A LATE QUATERNARY PALAEOENVIRONMENTAL INVESTIGATION OF THE FIRE, CLIMATE, HUMAN AND VEGETATION NEXUS FROM THE SYDNEY BASIN, AUSTRALIA

Manu P. Black

A thesis submitted under the requirements for the degree of Doctor of Philosophy, School of Biological, Earth and Environmental Sciences, University of New South Wales, Sydney, Australia.

October 2006

i

Abstract

It is widely believed that Australian Aboriginals utilised fire to manage various landscapes however to what extent this impacted on Australia’s ecosystems remains uncertain. The late Pleistocene/Holocene fire history from three sites within the Sydney Basin, Gooches Swamp, Lake Baraba and Kings Waterhole, were compared with archaeological and palaeoclimatic data using a novel method of quantifying macroscopic charcoal, which is presented in this study. The palynology and other palaeoecological proxies were also investigated at the three sites. The Gooches Swamp fire record appeared to be most influenced by climate and there was an abrupt increase in fire activity from the mid- Holocene perhaps associated with the onset of modern El Niño dominated conditions. The Kings Waterhole site also displayed an abrupt increase in charcoal at this time however there was a marked decrease in charcoal from ~3 ka. Lake Baraba similarly had displayed low levels of charcoal in the late Holocene. At both Kings Waterhole and Lake Baraba archaeological evidence suggests intensified human activity in the late Holocene during this period of lower and less variable charcoal. It is hence likely that at these sites Aboriginal people controlled fire activity in the late Holocene perhaps in response to the increased risk of large intense fires under an ENSO-dominated climate. The fire history of the Sydney Basin varies temporally and spatially and therefore it is not possible to make generalisations about pre-historic fire regimes. It is also not possible to use ideas about Aboriginal fire regimes or pre-historic activity as a management objective. The study demonstrates that increased fire activity is related to climatic variation and this is likely to be of significance under various enhanced Greenhouse scenarios. There were no major changes in the composition of the flora at all sites throughout late Pleistocene/Holocene although there were some changes in the relative abundance of different taxa. It is suggested that the Sydney Sandstone flora, which surrounds the sites, is relatively resistant to environmental changes. Casuarinaceae was present at Lake Baraba during the Last Glacial Maximum and therefore the site may have acted as a potential refugium for more mesic communities. There was a notable decline in Casuarinaceae during the Holocene at Lake Baraba and Kings Waterhole, a trend that has been found at a number of sites from southeastern Australia. ii

Acknowledgements Big thanks to Scott Mooney, my supervisor and friend, who has been a great source of inspiration, guidance and support over the years. Thanks to other staff members from UNSW for providing help over the years especially Helene Martin, David Edwards, Geoff Hyde and Paul Adam.

I appreciate the very helpful discussions with Simon Haberle (Australian National University, Canberra) and thank you Simon for the comments on my final draft. David Keith (National Parks and Wildlife Service, Sydney) and Marshall Wilkinson (Macquarie University, Sydney) also provided valuable advice. Thanks to NSW National Parks and Wildlife for allowing access to the sites.

Discussions with archaeologists Val Attenbrow, Robin Terrance and Peter Hiscock (Australian Museum, Sydney) helped clarify some important issues regarding past Aboriginal activity. I would also like to thank Rick Battarbee, John Dearing, Isabelle Larocque and Frank Oldfield for inviting Scott and I to contribute to the PAGES special issue to appear in Regional Environmental Change.

Thanks to the many people I shared an office with over the years but especially to Emma Haradasa and Joelle Gergis. Thanks to my friends Chris, Joe, Georgia and Emma who I was able to rope in to trudge through swamps with me… Sleeping in a cave for a week during the middle of winter in the Blue Mountains really ain’t that bad?

Much love to my friends and my dear family, Mum, Phil, Suriya and Jayanthi who, although I don’t really think they really understood what I was doing, were happy to be supportive and love me anyways. I would like to dedicate this thesis to my two grandmothers, Ameni and Thelma, who both passed away earlier this year.

Finally I would like to acknowledge the Traditional Owners of the land on whose past activity I have been investigating for the past several years. iii

Table of Contents

PAGE NUMBER

ABSTRACT i

ACKNOWLEDGEMENTS ii

LIST OF TABLES vii

LIST OF FIGURES viii

LIST OF ABBREVIATIONS xi

CHAPTER 1 1

INTRODUCTION

1.1 FIRE AND THE AUSTRALIAN ENVIRONMENT 1

1.2 A HISTORY OF FIRE IN AUSTRALIA 3

1.3 ABORIGINAL USE OF FIRE DURING THE LATE QUATERNARY 5

1.4 EUROPEAN IMPACTS TO FIRE REGIMES 8

1.5 THE SYDNEY BASIN 10

1.5.1 GEOMORPHOLOGY 10 1.5.2 CLIMATE 12 1.5.3 VEGETATION 13 1.5.4 ABORIGINAL HISTORY 14

1.6 LATE QUATERNARY CLIMATE CHANGE 19

1.7 RECONSTRUCTING FIRE HISTORY 24

1.8 REVIEW OF PREVIOUS PALAEOSTUDIES IN THE SYDNEY BASIN 27

1.9 THESIS AIM 31

1.10 THESIS OUTLINE 31

CHAPTER 2 34 A SIMPLE AND FAST METHOD FOR CALCULATING THE AREA OF MACROSCOPIC CHARCOAL ISOLATED FROM SEDIMENTS

2.1 INTRODUCTION 34

2.2 THE METHOD 35 2.2.1 STEP 1. ACQUIRING A DIGITAL IMAGE 36 iv

2.2.2 STEP 2. FORMATTING THE IMAGE FOR PROCESSING 36 2.2.3 STEP 3. CALIBRATION 36 2.2.4 STEP 4. SETTING MEASUREMENT PARAMETERS 37 2.2.5 STEP 5. ANALYSING AND SHOWING RESULTS 37

2.3 CONCLUDING REMARKS 38

CHAPTER 3 40 HOLOCENE FIRE HISTORY FROM THE GREATER BLUE MOUNTAINS WORLD HERITAGE AREA, NEW SOUTH WALES, AUSTRALIA: THE CLIMATE, HUMANS AND FIRE NEXUS.

ABSTRACT 40

3.1 INTRODUCTION 41

3.2 STUDY AREA 43

3.3 METHODS 45

3.4 RESULTS 48 3.4.1 CORE STRATIGRAPHY AND CHRONOLOGY 48 3.4.2 CHARCOAL AND OTHER ANALYSES 48 3.4.3 POLLEN 49

3.5 DISCUSSION 51

3.6 CONCLUSION 59

CHAPTER 4 61

THE ARCHAEOLOGICAL AND CLIMATIC IMPLICATIONS OF A 14 200 YR CONTIGUOUS FIRE RECORD FROM GOOCHES CRATER, BLUE MOUNTAINS, AUSTRALIA

ABSTRACT 61

4.1 INTRODUCTION 62 4.1.1 POSTGLACIAL CHANGES IN AUSTRALIAN ABORIGINAL TECHNOLOGIES (AND THEIR RELEVANCE TO POPULATION CHANGES) 64 4.1.2 CLIMATE CHANGE SINCE THE LGM 65

4.2 SITE DESCRIPTION 67

4.3 METHODOGY 70

4.4 RESULTS 71 4.4.1 CORE STRATIGRAPHY AND CHRONOLOGY 71 4.4.2 MACROSCOPIC CHARCOAL 72 4.4.3 HUMIFICATION, LOSS OF IGNITION AND GEOCHEMISTRY 73

4.5. DISCUSSION 75 v

4.5.1 THE LATE PLEISTOCENE-EARLY HOLOCENE (~14.2 TO 9.1 KA) 78 4.5.2 THE EARLY-TO-MID-HOLOCENE (9.1 TO 5.7 KA) 79 4.5.3 THE MID-TO-LATE-HOLOCENE (FROM 5.7 KA) 80

4.6 CONCLUSIONS 84

CHAPTER 5 86

A >43 000 YEAR VEGETATION AND FIRE HISTORY FROM LAKE BARABA, NEW SOUTH WALES, AUSTRALIA

ABSTRACT 86

5.1. INTRODUCTION 87

5.2. THE ENVIRONMENT 89

5.3. METHODOLOGY 90

5.4. RESULTS 93 5.4.1. CORE STRATIGRAPHY AND CHRONOLOGY 93 5.4.2. POLLEN 94 5.4.3. MACROSCOPIC CHARCOAL 97

5.5. DISCUSSION 98 5.5.1. SEDIMENTARY HISTORY 98 5.5.2. THE VEGETATION HISTORY 100 5.5.3 FIRE HISTORY 104

5.6. CONCLUSION 107

CHAPTER 6 109

A ~6,100 YR VEGETATION AND FIRE HISTORY FROM KINGS WATERHOLE, WOLLEMI NATIONAL PARK, NEW SOUTH WALES

ABSTRACT 109

6.1 INTRODUCTION 110

6.2 SITE DESCRIPTION 113

6.3 METHOD 115

6.4 RESULTS 116

6.5 DISCUSSION 119

6.6 CONCLUSIONS 125

vi

CHAPTER 7 127

THE FIRE, HUMANS AND CLIMATE NEXUS DURING THE LATE QUATERNARY IN THE SYDNEY BASIN, AUSTRALIA

ABSTRACT 127

7.1. INTRODUCTION 128

7.2. SITE DESCRIPTION AND SELECTION 132

7.3 METHOD 138

7.4 RESULTS 140 7.4.1 STRATIGRAPHY AND CHRONOLOGY 140 7.4.2 CHARCOAL ANALYSIS 141 7.4.3 STATISTICAL ANALYSIS 142 7.4.4 COMPARISON TO ARCHAEOLOGICAL RECORDS 145 7.4.5 COMPARISON WITH CLIMATIC DATA 147

7.5 DISCUSSION 149

7.6 CONCLUSION 154

CHAPTER 8 156

SUMMARY OF CONCLUSIONS

8.1 NEW METHOD FOR ANALYSING CHARCOAL 156

8.2 THE LATE QUATERNARY VEGETATION HISTORY OF THE SYDNEY BASIN 156

8.3 THE LATE QUATERNARY FIRE HISTORY OF THE SYDNEY BASIN 157

REFERENCES 159

vii

List of Tables

Table 3.1: The rationale for quantifying the target palynomorphs…………………....47

Table 3.2: Radiocarbon dates from GCR……………………………………………...48

Table 4.1: Basal dates and artefact accumulation through time for published archaeological sites located in the Blue Mountains……………………...... 69

Table 4.2: Radiocarbon dates and calibration for GCR sediments……………………72

Table 5.1: The pollen and spores quantified in the Lake Baraba sediment, and their indicative value…………………………………………………………….92

Table 5.2: Radiocarbon dates and calibration for Lake Baraba sediments……………93

Table 5.3: A comparison between the average percentages for the south-eastern mainland pollen data-set for major taxa at selected time-slots from Kershaw (1995: 661) with the pollen record from Lake Baraba……………………101

Table 5.4: The key Australasian sites that have a record of charcoal throughout the Last Glacial Maximum…………………………………………………………105

Table 6.1: Radiocarbon dates and calibration for Kings Waterhole…………………116

Table 7.1: Radiocarbon dates and calibration for Gooches Swamp, Lake Baraba and Kings Waterhole sediments……………………………………………….140

Table 7.2: Statistical analysis results for the three sites……………………………...143

Table 7.3: Results of the correlation for the three sites………………………………143

viii

List of Figures

Figure 1.1: The major geomorphic units of the Sydney Basin and the location of the three sites………………………………………………………………...... 11 Figure 1.2: The territory of the different Aboriginal language groups and the key archaeological sites in the Sydney Basin………………………………...... 18 Figure 1.3: Location of palaeoenvironmental sites within the Sydney Basin………... 28 Figure 2.1: The tool-box in Scion Image with the necessary tools highlighted……...... 37 Figure 2.2: A comparison of the count and image analysis-derived charcoal data from the Gooches Crater (Right) site………………………………………39 Figure 3.1: The location of Gooches Crater Right (GCR)…………………………...... 44 Figure 3.2: Charcoal analysis results and the fire index (GCR)……………………...... 49 Figure 3.3: Palynological results (GCR)……………………………………………… 50 Figure 3.4: Annotated charcoal diagram, depicting possible influences on past fire activity at GCR…………………………………………………………….52 Figure 4.1: The location of Gooches Crater Right……………………………………..67 Figure 4.2: The results of the charcoal analyses, palynology, humification and loss-on- ignition at Gooches Crater Right…………………………………………..73 Figure 4.3: The results of the geochemical analyses at Gooches Crater Right………...74 Figure 4.4: The charcoal concentration expressed against time, highlighting the difference between the Late Pleistocene, early-to-mid and mid-to-late Holocene…………………………………………………………………...76 Figure 4.5: A comparison of the composite charcoal results from several sites in tropical Sahul, from Haberle et al. (2001), and the results from Gooches Crater Right………………………………………………………………………..77 Figure 4.6: Artefact discard rates (Capertee 3 site) versus average charcoal concentration (GCR) throughout the Holocene……………………………84 Figure 5.1: Location of Lake Baraba………………………………………………….. 89 Figure 5.2: Pollen, charcoal, organic matter and pollen zonation for Lake Baraba……94 ix

Figure 5.3: Macroscopic charcoal curve versus age for the radiocarbon dated section of the Lake Baraba sequence………………………………………………….97 Figure 5.4: A comparison between the rates of sedimentation, sediment type and the pollen zones………………………………………………………………..98 Figure 5.5: A comparison between the Myrtaceae/Casuarinaceae ratio and the organic content of the sediment from loss-on-ignition……………………………103 Figure 5.6: A comparison between the average charcoal values for Lake Baraba and the number of habitation sites used in the region for millennial time scales…106 Figure 6.1: Location of Kings Waterhole……………………………………………..114 Figure 6.2: Results of the palynological analysis of Kings Waterhole………………..117 Figure 6.3: Results of the charcoal analysis from Kings Waterhole………………….119 Figure 6.4: A comparison of the Upper Mangrove archaeological data (Attenbrow, 2003; 2004) and the Kings Waterhole charcoal curve……………………125 Figure 7.1: The major geomorphic units of the Sydney Basin and the location of the three sites……………………………………………………………….....133 Figure 7.2: The location of the territory of the major Aboriginal language Groups and the key archaeological sites in the Sydney Basin relevant to this study… 136 Figure 7.3: The results of the macroscopic charcoal analysis for Gooches Swamp, Lake Baraba and Kings Waterhole……………………………………………..141 Figure 7.4: The results of the spectral analysis (Fourier’s analysis). Results are significant at the 0.05 level……………………………………………….144 Figure 7.5: A comparison of the Gooches Swamp macroscopic charcoal curve with the number of backed and non-backed artefacts from Capertee 3……………145 Figure 7.6: A comparison of the Lake Baraba charcoal curve with a summary of the archaeological data from the southern Sydney Basin…………………….146 Figure 7.7: A comparison of the Kings Waterhole charcoal curve with the archaeological data from the Upper Mangrove Creek catchment………...147 Figure 7.8: a) Seasonality at 30oS based on the difference in insolation between summer and winter (Berger, 1992) b) The frequency of ENSO events per 100 yrs based on Moy et al. (2002) c) The climatic summary of south-eastern Australia (Kershaw et al., 2002; Lees, 1995; Shulmeister, 1999 etc) d) The x

smoothed Gooches Swamp charcoal record constructed by summing the 200 yr values e) The smoothed Kings Waterhole charcoal record constructed by summing the 200 yr values f) The smoothed Lake Baraba constructed by summing the 200 yr values……………………………………………… 148 Figure 7.9: A comparison of the composite charcoal results from several sites in tropical Sahul, from Haberle et al. (2001), and the charcoal curve from Gooches Swamp…………………………………………………………………… 151

xi

List of Abbreviations . km Kilometres ha Hectares μm Micrometres cm Centimetres mm Millimetres m Metres m asl Metres above sea level ka Thousand years ago BP Radiocarbon years before present (present taken as AD1950) cal. yr BP Calibrated radiocarbon years before present (present taken as AD1950) NSW New South Wales ENSO El Niño Southern Oscillation LGM Last Glacial Maximum MIS Marine Isotope Stage no. Number pa Per annum mg Milligrams

ppm Parts per million

L Litres

Chapter 1 1

Chapter 1: Introduction

1.1 Fire and the Australian Environment

Australian plants have adapted to increased aridity and low nutrient soils over the past 15 million years (Benson, 1988). The evolutionary changes associated with these adaptations have enabled most Australian plant communities to survive periodic burning, and in some cases, require fire to reproduce. Fire plays a crucial role for many of Australia’s terrestrial ecosystems (Benson, 1998) since it can either promote or adversely affect biodiversity (Fox and Fox, 1987; Bradstock, 1993; Conroy and Sanders, 1993; Williams et al., 1994; Gill and Bradstock, 1995; Williams and Gill, 1995; Keith, 1996; Bowman, 1998; McLoughlin, 1998; Keith et al., 2002).

A fire regime is composed of the following components: frequency, intensity, type and seasonality (Gill, 1975; 1977) and it has been shown that changing any of these components can affect species composition and biodiversity (e.g. Clark 1988; Cary & Morrison 1995; Morrison et al. 1995; Conroy 1996). The state of the vegetation affects fire, and the inter-fire interval and frequency regulates the composition, structure and quantity of living vegetation and dead fuels (Cheney, 1981; Sousa, 1984; Gill, 1996; Ogden et al., 1998).

The local fire regime of a plant community is determined by complex interrelations between its climate, topography, frequency and seasonality of ignition sources, fuel characteristics and cultural activities (Gill, 1977; Clark, 1983; Sousa, 1984; Cope and Chaloner, 1985; Clark and Robinson, 1993; Williams and Gill, 1995). These factors vary spatially and temporally, and hence there is considerable heterogeneity in local fire regimes and consequently the effects of fire on vegetation (Sousa, 1984; Williams and Gill, 1995). Fire regimes are also determined by the existing vegetation communities of a site and this is complicated since the plants supply the fuel for the fire which, in turn, affects the plants (Jackson, 1968; Gill, 1994). This complex interaction between fire, vegetation, humans and

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 2

climate is poorly understood in a modern ecological sense. The longer temporal perspective afforded by palaeoecological studies mean that they have a potentially important role in untangling this ‘nexus’.

Biotic responses to fire are complex (Conroy and Sanders, 1993; Whelan, 1995; Williams and Gill, 1995; McLoughlin, 1998; Whelan et al., 2002), however, individual plant species and communities are thought to be adapted to a particular fire regime or range of fire regimes (Gill, 1975; 1977; Fox and Fox, 1986; Williams and Gill, 1995; Bradstock et al., 1996; Clark, 1988; Dodson, 1994; Benson, 1991; Benson and Redpath, 1997).

In fire prone vegetation, survival and reproduction are closely related to the occurrence of fire, with the fire interval ultimately determining whether a species will persist at a site (Gill, 1977; Keith, 1992). Fire regimes directly influence several plant processes such as survival, germination, establishment and dispersal, and can therefore alter the structural and floristic composition of the vegetation (Bradstock, 1993; Whelan, 1995; Williams and Gill, 1995; Morrison et al., 1996; Bowman, 1998; Clark et al., 2002; Keith et al., 2002). Particular life-history traits will determine the kinds of fire regimes likely to cause population decline by influencing the responses of particular life stages and rates of transfer between them (Gill and Bradstock, 1995; Keith, 1996).

Australian plants can be divided between those that are killed by fire and those that survive. Plant species that are killed by fire reproduce predominately from seed are called obligate seeders and are classed as fire sensitive (Gill, 1975; 1977). Plant species that are not usually killed by fire are called vegetative regenerators and are classed as fire resistant (Gill, 1975; 1977). Changes in fire regimes, especially the frequency component can result in a change in dominance between fire sensitive and fire resistant species over time (Purdie and Slater, 1976; Fox and Fox, 1986; Nieuwenhuis, 1987; Siddiqi et al., 1975).

The components of fire regimes can be easily manipulated by anthropogenic ignition (e.g. Williams and Gill, 1995; Keith, 1996) and hence humans can drastically affect the biodiversity of Australia’s terrestrial ecosystems. Floristic diversity may be maintained by

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 3

promoting fire intervals that are intermediate between the time of maturation and the time of senescence (Keith, 1996). It has been shown that fire frequency is a major source of within community variability in the relative abundances of the component plant species in the fire-prone sandstone communities of the Sydney Basin (e.g. Clark 1988; Cary & Morrison 1995; Morrison et al. 1995; Conroy 1996).

1.2 A history of fire in Australia

The Quaternary vegetation and fire history of Australia was first reviewed by Singh et al. (1981) and published in Gill et al.’s (1981) Fire and the Australian Biota. Subsequently this topic has been reviewed by a number of authors (e.g. Clark, 1989; Pyne, 1991; Dodson, 1994; Kershaw, 1995; Kohen, 1995; Bowman, 1998). Singh et al.’s (1981) review focused on the pollen and charcoal records of Lake George, Lynchs Crater and Lashmars Lagoon. The Lynchs Crater record was first published by Kershaw (1974; 1978) and the Lake George record was later revised by Singh and Geissler (1985).

Lake George is an extensive internal drainage basin located on the Southern Tablelands of New South Wales that contains sediments dating to ~800 ka (Singh and Geissler, 1985). An increase in charcoal at 128 ka, which according to Singh et al. (1981) represented an increase in fire frequency, coincided with a change from a interglacial dominated by (reportedly) fire-sensitive Casuarina woodlands to open woodland or forest of fire adapted Eucalyptus. Singh et al. (1981) suggested that this increase in charcoal, and shift in vegetation, was the result of changes in fire regimes caused by the arrival of Aborigines, since there were no notable climate changes at the time. Bowman (1998), among others, criticised the study for basing such controversial conclusions amongst too much uncertainty. Uncertainty surrounded the chronology of the core for example, as it was inferred from correlation with oxygen-isotope records preserved in deep-sea cores (Bowman 1998; Horton, 1982). Singh et al.’s (1981) conclusion that Casuarina is fire sensitive has also been questioned (e.g. Ladd, 1988). The 128 ka date of Aboriginal arrival

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 4

predates archaeology evidence by at least 80 000 yr, and has also been a source of substantial controversy (Bowman, 1998).

Lynchs Crater is found on the Atherton Tableland, Queensland and the palaeoecological record from this site dates back to ~200 ka and is one of the longest records of vegetation and climate change in the low altitude tropics. The Lynchs Crater record revealed a transition from Araucaria dominated rainforest to Eucalyptus at 38 ka (Kershaw, 1978; 1983). This transition, similar to Lake George, coincided with a dramatic increase in microscopic charcoal. Kershaw (1978; 1983) argued that this transition did not occur due to climatic change because the transition is unlike all the others in the core. The transition is believed to have been due to the increase in burning by colonising Aborigines (Kershaw, 1978; 1983). However, Clark (1983) suggested that this increase in charcoal could have been due to the drying of the lake and subsequent fires on its surface or increased fires in the region caused by climatic drying. The 38 ka date given by Kershaw’s (1983) study was much more consistent with archaeological evidence, and hence was widely more accepted than the Lake George interpretation. More recently Turney et al. (2001) better constrained the chronology of the Lynchs Crater sediments to find that the increase in charcoal occurred at ~45 ka.

