the Covid-19 Pandemic. I would like to thank Shellie Cash, Moc Parker and Kevin Bradley at Currawinya National Park for providing assistance and kind friendship during my fieldwork. Their local knowledge and experience were critical to refining my study sites across this unique human and natural landscape. I thank Peter Connelly who gave up his time and resources to help me during my fieldwork and, helping to collect undisturbed soil samples and optically stimulated luminescence dates. I also thank Indigenous Scholars Vince Backhaus, James Beaufils, Nina Cass and partner Ben Thompson and, Olivia Slater and her family, Samara Hand, Graham Akhurst, and Nathan Canuto for being part of larger cohort of Aurora scholars that was part of the 2015 Study Tour and studying at Cambridge and Oxford universities. And, a final thanks to Sean and Sally Finlay, Steve and Troy Baxter, Tim and Elissa Robins, Robin and Sonja Street, Alan Baxter, Alva and Ilma Jeffs, and Jessica and Jamie Smith, and Kate and Meg Heuschele for their personal support and encouragement throughout the completion of this thesis. Table of contents List of figures 15 List of tables 21 1 Introduction to the thesis 25 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 1.2 Study motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 1.3 Organisation of this thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 2 Eulo Ridge Setting 35 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 2.2 Study setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Budjiti People . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Previous archaeological research . . . . . . . . . . . . . . . . . . . . . . . . . 36 2.3 Palaeoenvironmental setting . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 2.4 Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 2.5 Geology, Geomorphology and Soils . . . . . . . . . . . . . . . . . . . . . . . 44 2.6 Hydrology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 2.7 Landforms and land systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 2.8 Vegetation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 2.9 Fauna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 2.10 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 3 Living near water sources 59 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 3.2 Aboriginal Setting: Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Settlement of Sahul (Australia) by early peoples . . . . . . . . . . . . . . . . . 60 How many languages were spoken in upper Murray-Darling Basin? . . . . . . 61 How mobile were early peoples in the northwest MDB? . . . . . . . . . . . . . 65 What behaviours occurred near water? . . . . . . . . . . . . . . . . . . . . . . 66 What types of social behaviours occurred near water? . . . . . . . . . . . . . . 68 12 | Table of contents What types of intangible behaviours occurred near water? . . . . . . . . . . . . 69 Increase Ceremonies: How to bring rain and resources? . . . . . . . . . . . . . 70 3.3 How did early peoples cope during extreme aridity? . . . . . . . . . . . . . . . 72 What were the subsistence strategies? . . . . . . . . . . . . . . . . . . . . . . 74 How did early peoples cope with dryland conditions? . . . . . . . . . . . . . . 75 Coping with Temperature Extremes . . . . . . . . . . . . . . . . . . . . . . . 76 3.4 What happened when Europeans arrived? . . . . . . . . . . . . . . . . . . . . 76 Pastoralism period (1870 to 1991) . . . . . . . . . . . . . . . . . . . . . . . . 76 Post-colonial contact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 3.5 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 4 Case Studies: A multi-proxy approach 81 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 4.2 Preliminary fieldwork and setting up this study . . . . . . . . . . . . . . . . . 82 4.3 Overview of the case studies . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 4.4 Case Study 1: Granites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 4.5 Granites East Study Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Archaeological material: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 4.6 Granites Relict Dune Study Site . . . . . . . . . . . . . . . . . . . . . . . . . 89 Archaeological material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 4.7 Case Study 2: Hardpan Creek and Basin Gully . . . . . . . . . . . . . . . . . . 91 4.8 Hardpan Creek North Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Archaeological material: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 4.9 Hardpan Creek South/Centre Study Site . . . . . . . . . . . . . . . . . . . . . 95 Archaeological materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 4.10 Basin Gully Study Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Basin Gully 1 Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Archaeological materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Basin Gully 2 Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Archaeological materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 4.11 Case Study 3: Double Well and Tunkana Well . . . . . . . . . . . . . . . . . . 100 4.12 Double Well Study Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Archaeological material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 4.13 Tunkana Well Study Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Characteristics of the archaeological record . . . . . . . . . . . . . . . . . . . 107 4.14 Research design and preliminary study methods . . . . . . . . . . . . . . . . . 108 4.15 Field Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Micromorphology and Sediment Sampling . . . . . . . . . . . . . . . . . . . . 109 Optically Stimulated Luminescence Dating . . . . . . . . . . . . . . . . . . . 110 Table of contents | 13 Archaeological materials: identification and assessment . . . . . . . . . . . . . 111 4.16 Laboratory methods and procedures . . . . . . . . . . . . . . . . . . . . . . . 112 Magnetic Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Loss on Ignition (LOI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 4.17 Soil micromorphology methods and techniques . . . . . . . . . . . . . . . . . 113 Textural pedofeature records . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Use of opal phytoliths for site interpretation . . . . . . . . . . . . . . . . . . . 118 4.18 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 5 The Geoarchaeology of Eulo Ridge 121 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 5.2 Case Study 1: Granites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 5.3 Granites East Study Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Micromorphological characteristics . . . . . . . . . . . . . . . . . . . . . . . 122 Sediment/soil characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Granites East Site Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . 130 5.4 Granites Relict Dune Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Micromorphological characteristics . . . . . . . . . . . . . . . . . . . . . . . 132 Age of the deposit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Sediment/soil characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Granites Dune Site Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . 136 5.5 Case Study 2: Hardpan Creek and Basin Gully . . . . . . . . . . . . . . . . . . 140 5.6 Hardpan Creek North Study Site . . . . . . . . . . . . . . . . . . . . . . . . . 140 Micromorphological characteristics . . . . . . . . . . . . . . . . . . . . . . . 140 Age of the deposit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 Sediment/soil characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Hardpan Creek North Site Interpretation . . . . . . . . . . . . . . . . . . . . . 152 5.7 Hardpan Creek Centre and South Study Sites . . . . . . . . . . . . . . . . . . 154 Micromorphological characteristics . . . . . . . . . . . . . . . . . . . . . . . 154 Age of the deposit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Sediment/soil characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Hardpan Creek South Site Interpretation . . . . . . . . . . . . . . . . . . . . . 160 5.8 Basin Gully Study Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Micromorphological characteristics . . . . . . . . . . . . . . . . . . . . . . . 161 Age of the deposits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Sediment/soil characteristics: . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Basin Gully Sites Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . 166 5.9 Case Study 3: Double Well and Tunkana Well . . . . . . . . . . . . . . . . . . 168 5.10 Double Well Study Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 14 | Table of contents Micromorphological characteristics . . . . . . . . . . . . . . . . . . . . . . . 168 Sediment/soil characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Double Well Site Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . 176 5.11 Tunkana Well Study Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 Micromorphological characteristics . . . . . . . . . . . . . . . . . . . . . . . 178 Sediment/soil characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 Tunkana Well Site Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . 182 5.12 Chapter summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 6 Investigating palaeoenvironments: a human occupation model 187 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188 6.2 What evidence explains episodic landscape change? . . . . . . . . . . . . . . . 188 ~60 Ka to ~25 Ka (MIS 3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188 ~25 Ka to ~20 Ka (MIS2) (LGM) . . . . . . . . . . . . . . . . . . . . . . . . 189 ~20 Ka to ~15 Ka (MIS 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 ~15 Ka to ~12 Ka MIS 2–MIS 1 . . . . . . . . . . . . . . . . . . . . . . . . . 191 ~12 Ka to ~5 Ka (MIS 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192 ~5 Ka to present (late MIS 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 6.3 Why did these landscapes change over time? . . . . . . . . . . . . . . . . . . . 196 6.4 What do stone artefacts reveal about site formation processes? . . . . . . . . . 198 6.5 How have opal phytoliths and inorganics informed site interpretation? . . . . . 200 6.6 How has landscape change impacted on early peoples? . . . . . . . . . . . . . 203 6.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 6.8 Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 References 211 Appendix A Field and Laboratory Methods 237 Appendix B Soil Analyses Data and Archaeological Material 251 Appendix C Micromorphology descriptions and Opal phytoliths micrographs 269 List of figures 1.1 Eulo Ridge site map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 1.2 Arroyos found within range landscapes . . . . . . . . . . . . . . . . . . . . . . 28 1.3 Three main types of geomorphic surfaces . . . . . . . . . . . . . . . . . . . . 29 2.1 Selected prominent archaeological sites in the MDB . . . . . . . . . . . . . . . 38 2.2 Location of Eulo Ridge within semi-arid zone . . . . . . . . . . . . . . . . . . 43 2.3 Australian climate systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 2.4 Eulo Ridge is located within the southeast of the Great Artesian Basin . . . . . 47 2.5 GAB aquifers located within the MDB . . . . . . . . . . . . . . . . . . . . . . 48 2.6 Two examples of springs along Eulo Ridge . . . . . . . . . . . . . . . . . . . . 49 2.7 Granite outcrop along southern end of Eulo Ridge . . . . . . . . . . . . . . . . 51 2.8 M4 and H2 land systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 2.9 Landform cross section: Eulo Ridge . . . . . . . . . . . . . . . . . . . . . . . 53 2.10 Distribution of Acacia aneura and Eucalyptus populanea across Australia . . . 55 3.1 Distribution of language groups in the northwest Murray Darling Basin . . . . 63 3.2 Aboriginal group photographed by Frederick Bonney (1884) . . . . . . . . . . 64 3.3 Burials observed by the Mitchell expedition in 1838 . . . . . . . . . . . . . . . 64 3.4 Painting of a family group camped (Charles Sturt expedition 1849) . . . . . . . 68 3.5 Drawing of a man with body scars (Major Mitchell expedition) . . . . . . . . . 72 3.6 Man bathing in a water bore near Cunnamulla, Queensland (1894) . . . . . . . 79 4.1 Eulo ridge case study areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 4.2 Granites East case study area . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 4.3 Granites East: shallow arroyo located on outer bend of channel . . . . . . . . . 87 4.4 GE: stone artefact examples . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 4.5 Granites East: heat retainers . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 4.6 Granites East cobbles cemented into arroyo profile . . . . . . . . . . . . . . . 88 4.7 Granites Dune arroyo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 4.8 Granites Dune stone artefacts . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 4.9 Hardpan Creek and Basin Bore study area and sites . . . . . . . . . . . . . . . 92 4.10 Hardpan Creek North, east bank of main channel . . . . . . . . . . . . . . . . 93 16 | List of figures 4.11 Hardpan Creek North: water flow directions . . . . . . . . . . . . . . . . . . . 94 4.12 Hardpan Creek North site: stone artefacts . . . . . . . . . . . . . . . . . . . . 95 4.13 Hardpan Creek South profile . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 4.14 Hardpan Creek South: yellow-brown stone artefact embedded in sediments . . 96 4.15 Basin Gully 1 site: an 80 cm eroded terrace . . . . . . . . . . . . . . . . . . . 98 4.16 Basin Gully 1 site: stone artefact embedded in deposit . . . . . . . . . . . . . . 99 4.17 Basin Gully 2 site: an exposed terrace with stone artefacts . . . . . . . . . . . . 100 4.18 Double Well (DW) and Tunkana Well (TW) study sites . . . . . . . . . . . . . 101 4.19 Double Well site profile (arroyo) . . . . . . . . . . . . . . . . . . . . . . . . . 103 4.20 Double Well site: gravel base of channel . . . . . . . . . . . . . . . . . . . . . 103 4.21 Double Well site: conglomerate feature upstream . . . . . . . . . . . . . . . . 104 4.22 Double Well site: stone line, which dissects arroyo . . . . . . . . . . . . . . . 104 4.23 Double Well site: stone artefact . . . . . . . . . . . . . . . . . . . . . . . . . . 105 4.24 Double Well site: archaeological material . . . . . . . . . . . . . . . . . . . . 105 4.25 Tunkana Well profile and spring . . . . . . . . . . . . . . . . . . . . . . . . . 106 4.26 Tunkana Well site: edge-ground basalt axe . . . . . . . . . . . . . . . . . . . . 107 5.1 Granites East site: undisturbed sediment sample locations . . . . . . . . . . . . 123 5.2 Granite East Sample 1 (0-7 cm), fabric pedofeatures . . . . . . . . . . . . . . . 124 5.3 Granites East Sample 2 (30–39 cm), fabric pedofeatures . . . . . . . . . . . . . 125 5.4 Granite East Sample 3 (50–57 cm), fabric pedofeatures . . . . . . . . . . . . . 126 5.5 Granite East Sample 4 (040818), fabric pedofeatures . . . . . . . . . . . . . . 127 5.6 Granites East particle size analysis plot: sediment and soil characteristics . . . . 128 5.7 Granite East loss on ignition plot . . . . . . . . . . . . . . . . . . . . . . . . . 129 5.8 Granite East frequency-dependent susceptibility plot . . . . . . . . . . . . . . 130 5.9 Granites Dune: micromorphology and optically stimulated luminescence sample collected from the Static layer . . . . . . . . . . . . . . . . . . . . . . 132 5.10 Granites Dune (X-layer), fabric pedofeatures . . . . . . . . . . . . . . . . . . 133 5.11 Granites Dune (X-layer), micromorphological characteristics . . . . . . . . . . 133 5.12 Granites Dune (Static layer), fabric pedofeatures . . . . . . . . . . . . . . . . . 134 5.13 Granites Dune particle size analysis plot . . . . . . . . . . . . . . . . . . . . . 136 5.14 Granites Dune loss on ignition and bulk density plot . . . . . . . . . . . . . . . 137 5.15 Granites Dune frequency-dependent susceptibility plot . . . . . . . . . . . . . 138 5.16 Granites Dune: relict dune and possible colluvial event . . . . . . . . . . . . . 139 5.17 Micromorphology and optically stimulated luminescence sample locations . . . 141 5.18 HPCKN Sample 1 (Mulga), fabric pedofeatures . . . . . . . . . . . . . . . . . 142 5.19 HPCKN Sample 2 (2–9 cm), fabric pedofeatures . . . . . . . . . . . . . . . . 143 5.20 HPCKN Sample 3 (29–38 cm), fabric pedofeatures . . . . . . . . . . . . . . . 144 5.21 HPCKN Sample 4 (45 cm), fabric pedofeatures . . . . . . . . . . . . . . . . . 145 List of figures | 17 5.22 HPCKN Sample 5 (90–97 cm), fabric pedofeatures . . . . . . . . . . . . . . . 146 5.23 HPCKN Sample 5 (90–97 cm), fabric pedofeatures – 2 . . . . . . . . . . . . . 147 5.24 HPCKN Sample 6 (97–106 cm), fabric pedofeatures . . . . . . . . . . . . . . 148 5.25 Hardpan Creek North Sample 6 fungi micro-fossils . . . . . . . . . . . . . . . 148 5.26 Hardpan Creek North: plot of particle size distribution and soil characteristics . 150 5.27 Hardpan Creek North frequency-dependent susceptibility plot . . . . . . . . . 151 5.28 Hardpan Creek North loss on ignition and bulk density plot . . . . . . . . . . . 152 5.29 Hardpan Creek Centre undisturbed sediment samples . . . . . . . . . . . . . . 154 5.30 HPCKC Sample 1 (0–8 cm), fabric pedofeatures . . . . . . . . . . . . . . . . . 155 5.31 HPCKC Sample 2 (Mid), fabric pedofeatures . . . . . . . . . . . . . . . . . . 156 5.32 HPCKS Sample 3 summary descriptions . . . . . . . . . . . . . . . . . . . . . 157 5.33 HPCKS Sample 3 (OSL), fabric pedofeatures . . . . . . . . . . . . . . . . . . 158 5.34 HPCKS particle size analysis plot . . . . . . . . . . . . . . . . . . . . . . . . 159 5.35 HPCKS loss on ignition and bulk density plot . . . . . . . . . . . . . . . . . . 159 5.36 HPCKS frequency-dependent susceptibility plot . . . . . . . . . . . . . . . . . 160 5.37 Basin Gully profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 5.38 Basin Gully 1 wood nodule with phytoliths . . . . . . . . . . . . . . . . . . . 162 5.39 Basin Gully 1 fabric pedofeatures . . . . . . . . . . . . . . . . . . . . . . . . 163 5.40 Basin Gully 2 fabric pedofeatures . . . . . . . . . . . . . . . . . . . . . . . . 164 5.41 Basin Gully 2 phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 5.42 Basin Gully frequency-dependent susceptibility plot . . . . . . . . . . . . . . . 166 5.43 Basin Gully rhythmite features . . . . . . . . . . . . . . . . . . . . . . . . . . 167 5.44 Double Well location of undisturbed sediment samples . . . . . . . . . . . . . 169 5.45 Double Well Sample 1 (1–10cm), fabric pedofeatures . . . . . . . . . . . . . . 170 5.46 Double Well Sample 2 (Top–mid), fabric pedofeatures . . . . . . . . . . . . . 171 5.47 Double Well Sample 3 (Middle), fabric pedofeatures . . . . . . . . . . . . . . 173 5.48 Double Well Sample 4 (Mid-Low) micromorphological characteristics . . . . . 173 5.49 Double Well particle size analysis plot . . . . . . . . . . . . . . . . . . . . . . 174 5.50 Double Well loss on ignition and bulk density plot . . . . . . . . . . . . . . . . 175 5.51 Double Well frequency-dependent susceptibility plot . . . . . . . . . . . . . . 176 5.52 Double Well site: stone line, which dissects arroyo . . . . . . . . . . . . . . . 177 5.53 Tunkana Well: location of undisturbed samples . . . . . . . . . . . . . . . . . 180 5.54 Tunkana Well Sample 1 (Mid-Low) micromorphological characteristics . . . . 181 5.55 Tunkana Well Sample 2 (138-144 cm), fabric pedofeatures . . . . . . . . . . . 181 5.56 Tunkana Well particle size analysis and sediment/soil texture . . . . . . . . . . 182 5.57 Tunkana Well frequency-dependent susceptibility plot . . . . . . . . . . . . . . 183 5.58 Tunkana Well loss on ignition plot . . . . . . . . . . . . . . . . . . . . . . . . 184 6.1 Hardpan Creek North (90-97 cm), dusty clay coatings . . . . . . . . . . . . . . 190 18 | List of figures 6.2 Double Well site: stone line, which dissects arroyo . . . . . . . . . . . . . . . 192 6.3 Hardpan Creek North (45 cm, OSL) fabric pedofeatures . . . . . . . . . . . . . 193 6.4 Hardpan Creek South (OSL) phytoliths, bilobate opal phytolith or diatom . . . 194 6.5 Hardpan Creek North (2–9 cm), lacustrine features (varves) . . . . . . . . . . . 195 6.6 Hardpan Creek South profile . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 6.7 Stone artefact burial and dispersal model . . . . . . . . . . . . . . . . . . . . . 200 6.8 Ficus and Araucaria phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . 202 6.9 Hardpan Creek North (97 cm) fungal micro-fossils . . . . . . . . . . . . . . . 202 B.1 Granites East soil classification . . . . . . . . . . . . . . . . . . . . . . . . . . 254 B.2 Granites Relict Dune soil classification . . . . . . . . . . . . . . . . . . . . . . 254 B.3 Hardpan Creek North soil classification . . . . . . . . . . . . . . . . . . . . . 255 B.4 Hardpan Creek South and Centre soil classification . . . . . . . . . . . . . . . 256 B.5 Basin Gully soil classification . . . . . . . . . . . . . . . . . . . . . . . . . . 257 B.6 Double Well soil classification . . . . . . . . . . . . . . . . . . . . . . . . . . 258 B.7 Tunkana Well soil classification . . . . . . . . . . . . . . . . . . . . . . . . . 259 B.8 Soil classification for all sites in this study . . . . . . . . . . . . . . . . . . . . 260 B.9 GE and GD Sites: Bulk density and LOI . . . . . . . . . . . . . . . . . . . . . 261 B.10 HPCKN bulk desnity and loss on ignition . . . . . . . . . . . . . . . . . . . . 262 B.11 HCKS Site and Basin Gully LOI . . . . . . . . . . . . . . . . . . . . . . . . . 263 B.12 DW and TW Sites LOI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264 B.13 All Sites: Bulk Density and loss on ignition . . . . . . . . . . . . . . . . . . . 265 B.14 Stone artefacts embedded into sediments features . . . . . . . . . . . . . . . . 266 B.15 HPCKN profile: Stone artefact found at 95 cm (base of profile) . . . . . . . . . 267 C.1 BG1 phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296 C.2 BG2 phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296 C.3 GE0–7 cm phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 C.4 GE30 phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 C.5 GGE40 and E57 phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298 C.6 GD X-layer phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298 C.7 Hardpan Creek Mulga phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . 299 C.8 HCKN 2–9 cm phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 C.9 HPCK 29-38 cm phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 C.10 HCKN OSL (45 cm) phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . 301 C.11 HCKN 90 cm phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301 C.12 HCKN 97 cm phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 C.13 HCKN97 cm phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 C.14 Hardpan Creek South 0–8 cm phytoliths . . . . . . . . . . . . . . . . . . . . . 302 C.15 Hardpan Creek Centre phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . 303 List of figures | 19 C.16 Hardpan Creek South (OSL) phytoliths . . . . . . . . . . . . . . . . . . . . . 303 C.17 DW 1–10 cm phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304 C.18 Double Well Top/Middle phytoliths . . . . . . . . . . . . . . . . . . . . . . . 304 C.19 Double Well Middle phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . 305 C.20 DW Mid-Low phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306 C.21 Double Well 162 cm phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . 306 C.22 TW 138 cm phytoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 List of tables 4.1 Diagnostic pedofeatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 4.2 Phytolith morphotypes expected to be found in this study . . . . . . . . . . . . 119 5.1 Granites East Sample 1 summary descriptions . . . . . . . . . . . . . . . . . . 124 5.2 Granites East Sample 2 summary descriptions . . . . . . . . . . . . . . . . . . 125 5.3 Granites East Sample 3 summary descriptions . . . . . . . . . . . . . . . . . . 126 5.4 Granites East Sample 4 summary descriptions . . . . . . . . . . . . . . . . . . 127 5.5 Granites Dune Sample 1 summary descriptions . . . . . . . . . . . . . . . . . 133 5.6 Granites Dune Sample 2 (Static layer) summary descriptions . . . . . . . . . . 134 5.7 OSL dates for Granites Relict Dune . . . . . . . . . . . . . . . . . . . . . . . 135 5.8 HPCKN Sample 1 (Mulga), summary descriptions . . . . . . . . . . . . . . . 140 5.9 HPCKN Sample 2 summary descriptions . . . . . . . . . . . . . . . . . . . . . 142 5.10 HPCKN Sample 3 summary descriptions . . . . . . . . . . . . . . . . . . . . . 143 5.11 HPCKN Sample 4 (45 cm) summary descriptions . . . . . . . . . . . . . . . . 145 5.12 HPCKN Sample 5 summary descriptions . . . . . . . . . . . . . . . . . . . . . 146 5.13 HPCKN Sample 6 summary descriptions . . . . . . . . . . . . . . . . . . . . . 147 5.14 OSL dates for all sites for Hardpan Creek North . . . . . . . . . . . . . . . . . 149 5.15 HPCKS Sample 1 (0–8 cm) summary descriptions . . . . . . . . . . . . . . . . 155 5.16 HPCKC Sample 2 (Mid) summary descriptions . . . . . . . . . . . . . . . . . 156 5.17 HPCKS Sample 3 (OSL) summary descriptions . . . . . . . . . . . . . . . . . 157 5.18 OSL dates for Hardpan Creek South . . . . . . . . . . . . . . . . . . . . . . . 158 5.19 Basin Gully 1 summary descriptions . . . . . . . . . . . . . . . . . . . . . . . 162 5.20 Basin Gully 2 summary descriptions . . . . . . . . . . . . . . . . . . . . . . . 163 5.21 OSL dates for Basin Gully sites . . . . . . . . . . . . . . . . . . . . . . . . . . 165 5.22 Basin Gully particle size analysis . . . . . . . . . . . . . . . . . . . . . . . . . 165 5.23 Double Well Sample 1 summary descriptions . . . . . . . . . . . . . . . . . . 170 5.24 Double Well Sample 2 summary descriptions . . . . . . . . . . . . . . . . . . 171 5.25 Double Well Sample 3 summary descriptions . . . . . . . . . . . . . . . . . . 172 5.26 Double Well Sample 4 summary descriptions . . . . . . . . . . . . . . . . . . 172 5.27 Double Well Sample 5 summary descriptions . . . . . . . . . . . . . . . . . . 174 5.28 Tunkana Well Sample 1 summary descriptions . . . . . . . . . . . . . . . . . . 179 22 | List of tables 5.29 Tunkana Well Sample 2 summary descriptions . . . . . . . . . . . . . . . . . . 179 B.1 Granites East site particle size analysis results . . . . . . . . . . . . . . . . . . 253 B.2 Granites Dune site particle size analysis. . . . . . . . . . . . . . . . . . . . . . 253 B.3 Hardpan Creek North site particle size analysis . . . . . . . . . . . . . . . . . 255 B.4 Hardpan Creek centre and south sites particle size analysis. . . . . . . . . . . . 256 B.5 Basin Gully sites particle size analysis. . . . . . . . . . . . . . . . . . . . . . . 257 B.6 Double Well site particle size analysis. . . . . . . . . . . . . . . . . . . . . . . 258 B.7 Tunkana Well site particle size analysis. . . . . . . . . . . . . . . . . . . . . . 259 B.8 OSL dates for all sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 List of tables | 23 Chapter 1 Introduction to the thesis Michael Connolly, Artist “To dream of deep time is not to dig in search of treasure; it is to seek to understand and revivify the human history of a place from the fragments that have survived the vicissitudes of time. It is an act of wonder – a dilation of the commonplace – that challenges us to infer meaning from the cryptic residue of former worlds. It is a scale of thinking that propels us into a global perspective and allows us to see ourselves as a species (Griffiths 2018: 8–9)". 26 | Introduction to the thesis 1.1 Introduction This thesis examines arroyos near artesian springs to address the question, to what extent has continental and global climate change influenced the behaviours of early peoples living near reliable water sources along Eulo Ridge (eastern Australia)? We know that during hyper arid phases early peoples abandoned vast areas of inland Australia to occupy inland lakes, river systems and major range complexes (cf. Hughes et al. 2017; Smith 2013; Veth 1989; Williams et al. 2013). Whereas others argue that early peoples adapted to drylands with flexible and broad spectrum foraging techniques to access resources across drylands (Hiscock and Wallis 2005; Horton 1981). To test these hypotheses, a fundamental challenge concerns the limited number of stratified archaeological sites to align human behaviours with oscillating glacial and de-glacial phases (cf. Bowler 1973; 1976; Bowler et al. 2003; Denham 2008; Field and Dodson 1999; Smith 2013). Further challenges arise from problems associated with finding suitable research sites, and use of specific geoarchaeological methods and techniques to reconstruct human behaviours and the many intersects with palaeoenvironmental change (cf. Denham 2008; French 2015; Goldberg and Macphail 2006; Macphail et al. 2006). Over the last four decades, a multi proxy approach with advances in soil micromorphology and dating techniques reconstructed prehistoric land uses across Europe, Britain and Australia (Dejmal et al. 2014; Devos and Vrydaghs 2011; French 2003; 2015; French et al. 2009; Kooistra and Kooistra 2003; van Mourik 1999; Vannieuwenhuyse et al. 2017; Ward, Veth, Prossor, Denham, Ditchfield, Manne, Kendricke, Byrne, Hook and Troitzsch 2017). The work of micromorphologists has identified an array of soil related signatures to reveal anthropogenic and environmental contexts (cf. Devos and Vrydaghs 2011; Macphail and Goldberg 2018; Stoops et al. 2018). Micromorphology has often been utilised alongside soil analyses, optically stimulated luminescence dating, and phytolith analysis to bolster these approaches and develop an understanding of site interpretation and environmental frameworks (Canti and Huisman 2015; Courty 1992; French 2003; 2015; Macphail and Goldberg 2017). The central motivation of this study stems from this previous geoarchaeological research and the natural accumulation of horizontal layers within arroyos which continue at variable rates over time (cf. Butzer et al. 2008; French 2003; 2015; Goldberg 1980; Goldberg and Macphail 2006; Macphail and Goldberg 2017; Waters 1991). During the Quaternary, gully and arroyo environments went through periods of stability but were interrupted by channel change and eventual backfilling of the deposits (Waters 1991: 144). The study region, named ’Eulo Ridge’, is a potentiometric surface which is located west of the Paroo River between 144°E and 145°E, and 28°S and 29°S in southwest Queensland, Australia (Fig. 1.1). The southern end of the study area is located ~20 Km north of Hungerford township at the intersection of Boorara Creek and Eulo Ridge. Groups of mound springs are randomly dispersed along its ~125 Km extent (Fairfax and Fensham 2003; Ransley et al. 1.1 Introduction | 27 2015). Strong correlations exist between springs and high density stone artefacts scatters (cf. Robins 1997; 1998). Figure 1.1: Eulo Ridge is located west of the Paroo River (mAHD = Mean Australian Height Datum). Spring complexes occur at random intervals along Eulo Ridge, modified from Ransley et al. (2015) The geoarchaeological research reported in the following pages focuses on soil micromorphology and pedology, and optically stimulated luminescence (OSL) dating to examine soils and/or sediment exposed by gully erosion (arroyos) near springs (cf. French 2015). It is hypothesised that arroyos conceal evidence of early peoples living near water during the late Pleistocene and Holocene epochs. Fast flowing streams have cut through relict fan terraces and basin floors to reveal a dynamic palaeoenvironmental record (Fig. 1.2) (QDPI 1974). It is worth pointing out that no formal geoarchaeological studies had been carried out along Eulo Ridge. Investigations of the deposits are pursued at several levels. First, it is well known that pastoralism in semi arid Queensland and northwest New South Wales has caused extensive land degradation and deflation of loams and sandy loam soils to expose the B Horizon in some 28 | Introduction to the thesis Figure 1.2: Arroyos are found within basin and range landscapes of Eulo Ridge. This study focuses on relict fan terraces located within the basin floor, diagram modified from Schaetzl and Anderson (2005: 408) instances (Fanning 1999; QDPI 1974). Archaeological materials buried within soils and soil features are now dispersed across land surfaces and have poor spatial and temporal resolution (cf. Fanning 1999; Fanning et al. 2008; Holdaway and Fanning 2014; Rhodes et al. 2009; Robins 1997). Previous research shows that within the valleys, fluvial activity has exposed multiple archaeological surfaces that are of different ages and stacked one on top of the other (Rhodes et al. 2009: 191). Rhodes et al. (2009: 191) identified that, "a new land surface is constructed from the deposition of alluvial (water-lain) and/or colluvial (slope) sediments within part of a catchment, usually the valley floors, following, for example, a flood event". In northwest New South Wales, valley-bottom sediments comprise poorly-sorted silts (including pelletal silts) and sands with occasional gravely units (Fanning 1999; Fanning and Holdaway 2001). Coarse archaeological material may be redeposited within the gravel units but rarely within sand or silt-dominated contexts (Rhodes et al. 2009: 191). Second, large scale developments cause major disasters to the smallest organism and a state of equilibrium (Butzer 1982; French 2003; Gerrard 1992; Macphail and Goldberg 2017; Schaetzl and Anderson 2005). Eldredge and Gould (1972: 84) coined ’punctuated equilibria’ as an alternative to the Darwinian model of slow and steady transformations of entire populations. They argued that the history of life is more adequately determined by rapid and episodic events of speciation rather than the fullness of time. French (2003: 34) adopted this theory to geoarchaeology and noted that when erosion occurs it is both episodic and fast. He coined the term ’punctuated equilibrium’ to describe the episodic changes to landscapes. An examination 1.1 Introduction | 29 of landscape evolution identified rare periods of tectonic unconformities, colluvial aggradation, channel incision, gullying, pedogenesis, soil creep, and terrace collapse (French 2003; French et al. 1998). Once top soils are removed, subsoils are highly erodible but may remain relatively stable for long periods of time which is then interrupted by rapid episodic change. For landscapes to change, soil infiltration must exceed soil capacity (Dunkerley 2010). Further, sub-surface water erodes about 1% of the total soil material from a hillside (French 2015: 12). The order and arrangement of strata and time intervals provide a basic framework for identifying material remains, sediment processes, and phases of deposition (Fig. 1.3) (Stern 2008: 365). The diagnostic features are surface deflation, colluvial and alluvial processes, pedogenesis, and anthropogenic behaviours such as fire use (French 2003; 2015; 2017; Mentzer 2014). Figure 1.3: Three main types of geomorphic surfaces showing position of each surface at time 1 and time 2 (modified from Schaetzl and Anderson (2005: 467)) Third, a major phase of inquiry proposes soil pedology and micromorphology, and optically stimulated luminescence dating to bolster our knowledge of landscape and anthropogenic changes. A mixed-method greatly improves interpretative resolution (cf. Canti 1995). Recent developments and methodologies allows the study of soil microstructures and other relatively large archaeological features (cf. Courty 1992; Macphail and Goldberg 2018; Stoops et al. 2018: 1). Reliable diagnostics for sediments include macroscopic, microscopic and granulometric criteria to assist in the reconstruction of past processes (Stoops et al. 2018: 22). Four, high density stone artefact scatters found near springs in Diyari and Arabana country of the Lake Eytre Basin demonstrate that springs play a vital role to investigate links between early peoples and permanent water sources (Florek 1993; Smith 2013). The assemblages reveal that lithics provide insights into the behaviours of early peoples and their environment (Florek 1993). At Eulo Ridge, strong correlation between springs and stone artefacts. 30 | Introduction to the thesis Therefore the expectation is that stone artefacts should present an ideal situation to investigate similar patterns. A further consideration is that catchment size could influence sediment movement and/or burial of palaeo-land surfaces. Proxy data may not be able to identify the full extent and magnitude or intensity of landscape changes. The removal of palaeosols poses problems for gaining a complete understanding of the landscape change and the extremities of depositional events (Fitzsimmons et al. 2013: 80). 1.2 Study motivation The motivation for this research emerges from an urgent need to learn more about early peoples’ occupation of water sources (springs) in semi arid environments. Springs within this dryland have been flowing for at least 800,000 years (cf. Ransley et al. 2015), well before early peoples walked throughout the Australian continent. Thus, springs and the surrounding environments provided an ideal opportunity to investigate the spatial and temporal origins of early peoples. Artesian springs located in Australia’s drylands provided reliable water sources for early peoples (Smith 2013: 144). Spring environments therefore have the potential to conceal evidence of palaeoenvironmental change, and intermittent or permanent use by people and animals. This scenario implies disturbances such as trampling, vegetation clearing, modifications to the landscape, and added erosion because of human or animal activity this evidence could be found in thin section. The importance of Eulo Ridge rests with its large numbers of springs and spring complexes, and its ability to support early peoples for a long period of time. This study developed because arroyos have the potential for stratified deposits containing archaeological material to uncover the implications for site formation processes, and reveal how these hunter gatherers coped with climatic instability. Second, a lack of ethnographic and anthropological knowledge of early peoples in the study region restricts the types of questions being asked in these regions. The history of assimilation and the removal of Budjiti people to government managed institutions in eastern Australia depopulated the region after the turn of the 20th Century (McKellar 1984). There is an urgent need to better understand the past and ask how they coped with boom (wet) and bust (dry) weather cycles. Third, motivation for this study derives from known issues with poor site integrity across open air archaeological sites across northwest Murray Darling Basin (cf. Allen and Holdaway 2009). The concerns of this study stem from a lack of stable stratified open air sites and/or deep rockshelters deposits that conceal an extensive record of human occupation (Robins 1995). Thus, geoarchaeological research concentrates on surface scatters and heat retainers that reveal that these sites have temporal and spatial integrity (Fanning et al. 2009; Holdaway 1.2 Study motivation | 31 and Fanning 2014; Holdaway et al. 2008). This results in a limited understanding of early peoples and occupation of the region. Fourth, this study is inspired by the need to appraise new geoarchaeological techniques and approaches that would help us explore spatial and temporal integrity of archaeological sites. A multi proxy approach has been used to great effect across Europe and the United Kingdom to reconstruct human behaviours and palaeoenvironments (cf. Conesa et al. 2017; French 2003; 2015; Macphail and Goldberg 2017). Micromorphological research of arroyo fill and cut sequences reveal both palaeoenvironments and human use of palaeo-landscapes (French 2010; 2015; Macphail and Goldberg 2017). Pedogenic signatures aim to reconstruct wet and dry conditions to reveal natural and/or anthropogenic impacts (French 2003; 2015). Micromorphological research in Australia to date includes investigations of stratified deposits within rockshelters of northwest Australia (Jankowski et al. 2015; Vannieuwenhuyse 2008; Vannieuwenhuyse et al. 2017; Ward, Merigot and McInnes 2017; Ward, Veth, Prossor, Denham, Ditchfield, Manne, Kendricke, Byrne, Hook and Troitzsch 2017). The rationale for a multi proxy approach arose from previous geoarchaeological research conducted in Europe, Africa, the Americas, and United Kingdom (cf. Butzer et al. 2008; Devos and Vrydaghs 2011; French 2015; Goldberg 1980; Goldberg and Macphail 2006). Geoarchaeologists have used soil analysis techniques and microscopic analyses to decipher the arrangement and nature of sediment/soil components within archaeological deposits to model human settlement models and behavioural activities (Boixadera et al. 2016; Canti and Huisman 2015; Courty 1992; French 2015; French et al. 2009; Friesem et al. 2016; Goldberg 1980; Kooistra and Kooistra 2003; Macphail and Goldberg 2018; McKeague et al. 1978; Nicosia and Stoops 2017; Shahack-Gross and Finkelstein 2008; Stoops et al. 2018; Triana-Vega et al. 2019). By using these approaches, geoarchaeological research in drylands has the ability to reveal episodic erosion and reactions of early peoples to climatic variability. The state of pre-colonial landscapes and the impacts of early peoples is unknown. A final motivation concerns human movement models in response to extreme aridity such as the LGM (Smith 2013; 1989; Veth 1989). Australian archaeologists continue to debate the peopling of drylands and their behaviours (Hiscock and Wallis 2005; Smith 2013; Veth 1989; 1995)? The expansion of early peoples across Australia requires knowledge of access to water and endemic food and animal resources (Birdsell 1957; Horton 1981; Smith 2013; Williams et al. 2015; Williams 2015a). Early peoples abandoned large areas of drylands and relocated to refugia such as lakes and river systems along the Murray Darling basin, and Cooper Creek to the north west of this study area (Smith 2013; Veth 1989). A consideration of this study is whether Eulo Ridge acted as a refugia? Site integrity and scale of analysis is essential for understanding the ethos of human mobility and human reactions to landscape change (David and Thomas 2016). 32 | Introduction to the thesis The motivations and study aim and objectives of this research demonstrate that there are many facets of landscape change that help to understand the past. But problems arise with the lack of spatial and temporal data of open air archaeological sites, and the ability to connect the behaviours of early peoples to an array of landscape transformations that have occurred over time. This thesis attempts to question and investigate these issues. 1.3 Organisation of this thesis The multiple objectives necessitate introductory chapters to set the broader scene of the study region and its known climate and landscape characteristics as well as its early peoples. In Chapter 2, the aim is to discuss the archaeological and palaeoenvironments background, climate, and physical landscape characteristics. An attempt is made to unravel environmental conditions at a national, regional and local scale, and to anticipate correlations between landscapes and the past. The palaeoenvironmental focus draws on Fitzsimmons et al. (2013) and others such as Cohen et al. (2012; 2011); Dunkerley (2010); Hughes et al. (2017); Wallis (2001) to present an overview of what could be expected across the study regions in terms of the impacts of glacial and deglacial phases. Early peoples are presented in Chapter 3 and questions are framed to detail the arguments about the human setting. The objective here was to identify and describe Aboriginal behaviours with the view to understand early peoples-related evidence contained with the sediments/soils. The discussion highlights the importance of water as a component of Aboriginal belief systems, their behaviours near water sources, diet, subsistence strategies and abilities to cope with both dryland and wetland environments. Chapter 4 deals with the case studies and creates the conditions and basis for this study. Following on, the chapter discusses the project design, and field and laboratory methods used to collect and analyse each data set. The methods draw on geoarchaeological methods and techniques published by Goldberg and Macphail (2006) and French (2015) to investigate unconformities and buried soils. No technological study can avoid detailed characterisation of the microscopic pedofeatures and fossilised particles of plant tissue such as phytoliths. Phytolith classes are obtained from Neumann et al. (2018) and presented in a table of what could be expected to be recorded in this study. All of these methods and techniques are crucial to the presentation and analysis of these results. Chapter 5 focuses on the soil pedology and micromorphology results across the 3 case studies and their respective sites and features. Micrographs and photographs of key features help to describe and analyse the results. The objective is to focus on the multi proxy approach and build a chronology where possible to align the unconformities with known palaeoenvironmental evidence discussed in Chapter 2. Charts and micrographs present datasets and features to analyse and interpret each stratum, site and study area. At the end of 1.3 Organisation of this thesis | 33 each section, a summary condenses the information and includes some minor interpretations of the results. Additional data are presented in Appendices B and C. A synthesis and understanding of the results are presented in Chapter 6. Detailed reconstructions of the findings are discussed by exploring the implications of a multi proxy approach and what this means for interpreting palaeoenvironmental change. Punctuated equilibrium is examined to model wet and dry phases and its impacts on early populations. Chronological changes account for the timing of individual events and the timing of rapid erosional events. OSL dating also describes the ages of sediments and temporal changes to estimate the frequency of landscape change and how this aligns with known glacial and deglacial events. A range of archaeological materials found present within unconformities identify the presence of early peoples. On the basis of this evidence, the behaviours of early peoples reconstructed in conjunction with landscape data. The objectives raised early are examined and discussed on a level that considers the quantitative evidence. The analyses are summed up and conclusions drawn from the heuristic and empirical findings. The summary focuses on the findings and how the timing and evidence for landscape change could contribute to links between early peoples and palaeoenvironments. The chapter and thesis is concluded with a discussion of limitations of this study and future research directions. The aim is to recommend how other geoarchaeological studies could progress a multi proxy approach to understand stratified archaeological deposits in Australia’s drylands. Chapter 2 Eulo Ridge Setting Michael Connolly (artist) For Aboriginal peoples, country is much more than a place. Rock, tree, river, hill, animal, human—all were formed of the same substance by the Ancestors who continue to live in land, water, sky. Country is filled with relations speaking language and following Law, no matter whether the shape of that relation is human, rock, crow, wattle. Country is loved, needed, and cared for, and country loves, needs, and cares for her peoples in turn. Country is family, culture, identity. Country is self. Source: Meaning of land to Aboriginal people—Creative Spirits, retrieved from https://www.creativespirits.info/aboriginalculture/land/ 36 | Eulo Ridge Setting 2.1 Introduction Chapter 2 describes the study setting by discussing the previous archaeological and palaeoenvironmental research, regional climate, geology, hydrology, landforms, soils, vegetation, and fauna. These features help to provide a physical background to this study and introduce the types of landscapes along Eulo Ridge. Eulo Ridge today aligns with Hoods Range and the upper catchment of Boorara Creek. It’s extent in these lower sections dissect mulga landscapes, and the Tertiary age ranges. Groups of springs represent the aquifers and present scenarios for the reliability of water along its extent. This makes the region a human environment which the Budjiti people called their home prior to pastoralism in the late 1800s. The following characterises the Eulo Ridge landscape and sets up the background of this study. 2.2 Study setting Budjiti People The Budjiti people occupied the greater study region from Lake Thorlindah (east of the Paroo River), to west to the Bulloo River, north to Dynevor Lakes, and south to an unknown parallel within what is now New South Wales (McKellar 1984: 43–46). The Budjiti lived among the lakes and river systems, sandplains and mulga lands. Their culture centres on the land and a diversity of plants and animals such as reptiles, fish and aquatic species, kangaroos and emus, and large numbers of birds as populations escalated during extensive flooding of the lakes and Paroo River (cf. Allen 1972). Budjiti descendants connect with their homelands through cultural knowledge and relationships with people and the land. They continue to call themselves the "Paddymelon people" after a small football size marsupial that once lived throughout the region (Elizabeth, nee McNiven, Adelaide pers comm. 24 July 2018). It is worth noting that paddymelons are not listed on the species list for this region (QPWS 2001). Therefore it is assumed that this small marsupial is extinct. More will be revealed about the Budjiti and bordering language speakers in the subsequent chapter (Chapter 3). Previous archaeological research Previous archaeological research near Eulo Ridge comprised a PhD research project (Robins 1993). He excavated of an arroyo at Youlain Springs and recovered stone artefacts and marsupial bone from colluvium or aeolian deposits (Fig. 1.1) (Robins 1993; 1998). A basal thermoluminescence date of 13.23 Ka BP ± 0.44 Ka put this site toward the end of late Pleistocene post LGM. An analysis of the deposits suggested that these layers were laterally transported during a dry phase (Robins 1998: 72). Robins (1998) concluded that arroyos located near springs had the potential to provide spatial and temporal data that is otherwise 2.2 Study setting | 37 lacking in surface scatters (cf. Fanning et al. 2008; Holdaway and Fanning 2014; Holdaway et al. 2010). Studies of extensive surface stone scatters located in northwest New South Wales found multi-temporal deposits dispersed across the valley floors of open drylands (Fanning et al. 2008; Fanning and Holdaway 2001; Holdaway et al. 2017; Holdaway and Fanning 2014; Rhodes et al. 2009;?). Stone artefact numbers correlate with permanent water sources (particularly springs), but these are largely without reliable temporal and spatial resolution (Fanning et al. 2008; 2009; Robins 1997). Despite a significant amount of effort made to understand the geoarchaeology of these sites, surface scatters are complex and largely without reliable spatial and temporal resolution (cf. Fanning et al. 2008; Holdaway et al. 2017; Holdaway and Fanning 2014; Rhodes et al. 2009). To situate the archaeology of Eulo Ridge, 7 prominent archaeological sites located within the Murray Darling Basin (MDB) make a significant contribution to the archaeology of southeast Australia (Fig. 2.1). Cuddie Springs and the Willandra Lakes dominate this section because of their proximity to Eulo Ridge and these feature as significant sites in the greater region. Other archaeological sites are located to the very north and south of the basin and are better known for their Pleistocene antiquity and characteristic human remains and rock art (cf. Macumber and Thorne 1975; Mulvaney and Joyce 1965; Mulvaney et al. 1964; Pardoe 2003; Richards et al. 2007; Smith 1930). First, Cuddie Springs which is located about 330 km ESE of Eulo Ridge (Fig. 2.1, Site 2), is a 2–3 km wide lake situated within a riverine floodplain palaeo-channel on the Darling River floodplains. The site consists of archaeological material and megafaunal remains that were dated to 36 Ka (Field and Dodson 1999; Field et al. 2006). The stratigraphy comprises a mix of clays, deflation pavements, clays/silts, and a beach lag deposit. They found that marshy conditions persisted from about 36 Ka and people were preying on megafauna as the water levels in the lake was diminishing (Fillios et al. 2010: 136). Palaeo-ecologists challenged the notion in the literatures which resulted in a debate that failed to support the results of this work (Brook et al. 2006; Field 2006; Field et al. 2006; Fillios et al. 2010: cf.). Regardless, this intact stratigraphic record suggests the presence of early peoples before the commencement of LGM (~24 Ka) to 12,000 y BP (Field et al. 2001: 701). Second, the Willandra Lakes are located 560 kilometres (348 miles) south of Eulo Ridge (Fig. 2.1, Site 3). Lake Mungo is the best known for its culturally advanced cremation characteristics and human remains that date to 42 Ka ± 3 Ka BP (Bowler et al. 2003). Major layers dominate the exposed lunette deposits. The following briefly outlines the major deposits, 1) Golgol unit (excess of 100 Ka, culturally sterile) comprised sand clays deposited onto the floodplain margin by deflation followed by a long stable period and the formation of deep red calcareous soils, 2) Mungo unit (30 Ka), dark grey pelletal sandy clays overlying quartz sands of the Lower Mungo Unit, active channel development and overflows flooding 38 | Eulo Ridge Setting Figure 2.1: Selected prominent archaeological sites located in the Murray-Darling Basin (1,061,469 km2 (409,835 sq mi) area): 1) Eulo Ridge; 2) Cuddie Springs (Field and Dodson 1999); 3) Willandra Lakes (Bowler and Magee 2000); 4) Menindee Lakes (Pardoe 2003); 5) Fromms Landing (Mulvaney et al. 1964); 6) Box Gully (Richards et al. 2007); 7) Kow Swamp (Macumber and Thorne 1975); 8) Talgai skull (Smith 1930) and; 9) Mount Moffatt Station (Mulvaney and Joyce 1965) the swamps with early people harvesting resources of freshwater shellfish, 3) Arumpo and Zanci units (17 Ka to late glacial period), low lake levels with thick pelletal clays a thick sequence of pelletal clay facies and, 4) Zanci unit (upper most unit, clay dune building), laminated clayey sands, stable period with the formation of brown soil over earlier dune sediments (see Bowler et al. (2003) for more information). Within the broader site complex, research identified stone artefact, shell middens, heat retainers, and fossilised human footprints embedded in the ancient lake beds (Balme 1991; Bowler and Magee 2000; Fitzsimmons et al. 2013; Webb et al. 2006; Westaway et al. 2017). The remaining sites in Figure 2.1 sites highlight the diversity of significant sites and distribution of skeletal material within lake and riverine environments (Macumber and Thorne 1975; Pardoe 2003; Richards et al. 2007; Smith 1930). It is worth noting that Kenneth Cave (Fig. 2.3 Palaeoenvironmental setting | 39 2.1, Site 9) was the first significant archaeological site excavated in the MDB (Mulvaney et al. 1964). Dating of the deposits returned an age of 22,000 y BP which was the old known age for occupation at that time. Other rockshelter excavations at Fromms Landing identified stone tools, animal bones, shellfish and other organics (Mulvaney et al. 1964). At this site, a dingo skeleton was dated to around 3.3 Ka making it the oldest of its kind in Australia. Other sites across the MDB comprise shell middens, heat retainer features, scarred/carved trees, fossilised footprints, Pleistocene occupation sites, skeletal material, and significant human burials, which are considered the oldest in Australia (Bowler 1998; Field and Dodson 1999; Fitzsimmons et al. 2014; Pardoe 2003; Richards et al. 2007; Thorne et al. 1999; Webb et al. 2006; Westaway et al. 2017). These are simply a small number of known significant archaeological sites found in the MDB. It is beyond the aims of this study to describe all of these sites in detail. 2.3 Palaeoenvironmental setting Previous palaeoenvironmental research reported near Eulo Ridge comes from a PhD study of selected lake deposits (Gayler 2008). The findings identify humid with lake full levels prior to 30,000 years ago (30 Ka). This was followed by dune mobilisation during the Last Glacial Maximum (LGM) (24–18 Ka), and periodic humid and dry phases during the de-glacial phase 16–14 Ka. By 11 Ka – 10 Ka (Terminal Pleistocene and Early Holocene boundary), humid conditions had stabilised the region. These wet conditions weakened with episodic dry phases and intermittent filling of the lakes. By the late Holocene, pollen data demonstrated a transition from an environment dominated by herbs and grasses (Poaceae) to a landscape largely comprised of Chenopodiaceae, which is much similar to the present (an arid dryland shrub family) (QDPI 1974). Research in south-eastern Australian provides a backdrop for assessing significant climate trends along Eulo Ridge. Research of synoptic conditions by Cohen et al. (2012) identifies that Indo-Australian monsoon and mid-latitude westerlies dominated the region. During the marine isotope stage (MIS) 5, tropical lows dominated the climate which was modulated by El Niño Southern Oscillation (ENSO). Such conditions favoured sediment transport within the Murray Darling Basin tributaries, and during late MIS 5 (130–71 Ka), the drylands had received abundant moisture from these climatic sources. From the end of MIS 3 (65 Ka–57 Ka) to about 40 Ka, lower temperatures, reduced evaporation, and higher amounts of rainfall than the present dominated large areas across the MDB. By early MIS 3 (50–45 Ka), conditions favoured a southern ocean moisture source (Cohen et al. 2011: 169). From late MIS 3 or ~40 Ka to 30 Ka, the Darling River and Willandra Lakes were broadly influenced by ENSO with full lakes and high stream discharges (Bowler 1998; Cohen et al. 2012; Fitzsimmons et al. 2013). By 30 Ka to around 25 Ka, rivers flows dominated climate in 40 | Eulo Ridge Setting the western margin of the MDB. Palaeoecological data imply wetter vegetation assemblages during the late MIS3 in northern areas and regions of the MDB (Wallis 2001). This was also a period of complexity with both tropical and temperate oceanic moisture sources despite stable sea surface temperatures (Fitzsimmons et al. 2013: 90). Cohen et al. (2011) argues that amino acid racemisation analysis undertaken on emu egg shell suggests that temperatures in the MDB were 8-9° C cooler than the present. Lake levels fell earlier in the north as opposed to the slowing of conditions in southern Australia (Fitzsimmons et al. 2013: 90). Bowler (2012) suggested that lake levels within the Willandra Lakes were oscillating between about 30–18 Ka. In addition, the Murrumbidgee and Laughlin rivers flowed during this time but these are believed to be attributed to snow melt and then increased run-off in the highlands (Fitzsimmons et al. 2014). The reason is that the Darling River has its headwaters within a subtropical region and increased flows may be attributed to a wetter phase in the north. There may well have been precipitation in the Coral Sea and in the Fitzroy River in central Queensland to support this argument. During MIS 2 (24–18 Ka), the Last Glacial Maximum (LGM), conditions were much cooler and extreme dry conditions prevailed throughout arid regions. The LGM was a period of dune building and heightened dust transport. An onset of drying occurred throughout the period 20 Ka to 18 Ka with widespread is dune activity, dust transport that followed a period of pedogenesis and instability. Lake levels were much lower and river flows reduced which suggests that the monsoon during the LGM was relatively inactive. A lack of access to reliable data hinders the ability to track the episodic fluvial activity during the LGM. Formation of source bordering dunes along the Murrumbidgee River suggest seasonal flows and sediment supply from increased bedload around this time (Fitzsimmons et al. 2013). There is also evidence to suggest the lateral migration of the Darling River west of the Willandra Lakes, which increased during the LGM. Palaeoecological data identified that palm trees had disappeared from the phytolith record (Wallis 2001). From about 20 Ka to the early Holocene, MDB rivers were highly active. The monsoon had shifted southwards from about 15 Ka to 14 Ka and dated palaeo-hydrological records show increased sediment supplies in lake and river systems. This occurred in two stages, as early as about 17 Ka, which suggest lake level rises in the both the Northern Lakes in the Lake Eyre basin. This humid phase was followed by increasing aridity to the west within the arid centre. Between about 14 Ka and 10 Ka, there appears to be a decrease in fluvial activity which has also been noted as causing reduction in the growth of speleothems (Fitzsimmons et al. 2013: 92). Temperature at this time is also warmer than the LGM. By MIS 1 (14 Ka), the end of the LGM, the climate progressed to widespread dry conditions and dune building activity across the region. Palaeoenvironmental changes were effected by westerlies crossing the continent at temperate latitudes, and tropical moisture and humid conditions influenced climate in the north and interior (Bowler et al. 2003; Fitzsimmons et al. 2.3 Palaeoenvironmental setting | 41 2013; Hughes et al. 2017). Evidence of humid periods and global cold stages resulted in the lowering of lake and river levels with extreme aridity. Vegetation was also becoming more sparse and drylands were mostly void of vegetation (cf. Cohen et al. 2012; Dunkerley 2010; Fitzsimmons 2017; Fitzsimmons et al. 2013). The period after the LGM from about 18–12 Ka, is characterised by spatial variability with divergence between the monsoon in the westerly influences in the south and north of the continent (Fitzsimmons et al. 2013). Lake full conditions were common across the Murray Darling Basin at the end of LGM. By 17–14 Ka, humid phases were more common in the semi arid margins of the MDB but this slowly transitioned into dune building by around 14–10 Ka. These conditions are generally identified as the end of the LGM. Widespread humid conditions were followed by marked instability and fluctuations between each dry phase. Evidence suggests that between 14 Ka and 15 Ka the monsoon sifted southwards and resulted in weather conditions which caused rises in lake levels. Between 12 Ka and the present, there was conjecture about whether this was a continuation of the LGM with more aridification (Fitzsimmons et al. 2013). During the early to middle Holocene, conditions were moderately humid and rising lake levels and source bordering dunes were consistent across the continent. The Holocene climate was characterised by spatially divergent and abrupt changes across the continent (Fitzsimmons et al. 2013). Fitzsimmons et al. (2013: 93) argue that southern central Australia experienced increasingly humid and stable climatic conditions, with fluvial and pedogenesis on the Finders Ranges, fluvial fans and short lake level rises in Lake Frome around 7 Ka to 5 Ka years ago. River flows following the LGM declined sharply at the start of the Holocene but re-intensified with wetter conditions during the mid Holocene between ~7–5 Ka. The margins of the MDB however saw increased dune activity with peak dust output. But this is not confirmed because of limited data availability in these regions. At ~5 Ka, this was a period of intermittent peaks in aeolian activity with fluxes of dust. Between 5 Ka and 2 Ka, there was a greater trend towards aridity with much harsher conditions. Fitzsimmons et al. (2013: 93) point out that Holocene climatic conditions are not fully understood and more research is needed to better understand some of the complexities. After 5 Ka, increased aridity saw lower river discharges and aeolian activity within the MDB rivers (Fitzsimmons et al. 2013). There is evidence for a weak monsoon system that has been attributed to the late Holocene and awakening of the Walker circulation over the tropics (Shulmeister 1999). Between ~2–1.2 Ka, the evidence shows peaks in dust fluxes both in the east and south of the MDB, and a strong trend toward aridification (Fitzsimmons et al. 2013). The monsoon however was substantially less humid in the north during the Holocene. It is suggested by Fitzsimmons et al. (2013: 92) that the Holocene corresponds to a global interglacial phase but the monsoon and its effects on the climate appear less effective than 42 | Eulo Ridge Setting MIS 3. They suggest that more research is needed to examine landscape change to support climate models. 2.4 Climate Eulo Ridge is located in the semi-arid zone within the Köppen classification "BSh", Hot semi-arid with evaporation of 3,000 millimetres per annum and rainfall less than 280 millimetres. (Fig. 2.2). The region falls within a belt of high pressure which encircles the globe in the middle latitudes and creates a sub-tropical ridge which stretches across inland Australia from east to west (Fig. 2.3) (Bureau of Meteorology, "Australian Climate Influence", accessed 25 July 2019). Conditions are highly variable with hot dry summers, cloudless days and nights, cold winters with frosts, and significant wet periods or years without any noteworthy rainfall (Winkworth and Thomas 1974). Easterly troughs and blocking highs dominate the broader climate phases ("Easterly troughs", www.bom.gov.au, accessed 06 April 2019). Easterly troughs deepen and bring unstable air masses, and interact with southern cold fronts and bring heavy rainfall. ’Blocking highs’ form in the southern half of the continent and block weather systems. These remain stationary and promote dry conditions and southeast winds. These highs block frontal systems, which slow down, weaken and move south of the high pressure system. On occasions, strong southeast winds cause unstable conditions ("Blocking Highs", www.bom.gov.au, accessed, 25 July 2019). The region is well known for its unpredictable ’boom and bust’ cycles of droughts and floods dominate Australia’s drylands. In boom years, parched landscapes are transformed into a lush environment with significant increases in invertebrates, plants and animals. La Niña brings rainfall and tropical cyclones to northern Australia (Bureau of Meteorology, www.bom.gov., "El Niño Southern Oscillation (ENSO)", accessed 25 July 2019) and sometimes results in extreme wet periods with major flooding in the Paroo River and tributaries. Tropical lows deliver well above average rainfall to inland regions and rivers channels, lakes, claypans, and spring aquifers fill to transforms the landscape into a wetland with diverse ecologies bursting with life and energy (Kingsford et al. 2010). Wet phases align with La Niña, which brings rainfall and tropical cyclones to northern Australia (Bureau of Meteorology, www.bom.gov., "El Niño Southern Oscillation (ENSO)", accessed 25 July 2019). Tropical cyclones sometimes cross the coast and move inland and change to tropical lows, which brings equivalent of several year’s average annual rainfall in a few days. Soils become saturated and major flood waters spread out for several kilometres across the ancient floodplains and sometimes into lakes and other minor drainage systems (Bureau of Meteorology, http://www.bom.gov.au/, accessed 26 July 2019). Floods in the Murray-Darling Basin have been recorded in 1895, 1917, 1956, and 1974. Summer rainfall stimulates perennial grass growth and winter rainfall encourages growth of annual species. 2.4 Climate | 43 Figure 2.2: Eulo Ridge is located within the eastern hot semi-arid zone (modified from ’Köppen climate types in Australia’, https://en.wikipedia.org/wiki/Climate_of_Australia). Köppen climate classification divides climates into five main climate groups, A=Tropical, B=Arid, C=Temperate, D=Continental and, E=Polar Dry phases align with El Niño Southern Oscillation (ENSO), which results in extensive dry periods in northern and inland Australia, and refers to ocean temperature warming in the central and eastern tropical Pacific Ocean (ENSO, www.bom.gov.au, accessed 25 July 2019). High summer temperatures, mild winters days, and windy conditions prevail throughout spring and autumn dominate weather patterns (Winkworth and Thomas 1974). Long periods without effective rainfall (years) are common and plants enter hibernation or a dormant state. Dry phases continue for several years and sometimes a decade (or more). In winter, temperatures fall below freezing and frost are prevalent. Between 1890 and 1940 this was dry period with below average rainfall and since the 1970s there has been return to much drier conditions (Gayler 2008). Eulo Ridge is located within the northwest frost zone of southeast Australia. For frosts to occur, below freezing temperatures, low humidity and 44 | Eulo Ridge Setting Figure 2.3: Climate and weather systems that effect Australia’s climate. The study area lies in a region that is located above the Sub-Tropical Ridge (Winter), east of Upper Level Trough, and west of Easterly Trough ("Australian Climate Influences", http://www.bom.gov.au, accessed 25 July 2019). radiation, cloudless skies with little to no wind, which causes ice crystals or frozen dew drops to form in ground depressions and valleys (BOM 2014). 2.5 Geology, Geomorphology and Soils The geology forms part of the eastern margin of the Eromanga Basin, a sequence of conformable successions of Jurassic, Cretaceous, and Tertiary sediments (QDPI 1974). Ferricrete and silcrete were deeply weathered during the Tertiary period and since then, lateritic and Quaternary deposits have obscured the original sediments (Dawson 1974a). Devonian-age granite outcrops can be found in discrete places along the Eulo Ridge (Purdy et al. 2016: 4). The earlier sequences include Middle Devonian igneous and plutonic outcrops comprising granite, adamellite, quartz-muscovlte schist, minor quartz veins (Dawson 1974a) . The Upper and Lower Jurassic-Triassic; and Permian (0-300 m) comprises sandstone, siltstone, coal, and 2.5 Geology, Geomorphology and Soils | 45 conglomerates (Dawson 1974a). The Mesozoic and Cenozoic (Tertiary) periods dominate the geology of the region. Aspects of the geology include Lower/Upper Cretaceous fluviatile and lacustrine (of lakes) landforms. These are shallow marine mudstones and limestones. The most identifiable features in this study are the Granite outcrops that derived from the Palaeozoic (Purdy et al. 2016: 4). A geological conformity comprising a succession of Jurassic, Cretaceous and Tertiary sediments, which were deeply weathered during the latter period (Dawson 1974a; Ransley et al. 2015). Above this are the Cenozoic and Glendowner Formations is an unconformity of fluviatile sediments. These comprise quartz sandstone, sandy conglomerate, breccia, siltstone and silcrete, which are largely obscured by Cenozoic (Tertiary) laterite and Quaternary quartz-dominated sands (Dawson 1974a: 12). Faulting, rifting, and epeirogenic movements during the Mesozoic and Cenozoic caused sediment to move from the uplands to closed basins (Williams 2015b: 4). A regional southwestward tilt of the Eromanga Basin, away from the Great Dividing Range, activated denudation/erosional regions in the east (Ransley et al. 2015: 44). Parallel fracture systems controlled the colluvial and fluvial exhumation of the duricrusted tablelands. Riverine systems transported these sediments from the Surat Sub-basin into three large geographical areas. These are the Darling River system of the Murray Basin, eastern Eromanga Sub-basin and, west to the Callabonna and Tirari depocentres of the Lake Eyre Basin. The region’s topography consists of two main physiographic units, 1) Tertiary-age dissected residuals and, 2) Quaternary-age valleys, dunefields, and sandplains. The dissected residuals that comprise the Walters and Hoods ranges rise about 40–80 metres above the surrounding valleys and landscape. These extend north-north-east to south-south-west and form elevated linear landmarks in what is a generally flat landscape. Shallow rockshelters occur along the eroded margins where fissures and sandstone intrusions interrupt the silcrete capping. To the west of Hoods Range, a shallow valley contains palaeo-lakes and dunefields, and to the east, mulga woodlands that converge with eucalypt swamps, and the Paroo River floodplain. Past climate and drainage has had significant effects on soil development and chemical alteration of the Cretaceous sediments, during the Tertiary period, has been responsible for subsequent erosion of the profile. Quaternary deposits derived from erosion of the Tertiary land surface, loamy red earths with iron shot gravel, red earths and lithosols on upper slopes (QDPI 1974: vi). Soils correlate with the geology. Soils across the greater region include alkaline red Tenosols, red Kandosols, Calcarosols, and Sodosols. Tenosols and red Kandosols occur on the sandplains; Calcarosols disperse with saline landscapes; and Sodosols exist in the depressions and claypans (Biggs et al. 2010: 210). Dawson (1974a: 12-13) notes the following Quaternary (Depth: 0-170 m) sediment derivatives of weathering and geomorphic processes: • (Alluvial)—Clay, sand, silt, soil, minor gravel, locally gypsiferous; 46 | Eulo Ridge Setting • (Aeolian) (Superficial)—Quartz sand, mostly iron-stained; • (Eluvial and aeolian) (Superficial)—Red sandy soil, minor gravel; • (Colluvial and alluvial) (Superficial)—Gravel, mainly silcrete; • (Colluvial and alluvial) (Superficial)—Gravel, mixed clasts; • (Valley floodplain) (Superficial)—Limestone, chalcedony and; • (Colluvial and alluvial) (Superficial)–Limestone, chalcedony. The mulga soils are, "deep bright red soils of sandy loam to sandy clay loam texture" with ironstone gravel in the profile (Bastin and the ACRIS Management Committee 2008: 78), and organic staining in surface sediments (QDPI 1974: 21) (Fig. 2.8a and Fig. 2.8b). These soils are poor and very old and are a result of the dissection of the weathered mantle. These were formed by chemical alteration of the Cretaceous sediments and are closely related to water availability (Dawson and Ahern 1974: 22). Soils textures range from sandy-clay-loams to clay-loams and light to medium clays at depth. Soil depth varies considerably depending on position within the run-on areas and are free of gravel but ferruginous gravel occurs at the base of some profiles. The valleys are filled with Quaternary alluvium and sands that form the dunefields and sand loams on the sandplains. Low dunefields cover much of the surrounding alluvium. This alluvia is generally multi-channelled and cut through floodplain to form a dendritic river system. Boorara Creek flows southward between the Walters and Hoods Ranges and fills Lake Numalla from the northeast. In peak flood periods, floodwaters fill Lake Numalla from the south and as floodwaters recede, excess water syphons back into Paroo River (QDPI 1974: 16). Flooding in the Paroo River flows steadily southward to the Paroo River floodouts and Darling River. 2.6 Hydrology The Eulo Ridge is a hydrological sub-basin of the Great Artesian Basin (Fig. 2.4) (Ransley et al. 2015). The Great Artesian Basin (GAB) is Australia’s largest aquifer. This covers an area of more than 1.7 million square kilometres from eastern Australia to Cape York, northern New South Wales to central South Australia, and west of the Great Dividing Range to central Australia (Ransley et al. 2015: 2). The GAB recharges from the Great Dividing Range and from local recharge zones such as the Eulo Ridge, which is located in a sub-basin of the Eromanga Basin. Within the basin are aquitards which are charged from the upper water tables of the uppermost Lower Cretaceous aquifers (Ransley et al. 2015: 32). Eulo Ridge is located in the uppermost Lower Cretaceous aquifers and its basement high separates the Surat and Eromanga hydrogeological sub-basins. Many springs occur along the 2.6 Hydrology | 47 axis (Ransley et al. 2015). The axis of Eulo Ridge is coincident with the water table highs which passes through granite basement inliers. Eulo Ridge has some local groundwater flow systems with local recharge mounds. Two such mounds straddle the Paroo River and occur within a small granite intrusion. Most springs have ceased to flow and appears to strongly correlate with the drilling of water bores into the Doncaster Member of the Wallumbilla Formation (Ransley et al. 2015). Figure 2.4: Eulo Ridge is located within the south eastern region of the Great Artesian Basin (Modified from Ransley et al. (2015: 21)) The depth to hydrogeological basement diminishes at the Eulo Ridge and the water table forms an almost impermeable subsurface boundary (Ransley et al. 2015: 22). Groundwater permeates back and forth through this aquitard although its hydrogeological role is uncertain. The shallow depth of the water table at the Eulo Ridge, along this aquitard boundary, forms a potentiometric surface (Ransley et al. 2015: 32). This surface dissects the Eulo Ridge in the 48 | Eulo Ridge Setting vicinity of Hoods Range. Along this range, groups of springs are randomly distributed and form a mix of mounds, depressions and calcareous deposits created from intermittent phreatic events (Ransley et al. 2015: 33). Eulo Ridge is located in the the Hooray Sandstone of the Great Artesian Basin of the northwest Murray Darling Basin (Fig. 2.5) (cf. Ransley et al. 2015). Springs are known to have been active for hundreds of thousands of years (cf. Ransley et al. (2015)), well before early peoples had arrived in this region. Springs occur in a random distributions along the water-table axis of the Eulo Ridge (Fairfax and Fensham 2003). Major springs complexes are sometimes found at the base of the Hoods Range along the toe slopes and valley floors, in association with high to moderate runoff areas, and major sediment traps. A survey by Fairfax and Fensham (2003) found 58 known springs of which, 42 were inactive and 13 active. These combinations of springs occur as distinct spring groups. Figure 2.5: Eulo Ridge is located within the western extremities of the Hooray Sandstone and northwest region of the MDB The springs’ characteristics included mounds, shallow depressions with a central pool of water, and small seepages that flow across the ground surface (Fairfax and Fensham 2003; Fensham and Fairfax 2003). Active spring discharge varies from a few litres to many megalitres per day (Fairfax and Fensham 2003: 285). Radke et al. (2000) suggests that springs have flowed for about 800 Ka. 2.6 Hydrology | 49 (a) Spring (pool) (b) Mound spring (inactive) Figure 2.6: Two examples of springs located along Eulo Ridge 50 | Eulo Ridge Setting Aside from springs, permanent waterholes occur along the lower reaches of Boorara Creek. River redgum trees line these large waterholes, streams and tributaries. Ephemeral watercourses and small tributaries flow from escarpments and elevated areas of Hoods and Walters Ranges to the lower plains, swamps, dunefields, lakes and claypans. These hold water for weeks and sometimes months but this depends entirely on diurnal temperatures and evaporation rates (QDPI 1974; Timms 2001). River channels eventually dry-back and slowly evaporate. Permanent waterholes get dangerously low without intermittent flows. First and second order streams cut across shallow irregularly shaped valleys and flow into Boorara Creek, which eventually flows into a large freshwater lake. Arroyos are randomly dispersed within these gully and channels and vary in depth and length. Arroyos appear to correlate with distance to slope, sediment/soil type, and historical land use. 2.7 Landforms and land systems The geological landform correlates with the major soil and vegetation types (Dawson 1974a: 12). The landforms change with geological and topographical features from the western margin of the lakes to the dissected residuals of Hoods and Walters Range (Fig. 2.9). Erosion gullies and sheet wash is heightened from the lower slopes to the major streams. Its central features are dominated by dunefields, sandplains, acacia woodlands, artesian springs (Fig. 2.6), slopes and valleys and, granite outcrops (Fig. 2.7). Dramatic contrasts exist between erosion amounts in the mulga lands, sandplains, dunefields, and range land units. Each land systems contains unique topographical features, soils, vegetation, and ecological responses to extreme climate phases. and impacts caused by people and animals (QDPI 1974). The major land systems for this region are QDPI (1974): • R5 Residual: The dissected residual land systems form the backbone of the area in that they dominate the high areas forming the main ranges and catchment boundaries (Dawson 1974b: 79). The backslopes may grade into undulating plains. Ironstone gravel and ferricrete are common and small areas of silcrete occur. Soils are predominantly lithosols. Lancewood and bendee shrublands occur on the scarp and crests with mulga and bastard mulga (Dawson 1974b: 81). • H2 Hard Mulga: The landform is gently undulating plains with slopes commonly grading into dissected low hills (Dawson 1974b: VII-9). Soils are very shallow to moderately deep, acid, loamy, red earths with ironstone gravel throughout the profile. Textures range from sandy-clay-loam at the surface to light clay at depth. Iron shot is common on the surface, Gn 2.11, Gn 2.12, Um 5.51 (Dawson 1974b: VII-9). 2.7 Landforms and land systems | 51 • M2 Soft Mulga: The landform is flat to gently undulating plains with low slopes and few defined drainage lines (Dawson 1974b: VII-6). Soils are loamy, red earths with ferruginous gravel in the profile. Textures range from sandy-clay-loam to clay-loam at the surface to light and medium clays at depth (Dawson 1974b: VII-6). • S2 Sandplains: Very deep, alkaline, red, earthy sands on dunes; alkaline at depth Gn 2.13, Um 5.31 (Dawson 1974b: VII-5). Mulga, poplar box shrubby tall open shrubland to low open woodland, open tussock grassland (Dawson 1974b: VII-5) • D7 Dunefields: These are flat plains with low rounded dunes (<1 m high) and flat claypans and saltpans (Dawson 1974b: VII-7). The soils are very deep, alkaline, red, earthy sands on dunes, Uc 5.21 (Dawson 1974b: VII-7) • WWoodlandsWoodlands are mainly distributed in the braided channels and floodplains, poorly drained swamps and small inter channel plains. Soils are a complex of alluvial, grey clays, alluvial soils and texture contrast soils (Dawson 1974b: 82). Figure 2.7: Granite landforms Eulo Ridge: an outcrop located at the southern end of Eulo Ridge 52 | Eulo Ridge Setting (a) Soft mulga (M4 Land System). (b) Hard mulga (H2 Land System) Figure 2.8: M4 occurs within the gently undulating convex plains with slopes. H2 land systems persist throughout the undulating plains with slopes grading into dissected residuals (QDPI 1974) Quaternary deposits overlying weathered Cretaceous sandstones. Soils are very shallow to moderately deep loamy red earths with gravel in profile. Sandy clay loam occur at surface and grade into light clay at depth(QDPI 1974). The mulga sandplains derive from aeolian sediments, whilst soft mulga land zone originates from reworked alluvia, pediment mantles and fans. The hard mulga land systems occur on gentle to undulating plains, and grade into the Dissected Residual land systems (Fig. 2.8) Dawson (1974b: 77). 2.7 Landforms and land systems | 53 Fi gu re 2. 9: La nd fo rm cr os s- se ct io n de pi ct in g la nd sy st em s (w es t( le ft) to ea st (r ig ht )) :E ul o R id ge en vi ro nm en t 54 | Eulo Ridge Setting 2.8 Vegetation Vegetation is strongly correlated with the geology, topography and soil types (Dawson 1974b: 77). Along the Eulo Ridge, vegetation comprises Acacia aneura (mulga) and Eucalypt-dominant communities (Fig. 2.10). The understory and shrublands comprise Eromophila spp. and Dodonea spp. distributed throughout the sandplains and dunefields (Boyland 1974: 50). The main species along the drainage lines are Eucalyptus camaldulensis. The heavier clay soils are dominated by Acacia cambageii and Eucalyptus ochrophloia on the drainage areas and depression of the Paroo River floodplain. The vegetation around the phreatic zones (spring complexes) comprise mainly small endemic species including couch grasses, which grow near the vent and saturated zones of the springs (Fensham and Fairfax 2003). Gramineae and Leguminosae are the largest plant families across the entire region, followed by Chenopodiaceae, Compositae, Fabaceae, Malvaceae, Myoporaceae, Myrtaceae and, Proteaceae (Boyland 1974: 50). Plants possess simple structures such as low height with projective foliage, and an ability to withstand long periods between rainfall (Boyland 1974: 47). Mulga (Acacia aneura), which belongs to the Fabaceae family, is a dry climate tolerant tree species that occupies the loamy soils from eastern to western Australia (Fig. 2.10) (Dawson 1974b). Another common species is Eucalyptus populanea, which belongs to the Myrtaceae family, grows to 20 metres high and is distributed throughout eastern Australia (Fig. 2.10). A further tree species, Grevillea striata, which belongs to the Proteaceae family, occurs in small groves or as scattered trees in open Eucalyptus and Acacia woodland land systems of the region (Boyland 1974). Grass species, Eragrostis spp., Eriachne spp., and Aristida spp. belong to the Poaceae family, which occur throughout this region and in most land systems along the Eulo Ridge (Boyland 1974). Vegetation types vary as each land system changes. The following presents vegetation for the S2, M2 and H2 land systems (Dawson 1974b): • S2 Vegetation: Mulga, poplar box shrubby tall open shrubland to low open woodland, woollybutt grass, budda bush, hop bush low shrubland occurs on some low sloping edges of the plain. Mulga, poplar box tall shrubland is found on run-on areas (Dawson 1974b: VII-5). • M2 Vegetation: Mulga shrubby tall shrubland with poplar box to tall open shrubland. Mulga sparse in some areas (Dawson 1974b: VII-6). • H2 Vegetation: Mulga tall open shrubland and mulga, western bloodwood associated with small areas of gidgee (Dawson 1974b: VII-9). 2.8 Vegetation | 55 (a) Acacia aneura distribution (b) Eucalyptus populanea distribution Figure 2.10: Distribution of Acacia aneura and Eucalyptus populanea: these species overlap in eastern Australia and the MDB (Map sources, https://en.wikipedia.org/wiki/, accessed 16 December 2019) 56 | Eulo Ridge Setting In dry periods, plants become dormant but recover quickly soon after an extensive wet period (QDPI 1974). Long dry periods have been known to kill stands of Acacia aneura forests. In low rainfall areas, mulga becomes shorter or stunted, and grasses such as Eriachne spp. possess mechanisms to withstand long periods without rain. Plants are adapted to dry conditions and are fire tolerant (QDPI 1974). Grasses grow in response to summer rainfall but have adapted to long periods without rainfall. Grass growth commences soon after adequate summer rainfall, and is followed by a long period of cessation (Winkworth and Thomas 1974: 8). In contrast, mulga (Acacia aneura) requires winter rainfall to regenerate and set seed. Soil depth, texture, and nutrients constrain plant growth. Clays contain the most nutrients and sandy-loam textured soils are generally low in nutrients, but in these environments, soil moisture matters most for plant growth. Other influences include temperature, winds, and frosts (Boyland 1974: 47–48). 2.9 Fauna The fauna include the red kangaroo (Macropus rufus) and western grey kangaroo (Macropus fuliginosus), Emu (Dromaius novaehollandiae), possum (Trichosurus spp.), echidna (Tachyglossus spp.) and koala (Phascolarctos cinereus) (Miller and Worland 2017: 11-12). A large number of the smaller marsupial species are either under threat or extinct for example, bilby (Macrotis lagotis, burrowing bettong (Bettongia Iesueur) (Strahan 1983: 187). Birds included territorial and migratory species (Kingsford et al. 2010). In wet seasons, Lakes Numalla and Wyara fill with floodwaters and come to life with breeding colonies of pelican and other waterbirds. These lakes account for about 41 waterbird species (Kingsford et al. 2010; QPWS 2001). Reptiles are most active during the warmer months and hibernate in the cooler months. The largest of these are Varanus gouldii (Gould’s goanna) (Cogger 2014: 773), and medium to small species include snakes, for example, western brown snake (Pseudonaja nuchalis), eastern brown snake (Pseudonaja textilis), mulga snake (Pseudechis australis), and Burns’s dragon (Amphibolurus burnsi), shingle-back lizard (Tiliqua rugosa) and 15 species of frog (Cogger 2014). 2.10 Chapter Summary In summary, the aim of this chapter was to describe the past environments of the Eulo Ridge. This chapter has introduced many variables and situations that contributed to landscape change over time. The objective was to set the geoarchaeological context, and how this introduces site formation processes and former human landscapes. Southeast Australia has not remained static but has oscillated between arid and humid periods throughout the 2.10 Chapter Summary | 57 past. The palaeoenvironmental phases from 40 Ka BP to the present form a backdrop to this information. These include the humid phase from 40 Ka to 30 Ka, arid phase from 30 Ka to 18 Ka (LGM), humid phase from 18 Ka to 12 Ka, and variable conditions from 12 Ka to the present (Fitzsimmons et al. 2013). The physical processes shaped much of the current landscape and human impacts during the pastoral period have deflated the land surfaces and caused widespread erosion and impacted on the ecology, and hydrological system (Fairfax and Fensham 2003; Gawne et al. 2011; QDPI 1974). This sets a scene for investigation of the overall research question and how a multi-proxy approach can address questions about these palaeoenvironments and a human occupation model. The next chapter examines how humans have contributed to site formation processes and, how prehistoric and historic peoples have shaped the landscape near springs. These types of behaviours are often overlooked therefore the focus here is on how people reacted to palaeoenvironmental conditions during humid and arid phases, and how they managed the landscape and survived in sometimes difficult conditions. Chapter 3 Living near water sources Aboriginal law – Bunyip (Macfarlance 1890) The rainbow-serpent lives in deep permanent lagoons and waterholes and devour human beings who approach its home (Radcliffe-Brown 1930). 60 | Living near water sources 3.1 Introduction This chapter examines the behaviours of early peoples living near springs. The purpose of this chapter is to discuss these behaviours and understand how this could influence the results and analysis of this study. Several questions are raised to explore the evidence and draw on studies from selected inland regions. The objectives are to understand populations, their language, use of fire to manage the ecology of these dryland landscapes, subsistence strategies, and how they coped with changing climatic conditions such as mobility across inland regions (Bird et al. 2012; Cane 1987; Gould 1968; Jones 1969; Kimber 1983), (Florek 1993; Holdaway et al. 2008; Robins 1998; Smith 2013), (Cane 1987; Gould 1968; 1971; 1995). Questions in this chapter include, What do we know about the early peoples who lived near springs and waterways? When did people arrive in the region and how? Who were the early people of the northwest Murray Darling Basin? How did early peoples cope with extreme climate conditions? What kinds of evidence explain the types of behaviours near springs? What types of subsistence strategies did people use for their survival in drylands? How mobile were people in their language area? And, what types of belief systems did early peoples associate with water? Part of this chapter will look at the bigger picture such as when people first arrived in this region, how many languages were spoken in the greater region, and what was the significance of water in their belief systems? It is worth noting there is a focus on sources obtained from central Australia as this has landscape similarities to Eulo Ridge. These sources come from ethnoarchaeological and anthropological studies, and unsystematic observations made by explorers. 3.2 Aboriginal Setting: Overview Settlement of Sahul (Australia) by early peoples Archaeological research tells us that the first people to populate Sahul (Australia) arrived from Asia/Africa about 65,000 years ago (Clarkson et al. 2017; Hiscock 2015; O’Connell and Allen 2004). People are thought to have crossed about 50 kilometres of ocean via West Papua to reach northern Australia. From arrival, it is unknown how long it took to occupy the entire continent or whether they moved around the entire extent of the coastline or moved inland (Bird et al. 2019; Clarkson et al. 2017; Hiscock 2015). Smith (2013: 68) suggests that the initial occupation of the interior took place 45 ± 5 ka. People would have encountered extensive deserts and drylands, chenopod shrublands and open woodlands but people quickly adapted to the availability of its endemic plants and animals (Hiscock 2015: 440–443) and (Flood 2006; Horton 1981). To reach these areas, it has been hypothesised that early peoples travelled around the better resourced areas along the coasts and then inland along major river systems and into the marginal areas (Bowdler 1977). 3.2 Aboriginal Setting: Overview | 61 Bird et al. (2016) argue that water sources played an important role in mobility across vast distances of these dryland landscapes. The strategy was to move between water sources and gather resources along the way (Cane 1987; Gould 1968). During Last Glacial Maximum as the continent dried out, it is hypothesised that populations contracted into smaller areas as permanent water and food resources dwindled (Smith 2013; Veth 1989). They suggest that these were small isolated refugia that were distributed more widely across the content because of springs, billabongs (large waterholes along the river systems), and gnammas and rockholes found in the bed-rock that fill after adequate rainfall (Smith 2013; Veth 1989). Others argue for a possible, "central cultural attachment of Aboriginal Australians to ‘country’ may reflect the continuous presence of populations in discrete geographic areas for up to 50 Ka BP" (Tobler et al. 2017: 4). Tobler et al. (2017) noted localised, territorial, and language maintenance across broad environmental regions. They claim that limited geographical movement between the language groups were consistent with the fluidity of estate boundaries that change between competing knowledge holders (Sutton 2003). In addition, they demonstrated Holocene intensification around 9 Ka BP – 6 Ka BP (mid-Holocene) aligns with increasing accessibility to the arid regions and cultural change such as the spread of Pama-Nyungan languages, development of Panaramittee rock art, and seed-grinding techniques. In the Murray Darling Basin, archaeological and palaeoenvironmental evidence at Lake Mungo demonstrated that the first people’s had occupied the region between 45 to 42 Ka BP (Bowler et al. 2003). This coincided with high lake levels in the Wilandra Lakes and a major climatic shift across eastern Australia (Brook and Bowman 2002; Cohen et al. 2012; Field et al. 2006; Fitzsimmons et al. 2013). They arrived in the upper Darling River by at least 33 Ka BP to 36 Ka BP (Field et al. 2001). Field et al. (2001) research at Cuddie Springs provides evidence for interactions between people and megafauna. Robins (1998) places people in this study region by about 16 Ka BP (Richard Robins pers comm. 30 March 2018) prior to the extinction of some small marsupial species (Robins 1998). This puts arrival in the region after 20 Ka BP but this could be due to the small number of research projects undertaken so far and does not necessarily provide a definitive age for this region. How many languages were spoken in upper Murray-Darling Basin? About 250 separate languages were spoken on the Australian mainland and off-shore islands (Flood 2006: 16). About 45 of the 250 languages in Australia were spoken in the Murray-Darling Basin Basin (MDB) (Area = 1.059 Million square kilometres) (Fig. 3.1). The MDB languages fit within the Wiradhuric and Maric subdivisions. The Wiradhuric languages cover most of the MDB, and Maric/Karnic, the northwest region. The language of the study region is Budjiti or Badjiri. This language is a Karnic/Maric language—Karnic extends into the central Lake Eyre region and Maric, to central Queensland (Tindale 1974: 164). Most 62 | Living near water sources languages are extinct as a result of colonial policies that prevented people from speaking their languages. Kefous (1988: 225), and Smith (2013), Veth (1995) and Hiscock (2015) propose that small populations of people living in a large dry continent reliant on limited water supplies and highly variable resources. Kefous (1988: 225) estimates that 250,000 people occupied Australia prior to European arrival. The lowest population densities occurred in deserts and the highest in the ranges and, floodplain corridors abound with fish and riverine resources (Smith 2013: 10). Tindale (1974: 110) estimated an average of 450 to 500 people per language group and he related population size to the availability of resources. A typical family group in the MDB comprised several brothers with one or more wives, children, and an extended family of brothers/sisters and their wives/husbands (Fig. 3.2) (Allen 1972; Bonney 1884). Others describe between 10–12 people in a small family group, and >100 people (up to 150 people in northern part of the desert) when plant foods and animals were readily available (Cane 1987; Gould 1968). The volume of water controls the number of people living in a given area. In the Western Desert for example, a marginal area without major river systems and reliable rainfall, Cane (1987: 393) estimated carrying capacities near water to be one person per 150–200 square kilometres. Numbers of people were probably much higher along the river systems of the MDB but without previous research it is difficult to estimate or assume the exact number people. Inland populations in marginal areas managed their population through birth control (Roth 1897). Walter Roth (1897) and Frederick Bonney (1884: 126) provides a graphic account of how healers treated the sick and injured, how they dealt with death, abortion and infanticide. Among these was the termination of a twin baby as these were taboo as it meant mothers were able to travel further to obtain food, and survive during lean food periods. These events appear brutal but such decisions could be life saving for mothers and the family group. Others became ill and too weak to walk, and died (Lowe and Pike 2009). A deceased person was buried in a shallow grave and covered with soil and dead limbs, and the area cleaned around its perimeter, much like the one in Figure 3.3 (Mitchell 1838). Bonney (1884: 134) noted that graves were dug very close to their camp and describes: A grave 3 or 4 feet in depth is dug at a spot chosen not far from the camp where the death takes place, the digger using the sharp-pointed stick called pirrah to loosen the ground, and shovelling out the loose earth with the wooden bowl called yokudjah. The bottom of the grave is covered with boughs from the broom bush, and then the bundle containing the corpse, having been separated from the moolairee stick, is laid in the grave by two men who stand in... 3.2 Aboriginal Setting: Overview | 63 Figure 3.1: Distribution of language groups in the northwest Murray Darling Basin (NW MDB): about 10 languages were spoken in the NWMDB. All of these language belong to the Pama–Nyungan family, and the Maric, Karnic, Paakantyi, Wiradhuric, Yarli, and Lower Murray language classes. The Budjiti language was spoken in the study region (Figure modified from Aboriginal Language Group map compiled by David Horton (1996)) 64 | Living near water sources Figure 3.2: Upper Darling River people photographed in 1884 by Pastoralist, Frederick Bonney (Bonney 1884) Figure 3.3: Burials observed by the Mitchell expedition in 1838 (Mitchell 1838: 816). These burials appear in an unusual place—on top of a range or a very high sand dune? The un-named artist also captures what are likely to be two Aboriginal fires in the background 3.2 Aboriginal Setting: Overview | 65 Burial trees around in the northeast MDB relate to the burial practices of these lowland language groups. These were carved to show the sacred designs of the clan owners and to mark the importance of the area (Black 1941). It is unknown whether these across the study region. How mobile were early peoples in the northwest MDB? Large mobile populations were moving up and down the river systems and into the ranges and out onto the plains during wet or cool periods (Allen 1972). It was not uncommon for hundreds of different language speakers to congregate for large ceremonies Eyre (1845). There were also gender specific roles where, for example, men would move large distances to carry out ceremonies and rituals and return to the camp some weeks or months later (Bonney 1884). In most cases, they constantly moved from place-to-place to fulfil their social and cultural responsibilities of their culture, and their family traditions. Cane (1987: 428) noted that hunting and gathering activities were predictable and reliable, and not overly opportunistic. They obtained many species in a single hunt. The diet included a mix of terrestrial, aquatic, avian, and plant foods (Cane 1987; Clarke 2009; Florin et al. 2020; O’Dea et al. 1991). Food eaten in the camps came from hunting and foraging. They ate acacia and eucalypt seeds, and tubers stored in caches until they run out of food (Cane 1987: 394). Their food intake was about 800 calories per day. Women foraged and gathered seeds from grasses, shrubs and trees and sometimes processed these at the camp (Cane 1987; Gould 1968). In Little Sand Desert, Gould (1968: 105) found charred quandong seeds near most camp sites. Perennial grasses such as Eragrostis eriopoda produced large quantities of small seeds, which women harvested, winnowed, and made into a flat dough bread, almost like an Indian naan bread (Cane 1987; Zeanah et al. 2017). During wet periods, nardoo (Marsilea spp. was harvested and processed by soaking in water for a period of time to remove harmful toxins. The most sought-after food included energy rich foods such as depot fat, organ meats, fatty insects and honey (O’Dea et al. 1991: 78). Bonney (1884: 132) notes that upper Darling River people craved fatty foods and oils, for example, castor oil, which he gave them upon request. To store food such as grass seeds, these desert peoples dug storage pits 5-5.5 inches (10-11 cm) deep (Gould 1968: 112). This was a common practice among groups across the drylands (Smith 2013). Men hunted kangaroo such asMacropus rufus, Macropus fuliginosus, Macropus giganteus, small marsupials, for example, possum (Trichosurus vulpecula), burrowing bettong (Bettongia lesueur), bilby (Macrotis lagotis), long-haired rat (Rattus villosissimus), emu (Dromaius novaehollandiae), bustard (Ardeotis australis), and foraged on many plants (Cane 1987; Clarke 2009). Kangaroos, emus, and euros were more common in the ranges (Allen 1972). In summer, they hunted on full-moon nights when temperatures were much cooler (Cane 66 | Living near water sources 1987). Possums and birds were obtained from trees by climbing to cut a hole in the tree trunk. Large reptiles were chased down and sometime dug from a burrow with a digging stick or by hand. Reptiles are more common in summer and these were a reliable food source (Cane 1987). Hunting kangaroos and emu was sometimes a group effort. The main strategy involved tracking animals to a suitable location, ambushing them and selecting the easiest animals to spear (Cane 1987). Another strategy was to build a brush hide and wait for a kangaroo or emu to arrive at the waterhole. Individuals or an entire family group worked together to guide kangaroos or emus in the direction of waiting hunters. A hunt ended with men carrying a carcass(es) back to camp. These were cooked in shallow pits of hot coals and ashes. Once removed from the fire pit, the cooked animal was laid on a blanket of fresh leaves, usually a broad leaf species such as Eucalyptus populnea, and each part of the animal was cut and divided between the group (Bonney 1884; Clarke 2009). What behaviours occurred near water? Semi-permanent wooden structures across southwest Queensland were built from mulga Acacia aneura or gidgee Acacia cambadgeii and the frame was covered with bark, grass, brush, skins and, the roof sometimes sealed with mud to make it waterproof (Allen 1972; Memmott 2007; Mitchell 1838; Thomas 2007). The frame posts were stood in shallow holes and the fork-shaped tips interlocked to form a dome-shaped structure (Gould 1968). A small fire at the entrance provided warmth (Gould 1968). Shallow channels were dug around the perimeter of the structure to divert runoff water. It is worth noting that these structures are more common in marginal areas of the Murray Darling Basin, Lake Eyre Basin and Western Desert. Other forms of structures have been made with rock but distance to a sources and local material mostly effect the construction materials. Villages and hypothetical farming practices described by Bruce Pascoe (2018) have been part of a significant academic and national debate. Pascoe (2018) extrapolated on the earlier works of amateur archaeologist Rupert Gerritsen (2001) who suggested that early peoples were ’farmers’ and had an intimate knowledge of the Australian landscapes. Early people’s did not till the fields or modify plant species such as wheat, rather people harvested grass seeds, constructed eels and fish traps and through fire, developed advanced food production systems based on knowledge of fire use, plants, animals and, weather patterns (Gerritsen 2001). While this is not the place to enter into this debate because of the general focus of this chapter, it is noted that early explorers observed habitable structures arranged in groups, which they termed a village. Bruce Pascoe (2018) implied people were living semi-permanently in better resourced places without having to regularly move to find food and resources. In my view, the term farmers fails to describe people’s land management and food production systems. Farming is incorrect terminology and more research is needed to debate how to describe the 3.2 Aboriginal Setting: Overview | 67 behaviours of early peoples of Australia. The key, in my view, is to separate Australia from the rest of the world and to look whole-heartedly at the soils, vegetation, environment, climate and the former behavioural practices and technologies. The aim would be to develop a data dictionary that describes early peoples and their land management practices. Desert peoples made camps within a short distance of an ephemeral/permanent water source (Fig. 3.4) (Gould 1968: 112). They camped >250 metres from the only permanent and temporary water sources for both ecological and cultural reasons. Animals accessed the same water source overnight, and strict kinship rules meant that certain family members had to avoid kin. Accidentally meeting a forbidden relative was breaking the law and was punishable by the lawmakers (Gould 1968: 105). Activities around camping areas included clearing small stones and organics with makeshift brooms (Cane 1987). Firewood provided light at night, warmth (windbreak), and fuel for heat and cooking—cooking kangaroo/emu and other wildlife (Gould 1968). A couple, or mother and children, or group of young men or women slept in each shelter (Gould 1968). Mitchell (1838: 837) remarked that he saw a structure that was so large that it could house 15 individuals. Others appeared set up for permanent use (Mitchell 1838: 837). Grindstone slabs arranged beside each other indicated the location of a previous camp site with shelters (Gould 1968: 112). Mitchell (1838: 837) noted permanent camps on either side of large waterholes. Within the camp, men and women made stone tools, weapons, fishing objects, canoes, string bags, water-bags, skin cloaks, and cultural items, for example, headbands, headwear, and other sacred objects made from feathers of various bird species (Kenny 2013; Smith 2013; Thomas 2007). The unused pieces were discarded where the person sat. Allen (1972: 46) points out that stereotypical views of lone hunters armed only with a spear and his own resourcefulness appeared the norm in early Australia. But they used tools such as a net, made bird and animal snare, and sizeable groups hunted together to ensure everyone had access to food. Observations by Mitchell (1838) and members of his exploration party identified 36 men systematically swimming up and down a large waterhole to catch fish. It was also common to see just two men catching fish from canoes. Fish were caught by placing a brush dam across a creek and in the middle of a weir (Mitchell 1838). As the river dried back, quantities of fish were obtained by driving them into shallow pens made of mud. When groups camped beside a lake, they affixed a net in zig-zag pattern about 18 metres (20 yards) from the shore (Allen 1972: 47). Nets for catching ducks, pelicans, herons, and swans were up to 83 metres (90 yards) long and 6 metres (20 feet) wide (Sturt 1849). Poles were stood up and anchored, and one end of the net tied to it (Mitchell 1838). Up to 50 to 100 birds could be caught in a single haul. Aquatic bird hunting was a group activity with women going upstream to chase the birds, and the men used boomerangs to make birds swoop to be caught in the net (Eyre 1845). 68 | Living near water sources Figure 3.4: Painting of a family group camped (Charles Sturt expedition Sturt (1849: 276). It depicts timber-framed shelters covered with grass and brush. Each entrance faces away from the prevailing wind, and fires were made outside shelters. Spears were typically leaned against the shelter or a nearby tree (Gould 1968). Men are in the foreground and women and children standing away from the camp structures. This group appears to be camping on the upper floodplain a short distance away from a small channel (trees in the background) What types of social behaviours occurred near water? This section will discuss early belief systems and how knowledge connected people to water sources via their spiritual and natural worlds (Jordon 2014: 13). It is well known that Elders passed their knowledge down through generations using stories and ceremonies, re-enacting the fine details of events in actions and words (Kenny 2013). Mitchell (1838: 712) remarked that Upper Darling groups had expert tracking skills, and acute knowledge of the lands, and their ability to identify water source locations with pinpoint accuracy. The ability to remember and transfer knowledge was shared through songs, sand drawings, and by stories that were shared among the individuals. They needed this to travel vast distances with confidence in some of the driest areas in the world (Kenny 2013). Their stories focus on the natural environment: the moon, sun, stars, fire, rain, trees, landscape, and mythical creatures, for example, bunyips, biame, little people, and the rainbow serpent, a large snake that occupies river systems and waterholes (Clarke 2007). They believe that the rainbow serpent still lives in these large deep waterholes and controls inland waterways (Hercus 2009). Mythological beliefs are strongly associated with storylines, songs, rituals and behaviours (Kenny 2013). Their stories also describe catastrophic floods, fires, large meteorite impacts, and animals and beings that created the landscape as they travelled across it. 3.2 Aboriginal Setting: Overview | 69 The Budjiti people had many dreamtime stories or legends are known for a major waterholes of the Paroo River. Knowledge passed down through the generations speak of people being able to catch fish by hand. And a fateful story of a man who wrongfully married a woman and for his punishment the old men danced and created a massive rainstorm to drown the man as he swam across the river to escape the penalties of Budjiti law. He was never seen again. Legend has it that his body may be seen at the end of the waterhole and when it is exposed, rain comes and floods the river to hide his body again (McKellar 1984). What types of intangible behaviours occurred near water? The water culture of early peoples encompasses the songlines of spirit beings that describe water sources within the broader estate-owning groups: how they formed or what events occurred near them (Langton 2006: 154-155). These intertwine with complexity of connectedness through the songlines and stories that criss-crossed the continent (Atkinson 2002; Campion 2012; Kenny 2013; Norris and Harney 2014; Radcliffe-Brown 1930; Reay 1949). The mythical ancestors met at key water sources. From here the mythical ancestors went underground, while others disappeared into the skyworld and formed motifs in the star constellations and dark regions of the Milky Way (Kenny 2013; Langton 2006). Marcia Langton (2006: 139) in the volume, The social archaeology of Australian Indigenous Societies describes water as sacred and elemental source and symbol of life. Water cosmology holds great significance to early peoples. The mythical creatures: Rainbow Serpent, i.e. snake(s), bunyip, and wallaby and possum travelled across the landscape to create springs, waterholes, rivers, and waterfalls (David et al. 2006). On this journey, the ancestors created water sources and they sang songs and told everyone how these inter-connecting stories describe the journeys and places (Kenny 2013). The medicine men held the Rainbow Serpent as a personal Totem (Radcliffe-Brown 1926: 343). The Rainbow devours anyone that swims near their domain (Clarke 2007; Radcliffe-Brown 1930: 149-150). Radcliffe-Brown (1930: 342) describes the Rainbow Serpent as a deity, and perhaps the most important nature-deity. Radcliffe-Brown (1930: 347) further states: the rainbow-serpent may be said to be the most important representation of the creative and destructive power of nature, principally in connection with rain and water. It is apparently as such that it played a considerable part in the initiation ceremonies of some of the tribes of this region. In New South Wales, the snake is a crocodile-like creature with large eyes like crystals, the colour of a rainbow, and ears that stick up to warned them about the law associated with the waterhole (Radcliffe-Brown 1926; 1930). If anyone wrongfully swam in a waterhole, the snake devoured them. In the Murray-Darling Basin, two snakes travel across the landscape 70 | Living near water sources from waterhole to waterhole, and out to the desert regions. Barkanji men told ethnographers that their ancestor disappeared into the ground at Peery Lake, about 50 kilometres (30 miles) south of this study area, and continue to live under the lake (Allen 1972; Radcliffe-Brown 1930). Anna Kenny (2013: 42) describes ancestral animals creating a water source: wallaby ancestor joins some possum ancestors for a ceremony and go together into the ground there creating a water-source. [Further], "they live in the sky which is imagined as an eternal land with permanent water, trees, flora and fauna" (Kenny 2013: 34). Carl Strehlow in Kenny (2013: 42) discusses how language groups shared similar stories but with a different type of ancestor. Strehlow is quoted as saying that the wallaby ancestor joins some possum ancestors for a ceremony and they go together into the ground there creating a water-source. The Loritja has a wallaby ancestor and the Arrernte a possum. Their stories tell of both wallaby and the possum disappearing into the waterhole. Physical landscape features show that the Rainbow Serpent lived in deep waterholes or below the ground and carried out tasks sometimes only known by the medicine men. The story lines are sometimes associated with physical features in the landscape that resemble faces, body shapes and features, eggs or other things that relate to the creature is a phenomenon known as ’pareidolia’ (Blackburn 2016). The tracks and body parts of ancestral beings create a physical and cultural landscape. The events, which usually end with the ancestors metamorphosing into natural features (Kenny 2013: 40), hold physical, social and cosmological importance to early peoples. Children inherit these through stories and songs from birth and hold transmissible rights for social actions, rituals and etiquette, language and discourse strategies, and kinship (Langton 2006: 148). Western Arrernte speakers of central Australia identify with birds, eggs, lizards, sun/moon, clouds, rain, floods, trees, and believe that almost every physical thing shapes their lives (Kenny 2013). They walked these songlines and sang songs to maintain these cosmologies (Campion 2012: 31). The songlines, which cross the night sky and the landscape, describe mythical creatures, and animals. Transmission of these songs continue down through each generation and boys and girls born into this cosmic scheme continue to learn the ways of their ancestors and maintain songs and belief systems through dance, stories, songs, and rituals Kenny (2013). Increase Ceremonies: How to bring rain and resources? Increase ceremonies were performed to bring rain and renew resources (Clarke 2009; Kenny 2013). In good times, many ceremonies would be conducted at key points in the landscape. It is well known that early peoples travelled hundreds of kilometres along their ancestral 3.2 Aboriginal Setting: Overview | 71 routes and attended important ceremonies and performed rain-dreaming dances and rituals. Pitjantjarra men (central Australia) perform a rain-making ceremony to bring rain. Tindale (1974: 68) quotes this event: they shout out very loudly with special cries to tell the [wanambi] or giant carpet snake they need water and to come and fill the tjila again for them. Body scarification was common among inland groups (Flood 2006). Incisions made with stone tools (or marsupial teeth – lower jaw) in the upper arms and legs, chest and back and buttocks (Allen 1972: 130). Open wounds were filled with ash to aid in the healing process. This creates a raised scar to demonstrate their connections to land and culture (Akerman 1994; Flood 2006: 150). Body scars sometimes represent blood-letting and a ritual practice associated with rain-making. These scars represent the importance of rain and with magical powers of the rainmakers they are able to bring sometimes much needed rain to the region. Figure 3.5 shows the extent of scarification inflicted on the body (Allen 1972: 114). Substantial amounts of pain and blood must have been involved in these practices, which demonstrates the importance of ceremony ritual and it connection to the landscape and rain. In the MDB, rain features are piles of white stone, for example, gypsum, quartz crystals and standing stones (cylcons) (Black 1942). Other significant cultural features which relate to death are the carved trees of the upper eastern MDB (Black 1941). These were carved into the large Eucalyptus populnea trees to signify the death of relatives (Black 1941). Other death-related cultural items are the widow caps, which were made of gypsum and worn as a skull cap to mourn the loss of their husband (Kenny 2013). These cultural items relate to the diversity of cultural practices found across the northern MDB region (Allen 1972). Rock art sties, rock features, and trees comprise many stories of the mythical creatures and belief systems that control how they approach to water and/or walk through another language group’s country. TheWandjina, the rainbow/rain-being (a painted face with eyes and nose but no mouth) holds great significance to Kimberley groups (Cowan 1992: 48).Wandjina controlled seasonal regeneration and all natural resources. These relate to the lightning man and the rainbow serpent, which are painted in the rockshelters of the Kimberly, Western Australia. Cylcons or standing stones and carved trees found in the northwest MDB and were used for rituals (Black 1941; 1942). Cylcons were etched with sacred markings and were placed upright with the narrow end to the top for ceremonial purposes, and their significance was maintained by law-makers. These cylcons are about 250 mm – 300 mm (10 in. – 12 in.) long and 6 cm – 9 cm wide (2.3 in. – 3.5 in.) at the base. These relate to increase ceremonies described above (Allen 1972). 