Kershaw et al. (1993) examined a continuous pollen core from a deep ocean core, ODP 820, representing more than 1.5 million years of sediment. The site exhibits a number of changes in fire activity and vegetation, one at ~ 140 ka and the other at 40 – 35 ka. Kershaw et al. (1993: 113) claimed that the changes at ~140 ka “adds substantially to the case for Aboriginal burning and an early date for human arrival in Australia”. White (1994) criticised the charcoal evidence, pollen record and the depositional situation and hence described the claims by Kershaw et al. (1993) as insubstantial. Anderson (1994) and Hope (1994) had concerns regarding the sampling of the sequence.

Kershaw et al. (2002) comprehensively reviewed the history of fire using long-term charcoal records of Australia. They asserted that climate has been the major control over both fire activity and vegetation change. Kershaw et al. (2002) identified a general increase

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 5

in fire activity throughout the Holocene with a slight decrease between 7 and 5 ka associated with an increase in rainfall and reduction in seasonality (Kershaw, 1995). Kershaw et al. (2002) also noted a significant increase in fire from the mid-Holocene in all Australian vegetation types, excluding wet forests, and this was maintained until ~2 ka. Although Kershaw et al. (2002) considered the influence of people to explain this increase in fire activity they thought it was better explained by decreased precipitation and the onset of El Niño events in the region.

1.3 Aboriginal use of fire during the Late Quaternary

The timing of the arrival of Aboriginal people to Australia remains unclear and is subject to much debate. It has long been accepted that Aboriginal people arrived some 40 000 yr ago and this date continues to be asserted by a number of authors (e.g. Allen and Holdaway 1995; O’Connell and Allen 1998). However, ~40 ka is the upper limit of conventional radiocarbon dating and hence this arrival date may be biased due to limitations of this dating technique (Fifield et al., 2001). New dating techniques such as step combustion accelerator mass spectrometry of 14C, thermoluminescence and optically stimulated luminescence have been used to suggest Aboriginal arrival dating back to 60 ka (Roberts et al. 1990, 1994) and this earlier date has been widely accepted (Attenbrow, 2002).

The earliest undisputed dates of Aboriginal arrival to southern Australia range between 40 and 30 ka (Attenbrow, 2002) and this is based on archaeological sites from Lake Mungo (Bowler et al., 1970) and Cuddie Springs (Furby, 1995; Field and Dodson, 1999) in western New South Wales, Keilor in Victoria (Bowler, 1976), Devils Lair in southwestern Western Australia (Dortch and Merrilees, 1973; Dortch, 1979), and in Tasmania (Cosgrove, 1995). The extent of the impact of the arrival of people had on Australia’s ecosystems remains speculative and controversial.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 6

In 1969 the theory of ‘Fire-stick Farming’ (Jones, 1969), where Aboriginal people were thought to have used fire to manipulate and thereby increase the availability of resources, changed European Australia’s perception of how Aboriginal people had managed the land in the past. This also had implications for environmental management issues and had political implications. For Aboriginal people to have ‘farmed’ the land suggested some sort of custodianship and this challenged the notion of Terra nullius (whereby European colonists justified the invasion of Australia under the pretence that there was no ownership by Aboriginal people and therefore no legal rights). Head (1989: 41) outlined some of the legal and political ramifications of Aboriginal use of fire to challenge the notion of Terra nullius stating “Aboriginal ways of using the land had to appear as much like farming and as little like hunting and gathering as possible” (Head, 1989: 41) in order for non- Aboriginal Australians to accept land rights.

Aboriginal people used fire for hunting (Clark and McLoughlin, 1986; Kohen, 1993) and to stimulate the growth of young shoots to attract game (Lewis, 1982; Christensen and Burrows 1986; Benson, 1991; Dodson, 1994; Benson and Redpath, 1997). However, this was not a universal practice, with evidence of Aboriginal people hunting without the use of fire (Harvard, 1944; Gunston, 1974; McLoughlin, 1998). Fire was also used to regenerate food and useful plants (Jones, 1969; Clark and McLoughlin, 1986; Benson, 1991; Gott, 2005). In some areas there is evidence that the Aboriginal people would have protected areas from frequent fire to protect areas of cultural significance or resources such as the yam in wet sclerophyll forests and rainforests (Dodson et al., 1994; Kohen, 1996; Benson and Redpath, 1997). Fire was also employed for ease of travel (Jones, 1969; Christensen and Burrows, 1986; Clark and McLoughlin 1986; Benson, 1991; Kohen, 1993), signalling (Jones, 1969; Benson, 1991), fashioning weapons (Christensen and Burrows, 1986), cooking (Nicholson, 1981; Dodson, 1994), warmth (Latz and Griffin, 1978; Nicholson, 1981; Dodson, 1994) and ceremonial purposes (Hallam, 1975; Dodson, 1994).

Jones’ (1969) thesis also influences ecological management strategies with some believing that “Aboriginal” fire regimes were applied across the entire continent and hence similar fire regimes could be used for contemporary management of nature reserves (e.g. Rolls,

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 7

1981; Flannery, 1994; Ryan et al., 1995; Horton, 2001). It has been suggested that the application of this type of fire regime will “return” the landscape to what it was under Aboriginal land management.

Flannery (1994) claimed that the Australian landscape was changed by the impact of huge uncontrolled fires which was a direct consequence of the build-up of uneaten vegetation following the demise of the Australian megafauna. Flannery (1994) believed that the vegetation patterns of Australia were greatly changed by Aborigines, and that Aborigines were particularly responsible for open, grassy vegetation communities on a continental scale resulting in the ‘park-like’ landscape discovered by European settlers. Flannery (1994) based his theories on evidence recorded in the writings of early settlers, which are contentious due to possible misinterpretations and exaggerations (Clark, 1983; Benson and Redpath, 1997). For example it is possible that researchers have failed to distinguish between smoke from campfires and smoke from broadacre fires, as noted in explorers accounts (Vigilante, 2001). Flannery (1994) has also been criticised for basing his argument on untestable hypotheses (Benson and Redpath, 1997).

Gill (1977) agreed that frequent low intensity fires were applied by Aboriginal people in some ecosystems, such as the northern savannas, but urged managers not to generalise this regime across the entire continent. Horton (1982) suggested Aborigines had little effect on vegetation zones when compared to changes caused by climate and formulated these conclusions primarily based on ecological evidence. Horton (1982) hence argued that the broad vegetation zones of Australian were determined by climate and that Aboriginal use of fire had very little role in shaping Australia’s ecosystems.

Benson and Redpath (1997) have also criticised the widespread acceptance of these hypotheses about Aboriginal use of fire as a management strategy, believing that it would not benefit conservation as frequent and uniform landscape firing has been proven to detrimentally affect biodiversity in Australian ecosystems. In fact too frequent fires have been listed as a ‘Key Threatening Process’ to biodiversity under the Threatened Species Conservation Act 1995 (NSW).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 8

The impact of Aboriginal use of fire most probably varied across Australia, with some ecosystems being more resilient than others, or more favoured for Aboriginal utilisation (Clark, 1983; Clark and McLoughlin, 1986; Head, 1989; Bowman, 1998). Clark (1983) suggested there would have been a large variety of fire regimes applied across Australia and would have changed with variations in climate, vegetation, population and technological changes, food resources and other cultural factors. This view suggests that Aborigines did have a drastic affect on the vegetation in some areas, as proposed by Flannery (1994), Flood (1986), Gott (2005) and others, whereas other landscapes may not have been burnt and climatic zones would have been the deciding factor in vegetation distribution, which supports the theories of Horton (1982) and others.

Head (1989: 31) noted that there is a common assumption that Aborigines “had a single ongoing impact”, potentially ignoring climatic change and population and cultural change. Head (1989) also suggested that Aboriginal people minimised their impact once they understood how the Australian environment worked, and hence used fire within a climatic framework. Clark and McLoughlin (1986) and Baker (1997) suggested that Aboriginal people applied fire to different vegetation types differently depending on what resources they were extracting. Unlike Europeans, Aboriginal people did not have the technological means to control or extinguish landscape fires and thus with respect to the use of fire would have relied on knowledge or prediction of fire behaviour, for preservation of resources and life (Bowman, 1998).

1.4 European impacts to fire regimes

The arrival of Europeans to Australia saw the introduction of large numbers of plants and animals, vegetation clearing, soil changes, extinction, changed fire regimes and the gradual removal of land use practices and influences of the Aboriginal people (Adamson and Fox, 1982; Dodson et al., 1994; Hope, 1999).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 9

Low intensity burning by Aboriginal people has been suggested as a reason for this vegetation structure and the subsequent changes (e.g. Ryan et al., 1995) however Horton (2001) recently suggested this to be untrue. Upon arrival to the Sydney Region early European settlers described the ‘park-like’ nature of some vegetation communities, such as the Cumberland Plain Woodland, because of widely spaced trees with a grassy understorey (Florence, 1994). After the arrival of Europeans, and the subsequent demise and displacement of Aboriginal people, the vegetation began to thicken with the growth of a forest of young trees.

European settlers cleared land for crop and pastures with the application of fire to the forest landscape (Gill, 1977; Christensen and Burrows, 1986; Pyne 1991). The changed fire regimes, naturalisation of exotic plants, animals and diseases, and habitat fragmentation and destruction led to the decline of native birds and mammals (Adamson and Fox, 1982).

In some areas European settlement seems to have resulted in increased fire frequency whilst other areas have seen a decreased fire frequency and these changes to the fire regime are often associated with vegetation changes (Head, 1989; Gill and Bradstock, 1995). There are a number of charcoal and pollen studies from south-eastern Australia that have suggested an increase in burning since the arrival of European people (e.g. Singh and Geissler, 1985; Dodson, 1986; Banks, 1989; Green et al., 1988; Dodson et al., 1994; Kodela, 1996; Mooney et al., 2001) and other studies that have shown a decrease (e.g. Kodela and Dodson, 1988; Johnson, 1994; Martin, 1994; Dodson et al., 1995). This contradiction can also partially be explained by unresolved issues associated with methods, as described in Section 1.7.

There was a major shift in fire management in southeastern Australia in the early 1940s due largely to the widespread conflagrations of 1939 (Pyne, 1991; Kirkpatrick, 1999). In the 1960s hazard reduction burning was implemented, which saw the application of low intensity fire during the winter months, with the aim of reducing the risks to people and property (Pyne, 1991). This practice has been identified as potentially impacting on biodiversity and ecosystem health (Benson and Redpath, 1997; Brandes, 2003).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 10

1.5 The Sydney Basin

1.5.1 Geomorphology The Sydney Basin lies on the central east coast of New South Wales and covers an area of approximately 3.7 million ha. It contains the towns of Wollongong and Nowra to the south, Newcastle to the north and Katoomba and Mt Victoria to the west. The Sydney Basin is a geological basin that consists largely of Permian and Triassic-aged sandstones and shales, with layers of early Permian coal, which overlay the older basement rocks of the Lachlan Fold Belt (Herbert, 1983). The Basin was subjected to an uplift that resulted in gentle folding and minor faulting and the formation of the Great Dividing Range. The major geological units in the Sydney Basin are Hawkesbury Sandstone, Narrabeen Group (composed of sandstone and shale), Wianamatta Shale and the Permian Coal Measures (Herbert, 1983).

The Basin consists of a number of discrete geomorphic units (Figure 1.1) that include the Blue Mountains Plateau to the west, the Wollemi-Colo and Hawkesbury Plateau to the north, the Woronora Plateau and Southern Highlands to the south and the central Cumberland Plain (Herbert, 1983). The western part of the Sydney Basin that is covered by the Blue Mountains Plateau has the highest altitude. These elevated sandstone plateaus have been subjected to erosion by coastal streams and this has highly dissected the plateaus. The post-glacial rise in sea level (between ~18 – 6 ka) resulted in the flooding of river valleys and coastal plains and resulting in estuaries and deep harbours. Much of the Sydney Basin is dominated by the sandstone plateaus with the exception of the shale derived Cumberland Plain and Hunter Valley. The soils derived from Hawkesbury Sandstone are strongly acidic and are deficient in phosphate and nitrogen (Bannerman and Hazelton, 1990).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 11

Figure 1.1. The major geomorphic units of the Sydney Basin and the location of the three sites which were used in this study.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 12

1.5.2 Climate The Sydney Basin regions lies on the Subtropical East Coast (Gentilli, 1972) and experiences a temperate climate characterised by warm summers and no distinct dry season, although a sub-humid climate can be found in the northern parts of the Basin and small areas of Montane climate are found in the Blue Mountains (NPWS, 2005). Average annual rainfall is variable across the Basin with some parts (e.g. south-west Cumberland Plain) receiving as little as 522 mm pa and other parts as much as 1614 mm (e.g. Watagan Mountains (BoM, 2005; NPWS, 2005). Rainfall varies due to altitude and distance from coast and there is an east-west rainfall gradient.

Rainfall is associated with the latitudinal shifts in the anticyclone sequence with the wettest months from March to June whilst the driest months of the year are September and October.

The mean annual temperature of the Sydney Basin is 10 - 17°C with minimum average monthly temperatures ranging from -1.4 - 8.1°C, and maximum average monthly temperatures ranging from 22.4 - 31.9°C. Winds are strongest in winter when they are westerly and north-westerly. In summer the winds are predominately south to south- easterly with a tendency for onshore afternoon northeasterlies on the coast (Forster et al., 1977).

The climate of the Blue Mountains is strongly controlled by altitude which directly affects temperatures and indirectly affects precipitation through orographic uplift and the development of rainshadows.

The natural fire season of the Sydney region occurs from October to February, with most fires occurring in December and January (Cunningham, 1984). The Sydney Sandstone Complex vegetation, which is found on the plateaus that surround Sydney, are particularly prone to fire and fire events in these vegetation communities can occur every 3-4 years apart, with major bushfire seasons occurring every 11 years (Cunningham, 1984).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 13

1.5.3 Vegetation The considerable variation in geology, soils, climate and topography has resulted in one of the most species diverse regions in Australia (NPWS, 2005) and includes communities of dry and wet sclerophyllous forests, Warm Temperate rainforests, heath, mangroves and swamps. Sydney Sandstone Complex vegetation dominates the sandstone environments of the Sydney Basin and such as the Woronora, Blue Mountains and Hornsby plateaus that surround Sydney (Benson, 1992).

The species composition and structural formation of the vegetation communities that occupy the sandstone plateaus of the Sydney Basin vary due to altitude, aspect, moisture status, soil depth and rainfall (Benson, 1986; 1992). Heath and low open woodland are commonly found on the rocky platforms and ridgetops whereas taller open forests occur on the deeper plateau soils and on the slopes (Benson, 1986; 1992). Benson (1992) identified two broad sub-units within the Sydney Sandstone Complex. These are a moist forest type that is associated with sheltered hillsides and moist gullies and a dry woodland type found on the dry plateaus and ridgetops. The Sydney Sandstone complex vegetation communities are dominated by Eucalyptus, Angophora, Corymbia, Syncarpia, Doryphora and Ceratopetalum (Benson, 1986; 1992).

The study sites used herein have been confined to swamp deposits on sandstone locations. These communities are found as isolated communities throughout the Sydney Basin and include swamps found on poorly drained Quaternary deposits and hanging swamps on the sandstone plateaus (Fairley and Moore, 2000). Floristic composition varies locally in relation to soil moisture gradients and the vegetation on these swamps can form monocultures of Common Reed or complex communities of Prickly-leaved Tea-tree (Melaleuca stypheloides) and Paperbark (Melaleuca quinquenervia) associations, with Swamp Mahogany (Eucalyptus robusta), Swamp Oak, Sedges, Tall Spike Rush (Elaeocharis sphacelata) and Juncus (Juncus sp.) (Keith and Benson 1988; Benson and Keith 1990).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 14

1.5.4 Aboriginal history Attenbrow (2002) used a timeline to summarise the Aboriginal history of the Sydney region and discussed the changes using the following periods: pre-glacial (60 – 25 ka), glacial (25 – 15 ka), post-glacial (15 – 10 ka), the early Holocene (10 – 5 ka), late Holocene (5 – 1.6 ka), from 1.6 to European arrival (AD1788) and the colonial period.

Mulvaney (1971: 378) challenged the notion of Aboriginal people of being “unchanging in an unchanging land” and those who viewed Aboriginal socio-economic and demographic changes as seemingly insignificant in comparison to prevailing environmental conditions (e.g. Birdsell, 1953). It is well accepted that there have been changes in Aboriginal technologies including changes to tool assemblages and artefact densities from the mid-late Holocene. Rowland (1999) summarised the debate as to whether these changes were in response to changes in environmental conditions (e.g. Bowdler, 1977; Jones, 1977; Horton, 1981) or whether internal changes occurred within Aboriginal populations (e.g. Lourandos, 1980; 1983).

Lourandos (1980; 1983) suggested that Aboriginal populations ‘intensified’ from the mid- late Holocene (from ~5 ka) and this was a period of continent-wide changes in terms of technology, socio-demographics, settlement patterns, social structures and population densities. During this time ‘environmental manipulation strategies’, particularly large-scale drainage systems, the use of fire to increase resources, and harvesting and processing of food plants are thought to have intensified populations (Lourandos, 1983: 81).

The theory of ‘intensification’ is controversial within Australian archaeology with some especially critical about the forms of evidence and sources used to justify the theory (Attenbrow, 2004). As an example Bird and Frankell (1991: 10), explained the archaeological changes identified by Lourandos (1980; 1983) as a series of short-term adjustments to local conditions rather than “cumulative directional change” or intensification.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 15

A major flaw in Lourandos’ ‘continent-wide’ intensification model is that while some regions have revealed an increase in site numbers from 5000 yr BP others have not registered a change until, for example, 2000 yr BP and others have seen a decrease in the last thousand years (Attenbrow, 2004).

Population change, behavioural change and natural processes have been used by archaeologists to describe the changes in archaeological sequences (Attenbrow, 2004). Attenbrow (2004) also suggested the possibility that behavioural or land-use changes, associated with climatic change, may be responsible for changes in the archaeological record from Upper Mangrove Creek, New South Wales, rather than increases in Aboriginal population numbers. Rowland (1999: 15) also suggested that “internal social dynamics, external cultural influences and environmental factors” should all be considered when interpreting the archaeological record.

Stockton and Holland (1974) dated an artefact (a pebble chopper) found in situ at the base of gravels from Cranebrook Terrace, near Penrith in western Sydney, at ~30 ka. Stockton et al. (2004) more recently suggested that the artefact may in fact be substantially older due to contamination of younger carbon contaminating the radiocarbon age.

Another claim for early human occupation in the Sydney Basin is ~22 ka from Kings Tableland in the Blue Mountains (Stockton and Holland, 1974), although this date has been disputed. There are no coastal archaeological sites in the Sydney region of this antiquity however there are other sites from coastal south-eastern Australia that date from 21 – 17 ka (e.g. Bass Point, Cloggs Cave, Burrill Lake). As mentioned in Section 1.3, the earliest dates of occupation in other regions of south-eastern Australia range from 40 to 30 ka hence it is very likely that the Sydney region was occupied much earlier than the current archaeological record suggests. Turbet (2001) has suggested that Aborigines were present in the Sydney region by at least 30 ka, and they were largely concentrated in coastal areas. Attenbrow (2002) suggested that during the glacial period (~25 – 15 ka) Aboriginal people were probably living adjacent to watercourses and the coast and also hunting and gathering in the hinterland.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 16

The earliest undisputed archaeological date in the Sydney Basin is 14 700 ± 250 yr BP from Shaws Creek K2 (Nanson et al., 1987). The stone artefact assemblages of the post-glacial period (~15 – 10 ka) in the Sydney region are known as Capertian, which is the earliest of the three phases defined the ‘Eastern Regional Sequence’ (MacCarthy, 1964). During this period the people who inhabited the coastal zone of the Sydney region would have been displaced westward as sea levels rose, inundating campsites and fishing and hunting grounds (Attenbrow, 2002) In fact some 1100 km2 of the continental shelf in the Sydney Region was inundated due to the rising sea level (Roy, 1998). From the early Holocene to the mid-Holocene Aboriginal people are believed to have continued to move to higher ground as sea levels continued to rise and eventually stabilise at ~6 ka (Attenbrow, 2002). The westward movement of Aboriginal people during the early to mid-Holocene may have resulted in the establishment of new occupation sites in Sydney’s west (e.g. Darling Mills SF2 rockshelter). During this time new types of stone implements, such as flakes and blades made using a technique called ‘backing’, make their first appearance (Attenbrow, 2002). Backed artefacts increase substantially, as do the number of sites being occupied, in the Sydney region during the mid-late Holocene however whether this reflects an increase in the Aboriginal population is uncertain (Attenbrow, 2002; 2004).

Historically the ‘intensification’ of Aboriginal populations has been associated with what was thought to be the most widely recognised technological break in Australian prehistory occurring from the mid-late Holocene. After this time a range of small, well-made flaked stone tools make an appearance in the archaeological record, referred to as the ‘Australian Small Tool Tradition’ (ASTT) (Mulvaney, 1975). Hiscock and Attenbrow (1998) established that backed artefacts, often associated with the ASTT, have been produced, albeit at low rates and in only a number of sites, from the early Holocene and did not appear suddenly. Hiscock and Attenbrow (2004) reanalysed the Capertee 3 artefact assemblage, originally excavated by McCarthy in the 1950s and 1960s, to find backed artefacts dating to 6000 to 7000 yr BP. Hiscock and Attenbrow (1998) suggested Australian archaeologists to cease employing the concept of a pan-continental ‘Small Tool Tradition’.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 17

Of the more than 27 sites that have been excavated from the Blue Mountains only 7 have information on artefact accumulation rates or densities: Kings Tableland (Stockton and Holland, 1974); Springwood Creek (Stockton and Holland, 1974); Walls Cave (Stockton and Holland, 1974); Capertee 1 (McCarthy, 1964); Capertee 3 (Johnson, 1979; Hiscock and Attenbrow, 1998); Shaws Creek K1 (Stockton, 1973) and Shaws Creek K2 (Kohen et al., 1981; 1984). Capertee 3 is perhaps the best resolved archaeological sequence in the Blue Mountains region.

Stockton (1970) described human occupation in the upper Blue Mountains as ‘spasmodic’ resulting from visitation during seasonal hunting trips during the milder periods of the Holocene. This conclusion follows implications that the upper mountains were too rugged with too few resources and that populations would be sensitive to climatic variations (Stockton and Holland, 1974). Stockton and Holland (1974: 60) suggested that the Capertian Industry (of McCarthy, 1964) “flourished” between 12 ka and 6050 +/- 170 BP followed by a hiatus that separated the Capertian and Bondiain industries between ~6 and 3.36 ka. Bowdler (1981) suggested that Aboriginal occupation of any intensity, in the eastern highlands of Australia, can only be dated to ~5 ka.

Attenbrow (2003; 2004) investigated the past habitation and landuse patterns in Upper Mangrove Creek catchment, in the central coast hinterland. She found evidence of increasing numbers of habitations from ~11 ka with a dramatic increase from ~ 3 ka and explained these changes in terms of changing habitation, mobility and land use patterns. Attenbrow (2004) also reviewed the archaeological records of the South Coast and Southern Tablelands and suggested that the establishment of Aboriginal sites increased from 8 ka with the habitation rates of these sites generally increasing until the arrival of European people.