72 | Living near water sources Figure 3.5: Drawing of a man with body scarring (scarification) on the upper arms, chest, and upper abdomen (observed and drawn by Major Mitchell and his party) (Mitchell 1838: 718) 3.3 How did early peoples cope during extreme aridity? The most common types of human water sources in the Murray-Darling Basin are riverine waterholes, freshwater lakes, springs, soaks, claypans, and rockholes. These vary in water quality, scale, spatial patterning, and permanency. The latter depends heavily upon weather patterns, water source characteristics, e.g. water-holding capacity; and replenishment source, for example runoff, and occurrence of subterranean aquifers. People of the the Murray-Darling Basin (MDB) made kangaroo skin waterbags to take with them on line walks (Flood 2006: 26). Riverine waterholes and lake waters slowly evaporate during major dry phases and once water levels are significantly reduced they become unfit for consumption or use (Allen 1972). But 3.3 How did early peoples cope during extreme aridity? | 73 fresh drinking water bubbles to the surface from sandy substrates of stream beds and lakes. It is well known that Aboriginal people and dingos excavate sands to expose water just several feet below the surface. Most Aboriginal groups knew where to find water in some of the driest regions (Allen 1972; Bayly 1999). Smith (2013: 4) noted, Australia’s deserts have changed over time and aridity has intensified or relaxed, and deserts have expanded and contracted with periodic reactivation across the millennia. Extreme aridity during the Last Glacial Maximum (LGM) forced the abandonment of the arid zone (Smith 2013; Veth 1989; 1995). Ten human refugium locations were identified across inland Australia such as the Australian ranges and uplands, Carpentaria gorges, lake systems, Cooper Creek (located to the west of this study), the river systems of the Murray-Darling basin (Smith 2013: 116). This left open the possibility of small local refugia outside these areas during extreme arid phases, for example during the LGM. The abandonment model was widely accepted by archaeologists but doubts have been cast about its specific detail (Smith 2013: 39). The persistence of people in spinifex and sandhill landscapes west of the main MacDonnell Ranges identified that the inland ranges functioned as a human refugia. These landscapes comprise small water sources and sufficient resources were able to support viable populations without the need for total abandonment (Smith 1989). Archaeologists have been attempting to track the movement of early peoples. Bird et al. (2016) modelled the movement in combination of studies including 14C chronometry and correlations with water sources (Bird et al. 2016). They modelled the movement of people across rough terrain and watered landscapes to reach all regions of the continent. Other methods include stratigraphic sequences contained within rockshelters (Clarkson et al. 2017; Smith 1989; Thorely 1998), and open air sites (Bowler et al. 2003). DNA research has also been used to examine the movement of early populations (Nagle et al. 2016; Tobler et al. 2017). After arrival in Sahul, people would have moved further into the interior but as they extended further inland, they would have been faced with open arid landscape with extensive sand dunes, salt lakes, and extensive drylands (Smith 2013: 69). Peter Hiscock and Lynley Wallis devised a desert transformation model and proposed that, "people colonised terrestrial landscapes using flexible and general broad spectrum foraging focused on available riverine and lacustrine resources" (Hiscock and Wallis 2005: 49). People’s economic behaviours were in response to changes to their familiar environment. The ethnographic and archaeological literature supports mobile populations in response to climate and seasonal variations across the millennia (Allen 1972; Pardoe 1994; Smith 2013). Mike Smith (2013: 154–155) argued that survivors of the first phase of the LGM continued to live in the region but stresses became greater as temperatures increased across dryland between 17 Ka BP to 16 Ka BP. He added though that there is nothing in the archaeological record to confirm long-term movement in response to LGM climate flux (Smith 2013: 155). Instead, they tethered themselves to a series of permanent water sources and limited their 74 | Living near water sources movement to seasonal responses to change (Smith 2013: 155). The final part of the model concerns ’cryptic’ refugia—microhabitats found across former landscapes (Smith 2013: 112). In this situation, small populations inhabited small areas, which encouraged social reproduction and demographic variability at a regional scale. Scott Cane (1995) suggests that we need to examine local and regional landscape variation to understand colonisation and settlement patterns. What were the subsistence strategies? Archaeological research at Youlain Springs by Richard Robins (1998) showed that the springs to the west of the two lakes had been occupied up until about 14,000 years ago. Burned marsupial bone found at the site implied feasting on these animals, some of which are now extinct (Robins 1993; 1998). These lakes provided resources all year round and these became meeting places for adjoining language groups. Previous research has shown that these are lakes were occupied 1700 years ago (McKellar 1984). In drylands, movement between waters allowed an extension of people into an array of environments (Cane 1987; Clarke 2009). Each season controlled distances travelled and the types of food resources that were available (Cane 1987; Clarke 2009). In the Upper Darling River region, Harry Allen (1972: 55) presented a subsistence strategy based on historical accounts and his doctoral archaeological research. He noted that people in early spring moved out into ranges and uplands to obtain terrestrial animals. In late spring, they moved back to the river for fish, mussels, crustaceans, and plants. From summer to autumn, they continued to live along the river and catch aquatic species, marsupials, and harvest plant species. In winter, they moved away from the river to hunt kangaroo, emu, and birds of the plains, harvest seeds and fruits. Some plants and roots of a tree provided water in drylands Mitchell (1838: 701) noted young men digging up tree roots and drinking its juice. Perhaps the most important subsistence strategies was the use of fire to manage extensive areas of land (Jones 1969). Jones (1969) reported that fire was essential for survival in many of the Australian landscapes including Tasmania. Jones (1969) coined it firestick farming and noted that fire was of great value and integral to the hunting and gathering economy. As grass biomass increased, it was time to burn and ‘clean’ the country (Kimber 1983: 39-40). Early peoples knew when, how and why to burn the landscape and it was performed with precision, knowledge and regularity (Gamage 2011: 167). Explorers equated the number and size of fire scars with population size, i.e. higher populations create large numbers of fire scars (Eyre 1845). Fire stimulates growth in plant species (Cane 1987; Jones 1969; Latz 1995; Smith 2013). Fire use across marginal landscapes allowed early peoples to control plant growth and to maintain a patchwork of landscapes at various stages of growth (Bird et al. 2012; Gamage 2011; Jones 3.3 How did early peoples cope during extreme aridity? | 75 1969). Jones (1969: 228) believed that this technique had the greatest effect on Australia’s ecology and he proposed that this style of land management may not yet be over. It is worth noting that Rhys Jones published this paper in 1969 and it is clearly relevant and used across Australia for managing spinifex biomass and marginal landscapes. Kangaroos, emus, and bustards were cooked in pits filled with hot coals and ashes (Gould 1968). Fish were wrapped in tea-tree barkMelaleuca linariifolia or coated in mud and placed in the hot ashes (Allen 1972). (Clarke 2009: 151–152). Small marsupials and reptiles were cooked in the hot ashes. Burning the landscape was not the only fire use. The Pintubi language speakers of the western desert for example, had 32 different uses for fire (Kimber 1983: 40). These included cooking, cleaning the landscape, hunting, and ceremonial uses such as performing a fire magic ceremony. In addition, Kimber (1983: 40) observed men being delighted when they saw large fire scars. But a smouldering large dead tree caused the men to throw sediment over the embers to prevent it from igniting and burning the landscape unnecessarily. How did early peoples cope with dryland conditions? To cope with dryland conditions, people moved between water sources to obtain an array of food that is available to them in these landscapes (Cane 1987; Gould 1968). They followed their songlines and, spatio-temporal routes which depended on resources, and whether they were confident that water was available when they arrived (Cane 1987; Gould 1968: 103). During dry periods they stayed near the water supply longer to forage for food and resources before walking to the next water source (Cane 1987; Smith 2013). Their subsistence strategies and tools for survival relied on a heightened senses to direction and distances between each water source and, resource use. This required knowledge of the landscape and its food and resource production (Cane 1987; Gould 1979; 1968; 1971). Once at a camp, they radiated out a short distance from water to the ranges to collect stone and hardwood and/or medicine plants (Allen 1972). The foraging range around a water source was about 10-15 miles (18 to 24 kilometres) (Gould 1968: 104-105). From March to May, women gathered tubers, and grass and acacia seeds (Ipomoea spp.) (Cane 1987: 394). June to July was a time to harvest solanum fruits and to relax and conduct ceremonies. August-October begins with the warmer winds and is called goanna get up time. November-February is summer. Temperatures range from 35 to >40 degrees Celsius, and evaporation rates are very high (Chapter 2). This impacted on surface water availability. Movement between distant water sources only took placed if they were confident it could sustain them during the hot days and nights. To cope with hot conditions, people only hunted and gathered when necessary but as water dried up, they retreated to reliable water sources (Gould 1968: 105). These were 76 | Living near water sources dynamic environments and behaviours changed in response to seasonal variations (Allen 1972: 51–64). Coping with Temperature Extremes How did early people’s cope with such extreme temperatures? Heat and cold weather stresses posed survival problems. November-February is summer, when temperatures range from 35 to >40C, and evaporation rates are very high (Chapter 2). From May-July, temperatures are below freezing overnight, and strong winds and frosts persist (Chapter 2). To cope with these extremes, shade shelters and windbreaks or sand dune depressions helped to manage body temperatures and reduce exposure to sometime relentless weather conditions (Lowe and Pike 2009: 43-46). For example, on cold winter nights, fires were arranged in a triangle to provide warmth (personal observation near Alice Springs, Northern Territory, June 2015). Tree/shrub limbs and/or grass hummocks placed upwind of sleeping areas provided warmth when overnight temperatures fell below 0C (Gould 1968: 112). In summer, people sat in the shade from mid-morning until late afternoon, or buried themselves in damp sand to reduce hidrosis (Lowe and Pike 2009: 43-46). Other strategies included the application of ochre and animal fats, for example, goanna and emu or grub fats to the body (Clarke 1989). In waterless areas, women carried water in coolamons (a wooden dish or bowl) or they made depressions in antbeds to catch and retain rainwater. These temporary sources provided a family with water until they could make it to the next source (Cane 1987). 3.4 What happened when Europeans arrived? It is well known that once colonisation occurred, the culture broke down enormously and people were forced from their traditional lands (McKellar 1984). During the 1919 flu pandemic, many Budjiti and traditional people died. Community Elders spoke of no medicines and/or treatment, and the police shot many of the people suffering from the virus. What followed were the dispersal of people to government missions and town camps in nearby towns such as Cunnamulla. Pastoralism period (1870 to 1991) The pastoral period for Eulo Ridge began circa 1870 and continued to 09 May 1991 when it was gazetted a National Park under the provisions of the National Parks and Wildlife Act 1975 (QPWS 2001: 2). Pastoralism represented a complete transformation in the use of the land/landscape. From the 1870s, tens of thousands of sheep, unknown numbers of goats, cattle and horses were introduced over a few decades (Torrance 1870). Local materials such as grasses were 3.4 What happened when Europeans arrived? | 77 used to thatch the original buildings, and to feed many of horses suffering from long periods without rainfall (Torrance 1870). Water bores were installed in waterless areas, to extend the grazing carrying capacity, and fences were erected to control the numbers of stock in each area. A major issue for the new land managers was the semi-arid environment with limited annual rainfall, very high summer temperatures and high evaporation rates, and major flooding in the Paroo River (Torrance 1870). Since the pastoral period, the rabbit, pig, goat, sheep, cattle, horse, cat, fox, and wild dogs populations have affected the local ecology and environment (NPRSR 2014). This has contributed to the loss of habitat and reduction and extinction of some reptile and marsupial species. In the 1960s, continual land management pressures such as low rainfall, gradual reduction in grassland biomass, increased shrublands and Acacia aneura forests, and dwindling stock losses, led to a pastoralist organisation lobbying the Queensland government to investigate the future of this region (QDPI 1974). The outcome was the Western Arid Land Use Study (WARLUS) and considerable effort by a group of multi-discipline specialist officers within the Development Planning, Botany and Agricultural Chemistry Branches within the Department of Primary Industries undertaking a study of some 15 million hectares (QDPI 1974). The Western Arid Land Use Study concluded that mulga sandplains have been degraded as a result of poor land management, fires and overgrazing. Soil erosion is a major problem across the region (QDPI 1974). Erosion has been largely due to excessive reductions in plant cover and animal trampling associated with overgrazing and intensive use of areas around water sources. This has been further heightened by the felling of mulga for fodder during droughts or periods without significant rainfall (QDPI 1974: 130–131). Post-colonial contact Post-colonial contact was the most severe and harshest period for people of the northwest MDB (McKellar 1984). This included an influx of diseases such as small pox (Variola major and Variola minor), influenza, pulmonary and sexually transmitted diseases which impacted heavily on earlier populations after the arrival of Europeans (Flood 2006: 88–90). Major Thomas Mitchell (1838: 861) on his expedition of the upper Darling River in 1838 found large numbers of graves and saw men and women with small pox scars. They deduced from this that diseases had reached these isolated groups well before Europeans had made it to the upper Darling River (Fig. 3.3) (Kefous 1988; Mitchell 1838). Many deaths occurred as a result of the 1919 Flu epidemic and starvation (McKellar 1984). As mentioned in the previous chapter, the police shot many sick people because of poor health, no medication, and starvation. It is for this reason, population estimates in northwest Murray-Darling Basin and other regions should not be considered as being accurate. The exact population of the northwest Murray Darling Basin may never be known. 78 | Living near water sources Starvation was most likely caused by the competition with sheep, goat and cattle grazing for both food and water. Large numbers of grazing animals ate the grasses and plants that the Budjiti relied on for food, and their hard-hooves disturbed land surfaces and silted the waterholes (Hercus 1980). Additional impacts were the introduction of European rabbit and cat that multiplied to uncontrollable numbers. Many small marsupials and reptiles disappeared and inadvertently contributed to the destruction of the Budjiti’s culture (McKellar 1984). An additional contributing factor was the introduction of large number of water bores installed in the late 1800s. Installation of water bores in the late 1800s and early 1900s encouraged excessive water use that resulted in lowering of local aquifers, and extinction of once active springs (Fairfax and Fensham 2003).These bores fed hundreds of kilometres of drains and carried water into once waterless areas. This extended the grazing areas for sheep, goats, cattle and horses, and kangaroos and emus populations increased, which essentially expanded the carrying capacity and reduced the pastures even further. Water usage appeared endless and water was allowed to flow freely (Fig. 3.6). Fairfax and Fensham (2003) argued that spring discharge has decreased by 62 percent in the 20th century, and the Great Artesian Basin (GAB) Hydrological Zone decreased by 39 metres. Many springs now cease to flow and are identified as extinct and all that remains are the spring mounds and depressions and an occasional spring seeping water or contained within a small pool. Those people that were able stay on their lands worked as shepherds and pastoral workers while others moved to Eulo and Cunnamulla (McKellar 1984; Torrance 1870). They moved to fringe camps near Eulo and Cunnamulla and afar to the Government Mission at Cherbourg, which is located many hundreds of kilometres from their homelands in eastern Queensland (McKellar 1984). In the late 1960s, graziers lobbied the Queensland government to address the continuing decline in wool and cattle production in these marginal landscapes (QDPI 1974). Consistent over-stocking for almost one hundred years had seen a major decline in pastures and increased non-productive species, for example turkey bush, other Eromophila spp. and Dodonea spp., and widespread erosion had denuded hundreds of thousands of hectares of land. These processes were destructive and this could be why Holocene archaeological sites are more prevalent in inland regions (Fanning 1999). 3.5 Chapter Summary This chapter presented an overview of early people’s attachment to water. The objective was to present the types of behaviours around water that could contribute to site formation processes. Behaviours such as the use of fire, camping activities, building structures (gunyas), and processing and cooking food entail the major activities that could be identified near springs. Knowledge of these are necessary to formulate questions about archaeological sites, and how this contributes to site formation processes and the interpretation of archaeological sites. 3.5 Chapter Summary | 79 Figure 3.6: Man bathing in the water from an artesian bore in the Cunnamulla District, Queensland, ca. 1894 (SLQld) Noorama No. 1 Bore. Depth 1,502 ft. Yield per 24 hours 2,304,000 gallons. Water temperature at outlet 110°F or 43°C (https://www.slideshare.net/slqlibrary2/artesian-basin-2010, accessed 03 December 2019) In dryland they adjusted their behaviours to live in some of the driest areas of the world (Cane 1987; Gould 1968). This links to highly adaptive subsistence strategies, the development of skills and technologies, and making use of the available food resources. These contribute to site formation processes and the archaeological record. It was also important to know that water is sacred and that most of their activities occurred within proximity to water sources. This is important for addressing the overall research question and developing a model for human settlement in the study region. Early peoples coped with climate phases through a belief system and survival strategies. These provided individuals, families, and the broader group with skills to obtain resources as conditions oscillated between arid and humid conditions. As John Mulvaney (2012: 916) states, "familiarity with their landscape and its resources, together with an ability to adapt 80 | Living near water sources to local conditions imply culturally determined patterns of behaviour, which facilitated language". In the following chapters, this thesis turns to the geoarchaeological evidence and the methodology used to test the aims and objectives of this study. One of the key questions at this point: could geoarchaeology help imply previous behaviours, and the types of palaeoenvironments early peoples endured in the past? Throughout there is a recognised limitation of either poor preservation of organic cultural items buried within a deposit or post-depositional processes in surface scatters are without spatial and temporal resolution (Fanning et al. 2009; Fanning and Holdaway 2001; Holdaway and Fanning 2014; Robins 1997). A further issue are the relatively minimalistic behaviours described above and whether it is possible to find evidence of people in open air archaeological sties along Eulo Ridge? The next chapter describes the approaches used, along with individual methods employed to address the overall research question. Chapter 4 Case Studies: A multi-proxy approach Michael Connolly (Artist) – Platypus The Ngatyi traverse country belonging to two groups of Paakantyi people: the Paaruntyi from the Paroo and Pantyikali people, also called ‘Wanyiwalku’, further west. Paaru-ntyi literally means ‘belonging to the Paroo’, and refers to both the language and the people. Pantyi-kali literally means ‘the Creek people’ (Hercus 2009: 14). 82 | Case Studies: A multi-proxy approach 4.1 Introduction This chapter describes the 3 case studies and research design developed to address the overall research question. Eulo Ridge contained evidence of complex interactions between site formation processes and archaeological material. A set of minor nested questions was asked of each case study: why was each study area chosen and, what geomorphic processes and environments could be identified? The foundation of this approach derives from Benedetti et al. (2011: 83) who point out that a geoarchaeological approach should incorporate remote sensing, geospatial analysis and, advanced sampling and analytical techniques such as micromorphology, sedimentology, geochemistry, and dating methods. Before the methods are presented, the first section characterises the 3 case studies to setup a data presentation and analysis framework for the following 2 chapters. This chapter addresses the questions; how was the data collected?, how was a chronology established for the study region?, and what types of method and techniques were used to reconstruct the palaeoenvironments? The research design objective is to locate three study areas and to implement a multi proxy approach to test and analyse the data at various scales including microscopic detail. Criterion includes distance to active/inactive springs, depth of arroyo, sediment types, and presence of stone artefacts. Unconformities elevate site importance and methods and techniques aim to reveal if sites formed as a result of colluvial, alluvial events or former surfaces had been buried by younger soils. 4.2 Preliminary fieldwork and setting up this study To carry out this study, a Scientific Purposes Permit, No. WITK18586717, was obtained from the Queensland Department of National Parks, Sports and Racing, in consultation with the Budjiti people. In addition, visits were made to Budjiti people in Adelaide and Eulo to explain the nature of this research. In the setup phase, QGIS (https://www.qgis.org/) and digital datasets and Google Earth Pro (https://www.google.com/earth/versions/) allowed the identification of areas of interest in point and line layers. A primary goal is to find arroyos with stratified deposits and archaeological material located near groups of artesian springs. Small valleys located along the margin of Hoods Range provides the backdrop of this study. The criterion for selecting a study site includes evidence of sediment deposition and distinct layers that reveals its previous site formation processes and disconformities (cf. French 2015). Shuttle Radar Topography Mission (SRTM) 90 metre pixel resolution DEM (digital elevation mode) obtained from EarthData (https://earthdata.nasa.gov/) assisted with modelling stream flows and the selection of arroyo locations. Vector data, topographical and landscape datasets, was obtained from Geoscience Australia (www.ga.gov.au). Other digital data obtained from 4.3 Overview of the case studies | 83 the Queensland government aligns this study with cadastral and environmental data (www. data.qld.gov.au). QGIS is used to visualise and process the data and export digital outputs to map layers for use in a Hema Map HX-1 device (https://shop.hemamaps.com) global positioning system (GPS) to take in the field for ground-truthing and field mapping. Google Earth Pro helped to check features and refine the location of search areas. 4.3 Overview of the case studies Three case studies present the framework of the following methods and analysis in subsequent chapters (Fig. 4.1). Case Study 1, the Granites, is located at the southern end of Hoods Range within a small valley; Case Study 2, Hardpan Creek and Basin Gully, is located on the southeast side of Hoods Range; and Case Study 3, Double and Tunkana Wells, is located along the eastern margin of Boorara Creek in the very north of the study region (Fig. 4.1). A set of criteria helped choose these case studies, is the arroyo located near springs?, are there visible unconformities within the profile?, is sand/silt/clay visible in the matrix?, and/or is archaeological material embedded within any of the unconformities? Positive results aim to test episodic erosion across each case study area. 4.4 Case Study 1: Granites The Granites study area measures about 1 Km2 and includes a small valley, relict dune, and phreatic zone (Fig. 4.2b). The phreatic area is dominated by a mix of mound springs, concave vents, and spring seepages level with the surface. The phreatic zone margins are covered by deep red sediments which appears to have been transport from upslope. It is likely that the phreatic zone was much larger in areas in the past. A small channel that cuts along the western margin of the phreatic zone exposes the sediments that encroach the phreatic area. 84 | Case Studies: A multi-proxy approach Figure 4.1: Oblique view of case areas: 1) Granites; 2) Hardpan Creek and Basin Bore and; 3) Double and Tunkana wells (image extracted from Google Earth, 09 March 2020) A terrain model characterises the topography of this small valley. It is located at the southern and western ends of Hoods Range and is surrounded by gently eroded slopes (Fig. 4.2a). Surface and gully erosion accounts for much of landscape modification in this area. Above the phreatic zone and relict dune, the range slopes away to the west and reveals the underlying geology including the large granite tors and gravels. These slopes meet the phreatic zone at the plain and range margin. The phreatic zone extends from the foothills of Hoods Range for 1.4 km in a south-westerly direction (Fig. 4.2b). The focus of this study is a 500 m by 350 m area at its widest point. The springs vary from pyramidal mounds to concave depressions, or raised flat mounds about 2 m high. These are mostly inactive except for small wet areas <1 m in diameter and 4.5 Granites East Study Site | 85 depressions filled with water. The main southwest-flowing channel begins at 166 m ASL and meets the Granites East channel at 133 m ASL. Water flows west and spreads out into the smaller channels and eventually into the larger creeks systems of the lakes and Paroo River floodplain (Fig. 4.2a). The Granites study area was chosen because of a large phreatic zone with relict dune with horizontal stratigraphic features and stone artefacts embedded the upper and lower deposits of phreatic sediments. These patterns infer multi phases of human activity and palaeoenvironmental change over time. This arroyo and nearby gullies provided an ideal opportunity to examine site formation processes within this study area. An additional reason for choosing this study area is provided by two outcrops of large Devonian-age granite tors located on the western slopes of Hoods Range. These tors probably symbolise the eggs of the ancestral snake that created the springs and water features in this area (Beckett 1967; Clarke 2007; Hercus 2009; Ingold 2000; Radcliffe-Brown 1926; 1930; Thomas 1923; Willcox et al. 1987; Witter 2007). People probably visited this area to conduct ceremonies and rituals and in doing so they camped on the nearby dune located on the western margin of the phreatic zone. If so, it implies that people visited the Granites for millennia. I have not provided a photo of these tors out of respect of seeing cultural places that belong to the Budjiti people. 4.5 Granites East Study Site The Granites East study site is located along a gentle meander of the southern channel at the base of Hoods Range (Fig. 4.2b). Run-off has destabilised the phreatic zone on its eastern margin and created a shallow channel (Fig. 4.3). Erosion has slowly cut into the phreatic zone and exposed its stratigraphic features. The focus is a 20 m section located on the outer bend of this channel. Stone artefacts are embedded within the arroyo profile. Archaeological material: Approximately 20 stone artefacts per square metre was found on the surface above the channel. Raw materials includes cream to brown silcrete and chert. These were typical of stone artefacts previously recorded in this region (Robins 1997). Sizes were <60 mm long and the specimens was mostly debitage and flaked pieces and embedded horizontally or angled at ~45 deg and within <5 mm wide fissures. Figure 4.4 shows examples of the stone artefacts. 86 | Case Studies: A multi-proxy approach (a) Terrain model: study area located within small valley (b) Phreatic zone with 2 minor channels. GD and GE study sites (Google Maps, accessed 30 November 2019) Figure 4.2: Granites Case Study characteristics 4.5 Granites East Study Site | 87 Figure 4.3: Arroyo located on the outer bend of small channel (2m scale) Figure 4.4: GE: stone artefact embedded in phreatic sediments. Discolouration shows burial depth Four heat retainers was haphazardly dispersed near the channel. Small clumps of rounded to sub-angular blocky iron-rich rocks <0.9 m in diameter, sometimes charred and tarnished or discoloured by firing or combustion was found on the surface (Fig. 4.5). Previous research implies that these heat retainers are <1,200 years (Robins 1996). Exposure of intact heat retainers suggests gentle surface deflation and site stability. This implies gentle erosion with good preservation of heat retainers. 88 | Case Studies: A multi-proxy approach (a) Two stone heat retainers (b) Heat retainer (~70 cm dia.) Figure 4.5: Granites East: heat retainers a) two heat retainers (arrows) have been exposed by surface deflation and sheet wash (25 cm scale increments) Non-cultural quartz cobbles was found embedded in the channel wall upstream (Fig. 4.6). Large rounded to sub-angular cobbles implies evidence for a colluvial event or spring activity dispersing rocks during major groundwater events which re-distributed these cobbles and sediments. Figure 4.6: Granites East quartz cobbles cemented into upstream arroyo profile (30 cm scale) 4.6 Granites Relict Dune Study Site | 89 4.6 Granites Relict Dune Study Site This Granites Relict Dune (GD) site is located on the western side of the phreatic zone, at the margin of a relict dune and small flat-bottom channel. This west-flowing channel cuts along the southern margin of this relict dune to expose its deposits and archaeological features. The site is about 10 m long and has an average depth of 1.1 m (Fig. 4.7). Low shrubs and dry grass tussocks, probably Eriachne helmsii, and its fine roots penetrate deep into the dune profile. The shrub and tree canopy comprises Dodonaea spp. and Eremophila spp., and Eucalyptus populnea trees (Boyland 1974). This site was chosen because of stone artefacts found embedded in a lateral indurated layer at the base of the profile A flat basal layer assumed a separate climatic phase to the above layer that is loose and vulnerable to erosion on the dune margin. The base lateral layer assumed stable or static period without change followed by a dune-building phase. This scenario provided an ideal opportunity to investigate site formation processes and infer a chronological sequence of palaeoenvironmental events. It is hypothesised that significant palaeoenvironmental change following a long period of stability and more recent dune building phase. Figure 4.7: Granites Dune arroyo exposed by channel flow 90 | Case Studies: A multi-proxy approach Archaeological material Channel flow has undercut and eroded the face of the relict dune and exposed stone artefacts embedded in the arroyo. Archaeological material includes a core and flakes cemented into the base or Static layer along a buried soil layer (Fig. 4.8a and Fig. 4.8b). Appendix B.14) provides photographs of other stone artefacts found throughout this layer and dune profile. Cream-coloured crypto-crystalline cores and flakes dominated the assemblage. (a) Stone artefact locations (b) Cores and flakes embedded within Static layer Figure 4.8: GD: location of stone artefacts within the relict dune) 4.7 Case Study 2: Hardpan Creek and Basin Gully | 91 4.7 Case Study 2: Hardpan Creek and Basin Gully Hardpan Creek and Basin Gully study area is situated at the south-eastern edge of Hoods Range, a relict Tertiary-age range capped with silcrete and eroded along its slopes and lower margins. Hardpan Creek is situated 3.4 kilometres from the base of Hoods Range, and Basin Gully is located about 500 metres to southeast of Hardpan Creek (Fig. 4.1). At its highest point of 188 m ASL, it drains in a southeast direction and flows out into a lowland swamp at 136 m ASL. The study area (dashed square box, Fig. 4.9a) is located between 144 m ASL to 142 m ASL on a flatter land surface than the surrounding features. The Hardpan Creek and Basin Gully study sites was chosen because of linear stratigraphic features and stone artefacts embedded in two profiles (middle and base layers). Richard Robins (pers comm., 29 March 2018) noted he has recorded stone artefacts eroding from this arroyo and implies a human presence of some antiquity. With this evidence, and observations from this study, this arroyo and nearby gullies provided an ideal opportunity to examine climatic shifts and an ecological baseline for this area. A presence of archaeological material and stratigraphic change is likely to indicate connections to palaeoenvironmental change in eastern Australia. It is worth noting that the nearest known permanent water sources is a group of extinct springs (convex shaped depressions) located within a grove of western bloodwood trees (Grevillea striata), some 400 m east of these channels. The Hardpan Creek (HPCK) site is located on the northern side of this valley at ~144–146 m ASL, which is about 6 m below the base of Hood Range. The site is about 8–10 m above the valley floor and surrounding swamps (Fig. 4.9a). Hardpan Creek North (HCKN), Central (HCKC), and South (HCKS) sites comprise the study sites for this research (Fig. 4.9b). The HCKS and HCKC sites is located about 65 m to 75 m downstream of the HCKN site. Channel erosion has partially exposed all sites and appears to shape the arroyo and surrounding landscape. Erosion has clearly exposed the relict terraces and mulga soils have partially covered the HCKS and HCKC sites. Stone artefacts was found cemented into the terrace sediments and in some instances, gravel is incorporated into the matrix. These three sites are discussed in more detail below starting from the north site and continuing to the centre and south sites. 92 | Case Studies: A multi-proxy approach (a) Terrain model: sites located within a small valley (b) Hardpan Creek and Basin Gully sample site locations Figure 4.9: Hardpan Creek and Basin study area and sites: a) terrain of the Hardpan Creek and Basin Bore study area, 142 m to 144 m ASL and, b) Hardpan Creek North, Centre and South, and Basin Gully 1 and 2 sites. Water flows from NW to SE 4.8 Hardpan Creek North Site | 93 4.8 Hardpan Creek North Site Hardpan Creek North site is a relict terrace located on the outer bend of a south-flowing channel. The arroyo measures 7.5 m long and 1.4 m deep and its stratigraphic units comprises disconformities and lenticular layering, which implies a vivid past of possibly colluvial and alluvial processes (Fig. 4.10). Figure 4.11 shows channel movement and "S–bend" around the relict terrace with sediments being deposited on the inside of the bend. Overbank flows have partially eroded the terrace and removed the upper mulga layer above the terrace. It is inferred that soil texture and compaction of the sediments on the terrace was responsible for the channel shape. These relict terrace layers was covered by H4 Mulga land system, which are shallow to moderately deep, stony, acid, red earths with silcrete boulders intermixed with and overlie the soil mass (QDPI 1974: VII-10). Intact terrace located on the far right of Figure 4.10 allows collection of sediment and OSL dating samples. Channel erosion has undercut a small section of this terrace and exposed the lateral features. Figure 4.10: Hardpan Creek North, east bank of main channel: terrace is undercut by stream flows and has exposed this relict terrace 94 | Case Studies: A multi-proxy approach Figure 4.11: Hardpan Creek North: water flow is unable to carve its way into the relict terrace (arrows show incision direction water flow around the terrace) Archaeological material: An inspection of this site in April 2018 found stone artefacts embedded in two visible sediment/soil discontinuities located in the centre and base of this profile respectively. The characteristics of the archaeological record include stone artefacts cemented into indurated sediments/soils along the margin of the Hardpan Creek channel. Artefacts was found in layer two at 45 cm deep, and layer five at 95 cm deep. At a depth of ~45 cm, stone artefacts was found embedded in sandy loam sediments (Fig. 4.12a). At 95 cm deep, a weathered stone artefact is found embedded between lateral deposits. This stone tool has a platform, bulb of percussion and scars on the dorsal and vental surfaces (Fig. 4.12b). The artefact is found embedded between two lateral layers or lenticular deposits (associated with wave action). It was prised out with screwdriver to free it from the two horizontal layers (Appendix B, Fig. B.14 (h)). Weathering of the ventral surface suggest that chemical reactions occurred between stone artefact and the surrounding sediments causing the silica to breakdown and flake from the artefact. Ground surveys downstream found large numbers of stone artefacts embedded in indurated features on both sides of this small channel. 4.9 Hardpan Creek South/Centre Study Site | 95 (a) Two artefacts, 45 cm (b) Weathered ventral surface (arrows), base Figure 4.12: Hardpan Creek North site: a) two stone artefacts found embedded 45 cm below surface; b) a weathered surface through oxidisation and continuous contact with Fe-oxides in soil 4.9 Hardpan Creek South/Centre Study Site Hardpan Creek South (HPCKS) and Centre (HPCKC) are located about 65 m (HPCKC) and 75 m (HPCKS) downstream of Hardpan Creek North site. These two small terraces about 8.0 m long and <0.5 m deep (Fig. 6.6). Dark alluvial sediments gave the impression it has formed by spring deposition or alluvium and imply organics and a freshwater environment. Figure 4.13: Hardpan Creek South site: mulga soils overlie this buried terrace (40 cm and 2 m scales) 96 | Case Studies: A multi-proxy approach Archaeological materials A survey of the terrace both up and downstream of these terraces found stone artefacts embedded within the indurated surfaces (Appendix B, Fig. B.14). Stone artefacts embedded in these features also suggest buried features align with the occupation of early peoples in this area. A yellow-brown stone artefact is found in the Hardpan Creek Central terrace about 15 cm below the surface (Fig. 4.14). This stone artefact has become incorporated into the sediments from either phreatic site formation or alluvial processes. Episodic and direct incorporation into the sediments follows site stability. More will be revealed about these processes in the ensuing chapters. Figure 4.14: Yellow-brown stone artefact (star) embedded in buried feature. An OSL sample (circle) was collected next to the artefact 4.10 Basin Gully Study Site The Basin Gully sites is located in a southeast-flowing erosion gully (Fig. 4.9b). Two sites was chosen at the upper end of Basin Gully, BG1 and BG2. Both sites are located in differing features, BG1 is a shallow arroyo near heavily eroded by sheet and gully erosion and, BG2 is a shallow arroyo on the northern side of the channel within red loam sediments. Active and extinct springs located southeast of these sites was not included in this study. The Basin Gully sites was chosen because of laminar sediment features that implies water-borne deposits and thus possible palaeo-fluvial aggradation conditions (Fig. 4.15). Stone artefacts was embedded within the base of a laminar feature and in a layer of the BG2 site and has been previously mobilised and buried during an unknown period of time (Appendix, Fig. 4.10 Basin Gully Study Site | 97 B.14). These features provided an opportunity to investigate site formation processes and palaeoenvironmental conditions, and the location of archaeological evidence within these features. It is hypothesised that ancient peoples has occupied the Basin Gully area during phases of palaeoenvironmental change. Basin Gully 1 Site The Basin Gully 1 (BG1) site is located southeast of the Hardpan Creek sites. This site is located in a small eroded gully southeast of the Hardpan Creek sites (Fig. 4.9b). Surface deflation and incision has exposed a former land surface and an arroyo. Its linear structure and alignment infers that this is a former track that has been deepened by gully erosion. Surface deflation and gullying dominates the site on its northern margin. Reduced ground cover encourages sheet erosion and constant use of tracks by vehicles creates two depressions and over time and runoff water deepens these tracks. Double and parallel wheel ruts soon become deep gullies and are impassable by vehicles. In turn, old vehicle tracks are transformed to small erosion gullies. High intensity rainfall and significant sheet wash and gullying deepens and widens these former tracks. This scenario could explain the location and depth of Basin Gully. A significant feature of this site is its linear depositional features (Fig. 4.15). These suggest a previous shoreline with continuous lapping of the water’s edge and deposition of sediments. The depth of this site implies extensive site formation processes and unknown timeframes to identify the timing of past depositional events. The focus of this study will be to examine the maximum age of the site and its palaeoenvironment. Archaeological materials Off-white/cream crypto-crystalline stone artefacts was found embedded in the indurated sediments of BG1 (Fig. 4.16). A density of about 15 per sq. m exists across the site. Some artefacts are cemented into similar sediment features and others occur to the north of BG1 site along its channel margin. 98 | Case Studies: A multi-proxy approach Figure 4.15: Basin Gully 1 site: an exposed terrace <80 cm high shaped by gully erosion. Lateral deposits show evidence of wave actions along a shoreline. CaCO3 deposits fill the cracks and voids of the deposit Basin Gully 2 Site The Basin Gully 2 (BG2) is located about 20 m downstream of the BG1 site on the east bank of the gully (Fig. 4.17). A capping of sediment/soil <25 cm deep sits above the layer with stone artefacts. Gully incision suggest that this site is a former station track but severe overland flow removed the top soil and deepened the track. Vegetation includes low shrubs (Eremophila spp.) and trees (Acacia aneura and Eucalytus populnea). The ground cover is sparse and generally very low biomass. 4.10 Basin Gully Study Site | 99 Figure 4.16: Basin Gully 1 site: stone artefact embedded in the base of deposit Archaeological materials Stone artefacts at the BG2 site was found embedded between the sediment/soil capping of the underlying sediment layer about 40 cm below the surface (Fig. 4.17). On the surface, stone artefacts was found along the upper surface of the gully in densities of <5 per square metre. Stone artefacts densities increase toward the BG1 site and closer to the extinct springs and Hardpan Creek sites. No large stone artefacts such as grindstones or hammerstones, or cores was found in proximity to this site. 100 | Case Studies: A multi-proxy approach Figure 4.17: Basin Gully 2 site: an exposed terrace <65 cm high with stone artefact (Insert) and OSL sample location (circle) 4.11 Case Study 3: Double Well and Tunkana Well Double Well (DW) and Tunkana Well (TW) study area is located in the upper reaches of Boorara Creek and within the north-western margin of Hoods Range (Fig. 4.1). These sites area located in a small valley (Fig. 4.18a). An un-named channel gently flows west from Hoods Range and splits into two minor channels above each study site (Fig. 4.18b, identified by a yellow dashed line). Downstream, these channels flow into Boorara Creek and through the Walters/Hoods Ranges valley. Boorara Creek drains southward into Lake Numalla, a large freshwater lake (NPRSR 2014). The surfaces was void of grasses and ground cover is extremely low. The mulga Acacia aneura land H4 system encroaches near the main channel and along its east bank. Large river red gums line the channels and sometimes show the margin of the previous channel. The terrain varies from gentle rises and slopes that are dissected by third order streams (Fig. 4.18a). 4.11 Case Study 3: Double Well and Tunkana Well | 101 (a) Terrain Model: study sites located in small valley (b) Study area: location of DW and TW sites Figure 4.18: Double and Tunkana Well study area, a) terrain model: study area (dashed box) 162 m – 166 m ASL and, b) study sites: dashed line with arrows identify channel flow directions 102 | Case Studies: A multi-proxy approach The Double and Tunkana Well’s sites comprise <2 m high arroyos that are located within close proximity to active and inactive springs. Discontinuities at the Double Well site has stone artefacts deposited within a central stone line. The Tunkana Well site has no visible stone artefacts but a sequence of fresh spring sediments and several discontinuities that suggest significant palaeoenvironmental change over time—evidence of humid and arid phases, for example, possibly the LGM. It is hypothesised that this site could provide additional information about palaeoenvironmental change within the study region. The Double and Tunkana Wells study area was chosen because of it proximity to artesian springs, one arroyo has stone artefacts embedded in a stone line 1 m below the current land surface and, linear stratigraphic features probably represented phases of palaeoenvironmental change. Discontinuities represented abrupt changes in climate and this implies a human presence is likely and this deduced links to human antiquity. With this evidence, and observations from this study, this arroyo and nearby gullies provided an ideal opportunity to examine climatic shifts and an ecological baseline for this area. Archaeological material and stratigraphic change is likely to position this human landscape with palaeoenvironmental change across eastern Australia. 4.12 Double Well Study Site The Double Well site is located on the northern face of a west-flowing channel that is lined with large river red gums (Eucalyptus camaldulensis) (Fig. 4.18). The landscape features includes an alluvial plain, springs near the channel and, minor slopes. The surface of the arroyo is without ground cover and most likely eroded by overgrazing (cf. QDPI 1974). Channel incision cut 1.7 m into the surrounding floodplain and exposes four discontinuities (Fig. 4.19). The channel base near the DW study site is covered by large to medium-sized rounded quartz cobbles (Fig. 4.20). A survey of the channel upstream noted conglomerate features located about 400 m from the site. At first glance and walking upstream from the arroyo, I identify channel flows and erosion cutting into the conglomerate rocks upstream has transported cobbles and gravel downstream (Fig. 4.21). A stone line is noted in Layer 3, which extends along the length of the exposure (Fig. 4.22). Gravel and cobbles and, stone artefacts found within this layer has implications for site interpretation. It implies unique climatic or storm event and, depositional site formation processes. 4.12 Double Well Study Site | 103 Figure 4.19: Double Well site profile: 4 major disconformities/depositional layers (scale = 25 cm increments) separated by colour and texture Figure 4.20: Double Well site: gravel base of channel 104 | Case Studies: A multi-proxy approach Figure 4.21: Double Well site: conglomerate feature upstream Figure 4.22: Double Well site: stone line, which dissects the arroyo has gravel and stone artefacts within the matrix 4.12 Double Well Study Site | 105 Archaeological material Stone artefacts was embedded in a 12 cm wide stone line ~75 cm below the surface (Fig. 4.23). Medium-size stone artefacts was found embedded below and within this stone line (Fig. 4.24). A survey both upstream and downstream of the site found scatters of 15–20 stone artefacts dispersed along the minor slopes. Flakes and flaked pieces made from a crypto-crystalline material dominated the assemblage. Similar patterns occur long the slopes of the H4 land systems on the southern side of the Double Well Creek. Several heat retainers in poor condition was found on the upper banks of the channel. Figure 4.23: Double Well site: stone artefact partially exposed from arroyo (a) Medium sized stone artefact (b) Side view Figure 4.24: Double Well site: example of a medium-sized stone artefact located 0.75 m below the surface 106 | Case Studies: A multi-proxy approach 4.13 Tunkana Well Study Site Tunkana Well arroyo is located about 600 m north of the Double Well site on the west bank of the channel (Fig. 4.18b). The main channel banks is lined with large river red gums trees. A concave-shaped spring with a central pool is located about 12 m from the arroyo edge (Fig. 4.25b). An unusual stratified deposits indicates previous erosion and transport of material downstream (Fig. 4.25a). No stone artefacts was found in this arroyo. (a) TW Profile (scale = 25 cm increments) (b) Spring above arroyo Figure 4.25: Tunkana Well profile and spring (concave) 4.13 Tunkana Well Study Site | 107 The arroyo is about 1.75 m deep and has five visible layers. The upper layer shows the lighter-coloured sediment above a reddish brown layer stained with Fe-oxides. Visible disconformities includes what apears to be colluvium and alluvium (Fig. 4.25a). The majority of these deposits suggest episodic and short sharp events. Characteristics of the archaeological record The TW arroyo is without any archaeological material. Stone artefacts was found on the surface of the spring. A survey of the channel edges and onto the floodplain found surface stone artefacts scatters upstream and downstream of the arroyo. About 10 stone artefacts per square metre is estimated for areas up and downstream of the arroyo. These patterns correlate with greater numbers of stone artefacts found near water sources (Robins 1997). In addition, an edge-ground basalt axe found 400 m downstream next to small gravel heat retainer identifies the diversity in stone tools and previous trading of hafted stone axes (Fig. 4.26). The Selwyn Ranges, northwest Queensland greenstone axes have been found as far south as the Murray Darling and Lake Eyre basins (McBryde 1987; Roth 1897; Smith 2013; Tibbett 2002). The axe found here is located ~860 Km from its original quarry. Negative flaking scars centrally located on the axe implies hafting of a wooden handle midway between the cutting edge and its upper edge. Figure 4.26: Edge-ground basalt axe found 400 m downstream from study site 108 | Case Studies: A multi-proxy approach 4.14 Research design and preliminary study methods The research design focused on palaeoenvironmental proxies for example, micromorphological pedofeatures in soils, chemical and physical sediment characteristics and, optically stimulated luminescence dating (OSL) to reconstruct site formation processes and the past. The research objective is to dissect arroyo deposits into manageable datasets that could imply broad scale living spaces and palaeoenvironmental conditions (French 2015). This approach includes exploring explanations for episodic variations in the appearance of archaeological materials and, how each site and feature formed or has been deposited in each site environment. These offered ideal situations to identify how ancient peoples coped with episodic environmental change and, whether humid or arid phases has dominated each event and in what time frame. The methods identified for this study aimed to address these strategies and explain them in detail. The following discussion explains the field methods, laboratory procedures and, soil micromorphology used in this study. The research design identified four research priorities. The first, to locate suitable landscapes with arroyos near springs and archaeological sites. Second, collect soil samples that could determine the types of landscape processes that was operating in this region. Third, create micromorphological thin sections, date the sediment profiles from selected OSL samples, and examine the sediments in detail to understand their texture and chemical and physical properties. Fourth, examine the soil properties to reconstruct the palaeoenvironmental conditions and provide a chronology for the arroyo sequences, and link these to the archaeological record, environmental frameworks, and examine the impacts of changing climatic conditions along Eulo Ridge. The following outlines the approach used to address these priorities. This study adopts the Australian Archaeological Association Code of Ethics and these guided the research principles (https://australianarchaeologicalassociation.com.au/about/) : 3.1 Members acknowledge the importance of cultural heritage to Indigenous communities. 3.2 Members acknowledge the special importance to Indigenous peoples of ancestral remains and objects and sites associated with such remains. Members will treat such remains with respect. 3.3 Members acknowledge Indigenous approaches to the interpretation of cultural heritage and to its conservation. 3.4 Members will negotiate equitable agreements between archaeologists and the Indigenous communities whose cultural heritage is being investigated. AAA endorses and directs members to the current guidelines regarding such agreements published by the Australian Institute of Aboriginal and Torres Strait Islander Studies. 4.15 Field Methods | 109 4.15 Field Methods Local knowledge provided invaluable information about road conditions and access to proposed sampling areas. Site access is dependent on whether sites could be driven to by four-wheel drive vehicle. Field methods involved describing each profile and its stratigraphic features, bulk sediments, Munsell® soil colour (Munsell 1992) and soil pH was collected for analysis. Optically stimulated luminescence (OSL) dates was collected in seven stratum across the three main study areas. Profile drawings and descriptions documented disconformities and features. Bulk sediment/soil samples aims to address questions about the deposits such as the types of geomorphic processes, and previous environmental controls (humid or arid conditions). OSL dates aims to present a chronology for the study areas. The expected broader data outputs includes explanation for site formation processes, truncation and environmental change over the millennia, and a chronology for these spring environments. A Garmin GPS60 Global Positioning System (GPS: using Universal Transverse Mercator, Zone 55, WGS84 as a Datum ) is used to collect locations for each site and points of interest. A personal drone captured aerial views of profile features and site surrounds at various heights: DJI Spark Drone camera resolution – 1080p video and 3968 x 2976, aperture 2.757, exposure time 1/2,000; focal length 4.49 mm; ISO 100; F number f/2.6) (https://www.dji.com). Micromorphology and Sediment Sampling Micromorphology and sediment samples were collected to reconstruct palaeoenvironments and site formation processes across the study areas. The priority is to take samples next to stone artefacts; above and below a sediment/soil colour or textural change and within a homogenous stratum with no visible change in colour or texture. The objective is to sample features that has the potential to reveal information about palaeoenvironmental change and the associated processes, including whether a previous landscape is stable or unstable. Forty-two intact block sediment/soil samples were collected as representative samples of major stratigraphic units from each arroyo profile. Each block measured ⇠7 cm x ⇠5 cm x ⇠5 cm but varied in size because of soil texture and its location. Slight adjustments has to be made to cope with friable soils—more glue and patience is needed to extract samples from crumbly or very soft sands. In addition, about 200 grams of sediment/soil was collected next to the bulk sample for soil analysis, for example, particle size distribution, bulk density, magnetic susceptibility. Munsell colour, soil pH, and sediment/profile descriptions was completed during these sample collections. Detailed methods are discussed in Appendix A. 110 | Case Studies: A multi-proxy approach Optically Stimulated Luminescence Dating The objective of developing a chronology for the study area centres around previous archaeological and palaeoenvironmental data, and known site formation processes (Chapter 2) (Fitzsimmons et al. 2013; Rhodes et al. 2009; Robins 1998). Dr Richard Robins, Everick Heritage Consultants, Brisbane, Queensland funded six OSL dates. OSL samples was collected in March 2018 from five arroyos. The Victoria University of Wellington, New Zealand (https://www.wgtn.ac.nz/sgees/research/facilities/luminescence-dating-facility) was chosen to process the samples. This dating laboratory uses the single-aliquot regenerative-dose method (SAR) (Murray and Wintle 2000), which requires a quartz-dominated sample. Dating reliability and OSL age determination depend on the luminescence characteristics of the quartz and other minerals present (Rhodes et al. 2009: 194). OSL dating makes a small but significant contribution to this study. But before discussing this in more detail, it is important to point out a question that is often asked concerning its accuracy and collection methods. OSL dating is considered a reliable dating method and ideal for situations that require minimal disturbance of the sediment/soil features (Rhodes et al. 2009). Nelson et al. (2015: 166) points out that luminescence dating has become the cornerstone of archaeological, Quaternary geology, and palaeoenvironmental reconstruction. Geoarchaeologists utilise luminescence dating to secure temporal ranges for stratified deposits (Roberts et al. 2015). Nelson et al. (2015: 166) note, "luminescence dating provides a direct age estimate of the time of last exposure of quartz or feldspar minerals to light or heat and has been successfully applied to deposits, rock surfaces, and fired materials in a number of archaeological and geological settings". OSL dating has been used since the mid-1980s to date sediments, for example, directly date desert deposits, in particular, quartz grains (Krapf et al. 2018). The age range for luminescence dating is between ca. 100 and 200,000 years. This date range is dependent on a maximum attainable signal of the target minerals (saturation level), and the dose-rate of the environment. Higher dose-rate environments limit the upper age range for older samples (>10,000 years) but allow greater signal resolution for younger samples (<1,000 years). In terms of accuracy, Fitzsimmons et al. (2013: 82) identified that there is a 5 % – 10 % uncertainty within OSL dating and argue that it is unlikely ever to reach the precision of radiocarbon dating, which reduces the resolution of palaeoenvironmental reconstruction. Dose rate heterogeneity is also an issue for age estimate precision (Roberts et al. 2015). Nevertheless, the optically stimulated luminescence SAR method offers an effective approach for the development of a regional chronology. This method is therfore suitable for this study. Each OSL sample is collected from a layer of homogenous sediment/soil within a 20 cm to 30 cm radius of discontinuity (Nelson et al. 2015). This strategy aimed to avoid inaccuracies in 4.15 Field Methods | 111 gamma dose rates and obtain laboratory-based dosimetry measurements. In addition, because a portable gamma spectrometer was unavailable, about 100 grams of sediment was collected beside the OSL sample. OSL dating materials included the following, a 40 mm diameter galvanised water pipe 300 mm long with end caps, a thread cap to protect the end of the pipe while hammering it into the stratigraphy, spray foam—to fill the voids at either end of the pipe, a tape to completely seal each cap, and labelling to identify collection site, date, direction of the sample, and recorder. Archaeological materials: identification and assessment Archaeological material plays a small but significant role in this human occupation model. Stone artefacts recovered by Robins (1998) from the Youlain Springs site placed Aboriginal occupation to about 13 Ka BP. This is the oldest dated evidence in this region. At Peery Lake near Wilcannia, New South Wales, located about 280 Km south of this study area, Rhodes et al. (2009: 191) found stone artefacts within gravel units. They report that stone artefacts are rarely found in sand or silt-dominated contexts in these regions. The archaeology in this region comprise almost entirely palimpsests of stone artefact scatters and heat retainers more commonly found close to water sources, for example springs (Holdaway et al. 2017; Holdaway and Fanning 2014; Robins 1996; 1997). These are usually exposed on bare deflated surfaces and concentrated in clusters and/or as lag deposits. Formal tool types such as backed blades, tulas, pirri-points, and thumbnail scrapers dominate the assemblage. These tools represent a very small percentage of the archaeology. Spatial and temporal resolutions of these deposits are very poor. Holdaway and Fanning (2014) identified that stone artefacts <30 mm long, are easily mobilised by gravity, water, and trampling by people and/or animals. Stone artefacts >30 mm however tend to remain in situ whereas stone artefacts <3 centimetres long move large distances (Holdaway and Fanning 2014; Robins 1997). This has major implications for site interpretation. A general consensus in Australian archaeology is that palimpsests of Late to Middle Holocene age stone artefact lag deposits have little spatial and temporal resolution. Intense studies of these sites have attempted to understand their contexts in association with heat retainers and from this, infer relevant behaviours (Fanning et al. 2009; Holdaway et al. 2017; Holdaway and Fanning 2014; Holdaway et al. 2008; 2010; Robins 1996; 1997). In this study, it is important to positively identify stone artefacts. To do this, standard archaeological stone artefact diagnostic features was used to discriminate stone tools from natural stone (Hiscock and Clarkson 2000). The diagnostics used were: a platform, bulb of percussion, positive and negative scars on the dorsal and ventral surfaces, and possible retouch along the tool edges. In some cases, stone artefacts was removed from the deposit to check their diagnostic features of being manufactured by people. But minimal disturbance is 112 | Case Studies: A multi-proxy approach maintained to retain site integrity for the future research prospects. No records or inferences are made about stone tool making techniques. This is outside the scope of this study. Heat retainers was identified by clusters of gravel/rocks or baked clay >1 meter in diameter (Holdaway et al. 2017; Rhodes et al. 2009; Robins 1996). Charcoal pieces or charring was not expected because of poor preservation conditions across these deflated surfaces. But I expected to find evidence of firing and heating on the surfaces such as rocks and baked clay (Holdaway et al. 2017; Robins 1996). A further problem has been the removal of grindstones and cylcons, and the placement of these in personal and museum collections. It is well known that an absence of these stone artefacts from the archaeological record prevents interpretation of the spatial and behavioural studies across these regions. Ultimately, post depositional disturbance of sediments reduces the probability of interpreting human landscapes. 4.16 Laboratory methods and procedures This section presents the laboratory methods and procedures used to analyse 56 soil samples. It includes particle size analysis, loss on ignition, magnetic susceptibility, and micromorphology. These variables aim to question how sediment/soil evidence relates to site formation processes and palaeoenvironments. Disconformities and changes in particle size distribution or geochemical data are likely to imply environmental changes. Soil samples were air-dried and ground lightly in a mortar and pestle to disperse larger clods, organic matter, and gravel components. Sediments were passed through 2 mm and 1 mm Endicott sieves to remove the coarse fraction and large pieces of organic matter. Geochemical analyses includes; particle size analysis, soil pH, magnetic susceptibility, loss of ignition (cf. French (2015: 91-94) and Goldberg and Macphail (2006: 391-395). A Malvern Mastersizer 2000 is used to obtain the particle size distribution for samples within the 0.02µm to 2mm size range. The details of this process are presented in Appendix A. These results and stratigraphic profiles will be presented in the next chapter to evaluate process-driven variations down the profile. Sand, silt and clay proportions identify discontinuities and imply probable changes in environmental conditions. Soil pH ranges from 3 to 10, with 6–7 being neutral, acid <6 and, alkaline >7. Ultra-acidic soils are < 3.5 and very strongly alkaline soils are >9. Soil pH is of little value because the soils in this region are very strongly alkaline, normally >9 (Dawson and Ahern 1974). Little emphasis will be placed on soil pH unless there is an unusually low pH reading. For an explanation of the pH reading methods, see Appendix A. 4.17 Soil micromorphology methods and techniques | 113 Magnetic Susceptibility Magnetic susceptibility is a measure of how ‘magnetisable’ a material is (Dearing 1999: 5), and is used as an indicator of surface weathering/exposure, and natural and human burning activities (Dearing 1999; Dearing et al. 1996; Derbyshire et al. 1995). The production of large amounts of silt is considered an arid and cold environment, and the opposite is true for warmer and more humid environments (Liu and Yuan 1982). Measurements detect the presence of ultra-fine (<0.03 µm) superparamagnetic ferrimagnetic minerals occurring as crystals produced largely by biochemical processes in soil (Dearing 1999: 17). Soil samples was collected and magnetic susceptibility results was generated in the laboratory using the methods set out in Appendix A. The Bartington MS2 Susceptibility System meter expresses magnetic susceptibility in SI (standard international) units. Frequency dependent susceptibility measurements involve taking two K readings in magnetic fields created at low frequency (470 Hz) (KLF) and high frequency (4700 Hz) (KHF). Measurement of a strong frequency-independent paramagnetic salt at both frequencies allows inter-calibration of the circuits to within ±0.1 percent (see equation below) (Dearing et al. 1996: 229). The low frequencies was measured first and high frequencies second (see Appendix A). Each low value is subtracted from the high value to obtain a frequency dependent susceptibility measurement, and this is expressed in XFD % (see below). XFD % = 100(KLF – KHF)/KLF (Dearing et al. 1996: 229) Loss on Ignition (LOI) Loss on ignition samples was processed in the Department of Geography Laboratory, University of Cambridge. The procedure is to subject a small amount of sediment to temperatures of 105° C, 400° C, 480° C, 550° C and 950° C for a minimum of eight hours. The ouputs are percentages of water/moisture content, charcoal and organic matter, calcium carbonate and residual silicate amount in sediments (Robertson 2011). These procedures are explained in more detail in Appendix A. The importance of LOI is to identify the organic components as an indicator of humid conditions and development of humid topsoils, and pedogenesis more generally(Jackson 1958). The output data could reveal water flows=high percentages of calcium carbonate, organic matter=vegetated environments, and charcoal=fires. Bulk density is collected as part of the LOI process to examine soil porosity. 4.17 Soil micromorphology methods and techniques The overall research design aims to understand the lithologies and micro features found within the study sites. Courty (1992: 39) notes that micromorphology "consists of the 114 | Case Studies: A multi-proxy approach integrated use of various microscopic techniques for studying the arrangement and the nature of components that form sediments and soils". In archaeology, these techniques are used to identify human-related activities. The composition of constituents and their spatial relationship to each other reveals more about their genetic and chronological relationships (Stoops and Nicosia 2017: 1). Micromorphology is used for the recognition of mineral composition including sediment texture, textural pedofeatures (dusty clay void coatings and silty clay infillings), coprolites, bones, and rock fragments (cf. Goldberg (1980: 161) and Stoops et al. (2018: 781). These micromorphological investigations aim to examine buried soils and evidence of humid and arid phases to develop a model of human occupation model. The objective is to find the origin and source of sediments or materials and influences of biological agents and if possible, any associated human activities (Goldberg 1980: 161). The techniques used to observe the sediment/soil fabric and pedofeatures involves a petrographic microscope to view thin sections under plain polarised (PPL) and cross-polarised light (CPL). Under these PPL/CPL, it is possible to see depositional processes such as bedding which imply water-related features and clumps which could mean biological or human related activities (Goldberg 1980: 161). The secondary effects include cementation, compound grains, and carbonate nodules. The primary features are the rocks, and secondary features are the pedofeatures, for example, iron- or manganese-oxide nodules or calcium carbonates. These aid in the interpretation of stratigraphic discontinuities and attempts to recognise a sequence of events. Other considerations will be climate and local factors. All of these lead to an approach that reconstructs the sedimentary history and the entire deposition including the geogenic, biogenic, or anthropogenic factors. The challenge will be to identify human evidence in these arroyos. For the most part, micromorphology reveals micro-scale evidence for large impacts to the landscape, for example, ploughing and heavy trampling, or the features of building foundation or flooring (Zaidner et al. 2016). Other evidence includes fire features, such as hearths or heat retainers and evidence such as micro-charcoal , ash or burned phytoliths (Mallol et al. 2007; Matthews 2016; Mentzer 2014; Parr 2006). It is unclear if these will be found in this instance as hunter-gatherer activities may easily disappear with environmental change over time. Thin sections was made in the Charles McBurney Laboratory, Department of Archaeology, University of Cambridge. Twenty-six thin sections was polished to approximately 30 µm (see Appendix A for methods). Undisturbed sediment/soil samples were impregnated with resin; dried for four weeks or more; a thin slice is cut from the resin block and a glass thin section made within two to three stages. First, the sample is mounted on a glass plate and ground to about 3 mm on a Brot Lapping Machine (http://www.brot.fr/en/index.php). Second, samples were glued to a glass plate and ground to a nominal thickness of ~30 µm. This thin section is checked under a microscope to ascertain its overall thickness, it is then placed back on the Brot and re-ground if needed. Finally, a protective cover-slip is glued over the thin section 4.17 Soil micromorphology methods and techniques | 115 to improve its visibility and analysis under the polarising microscopic. Each thin section is then viewed under a Leitz 12 Pol S polarizing microscope and micrographs were taken on a Leitz 12 PolS polarizing microscope using a Q-Capture imaging system in plane and crossed polarizing light and oblique incident light (OIL or RL) at x4 and x10 magnifications. Textural pedofeature records Stoops et al. (2018: 378) described textural pedofeatures as being "characterised by a difference in grain size with the adjacent groundmass and comprise coatings, hypocoatings, infillings and intercalations". These form as clay coatings in B horizons and are related to clay eluviation/illuviation processes. These are generally attributed to the vertical translocation of fine clay suspended in percolating soil water. Changes in colour of the clay coatings, for example, black layer can imply these are attributed to the movement of organic matter and deforestation. The destruction of the clay coatings are attributed to mechanical processes which lead to fragments and deformation of the structures. Channel microstructures are likely to mean bioturbation is responsible. A further type is colluvial transport which can induce the fragmentation of clay coatings or clayey aggregates in the soil. All of the above have implications for landscape reconstruction and palaeoenvironments in this study. Table 4.1 provide samples of the expected textural pedofeatures. For thin section descriptions, the goal is to describe the microfabric, heterogeneity of the matrix, structure and voids, coarse and fine minerals, clay components, organics, pedofeatures, and amorphous features (Goldberg and Macphail 2006: 336). At a microscopic scale, the presence of fragments of textural pedofeatures for clay coatings and crusts, furrigenous nodules with sharp boundaries, horizontal root fragments, and rounded aggregate composed material derived from other soil horizons, relict palaeosol, bounded and sub rounded rock fragments free of weathering (Table. 4.1) (Stoops 2003: 23). Other features include evidence of rainfall being higher during interglacial and interstadial periods than during glacial or stadial periods (Stoops 2003: 593). The objective is to identify processes occurring in the soils/sediments, which will relate to changes in site formation, and post-depositional distortions identified within the archaeological record (French 2015; Goldberg and Macphail 2006). These aimed to relate these results to sediments/soils, palaeoenvironments, and post-depositional distortions of the archaeological record. A standard soil texture diagram is used to classify sediments into soil classes. Clay analyses included; 1) percent clay and silt, and whether it is clean/pure; strongly or moderately birefringent, 2) strong lines of extinction, 3) voids coated with either clays or organics, 4) clays were linked or bridged or cemented and, 5) features were moderate (50 percent) to strongly stained with clays (Nicosia and Stoops 2017). Nicosia and Stoops (2017: 124) suggest dusty clay bedding indicates cycles of rain and, the accumulation of iron and 116 | Case Studies: A multi-proxy approach manganese oxides could mean that water has been deposited in an area for a sufficient amount of time causing chemical reactions within the soils. For the analysis, an objective is to identify textural pedofeatures and patterns that imply humid or arid environments and/or cold/frost conditions (Table 4.1) (Stoops et al. 2018). Standard micromorphology techniques was used to describe each thin section (Macphail et al. 2006). These results are presented in tables, graphics, and photo micrographs to visualise each significant feature. A table structure entails: sample name/number; a description of the major characteristics of the thin section; and a first level interpretation of the evidence contained in the slide descriptions (French 2015) (see Appendix A for more details). 4.17 Soil micromorphology methods and techniques | 117 Table 4.1: Diagnostic pedofeatures, images modified from Stoops et al. (2018) Structure Microgram Description Silt capping Frost activity on coarse mineral grain, coated with a thin coating rich in organic material (p. 594) Micro-laminated clay coatings palaeosoil consisting of a Bt horizon with Fe-rich microlaminated clay coatings in voids (PPL) (p. 870) Kaolinite or haematite clay coatings presence of several generations of clay coatings; allows correlation between the oldest member of the chronosequence and relict palaeosoil (p. 879) Argilloturbation: Mechanical fragmentation crescentic pure clay infill of a soil pore which is then fragmented; fragmentation of clay coating (mechanical) and disorganisation of clay coatings as a result of micritic carbonate precipitation (chemical); colluvial processes (CPL) (p. 390) Calcite hypocoating micritic or microsparitic carbonates (PPL); formed from soil solutions percolating along the pores or fissures and penetrating into soil matrix; considered rapid precipitation of calcium carbonate due to root metabolism (p. 217) Fe-oxide pedofeature iron-oxide nodule containing subangular quartz grains (PPL) (p. 678) Carbonate nodule geodic disorthic complex microsparitic carbonate nodule with Fe and Mn oxide impregnations (CPL) (p. 617) 118 | Case Studies: A multi-proxy approach Use of opal phytoliths for site interpretation As discussed in Chapter 2, Acacia aneura (mulga) woodlands dominate the vegetation assemblage across these landscapes. These species are highly adapted to low rainfall and high evaporation rates (Dawson 1974b). But it is unknown whether other plant families and genus existed in this region and if early peoples modified the landscape with fire? To address these questions, the focus is to identify phytoliths in thin sections and reconstruct past vegetation types (cf. Chen and Smith 2013; Devos and Vrydaghs 2011; Devos et al. 2009; Parr 2006; Piperno 1988). Table 4.2 presents examples of phytolith morphotypes that are expected to be identified in this study. Opal phytoliths are microscopic silica/opal fossils that are found in some plant tissues that exist after the decay of the plant (Piperno 1988: 1). These can be introduced into archaeological sites by water, as residues from animal faecal pellets, in situ weathering, wind, and water (Wallis 2001: 109). An International code for phytolith nomenclature is used to identify phytoliths in thin section (Neumann et al. 2018). An objective is to complement the soil properties and micromorphology results and reconstruct site formation processes. Phytoliths are sometimes preserved in the soils and thus provide evidence of the former vegetation types (Madella and Lancelotti 2012; Nawaz et al. 2019). In an archaeological context, phytoliths can help to reconstruct palaeoenvironments (Cabanes et al. 2012; Devos et al. 2009; Esteban et al. 2018; Nicosia and Stoops 2017; Shahack-Gross and Finkelstein 2008; Stoops et al. 2018; Wallis 2001). The method is to locate and photograph phytoliths and add them to a database for identification and analysis. To identify morphotypes, there is a reliance on Australia research such as the work of Doreen Bowdery (1998) at Puritjarra Rockshelter with Mike Smith (1989), and research by Lynley Wallis (2001) in tropical northwest Australia. But mostly phytolith morphotypes was identified by the descriptions and photographs contained in Neumann et al. (2018). Evidence of fluvial or wet surfaces could imply array of vegetation dominated by grasses. It is highlighted that phytoliths are used to identify the presence of vegetation and to infer the types of plants in each palaeoenvironment. Organic components are rarely preserved in arid environments. Highly alkaline sediments/soils are strongly associated with poor preservation and increase the chances of phytolith dissolution via a purely chemical mechanism (Shahack-Gross 2016: 11). For this reason, I did not expect to find preserved organic matter in any of these soils. The most likely scenario in these areas will be endemic plants highly adapted to local soils and the climate. 4.17 Soil micromorphology methods and techniques | 119 Table 4.2: Phytolith morphotypes expected to be found in this study (images modified from (Neumann et al. 2018) Family Morphotype Sample Description Poaceae Acute bulbosus 20 µm Size ~25-100 µm; characterized by their acute apex and more bulbous antapex, found in many grasses and some sedges; used as diagnostic of grasses Panicoideae and PACMAD clade Polylobate Size 20-40 µm; end-lobes separated by acastula; large family of grasses Cyperaceae and Poaceae Blocky 50 µm Size 40-150 µm; very common in the leaves of Cyperaceae and Poaceae; surface decorated with convex discs centrally located; low diagnostic value Cyperaceae and Poaceae Bulliformflabellate 50 µm Size 40-200 µm; heavily built, solid phytolith, tabular, epidermal cells, furrows of the leaf surface in Poaceae and Cyperaceae, usually related to high water availability Conifers, lycophytes, eudicotyledons, monocotyledons Elongate Size 20-700 µm; plant tissues and organs, such as epidermis, subepidermal tissue, lycophytes, conifers, monocotyledons, and eudicotyledons Poaceae (Grasses) Elongate sinuate 20 µm Size 50-200 µm; leaf epidermis in plants; diagnostic type in archaeological sites; Euphorbiaceae Poaceae (Grasses) Elongate dentate 20 µm Size 20-250 µm; long cells in the epidermis of Poaceae leaves Sponge spicule Non-phytolith 20 µm Positive evidence of saturated environments 120 | Case Studies: A multi-proxy approach Other possibilities might be finding evidence of human fires and fire management practices (Bird et al. 2012; Gamage 2011; Jones 1969; Kimber 1983). It is well known that human fires burn at 1,200 degrees Celsius and natural fires <600 degrees Celsius. Parr (2006: 183) found that under the oxidative conditions, phytoliths could vary in colour from black or brown. It is expected that burned phytoliths are likely to be found in these deposits because early peoples are likely to have lived near spring for a very long time. The problem that arises is that it may not be possible to determine if a burned phytolith was a result of a cooking fire or a natural bush fire. 4.18 Chapter Summary This chapter has presented each case study and why it was chosen. Seven study sites were selected within the 3 case studies in that had archaeological materials or potential to understand each unconformity. The methods and techniques above aimed to investigate these soils and through analysis of the data, reconstruct palaeoenvironments and the behaviours of early peoples. The outcomes are likely to have implications for investigating episodic change and peoples behaviours. The next chapter, sites within the case study areas will be broken down into tthe methods and techniques used above, and discussed according to the characteristics and patterns exposed through investigation. The use of soil micromorphology is new to this region and when combined with other methods and techniques, it provides a means to focus in on links between palaeoenvironments and the behaviours of early peoples. Chapter 5 The Geoarchaeology of Eulo Ridge Michael Connolly (Artist) ...environment governs human life: latitude, altitude, landforms, and climate determine vegetation, which in turn determines animal life. And all these things taken together determine how and where humans have lived (Renfrew and Bahn 2016: 233). 122 | The Geoarchaeology of Eulo Ridge 5.1 Introduction This chapter presents the results of a multi proxy approach that aimed to investigate arroyos located near springs along Eulo Ridge. These results are presented in 3 major sections and discuss the soil micromorphology and soil characteristics for example; magnetic susceptibility, loss on ignition, partical size distribution of the sediments, and optically stimulated luminescence data. A set of minor nested questions were asked in this chapter: what does the micromorphological evidence, soil characteristics and the presence of stone artefacts reveal about episodic change? What features correlate with the chronology and ultimately, how does it align with palaeoenvironments and the prehistory of early peoples? The discussion is guided by each of the 3 Case Studies, 1) Granites, 2) Hardpan Creek and Basin Gully and, 3) Double and Tunkana Wells (Chapter 4). 5.2 Case Study 1: Granites 5.3 Granites East Study Site Four micromorphology samples were collected from the Granites East study site; sample 1 (0–7 cm), sample 2 (30–39 cm), sample 3 (50–57 cm), and sample 4 (GE 040818) collected next to a stone artefact (Fig. 5.1). Sediment samples were collected beside each of these blocks. These are discussed in more detail below. Micromorphological characteristics Table 5.1 and 5.2 and Appendix C provide summaries of each micromorphological analysis. These are discussed in more detail below and where appropriate, the text will refer back to the main profile (Fig. 5.1). Sample 1 (0-7 cm) (Table 5.1) comprises a sandy clay loam. The groundmass is characterised by poorly sorted and sub-angular blocky structure, and channel microstructure. The pedofeatures consists of dusty clay coatings, sediment fines (parent material) filling the voids and root channels, linked/bridged coatings, moderate birefringence, and weak crescentic coatings. The amorphous features comprise Fe- and Mn-oxide nodules. Slt cappings of the quartz grains imply cold frosty conditions (Table 5.1) (cf. Van Vliet-Lanoe and Fox 2018: 591). Opal phytoliths included Blocky, Elongate, Acute bulbous (some burned), Elongate entire/sinuate morphotypes. 5.3 Granites East Study Site | 123 Figure 5.1: Granites East site: undisturbed sediment sample locations: Sample 1 (0–7 cm), Sample 2 (30–39 cm), Sample 3 (50–57 cm), and Sample 4 (GE 040818) collected next to a stone artefact 124 | The Geoarchaeology of Eulo Ridge Table 5.1: Granites East Sample 1 (0–7 cm) summary descriptions and first level interpretations Major Characteristics Interpretation Channel to vughy microstructure, compound packing voids; root voids; surface crusts; channel coatings and infillings of illuvial clay; interpedal clay coatings of voids; abundant dusty to pure clay coatings (<75%); linked/bridged coatings; sesquioxide coatings of sand grains; moderate birefringence; weak crescentic coatings; complete to <70% infills of voids and channels with parent materials; reticulate striation in groundmass; fine porphyric structure; moderate silt cappings of the coarse grains; bioturbation; <10% Fe- and Mn-oxide nodules <0.15 cm within the groundmass and voids; Blocky, Elongate, Acute bulbous (some burned), Elongate entire/sinuate Groundwater origin; repeated wet and dry phases; tree, shrub and grass environment (a) (b) (c) (d) Figure 5.2: Granite East Sample 1 (0–7 cm): a) to d) void infillings with parent materials and Fe- and Mn-oxide nodules (flatbed scanner, natural light) Sample 2 (30–39 cm) sediments comprise a sandy clay. The groundmass is characterised by poorly sorted and sub-angular blocky structure with an angular blocky microstructure. Pedofeatures included <75% dusty clay coatings, root voids, linked/bridged coatings, hypocoatings, weak to moderate birefringence, and partial depletion of the groundmass. Amorphous features included Fe- and Mn-oxide nodules. The crystalline features included CaCO3 in the voids (Fig. 5.3) (Table 5.2). Organics included plant tissue and micro charcoal. Elongate, Blocky, and unknown phytoliths. Inorganics were represented by diatoms (Appendix C, (Fig. C.4(g)) (cf. Stoops et al. 2018: 165–166). Freshwater environment with regular raised groundwater table (cf. Round et al. 1990). 5.3 Granites East Study Site | 125 Table 5.2: Granites East Sample 2 (30–39 cm) summary descriptions and first level interpretations Major Characteristics Interpretation Sub-angular blocky microstructure; bioturbation; root voids; compound packing voids; abundant dusty to pure clay coatings of <75% of sand grains; intrapedal clay coatings, linked/bridged coatings; hypocoatings; irregular zones of groundmass implies partial depletion; weak to moderate birefringence of the clays; weak crescentic coatings; no visible sesquioxides; Fe/Mn-oxide coatings of the larger sand grains; intrapedal diffuse Mn-hydroxide nodules <0.15 cm; irregular CaCO3, discontinuous in voids; inherited nodules 0.3 cm long and 0.2 cm wide; partial and filling of voids with CaCO3 nodules/aggregates; Elongate, Blocky and unknown morphotypes; diatoms Groundwater origins, freshwater environment; post-LGM soil stabilisation and illuviation; tree and grass environment; natural and/or human fires; remnant root systems (a) (b) (c) (d) Figure 5.3: Granites East Sample 2 (30–39 cm) fabric pedofeatures: a) and b) Fe- and Mn-oxide nodules (flatbed scanner, natural light); c) CaCO3 nodules/aggregates (flatbed scanner, natural light); d) Calcite coatings (CPL) (cf. Stoops et al. 2018: 217–218) Sample 3 (50–57 cm) sediments comprise a sandy clay loam. The groundmass is characterised by poorly sorted and sub-angular blocky structure with a sub-angular blocky microstructure. Pedofeatures are dominated by linked/bridged coatings of the coarse fraction, weak birefringence of the clays, weak crescentic coatings and moderate sesquioxides staining of the quartz grains. Amorphous features include Fe- and Mn-oxide nodules. The crystalline features included CaCO3 nodules/aggregates and infillings of the voids (Fig. 5.4). Opal phytoliths comprised Blocky and unknown morphotypes (Table 5.3) (Appendix C, Fig. C.5). 126 | The Geoarchaeology of Eulo Ridge Table 5.3: Granites East Sample 3 (50–57 cm) summary descriptions and first level interpretations Major Characteristics Interpretation Sub-angular blocky microstructure; dusty to pure clay coatings of <40% of sand grains; linked/bridged coatings; irregular zones of groundmass; linked/bridged coatings; weak birefringence of the clays; weak crescentic coatings; moderate sesquioxides staining of the quartz grains; Fe- to Mn-oxide cappings on the larger sand grains; infills of the channels complete to <50% discontinuous/irregular of the fabric and Fe/Mn-oxide; Fe-oxide nodules <0.7 cm; CaCO3 nodules/aggregates–<0.7 cm long and 0.4 cm wide; Blocky and unknown morphotypes Fluvial origin; limited groundwater fluctuations and translocation of the fines; strong evaporation/drying phase; instability of the coarse-fine fractions; Fe- and Mn-oxide and CaCO3 formation; strong evaporation and drying phases; Cyperaceae and Poaceae environment; implied LGM conditions (a) (b) (c) (d) Figure 5.4: Granite East Sample 3 (50–57 cm) fabric pedofeatures: a) and b) Fe- and Mn-oxide nodules (flatbed scanner, natural light); c) and d) CaCO3 nodules/aggregates (flatbed scanner, natural light) Sample 4 (GE040818) sediments comprised clay loam. The groundmass is characterised by poorly sorted and sub-angular blocky structure with a channel microstructure. Pedofeatures were dominated by <75% clay coatings, linked and bridged coatings, poor birefringence, surface crusts, and complete to partial infills of the voids with parent material (Fig. 5.5) (Table 5.4). Amorphous features were Fe- and Mn-oxide nodules and account for <20% of the pedofeatures. Opal phytoliths included Elongate, Elongate sinuate, Blocky morphotypes. Poaceae (grass) and trees dominated the local environment. Inorganics included diatoms and confirmed that this was a groundwater dominated environment (Appendix C, Fig. C.4). 5.3 Granites East Study Site | 127 Table 5.4: Granites East Sample 4 (040818) summary descriptions and first level interpretations Channel microstructure; accommodating planes; vughy; abundant dusty clay coatings of <75% of the sand grains; complete to <50% infills of <0.6 cm voids with parent material; continuous zones of groundmass; linked to bridged coatings; poor birefringence; moderate to strong sesquioxides stain on the fine fabric; dusty to pure clays, partial to continuous to complete; <50% reticulate striation of groundmass; surface crusts and nodules; a manganese nodule 0.6 cm and Fe–oxide nodule <0.1 cm; CaCO3 nodules/aggregates discontinuous in voids; Opal phytoliths: Elongate, Elongate sinuate, Blocky; Diatoms Fluvial origin; bioturbation and root penetration; pedogenesis; groundwater fluctuations and translocation of the fines; illuviation/clay coatings; reticulate striations and stability of soils; poor birefringence; CaCO3 formation; evaporation phases (a) (b) Figure 5.5: Granite East Sample 4 (040818) fabric pedofeatures: a) clay and Fe- and Mn-oxide coatings, and parent material infilling of voids and; b) coatings of large quartz grain and voids (flatbed scanner, natural light) Sediment/soil characteristics Eleven samples were analysed for soil pH, particle size distribution, loss on ignition and magnetic susceptibility. Soil diagrams for each layer are presented in Appendix B, Fig. B.1. All sediments/soils were extremely alkaline (>9). The site sequence was dominated by lenticular layering. Precipitated CaCO3 filled the fissures which suggests that this layer had dried out following a humid phase (cf. Goldberg and Macphail 2006: 175). Layer 1, the phreatic zone surface layer was largely without vegetation cover. It comprised a dark red sandy loam (2.5YR 3/6) and was dominated by silt, very fine, and fine sand, 27/27/25% respectively (Fig. 5.6). The bulk density measurement of 1.68 g/cc showed these sediments were less porous and more compacted than all other layers. In addition, loss on ignition 128 | The Geoarchaeology of Eulo Ridge residue of 4% suggested that water deposition had deposited other materials such as CaCO3 and organic matter. Layer 2 or the middle layer was a reddish brown (2.5YR 4/4) sandy loam. Silt, very fine sand, and fine sand proportions were 32/30/24% respectively. CaCO3 filled small fissures between the soil nodules. These sediments were compacted and crumbled easily when disturbed. (Fig. 5.6). Layer 3 (base layer) was a red (2.5YR 4/6) sandy loam and was dominated by silt, very fine sand, and fine sand, 28/21/24% respectively. Sediments were compacted and stable and associated with less humid conditions and thus intermittent rainfall. CaCO3 deposits within the matrix suggested these sediments had been saturated but a significant drop in groundwater or raised temperatures dehydrated these sediments. Figure 5.6: Granites East particle size analysis plot: 6 samples showing percent clay, silt, very fine sand, fine sand, and medium sand in association with Munsell colour and the arroyo profile. Horizontal colour bars represent particle size percentages that total 100% 5.3 Granites East Study Site | 129 Physical and chemical characteristics Layer 2, 10 cm to 27 cm, contained the majority of stone artefacts. Sediments surrounding the artefacts were porous (bulk density: 1.1 g/cc – 1.2 g/cc) and had higher LOI residues of 4.3% – 5.3% (Fig. 5.8). The bottom layer had medium porosity and the lowest LOI residues (2.6%). Porous sediments indicate mixing was prevalent in this layer. Stone artefacts would have moved while sediments were pliable and added to by the expulsion of sediment from the springs and/or from the surrounding slopes. Frequency-dependent susceptibility (XFD%) results ranged from >10 % to <13 % for all layers (Fig. 5.8). Both the upper and bottom layers had the lowest XFD% and sediments near stone artefacts had the highest measurement (Fig. 5.8). Figure 5.7: Granite East: loss on ignition and bulk density plot 130 | The Geoarchaeology of Eulo Ridge Figure 5.8: Granite East frequency-dependent susceptibility and particle size <63 µm plot Frequency-dependent susceptibility (XFD%) results showed high percentages of superparamagnetic (SP) grains next to stone artefacts. Correlations appear to exist between stone artefact locations and Frequency-dependent. It suggests intermittent raised water tables and spring activity and/or hillwash (cf. Morton 2016). The data suggests that natural or human fires around springs. Granites East Site Interpretation The Granites East site revealed phases of groundwater fluctuations which suggested a stable palaeoenvironment with a constant supply of water. These results suggested stable conditions dominated the environment and this was followed by drying-out of the landscape possibly due to lowering of the water table. Waterborne sediments and spring activity are more likely responsible for the movement of sediments and burial of stone artefacts on an angle (cf. Morton 2016). Vegetation cover possibly comprised Cyperaceae and Poaceae, and some unknown plant families. These were likely sedges, which are associated with tropical and 5.4 Granites Relict Dune Site | 131 temperate climates. The occurrence of micro-charcoal within this layer suggested that it was likely washed into the site from the nearby slopes (cf. Bird et al. 2012; Jones 1969). Layer 2 data suggested raising and lowering of the water table. Diatoms found in this layer support freshwater conditions and thus raised water tables (Appendix C, Fig. C.4(g)) (cf. Round et al. 1990; Stoops et al. 2018: 165–166). Stable water supplies transitioned into a destabilised groundwater system and subsequent drying-out of these sediments was evidenced by high proportions of Fe- and Mn-oxide nodules and formation of secondary CaCO3. Groundwater conditions encouraged vegetation growth and a presence of plant tissue and opal phytoliths such as Elongates and other unknowns morphotypes suggested plant species uniquely correlated with phreatic conditions. An accumulation of micro-charcoal within these sediments suggested either natural or human fires. It would be reasonable to imply that early peoples burned the landscape (cf. Bird et al. 2012; Jones 1969). No chronology was obtained for this layer and thus the timing of this event is not estimated or known. Layer 1 showed recent surface deflation and soil compaction (raised bulk density) which is likely associated with the introduction of sheep, cattle, horses, and goats during the pastoral period (cf. Fanning 1999; QDPI 1974). But below this surface, translocation of the fine fraction suggested sporadic groundwater levels and/or intermittent rainfall events. The interpretation suggests heightened boom periods which aligns with La Nina weather patterns. Local aquifers were recharged and springs activated which was followed by subsequent drying out of the profile to form secondary carbonates. 5.4 Granites Relict Dune Site Two micromorphology samples were collected from GD-X (Sample 1) and, Static (Sample 2) locations (Fig. 5.9). These are discussed in more detail below and, where appropriate, the text will refer back to the main profile. Appendix C provide summaries of each micromorphological analysis. Note, the bottom two layers were visibly different to the loose sediment of the upper dune. Sediments had been deposited under unusual circumstances and stone artefacts within this layer helped to identify a significant geomorphic or climatic event. The micromorphology analyses, sediment analysis, and OSL date are used to investigate and determine site formation processes. Sediment samples were collected next to the micromorphology samples. These are discussed in more detail below. 132 | The Geoarchaeology of Eulo Ridge Figure 5.9: Granites Dune: micromorphology and optically stimulated luminescence sample collected from the Static layer (25 cm scale increments) Micromorphological characteristics Two micromorphology samples were collected from the lower section of the profile. Sample one was named ’GD-X’ as it was collected across layers 2 and layer 3. Sample two was named the ’Static’ because of its lateral layering that suggested a period of stable conditions with continuous laying down of sediments (Fig. 5.9). Layer GD-X (Sample 1) comprised a sandy loam. The groundmass was characterised by a strongly sorted and sub-angular blocky to sub-rounded structure with channel to vughy microstructure. The pedofeatures were dominated by dusty to pure clays, infilling of the voids, depletion of the fabric, and vertical accommodating planes (Fig. 5.10) (Table 5.5). Several sedimentary micro-layers extend across the Static layer and suggest phases of illuviation (cf. Fedoroff et al. 2018). Sponge spicules and vivianite are associated with wet/saturated conditions and probably relate to raised water tables (Fig. 5.11). Organics 5.4 Granites Relict Dune Site | 133 included micro-charcoal and suggested natural or human fires (Fig. 5.10). Other organics included roots from grasses on the upper surface. An absence of phytoliths highlighted that sandy sediments are less likely to preserve organics or these were largely bare surfaces. Table 5.5: Granite Dune Sample 1 (GD-X) summary descriptions and first level interpretations Major Characteristics Interpretation Sandy loam, channel to vughy microstructure; sub-angular blocky to sub-rounded; weakly orientated; occasional clusters; strongly sorted; chitonic; dusty to pure clay coatings across <50% of sand grains; moderate silt coatings of large grains; depletion of fabric; linear and sand grain bridging; three horizontal micro-layers composed of parent material <0.5 cm wide; infillings of voids with parent material; moderate vertical accommodating channels in upper half of section; Fe- and Mn-oxide nodules <0.15 cm diameter; micro-charcoal, root tissue; sponge spicules Loess origin; micro-layers; saturation, translocation of the fines filling of the voids, intermittent water infiltration; illuviation/clay coatings; Fe- and Mn-oxide formation; evidence of frosts; post-LGM unstable dune (a) (b) Figure 5.10: Granites Dune (X-layer) fabric pedofeatures: a) soil fragment with dusty clays (PPL) (cf. Stoops et al. 2018: 27,389) and, b) micro-charcoal (PPL) (a) (b) (c) Figure 5.11: Granites Dune (X-layer): a) and b) vivianite (PPL); c) micro-layering 134 | The Geoarchaeology of Eulo Ridge The Static layer (Sample 2) comprised a sandy loam. The groundmass consists of well sorted structure with angular to sub-angular groundmass, and channel to vughy porosity with large vughs (voids). Pedofeatures included pure to dusty clays (<50%) bridge the sand grains and silt coats the larger sand grains. Partial infilling of the voids with clays occurs but mostly the clays are depleted from the matrix. The organic components included plant tissues (Fig. 5.12), and micro-charcoal. Phytloliths included blocky, spheroidal, dentate, and elongate forms (Table 5.6) (Appendix C, Fig. C.6). Table 5.6: Granites Dune Sample 2 (Static layer) summary descriptions and first level interpretations Major Characteristics Interpretation Sandy loam, channel to vughy microstructure; weakly oriented; well sorted; accommodating planes; 40% vughs <0.4 cm diameter; pure to dusty clays (<50%); moderate silt coatings of coarse fabric; minimal laminar micro-features; strong voids to weak accommodating channels; silt coatings of the large grains; depletion of fabric; linear and sand grain bridging of the larger sand grains; partial infilling of voids with medium-sized quartz grains; Fe/Mn-oxide nodules; three small Fe–oxide nodules;infill voids; deep-rooted plant tissues, micro-charcoal; Opal phytoliths: Blocky, Spheroidal, Dentate, Elongate entire Loess origin; root penetration; limited groundwater fluctuations and translocation of the fines; stability in the coarse-fine fractions; Fe- and Mn-oxide formation; natural or human fires; former low dune; post-LGM soil stabilisation (a) (b) Figure 5.12: Granites Dune (Static layer) fabric pedofeatures: a) Fe- and Mn-oxide and root tissue; b) clay coatings around void (flatbed scanner, natural light) 5.4 Granites Relict Dune Site | 135 Age of the deposit An optically stimulated luminescence sample extracted from the Static layer returned an age of 14.6 Ka BP ± 1.5 Ka BP (Wang 2018) (Table 5.7). This puts the maximum age of the relict dune toward the end of the Last Glacial Maximum. Stone artefacts found in this layer suggest that early peoples were living near springs during the LGM. Table 5.7: Optically stimulated luminescence dates for Granites Dune (GD) (Wang 2018; 2019) Code a-Value* De (Gy) DR* (Gy/ka) Age (ka) Site ID WLL1348 0.07 ± 0.03 16.19 ± 0.055 1.11 ± 0.11 14.6 ± 1.5 GD Static (*) estimated a-value; DR = dose rate; Sediment/soil characteristics Nine sediment samples were analysed for particle size analysis, loss on ignition and magnetic susceptibility (Fig. 5.13). All sediments/soils were extremely alkaline (>9). Layer 1 (0–5 cm) begins with organic material, such as plants, leaves, and root systems. The top sediment/soil layer was a reddish brown (2.5YR 4/4) loamy sand and was dominated by erosion, long periods without rainfall and destabilisation of the surface. Near the dune edge, gravity has caused loose sands to fall into the channel below and has created a sawtooth pattern along the upper surface. Small dune crests are held together by the root systems of dormant grass species, for example, Eragrostis spp. and Eriachne spp. and shrub species, for example, Dodonea spp. and Eromophilla spp.. Layer 2 comprised a 2.5YR 3/6 dark reddish grey loamy sand. The matrix included charcoal pieces and root systems. At the top of layer 2, sediments were 10R 3/6 dark red loamy sand. These graduated into dull reddish brown (2.5YR 4/4) loamy sand at the base of Layer 2 (Fig. 5.13). Layer 3 comprised a dark reddish grey (2.5YR 3/6). Sediment size varied from a coarse sand at the top and very fine sand at the base. The dune sediments graduate from a fine sandy loam to a loamy sand at the base of the profile (S9), 60–70% fine to medium sands (Fig. 5.13). Sediments were brittle and easily fractured into angular blocky peds when disturbed. Silt and clay components were likely held together by poorly crystalline silica (or carbonate) cement (cf. Schaetzl and Anderson 2005: 422). 136 | The Geoarchaeology of Eulo Ridge Figure 5.13: Granites Dune particle size analysis plot: 9 samples showing percent clay, silt, very fine sand, fine sand, and medium sand in association with Munsell colour and the arroyo profile. Horizontal colour bars represent particle size percentages that total 100%, Layer 1) soft sands with grasses and root systems; Layer 2) increases in medium and fine sands; Layer 3) An indurated static layer with visible stone artefacts Physical and chemical characteristics From 0-63 cm, bulk density was 1.4 g/cc to 1.65 g cu. cum, and loss on ignition 0.7 % to 1.8 %. Bulk density for the Static (100–107 cm ) layer ranged from 1.34 g/cc to 1.42 g/cc and loss on ignition 2.0% to 2.8%. Frequency-dependent susceptibility (XFD%) results ranged from >7 % to <11 % for all layers (Fig. 5.15). The 47 cm layer had the lowest with ~7.5 XFD% and the Static GD-X layers (75-83 cm) had the highest results with ~10.5 XFD% (Fig. 5.15). Granites Dune Site Interpretation The Granites Dune site analyses revealed phases of stability, instability, and dune building. The Static layer represents a stable and/humid phases from 13.1 Ka to 16.1 Ka (14.6 Ka BP ± 1.5 Ka BP) (Wang 2018) The Static layer is interpreted as a former low dune with grasses, shrubs and trees. Frequency-dependent susceptibility results showed two unique data clusters. People were living in close to springs and conditions were stable. Diatoms and vivianite support a freshwater environment with rain and ideal conditions for human occupation. The vegetation included Cucurbita plants Cucumis myriocarpus (prickly paddy melon) orMelothria maderaspatana growing along this low dune. Cucurbita are thought to be of medicinal value to early peoples (QDPI 1974: VIII-7). Micro-charcoal found in this layer implied either natural or human fires. A human presence was further represented by a core and 5.4 Granites Relict Dune Site | 137 flakes cemented into these indurated sediments (Fig. 4.8b). These are interpreted as being in situ and associated with people and the de-glacial phase. The Early Holocene and Late Pleistocene boundary suggest conditions were unstable. Intermittent humid phases mobilised sediments. Laminar deposits document these short and abrupt change. During the Late Holocene, a dune was formed along the western margin of the phreatic zone. It is implied that sediments were transported downslope from Hoods Range. The timing of this event is unknown but it is thought that this occurred around 5 Ka. These sands are stabilised by grasses, shrubs and trees but at the time of this study conditions were very dry and ground cover was very low. Charcoal pieces, organic matter, and stone artefacts were haphazardly dispersed throughout the matrix. Early peoples occupation of the dune is likely but this could also natural fires. More research regarding these findings are needed. Figure 5.14: Granites Dune loss on ignition and bulk density plot Loss on ignition results demonstrated that the base layers, for example, Static, and adjoining layer had the highest percentages of organic and carbonate residues (Fig. 5.14). The lowest percentage of superparamagnetic grains (SD) were found between 0–63 cm, for example, <9 XFD% and, highest SD were found >70 cm deep and had >10 XFD%. It is hypothesised 138 | The Geoarchaeology of Eulo Ridge Figure 5.15: Granites Dune frequency-dependent susceptibility and particle size <63 µm plot that finer-grained sediments (10–20 %) appear correlated with SD percentages >10 XFD% and a truncated landscape. The base of the dune remained much more stable than the upper dune sediments. Thus, the dune sediments are being leached during pluvial periods and finer sediments filtered to the base of the profile. This phreatic zone/relict dune boundary was irregular and appeared to be controlled by channel flow and runoff. Its formation though is hypothesised to have formed by land slide which caused saturated overland flow and re-deposition to the base of the valley (cf. French 2003: 185). Sediments were transported downslope from Hoods Range and onto the valley floor to create this low dune. A wetting of the slopes assisted in the transportation of sediments downslope to their current position. This is interpreted as a significant humid period. The evidence suggests a movement of material more than 500 m to a kilometre to partially bury the western section of the phreatic zone (Fig. 5.16). A small area of phreatic zone sediments remained exposed and was lined by Eucalyptus populnea trees on its eastern margin (Fig. 5.16). The relict dune post-dates the phreatic zone and suggests unstable conditions with a dune building phase. 5.4 Granites Relict Dune Site | 139 Figure 5.16: Possible colluvial event transported sediment downslope from Hoods Range to partially cover the phreatic zone. An unburied area (dashed rectangle) provides evidence of this event (Google Maps, accessed 30 November 2019) Sediment micro-laminations at the base of Layer 2 and above Layer 3 (Static) were interpreted as a transitional phase or multiple wetting and drying phases. Soil fragments and sponge spicules supported evidence for freshwater fluctuations with intermittent high groundwater levels. 140 | The Geoarchaeology of Eulo Ridge 5.5 Case Study 2: Hardpan Creek and Basin Gully 5.6 Hardpan Creek North Study Site Micromorphological characteristics Six micromorphology samples were collected from 5 terrace layers: 1) Sample 1 (Mulga layer is setback above the main terrace); 2) Sample 2 (2 – 9 cm); 3) Sample 3 (29 – 38 cm); 4) Sample 4 (45 cm); 5) Sample 5 (90 – 97 cm) and; 6) Sample 6 (97 – 106 cm) (Fig. 5.17). Sediment samples were collected next to the micromorphology samples. These are discussed in more detail below. Sample 1 (Mulga) (Table 5.8) comprised a sandy loam. The groundmass was characterised by well sorted sub-angular blocky structure with a channel to vughy microstructure dominated by surface crusts. Pedofeatures included dusty to clean clays, irregular clay zones, hypocoatings, low to no birefringence, and sesquioxide staining of the fine/coarse fabric. Voids were partially to completely infilled with 30% excremental fabric. Amorphous components included Fe- and Mn-oxide nodules (Fig. 5.18). The organics included micro-charcoal and wood fragments. Opal phytoliths included Elongate to Elongate entire (Poacae), Polylobate (grasses), Blocky and Globular (Shrubs and Trees) morphotypes (Appendix C, Fig. C.7). The results suggest either a buried A or B horizon (cf. Kuhn et al. 2018: 401). Table 5.8: Sample 1 (Mulga) summary descriptions and first level interpretations Major Characteristics Interpretation Channel to vughy microstructure; sub-rounded to rounded peds <0.4 cm at the base of the slide, irregular clay zones in the groundmass, discontinuous <30% and weak to absent clay-bridge within the voids; dusty to clean clays, discontinuous to incomplete, low to no birefringence; sesquioxide staining with streaks across the coarse fabric/sand grains; partial to complete infills of channels and voids; 30% excremental fabric; soil surface nodules; hypocoatings of voids. Parent material partially to completely infill voids. 30 % excremental fabric; Fe- and Mn-oxide formation; CaCO3 nodules/aggregates in voids; Micro-charcoal; wood fragments; Opal phytoliths: Elongate (burned), Elongate entire, Arcuate, Fusiform, Blocky, Polylobate, Spheroid Colluvial origin; disturbed soils; bioturbation; translocation of the fines filling of the voids, soil surface nodules from under tree canopy; frost heave; root systems; vegetated surfaces 5.6 Hardpan Creek North Study Site | 141 (a) Micromorphology sample 1 and optically stimulated luminescence sample (b) Micromorphology samples 2 to 6 and 2 optically stimulated luminescence dates Figure 5.17: Micromorphology and OSL dating sample locations, a) mulga layer is setback and above the main terrace, b) main terrace with 4 layers Sample 2 (2–9 cm) (Table 5.9) comprised a loamy sand. The groundmass characterised by a well sorted sub-angular blocky to sub-rounded structure with a channel microstructure. Pedofeatures were dominated by varves with graded couplets, irregular clay zones, clean 142 | The Geoarchaeology of Eulo Ridge (a) (b) (c) Figure 5.18: Hardpan Creek North Sample 1 (Mulga) fabric pedofeatures: a) Fe-oxide hypocoating; b) surface crust and; c) angular micro-charcoal fragment (flatbed scanner, natural light) to dusty clays, translocation of the fines and coarse fractions filling voids, medium to high birefringence, silt cappings. Amorphous components included Fe- and Mn-oxide nodules (Fig. 5.19). Organics included probable ash, crystallised organic matter, and micro-charcoal. Opal phytoliths included Blocky, Elongate entire, Elongate, and Acute bulbosus morphotypes (Appendix C, Fig. C.7). Table 5.9: Sample 2 (2-9 cm) summary descriptions and first level interpretations Major Characteristics Interpretation Channel to vughy microstructure; accommodating planes; vughy; surface crusts <3.9 cm long and 0.6 cm long; several sub-angular blocky peds <2.4 cm long and <1.8 cm wide; clay infills along horizontal planes; laminated colluvium <3.9 cm long and <0.05 cm wide; parent material nodules <0.15 cm; surface nodules washed in from surfaces under tree canopy;) silt cappings; irregular clay zones in the groundmass, discontinuous <30% and weak to no clay-bridge within the voids; medium to high birefringence of the fine fraction; sesquioxide staining of the fine fabric; clean to dusty clays, discontinuous with partial to complete void infills; strong crusting <1 cm long; Fe- and Mn-oxide formation; Crystalline: CaCO3; deep-rooted plant tissues crossing into soil layer; Opal phytoliths: Blocky, Elongate entire, Elongate, and Acute bulbosus Colluvium origins; pedogenesis; laminated colluvium; groundwater fluctuations and translocation of the fines; stable coarse-fine fractions; frosts; natural or human fires; illuviation 5.6 Hardpan Creek North Study Site | 143 (a) (b) Figure 5.19: Sample 2 (2–9 cm) fabric pedofeatures: a) and b) varves with graded couplets (flatbed scanner, natural light) (cf. van der Meer and Menzies 2011: 224–225) Sample 3 (29–38 cm) comprised sandy clay. The groundmass was characterised by well-sorted sub-angular blocky structure with a channel microstructure. Pedofeatures were dominated by large voids, <75% dusty clay coatings of the sand grains, linked and bridged coatings, moderate to strong birefringence and sesquioxides (Fig. 5.10) (Table 5.10). Parent material infilled large voids . The amorphous components included Fe- and Mn-oxide nodules. Crystalline component included crescentic CaCO3 infillings of voids (Fig. 5.20). Micro-charcoal comprised the entirety of the organic components. Opal phytoliths included Elongate dentate, Elongate entire, Blocky (some burned) and Spherical morphotypes (Appendix C, Fig. C.9). Table 5.10: Sample 3 (29–28 cm) summary descriptions and first level interpretations Major Characteristics Interpretation Channel microstructure; accommodating planes; vughy, large voids; <75% dusty clay coatings of the sand grains; regular and irregular zones of groundmass; linked and bridged coatings; moderate to strong birefringence; strong sesquioxides staining the fine fabric and infills of the channels and voids; partial reticulate striation to massive compound clay coatings of the coarse fabric; linear and rounded crusts <5.8 cm long and 0.2 cm wide; parent material infills of oblong voids <1.1 cm long and 0.4 cm wide; intrapedal diffuse Mn-hydroxide nodules <0.15 cm; bottom zone contains an abundance of CaCO3 nodules/aggregates that infill the channels; Elongate dentate, Elongate entire, Blocky (some burned), Spherical Alluvial origins; groundwater fluctuations and translocation of the fines and coarse fractions into voids; formation of surface crusts; bioturbation; both stability and instability in the coarse-fine fractions; evidence of fires with micro-charcoal, ash, and crystalised organic matter within soil fabric 144 | The Geoarchaeology of Eulo Ridge (a) (b) (a) Figure 5.20: HPCKN Sample 3 (29–38 cm) fabric pedofeatures: a) and b) void infillings and clay coatings (flatbed scanner, natural light), c) Ficus spp. phytolith (CPL, black/white) Sample 4 (45 cm) comprised a sandy clay. The groundmass was characterised by well sorted sub-angular blocky structure with a channel to lenticular microstructure. Pedofeatures were dominated by well-sorted dusty to pure clays, crescentic clay infills, and some clay coatings (Fig. 5.21) (Table 5.11). The amorphous component included <20 % Fe-oxide nodules. Crystalline components CaCO3 laminations and infills of voids. Micro-charcoal and crystallised plant tissues comprised the organic components. Opal phytoliths included a honey comb (Ficus spp.) and Elongate dentate morphotypes. Inorganics include diatoms and fungal micro-fossil (Appendix C, Fig. C.10). 5.6 Hardpan Creek North Study Site | 145 Table 5.11: Sample 4 (45 cm) summary descriptions and first level interpretations Major Characteristics Interpretation Channel to lenticular microstructure; accommodating planes; compound packing voids; vughy; stability in the coarse-fine fractions; dusty to pure clays; fragmented concentric void fills; abundant dusty clay coatings of <85% of sand grains—top of the profile; regular and irregular zones of groundmass; bridged coatings; moderate to strong birefringence; strong sesquioxide staining of the fine fabric; dusty to pure clay infills of channels and voids with fragmented concentric void fills; hints of reticulate striation; one large Fe/Mn-oxide nodule 0.5 cm long and 0.2 cm wide; other nodules >0.05 cm; strong laminated CaCO3 infills in the voids and channels; intrapedal CaCO3 nodules/aggregates; CaCO3 features <1.8 cm long and <0.4 cm wide; Micro-charcoal within soil fabric; crystalised plant tissues; Opal phytoliths: honey comb (Ficus spp.), Elongate dentate. Diatoms and plant spores, for example, Testate amoebae Alluvial origins; groundwater fluctuations and translocation of the fines; wetting and drying phases; strong CaCO3 layers imply long periods of wetting of the soil fabric; CaCO3 layers intermix with the fine fabric and implies wetting and drying; clay coating around CaCO3 nodules/aggregates implies illuviation post wetting period (a) (b) (c) (d) Figure 5.21: Sample 4 (45 cm) fabric pedofeatures: a) juxtaposed crescentic clay coatings with extinction bands (PPL) (cf. Kuhn et al. 2018: 379–380); b) degraded clay coatings (PPL); c) and d) CaCO3 nodule with clay coating (PPL and CPL) (cf. Kuhn et al. 2018: 389–390) Sample 5 (90–97 cm) comprised a sandy clay. The groundmass is characterised by a poorly sorted sub-angular blocky structure with a channel to vughy microstructure. Pedofeatures are dominated by dusty clays, moderate to strong sesquioxide staining of the fine fabric, and partial to complete infills, crescentic infills of the voids with clay and organics (Fig. 5.22) (Table 5.12). Amorphous features included Mn- and Fe-oxide nodules <1.2 cm long and 1.4 cm wide. Recrystallised carbonate and surface crust dominate the macro features. The crystalline features were characterised by CaCO3 recrystallised nodules/aggregates <0.4 cm long and <0.3 cm wide (Fig. 5.23). Opal phytoliths included Blocky and Elongate entire morphotypes (Appendix C, Fig. C.11). 