It is estimated that there were between 3000 and 5000 Aboriginal people occupying the Sydney region at the time of European arrival (Attenbrow, 2002). This consisted of several distinct Aboriginal language groups: the Darug (also spelt Dharug, Darak, Dharak)

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 18

Figure 1.2 The territory of the different Aboriginal language groups and the key archaeological sites in the Sydney Basin. The location of the sites investigated in this study are also shown.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 19

extended from the Hawkesbury in the north to Botany Bay in the south and westward to the northern Blue Mountains; and to the north and south of the Darug dialects, the Darginung, Dharawal (also spelt Tharawal) and the Gundungurra languages were spoken (Figure 1.2).

The coastal Aboriginal people predominately subsisted on fishing, shellfish, hunting terrestrial animals and gathering plant foods and shellfish and they utilised the resources of the land, river, estuaries and oceans. Hinterland groups focused their resource use on the plant and animals of the land and forests (Attenbrow, 2002). Very little stone use by Aboriginal people was observed in the Sydney Region with the exception of the hinterland populations (Attenbrow, 2002).

The arrival of Europeans to the Sydney Basin saw the westward displacement of the Aboriginal people as European colonies encroached from the east. There was a disastrous reduction of the Sydney Aboriginal populations due to small pox from mid 1789 (Turbet, 2001). It has been estimated that this epidemic killed 50 to 90% of the Sydney Aboriginal population (Kohen, 1993; Breckell, 1993) and within 40 years of European arrival the ‘pre- colonial’ Aboriginal way of life had disappeared from the Sydney region (Attenbrow, 2002).

1.6 Late Quaternary climate change

The Quaternary period (approximately the last 2 million years) is characterised by a number of glacial-interglacial episodes, which are believed to be influenced by orbital fluctuations and interactions between the ocean-atmosphere-cryosphere system (Hays et al., 1976). The last full glacial cycle occurred from ~130 ka until climatic amelioration at ~ 14 ka (Williams et al., 1998). This period, typical of the late Quaternary, saw progressive cooling with a number of relatively short periods of warmth (interstadials) and cold (stadials), eventually with a maximum cold period (Last Glacial Maximum) followed by deglaciation into the Holocene. Nanson et al. (1992) described Australia’s Quaternary

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 20

climate as oscillating between dry and wet periods and this is associated with glacials and interglacials, respectively.

The longest palaeoecological record in this study exceeds 43 ka, which is the limit of conventional 14C dating, and is likely to represent 50 – 60 000 yrs. Hence climatic changes experienced in south-eastern Australia dating encompassing marine oxygen isotope stages 1 – 3 will be discussed. Marine isotope stage 3 was an interstadial occurring between ~59 and 24 ka. This period corresponded with high lake water levels in southeastern Australia (e.g. Bowler, 1981; Chappell, 1991; Nanson et al., 1992). Lakes from the Australian interior, including Lake George, were more extensive during this period and Bowler (1981) argued that this was evidence for a different hydrological regime to the present day. The palaeoenvironmental record from Willandra Lakes in southwest New South Wales provided evidence of a wet period between ~50 and 35 ka (Bowler, 1981; 1986) however it has recently been suggested that between 50 and 40 ka was a period of fluctuation between high lake levels and drier periods (Bowler et al., 2003).

The climate of the Last Glacial Maximum (LGM) in Australia, ~18 – 21 ka, was inhospitable with cooler, drier and windier conditions (Markgraf et al., 1992; Allan and Lindsay, 1998). Barrows et al. (2002) used cosmogenic isotopes 10Be and 36Cl and the exposure of boulders on the terminal edges of moraines in Tasmania to date the LGM between 17.3 ± 1.1 ka and 20.1 ± 1.9 ka. Dust records from a Gulf of Carpentaria sediment core revealed the highest proportion of sand, representing peak dune movements and hence aridity, occurred at 21.5 ka, with a second peak at 24 ka (DeDeckker, 2001).

Glaciers in New Guinea, Tasmania and New Zealand reached their maximum extent at this time (Markgraf et al., 1992). The effects of snow and ice on the Australian continent were geographically limited (i.e. 30 – 50 km2 on the Australian mainland and ~5000 km2 in Tasmania) (Galloway, 1973), but the climate was significantly drier, and perhaps more seasonal. At the LGM, aridity was more widespread in the intertropical zone, the Trade Winds were stronger, dunes were active beyond their present limits and considerable

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 21

volumes of dust and loess were deposited on land and at sea (Williams et al., 1998). For most regions it is estimated that precipitation levels were about half present day values in the mid-latitudes (Allan and Lindsay, 1998) and mean annual temperatures were up to 10oC cooler in the southeastern Australia (Kershaw, 1995). Sea levels during the LGM were ~120-135 m lower than current levels (Yokoyama et al., 2001; Lambeck and Chappell, 2001) resulting in Australia being about 30% larger than it is today (Markgraf et al., 1992).

In Australia there are relatively few continuous pollen records that cover the LGM (Dodson, 1994; Kershaw, 1995, Pickett et al., 2004). In fact there are only 33 sites in the Australian, Southeast Asian and Pacific region (SEAPAC region) that have a record of vegetation during the LGM. At the peak of the LGM, about 20 ka, the vegetation was a semi-arid grassland-steppe, dominated by Asteraceae and Poaceae and with small patches of mesic communities (e.g. Dodson, 1994; Hope, 1994; Kershaw, 1995: 1998). More recently Pickett et al. (2004) have suggested xerophytic shrubs/woodlands characterised southern Australia during the LGM rather than steppe vegetation. The exposed continental shelf of southeastern Australia at the LGM, however, was covered in shrub, heath and woodland communities dominated by myrtaceous shrubs, Asteraceae and Chenopodiaceae (Harle, 1997). Previous studies suggest a relatively rapid climatic amelioration following the LGM with increases in arboreal taxa (e.g. Eucalyptus, Casuarina).

In the Northern Hemisphere several high-resolution palaeoenvironmental records such as ice and coral cores reveal abrupt and large amplitude climatic changes in between the LGM and the Holocene (Bond et al., 1993). In Europe the ‘classic’ post-glacial sequence includes the Oldest Dryas stadial (cool), the Bølling interstadial (warm), an Older Dryas (cool), the Allerød interstadial (warm) and then the Younger Dryas (cold) which then terminates at the start of the Holocene. Whether this series of climatic fluctuations of the late glacial were felt globally is still a matter of debate, however such events within the period (e.g. the Younger Dryas) are increasingly being described as such.

Much of the description of the climatic oscillations of the late glacial in the Northern Hemisphere comes from well-dated ice cores drilled in the Greenland ice-sheet. Both the

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 22

Greenland Ice Core Project (GRIP) and the US Greenland Ice Sheet Project (GISPII) have provided a clear sequence of rapid climatic shifts during this period (Daansgard et al., 1993; Bond et al., 1993). Antarctic ice cores reveal stadial/interstadial conditions: viz. the Antarctic Cold Reversal (ACR) from ~14-12.5 ka (Jouzel et al., 2001) and the Oceanic Cold Reversal from ~13.2-12.5 ka (Stenni et al., 2001) following the LGM.

The Bølling/Allerød was terminated with a dramatic decrease of temperatures to almost full-glacial values. The subsequent cool period, known as the Younger Dryas, was first detected in pollen assemblages from north-western Europe that revealed the re-occurrence of the high latitude wildflower Dryas octopetala (Kerr, 1993). The Greenland ice core record suggests a drop in temperature of approximately 7oC (Bond et al., 1993). Using precise, sub-annually-resolved measurements from the GISPII core, the Younger Dryas was an event of 1 300±100 years duration commencing at about 12 640±250 yr BP (Alley et al., 1993).

Several palaeoenvironmental studies have found evidence for a Younger Dryas in Australasia. This includes evidence of the re-advance of the Franz Josef Glacier in the Southern Alps of New Zealand (Denton and Hendy, 1994) and δ18O variations from a speleotherm in south-eastern Victoria, Australia by Goede et al. (1996). Both of these studies infer a Younger Dryas event coeval and of similar character to the event in the Northern Hemisphere.

The Pleistocene-Holocene boundary (~11 ka) has been clearly marked as a period of rapid climate change with temperatures and moisture regimes reaching modern values (Kershaw, 1995). Palynological records show reafforestation proceeded to roughly present day level from the beginning of the Holocene. Changes in the vegetation have been relatively minor throughout the Holocene, which is perhaps due to palynological invisibility of subtle changes within families (e.g. Myrtaceae, Casuarinaceae) (Clark, 1983). This and the relative magnitude of Quaternary glacial-interglacial cycles have lead to the impression of relative stability during the Holocene (Dodson and Mooney, 2002). This is in contrast to palaeoclimatic and palaeoenvironmental evidence from other sources indicating Holocene

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 23

variability (Bond and Lotti, 1995; Rodbell et al., 1999; Sandweiss et al., 1999 deMenocal et al., 2000). A number of records in south-eastern Australia show an increase in Eucalytpus at the expense of Casuarina starting from various stages during the Holocene (e.g. Ladd et al., 1992; Devoy et al., 1994; Harle, 1998). This has been attributed variously, but most often to the role of fire.

The period 7 – 5 ka has been described as the precipitation peak, or Holocene Climatic Optimum in Australia and the period 4 – 2 ka was perhaps cooler and drier (Kershaw et al., 2002). The mid-Holocene has been identified as a period of climate change in Australasia (e.g. Bowler et al., 1977; Shulmeister, 1999) and further afield (e.g. Rodbell et al., 1999; deMenocal et al., 2000; Sandweiss et al., 1999). Steig (1999: 1485) described the mid- Holocene as “a period of particularly profound change” and Hodell et al. (2001) noted an abrupt cooling of sea-surface temperatures, expansion of sea ice and increased ice-rafted detritus accumulation in the Southern Ocean, between 5.5 to 5 ka. Morrill (2003) examined 109 previously published palaeoclimatic records to assess evidence of abrupt climate change during the mid-Holocene and identified two periods of abrupt climate change occurring at 5.5 - 5.8 ka and 4- 4.8 ka.

In Australasia Shulmeister (1999) described a decoupling of the northern (tropical) and southern (temperate) climate systems of Australia at ~5 ka resulting in increased westerlies, the loss of summer monsoon rainfall and a sharp decline in effective precipitation in southern Australia. Lees (1992) linked climatic variability in northern Australia with the stabilisation of sea levels at ~5.5 ka. Longmore (1997) used sedimentological, palynological and palaeolimnological records from a perched dune lake on Fraser Island (SE Queensland) to describe a mid-Holocene dry period from about 8.5 ka, peaking at 5.5 ka and ending about 2.5 ka.

Further afield, deMenocal et al. (2000) used dust in tropical Atlantic deep-sea sediment cores to describe an abrupt end of the ‘African Humid Period’ at 5.5 ka. These authors described these changes as perhaps reflecting climatic re-organisation of a greater spatial

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 24

scale, involving the Asian monsoon. It is hence possible that this re-organisation influenced atmospheric pressure systems and hence the climate of eastern Australia. Rodbell et al. (1999) argued that ENSO progressively achieved modern characteristics by ~5 000 cal. yr BP although Moy et al. (2002) placed this earlier, at ~7 000 cal. yr BP. Sandweiss et al. (2001) also described ENSO events from ~5 800 yr BP, albeit at a low frequency until about 3 200 yr BP. Riedinger et al. (2002) also described an increase in the intensity and frequency of El Niño events from 3 100 cal. yr BP and Clement et al. (2000) described an increase in ENSO events during the 3-1 ka period.

1.7 Reconstructing fire history

Fire history can be reconstructed from written documents, dendrochronology, pollen analysis, accumulation rates from charcoal in soil profiles, chemical digestion of carbon in sediments, geochemistry, and the analysis of charcoal deposited in lake sediments and swamps (Patterson et al., 1987; Clark and Robinson, 1993; Clark, 1988; Morrison, 1994). Observations of early explorers and settlers may also be used but are potentially subject to exaggeration or used out of context by modern authors (Benson and Redpath, 1997). For example Captain James Cook, the first European to arrive in eastern Australia, described Australia as a “continent of smoke” and noted whilst sailing the eastern along the eastern coast that he “…saw smokes by day and fires by night…” (Benson and Redpath, 1997)

Dendrochronology can be used to reconstruct fire histories of the late Holocene, depending on the species of tree (Whitlock and Millspaugh, 1996; Clark, 1990) and limited to places where trees are long lived and can survive scarring from fires (Clark, 1988). The analysis of charcoal in terrestrial soil can provide evidence of fire but is limited. The most useful methods involve charcoal analysis from lake or swamp sediments because charcoal preserves well in these anoxic, waterlogged sediments (Jacobson and Bradshaw, 1981). Charcoal analysis can potentially provide information on both the timing and extent of past fires over a long period of time (Patterson et al., 1987; Morrison, 1994).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 25

Understanding the process of charcoal production, transportation and sedimentation has been poorly understood in the past (Patterson et al., 1987). However, recent studies, for example Whitlock and Millspaugh (1995), have shed some light on the dynamics of charcoal particles. A better understanding of these processes is needed to improve the efficiency of fire reconstruction.

The amount of charcoal preserved in lake and swamp sediments depends on a number of factors including climatic conditions, fuel types and fire intensity (Morrison, 1994). Clark (1983) has identified some of the problems of relating fires to the abundance of charcoal. Since charcoal is produced in discrete events averaging quantities over many years may result in a misleading record of past fire (Clark, 1983). Changes in sedimentation rates can result in a misleading record of past fires when charcoal is represented as a concentration per volume of sediment (Clark, 1983) however representing charcoal as an influx per year overcomes this problem. The deposition of secondary charcoal (i.e. material introduced during non-fire years as a result of surface run-off and lake sediment mixing) poses another problem in charcoal analysis as high levels of charcoal may represent material temporally stored in the catchment rather than fire events (Clark, 1983). Charcoal analysis techniques may result in the breakage of charcoal due to chemical and physical treatments (Clark, 1983). This is problematic since one large piece of charcoal can be broken into many smaller pieces, skewing the charcoal record.

Most studies which are to reconstruct fire aim to measure charcoal optically (Patterson et al., 1987). Different size particles are transported from fires and deposited in sediments differently and hence perhaps the most important difference within the optical measurement methodology is the choice of particle size (Clark and Royall, 1995; Morrison, 1994). The benefits of using different size classes of charcoal are still debated.

Traditionally, the most commonly employed technique to determine fire occurrence has been the quantification of microscopic charcoal on pollen slides (<100 μm in size). The

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 26

Point Count Method (Clark, 1982; 1983) involves the counting of microscopic charcoal on encountered on pollen slides. The method is therefore convenient since the pollen slides can be reused to count charcoal (Mooney et al., 2001). This method was previously the most widely used technique for reconstructing fire records in Australia largely because the study of fire history was historically an extension of palynology.

Regional fire activity is best represented by microscopic charcoal particles (Patterson et al., 1987; Long et al., 1998; MacDonald et al., 1991; Millspaugh and Whitlock, 1995; Clark and Robinson, 1993; Tinner et al., 1998; Clark, 1988; Clark and Royall, 1995; Morrison, 1994). Since microscopic charcoal is more likely to reflect regional fire activity, the identification of specific fires (which has been attempted by many researchers using microscopic charcoal) is more difficult and probably best reflected by macroscopic charcoal (Millspaugh and Whitlock, 1995). Microscopic charcoal ranging from 2-5 μm has been shown to have continental to global sources, due to transportation in convection columns (Clark, 1988; Tinner et al., 1998). Clark (1988) suggested that due to the way the different charcoal sizes were transported that pollen-slide (microscopic) charcoal not only represents the regional or larger source area but may also under-represent the local fires because of the large skip distances.

It is generally accepted that macroscopic charcoal (>100 μm) does not travel very far from its source (due to their high settling velocities) and hence reflect local, or catchment, fires (Whitlock and Millspaugh, 1996; Long et al., 1998; MacDonald et al., 1991; Millspaugh and Whitlock, 1995; Head, 1989; Clark and Robinson, 1993; Tinner et al., 1998; Clark, 1988; Clark and Royall, 1995; Morrison, 1994). Clark and Royall (1995) demonstrated that macroscopic charcoal does not correlate well with regional fire activity, whilst Whitlock and Millspaugh (1996) found that macroscopic charcoal was abundant in lakes that lay within a 10 km radius of a recent fire but were scarce at great distances. Similarly, Clark (1988) suggested that fragments (>50 μm) are far more useful in detecting local fires than the smaller fragments. Millspaugh and Whitlock (1995) suggested that the most useful size class for studying past fires fall within the 125-250 μm size group. The results provided from microscopic charcoal analysis are very often different to that obtained by macroscopic

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 27

charcoal (Morrison, 1994), and hence the size classes rarely co-vary. However other studies have shown a strong correlation between micro- and macro-scopic charcoal (e.g. Mooney et al., 2001; Gillson, 2004).

Macroscopic charcoal was first analysed using the “petrographic thin section” method by Clark (1988). The method proved advantageous in overcoming the problems with identifying microscopic charcoal (Clark, 1988). The method restricted analysis to charcoal of 50-10000 μm size range, which are not usually examined on pollen slides. The PCM and petrographic thin section method resulted in different charcoal records and Clark (1988) determined the difference was due to charcoal size.

Macroscopic charcoal analysis has been undertaken by a number of researchers including Tinner et al (1998), MacDonald et al (1991), Clark and Royall (1995), Long et al. (1998), Millspaugh and Whitlock (1995) and Mooney et al. (2001). Morrison (1994: 15) suggested that macroscopic analysis holds great hope for studies in fire history and would be complimentary to the “broad spatial perspective afforded by microscopic studies”. In this study a novel method of using image analysis software to quantify macroscopic charcoal has been used and is presented in Chapter 2.

1.8 Review of previous palaeostudies in the Sydney Basin

There have been several studies that have examined the vegetation and fire history within the Sydney Basin and the location of these studies are shown in Figure 1.3 ( i.e. Kodela and Dodson, 1988; Chalson, 1991; Dodson and Thom, 1992; Johnson, 1994; Martin, 1994; Dodson et al., 1995; Mooney et al., 2001).

Kodela and Dodson (1988) analysed charcoal and pollen from South Salvation Creek in Ku-Ring-Gai Chase National Park and the record was dated to ~6 – 5 ka. They recorded no significant changes in the composition of the flora and attributed this to the resilient nature

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 28

Figure 1.3 Location of palaeoenvironmental sites within the Sydney Basin. There are eight sites throughout the Blue Mountains that were investigated by Chalson (1991).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 29

of the sclerophyllous vegetation that surrounds the site. Kodela and Dodson (1988) suggested that the sandstone flora of the site was resilient to the nutrient-poor sandy soils, high insolation exposure, drought, and fire. However there were some minor changes in the pollen assemblage that may have indicated a drier period at ~2 ka. Fire was a persistent feature over the past ~6 ka although there was a decline in fire activity, indicated by a decline in charcoal, over the past 200 yr and this was attributed to the arrival of Europeans and the displacement of Aboriginal people and their burning practices (Kodela and Dodson, 1988).

Chalson (1991) reconstructed the vegetation history of the Blue Mountains based on the palynological of eight sites from the region. The longest record (Penrith Lakes Swamp) extended to ~32 ka although most of the records were Holocene in age. Chalson (1991) described a drier and colder climate during the LGM and identified a series of small fluctuations in effective precipitation throughout the Holocene. The amount of microscopic charcoal in the sediments were estimated, but not quantified, and the samples were analysed at a coarse resolution.

Dodson and Thom (1992) provided charcoal and pollen records extending to the early Holocene (~9.1 ka) from the junction of Mill Creek and the Hawkesbury River. Similarly to South Salvation Creek record (Kodela and Dodson, 1988), Dodson and Thom (1992) found that the sclerophyll taxon that currently grows on the site, dominated by Eucalyptus, have persisted throughout the Holocene. They suggested that the site may have acted as a refugium for rainforest species during the aridity of the LGM. There was a decline in rainforest taxa and a shift to more open vegetation from the mid-late Holocene and this was associated with elevated microscopic charcoal concentrations. This shift was hence attributed to increased fire activity either from climatic or anthropogenic sources (Dodson and Thom, 1992). Dodson and Thom (1992) interpreted low levels of charcoal in the sediment as a more frequent but less intense fire regime.

Martin (1994) and Johnson (1994) have both provided palaeo-records from the Kurnell Peninsula, south of Sydney. Martin (1994) analysed the macroscopic charcoal and pollen

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 30

from a fenland and discussed the natural and anthropogenic impacts on the vegetation and fire history of the site. Fluctuations in the sea level throughout the Holocene played a strong influence over the composition of the vegetation. For example woodland cover suffered losses at ~5 ka associated with the destabilisation of a coastal photobarrier due to sea level rise. Martin (1994) found an increase in macroscopic charcoal at 5.5 ka and this was attributed to an increase in burning associated with the arrival of Aboriginal people since there was no evidence of climatic change at this time. High values of charcoal were also found after 2 ka and Martin (1994) attributed this to heavier ecological pressures of the Bondaian people. Nonetheless Martin (1994) suggested that human occupation over the past 5 ka at the site had little impact on the vegetation composition but cautioned that the concept of Aboriginal people as ‘careful custodians of the land’ may have to be re- evaluated and suggested that low population numbers have been ‘more important than care’ (Martin, 1994: 331).

Johnson (1994) took a multi-disciplinary approach in his investigation of three perched swamps on the Kurnell Peninsula. The records covered the late Holocene with the longest sequence dating to 2.4 ka. Johnson (1994) attributed most of the variations in microscopic charcoal concentrations to changes in Aboriginal people’s use of fire and also found that the palynological record may reflect the plant succession following the establishment of the swamp at ~2.4 ka.

Dodson et al. (1995) investigated the fire and vegetation history from a peatland at Coogee in Sydney’s eastern suburbs. Microscopic charcoal was analysed and values were low during the European period which is not unexpected considering the site is now surrounded by urban development.

Mooney et al. (2001) examined the late Holocene fire and vegetation history from Jibbon Lagoon, another coastal site within Royal National Park, south of Sydney. They found consistently low levels of macroscopic charcoal during the Aboriginal (or pre-European) period with only one large conflagration occurring during the analysed 1600 years. Mooney et al. (2001) suggested that it may have been unnecessary for the coastal Aboriginal groups

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 31

that occupied the site to burn the catchment considering the abundance of marine resources. Charcoal concentrations were found to be higher in the European period and hence there were major differences between the pre- and post- European periods.

1.9 Thesis Aims

The study presents the palaeoenvironmental records from three sites, Gooches Swamp, Lake Baraba and Kings Waterhole, in the Sydney Basin. The overall aim of the study was to:

1) Examine the nexus of fire, climate, humans and vegetation during the late Quaternary in the Sydney Basin; 2) Reconstruct past climatic conditions at the three sites and investigate how fire changes with climate change; 3) Contribute to the understanding of pre-historic changes in vegetation, climate and fire activity in the Sydney Basin; and 4) Investigate how, if at all, fire activity has changed with European settlement of the Sydney region. 5) Examine the environmental consequences of any altered fire in this landscape; and 6) Investigate the links between climate and fire, which may be of relevance for future planning associated with El Niño-like climatic variability and/or any future climate change such as the enhanced ‘greenhouse effect’.

Thesis outline

The chapters of this thesis are presented as either published manuscripts or manuscripts that have been submitted for publication. These manuscripts have co-authors, namely my supervisor Dr. Scott Mooney, Dr. Helene Martin and Dr. Simon Haberle. Dr Scott Mooney provided a conceptual framework for the project and has therefore been included as a co-

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 32

author on all papers. Dr Helene Martin assisted with the interpretation of the Lake Baraba palynology and has been included as a co-author for that paper. Dr Simon Haberle assisted with the statistical analysis of the final paper and has been included as a co-author on that paper. The thesis is outlined as follows:

Chapter 2 presents a novel method for quantifying macroscopic charcoal isolated from sediments. This chapter has been published as:

• Mooney, S. and Black, M. (2003) A simple and fast method for calculating the area of macroscopic charcoal isolated from sediments. Quaternary Australasia 21(1): 18-21.

The 14,200 yr non-contiguous charcoal and pollen record from Gooches Swamp is presented in Chapter 3 and this is currently in press:

• Black, M.P. and Mooney, S.D., (in press) Holocene fire history from the Greater Blue Mountains World Heritage Area, New South Wales, Australia: the climate, humans and fire nexus. Regional Environmental Change.

The Gooches Swamp palaeoenvironmental record was re-examined in Chapter 4 and here I present a 14,200 yr contiguous record of fire from this site. Humification and more detailed comparisons to the archaeological record for Gooches Swamp are also presented in this chapter. Chapter 4 has been submitted to Quaternary Science Reviews as:

• Black, M.P. and Mooney, S.D. (submitted) The archaeological and climatic implications of a 14 200 yr contiguous fire record from Gooches Crater, Blue Mountains, Australia. Quaternary Science Reviews, submitted November 2005.

The >43,000 yr pollen and charcoal record from Lake Baraba is presented in Chapter 5. This chapter has been submitted to Quaternary Science Reviews as:

• Black, M.P., Mooney, S.D. and Martin, H. (submitted) A >43 000 year vegetation and fire history from Lake Baraba, New South Wales, Australia. Quaternary Science Reviews, submitted October 2005.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 1 33

In Chapter 6 a ~6,200 yr record of fire and vegetation from Kings Waterhole is presented and this has been submitted for publication in Australian Geographer as:

• Black, M.P. and Mooney, S.D., (submitted) A ~6,100 yr vegetation and fire history from Kings Waterhole, Wollemi National Park, New South Wales. Australian Geographer, submitted November 2005.

Finally, the three charcoal records are compared to each other, local archaeological data and regional palaeoclimatic data to untangle the nexus of fire, humans and climate (Chapter 7). This final chapter has been submitted to The Holocene for publication as:

• Black, M.P., Mooney, S.D. and Haberle, S. (submitted) A Late Quaternary palaeoenvironmental investigation of the fire, climate and human nexus from the Sydney Basin, Australia. The Holocene, submitted November 2005.

The key findings of this study are summarised in Chapter 8.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 2 34

Chapter 2: A simple and fast method for calculating the area of macroscopic charcoal isolated from sediments

2.1 Introduction Mooney and Radford (2001) presented a method for the quantification of larger charcoal fragments, which had been washed from volumetric sub-samples of sediment. As noted in that paper, there has been a distinct move towards the quantification of larger charcoal fragments overseas, as it is believed that these particles better reflect the occurrence of fire at the catchment (spatial) scale (Clark, 1988). ‘Larger particles’ or ‘macro-charcoal’ are charcoal particles that have at least one axis longer than those traditionally quantified in palynological studies (eg. Clark, 1982). In practice, this separation between ‘point-count charcoal’ and macro-charcoal probably occurs at about 50μm.

The method described in Mooney and Radford (2001) was only slightly modified from what has become known as the ‘Oregon sieving method’. Descriptions by Millspaugh and Whitlock (1995), Long et al. (1998) and Gardner and Whitlock (2001) provide details of the method as used by the members of the Environmental Change Research Group in the Department of Geography at the University of Oregon, USA. Mooney and Radford (2001) also noted that image-analysis software could be usefully applied to the quantification of macro-charcoal.

Charcoal is typically been expressed charcoal (of a particular size fraction, for example >250μm) as abundance (i.e. number of particles) per unit of (fresh) sediment volume (no./cm3), for example in Mooney et al. (2001). Nonetheless, there are several limitations with expressing charcoal in this way. Abundance of charcoal per unit volume may be most appropriate where the size of individual charcoal fragments remains relatively constant throughout a (sediment) profile, which is unlikely if the distance to the source of the charcoal (i.e. the fire), or fuel type, to give two examples, varies through time.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 2 35

It is also easy to imagine how portraying charcoal as a count per unit volume may result in erroneous trends. For example, one large charcoal fragment approaching 1.0 cm3 in a 1.0 cm3 sediment sub-sample will result in a count of 1.0. An adjacent sub-sample with 10 small fragments in an equivalent volume would appear to have 10 times as much charcoal. This is obviously a problem, considering that charcoal is generally brittle, so any breakage (due to handling etc) could result in a higher charcoal count (Clark, 1984). One obvious solution to this is to quantify charcoal in terms of area per volume of sediment.

As Clark and Hussey (1996) have previously noted, comparison between sites is best accomplished with standard units. In a review of Australian fire history Kershaw et al. (2002, p. 5) noted “there is also a lack of consistency in methods of counting and portrayal of charcoal data” and that this made their overview more difficult. It is hence also the aim of this paper to provide a simple method, thereby potentially facilitating future comparisons between sites in Australia.

2.2 The Method Follow Steps 1 –3 in Mooney and Radford (2001). This results in charcoal of a known size fraction sieved from a known volume of sediment dispersed in water in a petri-dish. The concentration of material should be such that very few charcoal fragments are overlapping: experimenting with the volume of sediment used in the sieving operation can reduce this concern. This study has used charcoal fragments greater than 250μm and so no comments on the efficiency of the method for smaller charcoal particles can be made. Nonetheless, it is suspected that the recognition of smaller charcoal particles (at Step 5 below) may be a problem.

The method described below uses Scion Image, which is free software available at www.scioncorp.com. Before downloading the software you are required to register, however, this is a simple and fast procedure. Scion Image for Windows is based on NIH Image, which runs on a Macintosh platform. The method described below uses Scion

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 2 36

Image release Beta 4.0.2. The method also requires a digital camera to capture images and Adobe Photoshop to save the images in the required format.

2.2.1 Step 1. Acquiring a digital image Using a digital camera take a fine-resolution image of the petri-dish with the collected material. The method works best if the charcoal fragments are towards the centre of the petri-dish. The use of a tripod to support the camera keeps the images at a set size and minimises movement. A scale (such as a ruler) and a label (including site name, depth of sample etc) should be placed next to the petri-dish. Adjust the zoom of the camera and/or the height of the tripod so that the entire petri-dish, scale and label are included in the image. If extra light is needed placing the petri-dish on a light table minimises shadows. Keep the petri-dish (with collected material) for further use (see Step 5 below).

Repeat Step 1 until a good image of all samples has been obtained. Download the images from the camera into a labelled folder on the PC with Adobe Photoshop and Scion Image already installed.

2.2.2 Step 2. Formatting the image for processing Open one of the images using Adobe Photoshop. First re-size the image, adjust the contrast and brightness and then save the image as a bitmap file (*.bmp). If you are familiar with Photoshop you can record all of this as an “Action”. The images have been reduced to 35% (with ‘constrain properties’ ticked and using the ‘bicubic interpolation’) before saving them. Scion Image analysis can also be used with *.tif formatting but not *.jpg.

2.2.3 Step 3. Calibration Open the *.bmp file in Scion Image. Calibration sets the scale for the image and can be done by selecting the tool that allows you to draw a straight line (5th tool down on the right in the tool box, see Figure 2.1). Go to the ruler (within your image) and draw a straight line over a known distance (X). Then select >Analyse >Set scale from the drop-down menus. Fill in the table that appears, making sure that you first change the units to “mm” and fill in

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 2 37

your known distance (X). To check if the calibration of the scale is correct select >Analyse >Measure and then >Analyse >Show results (this should tell you the distance in mm along your straight line, X). You should only need to do this calibration once in any session (i.e. while Scion Image is open), assuming that the size of your images does not change. Nonetheless, it may be a good idea to check the calibration every 10 samples or so.

2.2.4 Step 4. Setting measurement parameters Select >Analyse >Options from the drop-down menus and set the maximum measurement as 8000. Then use the circle tool (2nd tool on the right in the tool box, see Figure 2.1) to select the area to be analysed (remembering to hold the shift key down to make it a perfect circle). The area to be analysed should include all charcoal particles within the petri-dish. The selected area can also be moved around using the arrow keys.

Figure 2.1. The tool-box in Scion Image with the necessary tools highlighted.

2.2.5 Step 5. Analysing and Showing Results Select >Options >density slice from the drop-down menus. The density slice option allows all pixels between an upper and lower threshold to be selected. The LUT (Look-Up Table) tool bar, which should then appear on the left hand side of the screen, is used to set these

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 2 38

thresholds. This is the most subjective stage of the procedure, and hence it is handy to have the original sample (petri-dish with charcoal) on hand to compare with the selection from Scion. Moving both the upper and lower limit in the LUT tool bar up or down either selects or deselects material in the petri-dish (see Help in Scion). By comparison with your original all charcoal particles in the sample should be highlighted in the Scion image. This may take some practice: experimenting with charcoal in water solutions with and without other materials may help this process. The lower threshold is always set on the LUT tool bar to the base (256 or pure black) but vary the upper threshold so that all charcoal particles are selected.

Select >Analyse >Analyse particles from the drop-down menus, making sure the following boxes are ticked: “Label”, “Outline”, “Reset”. Also make sure that min = 1, max= 99999 are chosen.

Select >Analyse >Show results from the drop-down menus and a table will appear showing the results of the different parameters for each charcoal fragment. Select >Edit >Copy measurements from the drop-down menus and the results can then be pasted into an Excel workbook. Areas, expressed as mm2, can then be summed to get the total area of charcoal for that sample. The number of particles can also be recorded. Other useful statistics can also be calculated, such as modal charcoal size, standard deviation etc.

2.3 Concluding Remarks There are still several important issues regarding fire history in Australia that are contentious or poorly understood. This suggests that much work is still required. The image processing method described above is relatively simple and with practice, it can be extremely fast. This speed then allows more samples to be analysed, allowing either a finer temporal resolution or a better spatial resolution of study sites.

The optically measured charcoal data has been compared to the image analysis method described herein and there is a very high correlation. Figure 2.2 demonstrates this

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 2 39

correlation (r2= 0.71, p<0.01) using data derived from a site Gooches Crater (Right). Gooches Crater (Right) is an unnamed valley swamp in an adjacent catchment to Gooches Crater, which occurs to the west of Sydney (at approximately 33º27’255”S, 150º 16’020”E). In Figure 2.2 some differences between the charcoal counts and charcoal area are discernible, particularly in the top 130 cm. Such differences are to be expected and may reflect relevant variables such as changing fuel sources or fire intensity. Comparison between charcoal counts and area is possible with the method described herein, as the number of particles summed in Step 5 gives the total count per cubic centimetre of sediment.

Charcoal Count Charcoal Area (no./cm 3) (mm2/cm 3) 1000 6000

900

5000 800

700 4000

600

500 3000

400

2000 300

200 1000

10 0

0 0 -10 90 190 290 390 490 590

dept h ( cm) macro area macro count

Figure 2.2. A comparison of the count and image analysis-derived charcoal data from the Gooches Crater (Right) site.

Finally, the method described herein results in charcoal data expressed as area per cubic centimetre of sediment. This is similar to the results provided by ‘Point Count Charcoal’ (Clark, 1982) and hence provides a means for comparison with previous work. Furthermore, the method is simple and relatively fast, and is compatible with a method that is becoming standard overseas.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 40

Chapter 3: Holocene fire history from the Greater Blue Mountains World Heritage Area, New South Wales, Australia: the climate, humans and fire nexus.

ABSTRACT This study presents a reconstruction of the fire activity of the last ~14 200 cal. yr BP from Gooches Crater Right. Charcoal analysis and palynology were undertaken with the aim of untangling any inter-relationships between climate, humans and fire. The study also aimed at contributing to the management of fire in the contemporary environment. Gooches Crater Right is located on the Newnes Plateau, approximately 150km to the west of Sydney (~33º27’S, 150º16’E). The site is within the Blue Mountains National Park, which forms a part of the Greater Blue Mountains World Heritage Area. A sediment core was retrieved and macroscopic charcoal (>250 µm) was analysed using a modified version of the ‘Oregon sieving method’. Selected pollen and spores were also quantified to examine aspects of the vegetation through time. A chronology of the site was provided by radiocarbon dating. In the late glacial-Holocene transition charcoal was variable including a peak between the Antarctic Cold Reversal (~14 500 to 12 900 cal. yr BP) and the Younger Dryas stadial (~12 700 to 11 500 cal. yr BP). Charcoal was less abundant between ~9 000 and 6 000 cal. yr BP however this was followed by a relatively abrupt increase in charcoal at 5 500 cal. yr BP, after which charcoal then remains high until ~3 400 cal. yr BP. Another abrupt increase occurs at 3 000 cal. yr BP. The post-European period has witnessed charcoal accumulating at levels unprecedented in the previous ~14 200 years. The dominant control on fire in this environment during the Holocene appears to be climate. Periods of climate change, identified in previous studies, are associated with higher levels of fire activity. Fire was less ubiquitous between ~9 000 to 6 000 yr BP, a period normally described as having higher effective moisture in south-eastern Australia. The mid-Holocene fluctuations in charcoal may reflect anthropogenic fire, climatic forcing or alternatively human responses to any climate change. Coeval changes in palaeoclimatic sequences elsewhere and palynology at the site support a climatic explanation or that Aboriginal people used fire within a climatic framework.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 41

3.1 Introduction

There are several contentious and poorly understood issues regarding the history of fire in the humid environments of south-eastern Australia (e.g. Bowman, 1998). These include questions regarding how Aboriginal people utilised fire in various landscape contexts; whether their use of fire was constant through time; and how strongly anthropogenic activity has influenced fire history when compared to any climatic changes.

Fire was used by Aboriginal people, according to the popular ‘Fire-stick Farming’ thesis by Jones (1969), to acquire or to manipulate and thereby increase the availability of resources. The fire regimes of the various Aboriginal people are often depicted as being of high frequency and low intensity and being applied across much of the Australian continent. Head (1989, p. 41) noted that there is a common assumption that Aborigines “had a single ongoing impact”, potentially ignoring climatic change and population and cultural change.

Although poorly understood, and despite some cautionary caveats (e.g. Gill, 1977), assumptions about Aboriginal use of fire prior to European invasion are often used to justify contemporary intensive prescribed burning regimes. Frequent and uniform landscape firing has been found to detrimentally affect biodiversity in Australian terrestrial ecosystems (e.g. Gill and Bradstock, 1995) and has been listed as a ‘Key Threatening Process’ under the New South Wales' Threatened Species Conservation Act (1995).

Prior to 1970, Aboriginal occupation, demographics and socio-economics was viewed as static (e.g. see Mulvaney, 1971), despite early delineation of late Holocene change (e.g. McCarthy, 1964). Lourandos (1980; 1983; 1997) subsequently identified the mid-to-late Holocene as a period of continent wide changes in Aboriginal Australia. From the mid- Holocene various changes in technology, settlement patterns, social structures and population densities are thought to have intensified occupation (Lourandos, 1983). This change may have altered resource management strategies including the use of fire. It should

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 42

be noted that Lourandos’ (1980; 1983) intensification model has received some criticism (e.g. see Head, 1996).

Several researchers have described an increase in archaeological visibility and the use of sites in the late Holocene including in the Sydney Basin (e.g. McCarthy, 1964; Stockton, 1970; Stockton and Holland, 1974; Attenbrow, 1982; 2003) and further afield in south- eastern Australia (e.g. Hughes and Lampert, 1982; Smith, 1982; Beaton; 1983; Ross, 1985).

The mid-Holocene has also been identified as a period of climate change in Australasia (Shulmeister, 1999) and further afield (Rodbell et al., 1999; deMenocal et al., 2000; Sandweiss et al., 1999). In Australasia Shulmeister (1999) described a decoupling of the northern (tropical) and southern (temperate) climate systems of Australia at ~5 000 yr BP. In southern Australia this has been described as resulting in increased westerlies, the loss of summer monsoon rainfall and a sharp decline in effective precipitation (Shulmeister, 1999).

Increased seasonality in the south-western Pacific since ~5000 yr BP appears to be related to the El Niño-Southern Oscillation (ENSO) (Shulmeister, 1999). Rodbell et al. (1999) argued that ENSO progressively achieved modern characteristics by ~5 000 cal. yr BP. Sandweiss et al. (2001) described that ENSO events occurred at a low frequency from ~5 800 yr BP, followed by an increased frequency from 3 200 yr BP. Riedinger et al. (2002) also described an increase in the intensity and frequency of El Niño events from 3 100 cal. yr BP.

This study aimed to investigate the post-glacial history of fire at Gooches Crater Right, located in the Blue Mountains to the west of Sydney, using charcoal analysis and palynology. The objective was to examine the timing of any change in fire activity and to see if any change could be better related to proposed changes in Aboriginal occupation or climatic controls. This research forms a part of a larger study investigating this climate, humans and fire nexus in the Sydney Basin. An overall objective of this and the larger

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 43

study is to contribute to the management of fire in the contemporary environment by providing a longer temporal perspective of fire activity.

3.2 Study Area

The Blue Mountains form the elevated western edge of the Sydney Basin, which is a Triassic sandstone-dominated depositional basin in humid south-eastern Australia (Branagan, 1979). Gooches Crater is located on the Newnes Plateau, which has an altitude of between 900 to 1200m above sea level (asl), in the northwest of the upper Blue Mountains. The site (at 33º27’116”S, 150º16’020”E, ~960 m asl), approximately 150 km to the west of Sydney (Figure 3.1), is located within the Blue Mountains National Park, which is incorporated into the Greater Blue Mountains World Heritage Area.

Associated with Gooches Crater, and nearby, are several swamps. The work described herein is from Gooches Crater Right (GCR), which is a narrow, elongate swamp in a low slope headwater valley adjacent to Gooches Crater. GCR contains ~6 m of sandy organic sediments and is currently vegetated with a closed wet heath (dominated by Baeckea, Epacris, Gleichenia, Grevillea, Gymnoschoenus, Leptospermum). Eucalypt woodland and open heath surrounds the site (Benson and Keith, 1990). The Blue Mountains are notoriously fire prone with a natural fire season occurring from October to February (Cunningham, 1984).

The climate of the Newnes Plateau is temperate with an average minimum of -1oC in July and in the hottest month, January, an average maximum of 23.5oC (BoM, 2003). The site receives an average annual rainfall of 1047 mm, which is influenced by a mild orographic effect.

There is conflicting evidence for the timing of first Aboriginal occupation of the Blue Mountains. Based on archaeological evidence from a site on the Kings Tableland, Stockton

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 44

Figure 3.1. The location of Gooches Crater Right (GCR). (Photo source: Wallerawang Run 12, photograph 14 June 1998, Surveyor-General’s Department, New South Wales).

and Holland (1974) argued for 22 240 yr however the consensus suggests human arrival from the late glacial period (Bermingham, 1966; Johnson, 1979; Bowdler, 1981). Bowdler (1981) suggested that there was sporadic occupation of the Blue Mountains between 14 000 and 12 000 yr BP followed by a hiatus and then intensification of occupation associated with the Small Tool Tradition.

The Australian Small Tool Tradition saw an enormous expansion of flaking techniques and activity and the addition of smaller implements to the stone tool kit of Aboriginal people. Various authors have dated its introduction to the Blue Mountains at about 4 000 yr BP (Johnson, 1979; Bowdler, 1981; Stockton 1993), however Bermingham (1966) had previously suggested abundant Small Tool Tradition artefacts from 5 300 yr BP.

Ethnographic sources suggest that Dharug (or Daruk) people were the dominant Aboriginal group to have inhabited the Blue Mountains region (Gollan, 1987; Kohen, 1993). Nonetheless, the Newnes Plateau was probably the western margin of Dharug territory (Kohen, 1993) and it may have been a place of interaction or a corridor between the Dharug and people to the west or south (Gollan, 1987).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 45

The rugged terrain of the Blue Mountains was a constraint for early European exploration, settlement and development (Mackaness, 1965). By the time Europeans had found a route over the Blue Mountains in 1813 there were very few Aboriginal people remaining as a result of small pox epidemics and other diseases. Massacres and the destruction of traditional resources resulted in the further demise of Aboriginal populations such that traditional lifestyles had almost completely disappeared from the Blue Mountains region by 1820 (Breckell, 1993; Turbet, 2001). Currently national parks, forestry and sand mining exist on the Newnes Plateau.

3.3 Methods

A 3.55 m sediment core was extracted from GCR swamp using a Russian d-section corer (Jowsey, 1966) in June 2002. A further 2.35 m was extracted in September 2002 giving a sedimentary sequence of 5.9 m. Subsequent dating of this additional core has proven to be problematic, hence only the upper 3.55 m of the sedimentary sequence will be discussed in this paper. The stratigraphy of the core was described using a modification of the Troels- Smith method (Kershaw, 1997) and was photographed. Four sections of the core (48-53, 80-90, 150-156 and 295-307 cm) were submitted for radiocarbon dating. Macroscopic charcoal, which is thought to represent local or catchment fire events (Whitlock and Millspaugh, 1996), was analysed using a modified version of the ‘Oregon sieving method’ (Long et al., 1998) and image analysis. Volumetric sub-samples at 5 cm increments were dispersed for 24 hr in 8 % sodium hypochlorite (bleach) to remove the pigment from organic matter and hence aid in the identification of charcoal. This material was washed through a 250 μm sieve and the collected material was photographed in a petrii dish using a digital camera (Nikon Coolpix 4500). The area of charcoal was calculated using image analysis software (Scion Image Beta 4.02 for Windows) and charcoal was also counted using a dissecting microscope (40 X).

Pollen samples were prepared using standard palynological techniques (Faegri and Iverson, 1975). Volumetric samples were taken every 5 cm along the core and exotic pollen (Alnus)

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 46

was added as a ‘spike’. The samples were deflocculated with hot 10 % NaOH and then sieved through a 150μm mesh. Silicates were removed using heavy liquid (ZnBr2(aq.)) separation and organic matter with acetolysis. Samples were mounted in silicon oil and twelve pollen types/groups were counted at 400 X magnification until 200 target grains were identified. The pollen counts were expressed as percentages, with all palynomorphs contributing to the pollen sum.

Target palynomorphs were selected for their potential as climatic or fire indicators as described in Table 3.1. Asteraceae, ferns, Poaceae and Restionaceae were classed as ‘fire tolerant’ whereas Casuarinaceae, Epacridaceae, Leptospermum and Proteaceae (including Banksia) were classed as ‘fire sensitive’. A fire index was calculated as the ratio between ‘fire tolerant’ and ‘fire sensitive’ palynomorphs.

Pine plantations first occurred in the upper Blue Mountains from 1919 (State Forests NSW pers. comm.) and hence the depth of the deepest record of Pinus pollen was associated with this date. Full pollen counts were undertaken at five depths, three of which (150, 235 and 320 cm) are included in the core under investigation here. Pollen and charcoal diagrams were produced using the Tilia software package (Grimm, 1992).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 47

Table 3.1. The rationale for quantifying the target palynomorphs.