146 | The Geoarchaeology of Eulo Ridge Table 5.12: Sample 5 (90–97 cm) summary descriptions and first level interpretations Major Characteristics Interpretation Channel to vughy microstructure; accommodating planes; vughs; dusty clay coatings of <50% of sand grains; irregular to complete groundmass; bridged coatings; moderate to strong birefringence on the coatings; moderate to strong sesquioxides staining on fine fabric and within channels; dusty to pure clay, coatings of voids and channels and peds, and sesquioxides; more recent organic coatings of the crescentic coatings; hypocoatings around Fe/Mn-oxide nodules; minimal CaCO3 nodules/aggregates; Mn-hydroxide nodules <1.2 cm high and 1.4 cm wide; CaCO3 nodules/aggregates <0.4 cm high and <0.3 cm wide; Opal phytoliths: Blocky (some broken), Elongate entire; plant spores Alluvial origins; stability; limited groundwater fluctuations and translocation of the fines; stability in the coarse-fine fractions (a) (b) (c) (d) Figure 5.22: Sample 5 fabric pedofeatures: a) to c) crescentic coatings and extinction bands (PPL) (cf. Stoops et al. 2018: 379–380); sequence of various illuvial events (cf. Fedoroff et al. 2018: 848); d) soil fragment (PPL) 5.6 Hardpan Creek North Study Site | 147 (a) (b) (c) (d) Figure 5.23: Sample 5 fabric pedofeatures: a) Mn-oxide nodule; b) Fe-oxide nodule; c) Carbonate nodule recrystallised; d) surface crust (flatbed scanner, natural light for all scans above) Sample 6 (97–106 cm) comprised a clay loam. The groundmass is characterised by a poorly sorted sub-angular blocky to sub-rounded structure with a vughy to channel microstructure and was dominated by coarse to medium sands and dusty clays that are fractured and dispersed through the groundmass. This also comprised moderate to strong birefringence and sesquioxide staining, and silt cappings on the coarse fractions. The amorphous component included <50% Fe- and Mn-oxide nodules (Fig. 5.24(d)) (Table 5.13). These carbonates have been re-crystalised (cf. Stoops et al. 2018: 234–236) . Opal phytoliths include Blocky and Elongate dentate morphotypes (Appendix C, Fig. C.11). Inorganics included fungi microfossils (Fig. 5.25). Table 5.13: Sample 6 (97–106 cm) summary descriptions and first level interpretations Major Characteristics Interpretation Channel microstructure; accommodating planes; vughy; silt cappings of coarse grains; abundant dusty clay coatings of <75% of sand grains, and moderate to strong birefringence, orange, brown to black; moderate to strong sesquioxide staining of the fine fabric; strong fracturing and dispersal of the clay coatings; partial to complete voids and channels filled with dusty clays; silt cappings on coarse fraction; <50% of Fe/Mn-oxide nodules; <40% Fe/Mn-oxide nodules throughout groundmass; CaCO3 nodules/aggregates <0.8 cm long and <0.6 cm wide; intrapedal diffuse Mn-hydroxide nodules <0.6 cm long and 0.4 cm wide; porous Mn-oxide nodule 0.3 cm; Phytoliths—Blocky, Elongate dentate and fungi microfossil Alluvial origins; cool environment; stability; limited groundwater fluctuations and translocation of the fines; stability in the coarse-fine fractions; frost heave; pre-LGM soils 148 | The Geoarchaeology of Eulo Ridge (a) (b) (c) (d) Figure 5.24: Sample 6 (97–106 cm) fabric pedofeatures: a) CaCO3 infills (flatbed scanner, natural light); b) Fe- and Mn-oxide nodule (flatbed scanner, natural light); c) clay coating around a possible rhizolith (PPL) (cf. Stoops et al. 2018: 226–227) d) recrystallised carbonate nodules (cf. Stoops et al. 2018: 234–235) (flatbed scanner, natural light) (a) (b) (c) Figure 5.25: Hardpan Creek North Sample 6: a) to c) fungi microfossils Age of the deposit Three optically stimulated luminescence (OSL) samples were taken in this study site. An OSL sample collected from the upper mulga soils dated to 0.7 ± 0.1 Ka BP (Wang 2018) or about AD 1300. This age implies recent deposition of mulga sediments/soils and thus burial of older land surfaces. A second OSL sample was taken 45 cm below the surface of terrace, which was 5.6 Hardpan Creek North Study Site | 149 35 cm below the base of the mulga soil boundary. This returned an age of 11.5 Ka BP ± 1.2 Ka BP (Wang 2018) and aligns with the deglacial period or Pleistocene-Holocene boundary. South-eastern Australia was transitioning to a humid phase 10,000 years ago (Fitzsimmons et al. 2013: 91). A third OSL sample was taken at the base of the profile and returned an age of 55.9 Ka BP ±5.9 Ka BP (Wang 2019), which was a humid phase across south-eastern Australia and in inland areas (Fitzsimmons et al. 2013; Hughes et al. 2017). Table 5.14: Optically stimulated luminescence dates for Hardpan Creek North (HCKN) (Wang 2018; 2019) Code a-Value* De (Gy) DR* (Gy/ka) Age (ka) Site ID WLL1354 0.06 ± 0.03 72.67 ± 1.60 1.30 ± 0.13 55.9 ± 5.9 HPCKN OSL3 WLL1346 0.07 ± 0.03 21.46 ± 1.26 1.86 ± 0.16 11.5 ± 1.2 HPCK OSL WLL1347 0.07 ± 0.03 1.02 ± 0.04 1.86 ± 0.16 0.7 ± 0.1 HPCK Mulga (*) estimated a-value; DR = dose rate; Sediment/soil characteristics Eight samples were collected and analysed for particle size analysis, loss on ignition and magnetic susceptibility tests to characterise this arroyo. The channel at this point cuts through four visible sediments/soil layers (Fig. 5.26). Two samples were collected next to stone artefacts, Layer 2 and 4. All sediments/soils were extremely alkaline (>9). Five major layers wholly comprised sandy loams with a loam sediment/soil at the base of the profile (Fig. 5.26). The upper soil layer that was the most recent feature relates to the present day landscape and M2 land system (Dawson 1974b). The mulga soils above the terrace have receded to expose the relict terrace. The top layer was 2.5YR dark reddish gray sandy loam. This layer sits above the relict feature and contains root systems, charcoal pieces and insect burrows. This layer was <50 cm deep and comprises red-brown sandy loams. Layer 2 was about 30 cm deep and contains root systems, organic matter, and charcoal. The soils were a dark reddish brown (2.5YR 3/4) sand loam. Soil laminations occur throughout this layer, which terminates about 75 cm from the base of the profile. The sediments comprise higher amounts of fine and very fine sand than silt, which in this layer was the lowest amount for this site (Fig. 5.26). Layer 3 was a dark red (2.5YR 3/6) sandy loam. Soil laminations occur throughout this layer, and terminate about 50 cm from the base of the profile. Equal amounts of silt, very fine sand and fine sand dominate this layer (Fig. 5.26). Laminations and silts and fine sand suggested a settling out the sediments in multiple phases. This was likely caused by intermittent changes in water levels and suggest groundwater fluctuations over time. Layer 4 was a yellowish red (5YR 4/6) sandy loam. This layer was dominated 150 | The Geoarchaeology of Eulo Ridge by higher proportions of coarse and fine sands, and silt reduces to those similarly found in Layer 2 (Fig. 5.26). Gravel dominates the sediment matrix and it is likely to be colluvium. Layer 5 was a laminated layer comprised reddish brown (2.5YR 4/4) loam. Silt dominates this layer (Fig. 5.26). The bottom layer was undercut by water erosion and was strongly laminated with <4 cm wide lenticular deposits. This formed a veneer at the base of the profile. This was interpreted as an alluvial event with intermittent deposition of sediments as water levels receded. Figure 5.26: Hardpan Creek North plot: 5 samples showing percent clay, silt, very fine sand, fine sand, and medium sand in association with Munsell colour and the arroyo profile. Horizontal colour bars represent particle size percentages that total 100%, 1) mulga soil, 2) upper alluvial deposit, 3) Late Pleistocene alluvial deposits, 4) Pleistocene sandy loam gravels and, 5) Pleistocene loam alluvial deposits Physical and chemical characteristics Sediment were extremely alkaline (>9) for all layers. An xy-plot of soil bulk density and loss on ignition residues show two outliers with the 45 cm (OSL) layer had 11.2% residue and 90–97 cm sample had 8.2% (Fig. 5.28). The remainder of the data points were clustered between 4.8% and 6.8% and the upper layers including the mulga layer, had <3.2% residues. A colluvial event had likely deposited organics and carbonates when raised and stabilised water tables had settled out these sediments. An OSL date of 11.5 Ka BP ± 1.2 Ka BP (Wang 2018) suggests landscape change toward the Late Pleistocene and Holocene boundary. 5.6 Hardpan Creek North Study Site | 151 Frequency-dependent susceptibility (XFD%) results ranged from >8 % to <13 % for all layers (Fig. 5.27). The 29-38 cm and 90-97 cm layers had the lowest with ~8 XFD% and the 45 cm layer had the highest results with ~12.5 XFD% (Fig. 5.27). The 45 cm layer had the highest frequency-dependent susceptibility percentage and loss on ignition results (Fig. 5.27 and Fig. 5.28). This is interpreted as a colluvial event, which is likely correlated with a humid phase that occurred toward the end of last glacial maximum 11.5 Ka BP ± 1.2 Ka BP (Wang 2018). Figure 5.27: Hardpan Creek North frequency-dependent susceptibility and particle size <63 µm plot 152 | The Geoarchaeology of Eulo Ridge Figure 5.28: Hardpan Creek North loss on ignition and bulk density plot Hardpan Creek North Site Interpretation The Hardpan Creek North site provided evidence of punctuated change from 55.9 Ka BP ± 5.9 Ka BP (Wang 2019) to the present. Humid and arid phases oscillated and sediments were transported downslope during major humid events. At the base of the profile (97–106 cm), lenticular deposits and multiple phases of fluvial aggradation dominated these conditions. Regular sediment deposition created multiple phases along the margin of swamp, lake or watercourse. Micro features included dusty clay coatings of the sand grains, staining of the fine fabric, and fracturing of the clay coatings. Silt cappings on coarse grains implied cold conditions. A weathered stone artefact found wedged between lenticular layers about 102 cm deep suggested continual contact with the surrounding sediments (Fig. 4.12b). Fe-oxides are known to gradually erode silica particles and implies continual site stability. An absence of stone artefacts upstream relate to episodic change within a largely an unstable landscape. Sediment deposition was followed by a drying-out of the deposits possibly during the LGM. The base layer of the HPCKN site is the oldest known human-related deposit found in the northwest Murray-Darling Basin. An OSL date put the maximum age of 55.9 Ka BP ± 5.9 Ka BP (Wang 2019). The layer above (90–97 cm) suggests a transition to the last glacial maximum (LGM) with major changes to the landscape. Gravel and large rocks dominated this layer and it 5.6 Hardpan Creek North Study Site | 153 contained moderate amounts of organic matter and carbon residues but frequency-dependent susceptibility was lowest in this layer (~9 XFD%) compared to all other sediments in this arroyo. Soil colour changed to yellowish red (5YR 4/6) sandy loam. Soil particles, silt and clay, and organic matter particles in dusty black clay coatings in crescentic clay-illuvial features represent cyclical erosion from intermittent rainfall events (cf. Fedoroff et al. 2018: 848). These variables imply cycles of deforestation and intermittent rainfall on a bare surface and vegetated surfaces. A humid phase was followed by an arid phase or localised denudation of the land surface caused by consistently low rainfall or human modification. Sample 4 (45 cm, (OSL)) was marked by intense percolation of clay, silt and organic matter particles of dusty black clay coatings in crescentic clay-illuvial features. Distinct fragmented clay coatings had been modified by mechanical processes (cf. Kuhn et al. 2018: 390). Silt cappings of the coarse grains suggest cold frosty conditions with dry cloudless days (cf. Van Vliet-Lanoe and Fox 2018: 594). Grass phytoliths suggest that these were open grasslands. Fungal micro fossils found in this layer does not allow an interpretation because these occur in a wide range of habitats and environments. Micro-charcoal implies fires and possibly a human presence. A honey-comb phytolith (Fig. 5.20(c)) implied a fig tree (Ficus spp.) or woody plants growing in the valley about 11.5 Ka (cf. Collura and Neumann 2017: 156). Stone artefacts in this layer were weathered and buried post-LGM. Frequency-dependent susceptibility (XFD%) results did not correlate with <63 µm sediments, but the highest XFD% was recorded in this layer (45 cm). It is unclear what this means for these interpretations. Sample 3 (29–38 cm), a dark red sandy loam and its well-sorted structure with bridged clay coatings suggested stable conditions followed by drying-out of the landscape. Intrapedal diffuse Mn-hydroxide and CaCO3 nodules/aggregates support this hypothesis (cf. Vepraskas et al. 2018). Micro charcoal and burned opal phytoliths suggested either human or natural fires (cf. Parr (2006). Spherical (Cucurbita), Blocky and Elongate entire phytoliths implied a mixed landscape with grasses, shrubs/trees, and Cucumis myriocarpus. Cucurbita opal phytoliths suggested cold conditions or winter rainfall. It is well known that these species grow during winter rains (cf. QDPI 1974). The climate transitioned from humid to dry conditions and weathering of the profile during a dry phase. Sample 2, (2–9 cm), colluvial events were evident with laminated colluvium (Fig. 5.19) convey sheet wash (Mucher et al. 2018: 24–26). These red-brown sandy loams contained organic matter and sediment residues of 11.2%, much higher than other samples in this profile. It is likely seasonal waterlogging and a land slide caused a sudden movement of sediments downslope (cf. French 2003: 185). Transportation of this sediment is likely to have been associated with a saturated landscape and caused overland flow. Silt cappings of the coarse grains suggest cold and frosty conditions with clear skies. Micro-charcoal imply both natural and human fires. At this late stage in the sequence, there is high possibility that people 154 | The Geoarchaeology of Eulo Ridge managed these landscapes with fire to extend their ecological footprint (cf. Bird et al. 2012; Jones 1969). 5.7 Hardpan Creek Centre and South Study Sites Micromorphological characteristics Three micromorphology samples were collected from the HPCKC and HPCKS arroyos, 1) HPCKC Sample 1 (0-8 cm), 2) HPCKC Sample 2 (Mid) and 3) HPCKS Sample 3 (OSL) (Fig. 5.29). Table 5.17 and Appendix C describes the micromorphological results. Sediment samples were collected beside the micromorphology samples. These are discussed in more detail below. Figure 5.29: Hardpan Creek Centre: two undisturbed samples: 1) Sample 1 (0 – 8 cm) and; 2) Sample 2 (Mid) Sample 1 (0–8 cm) comprised a brownish black sandy loam. The groundmass was characterised by heterogeneous sub-angular blocky to sub-rounded microstructure dominated by strong vertical channels. Pedofeatures included clay coatings of coarse fraction, irregular zones in the groundmass, and weak crescentic infills in the voids (Table 5.15). Organics included 5.7 Hardpan Creek Centre and South Study Sites | 155 micro-charcoal, Blocky (some burned) and Elongate opal phytoliths. Figure 5.30 show sedimentary crusts, infillings of voids with parent materials, insect excrement and woody root epidermis (cf. Stoops et al. 2018). Table 5.15: Sample 1 (0–8 cm) summary descriptions and first level interpretations Major Characteristics Interpretation Channel microstructure; accommodating planes; vughy; clay coatings; <20% clay coatings of the sand grains; irregular zones of groundmass intermixed with CaCO3; bridged to partial coatings with clays and groundmass; weak to moderate birefringence; moderate to strong sesquioxides staining of the small amount of clays; dusty clays discontinuous; weak crescentic fills of the voids; diffuse Mn-hydroxide nodules; strong CaCO3 nodules/aggregates; organic matter fills the channels and voids; charcoal pieces within an organic fabric; Opal phytoliths: Blocky (some burned), Elongate Alluvial origins; groundwater fluctuations and translocation of the fines; stability in the coarse-fine fractions; saturation and drying processes; vegetated surfaces; Late Holocene alluvial surface buried by mulga soils and gradually exposed by erosion (a) (b) (c) Figure 5.30: Sample 1 (0–8 cm) fabric pedofeatures: a) soil surfaces (dusty clays); b) surface crust and Fe-oxides nodules (flatbed scanner, natural light); c) A) excrement (cf. Stoops et al. 2018: 457) and, B) woody root epidermis (PPL) (cf. Stoops et al. 2018: 477–480) Sample 2 (Mid) comprised sandy clay sediments. The groundmass was characterised by sub-angular blocky to sub-rounded poorly sorted blocky microstructure. Pedofeatures included a strong bridging of the sand grands, clay coatings, strong sesquioxides, and bridged to complete infills of the voids sometimes with CaCO3 (Fig. 5.31). Opal phytoliths included Spheroid, Arcuate, and Elongate morphotypes (Appendix C, Fig. C.16). Micro charcoal was found throughout the matrix. 156 | The Geoarchaeology of Eulo Ridge Table 5.16: Sample 2 (Mid) summary descriptions and first level interpretations Major Characteristics Interpretation Sub-angular blocky microstructure; strong bridging of the sand grains, accommodating planes; clay coatings; instability of coarse-fine fractions; dominant dusty clays with strong sesquioxides staining the fine fraction, weak to medium birefringence, yellow-orange-dark brown; bridged to complete infills; illuvial clay infillings; clean clays fill the 15 deg. channels; voids filled with surface materials; reticulate striations in the groundmass; diffuse Mn-hydroxide nodules <0.3 cm wide; CaCO3 coatings around voids and infills of vertical channels; CaCO3 nodules/aggregates <1.5 cm long and 0.5 cm high; Opal phytoliths: Spheroid, Arcuate, Elongate (Aristida spp.), Elongate geniculate Alluvial origins; strong groundwater fluctuations, saturation, evaporation and translocation of both the fines and amorphous fractions; Late Holocene alluvial surface buried by mulga soils and gradually exposed by erosion (a) (b) Figure 5.31: Sample 2 (Mid) fabric pedofeatures: a) Infillings of voids with organic matter and micro-charcoal (flatbed scanner, natural light); b) partial coating and infilling of void (flatbed scanner, natural light) Hardpan Creek South Sample 3 (OSL) (Fig. 5.32) comprised a sandy clay. The groundmass was characterised by poorly sorted sub-angular blocky to sub-rounded structure with a vughy to channel microstructure. Pedofeatures were dominated by strong vertical orientation of the sediments, reticulate striations, bridged to clay coatings, and medium to strong birefringence (Table 5.17). Figure 5.33 provides an example of the amorphous components including Fe-oxide nodules, carbonate recrystallisation, and CaCO3 infills. Opal phytoliths included Arcuate, Elongate (Aristida spp.) and Bilobate (Themeda spp.) morphotypes (Fig. 5.33), some of which were burned (Appendix C, Fig. C.15). 5.7 Hardpan Creek Centre and South Study Sites | 157 Figure 5.32: Sample 3 undisturbed summary descriptions Age of the deposit An OSL sample was collected beside a stone artefact cemented into the middle of the profile of the South terrace (Fig. 4.14 and Table 5.18. The sample returned an age of 4.6 Ka BP ± 0.4 Ka BP (Wang 2018) and aligns with a known humid phase in the MDB during the Late Holocene (cf. Fitzsimmons et al. 2013: 91). This chronology aligns with meandering rivers in the MDB and alluvial fans in the Flinders Ranges, South Australia (Fitzsimmons et al. 2013: 91). Table 5.17: Sample 3 (OSL): Summary descriptions and first level interpretations Major Characteristics Interpretation Vughy to channel microstructure; strong vertical orientation with channels; compound packing voids; vughs; <20% silt cappings; <40% pure clay coatings of sand grains, light brown to orange in colour; irregular zones of laminar clays to hints of reticulate striations of the groundmass; bridged to clay coatings; medium to strong birefringence; sesquioxides in the coatings and groundmass; Fe- and Mn-oxide nodules <1.2 cm wide and <1.5 cm high; CaCO3 infills to laminar deposits, CaCO3 nodules/aggregates <0.5 cm wide and <0.4 cm high; Organics: micro-charcoal; Opal phytoliths: Arcuate, Elongate (Aristida spp.), Elongate entire, Elongate genticulate, Elongate (burned), Bilobate (Themeda spp.), Blocky (some pitted) Alluvial origins; strong groundwater fluctuations and translocation of the fines; stability in the coarse-fine fractions; groundmass intermixed with CaCO3; mobilisation of the organics through groundwater flow; Late Holocene alluvial surface 158 | The Geoarchaeology of Eulo Ridge Table 5.18: Optically stimulated luminescence dates for Hardpan Creek South (HCKNS) (Wang 2018; 2019) Code a-Value* De (Gy) DR* (Gy/ka) Age (ka) Site ID WLL1345 0.07 ± 0.03 6.91 ± 0.27 1.50 ± 0.10 4.6 ± 0.4 HPCK C (*) estimated a-value; DR = dose rate; Sediment/soil characteristics All sediments/soils were extremely alkaline (>9). The upper layer was brownish black sandy loam (7.5YR 3/2), Layer 2 was dark (5YR 3/2) sandy loam (Fig. 5.34 and CaCO3 inclusions were visible in both layers. The sediments/soils were blocky and dispersed easily when disturbed. Physical and chemical characteristics An xy-plot of soil bulk density and loss on ignition residues show that LOI residues varied between the Centre and South terraces, for example, 4.7% for the Centre and 6.4% for the South site. This implies localised variations in water depth and/or vegetation growth in the immediate area. (a) (b) (c) (d) Figure 5.33: Sample 3 (OSL) fabric pedofeatures: a) Fe-oxide nodules (flatbed scanner, natural light); b) carbonate recrystalisation (flatbed scanner, natural light); c) CaCO3 infillings (flatbed scanner, natural light); d) Bilobate opal phytolith (possible Themeda spp.) (CPL) 5.7 Hardpan Creek Centre and South Study Sites | 159 Figure 5.34: Percent clay, silt, very fine, fine, and medium sand. Colour bars represent particle size distribution Figure 5.35: Loss on ignition and bulk density plot for HPCKS 160 | The Geoarchaeology of Eulo Ridge Frequency-dependent susceptibility (XFD%) results ranged from >6.1 % to <6.9 % for all layers (Fig. 5.36. The 40 cm layer had the lowest XFD% and 0-8 cm layer had the highest results (Fig. 5.36). Figure 5.36: Frequency-dependent susceptibility and particle size <63 µm plot for HPCKS Hardpan Creek South Site Interpretation The Hardpan Creek South site is a buried soil that formed between 5 Ka and 4.2 Ka. The 2 layers in this sequence exhibited a capping of brownish black alluvial sediment/soil that overlies a dark alluvial layer. The evidence suggests that this deposit formed during a humid period. In the upper layer, irregular zones found in the groundmass show crystalline CaCO3 and organics distributed throughout the channels and voids. Intermittent groundwater fluctuations likely caused the movement of organics and sediments (cf. Stoops et al. 2018: 853–854). In the lower layer, the groundmass comprises micro-charcoal, insect burrows/excrement, grass phytoliths, and and CaCO3 infills. These characteristics provide support for a water-dominated environment with a raised water table (cf. Stoops et al. 2018: 218). A Bilobate phytolith 5.8 Basin Gully Study Sites | 161 suggested that Themeda spp. grew in the area. Kangaroo grass was harvested and processed by early peoples (Latz 1995). An OSL age of 4.6 Ka BP ± 0.4 Ka BP (Wang 2018) centres this layer with the active rivers in the MDB during the mid Holocene (cf. Fitzsimmons et al. 2013: 92). Such conditions result from intermittent channel flows and movement of sediments downslope from Hoods Range. Surface deflation and overbank flows were likely responsible for landscape change over time. Stone artefacts and charcoal found in this deposit suggested early peoples lighting fires. Alternatively, natural fires from lightning strikes contributed to these fires. 5.8 Basin Gully Study Sites Micromorphological characteristics Two micromorphology samples were taken from the BG1 and BG2 arroyo sites. Both samples were collected next to OSL samples (Fig. 5.37). Sediments samples and OSL dates were collected next the micromorphology samples. These are discussed in detail below. (a) (b) Figure 5.37: Basin Gully profiles: a) BG1 site: location of undisturbed samples (rectangle), OSL samples (circle) and, b) BG2 site: location of undisturbed samples (rectangle), OSL samples (circle), and in situ stone artefact (arrow) BG1 site (Table 5.19) sediments comprised clay loam. The groundmass is characterised by poorly sorted sub-angular blocky to sub-rounded structure, and channel to vughy microstructure. Pedofeatures are dominated by <50% dusty clay coatings, linked/bridged, reticulate striations, partial/discontinuous infills, moderate to strong sesquioxides, crescentic 162 | The Geoarchaeology of Eulo Ridge void fills, channel and large quartz grain coatings (Fig. 5.39). Organic fragments were found in this layer. Opal phytoliths included Elongate and Elongate sinuate morphotypes, and sponge spicules suggested a freshwater environment (Appendix C Fig. C.1 and Fig. 5.38) (cf. Racek 1969). Table 5.19: Basin Gully 1 (BG1)summary descriptions and first level interpretations Major Characteristics Interpretation Channel to Vughy microstructure; fine porphyric c/f-related distribution pattern; <50% dusty clay coatings of sand grains with irregular zones; linked/bridged coatings; moderate birefringence; moderate to strong sesquioxides; dusty to pure clays, illuviation/clay coatings; partial depletion of fine fabric in upper zone and degradation of soil fabric in lower section; near complete infills of channels; silt cappings of larger grains; moderate Fe- and Mn-oxide nodules; organic fragments; Elongate sinuate, Elongate, Elongate dentate phytoliths and sponge spicules Commencement of the Last Glacial Maximum; freshwater conditions; cold period with frosts; stable LGM soils; bioturbation; groundwater fluctuations; saturation, limited translocation of the fines; drying out of the sediments Figure 5.38: BG 1: organic fragment with Elongate sinuate phytoliths (arrows) (PPL) 5.8 Basin Gully Study Sites | 163 (a) (b) (c) (d) Figure 5.39: BG 1, a) clay infills and coatings (PPL); b) Intense percolation and clay dispersal (cf. Stoops et al. 2018: 848) (PPL); c) silt cappings (cf. Van Vliet-Lanoe and Fox 2018: 593–594) (PPL); d) Fe-oxide nodule (flatbed scanner, natural light) BG2 (Table 5.20), sandy clay was poorly sorted and dominated by a channel microstructure with <75% dusty to pure clay coatings, weak crescentic coatings, reticulate striations, hypocoatings around rhizoliths, and partial to complete infills of the voids. Silt cappings around large quartz grain (Table 5.20) (cf. Van Vliet-Lanoe and Fox 2018). Micro-charcoal <4 mm long (Fig. 5.40), and Acute bulbosus morphotypes phytoliths (Fig. 5.41). Table 5.20: Basin Gully 2 (BG 2) summary descriptions and first level interpretations Major Characteristics Interpretation Channel microstructure; accommodating planes; <75% dusty to pure clay coatings of sand grains, linked/bridged; reticulate striations; moderate birefringence; weak crescentic coatings; hypocoatings; <20% silt cappings of large grains; partial to completely filled voids; <10% Fe- and Mn-oxide nodules; weak carbonates; micro-charcoal; Elongate and Acute bulbosus phytoliths Cold conditions with frosts; bare soils; limited groundwater fluctuations; stable environment 164 | The Geoarchaeology of Eulo Ridge (a) (b) (c) Figure 5.40: Basin Gully 2 fabric pedofeatures: a) hypocoating around root void; b) partial infilling of void; c) sub-angular blocky quartz clasts, vughs, and Fe/Mn-oxide nodules (All images, flatbed scanner, natural light) (a) (b) (c) (d) Figure 5.41: Basin Gully 2 phytoliths: a) to d) Acute bulbosus (PPL) strongly associated with Poaceae and Cyperaceae (Neumann et al. 2018: 2–3) Age of the deposits The BG1 OSL sample returned an age of 25.2 Ka BP ± 2.6 Ka BP (Wang 2018). This age aligns with the commencement of the LGM. Humid conditions and precipitation are likely to have prevailed throughout the study area at this time (cf. Fitzsimmons et al. 2013). The BG2 OSL sample returned an age of 19.6 Ka BP ± 2.0 Ka BP (Wang 2018). Site formation occurred during the height of the LGM. Clays where being mobilised by wind and Acute bulbosus phytoliths identified a grass dominated environment. 5.8 Basin Gully Study Sites | 165 Table 5.21: Optically stimulated luminescence dates for Basin Gully 1 (BG1) and Basin Gully 2 (BG2) (Wang 2018; 2019) Code a-Value* De (Gy) DR* (Gy/ka) Age (ka) Site ID WLL1349 0.07 ± 0.03 48.30 ± 0.03 1.92 ± 0.18 25.2 ± 2.6 BG1 WLL1350 0.07 ± 0.03 30.51 ± 0.085 1.56 ± 0.15 19.6 ± 2.0 BG2 (*) estimated a-value; DR = dose rate; Sediment/soil characteristics: Two sediment samples were analysed for particle size analysis and loss on ignition. All sediments/soils were extremely alkaline (>9). Table 5.22 shows the distribution of particle size for each sample site. Table 5.22: Basin Gully sites particle size analysis Clay Silt V F Sand F Sand M Sand C Sand F Gravel BG1 3.87 25.67 25.13 29.729 15.22 0.39 0.00 BG2 10.53 34.42 22.6 21.98 10.47 0.00 0.00 The BG1 site had been exposed by surface deflation, sheet wash, and channel erosion. The sediment/soil was friable and crumbled when disturbed. CaCO3 evaporites were found within the fissures and sediment/soil matrix. Sediments were a clay-loam with poorly sorted sub-angular blocky coarse fraction. The coarse/fine fraction comprises ~44%, and very fine sands and silts were ~50%. The BG2 site was exposed by gully erosion (Fig. 4.17). Water flows in an easterly direction from the BG1 site and out on to the nearby plain. The main erosion processes were surface deflation, sheet wash, and gully erosion. This gully was derived from water flows along the channel of what appears to be the old vehicle track, and the new track was situated almost parallel to this gully. The BG2 sediment/soil was a loam. Silt strongly dominates this layer at 34.4%, and there is a higher than normal clay component of 10.5%. The very fine and fine sand fractions have a combined total of 44.5%. Physical and chemical characteristics Soil bulk density and loss on ignition residues show moderate to high compaction in both sites. LOI residues for BG1 was ~6.1% and BG2, ~2.7% (Appendix B, Fig. B.5). These indicate that there has been more vegetation and water movement in the BG1 site than BG2 and thus a humid period existed here. Frequency-dependent susceptibility (XFD%) results ranged from >11.8 XFD% to <12.45 XFD%. The BG2 site had the lowest with ~11.8 XFD% and BG1 site had 12.45 XFD%(Fig. 5.42). 166 | The Geoarchaeology of Eulo Ridge Figure 5.42: Basin Gully frequency-dependent susceptibility and particle size <63 µm plot Additional features: BG1 site comprised sediment laminations or what is interpreted as rhythmite layering. Glacial floodwaters laid sediments down over a long period of time (Fig. 5.43). These features are common within the rising and fall of flood waters during a episodic glacial phases (cf. van der Meer and Menzies 2011). An alternative to these conditions is fluctuating groundwater tables. Pools fed by the aquifer could imply changing heights or levels implies topping up of system from distant intake areas of the Great Artesian Basin. Little information is available to determine if this was the case or not. Basin Gully Sites Interpretation Basin Gully formed during the last glacial maximum. BG1 site showed evidence of episodic groundwater flows which implied water lapping the vegetated margin of pool or waterbody. At BG2, conditions were arid and windblown sediments accumulated on a low vegetated dune ~22 Ka to ~18 Ka. Skies were clear with frosts dominating these hyper-arid conditions. The BG1 site aligns with the beginning of the Last Glacial Maximum. Its lenticular beds show a sedimentary archive of episodic water levels. Analysis of these sediments identified 5.8 Basin Gully Study Sites | 167 Figure 5.43: Rhythmite features were identified in the BG1 site. These infer episodic flooding with rising and falling water levels on the edge of a lake or pool. The deposition of sediment in multiple episodes likely occurred during a post-glacial phase bridged coatings, clay infills within the voids, and types of duricrusts support this observation (cf. Stoops et al. 2018: 138). In addition, high frequency-dependent susceptibility (XFD%) results (12.45 XFD%) and sponge spicules suggest a wet environment and aggradation of older sediments. Organic fragments and Elongate opal phytoliths imply a grassed environment. This layer dated to 25.2 Ka BP ± 2.6 Ka BP (WLL1349) (Wang 2018) and aligns with MIS 2, at approximately the commencement of the Last Glacial Maximum. Silt cappings on the coarse fraction implies frosty conditions and suggests a cold wet period at the commencement of 168 | The Geoarchaeology of Eulo Ridge the LGM. Stone artefacts embedded in this layer suggest people lived along the edge of this waterbody. Runoff are likely to have transported these stone artefacts to the water’s edge. BG2 site was dated to 19.6 Ka BP ± 2.0 Ka BP (Wang 2018). Clays and silts comprised 45 % and suggested that at the height of the LGM this was a loess. Silt capping of the coarse quartz grains suggested cold, frosty and dry conditions (cf. Van Vliet-Lanoe and Fox 2018: 593–594). Frequency-dependent susceptibility (XFD%) returned a ~11.8 XFD% result and suggested a stable homogenous environment. Low organic and carbonate residues identified an arid or dry phase, or this layer had been leached of its residues. Elongate and Acute bulbous opal phytoliths supported evidence for a grass dominated environment. This layer was interpreted as a palaeo-dune covered with grasses and continuous mobile surfaces. Sponge spicules suggested a freshwater environment but these were likely transported to the dune by wind or within faecal pellets (Wallis 2001: 114). Micro-charcoal in this layer implies firing of the landscape at the height of the LGM. 5.9 Case Study 3: Double Well and Tunkana Well 5.10 Double Well Study Site Micromorphological characteristics Five micromorphology samples were collected from the profile to obtain a representation of the micro-features (Fig. 5.44): 1) 1–10 cm, 2) Top–Mid, 3) Middle, 4) Mid-Low and, 5) 162–172 cm. These tags related to the position of the samples and points of interest down the profile. The ’stone line’ (Middle) is located in the approximate centre of the arroyo. Micromorphological samples were taken within the middle layer, and above and below it. It is assumed that this stone layer formed as a result of a major flooding event that flushed stone downstream and were deposited as the energy decreased and dumped the stone pebbles and gravel. All layers are discussed in more detail below. Sediment samples were collected next to the micromorphology samples. These are discussed in more detail below. Sample 1 (1–10 cm) sediments comprised a sandy silt loam. The groundmass is characterised by well sorted sub-angular blocky structure, and lenticular microstructure. Pedofeatures were dominated by <60% dusty clays, partial to complete infills of the voids, hypocoatings, surface crusts, and weak birefringence (Fig. 5.45) (Table 5.24). Its amorphous characteristics were dominated by intrapedal Fe- and Mn-oxide nodules. Opal phytoliths included Elongate entire (burned), Acute bulbosus, Spheroids, Blocky morphotypes and determines that this was a grass, shrub and tree environment. 5.10 Double Well Study Site | 169 Figure 5.44: Double Well location of undisturbed samples: 1–10 cm; 2) Top–Mid; 3) Middle; 4) Mid-Low and; 5) 162–172 cm Sample 2 (Top-Mid) sediments comprised a sandy silt loam. The groundmass is characterised by poorly to moderately sorted sub-angular blocky to sub-rounded structure, and sub-angular blocky to blocky microstructure. The pedofeatures were dominated by complete to bridged infills, sediment sorting with coarse at the top and fines at the base, redoximorphic silts, and <50% silt cappings (Fig. 5.46) (Table 5.24). The amorphous component is dominated by intrapedal diffuse Mn-oxide nodules and CaCO3 infills. The organic component comprised organic matter. Opal phytoliths included Elongate entire (Poaceae), Elongates (some burned), and Acute bulbosus morphotypes (Appendix C, which suggest a grassed environment C.18). Burned phytoliths suggest regular fires lit by either hunter gatherer or lightning strikes. Humid conditions and significant grass growth during wet periods could imply a naturally 170 | The Geoarchaeology of Eulo Ridge Table 5.23: Double Well (1–10 cm) summary descriptions and first level interpretations Major Characteristics Interpretation Lenticular microstructure; accommodating planes; vughs; soil nodules; surface crusts; abundant dusty clays of the sand grains <60%; weak pure clays in channels; partial to complete infills and coatings of the channels and voids; hypocoatings of the voids; surface crusts on the lenticular peds, top and bottom of peds—crusts <3.0 cm long and 0.15 wide; hyper-coatings; Intrapedal parent material infills, clay coatings; infills will groundmass; weak birefringence, brown to orange; some soil surface peds; sesquioxides staining within the groundmass; Intrapedal diffuse Mn-hydroxide nodules <0.15 cm; Elongate entire (Poaceae), Elongate (burned), Arcuate bulbosus, Blocky (some pitted) phytoliths, fungal morphotypes, and opalised organic matter Alluvial origins; movement of fines to form crusts; groundwater fluctuations and translocation of the fines; movement of fine fractions by groundwater (eluviation); Fe- and Mn-oxide formation; frost heave; wet environment (a) (b) (c) (d) Figure 5.45: Double Well Sample 1 (1–10cm) fabric pedofeatures: a) Fe-oxide coatings and infillings of channels; b) surface crusts and Fe-oxide nodules; c) and d) infilling of void with parent material (flatbed scanner, natural light) fired landscape (Bird and Frankel 1991; Kimber 1983). Sample 3 (Middle) sediments comprised a sandy clay. The groundmass is characterised by well sorted sub-rounded structure with a channel to vughy microstructure. Surface crusts were found throughout the matrix. 5.10 Double Well Study Site | 171 Table 5.24: Double Well Sample 2 (Top–Mid) summary descriptions and first level interpretations Major Characteristics Interpretation Angular to sub-angular blocky microstructure; highly separated packing voids; accommodating planes; vughy; silt cappings <50% and dusty clay in between sand grains; irregular groundmass with depletion of the groundmass; coarse sand fraction at the top zone, and fine groundmass fabric at the base; complete to bridged infills of the channels and voids; weak to moderate birefringence; strong sesquioxide staining of the groundmass, dark brown; partial to discontinuous to near complete infills of the voids and channels; weak pure clays; mineral grains capped with redoximorphic silts; intrapedal clay coatings; intercalations; large intrapedal diffuse Mn-hydroxide nodules; CaCO3 infills; Elongate entire (some burned), Elongate, Acute bulbosus (pitted) Alluvial origins; groundwater fluctuations and translocation of the fines; instability of coarse-fine fractions; groundmass depletion; Fe- and Mn-oxide and CaCO3 formation; frosty conditions; buried alluvial surface (d) (b) (c) Figure 5.46: Double Well Sample 2 (Top–mid) fabric pedofeatures: a) pure clays filling a channel (CPL); b) Fe-oxide nodule; c) re-crystalised carbonate (flatbed scanner, natural light) Pedofeatures included continuous to partially-filled voids, pure to dusty clays, soils nodules, and moderate to strong sesquioxides. Its amorphous component was dominated by Fe- and Mn-oxide nodules. Crystalline components were CaCO3 void infills (Fig. 5.47) (Table 5.25). The opal phytoliths were dominated by Polylobate, Dentate (possible Brachiaria spp.), Elongate dentate (Poaceae), Elongate, Acute bulbosus, Spheroidal (likely Cucurbita spp.) morphotypes and, fungal micro fossils (Appendix C, Fig. C.19). Sample 4 (Mid-Low) comprised a sandy clay. The groundmass was characterised by a poorly to moderately sub-angular blocky structure with a granular to vughy microstructure. The pedofeatures were dominated by <45% intrapedal dusty clays , <40% reticulate striations; moderate to strong birefringence, moderate to strong sesquioxides, and intradpedal infills (Fig. 5.48) (Table 5.26). Amorphous components were Fe- and Mn-oxide nodules. Crystalline components were <60% CaCO3 infills in the voids. Organics were dominated by plant fragments. Opal 172 | The Geoarchaeology of Eulo Ridge Table 5.25: Double Well (Middle) summary descriptions and first level interpretations Major Characteristics Interpretation Channel to vughy microstructure; accommodating planes; vughs; compound packing voids; sub-angular blocky peds <2.8 cm long and <1.2 cm wide; surface crusts <1.3 cm long and 0.15 cm wide; continuous to partially filled voids with pure to dusty clays; partial to complete coatings and infills of the voids, <10% dark brown; moderate to strong sesquioxides staining; medium to strong birefringence; pure clay infills of the vertical and horizontal channels; strong crusts in the middle section; silt cappings of sand grains; irregular zones of groundmass; Fe- to Mn-impregnated surfaces; intrapedal diffuse Fe- and Mn-hydroxide nodules; Wood fragments; Opal phytoliths: Elongate entire, Blocky (some broken), Bulliform flabellate, Arcuate bulbosus Colluvial origin; groundwater fluctuations/oxidation and translocation of the fines; instability in the groundmass; Fe- and Mn-oxide formation; phytoliths within soil fabric; illuviation; former floodplain buried by other alluviation phytoliths included Elongate entire, Blocky (some broken), Bulliform flabellate, and Arcuate bulbosus and fungal micro fossils, and sponge spicules were found in this layer (Appendix C, Fig. C.20). Figure 5.47 implies high water tables, illuviation and, grasses growing in the area. Sample 5 (162–172 cm) comprised a loamy sand. The groundmass was characterised Table 5.26: Double Well (Mid-Low) summary descriptions and first level interpretations Major Characteristics Interpretation Granular to vughy microstructure; accommodating planes; vughs; granular peds <4.1 cm high and <2.6 cm wide; <50% dusty clay coatings and irregular zones; intrapedal and bridged clays; pure clay coatings of channels; moderate to strong birefringence; moderate to strong sesquioxides staining; reticulate stratification; intrapedal diffuse Mn-hydroxide nodules toward base; CaCO3 nodules/aggregates within voids and groundmass; wood fragments; Elongate entire (Poaceae), Blocky (some broken), Bulliform flabellate, Acuate bulbosus phytoliths and sponge spicules Alluvial origin; groundwater fluctuation/oxidisation and translocation of the fines; instability in the coarse-fine fractions; Fe- and Mn-oxide formation; illuviation; former floodplain buried by other alluviation; grassed environment; wet/humid conditions (sponge spicule) by poorly to moderately sorted sub-angular blocky to sub-rounded structure with a channel to vughy microstructure. Pedofeatures were dominated by <50% silt cappings and coatings of the sand grains, moderate to strong birefringence, and sesquioxide staining of the fine fabric (Table 5.27). Amorphous components included diffuse Mn-hydroxide nodules, and 0.3 cm long and 0.6 wide . Crystalline components included CaCO3 infillings and aggregates <0.6 cm long and <0.4 cm wide. Opal phytoliths included Elongate (possibly Phragmites australis), 5.10 Double Well Study Site | 173 (a) (b) (c) (d) Figure 5.47: Double Well Sample 3 (Middle) fabric pedofeatures: a) CaCO3 infills (CPL); b) Pure clay infills (PPL); c) Dusty clay soil nodule (PPL); d) Elongate dentate phytoliths (Poacae) (PPL) (a) (b) Figure 5.48: Double Well Sample 4 (Mid-Low): a) and b) pure clay coatings (PPL) (possibly colluvial) (PPL) Tabular, Acute bulbosus, conical-single-perforate (Araucaria cunninghamii), and trough (Zingiberales) (Appendix C, Fig. C.21). 174 | The Geoarchaeology of Eulo Ridge Table 5.27: Sample 5 summary descriptions and first level interpretations Major Characteristics Interpretation Channel to vughy microstructure; accommodating planes; vughs; abundant clay coatings of <50% silt cappings and coatings of the sand grains and thus irregular zones of groundmass, red to light browns; moderate to strong birefringence; sesquioxide staining of the fine fabric; minimal dusty to pure clays infills; some reticulate striations light brown to dark brown; Fe/Mn-oxide nodues at the base of the channel and CaCO3 flow; diffuse Mn-hydroxide nodules, 0.3 cm long and 0.6 wide; stabilised CaCO3 flows within the voids, CaCO3 nodules/aggregates <0.6 cm long and <0.4 cm wide; Opal phytoliths: Elongate (possibly Phragmites australis or reeds), Tabular, Acute bulbosus, conical-single-perforate (Araucaria cunninghamii) and fungal morphotypes Alluvial origin; limited groundwater fluctuations and translocation of the fines; stability in the coarse-fine fractions; Fe- and Mn-oxide formation; Opal phytoliths; illuviation; frosts; former alluvial sediments buried by other alluvial events (above); wet or humid environment, reeds growing near a water source Sediment/soil characteristics Nine sediment samples were analysed for particle size distribution and loss on ignition. Figure 5.49 presents the results of these analyses. All sediments/soils were extremely alkaline (<9). Figure 5.49: Double Well particle size analysis plot: 9 sediment samples showing percent clay, silt, very fine sand, fine sand, and medium sand (total = 100%) with Munsell colour and sediment texture 5.10 Double Well Study Site | 175 Physical and chemical characteristics An xy-plot of soil bulk density and loss on ignition residues shows less compaction around the stone artefacts and gravel disconformity, for example, 1.06% to <1.09%, and high LOI residues, ~8.5% (Fig. 5.50). The upper and lower layers were between the 4.5% and 6.2% and bulk density was between 1.14 g/cc and 1.353 g/cc. The 18–26 cm sample had the most sediment/soil compaction and implies indurated sediments. Frequency-dependent susceptibility (XFD%) separated into two clusters. The upper cluster included the middle and middle-low layers and ranged from >10.5 % and <12 % and had >40 % fine sediments. Cluster 2 included all other layers and ranged from 8 % to <11.5 % and these finer sediments were <31 % and >16 %. The Upper/Mid layer had the lowest with ~8.5 XFD% and 1-10 cm layer had the highest percent with 12 XFD% (Fig. 5.51). The highest amount of residue correlates with the stone artefact layer and indicates micro organics and carbonate had been transferred as part of this probable colluvial event. Figure 5.50: Double Well loss on ignition and bulk density plot Additional features: The Double Well site comprised a 10 to 15 cm layer of sediments about midway between the surface and base of the profile (Fig. 5.52, see also Chapter 4). This 176 | The Geoarchaeology of Eulo Ridge Figure 5.51: Double Well frequency-dependent susceptibility and particle size <63 µm plot deposit was interpreted as an episodic flash flooding event that transported sediment, gravel, stone artefacts, and organics downstream. The upstream landscape comprised conglomerate deposits which derived from the Glendowner formation (cf. QDPI 1974). Landscape erosion and exposure allowed fast flowing water to erode and carry these gravels and materials downstream. The outcomes was this thin layer which provides evidence of this unique event. Ultimately, these intense events promote significant amounts of sheet wash and surface erosion which mobilises all materials in its wake. Double Well Site Interpretation Double Well site comprises a dynamic environment. At the earliest time available in this site, conditions were stable and wet with cold periods and frost. A pine forest grew nearby and reeds lined the margins of spring pools. This was followed humid conditions with flooding, pedogenesis and vegetated surfaces. Later, an abrupt change occurred with colluvium being deposit on the lower slopes and floodplain margin. The site was covered with vegetation and freshwater pools located across the site. The remaining periods to the Late Holocene included 5.10 Double Well Study Site | 177 Figure 5.52: Double Well site: stone line, which dissects the arroyo has gravel and stone artefacts within the matrix further colluvium and alluvium being deposited on the floodplain. Natural and/or human fires appeared more common in the upper layers. In the 162 cm – 172 cm sample or bottom layer, reticulate striations and silt cappings on the quartz grains identified that these were stable cold frosty conditions with active springs and a high groundwater table. Opal phytoliths included Elongate (possibly Phragmites australis or reeds), Elongate entire, Acute bulbosus, Tabular morphotypes, and fungal micro fossils such as a trough. Zingiberales or reeds were associated with permanent water and supported a stable wet environment. Reeds are commonly found in pools of water along the creeks and river systems of the northwest MDB (Boyland 1974). The conical-single-perforate morphotypes suggest a stand of Araucaria cunninghamii growing nearby. If so, the evidence points toward pine forests growing under cold conditions (cf. Parr 2006: 83) (Appendix C, Fig. C.21). These results suggested that this was a dynamic environment and very different to the present. In the Mid-Low layer, intermittent groundwater fluctuations appeared to modify this site. Dusty clay coatings, irregular zones, and pure clay coatings of the channels, translocation of the fines suggested pedogenesis or soil forming conditions. Mobile sediments are thought to exist because of high proportions of organic and carbon residues. Opal phytoliths included Bulliform flabellate (grasses) Polylobate, Dentate, Elongate, Acute bulbosus, and Spheroidal (likely Cucurbita spp. or wild cucumber) morphotypes suggested that this was a grassed 178 | The Geoarchaeology of Eulo Ridge environment with winter rains to stimulate growth of the wild cucumbers. Sponge spicules suggest that this was a freshwater palaeoenvironment (cf. Racek 1969). Wet or humid conditions align with pre-Last Glacial Maximum humid period (Fitzsimmons et al. 2013: 91). These conditions supported a diverse vegetated environment with intermittent flooding with the transport of sediment downslope and downstream. In the Middle layer, a <15 cm stone/gravel lens suggested a unique event such as colluvial processes. This layer contained surface crusts, dusty clay soil nodules, partial to complete infilling of the voids, strong sesquioxide staining, Fe- and Mn-oxide nodules, and wood fragments. Strong water flow is likely to have carried sediments, gravel, archaeological material, and organics downstream to this site. It is unclear when this occurred but it could relate to a Younger Dryas event about 14 Ka BP (Denniston et al. 2013). Opal phytoliths included Elongate entire, Blocky, Bulliform flabellate, and Arucate bulbosus morphotypes. Some phytoliths were broken which suggested that they had been broken as part of transportation by water and as part of a slurry of sediment (cf. Kaczorek et al. 2018: 163). The Upper-Middle layer had the lowest organic/carbon residues and frequency-dependent susceptibility result of all layers in this profile. Together with an irregular groundmass, silt cappings of the coarse grains, complete to bridged infills of the channels and weak to moderate birefringence, conditions in this layer were unstable frosty environment. Elongate and Acute bulbosus (some burned) opal phytolith morphotypes defines that this landscape was dominated by grasses. The Upper layer (0–10 cm) had friable lenticular beds and identified periodic overbank flooding and overland flow conditions (cf. French 2003: 125). These alluvial origins included surface crusts on the top/bottom of these lenticular peds, abundant dusty clays, weak clays in the channels, hypocoatings, and intrapedal diffuse Mn-hydroxide nodules (Fig. 5.45). Opal phytoliths included Elongate entire, Acute bulbosus, Spheroids, and Blocky morphotypes and suggested that this had a diverse grass, shrub and tree landscape. Burned phytoliths could suggest human fires (cf. Parr 2006). 5.11 Tunkana Well Study Site Micromorphological characteristics Two micromorphology samples were collected from this profile. Sample 1 (Mid-Low) and Sample 2 (138–144 cm) (Fig. 5.53). Sediment samples were collected next to the micromorphology samples. These are discussed in more detail below. Sample 1 (Mid-Low) comprised a clay loam sediment. The groundmass was characterised by poorly to moderately sub-angular blocky to sub-rounded structure with a granular to vughy microstructure. Pedofeatures were dominated by clay hypocoatings around the mineral grains, 5.11 Tunkana Well Study Site | 179 reticulate striations, and weak to moderate birefringence. CaCO3 infilled voids and channels and micro charcoal was dispersed throughout the matrix. Opal phytoliths included Elongate entire (Poaceae), Elongate, Blocky (some broken), Bulliform flabellate, Arcuate bulbosus morphotypes. Broken phytoliths assumed post depositional transport of sediments. Sponge spicules in the matrix suggested a raised water table and/or wet/humid conditions (Table 5.28) (Appendix C, Fig. C.22). Sample 2 (138–144 cm) comprised a sandy clay. The groundmass Table 5.28: Tunkana Well (Mid–Low) summary descriptions and first level interpretations Major Characteristics Interpretation Granular to vughy microstructure; accommodating planes; vughs; dusty clays; reticulate striation; dusty clay coatings <30% of sand grains; irregular zones in groundmass; weak to moderate birefringence; clay hypocoatings around large mineral grains; hints of reticulate striations; soil fragments; diffuse Mn-hydroxide nodules <0.2 cm; CaCO3 complete to partial filling of voids; carbonate nodules <0.4 cm; Micro-charcoal; Elongate entire (Poaceae), Elongate, Blocky (some broken), Bulliform flabellate, Arcuate bulbosus phytoliths and sponge spicules Colluvial origins; water table fluctuations (eluviation); drying phase; natural or possible human fires; grassed and tree environment was characterised by sub-angular blocky to sub-rounded poorly to moderately sorted structure with a vesicular microstructure. Pedofeatures were dominated by dark brown clay and dusty clay coatings, and the groundmass comprised <45% large mineral grains (Fig. 5.55) (Table 5.29). Fe-oxide staining of minerals and recrystallised CaCO3 were found throughout. Opal phytoliths included Acute bulbosus, Blocky, Elongate entire, Spheroid areolate morphotypes (Appendix C, Fig. C.21). These suggested that this was largely a grasses and woodland environment. Table 5.29: Tunkana Well Sample 2 (138–144 cm) summary descriptions and first level interpretations Major Characteristics Interpretation Vesicular microstructure; compound packing voids; vughy; dark brown clay coatings and dusty clays infill the compound packing voids and around mineral grains, moderate birefringence; irregular zones between large grains; moderate sesquioxide staining in groundmass; dark brown clay staining of large grains; large Fe/Mn-hydroxide nodules; CaCO3 nodules/aggregates <0.4 cm and voids completely infilled; Acute bulbosus, Blocky, Elongate entire, Spheroid areolate, and fungal forms Groundwater origins; fluctuation and translocation of fines; instability in coarse-fine fractions; groundmass depletion; leeching of the sediment; secondary CaCO3 formation; tree, shrub and grassed environment 180 | The Geoarchaeology of Eulo Ridge Figure 5.53: Tunkana Well Profile: 1) Mid-Low and; 2) 138–144 cm Sediment/soil characteristics Eleven random sediment samples were analysed (Fig. 5.56). All sediments/soils were extremely alkaline (>9). The upper layer had a pale brown (10YR 6/3) silt loam surface 5.11 Tunkana Well Study Site | 181 (a) (b) Figure 5.54: Tunkana Well Sample 1 (Mid-Low): a) CaCO3 infilling and Fe-oxide; b) Micro-charcoal (flatbed scanner, natural light) (a) (b) (c) Figure 5.55: Tunkana Well Sample 2 (138-144 cm): a) and b) recrystalised CaCO3; c) Fe- and Mn-oxide (flatbed scanner, natural light) above a reddish brown (2.5YR 3/4) silt loam. Layer 2 was a dark reddish brown (2.5YR 3/4) sandy loam. Layer 3 was a dull reddish brown (5YR 4/4) sandy loam, Layer 4 was a dull reddish brown (5YR 4/3) sandy loam, Layer 5 was a yellowish brown (10YR 5/4) loamy sand, and Layer 6 was a greyish brown (10YR 5/2) sandy loam. The boundary between Layer 3 and Layer 4 was characterised by ~63% silt and ~16% clay and ~19% very fine sand (Fig. 5.56). Distinct differences in sediment texture and colour suggested a deposition of sediments formed Layer 2, between Layer 3 and Layer 4, and Layer 5 and Layer 6. It is likely that groundwater fluctuation or flooding and settling out of sediments contributed to changes in particle size distribution (cf. Goudie 2013). Physical and chemical characteristics Frequency dependent susceptibility analysis separated samples into two main groups, 1) <6 XFD% and 2) >7 XFD% with two outliers in each group. These results implied raised water tables and possible humid conditions, which could be associated with colluvial and alluvial events. Loss on ignition residues identify that upper layers including the spring sediments contain the smallest small percentages of organics 182 | The Geoarchaeology of Eulo Ridge Figure 5.56: Tunkana Well particle size analysis plot: 11 samples showing percent clay, silt, very fine sand, fine sand, and medium sand in association with Munsell colour and the arroyo profile. Horizontal colour bars represent particle size percentages that total 100% and carbonates (Group 1) (Fig. 5.58). Data points between 3.2 % and 5.2 % account for the majority of the sediments. Stable environments without significant amounts of vegetation was interrupted by episodic flooding (Group 3). Further, Layer 1 had the highest bulk density percentages and is consistent with soil compaction from animal trampling during the pastoral period (cf. Fanning 1999). Tunkana Well Site Interpretation Tunkana Well site comprises a dynamic humid and arid phases with fluctuating groundwater tables throughout time. The sequences begin with an unstable environment with arid conditions and changing levels within the aquifer. Alluvium followed with high water tables, humid conditions, vegetation growing near springs, and natural and/or human fires. Within the estimated Late Holocene, the environment was dominated by springs and formation of an active spring at the surface. The springs continued to deposit fresh sandstone sediments on to the upper surface. These fluctuating conditions are presented below. Layer 7 was dominated by fluctuating groundwater conditions. The sediments comprised comminuted amorphous and crystalline pedofeatures and suggested an unstable environment with wetting and drying of the deposits. Subsequent erosion removed plants and exposed 5.11 Tunkana Well Study Site | 183 Figure 5.57: Tunkana Well frequency-dependent susceptibility and particle size <63 µm plot bare surfaces. The timing of this change is undated and therefore cannot be aligned with well known climatic events. Layer 6 to Layer 4 had a bi-modal particle size distribution with peaks at the boundary between layer 3 and 4 and between layer 5 and 6 respectively. Each fine fraction peak was followed by depletion of the matrix with formation of Fe- and Mn-oxide and secondary CaCO3. Layer 5 data implies either alluvial or intermittent raised groundwater tables dominant environment. Sediment texture change with almost equal proportions of silt, very fine sand, fine sand, and medium sand identifies an alluvial layer or raised groundwater tables. Layer 3 had an abrupt and significant increase in clay and silts percentages. This appeared to relate to a former spring and expulsion of spring sediments as the water tables increased. Its organic and carbonate residues were ~5 % and frequency-dependent susceptibility ~8 XFD%. Reworking of spring-related sediments resulted from a raised water table. Soil fragments, dusty clay coatings, hypocoatings, and reticulate striations in the matrix confirmed stable environments. Complete to partial infills of the voids with CaCO3 identified a gradual drying-out of the sediments. Micro-charcoal implied natural fires and possible human 184 | The Geoarchaeology of Eulo Ridge Figure 5.58: Tunkana Well loss on ignition and bulk density plot fire management. Elongate, Blocky, Bulliform flabellate, Acuate bulbosus opal phytoliths identified a grass and tree environment (cf. Cabanesa et al. 2011). Humid freshwater conditions were identified with sponge spicules environment. Layer 2 appeared heavily stained with Fe-oxides. A 5 % organic and carbonate residues result and, frequency-dependent susceptibility ~9.5 (XFD%) identified a dynamic hydrological environment at a local scale. Sediments had been transported downstream and redeposited by water and possibly a colluvial event. An OSL date could align this environment with other colluvial events in the greater study region. The upper layer comprised silt loam sediments which were deposited by the springs. Sediments were without any recognisable geomorphic modification and low organic and carbonate residues represented its spring dominated origins. Sediments were expelled onto the surface by spring water and deposited as water levels rise and fall. These light-coloured subterranean sediments were probably eroded from an underlying sandstone and shallow aquifer. 