PALYNOMORPH INDICATIVE VALUE1 (habitats, fire response) Asteraceae Generally in dry woodland communities. Tend to be opportunistic species, responding to opening of the canopy after fire. Banksia At GCR Banksia predominantly reflects dry woodland communities (B. marginata, B. spinulosa) but also as a moist heath shrub (B. ericifolia). Generally considered to be sensitive to fire and frequent fire can result in localised extinctions. Other Proteaceae Includes Hakea and Grevillea, the later dominated at the site by G. acanthifolia a swamp indicator. Hakea occurs in dry heath communities. Generally sensitive to fire. Casuarinaceae In the GCR region dominated by Allocasuarina, which occur in dry open heath communities. Controversially considered fire sensitive Chenopodiaceae In the Sydney region confined to near-coastal environments. At GCR likely to reflect long-distance transport from arid landscapes to the west. Epacridaceae Often occur on swamp margins or in wet heath but some members in woodlands (e.g. Monotoca). Likely to be sensitive to fire. Leptospermum Leptospermum-type may include Baeckea. Predominantly grows on swamps and in wet heath but also in woodland. Sensitive to fire. Other Myrtaceae Includes Eucalyptus, the dominant canopy-tree at GCR, and shrubs (Callistemon, Darwinia, Kunzea, Melaleuca) common to wetter habitats. Pollen rain probably dominated by Eucalyptus, a dry woodland indicator. Response to fire variable. Pinus Introduced genus, roughly indicative of the post-European period. In the GCR region Pinus radiata plantations were established in the 1919. Poaceae Occurs as understorey in woodland communities and swamp margins. Opening of canopy after fire promotes grasses. Restionaceae Generally swamp herbs but also in wet heath. Generally tolerant to fire. Ferns and Mosses Generally wet heath and swamp habitats but also a component of forest understorey. Pollen spectra dominated by Glechenia that regenerates rapidly from rhizomes after fire; favours higher fire frequencies. Also includes Pteridium which occurs in drier locations and responds to fire.

1 Botanical information comes from Fairly and Moore (2000), Benson and Keith (1990), P. Adam and D. Keith (pers.comm.).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 48

3.4 Results

3.4.1 Core stratigraphy and chronology The analysed sediment core was composed of humified peat interspersed with bands of clay, charcoal, wood and sand. The core description identified very high levels of sand and charcoal between 104 to 132 cm. In the subsequent deeper core the clay content of the sediment increased gradually below a depth of ~480 cm, such that the base of the core was heavy clay.

The 14C dating of the deposit (Table 3.2) implies a relatively constant rate of accumulation (~0.025 cm.yr-1) in the analysed core. A linear depth-age relationship (y = 41.218x – 485.98, R2 = 0.9931, where x= depth in cm and y= age in cal. yr BP) is used for all age calculations. Based on this relationship the analysed core (355 cm) represents ~14 200 cal. yr BP. Pollen analysis revealed the first appearance of the exotic taxon Pinus at 15 cm.

Table 3.2. Radiocarbon dates from GCR. All calibrations are at the 95% level and are calibrated using INTCAL98 Radiocarbon Age Calibration (Stuiver et al., 1998). 14 13 1 Depth (cm) C date BP Lab code δ C (‰) Cal. Yr BP 2 48-53 1 760 ± 60 β-169992 -25 1 700 80-90 2 450 ± 60 β-192605 -25 2 470 150-156 4 950 ± 130 β-169993 -25 5 560 295-307 10 360 ± 140 β-169994 -25 12 190 1 ratio estimated. 2 BP = before AD 1950.

3.4.2 Charcoal and other analyses Charcoal, expressed as abundance (no.cm-3) and area (mm2.cm-3), depict similar trends (Figure 3.2) although above ~150 cm charcoal area displays greater variability than the count. There are several limitations associated with the expression of charcoal as abundance, and since there is a very strong correlation between charcoal abundance and charcoal area, only the charcoal area results will be discussed here. Charcoal has not been expressed as an influx (e.g. mm2.cm-3.yr-1) due to the apparent near-linearity of the accumulation rates for the core under consideration here.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 49

The area of macro-charcoal (Figure 3.2) is relatively high between 0-25cm (0-~550 cal. yr BP), 55-85 cm (~1750-3000 cal. yr BP), 95-145 cm (~3 400-5 500 cal. yr BP), 230-245 cm (~9 000-9 600 cal. yr BP) and 255-290 cm (~10 000-11 450 cal. yr BP). There are low levels of charcoal between 160-225 cm (~6 100-8 800 cal. yr BP), 290-315 cm (~11 400-12 500 cal. yr BP) and 325-355 cm (~12 900-14 200 cal. yr BP). The highest concentration of charcoal was found from the surface sample where the area of charcoal (875 mm2.cm-3) is almost double that of the next highest peak (437 mm2.cm-3 at 120 cm depth).

Figure 3.2. Charcoal analysis results and the fire index.

3.4.3 Pollen Myrtaceae pollen (excluding Leptospermum) is well-represented throughout the profile (Figure 3.3), with ferns/mosses, Leptospermum, Restionaceae, Casuarinaceae and Poaceae

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 50

Figure 3.3. Palynology results (all units are in percentages, unless otherwise stated).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 51

also showing relatively high percentages. The remaining pollen types are poorly represented. There appears to be a moderate level of variability in pollen assemblages between 355 cm (~14 200 cal. yr BP) and 200 cm (~7 750 cal yr BP). The variability appears to increase between 200 and 95 cm (~7 750-3 400 cal. yr BP). Full pollen counts at 235 cm and 320 cm (~9 000 cal. yr BP and 12 500 cal. yr BP) suggest varying environments at GCR from a wet heath with semi-permanent to permanent water to a fern swamp. Between 200 and 155 cm (~7 750-5 900 cal. yr BP) the record is dominated by Myrtaceae-type pollen and Casuarinaceae and Asteraceae are also elevated. At ~155cm (~5 900 cal. yr BP) there is a decline in Myrtaceae, Casuarinaceae, Leptospermum and Asteraceae representation and a sharp increase in fern and moss spores that continue to dominate the record between 155 and 95 cm (~5 900-3 400 cal. yr BP), a period that coincides with high charcoal (Figure 3.3).

High charcoal found between 95 and 145 cm (~3 400-5 500 cal. yr BP) are matched by high values in the fire index (Figure 3.2). Conversely the low levels of charcoal occurring between 160 and 225 cm (~6 100-8 800 cal. yr BP) correspond with a low fire index. There are, however, a number of exceptions to this correlation, notably at 70 cm (~2 400 cal. yr BP), 90 cm (~3 200 cal. yr BP), 290 cm (~11 500 cal. yr BP) and 350 cm (~13 900 cal. yr BP) where a high index coincides with low levels of charcoal.

3.5 Discussion

The charcoal results from GCR can be interpreted as changing fire activity in response to known climatic events or to changes to Aboriginal society and so the charcoal curve from GCR has been annotated with these influences (Figure 3.4). Haberle and David (2004) have particularly emphasised the interplay between changes in climate, culture, resources and habitats, suggesting that a complex interaction between both climate and humans must also be considered.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 52

Figure 3.4. Annotated charcoal diagram, depicting possible influences on past fire activity at GCR.

At GCR several major trends in fire activity are obvious: this includes fluctuations between ~14 000 and 9 000 cal. yr BP; a period of low fire activity from about ~8 900 to 6 100 cal. yr BP; a dramatic increase in fire activity from ~5 500 cal. yr BP; and a period of increased fire activity between ~1 750 and 3 000 cal. yr BP. The fire activity of the later half of the Holocene appears distinctly different from the earlier Holocene, and this includes unprecedented levels of charcoal in the post-European period.

Using microscopic charcoal and a number of sites located in tropical eastern Indonesia and New Guinea (Melesia), Haberle et al. (2001) reconstructed the history of fire over the last 20 000 years at a century-scale resolution. In the period of overlap with this study Haberle et al. (2001) identified identical trends. This included high variability in charcoal for the

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 53

late-glacial-Holocene transition, low charcoal in the early Holocene and highly variable charcoal values in the later half of the Holocene.

The GCR charcoal record suggests that fire activity was extremely variable between ~14 200 and 9 000 cal. yr BP (Figure 3.4). This variability includes three peaks centred on ~12 700, ~10 600 and ~9 400 cal. yr BP and periods of low fire activity ~14 200-12 900 and ~12 500-11 400 cal. yr BP. This trend is reminiscent of palaeoclimatic records of the late glacial-Holocene transition which includes the Antarctic Cold Reversal (~12 900 to ~14 500 yr BP) (Blunier et al., 1997) and the Younger Dryas stadial dated between 12 700 ± 100 and 11 550 ± 70 ice core years in the GRIP ice core (Johnsen et al., 1992), and between 12 940 ± 260 and 11 640 ± 250 ice core years in the GISP2 ice core (Alley et al., 1993).

There has been conflicting evidence concerning the Younger Dryas (YD) event in the Southern Hemisphere. Nonetheless, Goede et al. (1996) found evidence for a YD in oxygen and carbon isotopes ratios of a speleothem from Buchan Cave in Victoria, south-eastern Australia. Haberle et al. (2001) also found a reversal of high charcoal values coeval with the YD and suggested that a relatively cool phase may have altered soil moisture and the vulnerability of the vegetation to fire.

Turney et al. (2003) described a cool oscillation that coincided with the ACR , based on palynological and geomorphological research from a number of sites throughout New Zealand. The phase between 14 000 and 11 500 yr BP is characterised with an initial cooling period followed by a sustained warming that almost exactly corresponded with the YD event in the Northern Hemisphere (Turney et al., 2003).

The GCR charcoal record reveals a peak in fire activity in the period between the ACR and YD, less fire during the YD, and higher activity at the end of the YD. Haberle et al. (2001) particularly highlighted the importance of the relative stability of climate to fire activity. It is hence possible that the increase in fire activity between ~12 900 and 12 500 cal. yr BP was associated with climatic instability during the transition from the ACR phase to the YD

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 54

phase. Likewise the increase in charcoal after ~11 400 cal. yr BP may reflect the climatic instability associated with the termination of the YD phase.

The time of the ACR corresponds with a slight increase in representation of ferns and mosses and a decreased representation of Leptospermum (Figure 3.3). Whereas the YD phase saw a slight increase in the representation of fire-sensitive Leptospermum and small reductions in the representation of fire-tolerant Asteraceae, Poaceae and Restionaceae, vegetational changes which are consistent with a reduction in fire activity. Despite these observations there are no dramatic changes in the pollen spectra coincidental with the YD. It is possible that the sclerophyllous sandstone vegetation that dominates the Sydney Basin, including around GCR, is relatively insensitive to climatic changes or that the climatic changes associated with the YD and ACR were relatively minor in this setting. Nonetheless the GCR record suggests a reduction in fire during the YD.

Alternatively, the arrival of Aboriginal people to the Blue Mountains may have affected fire activity (Figure 3.4). With climatic amelioration the Blue Mountains was likely to have increasingly become a favourable habitat for people between 12 000 and 10 000 years ago (Stockton, 1993). Bowdler (1977) earlier argued that the highlands were not used until after climatic amelioration.

Stockton and Holland (1974) suggested that the permanent habitation of the Blue Mountains, especially at higher altitudes, depended on a favourable climate. It is generally assumed that the climate of the Blue Mountains limited occupation of higher altitudes such that they were preferentially used in summer (Stockton, 1993). The variability in charcoal from ~14 200 to 9 200 cal. yr BP may therefore represent the displacement of Aboriginal people from the upper Blue Mountains as the climate became less suitable for habitation, perhaps during the ACR or YD events, followed by their return as the climatic conditions improved.

Haberle and Ledru (2001), in discussing Central and South American fire history, also found elevated charcoal values prior to the Holocene and attributed this to a combined

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 55

impact of rapid climate change and humans in the landscape. Complex feedback between climate, humans and fire is likely to be a better explanation for the fire activity at GCR between ~14 200 and 9 200 cal. yr BP.

The period, ~8 800 to 6 100 cal. yr BP, is characterised by low levels of charcoal coupled with a relatively low variability in the record. In south-eastern Australia Goede et al. (1996) suggested that summer temperatures were depressed by ~2oC between 9 000 and 8 000 yr BP and a period of maximum effective precipitation has been identified in southern Australia between about 7 000 and 5 000 yr BP (Bowler, 1981; Shulmeister, 1999). As noted, Haberle et al. (2001) also found low levels of charcoal associated with this period and attributed the low fire activity to climatic stability and reduced seasonality. Notably, Bowdler (1981) suggested a hiatus in the occupation of the Blue Mountains by Aboriginal people in the early-to-mid Holocene.

The fire activity and the fire index at GCR increases dramatically between ~6 000 and 5 000 cal. yr BP. The increase in the fire index is largely a result of the increased representation of ferns in the pollen assemblage. Gleichenia, the dominant fern found throughout the GCR record, grows on damp to wet ground (Fairley and Moore, 2000) so their increased abundance could represent hydrological changes or swamp development associated with a moister climate (David Keith pers. comm.). It is possible that increased effective precipitation could enhance fuel loads however this is inconsistent with Shulmeister’s (1999) suggestion that there was a sharp decline in effective precipitation in southern Australia at this time. But notably Pickett et al. (2004) suggested a wetter climate for some areas including east of the Great Dividing Range.

The increased representation of ferns between ~5 500 and 3 500 cal. yr BP may also reflect a response to the higher fire activity as suggested by the high levels of charcoal during this period. Gleichenia regenerates rapidly from rhizomes after fire and can form dense thickets (David Keith pers. comm.).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 56

Notably, a study by Martin (1994) from coastal Sydney (Kurnell Fen) also revealed an increase in macroscopic charcoal from 5 500 cal. yr BP. Martin’s (1994) palynological data provided no evidence of climate change at this time and hence the increase in charcoal was attributed to Aboriginal activities. This lack of evidence for climate change could, however, reflect the moderating influences of Martin’s (1994) coastal location. Further afield, at Lynchs Crater, Kershaw (1983) also found an increase in charcoal between 6 000 and 5 000 yr BP.

The increase in charcoal at GCR from ~5 500 cal. yr BP is approximately coeval with theorised changes in both climate and Aboriginal occupation. The hypothesised ‘intensification’ from the mid-Holocene ascribes significant change in Aboriginal Australia to altered social and economic systems (Lourandos, 1980; 1983). Changes in Aboriginal technology, resource use, settlement patterns, art, exchange systems and burial have been interpreted as supporting this paradigm (Rowland, 1999). Lourandos (1997 p. 299), however, clearly linked intensification with a progressive and continued increase in Aboriginal population levels from the mid-Holocene. Earlier Lourandos (1980) had also argued that increased use of fire might be one result of intensification.

Assuming this model of anthropogenic change is correct and that Aboriginal strategies were associated with fire (e.g. Jones, 1969), fire activity should continue to increase from the time of Holocene intensification. In the Blue Mountains region intensification occurred from about 4 000 yr BP (e.g. Stockland and Holland, 1974; I. Johnson, 1979; Bowdler, 1981; Flood et al., 1987). Bermingham’s (1966) earlier date for the Small Tool Tradition assemblage, 5 300 yr BP, may reflect the limitations of radiocarbon dating at that time. Overall, these results suggest that the mid-Holocene increase in fire activity at GCR precedes any significant intensification of Aboriginal society by at least 1 500 years.

Furthermore, the intensification model depicts a sustained increase in socio-economic activity from the mid-Holocene. The charcoal record at GCR shows a dramatic decline in fire activity centred on 3 500 cal. yr BP followed by relatively high activity between ~3 000 and 1 750 cal. yr BP. Sustained anthropogenic influence under the ‘intensification’ model

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 57

therefore cannot be the sole determinant of fire at Gooches Crater. Interestingly Hiscock and Attenbrow (2004) described a period of relatively high production of Aboriginal backed-artefact between ~1 500 and 3 500 yr BP from a site ~35 km north of GCR and Attenbrow (2003) found a dramatic increase in the number of artefacts from about 3 000 yr BP from a site ~100 km to the east of GCR.

As an alternative to anthropogenic influences the changed fire activity at Gooches Crater from ~5 500 cal. yr BP may be a response to climate. As described this timeframe is associated not just with the onset of modern ENSO (Rodbell et al., 1999; Sandweiss et al., 2001), but also with a reorganisation of climates in Australasia (Shulmeister, 1999). Riedinger et al. (2002) described considerable millennial-scale variability in El Niño events since the mid-Holocene and notably found few events between 5 000 and 4 000 14C yr BP when charcoal at GCR is high. In fact Riedinger et al.’s (2002) record of El Niño events, derived from the Galapagos Islands, is often out of phase with fire activity at GCR. Although large fire events in eastern Australia are often popularly linked with El Niño induced droughts, Cunningham (1984) has demonstrated that historic fires in the Blue Mountains normally occur in the first dry fire season following extended periods of above average rainfall.

Shulmeister (1999) argued that after about 5 000 yr BP a relatively sudden change to pressure systems saw the loss of summer monsoon rainfall in southern Australia and the strengthening of the mid-latitude westerlies. This hypothesis can be tested at GCR: Chenopodiaceae in the Sydney Basin, including Sarcocornia, Enchylaena, Suaeda and Rhagodia, are confined to coastal locations (Fairley and Moore, 2000) and so their representation at GCR is likely to represent long-distance transport from arid environments to the west of the site. There is a more consistent representation of Chenopods from about the mid-Holocene, although any difference between this period and the early Holocene is subtle (Figure 3.3). This avenue of investigation is currently under further consideration.

The high fire activity since the mid-Holocene at GCR therefore cannot be simply attributed to either innovations in human society or climatic change. Rather, the higher fire activity

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 58

may reflect increased climatic variability, as seems likely since ~5 000 yr BP, or a complex nexus between climate and human society. For example the D’harawals, the group of Aboriginal people from south-eastern Sydney, recognised three distinctive cycles that affected the weather and “used the indicators of those cycles to predict when to burn the bushland…even the burning of the bushland was not a haphazard exercise” (Bodkin, 2004).

The remarkable similarity between Haberle et al.’s (2001) eastern Indonesian and New Guinea sites and the record from GCR, despite the obvious differences in Holocene human occupation and subsistence strongly implies that climate has been the dominant influence over the history of fire. Haberle et al. (2001) suggested influences from the relative position and intensity of the Walker Circulation and the austral summer monsoon. ENSO phenomena are also a key factor (Haberle et al. 2001). Although these potential influences are predominantly tropical systems the Australian climate is strongly influenced by them (e.g. see Shulmeister, 1999).

These mid-Holocene climatic anomalies also approximately coincide with the abrupt cessation of the African Humid Period at ~5 500 cal. yr BP (deMenocal et al. 2000). deMenocal et al. (2000) suggested that this reorganisation involved teleconnections with the Asian monsoon and linked the cessation of the period with the crossing of a critical threshold in insolation.

Precessional influences on the climate of south-eastern Australia must also be considered in the GCR record of fire activity. Data contained in Berger (1992) indicates that at 30o S December insolation has been increasing over the Holocene whilst June insolation has been decreasing. In the early Holocene Haberle et al. (2001) attributed a lower fire activity to this slightly reduced seasonality and a relatively stable climate in the Melesian region. The history of fire activity at GCR raises the possibility of a non-linear response to the slowly increasing seasonality: again this may reflect direct climatic forcing or responses in human systems to accommodate such changes.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 59

3.6 Conclusion

Macphail (1983) suggested that Aboriginal use of fire may either reinforce or oppose trends in vegetation caused by climate change. Here another factor, the inter-relationships between climate, humans and fire, should be considered.

It is possible that climate may result in a change in fire activity directly, for example to ignition of fire via lightning, and at GCR any climatic system which includes dry electrical storms may be important. More probably climate is likely to influence fire indirectly via vegetation. At GCR changes in vegetation appear to be associated with fire activity, such that climate influences vegetation through the intermediary of fire. This is not surprising considering the fire-prone sclerophyllous vegetation of the site. At GCR the vegetation appears to be relatively resilient to climate but is more greatly affected by fire, suggesting that fire history may be a more sensitive index of environmental change than palynology.

This study has highlighted an apparent increase in fire activity during periods of climate change, for example during the ACR-YD transition and the mid-Holocene. Haberle et al. (2001) came to a similar conclusion in Melanesia.

Despite the potential interactions between climate, humans and fire over the last ~14 200 years climate appears to be the dominant control of fire activity at Gooches Crater Right. This is not true, however, in the recent historic past, which has a high fire activity without precedent in the previous ~14 200 years. The suggestion that climate is a dominant control of fire activity in south-eastern Australia is very much at odds with the prevailing paradigm which depicts Aboriginal people as controlling regimes in the pre-European period. This conclusion may also imply that the use of fire for resource manipulation by Aboriginal people in the Sydney Basin has been overstated. Bowman and Brown (1986: 166) have previously suggested that fire-stick farming had received “too little critical examination”, with attendant circular arguments and hence it had become a “self-fulfilling prophecy”.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 3 60

This research forms a part of a broader project investigating post-glacial palaeoecological records from several sites within the Sydney Basin. The sites have a similar flora, depositional environment and, presumably, have been subject to broadly similar climatic patterns in the past but have been chosen as they were occupied by different Aboriginal groups. Any synchroneity in the fire activity at these sites will provide a test of the hypothesis that climate appears to be the dominating factor on past fire activity in the Sydney Basin. Alternatively any differences could imply that the fire activity was controlled by people. The results of this broader study will potentially disentangle the nexus between climate, fire and humans.

In terms of management of fire in the contemporary environment, the palaeoenvironmental record at Gooches Crater Right reveals that fire is a significant variable in the environment. There is no single pre-European fire regime that can be recommended as a management target or applied: instead several regimes have existed, each tied to the prevailing climate of the time. Furthermore, the Gooches Crater Right study highlights the important influences of climate change including ENSO on fire activity. This suggests that fire activity is likely to become an increasing concern with projected rapid anthropogenic climate change in our near future. How ENSO responds to any anthropogenic climate change is likely to be critical to future fire regimes in south-eastern Australia.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 61

Chapter 4: The archaeological and climatic implications of a 14 200 yr contiguous fire record from Gooches Crater, Blue Mountains, Australia

ABSTRACT

This study presents a contiguous macroscopic charcoal record of the last ~14 200 cal. yr BP from Gooches Crater Right (GCR), located on the Newnes Plateau, approximately 150km to the west of Sydney (~33º27’S, 150º16’E). Macroscopic charcoal (>250 µm) was analysed using contiguous sub-samples and geochemistry, humification and loss-on- ignition were also analysed in a radiocarbon-dated sediment core. The overall aim of the research was to assess any relationships between humans and fire through time primarily by comparison of these palaeoenvironmental indices with archaeological information from the region. Climate can be used to explain all periods of change in the charcoal record with variable levels of charcoal associated with the late glacial-Holocene transition, low levels of charcoal associated with the relatively stable climate of the early Holocene and high and variable charcoal associated with the onset of the modern El Niño events from the mid- Holocene. Although the dominant control on fire in this environment during the Holocene appears to be climate, several fluctuations in charcoal may reflect anthropogenic fire or human responses to climate change. This includes increases in fire activity during the mid- Holocene and at ~3 000 cal. yr BP the later of which best corresponds to changes in archaeological records. Changes in the fire activity at Gooches Crater Right do not support a continuous late Holocene intensification of Aboriginal subsistence and the most parsimonious explanation for the data are that Aboriginal people used fire within a climatic framework.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 62

4.1 Introduction

This study aims to compare past fire activity at Gooches Crater Right, located in the Blue Mountains, Australia, ascertained from palaeoecological analysis of sediment and especially charcoal analysis, with regional archaeological data. This was undertaken in an attempt to better resolve any anthropogenic influences on fire. Recently Black and Mooney (in press) suggested that climate was a significant control of past fire activity at this site, an interpretation that conflicts with the dominant paradigm in Australia, which depicts Aboriginal people as strongly controlling prehistoric fire activity. In an extensive review of impacts associated with Aboriginal use of fire in Australia, Bowman (1998) highlighted that palaeoecological and archaeological information used together may lead to a better understanding of fire history.

The interpretation of the palaeoenvironmental record of fire is often difficult as either climate (directly or via vegetation) or human activity may be responsible for any change. Assessing the influence of climate and humans is a worldwide, significant endeavour in palaeoenvironmental studies (e.g. the IGBP PAGES Focus 5 project; HITE, 2003). This study presents a 14 200 cal. y BP contiguous charcoal record and aims to further investigate the nexus between fire, climate and humans.

In Australia the popular paradigm describes that fire was used by Aboriginal people to acquire or to manipulate and thereby increase the availability of resources (Jones, 1969; Nicholson, 1981). It has been argued that this ‘Fire-stick Farming’ (senso Jones, 1969 and hereafter FSF) lead to vegetation change and other environmental impacts (e.g. Singh et al., 1981; Hughes & Sullivan, 1981; Flannery, 1994; Miller et al. 2005). FSF also had socio- political ramifications (Head, 1989; Rose, 1998): the word ‘farming’ was deliberately used by Jones (1969) to imply that Aboriginal people actively managed the land and were not totally reliant on what nature produced (c.f. Elkin, 1954). This was a direct challenge to the notion of Terra nullius and made Aboriginal land rights more palatable for non-Aboriginal Australians.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 63

In Australia the FSF thesis also influences ecological management strategies. Prehistoric fire regimes are typically and uncritically related to human activity and described as frequent and of a low intensity across much of the continent. It has been suggested that a similar fire regime could be used for contemporary management of nature reserves (e.g. Rolls, 1981; Flannery, 1994; Ryan et al., 1995) to “return” the landscape to what it was under Aboriginal land management or as an argument for regular hazard reduction.

Gill (1977) conceded that frequent low intensity fires were applied by Aboriginal people in some Australian ecosystems but that it was not applicable across the entire continent. Clark and McLoughlin (1986) and Baker (1997) suggested that Aboriginal people variously used fire in different vegetation types depending on what resources they were extracting. Horton (1982; 2001) had also previously strenuously criticised FSF, arguing that a high fire frequency would not increase animal resources for Aboriginal people.

Clark (1981; 1983) and Horton (1982; 2001) argued that Aboriginal use of fire had a minor role in shaping Australia’s ecosystems, noting instead that vegetation zones broadly reflect climatic influences. Benson and Redpath (1997) critically evaluated the evidence for Aboriginal use of fire and any subsequent impacts on vegetation, and concluded that the evidence was misinterpreted and/or over-stated. They also believed that reintroduction of an ‘Aboriginal’ fire regime would not benefit conservation as frequent and uniform landscape firing has been proven to detrimentally affect biodiversity in Australian terrestrial ecosystems.

Mulvaney (1971: 378) challenged the fallacy of an “unchanging land and people” and those who viewed Aboriginal socio-economic and demographic changes as seemingly insignificant (e.g. Birdsell, 1953). Head (1989) also noted that there is a common assumption that Aborigines had a single ongoing impact, thereby potentially ignoring climatic change and population and cultural change.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 64

4.1.1 Postglacial Changes in Australian Aboriginal Technologies (and their Relevance to Population Changes)

In Australia it is well accepted that there have been changes in Aboriginal technologies including changes to tool assemblages and artefact densities from the mid-to-late Holocene. Rowland (1999) has summarised arguments as to whether these archaeological changes were a response to changing environmental conditions (e.g. Bowdler, 1977; Jones, 1977; Horton, 1981) or resulted from changes within Aboriginal populations (e.g. Lourandos, 1980; 1983).

Lourandos (1980; 1983) popularised the theory of ‘intensification’ and suggested that the mid-late Holocene (from ~5 ka) was a period of continent-wide changes in terms of technology, socio-demographics, settlement patterns, social structures and population densities. Lourandos (1983) described Aboriginal occupation as less intensive and more nomadic during the late Pleistocene/early Holocene and more sedentary and intensive from the mid-Holocene. From this time ‘environmental manipulation strategies’, particularly large-scale drainage systems, the use of fire, and harvesting and processing of food plants are thought to have intensified populations (Lourandos, 1983: 81).

The ‘intensification’ of Australian Aboriginal populations has historically also been associated with a technological break occurring from the mid- Holocene. After this time a range of small, well-made flaked stone tools referred to as the ‘Australian Small Tool Tradition’ (ASTT) (Mulvaney, 1975) make an appearance in the archaeological record. More recently, Hiscock and Attenbrow (1998; 2004) have established that backed artefacts, often associated with the ASTT, have been produced, albeit at low rates and in only a number of sites, from the early Holocene and did not appear suddenly.

The theory of ‘intensification’ is controversial within Australian archaeology with some especially critical about the forms of evidence and sources used to justify the theory (Rowland, 1999; Attenbrow, 2004). As an example Bird and Frankell (1991: 10), described the “cumulative directional change” in archaeological sequences used by Lourandos (1980;

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 65

1983) as a series of short-term adjustments to local conditions rather than intensification. A flaw in Lourandos’ ‘continent-wide’ intensification model is that some archaeological sequences do not show continuous unidirectional changes as should happen under intensification (Attenbrow, 2004).

Rowland (1999: 15) suggested that “internal social dynamics, external cultural influences and environmental factors” should all be considered when interpreting any archaeological record. Similarly, Attenbrow (2004: 13) identified three commonly used explanations to account for changes in archaeological records: 1. Population change (including changes in the number of people, redistribution of populations, increased intensity of site usage, and intensification/increased social complexity); 2. Behavioural change (e.g. tool manufacturing, subsistence practices); 3. Natural processes (including environmental changes or biological or geomorphological agents leading to misinterpretation).

The most detailed archaeological investigation in the Sydney Basin arguably comes from the Upper Mangrove Creek area (e.g. Attenbrow, 2004). Attenbrow (2004) interpreted changes in this archaeological record as resulting from land-use changes, triggered by climatic change. This work is significant as archaeological information, including site densities, site establishment rates and artefact accumulation rates at individual sites are often uncritically interpreted as a proxy for past human population. This is of relevance in any comparison of archaeological and palaeoecological information: archaeological changes may reflect factors independent of population change, however palaeoecologists are often most interested in some proxy of past human influences via population estimates.

4.1.2 Climate change since the LGM The last glacial maximum (LGM) in Australasia is generally dated between about 17 and 24 ka (Kershaw et al., 1991; DeDeckker, 2001; Barrows et al., 2002). Sea levels during the LGM were lower by ~135 m (Chappell and Shackleton, 1986; Yokoyama et al., 2000) resulting in Australia being about one-third larger than it is today (Markgraf et al., 1992).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 66

There is substantial evidence of dry or very shallow lakes (Bowler, 1981; Bowler and Wasson, 1984; DeDeckker et al., 1991; Nanson et al., 1992; Harrison and Dodson, 1993), expanded dunefields (Wasson et al., 1988), an increased flux of dust into the Tasman (Hesse, 1994) and marked changes in pollen records (e.g. Colhoun et al., 1982; Singh and Geissler, 1985; Kershaw, 1986; Hope, 1987; Colhoun and van der Geer 1988; Dodson and Wright, 1989; Luly, 1993; McKenzie and Kershaw, 1999; Moss and Kershaw, 1999; Hopf et al., 2000; Sweller and Martin, 2001).

The transition from the late glacial Pleistocene to the Holocene included a number of climatic oscillations that have been recorded in various palaeoenvironmental archives from both the Northern and Southern Hemispheres. This has included, most famously in the Northern Hemisphere several events including the Bölling/Alleröd inter-stadial (~14.7-12.9 ka); the Younger Dryas (YD) stadial (~12.9-11.6 ka); and the warm preboreal after ~11.6 ka (chronology based on the Greenland ice core GISP2). Research from the Southern Hemisphere has identified two overlapping cooling events: the Antarctic Cold Reversal (ACR) from ~14-12.5 ka (Jouzel et al., 2001) and the Oceanic Cold Reversal (OCR) from ~13.2 – 12.5 ka (Stenni et al., 2001). Based on palynological and geomorphological evidence from a number of sites throughout New Zealand, Turney et al. (2003: 223) described an initial cooling period during the first half of the period 14 – 11.5 ka followed by a sustained warming, with the latter event closely corresponding with the Northern Hemisphere YD event.

The Holocene period, from ~11 500 cal. yr BP has until recently been described as a stable, warm interglacial. Dodson and Mooney (2002) argued that relatively stable sea levels, atmospheric CO2 concentrations, global ice cover and the Earth’s orbital parameters within 5% of present contributed to a relatively stable period from the mid-Holocene. Others argued that the Holocene was highly variable (e.g. Maasch et al., 2003), and there is increasing evidence for abrupt changes (e.g. Bond and Lotti, 1995; deMenocal et al. 2000;). The Holocene history of El Niño-Southern Oscillation (ENSO) phenomena remains somewhat controversial (Moy et al., 2002). Kershaw et al. (2002) have identified the period of 7-5 ka as the Holocene precipitation peak (the Holocene Climatic Optimum), the period

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 67

between 4-2 ka as a drier and perhaps cooler period, and the last 2 000 years as a return to wetter conditions in south-eastern Australia.

4.2 Site Description

Gooches Crater Right (33º27’116”S, 150º16’020”E, ~960 m asl), hereafter GCR, is located on the Newnes Plateau in the Blue Mountains, on the western edge of the Sydney Basin (Figure 4.1). The site is ~150 km west of Sydney and located within the Blue Mountains National Park, which is incorporated into the Greater Blue Mountains World Heritage Area.

Figure 4.1. The location of Gooches Crater Right.

GCR is a narrow, elongate feature in the landscape in a low slope headwater valley (Figure 4.1). It is surrounded by sandstone cliffs and steep slopes, vegetated with eucalypt woodland and open heath (Benson and Keith, 1990). GCR is accumulating sandy organic sediments and is currently vegetated with a closed wet heath (dominated by Baeckea, Epacris, Gleichenia, Grevillea, Gymnoschoenus, Leptospermum). The timing of fieldwork

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 68

(2002) and subsequent visits has coincided with an extreme drought, and the watertable has been at or just below ground level: while is probable that standing vegetation may burn during a fire event, it is unlikely that any of the waterlogged sediment is lost.

The climate of the Newnes Plateau is temperate with an average minimum of -1oC in July and in the hottest month, January, an average maximum of 23.5oC (BoM, 2005). The site receives an average annual rainfall of 1047 mm, which is influenced by a mild orographic effect. In the recent historic period the Blue Mountains have been notoriously fire prone with a fire season occurring from October to February (Cunningham, 1984). The Blue Mountains are also subject to ENSO-related drought which can be associated with “high intensity wildfires during severe fire weather” may occur (BMBMC, 2000: 11). Cunningham (1984) observed that conflagration fires in the Blue Mountains during the recent historic period have been most commonly associated with El Niño-related droughts in the year following above average rainfall (and hence fuel loads) linked to La Nina events.

It has been suggested that the Newnes Plateau was a place of interaction or a transport corridor for various Aboriginal groups. Most sources suggest that the Dharug (or Daruk) and Gundungurra people were the dominant Aboriginal tribes to have occupied the Blue Mountains region (Gollan, 1987; Kohen, 1993; Horton, 1994).

Flood (1980) examined ethno-historic data to suggest a negative correlation between population size and altitude resulting in relatively low population densities in the upper Blue Mountains when compared to the coast. This is a common theme in archaeological research in the upper Blue Mountains and it is often implied that the altitude, rugged topography and limited resources meant that humans would be sensitive to climatic variations (e.g. Stockton and Holland, 1974). Stockton (1970) described human occupation in the upper Blue Mountains as ‘spasmodic’ with only seasonal hunting trips during the milder periods of the Holocene. In the eastern highlands of Australia, Bowdler (1981) described that Aboriginal occupation of any intensity only occurred after ~5 ka.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 69

Of the more than 27 sites that have been excavated from the Blue Mountains only 7 have information on artefact accumulation rates or densities (Attenbrow, 2004) (see Table 4.1). McCarthy (1964) named the early phase of the Eastern Regional Sequence as the ‘Capertian’ phase and this existed between 12 ka and 6 050 +/- 170 BP (Stockton and Holland, 1974). Following this was a hiatus, ~6 and 3.36 ka, that separated the Capertian phase with the later Bondian phase.

Site Initial habitation Artefact accumulation/densities Author

Kings Tableland 22 400 ± 1000 BP Highest at 1100 BP with a decrease Stockton and from 1000 BP Holland (1974) Springwood 8 563 ± 430 BP Highest at 100-600 BP Stockton and Creek Holland (1974) Walls Cave 12 000 ± 350 BP Steady rate over the past 4000 BP Stockton and Holland (1974) Capertee 1 7 360 ± 125 BP Decrease from 1000 BP McCarthy (1964); Capertee 3 (Capertee 3) Johnson (1979) Shaws Creek K1 No radiocarbon dates Decrease in uppermost levels Stockton (1973)

Shaws Creek K2 14 700 ± 250 BP Decrease in uppermost levels Kohen et al. (1981, 1984)

Table 4.1: Basal dates and artefact accumulation through time for published archaeological sites located in the Blue Mountains. (The information for Kings Tableland, Springwood Creek and Walls cave comes from Attenbrow (2004) and for the other sites from the cited authors.)

Europeans entered the Blue Mountains in the early 19th Century. By this time there were very few Aboriginal people living traditional lives largely due to the decimation of populations by small pox epidemics and displacement by European colonists (Breckell, 1993; Turbet, 2001). National parks, forestry and sand mining currently exist on the Newnes Plateau.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 70

4.3 Methodology

A 355 cm sediment core was extracted from GCR swamp using a Russian d-section corer (Jowsey, 1966) in June 2002. The stratigraphy of the core was described using a modified version of the Troels-Smith method (Kershaw, 1997) and was photographed. Four sections of the core (48-53, 80-90, 150-156 and 295-307 cm) were submitted for (bulk) radiocarbon dating. Radiocarbon dates were calibrated with CALIB v5 (Stuiver et al., 2005) using the IntCal04.14c (Reimer et al., 2004) and ShCal04.14c (McCormac et al., 2004) data sets.

Macroscopic charcoal, which represents local or catchment fire events (Whitlock and Millspaugh, 1996), was analysed using a modified version of the ‘Oregon sieving method’ (Long et al., 1998) and image analysis (Mooney and Black, 2003). Volumetric sub-samples were taken from contiguous 1cm sections of the core and were placed in 8% sodium hypochlorite (bleach) for 24 hrs to remove the pigment from organic matter and, hence, aid in the identification of charcoal. This was then carefully washed through a 250 µm sieve and the collected material was photographed in a petrii dish using a digital camera (Nikon Coolpix 4500). The area of charcoal was calculated using image analysis software (Scion Image Beta 4.02 for Windows).

Loss-on-ignition (LOI) analysis is commonly used to calculate the relative proportion of organic material contained in sediments. LOI was determined using sediment samples of a known volume and mass taken at 5 cm intervals through the sequence, oven-dried at 105oC for 24 hrs and combusted at 550oC for 4 hrs (based on Bengtsson and Enell, 1986). LOI was expressed as a percentage of combustible against the oven-dry mass of the sediment.

The humification of swamp sediments is used as an indication of the speed of accumulation over time with low values suggesting a rapid accumulation of organic material (Aaby and Tauber, 1975). Humification was analysed at every 10 cm following a modified version of Aaby (1986). Gravimetric oven-dried sub-samples were placed in a 0.5% NaOH solution, boiled for an hour, diluted with water and then filtered. The degree of humification was

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 71

quantified as percentage transmission in this supernatant using an EEI Colorimeter (using a 540 nm filter).

Samples for geochemical analysis were taken at 15 cm increments along the core. Gravimetric samples were combusted at 550oC, acid digested (using hydrochloric and nitric acids in a microwave oven) and after cooling, the solutions were analysed using an Ionically Coupled Plasma-Atomic Emission Spectrophotometer (ICP-AES). The geochemical analysis was undertaken using Ionically Coupled Plasma-Atomic Emission Spectrophotometer (ICP-AES). Four calibration standard solutions were prepared for concentrations of 0, 1, 10 and 100ppm (mg/L), which were then used to calibrate the background. Four blank solutions were also analysed, then averaged to quantify the background concentration error of the acid digest solution. Results were originally given in milligrams per litre (mg/L), or (ppm), and were then converted to milligrams per gram of oven dried sediment (mg/gOD). Iron, magnesium, manganese, calcium, aluminium, sodium, potassium, sulfur and phosphorous were analysed to better understand the depositional nature of the sediment.

4.4 Results

4.4.1 Core stratigraphy and chronology The GCR sediment core was composed of humified peat interspersed with clay, charcoal and sand. The core description identified very high levels of sand and charcoal between 104 and 150 cm. The 14C dating of the deposit (Table 4.2) implies a relatively constant rate of accumulation (~0.025 cm.yr-1) in the analysed core. A linear depth-age relationship (y = 41.224x – 471.74, r2 = 0.9936. where x= depth in cm, y= age in cal. yr BP) is used for all age calculations. Based on this relationship the analysed core (355 cm) represents ~14 200 cal. yr BP. Pollen analysis revealed the first appearance of the exotic taxon Pinus at 15 cm.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 72

Table 4.2. Radiocarbon dates and calibration for GCR sediments. Calibration results from CALIB v5 (Stuiver and Reimer, 2004). The mid-point of the entire calibrated year range is used in age-depth model calculations.

Sample depth (cm) 14C date BP with Cal. yrs BP* (2 σ error) Lab code 1σ error 48-53 1 760 ± 60 1419 – 1466 (4.3%) β-169992 1491 – 1497 (0.4%) 1509 – 1742 (91.4%) 1753 – 1785 (2.4%) 1790 – 1811 (1.4%) 80-90 2 450 ± 60 2333 – 2619 (82.2%) β-192605 2632 – 2708 (17.8%) 150-156 4 950 ± 130 5322 – 5418 (8.9%) β-169993 5440 – 5912 (91.1%) 295-307 10 360 ± 140 11646 – 11667 (0.5%) β-169994 11703 – 12737 (99.5%) BP = before AD 1950

4.4.2 Macroscopic charcoal The area of macroscopic charcoal (Figure 4.2) is relatively high between 0–6 cm (the late European period), 67-87 cm (~2 300-3 100 cal. yr BP), 97–150 cm (~3 500-5 700 cal. yr BP), 232-244 cm (~9 100-9 600 cal. yr BP) and 250-281 cm (~9 800-11 100 cal. yr BP). There are low levels of charcoal between 325-353 cm (~12 900-14 000 cal. yr BP), 287- 315 cm (~11 300-12 500 cal. yr BP), 150-232 cm (~5 700-9 000 cal. yr BP), 87-97 cm (~3 100-3 500 cal. yr BP), and from 40 to 25 cm (~1 100-500 cal. yr BP) and 6-13 cm (early European occupation). Given the near linear rate of sedimentation in the analysed core, the concentration of charcoal is a good reflection of charcoal accumulation.

The charcoal concentration/accumulation at GCR reveals three distinctly different periods: 232-355 cm (coinciding with the late Pleistocene-Holocene transition); 150-232 cm (coinciding with the early-to-mid-Holocene); and 0-150 cm (coinciding with the mid-to- late-Holocene). The most abrupt change in charcoal occurs at 150 cm (~5 700 cal. yr BP).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 73

4.4.3 Humification, loss of ignition and geochemistry Transmission levels give an indication of the degree of humification in the sediments with low values representing highly humified sediments. Throughout the sequence transmission levels averaged ~34% (Figure 4.2). There were slightly lower transmission levels (i.e.

m) μ n tio al (>250 a -on-igntion arco s umific Ch H Los 0

1 620 ± 80 BP 50

2 520 ± 170 BP 100

5 620 ± 200 BP 150

200 Depth (cm)

250

12 190 ± 300 BP 300

350 250 500 750 1000 20 40 60 80 10 20 30 40 50 60 70 Concentration (mm2/ cm3) Transmission (%) Organic matter (%) Figure 4.2. The results of the charcoal analyses, humification and loss-on-ignition at Gooches Crater Right. reduced humification) between 280 and 350 cm (~11 000 and 14 200 cal. yr BP) and highly elevated levels between 80 and 140 cm (~2 800 and 5 300 cal. yr BP) (with the exception of a drop at 100 cm), and at 0 to 10 cm (during the time since European settlement).

Loss-on-ignition averaged 37% throughout the sequence with a maximum value of 73% on the surface and minimum of 14% at 115 cm (~4 300 cal. yr BP) associated with a high sand content (Figure 2). Between 355 (~14 200 cal. yr BP) and 140 cm (~5 300 cal. yr BP) LOI values remain relatively constant with the exception of peaks at 140 (~5 300 cal. yr BP),

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 74

Figure 4.3. The results of the geochemical analysis at Gooches Crater Right (all concentrations in mg/L)

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 75

165 (~6 300 cal. yr BP), 270 (~10 600 cal. yr BP) and 295 cm (~11 700 cal. yr BP) associated with woody detritus. From 140 cm (~5 300 cal. yr BP) LOI values become more variable with troughs at 100-140 cm (~3 600-5 300 cal. yr BP) and 80-100 cm (~2 800- 3 600 cal. yr BP). There are steadily increasing values from 55 cm to the surface (~1 800 cal. yr BP to the present).

The results of the ICP-AES geochemical analyses are presented in Figure 4.3. Aluminium, barium, iron, potassium, lithium, magnesium, phosphorus and sulphur levels revealed very similar trends throughout the sequence with generally stable levels from 355-150 cm (~14 200-5 700 cal. yr BP), a significant decrease between 150 and 90 cm (~5 700 – 3 200 cal. yr BP) and increased variability in the top 90 cm (from ~3 200 cal. yr BP). Calcium, cobalt, manganese and sodium also revealed similar patterns with mostly low levels throughout the core with the exception of increases in the uppermost sediments

4.5. Discussion

Macroscopic charcoal persisted throughout the analysed GCR sequence although the abundance fluctuated greatly (Figure 4.2). It can therefore be inferred that fire has persisted in this environment for the last ~14.2 ka although the intensity and/or frequency must have varied. It is suggested that consistently low accumulation rates of macroscopic charcoal represent periods when intense fires were largely absent from the landscape/catchment since it is assumed that this would result in the deposition of macroscopic charcoal. During these periods of low charcoal accumulation low intensity fires may still be a feature of the landscape as burnt/unburnt patchiness could inhibit the delivery of charcoal to the sediment sequence.

Intervals of sustained high macroscopic charcoal are interpreted as periods when intense or frequent fires were common in the landscape. This interpretation is collaborated by the charcoal concentration in the recent European period when large, intense fires have been a

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 76

feature of the environment. Given that the charcoal of the last 5.5 ka is similar to this recent period, and that the vegetation has been relatively consistent (Black and Mooney, in press), it can be concluded that intense fires have been a feature of this environment over this period.