5.12 Chapter summary | 185 5.12 Chapter summary This chapter examined three case studies to explain site formation processes and climatic shifts that possibly affected the ecological baseline and hunter-gatherer populations along Eulo Ridge. Multi-proxy datasets revealed overland flow, stop/start erosion, and long phases of stability punctuated by brief periods of instability spanning approximately 55,000 years. These results set the scene for understanding foraging technologies and social relations among groups, and how people populated the semi-arid and arid interior. There is a possibility that people adapted to their local conditions by the use of technologies and localised knowledge of their environment (Hiscock and Wallis 2005: 49). This will be eCPLored further in the next chapter. Chapter 6 Investigating palaeoenvironments: a human occupation model Michael Connolly (Artist) If the Humanities teach anything, it is the diversity of the human cultural response and an ability to adapt innovatively to environmental conditions (Mulvaney 1990: 120). 188 | Investigating palaeoenvironments: a human occupation model 6.1 Introduction Chapter 6 synthesises the previous chapter to reveal the interplay between episodic landscape change and early peoples living near springs along Eulo Ridge. The first objective is to appraise and evaluate episodic landscape changes and its effects on peoples behaviours. A second objective is to examine proxies such as phytoliths and stone artefacts to evaluate the contribution to the overall research question. A third objective is to appraise when landscape changed over time and its likely effects on people. By understanding site formation processes and its timing, it could be possible to determine how early peoples along Eulo Ridge responded to climate variability throughout the Late Pleistocene to the Late Holocene. This chapter addresses the aim and objectives with six structured questions. What evidence explains punctuated climatic shifts? Why did these landscapes change over time? What do stone artefacts reveal about site formation processes? How have opal phytoliths and inorganics inform site interpretation? How has landscape change impacted on early peoples? The structure of this chapter follows the configuration of these questions. The thesis conclusion is presented at the end of this chapter and includes recommendations for future research. 6.2 What evidence explains episodic landscape change? This study revealed episodic change across the 3 case study areas. Unconformities and buried soils correlate with past climate conditions which are incremental over the past ~55,000 years. Sediment variations including textural pedofeatures align with the topography and distance to a slope and/or a spring(s). ~60 Ka to ~25 Ka (MIS 3) The basal layer of the Hardpan Creek North site suggests that the energy generated from an erosional event strips the landscape of its features, and transports the amalgam of materials downslope to its final resting place. While any preservation of living surfaces are compromised, the ensuing mix presents a time capsule to reconstruct the former surface and its general human use. A weathered stone artefacts embedded in this layer implied that people had been present during a humid phase. Hardpan Creek North site contains the oldest known sediment layer or unconformity along Eulo Ridge. An OSL date obtained from this basal layer, determined a maximum age of 55.9 Ka ± 5.9 Ka BP (WLL1354) (Wang 2019). The basal layer comprised laminated reddish brown loams and textural pedofeatures that imply a humid phase which transitioned into dry cold conditions. At Tunkana Well, thicker colluvial deposits were identified toward the profile base and included gravel and clay-rich sediments. Both basal layers infer that these formed as a result of short sharp humid event. These align with pre-glacial humid 6.2 What evidence explains episodic landscape change? | 189 conditions where active lakes and river systems dominated climate proxies across south eastern Australia (Bowler 1998; Cohen et al. 2012; Fitzsimmons et al. 2013; Hughes et al. 2017; Kemp et al. 2019). Mega-flooding in the Paroo River and/or continual groundwater discharge from artesian springs may also explain these conditions (cf. Bobrov et al. 1999; Macphail and Goldberg 2017). Phases of rainfall provided ideal conditions for people to extend their foraging range (cf. Bowler 1998). At a regional scale, humid conditions persisted within the Murray Darling Basin rivers carrying water more often, and air temperatures were 8-9 °C cooler than the present (Fitzsimmons et al. 2013). ~25 Ka to ~20 Ka (MIS2) (LGM) Basin Gully 1 site is represented by laminated deposits of alluvium indicated a settling out of sediments on the margin of a freshwater lake or large pool of water. An OSL date of 25.2 ± 2.6 Ka BP (WLL1356) (Wang 2018) suggested persisted until onset of the LGM. Evidence of sponge spicules and Elongate phytoliths suggested persistent freshwater conditions with grasses growing along the water’s edge. Clear skies and subzero temperatures caused frosts and cold conditions. These were followed by high summer temperatures and very high evaporation rates which formed calcium carbonate evaporites in the alluvium (cf. Schaetzl and Anderson 2005: 38). Intermittent water levels suggested frequented sub-tropical humid conditions and flooding in the upper Paroo River catchment. In contrast, spasmodic arid phases in the lower Darling River encouraged local dune building along the margins of freshwater ponds and swamps. All the while, water tables along Eulo Ridge fluctuated intermittently but the aquifers provided permanent water supplies for early peoples. It is implied that springs provided focal point for people to live within a home range of springs. At a broader regional scale (eastern Australia), the effectiveness of monsoons reduced significantly after 25 Ka BP and during the LGM (Fitzsimmons et al. 2013). Lakes dried out and dunes formed across Australia’s arid core with heightened dust transport (Bowler 2012; 1973; Fitzsimmons 2017; Fitzsimmons et al. 2013; Smith 2013). At the Willandra Lakes and along the upper MDB though, lake levels persisted as a result of seasonal snow melt and increased runoff in the highlands and, increased bedload in the upper Darling River and it is likely that the northern catchments of the Paroo allowed irregular stream flows (cf. Bowler 2012; Fitzsimmons et al. 2013). Therefore despite the LGM having a significant impact on the wider general landscape, springs and intermittent river flows appear a contributor to early peoples survival in these landscapes. ~20 Ka to ~15 Ka (MIS 2) The evidence at Basin Gully 2 site demonstrated a low grassed dune located beside an expanse of freshwater at the height of the LGM, 19.6 Ka ± 2.0 Ka (WLL1350) (Wang 2018). It is 190 | Investigating palaeoenvironments: a human occupation model (a) (b) (c) (d) Figure 6.1: Hardpan Creek North (90-97 cm) dusty clay coatings: a) to c) Crescentic pure and dusty clay coatings and lines of extinction (PPL) (cf. Stoops et al. 2018: 379-380); sequence of various illuvial events (cf. Fedoroff et al. 2018: 848); d) soil fragment (PPL) hypothesised that the palaeoenvironment comprised a pond margin with grass along the waters edge. Micro-charcoal in these deposits pointed to natural fires and possibly people living on a low dune near water during hyper-arid conditions. High proportions of the clay and silt fraction, and sponge spicules suggest that this material had blown on to the dune from the nearby water body (cf. Wallis 2001: 114). Illuviation and crescentic clays with organic infills indicated bare surfaces and vegetation clearance (Chapter 5, Fig. 5.22 and Fig. 6.1) (cf. Nicosia and Stoops 2017: 241). In addition, silt coatings on the coarse quartz grains implied cold overnight temperatures and sunny dry days. At a broader scale, Fitzsimmons et al. (2013) highlights that the LGM was characterised by influences from the monsoons and westerly influences with north-south variations in climate. Further, fire management was practised across eastern Australia when Europeans arrived and remained a common practice across the central and western deserts (Bird et al. 2012; Gamage 6.2 What evidence explains episodic landscape change? | 191 2011; Jones 1969; Kimber 1983). It is possible that fire was being used along Eulo Ridge, but this is unconfirmed due to factors that will outlined in the ’Future Work’ section below. ~15 Ka to ~12 Ka MIS 2–MIS 1 The Granites dune site demonstrated periods of instability that was replaced by stable conditions. The core site dated to 14.6 Ka ± 1.5 Ka BP (WLL1348) (Wang 2018). Shallow sediments were deposited by intermittent and increased precipitation. Humid conditions were likely reactivated by the south-moving monsoon and increased precipitation across the region (cf. Fitzsimmons et al. 2013: 92). This brought periods of colluvium deposition with episodic changes to the surrounding landscape. Sponge spicules and vivianite identified freshwater conditions and intermittent flooding. These erratic weather conditions aligned with the Younger Dryas’s across Australia (Denniston et al. 2013). Shallow horizontal layering strongly suggested a period of stability and short term settling out of sediments. High water tables and active springs dominated the Holocene with a subsequent drying out and dune-forming activity. At the Double and Tunkana Well sites, a sequence of deposits identified colluvium, and evidence of fluctuating groundwater tables. A 10 to 15 cm lens of gravel and stone artefacts set mid-way between the surface and the base of the arroyo defines a fast flowing erosive event (Fig. 6.2). This hill wash deposit formed by high intensity rainfall and mass movement of sediment. The erosive nature of this event stripped the upstream land surface of its features and suggested that this event occurred during the post-glacial phase. An OSL date was not taken for this site but it is implied that high intensity events of this nature likely occurred during a humid phase. After the LGM, an increased divergence between the monsoon and westerly influenced promoted instability across eastern Australia and heightened the possibilities of episodic humid phases (Fitzsimmons et al. 2013: 92). It is implied that this event coincides with landscape at Double Well. At Tunkana Well, a <30 cm deposit of Fe-rich sediments implied a colluvial event that stripped sediments from upstream and deposited them over a spring. These deposits imply mass movement of sediment from upstream and this aligns with other abrupt changes at Double Well and the other two study areas. In addition, micro-charcoal was found in the middle layer and suggested natural fires. As in all other sites, this also hinted at a human presence but small quantities of micro charcoal was inadequate to imply human-related fires. It is hypothesised that this event coincided with the monsoons that shifted southward around 14 Ka to 15 Ka BP or during a humid phase (cf. Fitzsimmons et al. 2013). The westerly winds encouraged rainfall, with active river flows along the MDB filling the lakes and rivers. This was followed by another dune building phase between 14 Ka-10 Ka BP, which aligns with stable conditions around 14.6 Ka BP at the Granites. 192 | Investigating palaeoenvironments: a human occupation model Figure 6.2: Double Well site: stone line, which dissects the arroyo has gravel and stone artefacts within the matrix ~12 Ka to ~5 Ka (MIS 1) The middle layer of Hardpan Creek North site revealed a humid phase interrupted a period of stability. This layer was dated to 11.5 Ka ± 1.2 Ka BP (WLL1346) (Wang 2018), which aligns with suggested aridification (cf. Fitzsimmons et al. 2013). Micro- structures in thin section suggested illuvial phases filling the voids with clays, while fractured crescentic clay coatings and displacement of clay coatings indicated a colluvial event followed this phase (Chapter 4, Fig. 5.21 and Fig. 6.3) (cf. Kuhn et al. 2018: 379-380). Micro-fossils identified freshwater environments with grasses, annuals, shrubs and trees Ficus (Fig tree) (Fig. 6.3) growing within close proximity to water. Fig trees are extant from the study region which implies post-Holocene landscape change. It is worth noting that in central Australia, fig trees grow near to small ponds and springs despite being in a region with much average lower rainfall on average than Eulo Ridge (Box et al. 2008). An environment of groundwater fed pools shaded by fig trees does not seem unreasonable in this region of eastern Australia. Unstable conditions around 11.2 Ka BP caused landslides and the deposition of colluvium in the lower slopes and plains. The Hardpan Creek North terrace provides evidence of these events which appeared to dominate the areas along the margins of Hoods Range. 6.2 What evidence explains episodic landscape change? | 193 (a) (b) Figure 6.3: Hardpan Creek North (45 cm, OSL) fabric pedofeatures: a) juxtaposed crescentic clay coatings with extinction bands and phases of clay illuviation (PPL) (cf. Kuhn et al. 2018: 379-380); b) degraded clay coatings (PPL) ~5 Ka to present (late MIS 1) At Hardpan Creek South site, a brownish black sandy loam deposit contained plant material, opal phytoliths, micro-charcoal, and insect activity which demonstrates an expanse of freshwater that was maintained by groundwater and/or humid conditions. This feature formed 4.6 Ka ± 0.4 Ka BP (WLL1345) (Wang 2018) and aligns with a humid phase across eastern Australia (cf. Fitzsimmons et al. 2013). A peanut shaped micro-fossil could represent a lake or freshwater swamp but evidence is needed to confirm this inference (Chapter 5, Fig. 5.33 and Fig. 6.4) (cf. Round et al. 1990). Or a bilobate phytolith could also infer it belonged to Themedia spp. or kangaroo grass that had washed or blown into this deposit. If so, this suggests that the spring was located in an open grassland (cf. Bowdery 1998). Further, micro-charcoal found in this layer suggests either natural fires or firestick farming practices in and around the springs (cf. Bird et al. 2012; Jones 1969; Kimber 1983). As stated throughout this chapter, these are hypothetical suggestions that are not confirmed because of the limitations with data analysis. During the Early to Middle Holocene, a landslide buried the western portion of the Granites phreatic zone. A low dune formed on the valley edge and which eventually a dune formed in the valley. Landslides and the burial of land surfaces had a significant impacts on the landscape changes at the Granites. The evidence suggested that people camped on this dune throughout the Late Holocene but because of landscape change they were forced to develop technologies and practices to meet the demands of the new environment. 194 | Investigating palaeoenvironments: a human occupation model Figure 6.4: Hardpan Creek South (OSL) phytolith or diatom: Bilobate opal phytolith (Themeda spp.) or peanut-shaped freshwater diatom On the eastern side of Hoods Range, a sequence of alluvial deposits was overlain with colluvium and was due to significant heavy rainfall that waterlogged and transported sediments to the toe slope. Downslope movement of sediments placed a thick layer of sediment over the former phreatic landscape. Contained within these sediments were varves which represent an annual layer of waterlogged sediments that in this instance appears to be deformed by post deposition (Chapter 5, Fig. 5.19 and Fig. 6.5) (cf. van der Meer and Menzies 2011: 224-225). Varves are the smallest known evidence of a climatic event and are usually associated with rhythmites which are laid down with constant regularity along the margin of lake. Varves also only form in fresh or brackish water and under anoxic conditions and frozen surface water (Thunell et al. 1995). Finding varves in this layer suggests significant landscape change and post depositional movement of sediments from upstream. It is clear that sediment has been transported downstream and deposited on the former Late Pleistocene and Early Holocene surface. Episodic change appears to interrupt stable environments and demonstrates the irregularity of short sharp episodic events. Dune activity increased around 5 Ka BP, which aligns with dune formation and burial of the phreatic zone at the Granites. The stable dune located west of the Granites Relict Dune site identified that this dune had remained stable and unchanged for some time. The dune is stabilised by trees, shrubs, and grasses but the sands remain locally mobile. This dune building phase was interrupted by a humid phase about 4.6 Ka. A buried soil located at the Hardpan Creek South and Centre site identified humid conditions and a productive landscape with grasses. Micro charcoal found in this feature suggest natural or human fires. But given that the site is located so close to water, it is implied that early peoples were burning the landscape around these springs. 6.2 What evidence explains episodic landscape change? | 195 (a) (b) Figure 6.5: Hardpan Creek North (2-9 cm): a) and b) varves with graded couplets and micro faulting (flatbed scanner, natural light) (cf. van der Meer and Menzies 2011: 224-225) Figure 6.6: Hardpan Creek South site: mulga soils overlie this buried terrace (40 cm and 2 m scales) The humid event was followed by the deposition of mulga soils on the terraces and buried soils within the last 700 years. Silt cappings on the coarse quartz grains suggested frosty conditions (cf. Van Vliet-Lanoe and Fox 2018). One scenario could be bare surfaces as a result of aridity, which was followed by intense rainfall that saturated soils and caused major landslides along the toe slopes of Hoods Range. This was followed by dry spells and subsequent humid climate that resulted in further colluvial/alluvial events and the spread of sediments across the toe slopes. This could well have been the beginning of ENSO and the sees awing humid and arid phases that often account for boom and bust weather conditions across inland regions of Australia. 196 | Investigating palaeoenvironments: a human occupation model As discussed in Chapter 2, weather variants such as ’easterly troughs’ deepen and bring unstable air masses and with them, rainfall to the inland (BOM 2014) This is usually followed by ’blocking highs’ that form in the southern half of the continent and push low pressure off-shore and encourage dry conditions across inland regions. These highs will often remain stationary and promote dry air masses and sometimes strong winds. In addition, a belt of high pressure that encircles the globe in middle latitudes creates a sub-tropical ridge, which moves north and south along central latitudes of the continent. On its southerly track, it brings moist air masses and rainfall to the southern half of the continent, and on its northerly movement, it brings dry conditions to southern Australia. El Niño Southern Oscillation (ENSO) effects the current climate. As discussed in Chapter 2, ocean temperature warming in the central and eastern Pacific Ocean leads to dry conditions in eastern Australia. Low rainfall and dry conditions persists for up to 10 years causing droughts and major land degradation across much of the inland (BOM 2014). Conversely, La Niña conditions sometimes promotes the movement of tropical lows and low pressure systems across inland regions which causes flooding in the MDB rivers. These heightened conditions can vary between districts and in some circumstances rainfall amounts can surpass long term climate averages. It is highly likely that ENSO conditions influenced weather patterns along Eulo Ridge in the past. As noted in this thesis, punctuated conditions appear the norm rather than being unusual and it is implied that ENSO contributed to landscape change. About 0.7 Ka ± 0.1 Ka BP (WLL1347) (Wang 2018), overland flow caused the subsidence of mulga soils to bury the former terraces at Hardpan Creek North and South. An analysis of these sediments reveals colluvium which is implied from micro-layers and soil nodules contained within the thin section groundmass (cf. Nicosia and Stoops 2017: 241). Colluvial activity was prompted by large weather systems and flooding in the Paroo River and rivers across eastern Australia (cf. Fitzsimmons et al. 2013). Opal phytoliths were abundant in these deposits and represented local grasses, shrubs, and tree genera (Appendix C, Fig. C.7). Within the last 1,000 years, early people used adzes and made composite tools which coincided with the ENSO-dominated environment (Smith 2013: 209). Further, population controls and residential mobility led to early peoples occupying vast areas of the inland (cf. Cane 1987; Gould 1968; 1971). 6.3 Why did these landscapes change over time? The evidence presented above implied that episodic landscape change shaped the environment within close proximity to springs and Hoods Range. Formation of unconformities and buried soils derived from spasmodic episodes of colluvium, alluvium, illuvial, phreatic events, and period deposition of sediments along the margin of a body of water such as pool or lake. In 6.3 Why did these landscapes change over time? | 197 other words, climatic uncertainty shaped these landscapes and modified the types of resources that were available to early peoples. As described in Chapter 5, mulga soils buried or partially blanketed former terraces and buried soils along the foot slopes of Hoods Range. It was inferred that mass movement of sediments buried older land surfaces and stripped the landscape of its features and archaeological material. As mentioned above, burial of the Hardpan Creek North and South sites occurred about 700 years BP. This has had implications for landscape modification and destruction of former cultural areas. Its erosive energy destroys the organics and microscopic residues, and disperses cultural features within its path. Rates of colluvial movement is unknown. But it is likely rates of sediment/soil movement was dependent on each climate event, topography, and environment including ground cover. The final resting place of these sediments contains a record of the this event and the former landscape. Soils and the former landscape control preservation and mixing of the sediments. In post-depositional environment, chemical reactions effect the sediments and materials. By understanding site formation processes it is possible to reconstruct the palaeo-human landscape in regions that are not well known for their site integrity. An accumulation of cobbles, gravel, and sediments in distinct layers highlighted fast moving water from a flood or high intensity rainfall over a short period of time. Erosion and episodic change indicated sites located in small valleys on the lower slopes within 3 km of Hoods Range changed because of colluvial and alluvial aggradation, channel incision or gully erosion, and accumulations of these to form a major terrace (cf. French 2017). Processes and landscape types are identified from sediments sourced from nearby slopes, floodplains, and localised phreatic zones. Alluvial events are identified by lenticular layers and are much more subtle and take place over differing periods of time. The Granites East site demonstrated links between phreatic events and site formation processes. Thick deposits of clay-silt sediments derive from spring eruptions which dispersed subterranean sediments onto the surface. Sediments naturally settle by precipitation and likely runoff from the slopes of Hood Range. An arid phase resulted in a drying out of the deposits which formed CaCO3 precipitates in the sediments. Additional clues to site formation processes were observed in heat retainers at the Granites East site. Intact heat retainer peds provided clues to the intensity of surface deflation and erosion. It is implied that conditions were gentle and slow. The original surface above the heat retainer had been slowly removed and over time, small amounts of sediments are removed to expose the layer of stone that was laid at the bottom of the former pit. This process took place over hundreds and possibly thousands of years. They thus present a proxy for site formation and ground disturbance. Five heat retainers were found within close proximity to the Granites East study site. The evidence implied that these heat retainers had been gradually exposed by 198 | Investigating palaeoenvironments: a human occupation model low energy surface deflation and suggested stable environments without major interruption or mass movement of sediments. In other sites such as Hardpan Creek, Basin Gully, and Double and Tunkana Well, an absence of heat retainers suggested high energy erosion had dispersed any evidence of these fireplaces. It was implied that overland flow and sheet wash dispersed these features into nearby channels and gullies and subsequently intermittent stream flows have washed sediments and cultural features downstream. This possibly accounts for micro charcoal being washed into the deposits from afar. Strong winds and sheet wash are also likely processes to transport micro charcoal into the deposits. Finding charcoal in association with stone artefacts near permanent water sources implies human made fires rather than natural. But as will presented in the future research section, experimental and geoarchaeological research is needed to test these findings. During the pastoral period, surface deflation, rilling, and gullying stripped the A and H horizons, and exposed the B horizons and buried soils and terraces (Fanning 1999; QDPI 1974). Topography, sediment/soil characteristics, and climatic instability/stability accounts for site formation along Eulo Ridge. Climate intensity and rate of erosion were dependent on catchment size and the abatement of individual soils. It is implied that significant amounts of rainfall within small valleys resulted in the mass movement of sediments downslope to the lowest disposal point in the landscape. Storm intensity, slope, parent materials, and soil strength had the potential to affect site formation processes in the three case study areas (cf. French 2003). Each one of these categories affected sites and features in the case study areas. Mass movement of sediments downslope buried all relict terraces and soils. 6.4 What do stone artefacts reveal about site formation processes? Stone artefacts found embedded within arroyos had the potential to pose questions about site formation processes and, how this relates to past occupation of Eulo Ridge? Some doubt was cast over hypothesis with Rhodes et al. (2009) reporting an absence of stone artefacts in sand or silt-dominated deposits in similar environments to the south of Eulo Ridge. But it was demonstrated here that stone artefacts concealed within sandy loams, loams, and sandy clay loam of arroyos. This study found a correlation between both high and low energy environments and the burial of stone artefacts. Within high energy environments, colluvial processes strip the landscape of its natural and cultural features, and lithics downslope hundreds of metres away from its original position (cf. Morton 2016). Among these residuals were microscopic plant remains such opal phytoliths and other micro fossils which had been incorporated into the colluvium and alluvium as its mobilised and deposited elsewhere. As a result, it is implied that stone 6.4 What do stone artefacts reveal about site formation processes? | 199 artefacts were incorporated into feature as energy decreased and sediment bedload settled in its new location. As time since burial increases, the assemblages are cemented into the matrix through chemical changes. Sediment dispersal requires moderate to high energy environments with significant sheet wash and/or mass movement stripping the landscape of the A, H horizons. As energy increases, stone artefacts and materials are buried by further sedimentation. Stone artefacts become entombed and cemented by chemical reactions to conceal the final resting place of stone artefacts (Fig. 6.7). A cream coloured cryptocrystalline flake found in the basal layer of Basin Gully 1 site suggested acquisition of stone artefacts from a quarry (a large stone quarry is located on the western side of Hoods Range). It is implied that early peoples were utilising these raw materials within at least a 4 Km radius of the main quarry. During the LGM, arid conditions controlled the movement of peoples and with the certainty of water within springs, people lived within a home range that included the extend of spring water and access to resources. Trading materials were also limited during the LGM with less people moving any great distance and thus a reliance on local materials and resources. It is implied that Eulo Ridge functioned as a human refugia with access to water, food, and materials that existed across a diverse landscape including variations in topography, access to lakes and intermittent flows within the Paroo River. A high probability of stone artefact being found buried in unconformities exists where stone artefacts are buried deep within a profile. Low energy processes are likely to be buried by post-deposition and sediments covering stone artefacts after discard. Whereas lag deposits correlate with cemented weathered surface sediments and poor spatial and temporal resolution. Depth of stone artefact burial and sediment texture are strong indicators of specific climatic events and post depositional movement of stone artefacts through colluvial and alluvial processes (Fig. 6.7). By around ~14 Ka BP, stone artefacts were incorporated into the base of the relict dune. A subsequent dry phase dessicated the sediments and cemented stone artefacts into Static layer. At the Granites study area, stone artefacts embedded in sediments on various angles provided a proxy for episodic spring eruptions. Stone artefacts embedded in sediments on various angles reflected the intensity and processes involved in these deposits. The episodic displacement of artefacts during landscape change events strongly correlates with the former climate (cf. Butzer 1973: 315). The immediate burial of stone artefacts is dependent on scale and energy contained within each event. The timing of these events are not known and little is known about the effects of spring energy as a mechanism for stone artefact burial. Stone artefact numbers within these deposits were expectedly limited due to smaller populations and the nature of archaeological deposits within drylands. Correlations tend to exist between stone artefact numbers and human populations. Stone artefacts dispersed throughout the deposit and only small number visible from the face of an arroyo. For this to 200 | Investigating palaeoenvironments: a human occupation model Figure 6.7: A model for site formation processes and the dispersal and burial of stone artefacts within arroyos occur, stone artefacts had to have been transported by overland flows and/or mass movement and deposited downstream as water energy dropped and sediment ceased to flow (cf. French 2017). 6.5 How have opal phytoliths and inorganics informed site interpretation? At the outset of this study, there was no expectation to find opal phytoliths within thin section. Phytoliths are known not to survive in alkaline environments such as those that have been identified in this study. Phytoliths are known to dissolve phytoliths (cf. Cabanesa et al. 2011) but this study has confirmed highly alkaline with pH values of >9 do in fact provide suitable conditions to investigate the palaeo-vegetation of these palaeoenvironments. Sediment alkalinity did not effect preservation and phytoliths allow an insight into the past. But also, an absence of phytoliths in some contexts could imply poor preservation or it may suggest arid environments with bare soils or alkaline soils found within the dune systems (cf. Cabane and Shahack-Gross 2014; Madella and Lancelotti 2012). Opal phytoliths identified in thin section helped to reconstruct the palaeo-vegetation across the case study areas. Morphotypes imply an array of plants possibly existed in various landscapes over time. As noted by Gao et al. (2018: 764), opal phytoliths, "can reliably 6.5 How have opal phytoliths and inorganics informed site interpretation? | 201 differentiate samples from herbaceous and woody communities, and samples from Poaceae and non-Poaceae communities". Further, inorganics such as sponge spicules, diatoms, and fungi micro-fossils presented an opportunity to provide certainty about freshwater and marine waters (cf. Racek 1969). Small isolated clusters of phytoliths and inorganics were commonly found in clay features and along void margins. It is suggested that mass movement on the floodplains and the lower slopes incorporated phytoliths into the clays which in turn were transported in suspension. In other instances, wind and water transport had redeposited grass residues onto surfaces, or phytoliths were moved and deposited in the faecal pellets of herbivores such as kangaroos (cf. Wallis 2001). A further hypothesis is that early peoples processed grass seeds on the edge of this water source. Opal phytoliths helped to broadly place palaeo-vegetation types along Eulo Ridge in the Poaceae and Cyperaceae families (cf. Neumann et al. 2018: 2). Acute bulbosus, which is the most common diagnostic for grasses and sedges were found in upper and middle layers at Basin Gully 2, Double Well and, Granites sites. These represent a an open grassed environment or sedges growing near the waters edge during the Middle to Late Holocene. Elongates and Acute bulbosus morphotypes are typically consistent across all layers and imply a grass dominated environment. A honeycomb phytolith found in layer HCKN-OSL probably belonged to the Moraceae Family, a Ficus spp. or fig (Chapter 5, Fig. 5.20, and Fig. 6.8). Ficus spp. are endemic to inland Australia and grow near permanent water sources such as springs (Box et al. 2008: 1404). The sediments around this phytolith dated to approximately 11.5 Ka and suggest that conditions during the Late Pleistocene were different to the current mulga dominate landscapes. These findings have implications for interpreting vegetation change and transitions from fig trees growing near water to a homogenous mulga woodlands environment. Cavate phytoliths imply that conifers or Araucaria existed in the upper Boorara Creek area (Appendix C, Fig. C.21(h) and Fig. 6.8) at the base of Double Well site (162 cm). Evidence of large Gymnosperms indicate humid conditions and a new landscape from an earlier Epoch (cf. Parr and Watson 2007). If this is correct, it poses a new phase of inquiry regarding the types of research that could inform landscape change in drylands. Burned phytoliths found throughout the upper to middle layers of Granites East, Hardpan Creek North and South, and Double Well sites. These results suggest either good preservation within the more recent sediments and/or either human and/or natural fires. Similar problems occurs with micro-charcoal because it is difficult to confidently diagnose and determine the causes and origins of previous fires. This problem presents an issue for further research. More confidence in these results could be obtained from sediments with rubification and/or near a hearth with charcoal and ash deposits (Mentzer 2014). 202 | Investigating palaeoenvironments: a human occupation model (a) (b) Figure 6.8: Phytoliths, a) possible Ficus found in HPCKN OSL layer (PPL, black and white); b) possible Araucaria phytolith found in Double Well 162 cm layer (see Parr and Watson (2007) for a comparison) (a) (b) (c) Figure 6.9: Hardpan Creek North (97 cm): a) to c) fungi micro fossils. Unable to confirm the environmental conditions for these micro fossils In addition, inorganics such as diatoms, sponge spicules and unusual fungal micro fossils provided evidence of freshwater conditions and diagnostic features of the former humid and/spring environments (cf. Bobrov et al. 1999; Fitzsimmons et al. 2013; Racek 1969; Round et al. 1990). Fungi often grows in moist conditions but it was not possible to verify the habitat (Fig. 6.9) (cf. Douglass 1988). Swamps and/or pools of water maintained by springs provided sufficient wet conditions to encourage a diversity of plants. Thus variation in fungus micro fossils (Fig. 6.9) adds further weight to vegetation change and a diverse ecological environment. In sum, the extent of change and variance in vegetation types within the former environments requires a more focused look at landscape change. New methods that target phytoliths in the study area are needed to reveal more about Australia’s drylands. small numbers of phytoliths and inorganics limited the interpretation of vegetation communities and/or freshwater or marine environments. These findings however, provide a methodology to interpret human dryland environments at a broad scale (cf. Devos and Vrydaghs 2011; Devos et al. 2009). 6.6 How has landscape change impacted on early peoples? | 203 These results offer an introduction to the use and power of micro fossils to inform these types of studies. Phytoliths within thin section help broaden our knowledge of the past and if done well, phytoliths can make a valid contribution to understanding palaeoenvironments and other behaviours such as evaluating human resources such as seed processing and food types. As suggested in the Future Research section below, more research is needed to help improve these methods for Australian sites and to create and maintain an archaeological database for phytolith morphotypes that are indicated of landscape across these regions. Adding OSL dating to these methods could maintain chronological control to the deposits and landscape change over time. 6.6 How has landscape change impacted on early peoples? Landscape change along Eulo Ridge impacted on early peoples during episodic landscape change. Both arid and humid phases supported people over long periods of time. But in response to dry conditions, early peoples abandoned waterless areas to occupy the springs along Eulo Ridge. They responded to change with adjustments to their technologies and social systems, and they managed populations to ensure their survival in all periods of climate change. Life in marginal landscapes required behavioural change to counteract the impacts of climate and its effects on surface water supplies and dwindling resources (cf. Cane 1995; Gould 1968). To early peoples advantage, endemic plants and animals had adapted to the changing conditions and provided ample resources during humid and arid phases. As described in Chapter 3, early peoples controlled populations with birth control and increase ceremonies helped to stimulate environmental productivity when resources were low (David and Thomas 2016: 16). Managing the landscapes with patch mosaic burning helped to control biodiversity and in turn encourage plant diversity at small scales (Bird et al. 2012). People developed techniques to manage a marginal landscape to ensure its productivity and stability. As David and Thomas (2016) point out, the environment adapted to the people and not the other way around. During late MIS 3 and before the LGM, 55 Ka to 24 Ka, humid conditions transitioned to a dry phase with clear skies and frosty conditions. The initial humid phase was dominated high intensity rainfall and flooding which effected Eulo Ridge and large amounts of sediment was transport downslope from Hood Range. Thus, these humid conditions extended their geographical home range and contributed to a surge in population growth. Further, it allowed people interact with neighbouring language speakers to the north and south, east and south to the Warrego and Darling rivers, and west and north to the Bulloo River peoples. The premise is that they withstood these extended dry periods by implementing a richer culture with deeply etched mythologies and ceremonial song cycles (cf. Smith 2013: 10). A strong culture helps to 204 | Investigating palaeoenvironments: a human occupation model sustainably manage an array of resources over long periods of time. Sharing resources through trade and reciprocity functioned as a means to maximise human survival across large tracts of lands (Ingold 2000; Kelly 2013). Previous research suggests that significant modifications to productive landscapes forced an expansion of new technologies and social change among the groups (cf. Lourandos 1997; Smith 2013). During the LGM or from 24 Ka to ~14 Ka, life for early peoples changed dramatically. To survive, it is hypothesised that lived off the endemic species and continually moved between springs and reduced the size of their home range to move between springs. Movement included use of riverine waterholes of the Paroo River, and lakes and swamps until surface water and resources had disappeared. These behaviours were similar to those that are well known among the desert peoples in Central Australia which were studied in greater detail through ethnographic research (cf. Cane 1987; Gould 1968). The deglacial phase however, ~15 Ka BP to ~12 Ka BP saw the monsoons shift southward (cf. Fitzsimmons et al. 2013). Intense rainfall recharged the river systems, lakes, and groundwater systems and triggered a new phase of life along Eulo Ridge (cf. Smith 2013: 160). A stone line filled with gravel and occurrences of stone artefacts, and ample amounts of micro charcoal identifies a high energy environment that stripped the landscape of its natural and cultural features. These episodic changes resulted in people moving temporarily to rockshelters, the outer plains of the Paroo River, and the margin of the nearby lakes. People adjusted to these dynamic landscapes with new tools and changes in their social structure and organisation (Smith 2013). In addition, these created grass-rich environments and people were able to take advantage of increased grass biomass to support larger populations. Population increases promoted cultural and behavioural change and stability among the groups. These humid conditions transitioned into an oscillating humid and dry periods through the Middle to Late Holocene (cf. Fitzsimmons et al. 2013). Buried soils at Hardpan Creek South comprised micro charcoal, grass phytoliths, and dark organic rich sediments. The evidence implied that people were semi permanently living near springs and had ample resources to support larger populations. Humid conditions persisted around 5 Ka BP to 4 Ka BP and reliable rainfall encouraged people to harvest seed and to burn the surrounding landscape. Fire stimulated plant growth and helped to bring animals closer to their home range and in doing so, extend the extent of their available resources (cf. Bird et al. 2012; Gamage 2011; Jones 1969). From 2 Ka to 0.6 Ka, summer monsoons dominated the environment. Intense rainfall events moved large amounts of sandy loams and loams downslope to bury the former terraces. A transformation of the landscape resulted in significant changes to food resource areas and surface water. Peoples required access to new tools and technologies to cope with poorer soils of the new landscape. Previous research demonstrates that the Late Holocene developed extensive trade and exchange networks to transport stone and wooden tools, and organic 6.7 Conclusions | 205 materials from southern and northern Australia to inland regions (cf. Smith 2013: 297). Archaeologists argue that the Late Holocene was the ’intensification’ period (Lourandos 1993). Larger populations required more materials and new information to cope with significant change. The irregular boom and bust phases modified the landscape and people had to adjust their behaviours to meet the ensuing challenges of landscape change (Smith 2013). An unchanging landscape and culture needs further scrutiny to better understand early peoples and their behaviours. 6.7 Conclusions The aim of this study was to answer the question, to what extent has continental and global climate change influenced the behaviours of early peoples living near reliable water sources along Eulo Ridge (eastern Australia)? This study built on the previous research undertaken by Richard Robins (1993; 1998) and soil micromorphology studies undertaken in Britain and Europe to reconstruct human landscapes (cf. Devos and Vrydaghs 2011; French 2003; 2015; Goldberg and Macphail 2006; Macphail et al. 2006). This multi proxy approach aimed to examine soils and/or sediment features exposed by gully erosion (arroyos) near springs to reveal what caused episodic landscape change along Eulo Ridge (cf. French 2015). We know that early peoples abandoned drylands during hyper arid conditions (cf. Veth 1989; 2005) but it is unclear how people living near springs reacted to climate interruptions. The central motivation of this study stems from this previous geoarchaeological research that focused on natural accumulation of horizontal layers within arroyos (cf. Butzer et al. 2008; French 2003; 2015; Goldberg 1980; Goldberg and Macphail 2006; Macphail and Goldberg 2017; Waters 1991). It is hypothesised that arroyos build up at variable rates over time and thus conceal episodic events and the behaviours of early peoples living near water during the late Pleistocene and Holocene epochs. During the Quaternary, climate oscillated between humid and dry phases and periods of stability were interrupted by channel change and eventual backfilling of the deposits (Waters 1991: 144). Previous research in the study region shows that within the valleys, fluvial activity has exposed multiple archaeological surfaces that are of different ages and stacked one on top of the other (Rhodes et al. 2009: 191). Once top soils are removed, subsoils are highly erodible but may remain relatively stable for long periods of time which is then interrupted by rapid episodic change. Pastoralism in semi arid Queensland and northwest New South Wales caused extensive land degradation and deflation of loams and sandy loam soils to expose the B Horizon in some instances (Fanning 1999; QDPI 1974). These conditions made it ideal to undertake this study and to obtain knowledge about the occupation of Eulo Ridge. The multi proxy approach used here deciphered sediment/soil features and provided a chronology for various colluvial, alluvial, and illuvial processes. Landscape stability was 206 | Investigating palaeoenvironments: a human occupation model interrupted by erosion and aggradation at observable time intervals which aligned processes in the region with a punctuated equilibrium (cf. Eldredge and Gould 1972; French 2003). It was established that multiple arroyo deposits represent episodic landscape change which took place from MIS 3, ~55,000 years BP to MIS 1, up to 600 years BP. Stone artefacts played a vital role in understanding links between the timing of episodic landscape change and the presence of early peoples. Stone artefacts buried within unconformities appears to strongly correlate with each disruptive sediment transportation event. The final resting place of stone artefacts were indicative of the depositional energy and cementation processes operating in these local environments over time. Stone artefacts embedded within sand and silt rich sediments implied people had occupied these landscapes prior to a major climate event. The mass movement of colluvium and alluvium was indicative of episodic climate events and each phase created new challenges for early peoples. Humid conditions supported larger populations and a socially dynamic culture focused on adaptation to landscape change. Major climate variations implied that springs are durable nodes for long term occupation, and focal points for people to cope with aridity. Springs became the focal points for occupation and a continuous supply of freshwater provided reliable water sources within relative proximity to an array of resources. It is concluded that the Eulo Ridge springs functioned as a human refugia for ~50 millennia. The implications of landscape change sometimes included being tethered to springs, reductions in food and water resources, and technological change to cope with the new environment. The findings identified that high energy colluvium strips the sediments and human-related materials from the slopes and deposits sediments, organics and cultural materials as it loses energy, and provides a sedimentary time capsule of the past. Deposition appeared immediate, final, and unaffected by post depositional processes including bioturbation. Of note, significant climate events occurred ~55 Ka, ~24 Ka to ~19 Ka, ~14 Ka to ~11 Ka, ~4.6 Ka, and ~1 Ka to ~0.6 Ka to reveal episodic deposition of colluvium, alluvium, illuvial processes, and phreatic events. Of note, a major colluvial event occurred between 1,000- and 700-years BP which significantly changed the Eulo Ridge to a marginal landscape dominated by long dry periods and short sharp humid phases. Colluvium or hillwash blanketed the lower slopes, floodplain margins, and former alluvial and phreatic land surfaces. Burial of the former landscape features that provided water and food resources resulted in a less productive environment. These changes forced early peoples to modify their behaviours and adjust to a new marginal landscape which became more unpredictable. Boom and bust climates dominated the region which continues through to the present. 6.8 Future Research | 207 Opal phytoliths broadly placed the palaeo-vegetation in the Poaceae and Cyperaceae families. The results imply that vegetation changed over time and in some cases plant assemblages were replaced following landscape change. An array of phytoliths contained within the lower sediment units demonstrate the presence of species from the former landscapes. An appearance of honeycomb and cavate morphotypes imply extinct tree families are likely to have been present in the region. While there is some hesitance to reconstruct palaeoenvironments with a small number of phytoliths, a presence of morphotypes highlight the importance and power of micro fossils to help decipher the past. It is concluded that environmental and temporal proxies offer several analytical methods and techniques to address archaeological research questions in Australia. In particular, the arrangement of particles and pedofeatures help to reconstruct the human environment and in some cases, their behaviours and response to episodic landscape change. Micromorphology is reliable and repeatable, and when combined with OSL dating and archaeological data, it possible to refine interactions between continental and global climate change and the behaviours of early peoples. These outcomes present an opportunity for geoarchaeologists to better understand site formation processes in dryland environments across Australia. It is recommended that future research carefully considers arroyos and a multi proxy approach as a reliable environment and methodology to reveal more about the early peoples of Australia. 6.8 Future Research One final question is, where to from here? The outcomes of this study poses further questions concerning the significance of punctuated landscape change and its correlations with the early peoples of Australian drylands. Variables such as proximity to permanent water sources, deep arroyos with stratified features will be critical to new investigations. Discrete valleys, lake bordering dunes, and alluvial plains feature as prominent features to gather evidence of when and why people had populated vast areas of drylands. The multi-proxy research approach in this study helped to link palaeoenvironments with early peoples. Going forward, inland regions of Australia require a greater focus on archaeological sites using micromorphology and dating techniques. More work is needed to develop a comprehensive human behavioural model for inland Australia. A series of considerations should include the investigations of stratified open site, physical and chemical properties of soils, characteristics of stone artefacts and cultural materials, and securing a chronology for discrete areas. Soil horizons and/or unconformities offer a unique opportunity to better understand interactions between changing climates and early peoples. Micromorphology is seldom used in Australia to address questions about the past. More work should be done to apply it across a wide range of environments and chronologies. 208 | Investigating palaeoenvironments: a human occupation model But micromorphology is not without its limitations. The use of optical microscopy limits soil analysis and therefore future work might use scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDS) to undertake a targeted analysis of the thin section surfaces. This would allow greater surface analysis such as particle size analysis, a presence of organics, and other unknown materials. In addition, Fourier Transform Infrared Spectroscopy (FTIR) could also be used to identify particles, fibres, or liquids to reveal more about archaeological sites. Finally, it is worth pointing out that the thin section analysis technology is improving and thus new research should consider new and improved techniques to examine thin sections in detail and with ease. Further, a known absence of micromorphological studies in archaeology across eastern Australia has restricted the interpretations of archaeological sites. Going forward, comparisons should be made with a wider range of stratified archaeological sites to characterise human and environment related signatures. A known stratified feature such as the European dark earths would greatly assist archaeologist to better understand human and palaeoenvironmental studies. Micromorphological research could help to better understand commonalities that occur in multiple dryland sites. More phytolith research could help assemble a phytolith database to better understand early peoples landscapes from 60,000 y BP to the Late Holocene. Knowledge of palaeo-vegetation could help reconstruct past environments and contribute to a model of early peoples behaviours in all regions of the continent. More research is needed to determine rates of decay for organic and bone buried in highly alkaline soils verse those in anaerobic conditions. A strategic focus should investigate human behaviours such as Yandying, food discard and activities associated with early peoples. A behavioural-evidence-based databases could be developed for use by geoarchaeologists and micromorphologists. Early peoples fire use in Australia are not well understood. In this study, problems arose because it is difficult to categorise natural and/or human made fires. Micro charcoal and ash appeared the same and without diagnostics which makes it difficult to determine the origins of residues (but see Mentzer 2014). What is needed are experimental research studies that describe residues at this minute scale, for example, similar to those carried out by Mentzer (2014). This focus should examine fire use at hearth and landscape scales, and assign these outcomes to cultural/language group areas and traits. Kimber (1983) recorded large number of fire uses for Warlpiri and Pintubi people across the western Desert. As Mentzer (2014) points out, fire related evidence found in micro-archaeological situations helps to reconstruct human intensification behaviours. Micro-charcoal found within archaeological sites is dependent on fuel type and preservation conditions, and methods of combustion. An 6.8 Future Research | 209 understanding of the micro- archaeology of fire use is crucial if we are to understand the earliest phases of fire use in a range of palaeoenvironments. OSL dating developed a chronology for the Eulo Ridge and demonstrated that human occupation extends to at least 55,000 years ago. More work is needed to locate and date stratified archaeological sites to develop a better understand the value of open air sites in Australia. We need to know when people lived across these vast dryland landscapes to develop a detailed occupation model. Gaps appeared in the results because it was not possible to include all layers and features in this study. Future research opportunities should carefully examine all aspects of the topographical and ecological landscape to obtain higher resolution datasets. The refinement of these investigations will help fine tune methods and techniques to allow better outcomes for new geoarchaeological research. It is anticipated that this study should be discussed with the current Budjiti Elders to identify how these outcomes should be communicated to everyone. Ultimately, the objective would be to teach everyone about their homelands and help better understand the surroundings experienced by their ancestors. It is anticipated that the outcomes of this study could contribute to cultural knowledge of the broader region. A final point is that modern climate indices are a poor guide to understanding desert and marginal areas of the past (Smith 2013: 5). Future research should consider site formation processes and the chronology of sites before assumptions are made about deep human history of Australia. 