Black and Mooney (in press) concluded that climate was the dominant control on fire activity at GCR using a non-contiguous analysis of charcoal. Figure 4.4 presents a new contiguous charcoal record from Gooches Crater Right re-expressed against time. This diagram suggests that fire activity at this site can be characterised within three time intervals: the late Pleistocene-early Holocene period, the early-to-mid Holocene and the mid-to-late Holocene.

Figure 4.4. The charcoal concentration expressed against time, highlighting the difference between the Late Pleistocene, early-to-mid and mid-to-late Holocene.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 77

In Figure 4.5 the new contiguous GCR charcoal record, smoothed using a 41 point running average, is compared with Haberle et al.’s (2001a) regional cumulative charcoal curve (corrected) constructed by summing the 200 year values for each of ten sites throughout Indonesia and Papua New Guinea against age. Haberle et al. (2001a) described a general synchronicity between their cumulative charcoal curve and regional climate proxies, noting that charcoal increased during times of climatic instability such as during the glacial transition (~17-9 ka) and from the mid-Holocene when ENSO events became stronger (~5 ka to present). The strong association between the GCR and the composite tropical record of Haberle et al. (2001) strongly supports the hypothesis that climate had a profound influence on fire at GCR.

300 400 GCR charcoal curve* Haberle et al. 2001 curve

350 250

300 ) 3

/cm 200 2 250

150 200

150 . (2001) corrected cumulative charcoal 100 et al Gooches charcoal (mm Crater 100 Haberle Haberle 50 50

0 0 250 1500 2750 4000 5250 6500 7750 9000 10250 11500 Age (cal. yr BP)

Figure 4.5: A comparison of the composite charcoal results from several sites in tropical Sahul, from Haberle et al. (2001), and the results from Gooches Crater Right.

Haberle and Ledru (2001) found a degree of consistency in comparing charcoal records from Indonesia and PNG with Central and South America. The obvious differences in

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 78

human populations and cultures between the Americas, Malesia and Australia and the similarities in the charcoal records makes it more likely that some climatic signal is influencing fire activity. Haberle and Ledru (2001: 97) suggested that this broad correlation “demonstrates that fire is promoted during periods of rapid climate change and high climate variability, regardless of the presence or absence of people”. The GCR record supports this statement. Haberle and Ledru (2001) also noted that the strongest correlation between the charcoal records of Indonesia, PNG, Central and South America postdates 5 000 yrs BP when El Niño-related variability intensified.

4.5.1 The late Pleistocene-early Holocene (~14.2 to 9.1 ka) At GCR the charcoal record during the late Pleistocene-early Holocene is highly variable (Figure 4.5). This variability includes four major peaks centred on ~14 000, ~12 660, 10 640 and ~9 400 cal. yr BP and periods of low charcoal accumulation ~14 000-12 800, ~12 500-11 300 and ~10 100 to 9 600 cal. yr BP. These trends are reminiscent of palaeoclimatic records of the late glacial-Holocene transition, however caution is warranted given that calibration of radiocarbon dates during this period results in relatively large ranges (of ~1 200 years) and that there is contradictory evidence of a ‘southern’ YD event (e.g. Andres et al., 2003; Turney et al., 2003).

At GCR it hence appears that fire activity increases, probably associated with the occurrence of intense fires, during periods of climatic instability. In Indonesian and New Guinea Haberle et al. (2001) also found a reversal of high charcoal values coeval with the YD and suggested that a relatively cool phase may have altered soil moisture and the vulnerability of the vegetation to fire. The pollen record from GCR (Black and Mooney, in press) registered no significant changes during the late glacial-Holocene transition. Clark (1983) has previously described an inability of palynology to detect subtle changes in fire- adapted vegetation. In this regard the substitution of species within genera may make any subtle vegetation change invisible in the pollen record.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 79

The degree of humification at GCR was relatively low between ~14 200 and 11 100 cal. yr BP and slightly increases from ~11 100 cal. yr BP, associated with a brief period of increased charcoal content. Kershaw (1995) suggested that the Pleistocene-Holocene boundary (~11.5 ka) was a time when peat started forming in peatlands in southeastern Australia. Associated with peat formation is an increased degree of humification in sediments. The increase in the degree of humification at GCR from the early Holocene may be reflecting the change from organic clays to peat.

The earliest evidence of human occupation in the Blue Mountains is dated to ~22.4 ka (Stockton and Holland, 1974). Although often discounted (e.g. Johnson, 1979; Bowdler, 1981; Lennon, 1983) this date is based on an artefact in a King’s Tableland sequence found below a ~14.5 ka 14C-dated layer, where 8 artefacts were found and just below an undated layer with 10 artefacts.

Given this likelihood, and the often-stated assumption that the climate of the higher altitudes of the Blue Mountains limited occupation, another possible scenario is that the charcoal variability at GCR between ~14 200 to 9 100 cal. yr BP may reflect the comings and goings of people to the region as climatic variability made it more or less favourable for human occupation. Attenbrow (2002; 2004) has particularly highlighted the importance of the ~11-10 ka period in the archaeological records of the Sydney Basin, arguing that topography and postglacial sea level changes played an important role. Hence human activity may explain the sustained increase in charcoal from ~11.1 to 9.78 ka at GCR, however, such an explanation does not fit well with the drastically reduced fire activity during the early Holocene when environmental conditions were also presumably favourable.

4.5.2 The early-to-mid-Holocene (9.1 to 5.7 ka) Harrison and Dodson (1993: 282) identified the transition to wetter climates at 10- 9ka in coastal south-eastern Australia and attributed this to the “equatorward advance of the southern margins of the subtropical high-pressure belt and the westerlies”. A period of maximum effective precipitation has been identified in southern Australia between about

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 80

7 000 and 5 000 yr BP (Bowler, 1981; Shulmeister, 1999) with McGlone et al. (1996) suggesting that average annual temperatures were 0.5-2oC warmer and precipitation was up to 50% higher during this period. Between 8 and 5 ka southeastern Australia also witnessed the expansion of wet sclerophyll and rainforest taxa (e.g. MacPhail and Hope, 1984; Dodson et al., 1986). These conditions mirror more widespread climatic trends (e.g. COHMAP, 1988) including the early Holocene ‘hypsithermal’ or ‘Holocene Climatic Optimum’ followed by the deterioration of climate and the re-advance of most alpine glaciers during the last 5 000 years (the ‘neoglacial’).

In tropical conditions Haberle et al. (2001a) attributed low fire activity during the early Holocene to climatic stability and reduced seasonality. Similarly, in a review of humid Australian landscapes Kershaw et al. (2002) identified a slight reduction in fire between 7 and 5 ka and attributed this to an increase in rainfall and reduction in seasonality. The relatively stable climate, reduced seasonality and weakened ENSO (Moy et al., 2002) during this period may have resulted in the suppression of fire activity.

At GCR charcoal accumulation is low and extremely consistent in the early-to-mid Holocene (Figure 4.5), suggesting low fire activity. Chen (1986) has suggested, however, that a low and consistent charcoal record may indicate a regime of high frequency but low intensity fires: a scenario that is consistent with often-presumed Aboriginal use of fire. The warmer, wetter and more stable conditions associated with the Holocene Climatic Optimum may have been more favourable for human occupation in the upper Blue Mountains.

4.5.3 The mid-to-late-Holocene (from 5.7 ka) The charcoal curve at GCR from the mid-Holocene to the present is highly variable and charcoal concentrations are much higher than the preceding analysed record. The abundance of charcoal increases abruptly at 5.7 ka and remains high until ~3.5 ka. After a decrease in charcoal at 3.5 ka concentrations increase from 3.1 and remain high until 2.3 ka. Another abrupt increase above 6cm is thought to reflect recent conflagrations of the later part of European occupation. From 5.7 ka intense fires were likely to be a persistent feature of this environment.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 81

It is possible that the increase in charcoal (and variability) since the mid-Holocene is related to changes associated with the dynamics of the swamp. It is plausible that as the sediment has built up through time the aerial vegetation on the swamp surface has burnt more frequently. This scenario, while plausible, is not supported by the palynology at the site (Black and Mooney, in press) which demonstrates little vegetation change during the Holocene and thereby does not support the development of the wet-heath vegetation from this time. Furthermore, the sediment at the GCR site is accumulating as a result of a rock- fall dammed constriction, where it is likely that the water-table has risen concurrently with the accumulation of sediment.

Kershaw et al. (2002) noted a significant increase in fire from the mid-Holocene in all Australian vegetation types, excluding wet forests, and this was maintained until ~2 ka. Although Kershaw et al. (2002) considered the influence of people to explain this increase in fire activity they thought it was better explained by decreased precipitation and the onset of ENSO in the region. This is supported by approximately coeval increases in fire activity in New Zealand, which is similarly influenced by ENSO, but was not colonised by people at that time (Kershaw et al., 2002).

The mid-Holocene has been identified as a period of climate change in Australasia (Bowler et al., 1977; Shulmeister, 1999) and further afield (Rodbell et al., 1999; deMenocal et al., 2000; Sandweiss et al., 1999). Steig (1999: 1485) described the mid-Holocene as “a period of particularly profound change” and Hodell et al. (2001) noted an abrupt cooling of sea- surface temperatures, expansion of sea ice and increased ice-rafted detritus accumulation in the Southern Ocean, between 5.5 to 5ka. Morrill (2003) examined 109 previously published palaeoclimatic records to assess evidence of abrupt climate change during the mid- Holocene and identified two periods of abrupt climate change occurring at 5.5-5.8 ka and 4- 4.8 ka.

In Australasia Shulmeister (1999) described a decoupling of the northern (tropical) and southern (temperate) climate systems of Australia at ~5 000 yr BP resulting in increased westerlies, the loss of summer monsoon rainfall and a sharp decline in effective

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 82

precipitation in. southern Australia. Increased seasonality in the south-western Pacific during the mid-to-late Holocene appears to be related to El Niño -Southern Oscillation (ENSO) phenomena. Rodbell et al. (1999) argued that ENSO progressively achieved modern characteristics by ~5 000 cal. yr BP although Moy et al. (2002) placed this earlier, at ~7 000 cal. yr BP. Sandweiss et al. (2001) also described ENSO events from ~5 800 yr BP, albeit at a low frequency until about 3 200 yr BP. Riedinger et al. (2002) also described an increase in the intensity and frequency of El Niño events from 3 100 cal. yr BP and Clement et al. (2000) described an increase in ENSO events during the 3-1 ka period. The GCR charcoal record shows an initial increase in charcoal from ~5.7 ka and another sustained increase from ~3.1 ka that are roughly coeval with the pattern of ENSO described above.

Haberle et al. (2001a) found an abrupt increase in fire activity in the uplands of New Guinea from ~6ka, with a peak between 4.5 and 1 ka and attributed this to climatic variability. Chalson (1991) examined the palynology of eight swamps in the Blue Mountains region and identified 6.5 to 5.5 ka as having a particularly variable climate. Although of a low resolution and not chronologically well-constrained, Chalson (1991) also found a general increase in charcoal from the mid-Holocene. Similarly, at a coastal site near Sydney, Martin (1994) found an increase in charcoal at 5.5 ka, which was attributed to the use of fire by Aboriginal people.

The proposed mid-Holocene shift in climate and particularly the onset of modern ENSO character appear to have greatly influenced fire regimes at GCR. Charcoal from the mid Holocene suggests a more frequent and probably more intense fire regime which makes sense in an ENSO-dominated climate, which allows the accumulation of biomass during the wetter La Niña events and the drying and subsequent burning of this biomass during El Niño events. Similarly changes in humification and loss-on-ignition are best related to post- fire erosion events and particularly the mobilisation of sand. Aluminium, barium, iron, potassium, lithium, magnesium, phosphorus and sulphur levels decrease significantly during this period of increased charcoal.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 83

The abrupt increase in charcoal at GCR from the mid-Holocene could also potentially be related to the ‘intensification’ of Aboriginal populations at this time (Lourandos, 1980: 1983). This scenario is not supported by the archaeological record of the Blue Mountains, however, as the increase in charcoal pre-dates any evidence of significant changes by at least 1 000 yrs. The timing of any increase in Aboriginal populations at this time is also at odds with Stockton and Holland (1974) who suggested a hiatus in the Blue Mountain’s archaeological record between 6 000 and 3 360 yr BP.

Another plausible scenario is that the mid-Holocene increase in charcoal at GCR resulted from increases in fuel in the absence of regular low intensity anthropogenic fire. To test this hypothesis the archaeological record of the Capertee 3 site, which is the closest (~35 km away) and best resolved archaeological site to GCR is compared to the charcoal data. Figure 6 shows the discard rates of backed and non-backed artefacts at Capertee 3 (Hiscock and Attenbrow, 2004) versus the average charcoal concentration at GCR for each corresponding time period. Discard rates and average charcoal concentrations were both relatively low between ~9.7 and 6 ka however average charcoal concentrations were high between ~6 and 3.6 ka whilst discard rates remained low. Discard rates were highest between ~3.6 and 1.7 ka with the layer ~3-2.3 ka having the highest discard rates. The latter layer corresponds with the very high accumulation of charcoal. The period ~1.7 ka to present has very low artefact discard rates corresponding with relatively high levels of charcoal accumulation. This pattern suggests that climate and Aboriginal people’s influence on fire activity at GCR may have varied throughout time. For example during the period ~6-3.6 ka, with very high average charcoal but low artefact discard rates, the onset of an ENSO dominated climate may have been responsible for the high fire activity whereas between ~3 and 2.3 ka, where both artefact discard rates and charcoal accumulation rates are very high, the fire activity that existed during this time may reflect a combination of a climate influence by ENSO and the management of the land by Aboriginal people. However, these scenarios assume that the archaeological information reflects human activity at the broader landscape scale, which is controversial (e.g. Attenbrow, 2004).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 84

450 Artefact discard rate Average charcoal concentration 400

350

300

250 age charcoal concentration (area)

200

150

100

50 artefact discard rate (#/1000yr) / aver

0 0-1700 1700-2300 2300-3000 3000-3600 3600-4800 4800-6000 6000-7200 7200-9700 Time periods (yrs BP)

Figure 4.6. Artefact discard rates (Capertee 3 site) versus average charcoal concentration (GCR) throughout the Holocene (Source: Archaeological data adapted from Table 94 in Hiscock and Attenbrow, 2004)

4.6 Conclusions

A climatic solution can be used to explain all periods of change in the charcoal record: climatic variability associated with deglaciation causing variable charcoal; climatic stability during the early-mid Holocene resulting in low levels of charcoal; and high and variable levels of charcoal associated with ENSO-like climatic variability from the mid-Holocene. Prolonged periods with low periods of charcoal may represent times when intense fires were not a feature and during times of high macroscopic charcoal output intense fires were responsible.

The similar trends with Haberle et al.’s (2001a) cumulative charcoal curve, from PNG and Indonesia, suggests that perhaps fire activity is responding to a regional climatic signal.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 4 85

Further investigation into this will hopefully help to resolve this. The most significant change in the record occurs from the mid-Holocene (5-6 ka) with an abrupt increase in charcoal, as well as changes in loss-on-ignition, Mn:Fe ratio and humification levels. The fact that all proxies are changing at this time suggests climatic change. The mid-Holocene is becoming recognised as a period of abrupt and substantial climatic change (e.g. Rodbell et al., 1999; Sandweiss et al., 1999; Shulmeister et al., 1999; Steig, 1999; deMenocal et al., 2000; Hodell et al., 2001; Morrill, 2003).

The archaeological record can be inferred to explain some changes in the charcoal record however inconsistencies exist. For example a hiatus in the archaeological record in the Blue Mountains correlates with very high levels of charcoal at GCR but maximum artefact accumulation rates at Capertee 3 also correspond with very high levels of charcoal. Although there were apparently changes in the human populations it did not seem to affect fire activity of GCR.

It appears that frequent Aboriginal use of fire to mitigate intense fires, as suggested by Flannery (1994), was not the case at GCR. It appears that intense fires have persisted at GCR since the mid-Holocene and this is most likely associated with the establishment of modern ENSO rather than any change with Aboriginal populations. The unsupported and simplistic notions of pre-European fire frequency are challenged by the fire history from Gooches Crater Right, the second longest contiguous record in Australia to date (Haberle’s in press Lake Euramoo record spans to ~20 ka). The record indicates that fire regimes are dynamic and change temporally as well as spatially.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 86

Chapter 5: A >43 000 year vegetation and fire history from Lake Baraba, New South Wales, Australia

Abstract Continuous palaeoenvironmental sequences that describe the vegetation of the last glacial maximum (LGM) and of the subsequent climatic amelioration are relatively rare in the Australian, Southeast Asian and Pacific region (SEAPAC region). The results of a palynological investigation from Lake Baraba, located in eastern Australia, which extends beyond 43 ka are presented in this study. Bands of oxidised sediment prior to the LGM suggest lake level fluctuations, however lacustrine clays continued to be deposited throughout the LGM and into the early Holocene when the deposition of peat was initiated. The vegetation, a Casuarina woodland/shrubland with a mixed understorey, remained relatively stable from >43 ka to the early Holocene, suggesting that this sclerophyllous vegetation was resilient to changes in climate. The vegetation of the LGM at Lake Baraba does not conform to previous descriptions of a treeless south-eastern Australia, and it is possible that it was a refugium for woodland. Myrtaceae expanded at the expense of Casuarinaceae from the early Holocene, with charcoal analyses suggesting that fire was an unlikely explanation. There was no apparent relationship between Aboriginal site usage and fire activity although low levels of charcoal in the late Holocene corresponded with high use of Aboriginal sites. How Aboriginal people used fire at Lake Baraba remains speculative.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 87

5.1. Introduction

This paper describes a > 43 000 yr pollen and charcoal record from Lake Baraba, New South Wales, to investigate the nature of vegetation and fire in the landscape over a period which includes the last glacial maximum (LGM) and subsequent climatic amelioration. This palaeoecological record, compared and contrasted to existing studies from southeastern Australia, is used to better understand the history of our extant vegetation and to provide a longer temporal perspective for the discussion of fire management.

The climate of the LGM in Australia, ~18 – 21 ka, was inhospitable with cooler, drier and windier conditions (Markgraf et al., 1992; Allan and Lindsay, 1998). For most regions it is estimated that precipitation levels were about half present day values in the mid-latitudes (Allan and Lindsay, 1998) and mean annual temperatures were up to 10oC cooler in the southeastern Australia (Kershaw, 1995). Sea levels during the LGM were ~120-135 m lower than current levels (Yokoyama et al., 2001; Lambeck and Chappell, 2001) resulting in Australia being about one-third larger than it is today (Markgraf et al., 1992).

There are only 33 sites in the Australian, Southeast Asian and Pacific region (SEAPAC region) that have a record of vegetation during the LGM (Dodson, 1994; Kershaw, 1995, Pickett et al., 2004). At the peak of the LGM, about 20 ka, previous studies indicate the vegetation of southern Australia was a semi-arid grassland-steppe, dominated by Asteraceae and Poaceae and with small patches of mesic communities (Dodson, 1994; Hope, 1994; Kershaw, 1995: 1998). More recently Pickett et al. (2004) have suggested xerophytic shrubs/woodlands characterised the LGM rather than steppe vegetation. The exposed continental shelf of southeastern Australia at the LGM, however, was covered in shrub, heath and woodland communities dominated by myrtaceous shrubs, Asteraceae and Chenopodiaceae (Harle, 1997).

Palynological studies suggest a relatively rapid climatic amelioration following the LGM with increases in arboreal taxa (e.g. Eucalyptus, Casuarina). The Pleistocene-Holocene

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 88

boundary (~11 ka) has been described as a period of rapid climate change with temperatures and moisture regimes beginning to reach modern values (Kershaw, 1995). The re-expansion of vegetation communities dominated by trees also dates to about this time. Changes to the vegetation have been relatively minor throughout the Holocene, but this perceived stability may be related to palynological invisibility of subtle changes within families (e.g. Myrtaceae, Casuarinaceae) (Clark, 1983). A number of records in south- eastern Australia show an increase in Eucalytpus at the expense of Casuarina starting from various times during the Holocene (e.g. Ladd et al., 1992; Devoy et al., 1994; Harle, 1998; Gale and Pisanu, 2001). The cause of this has been attributed variously, but most often to the role of fire.

The period 7 – 5 ka has been described as the precipitation peak, or Holocene Climatic Optimum in Australia and the period 4 – 2 ka was perhaps cooler and drier (Kershaw et al., 2002). Lees (1992) linked the changes to enhanced climatic variability from 5.5 ka onward, suggesting this may be associated with sea level stabilisation and consequently, ENSO fluctuations. More recently Gagan et al. (2004) suggested that the onset of modern ENSO periodicities, identified by palaeo-records from the tropical Pacific, began ~ 5 ka with an abrupt strengthening in magnitude from ~3 ka.

The mid-late Holocene is also believed to have been an important period of change for humans in south-eastern Australia with technological changes in the archaeological record perhaps associated with ‘intensification’ (e.g. Lourandos, 1980: 1983). The ‘intensification’ of Australian Aboriginal populations during the mid-late Holocene, however, is controversial (e.g. Rowland, 1999) and may involve increased archaeological visibility. Often-quoted archetypes of technological ‘breaks’ have recently been demonstrated to be less abrupt ‘introductions’ (Hiscock and Attenbrow, 1998) and rising sea levels and the subsequent loss of the continental shelf for living space must have had an impact on the Aboriginal people (Attenbrow, 2004). The ecological impact of the past use of fire by Aboriginal people is controversial (e.g. Bowman, 1998).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 89

5.2. The Environment

Lake Baraba is one of the Thirlmere Lakes (34o13’S 150o13’E), which are an upland fluviatile system contained in an entrenched meander (Timms, 1992) at an altitude of 305 m asl (Figure 5.1). Tectonic activity beheaded a river that probably originally flowed westwards, leaving the isolated, sinuous channel that now contains the lakes (Timms, 1992). The valley system and the lakes may potentially be up to 15 million yrs in age which is unusually old for such small lakes (Horsfall et al., 1988).

Figure 5.1. Location of Lake Baraba.

The flora within TLNP is typical of the Sydney Sandstone complex and is particularly diverse with over 400 species from 250 genera represented (Benson and Howell, 1994; NPWS, 1995). It is predominantly dry sclerophyll woodland and forest formations of mixed Eucalyptus/Corymbia/Casuarina. Sydney Peppermint (Eucalyptus piperita) and Red Bloodwood (Corymbia gummifera) are common. Lake Baraba is largely infilled and is mostly covered by swamp species dominated by sedges. Species found on the swamp surface include Pithy Sword-sedge (Lepidosperma longitudinale), Scale-rush (Lepyrodia muelleri) and Zig-zag Bog-rush (Schoenus brevifolius) (Fairley, 1978). Aquatic vegetation of the site includes Tall Sedge (Lepironia articulata), Tall Spike-rush (Eleocharis

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 90

sphacelata), Woolly Frogmouth (Philydrum lanuginosum), Water Lily (Brasenia schrebi) and the algae Chara fibrosa. The paperbark tree, Maleleuca linariifolia, also grows on the swampy edges of the site.