210 | Investigating palaeoenvironments: a human occupation model Post scriptum Readers of this thesis need to be aware that the author is of Australian Aboriginal descent and any biases noticed in the text are due to his cultural beliefs and attitudes toward the well being of Aboriginal families now and in the future. These views may not always be in line with European constructs but where necessary attempts are made to insert the Aboriginal voice among these discussions. A final note is that this thesis was completed in Australia during the Covid-19 pandemic, 2020-2021 inconclusively, which interrupted my laboratory work and micromorphological data collection. But this did not effect the final outcomes of this thesis. The discussion above was completed in the spirit of Australian archaeology and its diversity of landscapes and arguments about people. I have enjoyed this process and I hope that this thesis makes a small contribution to the future of Aboriginal and Torres Strait Islander communities. References Akerman, K. (1994), ‘Riji and Jakuli: Kimberley pearl shell in Aboriginal Australia’, Monograph. Northern Territory Museum of Arts and Science. Allaby, M. (2015), A Dictionary of Ecology (5th Edition), Oxford University Press. Allen, H. (1972), Where the crow flies backwards: man and land in the Darling Basin, Unpublished PhD Thesis, Australian National University. Allen, H. and Holdaway, S. (2009), ‘The archaeology of Mungo and the Willandra Lakes: looking back, looking forward.’, Archaeology in Oceania 44, 96–106. Atkinson, J. (2002), Trauma trials, recreating song lines: the transgeneration effects of trauma in Indigenous Australia, Melbourne: Spinifex Press. Balme, J. (1991), ‘The antiquity of grinding stones in semi-arid western New South Wales.’, Australian Archaeology 32, 3–9. Barrows, T. T., Stone, J. O., Fifield, L. K. and Cresswell, R. G. (2002), ‘The timing of the Last Glacial Maximum in Australia’, Quaternary Science Reviews 21(1), 159–173. URL: http://www.sciencedirect.com/science/article/pii/S0277379101001093 Bastin, G. and the ACRIS Management Committee, R. . (2008), Taking the Pulse, Published on behalf of the ARCRIS Management Committee by the National Land & Water Resources Audit, Canberra. Bayly, I. A. E. (1999), ‘Review of how Indigenous people managed for water in desert regions of Australia’, Journal of the Royal Society of Western Australia 82, 17–25. Beckett, J. (1967), ‘Marriage, circumcision and avoidance among the Maljangaba of north-west New South Wales’, Mankind 6(10), 456–464. URL: https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1835-9310.1967.tb01348.x Benedetti, M. M., Cordova, C. E. and Beach, T. (2011), ‘Soils, sediments, and geoarchaeology: Introduction’, Catena 85(2), 83–86. URL: http://www.sciencedirect.com/science/article/pii/S0341816210001529 212 | References Biggs, A., Bryant, K. and Watling, K. (2010), ‘Soil chemistry and morphology transects to assist wetland delineation in four semi-arid saline lakes, south-western Queensland’, Australian Journal of Soil Research 48, 208–220. Bird, C. and Frankel, D. (1991), ‘Problems in constructing a prehistoric regional sequence: Holocene southeast Australia’,World Archaeology 23(2), 179–192. Bird, M. I., Condie, S. A., O’Connor, S., O’Grady, D., Reepmeyer, C., Ulm, S., Zega, M., Saltré, F. and Bradshaw, C. J. A. (2019), ‘Early human settlement of Sahul was not an accident’, Scientific Reports 9(1), 1–10. Bird, M., O’Grady, D. and Ulm, S. (2016), ‘Humans, water, and the colonization of Australia’, Proceeding of the National Academy of Sciences pp. 1–6. Bird, R. B., Codding, B. F., Kauhanen, P. G. and Bird, D. W. (2012), ‘Aboriginal hunting buffers climate-driven fire-size variability in Australia’s spinifex grasslands’, Proceedings of the National Academy of Sciences of the United States of America 109(26), 10287–10292. Birdsell, J. B. (1957), ‘Some population problems involving Pleistocene man’, Cold Spring Harbour Symposium on Quantitative Biology 22, 47–69. Black, L. (1941), Burial trees, Robertson and Mullens Ltd. Black, L. (1942), Cylcons. The mystery stones of the Darling River valley, Anthropological Society of New South Wales. Blackburn, S. (2016), The Oxford Dictionary of Philosophy, Oxford University Press. URL: https://www.oxfordreference.com/view/10.1093/acref/ Bobrov, A. A., Charman, D. J. and Warnter, B. G. (1999), ‘Ecology of testate amoebae (Protozoa: Rhizopoda) on peatlands in western Russia with special attention to niche separation in closely related taxa.’, Protist 150, 125–136. Boixadera, J., Riera, S., Vila, S., Esteban, I., Albert, R. M., Llop, J. M. and Poch, R. M. (2016), ‘Buried A horizons in old bench terraces in Les Garrigues (Catalonia)’, Catena 137, 633–650. BOM (2014), About frost, online resource, Bureau of Meteorology (BOM), Australian Government. URL: http://www.bom.gov.au/climate/map/frost/what-is-frost.shtml Bonney, F. (1884), ‘On some customs of the Aborigines of the River Darling, New South Wales’, The Journal of the Anthropological Institute of Great Britain and Ireland 13, 122–137. References | 213 Bowdery, D. (1998), Phytolith analysis applied to Pleistocene–Holocene archaeological sites in the Australian arid zone, 695, British Archaeological Reports, Oxford, UK. Bowdler, S. (1977), The coastal colonisation of Australia, in F. J. Allen, J. Golson and R. Jones, eds, ‘Sunda and Sahul, prehistoric studies in Southeast Asia, Melanesia and Australia’, Academic Press, London, pp. 205–246. Bowler (2012), Wind v water: glacial maximum records from the Willandra Lakes, in S. Haberle and B. David, eds, ‘Peopled landscapes: Archaeological and biogeographic approaches to landscapes’, Vol. 34, ANU Press. URL: http://www.jstor.org/stable/j.ctt24h85b Bowler, J. M. (1973), ‘Clay dunes: Their occurrence, formation and environmental significance’, Earth-Science Reviews 9, 315–338. Bowler, J. M. (1976), ‘Aridity in Australia: Age, origins and expression in aeolian landforms and sediments’, Earth-Science Reviews 12, 279–310. Bowler, J. M. (1998), ‘Willandra Lakes revisited: Environmental framework for human occupation’, Archaeology in Oceania 33(3), 120–155. URL: http://www.jstor.org/stable/40386901 Bowler, J. M., Johnson, H., Olley, J. M., Prescott, J. R., Roberts, R. G., Shawcross, W. and Spooner, N. A. (2003), ‘New ages for human occupation and climatic change at Lake Mungo, Australia’, Nature 421, 837–840. Bowler, J. M. and Magee, J. (2000), ‘Redating Australia’s oldest human remains: a sceptic’s view’, Journal of Human Evolution 38, 719–726. Box, J. B., Duguid, A., Read, R. E., Kimber, R. G., Knapton, A., Davis, J. and Bowland, A. E. (2008), ‘Central Australian waterbodies: The importance of permanence in a desert landscape’, Journal of Arid Environments 72(8), 1395 – 1413. URL: http://www.sciencedirect.com/science/article/pii/S0140196308000426 Boyland, D. (1974), Vegetation, Technical Report, Western arid region land use study–Part 1. Queensland Department of Primary Industries. Brook, B., Gillespie, R. and Martin, P. (2006), ‘Megafauna mix-up’, Australasian Science 27(5), 35–37. Brook, B. W. and Bowman, D. M. J. S. (2002), ‘Explaining the Pleistocene megafaunal extinctions: Models, chronologies, and assumptions’, PNAS 99(23), 14624–14627. Butzer, K. (1982), Archaeology as human ecology: method and theory for contextual approach, Cambridge University Press. 214 | References Butzer, K. W. (1973), ‘Springs sediments from the acheulian site of Amanzi (Uitenhage District, South Africa)’, Quaternaria 17, 299–319. Butzer, K. W., Abbott, J. T., Frederick, C. D., Lehman, P. H., Cordova, C. E. and Oswald, J. F. (2008), ‘Soil-geomorphology and “wet” cycles in the Holocene record of North-Central Mexico’, Geomorphology 101, 237–277. Cabane, D. and Shahack-Gross, R. (2014), ‘Understanding fossil phytolith preservation: The role of partial dissolution in paleoecology and archaeology’, PLoS ONE 10(5), 1–16. Cabanes, D., Gadot, Y., Cabanes, M., Finkelstein, I., Weiner, S. and Shahack-Gross, R. (2012), ‘Human impact around settlement sites: a phytolith and mineralogical study for assessing site boundaries, phytolith preservation, and implications for spatial reconstructions using plant remains’, Journal of Archaeological Science 39(8), 2697–2705. URL: http://www.sciencedirect.com/science/article/pii/S0305440312001288 Cabanesa, D., Weinera, S. and Shahack-Gross, R. (2011), ‘Stability of phytoliths in the archaeological record: a dissolution study of modern and fossil phytoliths’, Journal of Archaeological Science 38, 2480–2490. Campion, N. (2012), Astrology and cosmology in the world’s religions, New York University Press. Cane, S. (1987), ‘Australian Aboriginal subsistence in the Western Desert’, Human Ecology 15(4), 391–434. Cane, S. (1995), Nullarbor antiquity: archaeological, luminescent and seismic investigations on the Nullarbor Plain, Report to the National Estate Grants Program, Australian Heritage Commission, Canberra. Canti, M. (1995), A mixed-method approach to geoarchaeological analysis, in ‘Archaeological sediments and soils: analysis, interpretation and management’, UCL Press, Institute of Archaeology, London, pp. 183–190. Canti, M. and Huisman, D. J. (2015), ‘Scientific advances in geoarchaeology during the last twenty years’, Journal of Archaeological Science 56, 96–108. Chen, S. T. and Smith, S. Y. (2013), ‘Phytolith variability in Zingiberales: A tool for the reconstruction of past tropical vegetation’, Palaeogeography, Palaeoclimatology, Palaeoecology 370, 1–12. URL: https://www.sciencedirect.com/science/article/pii/S0031018212006098 Clarke, P. (1989), ‘Aboriginal non-plant medicines in soutern South Australia and western Victoria’, Journal of the Anthropological Society of South Australia 27(5), 1–10. References | 215 Clarke, P. (2009), Natural history of the riverlands and murraylands, in J. T. Jennings, ed., ‘Aboriginal culture and the riverine environment’, Royal Society of South Australia Inc., pp. 142–161. Clarke, P. A. (2007), ‘Indigenous spirit and ghost folklore of "settled" Australia’, Folklore 118(2), 141–161. URL: https://www.jstor.org/stable/30035418 Clarkson, C., Jacobs, Z., Marwick, B., Fullagar, R., Wallis, L., Smith, M., Roberts, R. G., Hayes, E., Lowe, K., carah, X., Florin, S. A., McNeil, J., Cox, D., Arnold, L. J., Hua, Q., Huntley, J., Brand, H. E. A., Manne, T., Fairbairn, A., Shulmeister, J., Lyle, L., Salinas, M., Page, M., Connell, K., Park, G., Norman, K., Murphy, T. and Pardoe, C. (2017), ‘Human occupation of northern australia by 65,000 years ago’, Nature 547, 306–313. Cogger, H. G. (2014), Reptilies and amphipbians of Australia, seventh edn, CSIRO Publishing. Cohen, T. J., Nanson, G. C., Jansen, J. D., Jones, B. G., Jacobs, Z., Larsen, J. R., May, J. H., Treble, P., Price, D. M. and Smith, A. M. (2012), ‘Late Quaternary mega-lakes fed by the northern and southern river systems of central Australia: Varying moisture sources and increased continental aridity’, Palaeogeography, Palaeoclimatology, Palaeoecology pp. 89–108. Cohen, T. J., Nanson, G. C., Jansen, J. D., Jones, B. G., Jacobs, Z., Treble, P., Price, D. M., May, J.-H., Smith, A. M., Ayliffe, L. K. and Hellstrom, J. C. (2011), ‘Continental aridification and the vanishing of Australia’s megalakes’, Geology 39(2), 167–170. URL: https://doi.org/10.1130/G31518.1 Collura, L. V. and Neumann, K. (2017), ‘Wood and bark phytoliths of West African woody plants’, Quaternary International 434, 142–159. 9th International Meeting of Phytolith Research (IMPR). URL: http://www.sciencedirect.com/science/article/pii/S1040618215014652 Conesa, F. C., Lobo, A., Alcaina, J., Balbo, A., Madella, M., Rajesh, S. V. and Ajithprasad, P. (2017), ‘Multi-proxy survey of open-air surface scatters in drylands: Archaeological and physico-chemical characterisation of fossilised dunes in North Gujarat (India)’, Quaternary International 436, 57–75. URL: http://www.sciencedirect.com/science/article/pii/S1040618216303378 Courty, M. A. (1992), ‘Soil micromorphology in archaeology.’, Proceedings of the British Academy 77, 39–59. Cowan, J. (1992),Mysteries of the Dream-time. The spritual life of Australian Aborigines, Prism Unity. Revised Edition. 216 | References David, B., Barker, B. and McNiven, I. J., eds (2006), The social archaeology of Australian Indigenous societies, Aboriginal Studies Press. David, B. and Thomas, J., eds (2016), Handbook of landscape archaeology, Routledge, Taylor and Francis Group, London and New York. Dawson, N. M. (1974a), Geology and landform, in ‘Western Arid Land Use Study, Part 1’, Information and Extension Training Branch of the Queensland Department «rf Primary Industries, techreport 3, pp. 12–17. Dawson, N. M. (1974b), Land Systems, in ‘Western Arid Land Use Study, Part 1’, Division of Land Utilisation, Queensland Government, chapter 7, pp. 18–46. Dawson, N. M. and Ahern, C. R. (1974), Soils, in ‘Western Arid Land Use Study, Part 1’, Division of Land Utilisation, Queensland Government, chapter 4, pp. 18–46. Dearing, J. (1999), Environmental magnetic suseptibility: Using the Bartington MS2 system, second edition edn, John A. Dearing. Dearing, J. A., Dann, R. J., Hay, K., Lees, J. A., Loveland, P. J., Maher, B. A. and O’Grady, K. (1996), ‘Frequency-dependentsusceptibility measurements of environmental materials’, Geophysics Journal International 124, 228–240. Dejmal, M., Lisa, L., Nyvltova, M. F., Bajer, A., Petr, L., Kocar, P., Kocarova, R., Nejman, L., Rybnıcek, M., Suvova, Z., Culp, R. and Vavrcık, H. (2014), ‘Medieval horse stable; The results of multi proxy interdisciplinary research’, PLOS ONE 9(3), 1–12. URL: https://doi.org/10.1371/journal.pone.0089273 Denham, T. (2008), Environmental archaeology: interpreting practices-in-the-landscape through geoarchaeology, in B. David and J. Thomas, eds, ‘Handbook of landscape archaeology’, Routledge, Talylor and Francis Group, London, pp. 468–481. Denniston, R. F., Wyrwoll, K.-H., Asmerom, Y., Polyak, V. J., Humphreys, W. F., Cugley, J., Woods, D., LaPointe, Z., Peota, J. and Greaves, E. (2013), ‘North Atlantic forcing of millennial-scale Indo-Australian monsoon dynamics during the Last Glacial period’, Quaternary Science Reviews 72, 159–168. Derbyshire, E., Kemp, R. and Meng, X. (1995), ‘Variations in loess and palaeosol properties as indicators of palaeoclimatic gradients across the Loess Plateau of North China’, Quaternary Science Reviews 14(7), 681–697. Aeolian Sediments in the Quaternary Record. URL: http://www.sciencedirect.com/science/article/pii/0277379195000771 Devos, Y. and Vrydaghs, L. (2011), ‘Integrating micromorphology and phytolith studies to investigate medieval Dark Earth: the example of Brussels’, Archeo . References | 217 Devos, Y., Vrydaghs, L., Dgraeve, A. and Fechner, K. (2009), ‘An archaeolopedological and phytolitarian study of the "Dark Earth" on the site of Rue de Dinant (Brussels, Belgium)’, Catena 78, 270–284. Douglass, L. A., ed. (1988), Soil micromorphology: a basic and applied science, Vol. 19 of Proceedings of the VIIth International Working Meeting of Soil Micromorphology, San Antonio, Texas–July 1988, Elsevier. Dunkerley, D. (2010), ‘Ecogeomorphology in the Australian drylands and the role of biota in mediating the effects of climate change on landscape processes and evolution’, Geological Society, London, Special Publications 346(1), 87–120. URL: https://sp.lyellcollection.org/content/346/1/87 Eldredge, N. and Gould, S. J. (1972), Punctuated equilibria: an alternative to phyletic gradualism, in ‘Models in paleobiology’, Freeman, Cooper & Company, San Franciso. Esteban, I., Marean, C. W., Fisher, E. C., Karkanas, P., Cabanes, D. and Albert, R. M. (2018), ‘Phytoliths as an indicator of early modern humans plant gathering strategies, fire fuel and site occupation intensity during the Middle Stone Age at Pinnacle Point 5-6 (south coast, South Africa)’, PLOS ONE 13(6), 1–33. URL: https://doi.org/10.1371/journal.pone.0198558 Eyre, E. J. (1845),Manners and customs of the Aborigines of Australia, Journals of Expeditions of Discovery, Boone, London. 145–512. Fairfax, R. J. and Fensham, R. J. (2003), ‘Great Artesian Basin springs in southern Queensland 1911-2000’, Memoirs of the Queensland Museum 49(1), 285–293. Fanning, P. C. (1999), ‘Recent landscape history in arid western New South Wales, Australia: a model for regional change’, Geomorphology 29, 191–209. Fanning, P. C., Holdaway, S. J. and Rhodes, E. J. (2008), ‘A new geoarchaeology of Aboriginal artefact deposits in western NSW, Australia: Establishing spatial and temporal geomorphic controls on the surface archaeological record’, Geomorphology 101(3), 524 – 532. Geoarchaeology and Geomorphology: Soils, Sediments, and Societies. URL: http://www.sciencedirect.com/science/article/pii/S0169555X0800072X Fanning, P. C., Holdaway, S. J., Rhodes, E. J. and Bryant, T. G. (2009), ‘The surface archaeological record in arid Australia: Geomorphic controls on preservation, exposure, and visibility’, Geoarchaeology 24(2), 121–146. URL: https://doi.org/10.1002/gea.20259 Fanning, P. and Holdaway, S. (2001), Temporal limits to the archaeological record in arid western NSW, Australia: Lessons from OSL and Radiocarbon Dating of hearths and 218 | References sediments, inM. Jones and P. Sheppard, eds, ‘Australasian connections and new directions’, Department of Anthropology, The University of Auckland, pp. 85–104. Fedoroff, N., Courty, M.-A. and Guo, Z. (2018), Palaeosoils and relict soils: A conceptual approach, in ‘Interpretation of micromorphological features of soils and regoliths’, Elsevier, pp. 822–858. Fensham, R. and Fairfax, R. (2003), ‘Spring wetlands of the Great Artesisan Basin, Queensland, Australia’,Wetlands Ecology and Management 11, 343–362. Field, J. and Dodson, J. (1999), ‘Late Pleistocene megafauna and archaeology from Cuddie Springs, south-eastern Australia’, Proceedings of the Prehistoric Society 65, 275–301. Field, J., Fullagar, R. and Lord, G. (2001), ‘A large area archaeological excavation’, Antiquity 75. Field, J. H. (2006), ‘Trampling through the Pleistocene: Does taphonomy matter at Cuddie Springs’, Australian Archaeology 63, 9–20. Field, J., Wroe, S. and Fullagar, R. (2006), ‘Blitzberg: Fact or fiction at Cuddie Springs’, Australasian Science 27(6), 1–28. Fillios, M., Field, J. and Charles, B. (2010), ‘Investigating human and megafauna co-occurrence in Australian prehistory: Mode and causality in fossil accumulations at Cuddie Springs’, Quaternary International 211, 123–143. Fitzpatrick, E. A. (1984), Micromorphology of soils, Chapman and Hall. Fitzsimmons, K. E. (2017), ‘Reconstructing palaeoenvironments on desert margins: new perspectives from Eurasian loess and Australian dry lake shorelines’, Quaternary Science Reviews 171, 1–19. URL: https://www.sciencedirect.com/science/article/pii/S0277379116305388 Fitzsimmons, K. E., Cohen, T. J., Hesse, P. P., Jansen, J., Nanson, G. C., May, J., Barrows, T. T., Haberlah, D., Hilgers, A., Kelly, T., Larsen, J., Lomax, J. and Treble, P. (2013), ‘Late Quaternary palaeoenvironmental change in the Australian drylands’, Quaternary Science Reviews 74, 78 – 96. Linking Southern Hemisphere records and past circulation patterns: the AUS-INTIMATE project. URL: http://www.sciencedirect.com/science/article/pii/S0277379112003435 Fitzsimmons, K. E., Stern, N. and Murray-Wallace, C. V. (2014), ‘Depositional history and archaeology of the central Lake Mungo lunette, Willandra Lakes, southeast Australia’, Journal of Archaeological Science 41, 349–364. References | 219 Flood, J. (2006), The original Austraians: story of the Aboriginal people, Allen & Unwin, Crows Nest, NSW. Florek, S. M. (1993), Archaeology of the mound springs campsites near Lake Eyre in South Australia, Unpublished PhD Thesis, School of Archaeology, Classics and Ancient History, Unversity of Sydney. Florin, S. A., Fairbairn, A. S., Nango, M., Djandjomerr, D., Marwick, B., Fullagar, R., Smith, M., Wallis, L. A. and Clarkson, C. (2020), ‘The first Australian plant foods at Madjedbebe, 65,000–53,000 years ago’, Nature Communications . French, C. A. I. (2003), Geoarchaeology in action. Studies in soil micromorphology and landscape evolution, Routledge Talylor & Francis Group, London and New York. French, C. A. I. (2010), ‘People, societies, and landscapes’, Science 328, 443–444. French, C. A. I. (2015), A handbook of geoarchaeological approaches for investigating landscapes and settlement sites, Oxbow Books, Oxford and Philadelphia, 10 Hythe Bridge Street, Oxford OX1 2EW. French, C. A. I. (2017), Colluvial settings, in A. Gilbert, P. Goldberg, V. Holliday, R. Mandel and R. Sternberg, eds, ‘Encyclopedia of geoarchaeology’, Springer Reference, pp. 157–171. French, C. A. I., Passmore, D. and Schulte, L. (1998), Geomorphology and edaphic factors, in ‘Aguas project: Palaeoclimatic reconstruction and the dynamics of human settlement and land use in the area of the middle Agaus (Almería), in the south-east of the Iberian Penisula’, Luxembourg: European Commission, pp. 45–52. French, C. A. I., Sulas, F. and Madella, M. (2009), ‘New geoarchaeological investigations of the valley systems in the Aksum area of northern Ethiopia’, Catena 78, 218–233. Friesem, D. E., Lavi, N., Madella, M., Ajithprasad, P. and French, C. A. I. (2016), ‘Site formation processes and hunter-gatherers use of space in a tropical environment: A geo-ethnoarchaeological approach from south India’, PLos ONE 11(10), 1–30. Gamage, B. (2011), The biggest estate on Earth. How Aborigines made Australia, Allen and Unwin. Gao, G., Jie, D., Wang, Y., Liu, L., Liu, H., Li, D., Li, N., Shi, J. and Leng, C. (2018), ‘Do soil phytoliths accurately represent plant communities in a temperate region? A case study of Northeast China’, Vegetation history and archaeobotany 27(6), 753–765. Gawne, B., Butcher, R., Hale, J., Kingsford, R., Brandis, K., Leo, V., Wheaton, L. and Hladyz, S. (2011), A review of river ecosystem condition in the Murray-Darling Basin, Technical Report 1, The Murray-Darling Freshwater Research Centre. 220 | References Gayler, L. M. (2008), A palaeoenvironmental history of the Paroo and Warrego regions, Australia a multi-proxy, multi-site approach, Unpublished PhD Thesis, Environmental Science, University of Newcastle. Gerrard, J. (1992), Soil geomorphology: an integration of pedology and geomorphoogy, London: Chapman and Hall. Gerritsen, R. (2001), ‘Aboriginal fish hooks in southern Australia: evidence, arguments and ikmplications’, Australian Archaeology (52), 18–28. URL: http://www.jstor.org/stable/40287518 Goldberg, P. (1980), ‘Micromorphology in archaeology and prehistory’, Paléorient 6, 159–164. URL: http://www.jstor.org/stable/41432207 Goldberg, P. and Macphail, R. I. (2006), Practical and theoretical geoarchaeology, Blackwell Publishing, 350 Main Street, Malden, MA 02148-5020, USA9600. URL: www.blackwellpublishing.com Goudie, A. S. (2013), Arid and semi-arid geomorphology, Cambridge University Press. Gould, R. (1979), ‘Exotic stone and battered bones: Ethnoarchaeology in the Australian desert’, Archaeologhy 32(2), 28–37. URL: http://www.jstor.org/stable/41726355 Gould, R. A. (1968), ‘Living archaeology: The ngatatjara of western australia’, Southwestern Journal of Anthropology 24(2), 1. Gould, R. A. (1971), ‘The archaeologist as ethnographer: A case from the western desert of australia’, World Archaeology 3(2), 143–177. URL: http://www.jstor.org/stable/124070 Gould, R. A. (1995), ‘Islands in the interior: The dynamics of prehistoric adaptations within the arid zone of Australia. Peter Marius Veth., Archaeological Series 3. International Monographs in Prehistory, Ann Arbor, 1993. viii 144 pp., figures, tables, bibliography, appendixes. ’28.00 (cloth); ’17.50 (paper).’, American Antiquity 60(3), 562—-563. Griffiths, B. (2018), Deep time dreaming. Uncovering ancient Australia, Carlton, Victoria: Black Inc. Hercus, L. A. (1980), ‘How we danced the Mudlunga’: Memories of 1901 and 1902’, Aboriginal History 1/2, 4–31. URL: http://www.jstor.org/stable/24045631 References | 221 Hercus, L. A. (2009), The two rainbow serpents travelling: mura track narrative from the ’Corner Country’, Aboriginal History Monograph 18, Australian National University E Press and Aboriginal History Incorporated. Hiscock, P. (2015), The Pleistocene colonization and occupation of Australasia, Cambridge University Press. Hiscock, P. and Clarkson, C. (2000), ‘Analysing Australian stone artefacts: an agenda for the Twenty-First Century’, Australian Archaeology 50, 98–108. Hiscock, P. and Wallis, L. (2005), Pleistocene settlement of deserts from an Australian perspective, in P. Veth, M. Smith and P. Hiscock, eds, ‘Desert peoples: Archaeological perspectives’, Blackwell Publishing, pp. 34–57. Holdaway, S., Davies, B. and Fanning, P. (2017), ‘Aboriginal use of fire in a landscape context: Investigating presence and absence of heat-retainer hearths in western New South Wales, Australia’, Current Archaeology 58, 230–242. Holdaway, S. and Fanning, P. (2014), Geoarchaeology of Aboriginal landscapes in semi-arid Australia, CSIRO Publishing,. Holdaway, S., Fanning, P. and Rhodes, E. (2008), ‘Challenging intensification: Human-environment interactions in the Holocene geoarchaeological record from western New South Wales, Australia’, The Holocene 18(3), 411–420. Holdaway, S. J., Fanning, P. C., Rhodes, E. J., Marx, S. K., Floyd, B. and Douglass, M. J. (2010), ‘Human response to palaeoenvironmental change and the question of temporal scale’, Palaeogeography, Palaeoclimatology, Palaeoecology 292, 192–200. Horton, D. (1981), ‘Water and woodland: the peopling of Australia’, Australian Institute of Aboriginal Studies 16, 21–27. Horton, D. R. (1996), ‘Aboriginal Australia Wall Map’, The Spatial Divison of Sinclair Knight Mertz. Hughes, P. J., Sullivan, M. E. and Hiscock, P. (2017), ‘Palaeoclimate and human occupation in southeastern arid australia’, Quaternary Science Reviews 163, 72–83. URL: http://dx.doi.org/10.1016/j.quascirev.2017.03.014 Ingold, T. (2000), The perception of the environment. Essays on livelihood, dwelling and skill, Routledge, New York. Jackson, M. L. (1958), Soil chemical analysis, Verlag: Prentice Hall, Inc., Englewood Cliffs, NJ. 222 | References Jankowski, N. R., Jacobs, Z. and Goldberg, P. (2015), ‘Optical dating and soil micromorphology at MacCauley’s Beach, New South Wales, Australia’, Earth Surface Processes and Landforms 40, 229–242. Jones, R. (1969), ‘Fire-stick farming’, Australian Natural History 16(7), 224–228. Jordon, P. (2014), Structured worlds. The archaeology of hunter-gatherer thought and action, in A. Cannon, ed., ‘Material culture perspectives on the worldview of northern hunter-gatherers’, Routledge. Taylor & Francis Group, London and New York, pp. 11–31. Kaczorek, D., Vrydaghs, L., Devos, Y., Peto, A. and Effland, W. (2018), Biogenic siliceous features, in ‘Interpretation of micromorphological features of soils and regoliths’, second edn, Elseveir, pp. 157–176. Kefous, K. C. (1988), Butlin’s bootstraps: Aboriginal population in the pre-contact Murray-Darling region, in B. Meehan and R. Jones, eds, ‘Archaeology with ethnography: an Australian perspective’, Australian National University, pp. 225–237. Kelly, R. L. (2013), Group size and demography, Cambridge University Press, pp. 166—-213. Kemp, C. W., Tibby, J., Arnold, L. J. and Barr, C. (2019), ‘Australian hydroclimate during Marine Isotope Stage 3: a synthesis and review’, Quaternary Science Reviews 204, 94–104. URL: https://www.sciencedirect.com/science/article/pii/S0277379118306978 Kenny, A. (2013), The Aranda’s Pepa. An introduction to Carl Strehlow’s masterpiece, Australian National University Press. Kimber, R. (1983), ‘Black lightning: Aborigines and fire in central Australia and the western desert’, Archaeology in Oceania 18(1), 38–45. Kingsford, R. T., Roshier, D. A. and Porter, J. L. (2010), ‘Australian waterbirds - time and space travellers in dynamic desert landscapes’,Marine and freshwater research 61, 875–884. Kooistra, M. J. and Kooistra, L. I. (2003), ‘Integrated research in archaeology using soil micromorphology and palynology’, Catena 54(3), 603–617. URL: http://www.sciencedirect.com/science/article/pii/S0341816203001371 Krapf, C., Werner, M., Questiaux, D., Spooner, N., Williams, F. and Dutch, R. (2018), Optically stimulated luminescence dating revealing new insights into the age of major regolith units of the eastern musgrave province, south australia, resreport 2018/00004, Department for Energy and Mining. References | 223 Kuhn, P., Aguilar, J., Miedema, R. and Bronnikova, M. (2018), Textural pedofeatures and related horizons, in ‘Interpretation of micrmorphological features of soils and regoliths’, second edn, Elsevier, pp. 377–424. Langton, M. (2006), Earth, wind, fire and water: the social and spiritual construction of water in Aboriginal societies, in B. David, B. Barker and I. McNiven, eds, ‘The social archaeology of Australia’, Aboriginal Studies Press, pp. 139–160. Latz, P. (1995), Bushfires and bushtucker: Aboriginal plant use in central Australia., Alice Springs: IAD Press. Liu, T. S. and Yuan, B. Y. (1982), ‘Quaternary climatic fluctuation—a correlation of records in loess with that of the deep sea core V28-238’, Research on Geology 1, 113–121. Institue of Geology, Academia Sinica (IGAS). Cultural Relics Publishing House, Beijing (in Chinese). Lourandos, H. (1993), ‘Hunter-gatherer cultural dynamics: Long- and short-term trends in australian prehistory’, Journal of Archaeological Research 1(1), 67–88. Lourandos, H. (1997), Continent of hunter-gatherers: New perspectives in Australian prehistory, Cambridge University Press. Lowe, P. and Pike, J. (2009), You call it desert - we used to live there., Magaballa Books Aboriginal Corporation. Macfarlance (1890), ‘Late colonial period painter, political cartoonist and illustrator’, Online. URL: https://trove.nla.gov.au/people/604896 Macphail, R. and Goldberg, P. (2017), Applied soils and micromorphology in archaeology, Cambridge University Press. Macphail, R. and Goldberg, P. (2018), Archaeological materials, in ‘Interpretation of micromorphological features of soils and regoliths’, Elsevier, pp. 779–820. Macphail, R. I., Courty, M. and Goldberg, P. (2006), ‘Soil micromorphology in archaeology’, Endeavour, New Series 14(4), 163–171. Macumber, P. G. and Thorne, R. (1975), ‘The Cohuna cranium site — A re-appraisal’, Archaeology & Physical Anthropology in Oceania 10(1), 67–72. URL: http://www.jstor.org/stable/40386238 Madella, M. and Lancelotti, C. (2012), ‘Taphonomy and phytoliths: A user manual’, Quaternary International 275, 76–83. Site Formation and Postdepositional Processes In Archaeology (International Workshop, Barcelona, 24 June 2010). URL: http://www.sciencedirect.com/science/article/pii/S1040618211005210 224 | References Mallol, C., Marlowe, F. W., Wood, B. M. and Porter, C. C. (2007), ‘Earth, wind, and fire: ethnoarchaeological signals of Hadza fires’, Journal of Archaeological Science 34(12), 2035–2052. URL: http://www.sciencedirect.com/science/article/pii/S0305440307000313 Matthews, W. (2016), ‘Humans and fire: Changing relations in early agricultural and built environments in the Zagros, Iran, Iraq’, The Anthropocene Review 3(2), 107–139. URL: anr.sagepub.com McBryde, I. (1987), Goods from another Country: Exchange networks and the people of the Lake Eyre Basin., in J. P. White and D. J. Mulvaney, eds, ‘Australians to 1788’, Vol. 1, Broadway: Fairfax, Syme and Weldon, pp. 252–273. McKeague, J. A., Guertin, R. K., Page, F. and Valentine, W. G. (1978), ‘Micromorphological evidence of illuvial clay in horizons designated Bt in the field’, Canadian Journal of Soil Science 58, 179–186. McKellar, H. (1984),Matya Mundu: a history of the Aboriginal People of south west Queensland, Cunnamulla Australian Native Welfare Association, Cunnamulla, Queensland. Memmott, P. (2007), Gunyah, Goondie and Wurley: the Aboriginal architecture of Australia, 1st edn, University of Queensland. Mentzer, S. M. (2014), ‘Microarchaeological approaches to the identification and interpretation of combustion featuresin prehistoric archaeological sites’, Journal of Archaeological Method Theory 21, 616–668. Miller, G. and Worland, J. (2017), Information sheet on the EAA Flyway Network Sites (SIS) – 2017 version, Technical report, Environmental Protection Agency. Mitchell, T. L. (1838), Three expeditions into the interior of eastern Australia; with descriptions of the recently explored region of Australia felix, and of the present colony of New South Wales, Vol. 1, 2 edn, T & W Boone, New Bond Street, London. Morton, A. (2016), Archaeological site formation: Understanding lake margin contexts, Vol. BAR International Series of 1211, BAR Publishing, Oxford. Mucher, H., van Steiin, H. and Kwaad, F. (2018), Colluvial and mass wasting deposits, in ‘Interpretation of micromorphological features of soils and regoliths’, second edn, Elsevier, pp. 21–36. Mulvaney, D. J. and Joyce, E. B. (1965), ‘Archaeological and geomorphological investigations on Mt. Moffat Station, Queensland’, Proceedings of the Prehistoric Society . References | 225 Mulvaney, D., Lawton, G. and Twidale, C. (1964), ‘Archaeological excavation of rock shelter no. 6 Fromm’s Landing, South Australia’, Proceedings Royal Society of Victoria 77(479). Mulvaney, J. (1990), Visions of environment: an afterview, in ‘The humanities and the Australian environment’, Vol. 11, Papers from the Australian Academy of the Humanites Symposium, pp. 111–122. Mulvaney, J. (2012), ‘Human cognition: the Australian evidence’, Antiquity 86, 915–921. URL: http://antiquity.ac.uk/ant/086/ant0860915.htm Munsell (1992), Soil colour charts, Macbeth, New York (Revised edition). Murray, A. S. and Wintle, A. G. (2000), ‘Luminescence dating of quartz using an improved single- aliquot regenerative-dose protocol’, Radiation Measurements 32, 57–73. Nagle, N., Ballantyne, K. N., van Oven, M., Tyler-Smith, C., Xue, Y., Wilcox, S., Wilcox, L., Turkalov, R., van Oorschot, R. A. H., van Holst Pellekaan, S., Schurr, T. G., McAllister, P., Williams, L., Manfred Kayser, R., Mitchell, J. and Consortium, T. G. (2016), ‘Mitochondrial DNA diversity of present-day Aboriginal Australians and implications for human evolution in Oceania’, Journal of Human Genetics 2016, 1–11. Nawaz, M. A., Zakharenko, A. M., Zemchenko, I. V., Haider, M. S., Ali, M. A., Imtiaz, M., Chung, G., Tsatsakis, A., Sun, S., Golokhvast and Sergeyevich, K. (2019), ‘Phytolith formation in plants: From soil to cell’, Plants 2(249), 1–38. Nelson, M. S., Gray, H. J., Johnson, J. A., Rittenour, T. M., Feathers, J. K. and Mahan, S. A. (2015), ‘User guide for luminescence sampling in archaeological and geological contexts.’, Advances in archaeological practice 3(2), 166–177. Neumann, K., Strömberg, C., Ball, T., Albert, R. M., Vrydaghs, L. and Scott (Cummings), L. S. (2018), International code for phytolith nomenclature (ICPN) 2.0. Supplementary information: morphotype descriptions, International Committee for Phytolith Taxonomy (ICPT). Nicosia, C. and Stoops, G., eds (2017), Archaeological soil and sediment micromorphology, Wiley-Blackwell. Norris, R. P. and Harney, B. Y. (2014), ‘Songlines and navigation in wardaman and other australian aboriginal cultures’, Journal of Astronomical History and Heritage 17(2), 141–148. NPRSR (2014), Currawinya lakes ramsar site / currawinya national park: Ramsar management summary 2014., Technical report, Marine Resource Management Unit, Queensland Parks and Wildlife Service, Department of National Parks, Recreations, Sport and Racing. 226 | References O’Dea, K., Jewell, P. A., Whiten, A., Altmann, S. A., Strickland, S. S. and Oftedal, O. T. (1991), ‘Traditional diet and food preferences of Australian Aboriginal hunter-gatherers [and discussion]’, Philosophical Transactions: Biological Sciences 334(1270), 233–241. URL: http://www.jstor.org/stable/55460 O’Connell, J. F. and Allen, J. (2004), ‘Dating the colonization of Sahul (Pleistocene Australia – New Guinea): a review of recent research’, Journal of Archaeological Science 31, 835–853. Pardoe, C. (1994), ‘Bioscapes: the evolutionary landscape of Australia’, Archaeology in Oceania 29(3), 182–190. Pardoe, C. (2003), ‘The Menindee Lakes: A Regional Archaeology’, Australian Archaeology 57, 42–53. URL: http://www.jstor.org/stable/40287713 Parr, J. (2006), ‘Effect of fire on phytolith coloration’, Geoarchaeology: An International Journal 21(2), 171–185. Parr, J. and Watson, L. (2007), Morphological characteristics observed in the leaf phytolith of select gymnosperms of Eastern Australia, in ‘Plants, people and places. Recent studies in phytolith analysis’, Oxbow Books, pp. 79–84. Pascoe, B. (2018), Dark Emu. Aboriginal Australia and the birth of agriculture, new edition edn, Magabala Books. Piperno, D. R. (1988), Phytolith analysis: An archaeological and geological perspective, Academic Press Inc. Purdy, D., Carr, P., Brown, D., Cross, A., Bultitude, R. J., Lee, M. S. and Verdel, C. (2016), The Granite Springs granite, Technical Report 2016/02, Department of Natural Resources and Mines, Geological Survey of Queensland. QDPI (1974), Western arid region land use study – Part 1, Queensland Department of Primary Industries, Technical Report 12, Queensland Department of Primary Industries. QPWS (2001), Currawinya National Park Management Plan, Queensland Government, Queensland Parks and Wildlife Service, Technical report, Queensland Government. Racek, A. A. (1969), ‘The freshwater sponges of Australia (Porifera: Spongillidae)’, Australian Journal of Marine and Freshwater Research 20, 267–310. Radcliffe-Brown, A. R. (1926), ‘The Rainbow-Serpent myth of Australia’, The Journal of the Royal Anthropological Institute of Great Britain and Ireland 56, 19–25. URL: http://www.jstor.org/stable/2843596 References | 227 Radcliffe-Brown, A. R. (1930), ‘The Rainbow-Serpent myth in south-east Australia’, Oceania 1(3), 342–347. URL: http://www.jstor.org/stable/40327331 Radke, B. M., Ferguson, J., Creswell, R. G., Ransley, T. R. and Habermehl, M. A. (2000), Hydrochemistry and implied hydrodynamics of the Cadna-owie—Hooray aquifer, Great Artesian Basin, Technical report, Bureau of Rural Sciences: Canberra. Ransley, T. R., Radke, B. M., Feitz, A. J., Kellett, J. R., Owens, R., Bell, J., Stewart, G. and Carey, H. (2015), Hydrogeological Atlas of the Great Artesian Basin, Geoscience Australia, Canberra. Reay, M. (1949), ‘Native thought in rural New South Wales’, Oceania 20(2), 89–118. URL: http://www.jstor.org/stable/40328235 Renfrew, C. and Bahn, P. (2016), Archaeology. Theories, methods and practice. Seventh edition, seventh edn, Thames & Hudson, New York. Rhodes, E. J., Fanning, P., Holdaway, S. and Bolton, C. (2009), Archaeological surfaces in western NSW: stratigraphic contexts and preliminary OSL dating of hearths, in A. Fairbairn, S. O’Connor and B. Marwick, eds, ‘New directions in archaeological science’, Terra Australis, ANU E Press, Australia, pp. 189–200. Richards, T., Pavlides, C., Walshe, K., Webber, H. and Johnston, R. (2007), ‘Box Gully: new eveidence for Aboriginal occupation of Australia south of the Murray River prior to the Last Glacial Maximum’, Archaeology in Oceania 42(1), 1–11. Roberts, R. G., Jacobs, Z., Li, B., Jankowski, N., Cunningham, A. C. and Rosenfeld, A. B. (2015), ‘Optical dating in archaeology: thirty years in retrospect and grand challenges for the future’, Jornal of Archaeological Science 56, 41–60. Robertson, S. (2011), Direct estimation of organic matter by loss on ignition: methods, Technical report, University of Cambridge. Robins, R. (1993), Archaeology and the Currawinya Lakes: towards a prehistory of southwest Queensland, Phd thesis, Griffith University. Robins, R. (1995), ‘The results of test excavations in three rockshelters in southwest Queensland’, Memoirs of the Queensland Museum 38(2), 643–666. Robins, R. (1996), ‘A report on archaeological investigations of open hearth sites in southwest Queensland’, Queensland Archaeological Research 10, 25–35. Robins, R. (1997), ‘Patterns in the landscape: A case study in nonsite archaeology from southwest Queensland’, Memoirs of Queensland Museum 1(1), 23–56. 228 | References Robins, R. (1998), ‘Archaeological investigations at Youlain Springs, southwest Queensland.’, Memoirs of the Queensland Museum 1(1), 57–74. Roth, W. E. (1897), Ethnographic studies among the north-west-central Queensland Aborigines, digitally printed version edn, Cambridge University Press. Round, F. E., Crawford, R. M. and Mann, D. G. (1990), The diatoms: biology and morphology of the genera, Cambridge University Press. Schaetzl, R. J. and Anderson, S. (2005), Soils gensis and geomorphology, Cambridge University Press. URL: www.cambridge.org/9780521812016 Schmidt, D. A. (2012), Archaeology, magnetic methods, in D. Gubbins and E. Herrero-Bervera, eds, ‘Encyclopaedia of geomagnetism and paleomagnetism’, Springer. Shahack-Gross, R. (2016), ‘Archaeological micromorphology self-evaluation exercise’, Geoarchaeology 31(1), 49–57. URL: https://onlinelibrary.wiley.com/doi/abs/10.1002/gea.21536 Shahack-Gross, R. and Finkelstein, I. (2008), ‘Subsistence practices in an arid environment: A geoarchaeological investigation in an Iron Age site, the Negev Highlands, Israel’, Journal of Archaeological Science 35, 965–982. Shulmeister, J. (1999), ‘Australasian evidence for mid-holocene climate change implies precessional control of Walker Circulation in the Pacific’, Quaternary International 57-58, 81–91. URL: https://www.sciencedirect.com/science/article/pii/S1040618298000524 Smith, M. (2013), The archaeology of Australia’s deserts., Cambridge University Press. Smith, M. A. (1989), ‘The case for a resident human population in the central Australian ranges during full glacial aridity’, Archaeology in Oceania 24(3), 93–105. Smith, W. R. (1930), Myths and legends of the Australian Aboriginals, Harrap, Sydney. Stern, N. (2008), Stratigraphy, depositional environments, and palaeolandscape reconstruction in landscape archaeology, in B. David and J. Thomas, eds, ‘Handbook landscape archaeology’, Routhledge, Taylor and Francis Group, London and New York, pp. 365–378. Stevenson, A. and Lindberg, C. A., eds (2010), New Oxford American Dictionary, third (3rd) edn, Oxford University Press. Stoops, G. (2003), Guidelines for analysis and description of soil and regolith thin sections, Soil Science Society of America, Inc. Madison, Wisconsin, USA. Eds. Warren A. Dick and Lisa Al-Amoodi. References | 229 Stoops, G., Marcelino, V. and Florias, M., eds (2018), Interpretation of micromorphological features of soils and regoliths, Elsevier, Amsterdam. URL: ttp://www.sciencedirect.com/science/article/pii/B9780444531568000271 Stoops, G. and Nicosia, C. (2017), Introduction, in D. C. Nicosia and E. P. D. G. Stoops, eds, ‘Archaeological soil and sediment micromorphology’, John Wiley and Sons Ltd. Strahan, R., ed. (1983), The Australian Museum. Complete book of Australian mammals, Angus and Robertson Publishers. Sturt, C. (1849), Narrative of an expedition into central Australia. Performed under the Authority of Her Majesty’s Government during the years 1844, 5, and 6, T. and W. Boone, 29 New Bond Street, London. Sutton, P. (2003), Native Title in Australia: an ethnographic perspective, Cambridge University Press. Thomas, A., ed. (2007), Culture in translation: The anthropological legacy of R. H. Mathews, Vol. 15, ANU Press. URL: http://www.jstor.org/stable/j.ctt24hc1r Thomas, W. J. (1923), Some myths and legends of the Australian Aborigines, Whitcombe and Tombs, Melbourne. Thorely, P. B. (1998), ‘Pleistocene settlement in the Australian arid zone: occupation of an inland riverine landscape in the central Australian ranges’, Antiquity 72(275), 34–45. Thorne, A., Gu˝rn, R., Mortimer, G., Spooner, N. A., Simpson, J. J., McCulloch, M., Taylor, L. and Curnoe, D. (1999), ‘Australia’s oldest human remains: age of the Lake Mungo 3 skeleton’, Journal of Human Evolution 36(6), 591–612. URL: http://www.sciencedirect.com/science/article/pii/S0047248499903056 Thunell, R. C., Tappa, E. and Anderson, D. M. (1995), ‘Sediment fluxes and varve formation in Satnta Barbara Basin, offshore California’, Geology 23(12), 1083–1086. Tibbett, K. (2002), ‘Archaeological analysis of stone axe exchange networks in the Lake Eyre Basin during the Mid-to Late Holocene’, Australian Archaeology 55, 22–29. URL: http://www.jstor.org/stable/40287599 Timms, B. V. (2001), ‘Large freshwater lakes in the arid Australia: A review of their limnology and threats to their future’, Lakes & Reservoirs: Research and Management 6, 183–196. Tindale, N. (1974), Aboriginal tribes of Australia. Their terrain environmental controls, distribution, limits, and proper names., Australian National University Press. 230 | References Tobler, R., Rohrlach, A., Soubrier, J., Bover, P., Llamas, B., Tuke, J., Bean, N., Abdullah-Highfold, A., Agius, S., O’Donoghue, A., O’Loughlin, I., Sutton, P., Zilio, F., Walshe, K., Williams, A. N., Turney, C. S. M., Williams, M., Richards, S. M., Mitchell, R. J., Kowal, E., Stephen, J. R., Williams, L., Haak, W. and Cooper, A. (2017), ‘Aboriginal mitogenomes reveal 50,000 years of regionalism in Australia’, Nature . Torrance, J. (1870), ‘The currawinya diaries’, Transcription. Triana-Vega, A. V., Sedov, S., Salinas-Acero, J., Carvajal-Contreras, D., Moreano, C., Tova-Reyes, M., Solleiro-Rebolledo, E. and Daaz-Ortega, J. (2019), ‘Environmental reconstruction spanning the transition from hunter/gatherers to early farmers in Colombia: paleopedological and archaeological indicators from the pre-ceramic sites Tequendama and Aguazuque’, Quaternary International 516, 175–189. Climate change and human-environment interaction from Neolithic to historical times. URL: http://www.sciencedirect.com/science/article/pii/S1040618218302787 van der Meer, J. J. M. and Menzies, J. (2011), ‘The micromorphology of unconsolidated sediments’, Sedimentary Geology 238(3), 213–232. URL: https://www.sciencedirect.com/science/article/pii/S0037073811001254 van Mourik, J. M. (1999), ‘The use of micromorphology in soil pollen analysis: The interpretation of the pollen content of slope deposits in Galicia, Spain’, CATENA 35(2), 239–257. URL: http://www.sciencedirect.com/science/article/pii/S0341816298001039 Van Vliet-Lanoe, B. and Fox, C. A. (2018), Frost action, in ‘Interpretation of micromorphological features of soils and regoliths’, second edn, Elsevier. Vannieuwenhuyse, D. (2008), Mind the gap: Geoarchaeology and micromorphology of cave and rockshelter sequences from the Kimberley, north-west Australia, Unpublished PhD Thesis, School of Social Science. Vannieuwenhuyse, D., O’Connor, S. and Balme, J. (2017), ‘Settling in Sahul: Investigating environmental and human history interactions through micromorphological analyses in tropical semi-arid north-west Australia’, Journal of Archaeological Science 77, 172–193. Vepraskas, M., Lindbo, D. and Stolt, M. (2018), Redioximrophic features, in ‘Interpretation of micromorphological features of soils and regoliths’, second edn, Elsevier, pp. 425–446. Veth, P. (1989), ‘Islands in the interior: A model for the colonization of Australia’s arid zone’, Archaeology in Oceania 24(3), 81–92. Veth, P. (1995), ‘Aridity and settlement in northwest Australia’, Antiquity 69, 733–746. References | 231 Veth, P. (2005), Cycles of aridity and human mobility: Risk minimization among late Pleistocene foragers of the western desert, Australia, in P. Veth, M. Smith and P. Hiscock, eds, ‘Desert peoples: Archaeological perspectives’, Blackwell Publishing, pp. 100–115. Wallis, L. A. (2001), ‘Environmental history of northwest Australia based on phytolith analysis at Carpenter’s Gap 1’, Quaternary International 83-85, 103 – 117. Australian Quaternary Studies: A Tribute to Jim Bowler. URL: http://www.sciencedirect.com/science/article/pii/S1040618201000337 Wang, N. (2018), Luminescence dating technical report, Technical Report 20/18, Luminescence Dating Laboratory, School of Geography, Environment and Earth Sciences, Victoria University of Wellington Wellington, New Zealand. Wang, N. (2019), Luminescence dating technical report, Technical Report 3/19, Luminescence Dating Laboratory, School of Geography, Environment and Earth Sciences, Victoria University of Wellington Wellington, New Zealand. Ward, I., Merigot, K. and McInnes, K. (2017), ‘Application of quantitative mineralogical analysis in archaeological micromorphology: a case study from Barrow Is., Western Australia’, Journal of Archaeological Method Theory pp. 1–24. Ward, I., Veth, P., Prossor, L., Denham, T., Ditchfield, K., Manne, T., Kendricke, P., Byrne, C., Hook, F. and Troitzsch, U. (2017), ‘50,000 years of archaeological site stratigraphy and micromorphology in Boodie Cave, Barrow Island, Western Australia’, Journal of Archaeological Science: Reports 15, 344–369. Waters, M. R. (1991), ‘The geoarchaeology of gullies and arroyos in Southern Arizona’, Journal of Field Archaeology 18(2), 141–159. URL: http://www.jstor.org/stable/530254 Webb, S., Cupper, M. and Robins, R. (2006), ‘Pleistocene human footprints from the Willandra Lakes, southeastern Australia’, Journal of Human Evolution 50, 405–413. Westaway, M., Olley, J. and Grun, R. (2017), ‘At least 17,000 years of coexistence: Modern humans and megafauna at the Willandra Lakes, south-eastern Australia’, Quaternary Science Reviews 157, 206–211. Willcox, A. R., Layton, R. and Lewis-Williams, J. D. (1987), ‘The cultural context of hunter-gatherer rock art’,Man 22(1), 171–175. URL: http://www.jstor.org/stable/2802969 Williams, A. N., Ulm, S., Turney, C. S. M., Rohde, D. and White, G. (2015), ‘Holocene demographic changes and the emergence of complex societies in prehistoric Australia’, PLoS One 10(6), 1–17. 232 | References Williams, A., Ulm, S., Cook, A. R., Langley, M. and Collard, M. (2013), ‘Human refugia in Australia during the Last Glacial Maximum and Terminal Pleistocene: a geospatial analysis of the 25-12 ka Australian archaeological record’, Journal of Archaeological Science 40, 4612–4625. Williams, M. (2015a), ‘Hydrogeological-Landscapes system: a framework for managing water resources’, Catena 134, 4–13. Williams, M. (2015b), ‘Interactions between fluvial and eolian geomorphic systems and processes: Examples from the Sahara and Australia’, Catena 134, 4–13. Winkworth, R. E. and Thomas, P. R. (1974), Climate, Technical Report, Queensland Department of Primary Industries. Witter, D. (2007), Aboriginal dreaming and aridity, in ‘Animals of arid Australia’, Royal Zoological Society of New South Wales, chapter 4, pp. 14–29. URL: https://publications.rzsnsw.org.au/doi/abs/10.7882/FS.2007.038 Zaidner, Y., Frumkin, A., Friesem, D., Tsatskin, A. and Shahack-Gross, R. (2016), ‘Landscapes, depositional environments and human occupation at Middle Paleolithic open-air sites in the southern Levant, with new insights from Nesher Ramla, Israel’, Quaternary Science Reviews 138, 76–86. Zeanah, D. W., Codding, B. F. and Bird, R. B. B. . D. W. (2017), ‘Mosaics of fire and water: the co-emergence of anthropogenic landscapes and intensive seed exploitation in the Australian arid zone’, Australian Archaeology 83(1-2), 2–19. Glossary of Terms This Glossary of Terms has largely been sourced and compiled from French (2015: 115–118) and other listed sources. Aeolian: windblown sediment; associated with loess; Alluvium: well sorted, homogeneous, freshwater-borne sediment, generally silt- and clay-sized sediment, entrained and aggrading in a river valley floodplain; Argillic horizon: clay enriched lower B (or Bt) horizon of a soil; Arroyo: a steep-sided gully formed by the action of fast-flowing water in an arid or semi-arid region(Stevenson and Lindberg 2010) B horizon: below A horizon and above the C horizon; Bioturbation: associated with burrowing invertebrates and soil fauna mixing soil; Buried soil or old land surface: a soil formed under previous environmental conditions buried by later drift or erosion deposits; C horizon: geological hardrock or drift substrate that weathers to create the soil above; Calcareous: pH>7; Coarse component: gravel and stone (>2 mm in diameter) component of soil; Colluvium (hillwash): loose, non-stratified, ill-sorted, heterogeneous mixture of sediment and various particle sized derived from soil/subsoil erosion upslope and redeposited at the base of the slopes; Cutan: clay coating, generally forming on the walls of the capillary voids in the lower or B horizon of a soil; DEM: digital elevation model; 3-D terrain model of the landscape; Dirty clay: illuvial clay containing abundant silt-sized carbonised and/or amorphous organic matter fragments and punctuations; Dusty clays: illuvial clay containing micro-contrasted silt-sized particles of silt and fine organic matter; 234 | References Ecological baseline: risks and impacts that effect the overall conditions in which living organisms within their physical surroundings can support populations of hunter gatherers; Eluvial/eluviation: removal of silt/clay/fine organic matter from an upper soil horizon by leaching and groundwater percolation; Fines: Silt and clay component of a soil/sediment; Geoarchaeology: interlinked study of landforms, landscape change and human impact on landscapes through archaeological, geographical, geomorphological, and soil science approaches; Geomorphology: the study of landforms; GIS: Geographical Information System; Gleying: influence of groundwater leading to greater/lesser degree of waterlogging causing reduction of iron and manganese staining; Groundmass: the main mineral and organic components of soil and their arrangement and relationships to each other; generally composed of variable mixtures of very fine sand, silt and clay components as well as fine organic matter;’ Groundwater table: the interface between the saturation zone and the capillary fringe in soil/sediment/geological substrate sequence; Heat retainer: a group of stone or baked clay peds with charcoal (sometimes) and scorched earth; Humus: organic component of soils, as either plant tissue and/or amorphous matter; occurs in three types: mull (acidic), moder (neutral) and mor (basic); Hunter gatherers: this means the Budjiti people or Traditional Owners of lands not limited to the Eulo Ridge. And, it more generally refers to the Aboriginal people who hunted and gathered food resources and maintained a complex social system to occupy the Australian continent for the past 65,000 years; Illuvial/illuviation: mobilisation, removal and redeposition of fines (silt, clay and organic matter) towards the base of the profile, generally occurring in the lower or B horizons of soils leaving a depleted or illuvial soil horizon above; Infiltration capacity: the ability of a soil or sediment to absorb surface and rain-water; Landscape: our surroundings—natural, human-modified and constructed; Last Glacial Maximum (LGM): climatic changes that had a profound effect on the surficial geology of Australia and the rest of the world. Late Kosciuszko Glaciation and glaciers on the highlands of Tasmania. Global LGM: 21 ± 3 ka (Barrows et al. 2002); Leaching: removal of fines and nutrients of a soil through percolating groundwater; References | 235 Limpid clay: pure clay; Loess: windblown silt and very fine sand, usually 25–50 µm a size range; Loam: an equal mix of sand, silt, clay and organic matter; a common Mulga Lands soil (Dawson and Ahern 1974); Loss on ignition: method of determining the total organic content of a soil/sediment by burning off the organic content in a muffle furnace; Magnetic susceptibility: measure of magnetic enhancement of the soil/sediment generally caused by burning; Micromorphology: the study of soils/sediments in thin section using a polarising microscope; Microstructure: the within-ped organisation of a soil; Minerogenic: sand, silt, clay components of a sediment/soil; Optically stimulated luminescence (OSL): a dating method for sediment/soil horizons and surfaces using measurement of aliquots of medium-coarse quartz grains (Murray and Wintle 2000); Oxidation: opposite of reduction; oxygen domination soil system, leading to formation of iron oxides and hydroxides and destruction of organic component of the soil; ferrous iron (Fe 2+) becomes ferric (Fe 3+) resulting in the precipitation of iron, usually as an oxide; disadvantageous for organic preservation; Palaeochannel: Relict stream or river channel, usually braided, meandering or anastomosing type; Palaeosol: buried soi as found under either archaeological sites or monuments and under more recent drift deposits, and generally exhibiting soil characteristics no longer observed in the same locale today; Ped: main structural unit of organisation of a soil; Pedofeatures: discrete zones in soils distinguished from the groundmass and resulting from soil formation processes; Phreatic zone: relating to or denoting underground water in the zone of saturation (Stevenson and Lindberg 2010); Punctuated equilibria/equilibrium: a biological evolution theory developed by Eldredge and Gould (1972) to reveal that the history of life is more adequately determined by rapid and episodic events of speciation rather than gradual change in the fullness of time (Darwinism). Within geoarchaeology, this theory represents rapid erosion and episodic changes over time (French 2003); Phytoliths: microscopic silica bodies that denote the assemblages of trees, shrubs and grasses; 236 | References Potentiometric surface: a hypothetical surface defined by the level to which water in a confined aquifer rises in observation boreholes. In practice, the potentiometric surface is mapped by interpolation between borehole measurements. As with the water table in an unconfined aquifer, the slope of the potentiometric surface defines the hydraulic gradient and the horizontal direction of groundwater flow (Allaby 2015); Reduction: Removal of oxygen and the formation of ferrous iron (Fe2+); Refugia: an indiscrete area in which an Aboriginal population can survive through a period of glacial or extreme aridity (Smith 2013; Veth 1989; 1995); Rhizolith: organosedimentary structures that have been described from many parts of the world and are produced by the activity and decay of plant roots (Stoops et al. 2018: 226); Sediment: weathering, transport and redeposition by various geographic agencies, which may or may not exhibit horizonation upon deposition; Sesquioxides: oxides and hydroxides of iron and aluminium whose formation is generally associated with wetting and drying soil conditions; Soil: an inorganic/organic material developing through the weathering of earths mantle by physical and chemical processes and geographic agents through time such that distinct horizonation occurs; Soil texture: the relative proportions of sand, silt and clay in the soil; Testate amoeba: eukaryotic unicellular organisms occurring abundantly in freshwater and estuarine environments, peatlands (Bobrov et al. 1999); Textural pedofeatures: coatings and/or infillings of pure to intermixed clay, silt and very fine organic matter formed by illuvial deposition down-profile of fine materials eluviated from upper layers by the soil water; Valley system: an area of landscape from watershed to watershed across a river valley; Vughs: Relatively large voids, other than packing voids, usually irregular and not normally interconnected with other voids of comparable size; at the magnification at which they are recognized they appear as discrete entities (Fitzpatrick 1984: 412). Appendix A Field and Laboratory Methods Micromorphology and Particle Size Analysis (PSA) Collection of undisturbed micromorphology samples: Blocks of sediment were collected with an archaeological trowel, tape measure, aluminium trays spray glue and, plastic film wrapping. The process commenced with the removal of sediments from around the top and sides of the block leaving the bottom intact. In some cases, spray-on glue bonded loose sediments and a plastic film placed over the glue and sample to retain its shape. Next, the sediment was removed from around the base of the block and the foil tray (similar to a small oven tray) was pushed carefully over the block. The trowel was then inserted into the top about 5 cm to 7 cm deep and the sample prized into the block and removed with it intact within the foil tray. An “arrow” marked the 'top' of the sample. Plastic film was wrapped around the sample to hold it firmly within the foil tray to retain its shape. The sample was labelled with site and sample details. About 8 stone artefacts were included in the shipment. All samples were placed in a strong box and mailed to the McBurney Laboratory, Department of Archaeology, University of Cambridge, United Kingdom. Once the samples arrived in Cambridge, they were stored in the DEFRA room in the West Building. How to make a thin section The next task was to mount these bulk samples onto thin sections. I selected 16 samples and transferred into trays for airing and drying in the Micromorphology Laboratory, University of Cambridge. On 15 January 2019, I transferred the 16 samples into eight boxes, for processing. To prepare these samples, I removed the field label (this was kept separate for my storage), cut-off the plastic wrapping with a sharp razor and scissors to reveal the bulk sample resting in the oven tray. Eight boxes were placed on the shelf in the laboratory for the next step of the process—oven drying and filling the boxes with resin. 238 | Field and Laboratory Methods Resin impregnation methods for undisturbed samples On 18 January 2019 I worked with laboratory technician Dr Tonko Rajkovaca to impregnate 16 samples in 8 plastic tubs. To begin, all equipment and liquids were opened and placed on benches for easy access to develop a routine of making the resin and pouring it into sample trays. This included placing strips of alfoil paper on the bench in the working areas—the alfoil has been obviously used many times by other researchers. The resin materials included, tins of resin, hardener, acetone, pipette, 300 ml beaker, small container to hold the hardener, a two litre bucket, and wooden spoon. Step 2 was to place all sample tubs (containing the samples) in the fume cupboard. This are pushed close together so when pouring the resin into the tubs, it is easy to move to the next tub once it is full. A detailed description is described in, “A step-by-step guide to the making of soil thin sections” (Edited by Charly French and Tonko Rajkovaca, McBurney Geoarchaeology Laboratory, Division of Archaeology, University of Cambridge Url: https://www.arch.cam.ac.uk/files/guide-thin- sections.pdf, accessed 08 November 2021). Particle Size Analysis Methods The Department of Geography, University of Cambridge details a set of instructions. These are as follows: The samples are stirred to counteract settling during transit. 2-3 tablespoons of sediment are put into labelled 250ml beakers. Use two sandbaths in a fume cupboard on heat setting 2. Put the beakers onto the sandbath, ensuring they are well spaced to avoid contamination, about 15 beakers can be put on each sandbath. Pour 30 percent hydrogen peroxide into each beaker covering the sediment (wear lab coat, gloves and glasses when handling). Have wash bottles of meths (IMS) to hand, to control frothing of the sediments by squirting a small amount into the beaker. When the frothing has subsided then turn the sand baths up to 6 or 7. Pour more hydrogen peroxide onto the samples and control with meths. When all the organic material has been oxidised and there is no more frothing turn the heat off and leave to cool. Cover the sample with water to keep moist. Continue with analysis on Malvern as above. | 239 Gloves and glasses are required when handling these samples and solutions. Clean 50ml plastic tubes and numbered with permanent marker are required. A lump of sediment approx. 1-5g is put in each tube. These are topped up with 4.4\% sodium pyrophosphate and immersed in a water bath at 90\textdegree C for approximately three hours. This is stirred once with a glass rod. The samples are removed from the water bath and centrifuged at 3500 rpm for 13 minutes. It is important to use the scales to balance the tubes, topping up with water if necessary. These tubes are removed from the centrifuge and liquid decanted off. Samples are now ready for the Malvern. Switch on the power, check the drain tubes are in the sink and switch on the water supply (tap). Open the Mastersizer software on the PC. Glass beakers are filled with tap water to present the sample to the laser sizer. The pump rate and ultrasonic stir can both be set on the unit. Fill a beaker 4/5 full with water and lower the stirring head into place. The laser sizer automatically checks the water in the bath, measures the background and so on. When this is complete , you will have a background reading of the obscuration value of clean tap water (Source, https://www.geog.cam.ac.uk/facilities/laboratories/techniques/psd.html) Before adding the sample, mix it on the whirl-mixer. It is important to ensure samples are moving freely in the tube and that there are no lumps of clay stuck to the bottom. A few drops of the sample should be added to the beaker with a pipette from the tube on the whirlimixer so all the particles are evenly mixed and distributed. Check the obscuration value given on the screen. An ideal value is 15 percent, between 10 - 20 percent is acceptable. If the value is too low, add another drop or two of sample, if it is too high add more water using the fill switch to dilute the sample. Analysis of the sample will proceed when the correct value is maintained for 6 seconds. When analysis has finished and the graph is plotted, drain the sample in the sink and flush the unit three times with clean water to remove all excess particles. This prepares it for the next sample. Data is automatically saved during this analysis. A statistical summary can be exported to an Excel file and pdf files. 240 | Field and Laboratory Methods | 241 Magnetic susceptibility procedures The Bartington Magnetic Susceptibility Meters are designed to measure a variety of ferro magnetic minerals in the soils and sediments. Results depend on the amount, type and particle size. A well-type sensor measures each discrete sample and outputs data to a personal computer and Excel spreadsheet (SOURCE: https://www.geog.cam.ac.uk/facilities/laboratories/techniques/, 03 April 2019). The magnetic susceptibility output process begins with 15 mm diameter and 15 mm high plastic container (with a lid); set of balance scales; containers filled with sediment, and a Bartington instrument and personal computer. To obtain magnetic susceptibility for each sample, container tare weight (average), sediment weight (full container minus its empty tare weight), and ’air readings’ (no container) prepare the reader for low and high frequency recordings. Each result outputs in SI Units. Dr Armin Schmidt (2012) describes magnetic susceptibility as: "Thermoremanent magnetization is probably the best understood magnetic effect caused by past human habitation. If materials that are rich in iron-oxides are heated above their Curie temperature and then allowed to cool in the ambient Earth’s magnetic field they have the potential to acquire a considerable thermoremanence that is fixed in the material until further heating. Typical archaeological examples are kilns and furnaces, often built of clay, which during their heating cycles often exceed the Curie temperatures of magnetite and maghaemite (578 °C and 578–675 °C, respectively). Such iron-oxides are commonly found in the clay deposits that were used for the construction of these features. Even if the clays only contained weakly magnetic haematite or goethite, the heating and cooling cycles may have converted these into ferri-magnetic iron-oxides". The procedures to collect magnetic susceptibility readings were provided by David Redhouse, Department of Archaeology, University of Cambridge. These are as follows: Use of the Bartington Instruments MS3 magnetic susceptibility meter !"#$%&'(")*+,& & & Preparation -$&.+/&#"(&/)*,0&%1(&2+34/%("&4"+'*5(5&*,&678& 9()%&6/*:5*,0&+;%#*,&#,+/,%&#,5&4#))<+"5& +,&%1(&!(4#"%3(,%=)&2+34/%("&).)%(3&;.&)(,5*,0& #&"(>/()%&%+?& rt@arch.cam.ac.uk& -$&.+/&#"(&/)*,0&.+/"&+<,&:#4%+4&5+<,:+#5&%1(& 6#"%)+$%&)+$%<#"(&$"+3& http://www.bartington.co.uk/& #,5&*,)%#::&*%+"5*,0&%+&%1(&*,)%"/2%*+,)&+,&%1#%& <(;@)*%(A&&B+/&<*::&"(>/*"(&#&:#4%+4&"/,,*,0&)+3(& C*,5&+$&9*,5+<)?&D1(&)+$%<#"(&*)&,+%&)/*%#;:(&$+"& E#2)A& Assembly F::&+%1("&%1*,0)&;(*,0&(>/#:&.+/&<*::&"(>/*"(&+,:.& %1"((&+$&%1(&2+34+,(,%)&)1+<,&*,&%1(&4*2%/"(& ;(:+/*43(,%&#,5&"(%/",&*%&%+&*%)& ;+MA& Notes S/"%1("&*,$+"3#%*+,&#;+/%&%1(&+4("#%*+,&+$&%1(& EGH&3(%("8&%1(&6#"%)+$%&)+$%<#"(8&#,5& (,'*"+,3(,%#:&3#0,(%*2&)/)2(4%*;*:*%.&*)&#'#*:#;:(& +,@:*,(&#%?& http://www.bartington.co.uk/magneticsusceptibilityOM.cfm& Support O+34/%("&+$$*2(")&2#,&;(&2+,%#2%(5&;.&)(,5*,0& 3#*:&%+?& & 248 | Field and Laboratory Methods | 249 Loss on Ignition Loss on ignition commences with pre- and post-temperature treatments—place one cubic centimetre of sediment into a pre-weighed crucible; weigh crucible and sediment on an electronic balance and record results in an Excel Spreadsheet—a direct link between the balance and Excel reduces human error (https://www.geog.cam.ac.uk/facilities/laboratories/ techniques/, 03 April 2019) Soil pH procedures The pH readings method—place 10 grams of ground sediment into a small beaker, fill with 25 millilitres of deionized water, and stir to disperse sediments. Turn on hand-held pH meter, place first beaker near instrument, immerse pH reader in dissolved sediment, and wait for approximately five seconds to obtain reading, and record measurement. Appendix B Soil Analyses Data and Archaeological Material 252 | Soil Analyses Data and Archaeological Material Particle Size Analysis Plots: R Language descriptions [language=R - Create plot for particle size analysis] #Load libraries library(readxl) library(ggplot2) #Build particle size analysis graphs with depth ggplot() + geom_linerange(data=hckpsacols, mapping=aes(x=Percent, y=Depth, ymin=Up, ymax=Down), size=1.5, color="black") + geom_point(data=hckpsacols, mapping=aes(x=Percent, y=Depth), size=2, shape=21, fill="white") + scale_y_reverse() + theme_bw() + facet_wrap(. ~ ‘Sed‘, ncol = 5, scales = "free") + expand_limits(y = 1) + labs(x = "Percent (%)",y = "Depth (cm)") Bulk Density and Loss on Ignition Plots [language=R - Create plot for ’bulk density’, CaCO3, Carbon data] #Load libraries library(readxl) library(ggplot2) #Build graphs for bulk density and loss on ignition with free x-axis scale ggplot() + geom_linerange(data=dwbulkdensloi, mapping=aes(x=Sed, y=Depth, ymin=Up, ymax=Down), size=1.5, color="black") + geom_point(data=dwbulkdensloi, mapping=aes(x=Sed, y=Depth), size=2, shape=21, fill="white") + scale_y_reverse() + theme_bw() + facet_wrap(. ~ ‘LOI‘, ncol = 3, scales = "free") + labs(x = "Percent (%)", y = "Depth (cm)") | 253 Granites Particle Size Analysis Table B.1: Granites East site particle size analysis results Clay Silt V F Sand F Sand M Sand C Sand S1 2.32 27.32 27.29 24.55 16.55 1.96 S2 5.76 34.17 28.34 23.15 8.59 0.00 S3 4.74 38.38 27.96 19.95 8.90 0.07 S4 3.87 32.18 29.61 24.45 9.90 0.00 S5 10.26 37.86 18.74 19.72 13.29 0.13 S6 11.34 27.58 20.65 24.29 15.33 0.31 SA 4.65 30.21 25.42 24.39 14.20 0.63 Sapr 7.64 42.21 27.61 17.81 4.73 0.00 Slith 3.82 33.48 28.07 24.41 10.22 0.00 Slith2 6.23 28.84 21.20 24.86 17.70 0.66 Sup 3.37 32.86 29.96 23.39 10.31 0.11 Table B.2: Granites Dune site particle size analysis. Depth (cm) Clay Silt V F Sand F Sand M Sand C Sand S1 23 0.00 11.26 28.24 44.34 16.15 0.00 S2 41 0.00 8.50 22.93 45.16 23.41 0.00 S3 45 0.00 10.28 23.85 44.65 21.22 0.00 S4 53 0.00 5.73 19.51 47.60 27.11 0.05 S5 59 3.88 13.65 25.01 40.73 16.72 0.00 S6 71 4.60 8.74 19.39 43.75 23.52 0.00 S7 78 7.56 14.35 17.48 38.17 22.40 0.03 S8 104 3.99 10.94 22.04 41.43 21.53 0.07 S9 108 9.02 14.13 40.03 27.28 0.15 0.00 SA 0.00 0.98 10.15 21.21 44.96 22.71 0.00 254 | Soil Analyses Data and Archaeological Material 20 40 60 80 100 20406080100 20 40 60 80 10 0 cla y sand silt clay sand silt PSD GE Sites Figure B.1: Soil classification for Granites East site. 20 40 60 80 100 20406080100 20 40 60 80 10 0 cla y sand silt clay sand silt PSD GD Sites Figure B.2: Soil classification for Granites Relict Dune site. | 255 Table B.3: Hardpan Creek North site particle size analysis Depth (cm) Clay Silt V F Sand F Sand M Sand C Sand S1 28 6.96 30.78 33.92 23.87 4.46 0.00 S2 35 2.05 17.02 33.44 37.55 9.94 0.00 S3 55 3.46 24.78 29.33 30.60 11.33 0.00 S4 64 3.84 20.67 29.18 33.92 11.89 0.00 S5 124 2.83 21.34 28.82 33.30 13.71 0.00 S6 131 6.62 28.05 18.31 24.78 20.29 1.94 S7 145 7.71 40.21 27.85 19.72 4.51 0.00 Sgrav 75 3.29 22.57 32.11 31.89 10.14 0.00 Hardpan Creek 20 40 60 80 100 20406080100 20 40 60 80 10 0 cla y sand silt clay sand silt PSD HCKN Sites Figure B.3: Soil classification for Hardpan Creek North site 256 | Soil Analyses Data and Archaeological Material Table B.4: Hardpan Creek centre and south sites particle size analysis. Clay Silt V F Sand F Sand M Sand C Sand HCKS1 3.17 21.24 25.76 33.45 16.19 0.19 HCKS2 4.85 28.37 30.55 29.09 7.14 0.00 HCKC1 1.71 19.96 28.50 34.23 15.47 0.13 20 40 60 80 100 20406080100 20 40 60 80 10 0 cla y sand silt clay sand silt PSD HCKO Sites Figure B.4: Soil classification for Hardpan Creek South and Centre sites | 257 Basin Gully Table B.5: Basin Gully sites particle size analysis. Clay Silt V F Sand F Sand M Sand C Sand F Gravel BG1 3.87 25.67 25.13 29.729 15.22 0.39 0.00 BG2 10.53 34.42 22.6 21.98 10.47 0.00 0.00 20 40 60 80 100 20406080100 20 40 60 80 10 0 cla y sand silt clay sand silt PSD BG Sites Figure B.5: Soil classification for Basin Gully sites. 258 | Soil Analyses Data and Archaeological Material Double Well and Tunkana Well Sites Table B.6: Double Well site particle size analysis. Clay Silt V F Sand F Sand M Sand C Sand S1 2.99 33.14 35.26 21.47 7.14 0.00 S2 4.72 41.88 34.74 16.55 2.11 0.00 S3 3.35 23.47 19.74 23.18 25.18 5.07 S4 8.15 45.29 25.87 15.59 5.10 0.00 S5 6.95 33.59 22.29 21.49 14.91 0.78 S6 6.65 53.30 29.22 8.84 1.99 0.00 S7 5.21 37.13 30.27 20.43 6.96 0.00 S8 4.02 35.80 34.36 19.52 6.19 0.11 S9 4.42 41.31 33.64 16.75 3.88 0.00 20 40 60 80 100 20406080100 20 40 60 80 10 0 cla y sand silt clay sand silt PSD DW Sites Figure B.6: Soil classification for Double Well site. | 259 Table B.7: Tunkana Well site particle size analysis. Clay Silt V F Sand F Sand M Sand C Sand F Gravel S1 3.17 47.66 35.63 11.53 2.01 0.00 0.00 S2 4.05 52.21 32.63 9.60 1.50 0.00 0.00 S3 3.32 55.00 35.16 5.82 0.69 0.00 0.00 S4 4.75 44.54 35.47 13.62 1.61 0.00 0.00 S5 8.96 36.07 33.73 19.08 2.16 0.00 0.00 S6 15.70 63.44 19.13 1.72 0.00 0.00 0.00 S7 7.39 46.68 29.58 12.98 3.37 0.00 0.00 S8 6.48 27.15 25.95 26.51 13.67 0.24 0.00 S9 1.50 20.25 26.26 30.73 19.31 1.96 0.00 S10 32.31 35.29 7.30 9.65 13.63 1.81 0.08 S11 2.92 31.06 33.54 24.78 7.70 0.00 0.00 20 40 60 80 100 20406080100 20 40 60 80 10 0 cla y sand silt clay sand silt PSD TW Sites Figure B.7: Soil classification for Tunkana Well sites 260 | Soil Analyses Data and Archaeological Material All Sites 20 40 60 80 100 20406080100 20 40 60 80 10 0 cla y sand silt clay sand silt PSD All Sites Figure B.8: Soil classification for all samples | 261 Bulk Density and Loss on Ignition Granites East and Granites Dune Sites (a) GE: Bulk density and LOI (b) GD: BD and LOI Figure B.9: GE and GD Sites: Bulk density and loss on ignition 262 | Soil Analyses Data and Archaeological Material Hardpan Creek North Sites Figure B.10: HPCKN bulk desnity and loss on ignition | 263 Hardpan Creek South and Basin Gully sites (a) HPCKS: Bulk density and LOI (b) Basin Gully: Bulk density and LOI Figure B.11: HCKS and BG SiteS: Bulk density and loss on ignition 264 | Soil Analyses Data and Archaeological Material Double Well and Tunkana Well sites (a) DW: Bulk density and LOI (b) TW: Bulk density and LOI Figure B.12: DW and TW: Bulk density and loss on ignition | 265 All Sites: Loss on ignition Figure B.13: Bulk Density and loss on ignition comparison for all sites 266 | Soil Analyses Data and Archaeological Material The Archaeological Setting Stone artefacts found cemented into features (Fig. B.14). (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) (l) Figure B.14: Stone artefacts embedded into sediments features: a) GE (Arroyo): a small flake exposed in the upper layer; b) GE (Horizontal): small flake exposed from eroded layer; c) GE (Arroyo): flaked piece cemented in lower layer; d) GE (Arroyo): exposed piece within central layer of channel; e) GD (Arroyo): core exposed in ’static’ layer; f) GD (Arroyo): flake exposed in middle layer of dune; g) GD (Arroyo): flaked piece exposed from static layer; h) HPCKN (Arroyo): flake cemented horizontally in lower alluvial layer; i) HPCKS (Plan View): retouched flaked exposed within indurated feature; j) HPCKS (Plan View): flake exposed in indurated layer; k) HPCKS (Deposit): flake cemented vertically in central layer and; l) BG1 (Arroyo): weathered flaked piece cemented in lower layer Key: GE = Granties East, GD = Granites Dune, HPCKN = Hardpan Creek North, HPCKS = Hardpan Creek South, and BG = Basin Gully | 267 (a) Proximal view (b) Dorsal view Figure B.15: Stone artefact found 95 cm deep wedged between two laminar deposits. a) proximal view of the ventral surface with platform and distinct ventral surface and, b) dorsal view Optically Stimulated Luminescence (OSL) dates Table B.8: Optically stimulated luminescence dates for all sites: Hardpan Creek North (HCKN) and Hardpan Creek South (HCKNS), Basin Gully 1 (BG1) and Basin Gully 2 (BG2), and Granites Dune (GD) (Wang 2018; 2019) Code a-Value* De (Gy) DR* (Gy/ka) Age (ka) Site ID WLL1354 0.06 ± 0.03 72.67 ± 1.60 1.30 ± 0.13 55.9 ± 5.9 HPCKN OSL3 WLL1349 0.07 ± 0.03 48.30 ± 0.03 1.92 ± 0.18 25.2 ± 2.6 BG1 WLL1350 0.07 ± 0.03 30.51 ± 0.085 1.56 ± 0.15 19.6 ± 2.0 BG2 WLL1348 0.07 ± 0.03 16.19 ± 0.055 1.11 ± 0.11 14.6 ± 1.5 GD Static WLL1346 0.07 ± 0.03 21.46 ± 1.26 1.86 ± 0.16 11.5 ± 1.2 HPCK OSL WLL1345 0.07 ± 0.03 6.91 ± 0.27 1.50 ± 0.10 4.6 ± 0.4 HPCK C WLL1347 0.07 ± 0.03 1.02 ± 0.04 1.86 ± 0.16 0.7 ± 0.1 HPCK Mulga (*) estimated a-value; DR = dose rate; Appendix C Micromorphology descriptions and Opal phytoliths micrographs 270 | Micromorphology descriptions and Opal phytoliths micrographs | 271 BASIN GULLY SITE BASIN GULLY 1: Thin section 2: base of the profile next to stone artefacts; Size: 9 cm by 5.8 cm wide (52.2 sq. cm); thickness ~20–40 µm; MICROSTRUCTURE AND POROSITY: Channel to Vughy microstructure; partly accommodating planes <6 cm long and <0.1 cm wide with serrated to smooth edges; Vughs; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: ratio: 15/85; clay-loam; poorly sorted; sub-angular to sub-rounded coarse fraction; weakly oriented; porphyric; Colour: brown/orange PPL, brown/orange CPL; Mineral: Coarse Fraction: coarse sand (1000–2000µm) = 5%; medium sand (200–500µm) = 5%; fine sand (100–200µm) = 5%; Fine Fraction: very fine sand (20–100µm) = 50%; silt (2–20µm) = 5%; clay (<2µm) = 30%; sub-rounded to rounded; ORGANIC COMPONENTS: wood fragment 0.3 mm long, 0.1 mm wide with elongate phytolith feature running along length of wood fragment; isolated phytoliths; ash fragments; sponge spicules in Zone 1 (upper) and 2 (lower); PEDOFEATURES: Textural: ZONE 1 (upper); in this slide: 1) abundant dusty clay coatings of <50% of sand grains, and irregular zones of groundmass; linked/bridged coatings; moderate birefringence; moderate to strong sesquioxides staining the fine fabric; dusty to pure clays, partial/discontinuous to occasionally near complete infills of channels (also with discontinuous/irregular infills of the fabric); some pure clays, medium–large, and crescentic void fills with medium to high birefringence; ZONE 2 (lower); <10% of groundmass; hints of reticulate striation in groundmass, fine porphyric structure; infillings and clay coats of the voids; channel coatings with parent material and CaCO3; Amorphous: discontinuous clay coatings; 20% crusts INTERPRETATIONS: dry environment; transition from wet to dry conditions; illuviation; aligned with last glacial maximum; sponge spicules (inferred raised groundwater table); groundwater pool; water lapping pool edge (sediment laminations); vegetated environment; 272 | Micromorphology descriptions and Opal phytoliths micrographs BASIN GULLY 2: Thin section 9: centre of profile next to stone artefacts; Size: 10.9 cm by 5.7 cm wide (62.13 sq. cm); thickness ~40 µm; MICROSTRUCTURE AND POROSITY: Channel microstructure, accommodating, 2 cm long, mostly serrated; vughs <0.2 cm; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 45/55; sandy-clay; poorly sorted; weakly oriented; close porphyric; Colour: orange/brown PPL and orange/red/brown CPL; Coarse Fraction: coarse sand (1000–2000µm) = 5%; medium sand (200–500µm) = 20%; fine sand (100–200µm) = 20%; sub-angular to rounded; randomly sorted; weakly oriented; Fine Fraction: very fine sand (20–100µm) = 5%; silt (2–20µm) = 10%; clay (<2µm) = 45%; ORGANIC COMPONENTS: Charcoal pieces <4 mm long, intrapedal random orientation; small numbers of elongate phytolith, some molten/damaged by heat; PEDOFEATURES: Textural: abundant dusty to pure clay coatings of <75% of sand grains; linked/bridged coatings; moderate birefringence of the clays; weak crescentic coatings, dispersed within the clay groundmass; reticulate striation in groundmass; moderate silt cappings of the large grains; voids partial to completely filled with parent material; Amorphous: <10% Fe- and Mn-oxide nodules within the groundmass and minimal within voids; weak amorphous material with in low number of crescentic coatings; weak CaCO3 nodules; INTERPRETATIONS: dry environment; stability; frosts; natural and/or human fires; | 273 GRANITES DUNE SITE GRANITES DUNE X-Layer: Thin section 7: sampled across layers above static later; size; 12.5 cm by 5.7 cm wide (71.25 sq. cm); thickness ~40-50 µm; MICROSTRUCTURE AND POROSITY: Vughy to channel microstructure, vughs <0.3 cm, channels <1.2 cm long serrated to smooth and accommodating; small to medium-sized accommodating channels and small to medium-sized angular-shaped voids MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 70/30; sandy-loam; sub-angular to sub-rounded; weakly orientated; occasional clusters; strongly sorted; chitonic; Colour: orange/brown PPL; red/brown CPL; Coarse Fraction: coarse sand (1000–2000µm) = 5%; medium sand (200–500µm) = 60%; fine sand (100–200µm) = 5%; Fine Fraction: very fine sand (20–100µm) = 5%; silt (2–20µm) = 2%; clay (<2µm) = 23%; Micromass: dark to light brown, circular striated b-fabric ORGANIC COMPONENTS: plant tissue; charcoal fragments (medium to coarse sand size and one pieces 1.5 mm diameter); starch grains in voids; globular phytoliths PEDOFEATURES: Textural: abundant dusty to pure clay coatings of <50% of sand grains; moderate silt coatings of the large grains; depletion of fabric; linear and sand grain bridging of the larger sand grains; three horizontal laminar features composed of parent material <0.5 cm wide and length is width of slide 5.8 cm; infillings of voids with parent material and low; Amorphous: Fe/Mn-oxide nodules in the base of the profile; small iron-oxide nodules <0.15 cm; no amorphous material; weak crystalline materials INTERPRETATIONS: wet and dry conditions; dune-forming activity; unstable environment; multiple phases of sediment deposition; frosts; bioturbation; aligns with deglacial period; 274 | Micromorphology descriptions and Opal phytoliths micrographs GRANITES DUNE Static Layer: Thin section 5: indurated base layer with stone artefacts; Size: 12.5 cm by 5.6 cm wide (60.0 sq. cm); thickness ~20–30 µm; MICROSTRUCTURE AND POROSITY: Vughy to channel microstructure, vughs <0.65 cm oblong/rounded smooth to serrated; channels <1.5 cm serrated to smooth; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 70/30; sandy-loam; porphyric; angular to sub-angular; weakly oriented; well sorted; Colour: brown/orange PPL; red/brown/orange CPL; dark to light brown, circular striated b-fabric; Coarse Fraction: coarse sand (1000–2000µm) = 5%; medium sand (200–500µm) = 60%; fine sand (100–200µm) = 5%; Fine Fraction: very fine sand (20–100µm) = 5%; silt (2–20µm) = 2%; clay (<2µm) = 23%; ORGANIC COMPONENTS: possible plant tissue; charcoal fragments (medium to coarse sand size and one pieces 1.5 mm diameter); starch grains in voids; oval-shaped phytoliths; PEDOFEATURES: Texture: abundant dusty to pure clay coatings of <50% of sand grains; moderate silt coatings of the large grains; depletion of fabric; linear and sand grain bridging of the larger sand grains; partial infilling of voids with medium-sized quartz grains; three horizontal laminar features composed of parent material <0.5 cm wide and length is width of slide 5.8 cm; infillings of voids with parent material and low; Amorphous: Fe/Mn-oxide nodules; three small iron-oxide nodules; no CaCO3 INTERPRETATIONS: colluvium; stable conditions; pre-dune forming activity; stabilisation of the dune with chenopods and grasses; illuviation and vertical translocation of fines; natural or human fires; | 275 GRANITES EAST SITE GRANITES EAST 0–7 cm: Thin section 18: top layer of profile; size: 9.2 cm by 5.8 cm wide (53.6 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: channel to vughy microstructure; vughy, peds <1.5 cm, channels <2.5 cm long and <0.15 cm wide; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 50/50; sandy-clay-loam; porphyric; sub-angular; weakly oriented; poorly-sorted; Colour: brown/orange PPL; brown/orange/red CPL; Coarse Fraction: coarse sand (1000–2000µm) = 10%; medium sand (200–500µm) = 10%; fine sand (100–200µm) = 40%; Fine Fraction: very fine sand (20–100µm) = 20%; silt (2–20µm) = 2%; clay (<2µm) = 18%; ORGANIC COMPONENTS: phytoliths; micro-charcoal; PEDOFEATURES: Textural: bioturbation; compound packing voids; root voids; surface crusts; channel coatings and infillings of illuvial clay; interpedal clay coatings of voids; bioturbation; abundant dusty to pure clay coatings of <75% of sand grains; linked/bridged coatings; sesquioxides around sand grains; moderate birefringence of the clays; weak crescentic coatings; complete to <70% infills of voids and channels with parent materials; reticulate striation in groundmass; fine porphyric structure; moderate silt cappings of the coarse grains; voids partial to completely filled with parent material; Amorphous: <10% Fe- and Mn-oxide nodules within the groundmass and voids; Fe/Mn-oxide nodules <0.15 cm; INTERPRETATIONS: phreatic zone; fluctuating groundwater heights; dry cold conditions with frosts; bioturbation with translocation of the fines; illuviation; grasses and natural and/or human fires; 276 | Micromorphology descriptions and Opal phytoliths micrographs GRANITES EAST 30–39 cm: Thin section 19: mid/low of profile; Size: 9 cm by 5.8 cm wide (55.1 sq. cm); thickness ~30–40 µm MICROSTRUCTURE AND POROSITY: Angular blocky microstructure, accommodating planes, intrapedal channels, compound packing voids, vughy; peds <1.4 cm long and 0.7 cm wide; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 60/40; sandy-clay; sub-angular; weakly oriented; poorly-sorted; close porphyric; Colour: orange/brown PPL; brown/orange/red CPL; Coarse Fraction: coarse sand (1000–2000µm) = 10%; medium sand (200–500µm) = 20%; fine sand (100–200µm) = 20%; Fine Fraction: very fine sand (20–100µm) = 10%; silt (2–20µm) = 5%; clay (<2µm) = 35%; ORGANIC COMPONENTS: phytoliths <2%; crystalised organic matter <2%; PEDOFEATURES: Textural: bioturbation; root voids; compound packing voids; abundant dusty to pure clay coatings of <75% of sand grains; intrapedal clay coatings, linked/bridged coatings; hyper-coatings; irregular zones of groundmass implies partial depletion; weak to moderate birefringence of the clays; weak crescentic coatings; no visible sesquioxides; Amorphous: irregular CaCO3, discontinuous in voids; inherited nodules 0.3 cm long and 0.2 cm wide; intrapedal diffuse Mn-hydroxide nodules <0.15 cm; partial and filling of voids with CaCO3 nodules; Fe/Mn-oxide cappings on the larger sand grains; INTERPRETATIONS: phreatic zone; Illuviation; fluctuating wet and dry conditions and/or alternate groundwater flows with changing groundwater table; evaporation; bare surfaces; | 277 GRANITES EAST 50–57 cm: Thin section 20: base of profile; Size: 10.9 cm by 5.8 cm wide (63.22 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: Highly separated angular blocky microstructure; peds <4 cm long, <2.2 cm wide; accommodating channels <0.2 cm wide, serrated to smooth; compound packing voids <0.6 cm wide, vughs 0.4 cm; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 40/60; sandy-clay-loam; porphyric; Colour: Coarse Fraction: coarse sand (1000–2000µm) = 10%; medium sand (200–500µm) = 20%; fine sand (100–200µm) = 10%; sub-angular; weakly oriented; poorly-sorted; Fine Fraction:very fine sand (20–100µm) = 30%; silt (2–20µm) = 2%; clay (<2µm) = 28%; ORGANIC COMPONENTS: low number of bulliform, trichome/tracheid phytoliths PEDOFEATURES: Textural: dusty to pure clay coatings of <40% of sand grains; linked/bridged coatings; irregular zones of groundmass; linked/bridged coatings; weak birefringence of the clays; weak crescentic coatings; moderate sesquioxides staining of the quartz grains; Fe- to Mn-oxide cappings on the larger sand grains; infills of the channels complete to <50% discontinuous/irregular of the fabric and Fe/Mn-oxide and CaCO3 nodules; Amorphous: discontinuous to irregular CaCO3 in voids–strong evaporation/drying phase; channel infilling with parent materials; calcium carbonate nodule <0.7 cm long and 0.4 cm wide; iron oxide nodules <0.7 cm; INTERPRETATIONS: phreatic zone; groundwater fluctuations; partial depletion of fabric; limited grassed environment; 278 | Micromorphology descriptions and Opal phytoliths micrographs GRANTES EAST - 040818: Thin section 8: next to small stone artefact; Size: 11.9 cm by 5.8 cm wide (69.02 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: Channel to vughy microstructure; bioturbation; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 50/50; clay-loam; close porphyric; Colour: brown PPL; brown/orange/red CPL; Coarse Fraction: coarse sand (1000–2000µmm) = 5%; medium sand (200–500µm) = 5%; fine sand (100–200µm) = 40%; sub-angular; weakly oriented; poorly-sorted; Fine Fraction:very fine sand (20–100µm) = 10%; silt (2–20µm) = 2%; clay (<2µm) = 38%; ORGANIC COMPONENTS: <5% phytoliths; PEDOFEATURES: Textural: abundant dusty clay coatings of <75% of the sand grains; complete to <50% infills of <0.6 cm voids with parent material; continuous zones of groundmass; linked to bridged coatings; poor birefringence; moderate to strong sesquioxide stained on the fine fabric; dusty to pure clays, partial to continuous to complete; moderate amounts of reticulate striation of groundmass; Amorphous: CaCO3 nodules discontinuous in voids; a manganese nodule 0.6 cm and iron-oxide nodule <0.1 cm; INTERPRETATIONS: phreatic zone; stable conditions; groundwater; relatively bare surfaces; | 279 GRANITES EAST 2 Art: Thin section 17: next to two stone artefact; Size: 11.0 cm by 5.8 cm wide (63.8 sq. cm); thickness ~40–60 µm; thin section scrubbing during manufacture in the top-right corner; MICROSTRUCTURE AND POROSITY: Channel microstructure, vughy; channels <5.8 cm long and 0.25 cm wide; sub-angular nodules <0.4 cm long and <0.2 cm wide with compound packing voids <0.21 cm wide; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 65/35; sandy-clay-loam; porphyric; Colour: brown/orange PPL; brown/orange CPL; Coarse Fraction: coarse sand (1000–2000µm) = 5%; medium sand (200–500µm) = 20%; fine sand (100–200µm) = 40%; sub-angular; weakly oriented; well-sorted; homogeneous; Fine Fraction: very fine sand (20–100µm) = 5%; silt (2–20µm) = 2%; clay (<2µm) = 28%; ORGANIC COMPONENTS: PEDOFEATURES: Textural: abundant dusty clay coatings of <75% of the sand grains; continuous zones of groundmass; linked to bridged coatings; poor birefringence; moderate to strong sesquioxides stained on the fine fabric; dusty to pure clays, partial to continuous to complete; moderate amounts of reticulate striation of groundmass; complete to <50% infills of groundmass in the voids; Amorphous: CaCO3 nodules discontinuous in voids; evaporation phases; iron-oxide nodules <0.3 cm (sub-angular) scattered through micromass; crusts <2 cm long; coatings and infillings; INTERPRETATIONS: phreatic zone; groundwater fluctuations; post deposition; spring flows; 280 | Micromorphology descriptions and Opal phytoliths micrographs HARDPAN CREEK SITE HARDPAN CREEK NORTH Mulga: Thin section 4: 30 cm below mulga soil surface; Size: 11.4 cm by 5.8 cm wide (66.12 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: Channel to vughy microstructure; channels are <2 cm long, serrated to smooth; accommodating planes; compound packing voids; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 83/17; sandy-loam; sub-angular; weakly oriented; well-sorted; chitonic; Colour: brown/orange PPL; brown/orange/red CPL; Coarse Fraction: coarse sand (1000–2000µm) = 5%; medium sand (200–500µm) = 40%; fine sand (100–200µm) = 38%; Fine Fraction: very fine sand (20–100µm) = 5%; silt (2–20µm) = 2%; clay (<2µm) = 10%; ORGANIC COMPONENTS: <55% phytoliths; charcoal nodule <0.15 cm; organic material within soil nodules; PEDOFEATURES: Textural: sub-rounded to rounded peds <0.4 cm at the base of the slide; silt cappings; irregular clay zones in the groundmass, discontinuous <30% and weak to no clay-bridge within the voids; dusty to clean clays, discontinuous to incomplete, low to no birefringence; sesquioxide staining of the fine fabric and streaks across the coarse fabric/sand grains; partial to complete infills of the channels and voids; 30% excremental fabric; soil surface nodules with phytoliths and organic matter; hypercoatings of voids and coarse fraction; Amorphous: nodules of CaCO3 within the voids <5%; clay coatings and intrapedal diffuse Mn-hydroxide nodules <0.15 cm; INTERPRETATIONS: colluvial; wet and dry conditions; vegetated environment; soil depletion; natural and/or human fires; | 281 HARDPAN CREEK NORTH 2–9 cm: Thin section 18: top of relict soils; Size: 12.6 cm by 5.8 cm wide (73.08 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: Channel to vughy microstructure; channels < 2cm long, serrated to smooth; accommodating planes, vughs <0.5 cm; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 37/63; loamy-sand; sub-angular to sub-rounded; weakly oriented; well-sorted; porphyric; Colour: brown/orange PPL, brown/orange/red CPL; Coarse Fraction: coarse sand (1000–2000µm) = 2%; medium sand (200–500µm) = 30%; fine sand (100–200µm) = 5%; Fine Fraction: very fine sand (20–100µm) = 40%; silt (2–20µm) = 5%; clay (<2µm) = 18%; ORGANIC COMPONENTS: large wood fragments within the voids and groundmass; <10% phytoliths; <2% micro-charcoal; PEDOFEATURES: Textural: surface crusts <3.9 cm long and 0.6 cm long; several sub-angular blocky peds <2.4 cm long and <1.8 cm wide at base of slide; clay infills along horizontal planes; crusts <3.9 cm long and <0.05 cm wide; parent material nodules <0.15 cm; surface nodules washed in from surfaces under tree canopy;) silt cappings; irregular clay zones in the groundmass, discontinuous <30% and weak to no clay-bridge within the voids; medium to high birefringence of the fine fraction; sesquioxide staining of the fine fabric; clean to dusty clays, discontinuous with partial infills and complete infills of the voids; strong crusting on fragments <1 cm long ;Amorphous: weak CaCO3 nodules within groundmass; diffuse Mn-hydroxide nodules <0.05 cm; INTERPRETATIONS: illuviation under tree canopy; bioturbation and formation of surface crusts; sub-soil depletion; 282 | Micromorphology descriptions and Opal phytoliths micrographs HARDPAN CREEK NORTH 29–38 cm: Thin section 25: top section of relict soils; Size: 12.2 cm by 5.8 cm wide (70.76 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: Channel to vughy microstructure; channels < 3 cm long and <0.05 cm wide, serrated and accommodating; spheroidal peds; voids <1.2 cm wide; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 35/65; sandy-clay; porphyric; sub-angular; weakly oriented; well-sorted; Colour: brown/orange PPL; orange/brown CPL; Coarse Fraction: coarse sand (1000–2000µm) = 5%; medium sand (200–500µm) = 30%; fine sand (100–200µm) = 10%; Fine Fraction: very fine sand (20–100µm) = 20%; silt (2–20µm) = 5%; clay (<2µm) = 40%; ORGANIC COMPONENTS: <40% phytoliths within infills and surface crusts; rare molten phytoliths; probable ash?; and crystallised organic matter; micro-charcoal; PEDOFEATURES: Textural: abundant dusty clay coatings of <75% of the sand grains; regular and irregular zones of groundmass; links and bridges coatings; moderate to strong birefringence; strong sesquioxides staining the fine fabric and infills nodules; dusty to pure clays partial to near complete; infills of the channels and voids; several zones <40% of the groundmass; partial reticulate striation to massive compound clay coatings of the coarse fabric; Amorphous: bottom zone contains an abundance of CaCO3 nodules that infill the channels; upper zone contains CaCO3 nodules in the groundmass; linear and rounded crusts <5.8 cm long and 0.2 cm wide; parent material infills of oblong voids <1.1 cm long and 0.4 cm wide; intrapedal diffuse Mn-hydroxide nodules <0.15 cm; INTERPRETATIONS: wet and dry phases; illuviation; post-deposition; vegetated surfaces; natural and/or human fires; | 283 HARDPAN CREEK NORTH 45 cm OSL: Thin section 18: sample near 11.2 ka OSL date; Size: 11.8 cm by 5.8 cm wide (68.4 sq. cm); thickness ~30–40 µm MICROSTRUCTURE AND POROSITY: Channel to lenticular blocky microstructure; peds <1.4 cm long and <0.7 cm wide; vughy; strong bioturbation and lenticular to laminar peds; MINERAL COMPONENTS: Groundmass: c/f ratio5µm: 50/50; sandy-clay; sub-angular; coarse fraction moderately oriented with CaCO3 laminations; well-sorted; porphyric; Colour: orange/light brown PPL, orange/brown/yellow CPL; Coarse Fraction: coarse sand (1000–2000µm) = 5%; medium sand (200–500µm) = 5%; fine sand (100–200µm) = 40%; Fine Fraction:very fine sand (20–100µm) = 10%; silt (2–20µm) = 1%; clay (<2µm) = 39%; ORGANIC COMPONENTS: phytoliths; crystallised organic fragments; PEDOFEATURES: Textural: abundant dusty clay coatings of <85% of sand grains—top of the profile; regular and irregular zones of groundmass; bridged coatings; moderate to strong birefringence; strong sesquioxide staining of the fine fabric; dusty to pure clay complete to incomplete infills of channels and voids with fragmented concentric void fills and linear dusty clay coatings; hints of reticulate striation in the lower section; Amorphous: strong laminated CaCO3 infills in the voids and channels, which intermix with the coarse/fine fabric; one large Fe/Mn-nodule 0.5 cm long and 0.2 cm wide; other nodules >0.05 cm; calcium carbonate features <1.8 cm long and <0.4 cm wide, some are clay coated; interpedal clay coatings and intrapedal calcium carbonate nodules; Mn-hydroxide nodules; INTERPRETATIONS: humid to dry transitional conditions; illuviation; Ficus spp. and vegetation; 284 | Micromorphology descriptions and Opal phytoliths micrographs HARDPAN CREEK NORTH: 90–97 cm: Thin section 10 (1 of 2): low feature; Size: 13 cm by 5.8 cm wide (75.4 sq. cm); thickness ~0–40 µm MICROSTRUCTURE AND POROSITY: Channel and vughy microstructure; channels <1.3 cm long and <0.05 cm wide, serrated accommodating; coatings and infillings of illuvial clay; large compound packing voids; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 41/59; sandy-clay; sub-angular; weakly oriented; poorly-sorted; porphyric; Colour: orange/brown/black PPL; orange/brown/red CPL; Coarse Fraction: coarse sand (1000–2000µm) = 5%; medium sand (200–500µm) = 5%; fine sand (100–200µm) = 40%; Fine Fraction: very fine sand (20–100µm) = 10%; silt (2–20µm) = 1%; clay (<2µm) = 49%; ORGANIC COMPONENTS: <50% phytoliths within the groundmass; organic coatings above crescentic dusty clay coatings in the voids; PEDOFEATURES: Textural: dusty clay coatings of <50% of sand grains; irregular to complete groundmass; bridged coatings; moderate to strong birefringence on the coatings; moderate to strong sesquioxides staining on fine fabric and within channels; dusty to pure clay, coatings of voids, channels, and peds; sesquioxides; more recent organic coatings of the crescentic coatings; Amorphous: Mn-hydroxide nodules <1.2 cm high and 1.4 cm wide with halos around nodules; calcium carbonate nodules <0.4 cm high and <0.3 cm wide; INTERPRETATIONS: illuviation; wetting and drying periods; vegetated | 285 HARDPAN CREEK NORTH: 97–106 cm: Thin section 26: base of profile, laminated sediments; size: 12.0 cm by 5.8 cm wide (69.6 sq. cm); thickness ~30–40 µm MICROSTRUCTURE AND POROSITY: Vughy to channel microstructure; channels <2 cm long and 0.1 cm wide, serrated to smooth accommodating planes; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 41/59; clay-loam; sub-angular to sub-rounded; strongly oriented in a laminar pattern; poorly-sorted; chitonic to gefuric; Colour: orange/brown/black PPL; orange/brown/red CPL; Coarse Fraction: coarse sand (1000–2000µm) = 40%; medium sand (200–500µm) = 40%; fine sand (100–200µm) = 1%; Fine Fraction:very fine sand (20–100µm) = 0%; silt (2–20µm) = 5%; clay (<2µm) = 54%; moderately oriented clays and silts; ORGANIC COMPONENTS: <40% phytoliths randomly distributed throughout groundmass; PEDOFEATURES: Textural: abundant dusty clay coatings of <75% of sand grains, and moderate to strong birefringence, orange, brown to black; moderate to strong sesquioxide staining of the fine fabric; strong fracturing and dispersal of the clay coatings; silt cappings on coarse fraction; Amorphous: <50% of Fe/Mn-oxide nodules; <40% Fe/Mn-oxide nodules throughout groundmass; partial to complete voids and channels filled with dusty clays and CaCO3 coatings; calcium carbonate nodules <0.8 cm long and <0.6 cm wide; intrapedal diffuse Mn-hydroxide nodules <0.6 cm long and 0.4 cm wide; porous magnetite nodule 0.3 cm; INTERPRETATIONS: colluvium; humid conditions; post-deposition; frost heave; vegetation; 286 | Micromorphology descriptions and Opal phytoliths micrographs HARDPAN CREEK SOUTH: 0–8 cm: Thin section 27: Upper soil layer; Size: 11.4 cm by 5.7 cm wide (64.98 sq. cm); thickness ~30–40 µm MICROSTRUCTURE AND POROSITY: sub-angular blocky microstructure with compound packing voids; heterogeneous ped sorting <2.8 cm long and <2 cm wide; channels 4.5 cm long and 0.2 cm wide; vughy with coatings in scattered locations; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 45/65; clay-loam; open porphyric; sub-angular to sub-rounded; strong vertical orientation with channels; heterogeneous; Colour: grey/orange PPL, grey/brown CPL; Coarse Fraction: coarse sand (1000–2000µm) = 5%; medium sand (200–500µm) = 10%; fine sand (100–200µm) = 30%; Fine Fraction: very fine sand (20–100µm) = 0%; silt (2–20µm) = 20%; clay (<2µm) = 45%; ORGANIC COMPONENTS: <40% humus, dark-coloured and coating the quartz grains; <50% phytoliths and organic matter within fills, channels and voids, dark brown to black; micro-charcoal; rare sponge spicules; PEDOFEATURES: Textural: <20% clay coatings of the sand grains; irregular zones of groundmass intermixed with CaCO3; bridged to partial coatings with clays and groundmass; weak to moderate birefringence; moderate to strong sesquioxides staining of the small amount of clays; dusty clays discontinuous, birefringent, oriented top to bottom; organic matter fills the channels and voids; weak crescentic fills of the voids with organic matter and sesquioxide-stained clays; Amorphous: strong CaCO3 nodules and inflls of the groundmass; charcoal pieces within an organic fabric; diffuse Mn-hydroxide nodules; charcoal pieces; INTERPRETATIONS: humid conditions; pedogenesis; natural and/or human fires; grasses; | 287 HARDPAN CREEK SOUTH (OSL): Thin section 1: Size: 11.0 cm by 5.7 cm wide (62.7 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: Vughy to channel microstructure; channels <3.6 cm long and <0.1 cm wide, serrated to smooth edges; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 50/50; sandy-clay; sub-angular to sub-rounded; weakly oriented; poorly sorted; open porphyric; Colour: grey/brown PPL, orange/brown/black CPL; Coarse Fraction: coarse sand (1000–2000µm) = 10%; medium sand (200–500µm) = 30%; fine sand (100–200µm) = 10%; Fine Fraction: very = 5%; silt (2–20µm) = 10%; clay (<2µm) = 35%; ORGANIC COMPONENTS: Wood fragments in wash zone (bottom) Fe-oxide discolouration (red-orange); <20% phytoliths intermixed with the dusty clays and within voids; <10% micro-charcoal in the groundmass particularly in washed clay zones; <1% sponge spicules within dusty clays; PEDOFEATURES: Textural: channel coatings and infillings with illuvial clay; several stages of formation including CaCO3 nodules to laminar deposits of Fe- Mn-oxide nodules; <20% silt cappings; <40% pure clay coatings of sand grains, light brown to orange in colour; irregular zones of laminar clays to hints of reticulate striations of the groundmass; bridged to clay coatings; medium to strong birefringence; sesquioxides in the coatings and groundmass; Amorphous: near complete to complete infills of the channels with CaCO3 and surface materials; moderate sorting of CaCO3/Fe-Mn-oxide nodules; Fe- Mn-oxide <1.2 cm wide and <1.5 cm high, and CaCO3 nodules <0.5 cm wide and <0.4 cm high in specific layers; INTERPRETATIONS: humid conditions; rainfall and runoff; laminar deposits; saturated soils; grassed environment; 288 | Micromorphology descriptions and Opal phytoliths micrographs HARDPAN CREEK CENTRE: Thin section 23: top of buried feature; Size: 11.0 cm by 5.8 cm wide (63.8 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: Channel to vughy microstructure; channels 5.6 cm long and <0.1 cm wide, serrated to smooth and accommodating; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 50/50; sandy-clay; close porphyric; sub-angular to sub-rounded; weakly oriented; poorly sorted; Colour: grey, brown, orange PPL, brown, orange, yellow CPL; Coarse Fraction: coarse sand (1000–2000µm) = 10%; medium sand (200–500µm) = 30%; fine sand (100–200µm) = 10%; Fine Fraction: very fine sand (20–100µm) = 5%; silt (2–20µm) = 10%; clay (<2µm) = 35%; ORGANIC COMPONENTS: <1% phytoliths; <1% micro-charcoal <1 mm long distributed throughout the clay groundmass and in the vertically filled channels; PEDOFEATURES: Textural: strong bridging of the sand grains, close porphyric; dominant dusty clays with strong sesquioxides staining of the fine fraction, weak to medium birefringence, yellow-orange-dark brown; bridged to complete infills, CaCO3 to dusty clay infills of vertical channels, clean clays fill the 15 deg. channels; voids filled with surface materials; reticulate striations in the groundmass; Amorphous: CaCO3 coatings around voids; illuvial clay infillings; calcium carbonate nodules <1.5 cm long and 0.5 cm high; diffuse Mn-hydroxide nodules <0.3 cm; INTERPRETATIONS: illuvial environment; unstable conditions; pedogenesis; post-deposition; | 289 DOUBLE WELL and TUNKANA WELL SITES DOUBLE WELL: 1–10 cm: Thin section 21: upper horizon of arroyo—eroded and bare surface, possible floodplain; thin section 21; Size: 9 cm by 5.8 cm wide (52.2 sq. cm); thickness ~40–60 µmMICROSTRUCTURE AND POROSITY: Lenticular plates: <4.2 cm long and <2.5 cm wide, some angular blocky peds <2.5 cm long and <0.4 cm wide; channels <0.8 cm long and <0.15 cm wide; crusts line the top margins of the peds and range from <3.0 cm long and <0.1 cm wide; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 40/60; sandy-silt-loam; close porphyric; sub-angular; weakly oriented; well-sorted; Colour: light to dark brown PPL; and dark brown to orange CPL; Coarse Fraction: coarse sand (1000–2000µm) = 20%; medium sand (200–500µm) = 10%; fine sand (100–200µm) = 10%; Fine Fraction:very fine sand (20–100µm) = 25%; silt (2–20µm) = 15%; clay (<2µm) = 20%; ORGANIC COMPONENTS: <10% phytoliths in the groundmass and infills; PEDOFEATURES: Textural: abundant dusty clays of the sand grains <60%; irregular zones of groundmass; minor pure clays in channels; dusty clays, clay coatings; partial to complete infills and coatings of channels and voids; hyper-coatings of the voids (dark browns); abundant surface crusts; hypocoatings; weak birefringence, brown to orange; some soil surface peds; sesquioxides stain the ground mass; Amorphous: CaCO3 crusts <3.0 cm long and 0.15 wide; intrapedal diffuse Mn-hydroxide nodules <0.15 cm; INTERPRETATIONS: alluvial origins; wet and dry phases; periodic flooding; vegetation; 290 | Micromorphology descriptions and Opal phytoliths micrographs DOUBLE WELL: Top–Mid: Thin section 22: top of the gravel lens; Size: 8.1 cm by 5.8 cm wide (52.2 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: A mix of angular blocky, sub-angular blocky, prism peds <2.5 cm long and 1.6 cm high; and a large yellow to brown in colour sub-angular blocky nodule, 2.0 cm – 1.8 cm; channels <3 cm long and 0.25 cm wide; highly separated compact packing voids with some accommodating planes; large peds; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 35/65; sandy-silt-loam; close porphyric; sub-angular to sub-rounded; weakly oriented; poorly to moderately sorted; Colour: brown/orange PPL, brown/orange/yellow CPL; Coarse Fraction: coarse sand (1000–2000µm) = 5%; medium sand (200–500µm) = 10%; fine sand (100–200µm) = 20%; Fine Fraction: very fine sand (20–100µm) = 20%; silt (2–20µm) = 10%; clay (<2µm) = 35%; ORGANIC COMPONENTS: <15% phytoliths located in groundmass and infills PEDOFEATURES: Textural: silt cappings <50% and dusty clay in between sand grains; irregular groundmass depletion; coarse sand fraction at the top and fine fabric at the base; complete to bridged infills of the channels and voids; weak to moderate birefringence; strong sesquioxide staining (dark brown); partial to near complete infills of voids and channels; weak pure clays; mineral grains capped with redoximorphic silts; Amorphous: calcium carbonate peds; intrapedal clay coatings, large intrapedal diffuse Mn-hydroxide nodules intrapedal INTERPRETATIONS: alluvial origins; buried alluvial surface; groundwater fluctuations wetting and drying; frosty conditions; | 291 DOUBLE WELL: Middle Layer: Thin section 15: middle layer of arroyo near gravel layer; Size: 11.8 cm by 5.8 cm wide (68.4 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: Channel to vughy microstructure; channels <3.6 cm long and 0.25 cm wide; accommodating channels; vughy; compound packing voids; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 25/75; sandy-clay; open porphyric; sub-rounded; weakly oriented; well-sorted; Colour: brown PPL, orange/brown/yellow CPL; Coarse Fraction: coarse sand (1000–2000µm) = 0%; medium sand (200–500µm) = 5%; fine sand (100–200µm) = 20%; Fine Fraction: very fine sand (20–100µm) = 10%; silt (2–20µm) = 1%; clay (<2µm) = 64%; ORGANIC COMPONENTS: <30% phytoliths; PEDOFEATURES: Textural: sub-angular blocky peds <2.8 cm long and <1.2 cm wide; surface crusts <1.3 cm long and 0.15 cm wide; continuous to partially filled voids with pure to dusty clays; partial to complete coatings and infills of the voids, light browns, <10% dark brown; moderate to strong sesquioxides stain the groundmass; medium to strong birefringence of the clays; pure clay infills of the vertical and horizontal channels, light brown; strong crusts of the peds in the middle section, dark brown; silt cappings on the sand grains; irregular zones of groundmass; Amorphous: Fe- to Mn-impregnated surfaces; intrapedal diffuse Fe- and Mn-hydroxide nodules, one nodule in void coated with clays; CaCO3 absent; INTERPRETATIONS: alluvial origin; former floodplain; intermittent flooding; groundwater fluctuations/oxidation and translocation of the fines; illuviation; vegetation; 292 | Micromorphology descriptions and Opal phytoliths micrographs DOUBLE WELL: Mid–Low: Thin section 12: below gravel layer; Size: 12.7 cm by 5.8 cm wide (73.6 sq. cm); thickness 40–60 µm MICROSTRUCTURE AND POROSITY: Granular to vughy microstructure; channels <1.5 cm long and <0.2 cm wide, serrated to smooth; accommodating planes; intrapedal channels and chambers; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 30/70; sandy-clay; sub-angular; weakly oriented; poorly to moderately sorted; porphyric; Colour: brown/orange PPL, brown/orange/yellow CPL; Coarse Fraction: coarse sand (1000–2000µm) = 10%; medium sand (200–500µm) = 10%; fine sand (100–200µm) = 10%; Fine Fraction: very fine sand (20–100µm) = 20%; silt (2–20µm) = 10%; clay (<2µm) = 40%; ORGANIC COMPONENTS: a wood fragment; <40% phytoliths within groundmass some burned; PEDOFEATURES: Textural: granular peds <4.1 cm high and <2.6 cm wide; <50% dusty clay coatings and irregular zones of groundmass, light brown; intrapedal bridged clay coatings; pure clay coatings of the vertical channels, light browns; moderate to strong birefringence, dark browns; moderate to strong sesquioxides staining the groundmass, dark browns to black; <45% dusty to pure clays discontinuous to complete; reticulate stratification in the upper and lower zones; Amorphous: intrapedal diffuse Mn-hydroxide nodules and features toward the base of the thin section; Crystalline: CaCO3 nodules within voids and mixed with groundmass; INTERPRETATIONS: alluvial origin; wet/humid conditions (sponge spicules); groundwater fluctuation/oxidisation and translocation of the fines; former floodplain; grassed environment; natural and/or human fires; | 293 DOUBLE WELL: 162–172 cm: Thin section 16: base of the arroyo profile; Size: 12.3 cm by 5.8 cm wide (71.34 sq. cm); thickness ~40-60 µm MICROSTRUCTURE AND POROSITY: Channel to vughy microstructure; channels <3 cm long and 0.25 cm wide, serrated to smooth; accommodating planes; crusts (2.4 cm long – 0.2 cm high); MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 60/40; loamy-sand; sub-angular to sub-rounded; weakly oriented; poorly to moderately sorted; close porphyric; Colour: brown PPL, brown/orange CPL; Coarse Fraction: coarse sand (1000–2000µm) = 20%; medium sand (200–500µm) = 10%; fine sand (100–200µm) = 30%; Fine Fraction: very fine sand (20–100µm) = 20%; silt (2–20µm) = 5%; clay (<2µm) = 15%; ORGANIC COMPONENTS: wood fragments; <25% phytoliths, molten, burned, good condition; PEDOFEATURES: Textural: abundant clay coatings of <50% to abundant silt cappings and coatings of the sand grains and thus irregular zones of groundmass, red to light browns; moderate to strong birefringence; sesquioxide staining of the fine fabric; minimal dusty to pure clays infills; some reticulate striations light to dark brown; Amorphous: stabilised CaCO3 flows within the voids that transform to Fe/Mn-oxide nodules at the base of the channel and CaCO3 flow; diffuse Mn-hydroxide nodules, 0.3 cm long and 0.6 wide; calcium carbonate nodules <0.6 cm long and <0.4 cm wide; diffuse Mn-hydroxide nodules, 0.3 cm long and 0.6 wide; INTERPRETATIONS: alluvial origin; limited groundwater fluctuations and translocation of the fines; stability in the coarse-fine fractions; Fe- and Mn-oxide formation; Organics: phytoliths; illuviation; frosts; former alluvial sediments buried by other alluvial events (above); wet or humid environment, reeds growing near a water source. Some reticulate striations infers stability and pedogenesis; 294 | Micromorphology descriptions and Opal phytoliths micrographs TUNKANA WELL: Mid–Low: Thin section 14: mid-layer of profile; Size: 10.8 cm by 5.8 cm wide (62.64 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: Channel microstructure; channels <1.7 cm long and 0.2 cm, serrated, compound packing voids, vughy; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 25/75; clay-loam; sub-angular to sub-rounded; weakly oriented; poorly to moderately sorted; open porphyric; poorly to moderately sorted; Colour: light brown PPL, brown/orange CPL; Coarse Fraction: coarse sand (1000–2000µm) = 0%; medium sand (200–500µm) = 5%; fine sand (100–200µm) = 20%; Fine Fraction: very fine sand (20–100µm) = 5%; silt (2–20µm) = 20%; clay (<2µm) = 50%; ORGANIC COMPONENTS: <75% organics throughout groundmass and soil fragments; <50% phytoliths, mixed shapes, molten, burned; PEDOFEATURES: Textural: dusty clay coatings <30% of sand grains; irregular zones of the groundmass (possibly too thin to see obtain a true picture of clay coatings?); weak to moderate birefringence; clay halos around large mineral grains; hints of reticulate striation; soil fragments; Amorphous: groundmass mostly stained with CaCO3; complete to partial infills of the channels and voids; diffuse Mn-hydroxide nodules<0.2 cm; calcium carbonate nodules <0.4 cm; INTERPRETATIONS: alluvial origins; movement of fines to form crusts; groundwater fluctuations and translocation of the fines; movement of fine fractions by groundwater (eluviation); Fe- and Mn-oxide formation; natural fires or possible human fires?; buried alluvial surface; grassed environment; freshwater conditions | 295 TUNKANA WELL: 138–144 cm: Thin section 13: lower layer of profile; Size: 12.3 cm by 5.8 cm wide (71.34 sq. cm); thickness ~40–60 µm MICROSTRUCTURE AND POROSITY: Channel to vughy microstructure; channels <1.5 cm long and 0.15 cm wide, serrated, accommodating; vughs <0.4 cm, 20%; MINERAL COMPONENTS: Groundmass: c/f ratio100µm: 60/40; sandy-clay; sub-angular to sub-rounded; weakly oriented; poorly to moderately sorted; close porphyric; Colour: Coarse Fraction: fine gravel = 30%; coarse sand (1000–2000µm) = 20%; medium sand (200–500µm) = 10%; fine sand (100–200µm) = 0%; Fine Fraction:very fine sand (20–100µm) = 0%; silt (2–20µm) = 5%; clay (<2µm) = 35%; ORGANIC COMPONENTS: <5% phytoliths; organic matter not visible; PEDOFEATURES: Textural: dark brown clay coatings and dusty clays infill the compound packing voids and around the large mineral grains, moderate birefringence; irregular zones of the grounds in between large sesquioxide staining of the groundmass, dark browns; dark brown clay staining of the minerals; Amorphous: large Fe/Mn-hydroxide nodules; calcium carbonate nodules <0.4 cm; Fe/Mn-hydroxide nodules; voids completely infilled; INTERPRETATIONS: humid and arid phases; unstable conditions; buried alluvial surface; groundwater fluctuations and translocation of the fines; ground and canopy cover 296 | Micromorphology descriptions and Opal phytoliths micrographs 2. Opal phytolith micrographs (a) (b) Figure C.1: BG1: a) Elongate; b) sponge spicule (a) (b) (c) (d) (e) Figure C.2: BG2: a) Elongate; b) Acute bulbosus; c) Acute bulbosus; d) Acute bulbosus; e) Acute bulbosus | 297 (a) (b) (c) (d) (e) (f) Figure C.3: GE0–7 cm: a) Blocky (burned); b) Elongate and Acute bulbosus (burned); c) Elongate entire; d) Elongate entire/sinuate; e) Acute bulbosus; f) Acute bulbosus (a) (b) (c) (d) (e) (f) Figure C.4: GE30 cm and GE40 cm: a) unknown; b) Elongate; c) unknown; d) unknown; e) Elongate sinuate; f) Elongate and Blocky 298 | Micromorphology descriptions and Opal phytoliths micrographs (a) (b) (c) (d) (e) (f) Figure C.5: GE40 a) Blocky; b) Blocky; c) Blocky - diatom in background; d) Elongates (broken) and micro-charcoal; GE57: e) Unknown; f) Blocky (a) (b) (c) (d) (e) (f) (g) (h) Figure C.6: GD X-layer: a) Sponge spicule, GD Static: b) Blocky; c) Spheroidal (areolate) (Cucurbita); d) Spheroidal; e) Spheroidal (areolate) (Cucurbita); f) Elongates (leaf); g) Elongate entire; h) Organics | 299 (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) (l) (m) (n) (o) (p) Figure C.7: Hardpan Creek Mulga: a) Elongate (burned); b) Elongate; c) Arcuate (left) and Fusiform (right); d) Elongate (Poaecae); e) Elongate; f) Elongate; g) Elongate entire (Poaecae); h) Blocky; i) Elongate (Poacae); j) Elongate (damaged); k) Elongate; l) Blocky; m) Elongate (broken) (Poaecae); n) Blocky; o) Polylobate; p) Globular spheroid 300 | Micromorphology descriptions and Opal phytoliths micrographs (a) (b) (c) (d) (e) (f) (g) Figure C.8: HCKN 2–9 cm layer: a) Blocky; b) Elongate entire; c) Elongate; d) Elongate; e) Acute bulbosus; f) Acute bulbosus; g) Bulliform flabellate (Aristida spp.) (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) Figure C.9: Hardpan Creek 29-38 cm: a) Blocky; b) Elongate dentate; c) Elongate dentate; d) Blocky (burned); e) Unknown, possible Cosmarium spp. or algae; f) Blocky; g) Globular; h) Blocky; i) Blocky; j) Blocky; k) Elongate; l) Elongate entire | 301 (a) (b) (c) (d) (e) (f) (g) (h) Figure C.10: HCKN OSL (45 cm) layer: a) Honey comb, Moraceae, possibly Ficus spp.; b) Testate amoebae; c) Elongate dentate; d) Hair phytolith, Elongate cluster; e) Elongate entire; f) Elongate (burned); g) Elongate; h) fungus micro-fossil (a) (b) (c) Figure C.11: HCKN 90 cm: a) Blocky (broken); b) Elongate entire; c) Blocky 302 | Micromorphology descriptions and Opal phytoliths micrographs (a) (b) (c) (d) (e) (f) (g) (h) (i) Figure C.12: HCKN 97 cm: a) Blocky; b)Blocky c) Blocky; d) Elongate sinuate; e) Elongate sinuate; f) Elongate (burned); g) Elongate entire; h) Elongate entire and Spheroid; i) Elongate (a) (b) (c) Figure C.13: HCKN97 cm: a) to c) fungus micro-fossils (a) (b) (c) Figure C.14: Hardpan Creek South 0–8 cm: a) Blocky (burned); b) Blocky; c) Elongate | 303 (a) (b) (c) (d) Figure C.15: Hardpan Creek Centre: a) Spheroid; b) Arcuate; c) Bulliform flabellate (Aristida spp.); d) Elongate geniculate (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) Figure C.16: Hardpan Creek South (OSL): a) Elongate; b) Elongate entire (broken); c)Elongate; d) Elongate dentate and Blocky; e) Blocky; f) Blocky; g) Blocky; h) Blocky; i) Blocky (pitted); j) Bilobate (Themeda spp.); k) micro-charcoal pieces 304 | Micromorphology descriptions and Opal phytoliths micrographs (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) Figure C.17: DW 1–10 cm: a) Elongate entire (Poaecae); b) Elongate entire (Poaecae); c) Elongate (burned); d) Elongate (burned) and Elongate; e) Elongate; f) Arcuate bulbosus, Blocky and Elongate; g) Blocky (pitted); h) Blocky; i) Unknown; j) Elongate entire (organic matter) (a) (b) (c) (d) (e) (f) Figure C.18: Double Well Top/Middle: a) Acute bulbosus (pitted); b) Elongate entire (burned); c) Elongate entire; d) Elongate; e) Elongate; f) Elongate entire (burned) | 305 (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) Figure C.19: Double Well Middle: a) Polylobate; b) Dentate; c) Elongate; d) Acute bulbosus; e) Spheroid echinate (Arecaceae); f) unknown; g) Diatoms; h) Dentate; i) unknown; j) unknown 306 | Micromorphology descriptions and Opal phytoliths micrographs (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) Figure C.20: DWMid-Low: a) Elongate entire (Poaceae); b) Elongate entire (Poaceae); c)Blocky; d) Blocky (broken); e) Elongate; f) Blocky; g) Blocky; h) Bulliform flabellate; i) Elongate; j) Arcuate bulbosus; k) sponge spicule (a) (b) (c) (d) (e) (f) (g) (h) Figure C.21: Double Well 162 cm (scale=100µm): a) Elongate, possibly Phragmites australis or reeds; b) Elongate entire; c) unknown; d) unknown; e) Tabular; f) Acute bulbosus; g) trough (Chen and Smith 2013: 3–4); and h) conical-single-perforate or cavate (Araucaria cunninghamii) (Parr 2006: 83) | 307 (a) (b) (c) (d) (e) (f) (g) Figure C.22: TW 138 cm layer: a) Acute bulbosus; b) Blocky; c) unknown; d) Elongate entire; e) unknown; f) Spheroid areolate; g) uknown cluster