Thirlmere Lakes NP experiences a warm temperate climate with average temperatures ranging from 25-33oC in summer and 5-15oC in winter. The average annual rainfall is 804 mm (BoM, 2005). Fire is a regular event in dry sclerophyllous vegetation and during the 20th Century at Thirlmere Lakes there has been an average of one conflagration every 7.5 years (Noakes, 1998).

The traditional custodians of the Thirlmere Lakes region are the D’harawal and Gundangarra people (NPWS, 1995). The lakes and wetlands of the Thirlmere Lakes were likely to represent a plentiful supply of food and ethnographic evidence suggests that the Aboriginal people of the region frequently applied fire to the landscape (NPWS, 1995). Evidence for the earliest occupation of the Sydney Basin is from Wentworth Falls, which has been dated at ~22 000 years BP (Stockton and Holland, 1974). Attenbrow (2004) has suggested that the establishment of Aboriginal sites increased from 8 ka with the habitation rates of these sites generally increasing until the arrival of European people. The Gundangurra and D’harawal populations declined in the 1800s largely due to conflict with and diseases introduced by European people.

5.3. Methodology

A 6.35 m sediment core was extracted from Lake Baraba using a Russian d-section corer (Jowsey, 1966) in June 2003. The stratigraphy of the core was described using a modified version of the Troels-Smith method (Kershaw, 1997) and was photographed. Five sub- samples of the core (147-153, 275-285, 347-353, 464-472 and 595-601 cm) were submitted for radiocarbon dating. Radiocarbon ages were calibrated with CALIB v5 (Stuiver et al.,

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 91

2005) using the IntCal04.14c (Reimer et al., 2004) and ShCal04.14c (McCormac et al., 2004) data sets.

Macroscopic charcoal, which is thought to represent local or catchment fire events (Whitlock and Millspaugh, 1996), was analysed using a modified version of the ‘Oregon sieving method’ (Long et al., 1998) and image analysis (Mooney and Black, 2003). Volumetric sub-samples were taken from contiguous 2.5 cm sections from the core and dispersed for 24 hr in 8 % sodium hypochlorite (bleach) to remove the pigment from organic matter and hence aid in the identification of charcoal. This material was washed through a 250 μm sieve and the collected material was photographed in a petrii dish using a digital camera (Nikon Coolpix 4500). The area of charcoal was calculated using image analysis software (Scion Image Beta 4.02 for Windows).

Pollen samples were prepared using standard palynological techniques (Faegri and Iverson, 1975). Volumetric samples were taken every 10 cm along the core and at every 5 cm between 600 and 635 cm, and exotic pollen (Alnus) was added as a spike. The samples were deflocculated with hot 10 % NaOH and then sieved through a 150 μm mesh. Silicates were removed using heavy liquid (i.e. ZnBr2(aq.)) separation and organic matter with acetolysis. Samples were mounted in silicon oil and the palynomorphs were counted at 400x magnification until 200 grains were identified. The pollen counts were expressed as percentages, with all grains contributing to the pollen sum. If the pollen count was very low (< 50 grains) they were omitted from the pollen diagram. The pollen data was stratigraphically grouped into different zones using the CONISS cluster analysis feature of Tilia Graph (Grimm, 1992). Significant changes in other analysed parameters were also considered when formulating the zones. The interpretation of the pollen diagram relies on the identification, representation and source habitat for each palynomorph, as described in Table 5.1.

Loss-on-ignition (LOI) analysis was used to calculate the proportion of organic material contained in sediments (Bengtsson and Enell, 1986). Sediment samples of a known volume

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 92

Table 5.1. The pollen and spores quantified in the Lake Baraba sediment, and their indicative value. Pollen or Spore Indicative Value Myrtaceae* • Predominantly shrubs and trees of natural vegetation communities but includes species used for wind-breaks. • Myrtaceae pollen thought to be well or over-represented. • Includes Eucalyptus, Corymbia and some Melaleuca of which some species are indicative of wet swampy conditions. Casuarinaceae • Indicates natural vegetation both shrubs and trees in a wide range of habitats. • Casuarinaceae pollen generally very over-represented (e.g. in coastal heaths if Casuarinaceae pollen represented >20% of the total count it can be assumed that it was growing locally whereas counts of <10% suggest that it was locally rare or perhaps coming from a regional source). • Includes Casuarina spp. and Allocasuarina spp.. Banksia • Common as an understorey in pre-disturbed woodland communities and also in heath formations and probably indicative of drier vegetation communities although it includes some swamp growing representatives. • Extra-local source and under-represented. Acacia • Generally indicative of drier native woodland communities but some members indicative of moist forests (e.g. A. melanoxylon). Thick stands of young Acacia may be indicative of recent disturbance, including fire. • Very poor dispersal characteristics and under-represented. Chenopodiaceae • A generally high salt tolerance means that the family is probably indicative of salt marshes around edge of the lake. • It is a well to over-represented pollen type. Poaceae • Often indicative of open and drier vegetation communities. Introduced pasture also represented by the family. • Regional pollen source that is well-to-over represented. • Includes many species of grasses. Asteraceae • Herb and shrub species of more open vegetation communities. • The family is a component of regional pollen rain and is over- represented in pollen diagrams. Pinus • Mostly P. radiata (Monterey Pine), an introduced tree indicating post-European sediments. • A well to over-represented pollen type. Cyperaceae • Sedges indicative of wet and swampy areas. • Considered to be well to over-represented in pollen diagrams. Restionaceae • Herbaceous species of a variety of habitats, however generally swamp taxa. • Restionaceae are often under-represented in pollen diagrams. Pteridium spp. • Wide range of habitats, usually well-drained. Also commonly associated with disturbance, either part of post-fire or old-field successionary sequence. • Well-to-over represented (Dodson, 1983). Other trilete spores and • Generally understorey species indicative of wetter patches in the landscape. other ferns • These spores are usually well-represented (Ladd, 1979). Haloragaceae • Either a herb of wetter areas (e.g. Haloragis spp.) or freshwater aquatic environment (Myriophyllum spp.) that is well-represented in pollen diagrams. Sources: Dispersal and Representation: Dodson (1983), Ladd (1979). Indicator value: Kodela and Dodson (1988). * Myrtaceae pollen was not identified to the component genera since separation is not reliable. were taken at 5 cm intervals through the sequence and their mass determined. After drying at 105oC for 24 hr, the samples were combusted at 550oC for 4 hr to estimate organic content. LOI was expressed as a percentage (of oven dried mass) that represents the proportion of organic (or combustible) matter in the sample.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 93

5.4. Results

5.4.1. Core stratigraphy and chronology Peat was found above 172 cm, a transition layer of peat and clay from 172 – 410 cm, which became more clayey with depth, and clay below 410 cm. Between 410 and 464 cm the dark organic clays became increasingly yellow and orange with depth. At 464 cm the clay changed abruptly to dark organic clay. Beyond this depth the sediments changed abruptly between dark organic and light yellow/orange bands.

LOI values were found to be very low, averaging ~8 %, between 425 and 635 cm. The values increase quite abruptly between 390 and 425 cm after which they remain very high (~90 %), with the exception of a drop at 360 cm. LOI values decrease to an average of ~67 % between 95 and 130 cm and between 0 and 95 cm they become variable with values ranging between 17 and 100%.

The results of the 14C dating of the deposit suggest that the base of the analysed profile is greater than 43 000 yr BP (Table 5.2). Pollen analysis revealed the first appearance of the exotic taxon Pinus, representing European occupation, at 10 cm.

Table 5.2. Radiocarbon dates and calibration for Lake Baraba sediments. Calibration results from CALIB v5 (Stuiver et al., 2005). The mid-points of calibrated year ranges are used in age-depth model calculations.

Sample depth (cm) 14C date BP Cal. yrs BP* (2 σ Lab code with 1σ error error) 147 – 153 4 130 ± 70 4421 – 4821 β-186144 275 – 285 5 950 ± 60 6549 – 6887 β-192607 347 – 353 6 750 ± 80 7433 – 7675 β-186145 464 – 472 19 411 ± 196 22541 – 23716 NZA-192608 595 – 601 > 43 630 N/A β-192608

5.4.2. Pollen Figure 5.2 shows the pollen counts and the zones identified.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 94

Figure 5.2. Pollen, charcoal, organic matter and pollen zonation for Lake Baraba. All pollen are represented as a percentage of the total count.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 95

Zone 1 (634 cm): The one sample in this zone is dominated (~80 % of the total count) by an unknown palynomorph that may possibly be an algal spore. Casuarinaceae pollen makes up ~10% of this zone with the remainder of the count composed of Halagoraceae, Myrtaceae, Poaceae, Podocarpaceae and Proteaceae. If the unknown palynomorph is disregarded, the composition of the pollen spectrum is similar to the zones above it. Total pollen concentrations are low in this zone.

Zone 2 (610-630 cm): Of the seven samples examined in this zone two were void of pollen (620 and 630cm). There were trace levels of Casuarinaceae, monolete spores, Podocarpaceae, Poaceae, Myrtaceae, Asteraceae, Chenopodiaceae, Cyperaceae and Dodonaea. Total pollen concentrations were very low in this zone and have been omitted from the pollen diagram.

Zone 3 (560-610 cm): This zone was dominated by Casuarinaceae (~60%) with Myrtaceae and Halagoraceae each representing ~10% of pollen counted. Poaceae, Asteraceae, Acacia spp. and Dodonaea spp. were represented although in small concentrations. The representation of monolete and trilete spores varied considerably in this section (~4 - ~25%). Several grains of Proteaceae, Podocarpaceae, Fabaceae, Euphorbiaceae, Cyperaceae and Chenopodiaceae were also seen. Total pollen concentrations were low in this zone.

Zone 4 (520-560 cm): Pollen was very sparse and has been omitted from the pollen diagram. Casuarinaceae was most common, followed by monolete/trilete spores, Halagoraceae and Podocarpaceae.

Zone 5 (490-520 cm): The pollen spectrum in this zone is very similar to Zone 3, with Casuarinaceae dominating, although Dodonaea was absent. Total pollen concentrations were slightly elevated in this section.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 96

Zone 6 (460-490 cm): Pollen was very sparse in this zone, however, Casuarinaceae (57%), trilete/monolete spores (10%), Poaceae (9%), Halagoraceae (8%), Myrtaceae (6%), Asteraceae (3%), Podocarpaceae (3%) and Chenopodiaceae (2%) were identified.

Zone 7 (410-460 cm): This zone is characterised by a high proportion of Casuarinaceae (~60%). Halagoraceae representation was high near the base, but decreased (from ~30% to 1%) through the zone. Myrtaceae (7.5%) and Poaceae (5%) are the only two other palynomorphs that are reasonably well represented. Total pollen concentrations are relatively high.

Zone 8 (230-410 cm): This zone is characterised by a substantial decrease in Casuarinaceae and an increase in Myrtaceae pollen. Poaceae concentrations are slightly elevated during this zone (~10%) and Cyperaceae, trilete spores and Restionaceae steadily increase through this zone. Dodonaea spp., Fabaceae and Halagoraceae also persist throughout this zone although in low concentrations. Chenopodiaceae increased through the zone and decreased towards the top. Total pollen concentrations are decreasing in this zone.

Zone 9 (170-230cm): There were very high concentrations of fungal spores throughout this zone and pollen was very sparse and actually absent in samples where fungal concentrations were especially high. The samples that did contain palynomorphs were dominated by Cyperaceae (24%), Myrtaceae (20%), Casuarinaceae (19%), Poaceae (18%) and trilete spores (5%).

Zone 10 (10-170 cm): Casuarinaceae and Myrtaceae each represented ~30% of the total pollen counted in this zone. Poaceae decreased from a high of ~50% and Halagoraceae increased abruptly to an average of ~30% in the upper part of this zone. Cyperaceae representation was generally low with the exception of two peaks at 60 and 80 cm. Trilete and monolete spores persisted throughout this zone. Total pollen concentrations were slightly elevated in the lower part of this zone and became very high in the upper part.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 97

Zone 11 (0-10 cm): Acacia spp., which is poorly dispersed (Dodson, 1983), represented ~5% of the total pollen sum. There were moderate concentrations of Myrtaceae, Casuarinaceae and Poaceae. Monolete spores peak at 10 cm representing ~35% of the pollen sum. Total pollen concentrations were relatively low when compared to the upper part of the previous zone.

5.4.3. Macroscopic charcoal Most of the analysed sediment profile from Lake Baraba has low levels of charcoal (Figure 5.3), with peaks at 240, 270 and 400 cm, a series of higher peaks at 230 – 200 cm and minor peaks at 170 – 130 cm. Macroscopic charcoal concentrations were extremely low between 470 and 455 cm and are very low until 405 cm. Charcoal is relatively low but variable between 405 and 275 cm and then increases abruptly and remains very high and variable until 200 cm. Between 185 and 55 cm there is a decreasing trend in charcoal concentrations, from being relatively high to moderately low, but it remains variable throughout this interval. The upper samples (50 to 0 cm) have very low charcoal concentrations with some samples almost void of charcoal. There is much less charcoal found in the clays when compared to the peat sediment.

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000 14000 Age yr BP) (cal. 15000 16000 17000 18000 19000 20000 21000 22000 23000 24000 50 100 150 200 250 300 mm2/cm3 Figure 5.3. Macroscopic charcoal curve versus age for the radiocarbon dated section of the Lake Baraba sequence.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 98

5.5. Discussion 5.5.1. Sedimentary history

Figure 5.4 presents a summary of the sedimentary history at the location of the sampled sediment core on Lake Baraba and gives the age-depth relationship. Sediments deeper than 595 – 601 cm are > 43 630 14C years in age. The site was initially a lake slowly accumulating clays (~0.04 mm per year). Extrapolation of the rate of sedimentation suggests that the base of the analysed profile could be 55 ka or older and hence it is speculated that the record may extend to the start of the Marine Isotope Stage (MIS) 3 (~59 to 24 ka). MIS 3 is an interstadial that corresponds with cool and moist conditions in

Figure 5.4 A comparison between the rates of sedimentation, sediment type and the pollen zones.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 99

south-eastern Australia suggested by high lake levels (e.g. Bowler, 1981: 1986; Wasson and Clark, 1988). During the latter part of the period conditions became increasingly drier and colder (Allan and Lindsay, 1998) culminating with the LGM.

The yellow/red/orange bands and mottling (605 to 635 cm, 510 to 555 cm and 410 to 262 cm) indicate oxidising conditions and are likely to reflect periodic drying of the lake. Uniform dark organic and grey clays (the colour representing iron in the ferrous state), however, indicate anaerobic conditions and were deposited when the lake was more permanent. Pollen is best preserved during permanently wet conditions and hence more pollen was recovered from the dark clays and very little from the yellow clays up to the LGM (Figure 5.4). Thus drier conditions alternated with wetter periods in the period leading up to the LGM.

It is difficult to assess just how ‘dry’ or permanent the lake may have been prior to the LGM. During historic times the lakes within Thirlmere Lakes National Park have been noted to recede significantly during times of drought, e.g. 1902, 1928 (NPWS, 1995), so it appears that they are hydrologically sensitive. Thus the sedimentary record of Lake Baraba, at a minimum, may indicate no more than the variation in moisture balance seen in historical times.

Reviews of lake levels in southeastern Australia (e.g. Harrison, 1993; Allan and Lindsay, 1998) indicate wetter conditions prevailed between 30 and 24 ka, and drier conditions from 24 ka, through the LGM until 12 ka (Harrison, 1993). Again, it is uncertain as to how ‘wet’ or how ‘dry’ these phases were. Reconstructed lake levels in the interior of southeastern Australia have also revealed mixed results, with some lakes being low and others either high or intermediate (Harrison and Dodson, 1993).

In the early Holocene (~8.5 ka), peat started accumulating at Lake Baraba and the rate of sedimentation increased markedly (~0.67 mm/yr). This indicates that at the site of the sampled core the lake had become shallow enough to facilitate swamp vegetation and peat formation. Peat accumulation has continued until the present although there was a hiatus,

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 100

perhaps in response to a dry period, between ~6 and 5.2 ka, indicated by high levels of fungal spores. Harrison (1993) described a dry phase between 6 and 5 ka and Pickett et al. (2004) reported increased moisture stress in the vegetation of the mid-Holocene. Chalson (1991) also identified the period ~6.5 – 5.5 ka as a time of oscillations between wet and dry conditions which is roughly coincidental with the Lake Baraba record. Allan and Lindsay’s (1998) synopsis of south-eastern Australian lake levels shows a marked drop between 5 and 4 ka which is asynchronous with the Lake Baraba record. Clearly regional differences should be taken into account.

5.5.2. The vegetation history The pollen profile suggests relatively stable vegetation with only one marked change, viz the decrease in Casuarinaceae and increase in Myrtaceae from the early Holocene. However, given the broad classification of the palynomorphs there may well have been undetected changes of species within the pollen groups.

The terrestrial vegetation of zones 2 – 7, from >43 ka to the early Holocene, was dominated by Casuarinaceae woodland/shrubland, at the time that the yellow/brown and dark grey clay bands were deposited. Halagoraceae is present in moderate frequencies in the dark clay bands, which could indicate the aquatic Myriophyllum in the lake. There is little evidence of any swamp vegetation (e.g. Cyperaceae and Restionaceae), suggesting that if it existed it must have been confined to a small zone around the lake. At this level of pollen identification, it appears that the composition of the terrestrial vegetation was much the same from >43 ka through the LGM, until the beginning of the Holocene. If shrubby forms of Casuarinaceae replaced trees under the harsher climate of the LGM the palynology could not detect this.

When the palynology of the LGM at Lake Baraba is compared with other records from southeastern Australia there are some major differences (Table 5.3). Notably, the results contrast with the general perception of treeless vegetation communities at the LGM. Casuarinaceae is prominent at Lake Baraba but uncommon in the ‘average’ and Poaceae is minor at Lake Baraba but common elsewhere in southeastern Australia. Lake Baraba,

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 101

however, is located on Hawkesbury Sandstone, a geological environment that is too infertile for grasslands. Chalson (1991) analysed a pollen record of the LGM in the Sydney Basin but the site was located such that Poaceae was more prominent. The comparatively low representation of Asteraceae and chenopods during the LGM at Lake Baraba further suggests that the site was not covered by steppe grassland. At Lake Baraba sclerophyllous vegetation dominated by Casuarinaceae appears to have continued to occupy the site throughout the LGM suggesting that the site acted as a possible refugium. Halagoraceae, dominated by Myriophyllum spp., also suggest shallow and fluctuating water levels during the LGM.

Table 5.3. A comparison between the average percentages for the south-eastern mainland pollen data-set for major taxa at selected time-slots from Kershaw (1995: 661) with the pollen record from Lake Baraba (Lake Baraba results are bolded). Kershaw (1995) used 71 pollen diagrams with only 12 of which covered all time periods.

Time Casuarina Eucalyptus Poaceae Asteraceae Chenopodiaceae Myriophyllum Restionaceae Slots Pre- 14 201 42 282 19 1 12 0 9 2 6 32 7 0 European 6 ka 20 261 37 182 21 7 9 1 5 1 6 3 6 12 9 ka 16 651 28 62 34 10 13 4 7 2 12 4 5 0 12 ka 4 541 17 142 45 5 23 1 15 1 3 13 3 0 LGM 3 571 7 62 44 9 35 3 11 2 21 8 1 0 1 Casuarinaceae, 2 Includes all Myrtaceae species

The climatic events that may have been experienced in Australia during deglaciation, such as the Antarctic Cold Reversal (~12.9-14.5 ka) (Blunier et al., 1997), are not evident in the Lake Baraba palaeoecological record. This is not to say that they were not experienced at the site but, more likely, the resolution of the record at this time was too coarse to register any changes. Furthermore if the site acted as some sort of refugium during the LGM it is unlikely that smaller changes such as the ACR would register in the record. Clark (1983) has previously suggested an inability of palynology to detect subtle changes in fire-adapted vegetation.

In southeastern Australia the Holocene generally depicts recovery of the vegetation and climate to its present day status. Patterns resembling the modern vegetation began to be

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 102

established from the late Pleistocene/early Holocene, when the climate approached that of today (Harrison and Dodson, 1993; Dodson, 1994; Kershaw, 1998). Dodson (1994) suggested that lowland vegetation was largely insensitive to environmental change during the Holocene, and our results concur with this overview. At Lake Baraba, vegetation similar to today began to establish from about 8.5 ka.

From ~8.5 ka, when peat accumulation started, Casuarinaceae declined and Myrtaceae increased. This change could indicate that Eucalytpus, the dominant genus of the vegetation today, increased at the expense of Casuarinaceae. In southeastern Australia, typical lowland forests of the early Holocene forests were dominated by Casuarinaceae (Lloyd and Kershaw, 1997) and a decline in Casuarinaceae and increase in Eucalyptus (or Myrtaceae) pollen is a feature of many Holocene pollen records (Dodson, 1974; Clark, 1983; D’Costa et al., 1989; Ladd et al., 1992; Devoy et al., 1994; Harle, 1998; Gale and Pisanu, 2001). The decline in Casuarinaceae has been dated as early as 7.5 ka (D’Costa et al., 1989; Aitken and Kershaw, 1993) and as late as 4.5 ka (Hooley et al., 1980). At Lake Baraba Myrtaceae, most likely eucalypts or paperbarks, expanded at the expense of Casuarinaceae from ~8.5 ka.

The reasons for this rise in Eucalyptus at the expense of Casuarinaceae throughout the Holocene have been debated (Dodson, 1974; Hooley et al., 1980; Clark, 1983; D’Costa et al., 1989) with competitive exclusion, increased moisture and altered fire regimes most commonly suggested (Lloyd and Kershaw, 1997). Crowley (1994) and Cupper et al. (2000) have suggested that the Holocene decline in Casuarinaceae was linked to soil salinity but this is an unlikely reason for the decline in Casuarinaceae at Lake Baraba considering the freshwater setting.

The fire sensitivity of the Casuarinaceae family has been controversial (e.g. see Clark, 1983), however, the Casuarinaceae palynomorph includes species that are fire tolerant and those that rely of fire for regeneration (Ladd, 1988). Fire activity at Lake Baraba gradually increased from ~7.5 ka and abruptly increased at ~6.7 ka where it remained high until ~5.5 ka. The decline in Casuarinaceae and associated rise in Myrtaceae precedes this increase in

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 103

fire activity by ~1 000 y. This suggests that fire activity was a response to the establishment of the more fire-prone Myrtaceae and swamp vegetation, rather than the cause of this change in vegetation.

Ladd (1988) has suggested that the morphology of Casuarina species means that they are poor competitors with broad-leaved species. At Lake Baraba it is suggested that the pre- swamp environmental conditions were more suitable for Casuarinaceae but this declined as a swamp formed in the early Holocene. The relationship between the Casuarinaceae/ Myrtaceae ratio and the evolution of the lake to a swamp (indicated by the organic content) is depicted in Figure 5.5. The lake environment corresponds with a low Myrtaceae/Casuarinaceae ratio between ~20 and 8.5 ka. Swamp conditions are indicated by a high Myrtaceae/Casuarinaceae ratio from ~8.5 ka to present.

100 1.8 LOI Myrtaceae:Casuarinaceae ratio

90 1.6

80 1.4

70 1.2

60 1

50

0.8 Unitless ratio Unitless 40 Organic matter (%)

0.6 30

0.4 20

10 0.2

0 0 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 Time (yr BP)

Figure 5.5. A comparison between the Myrtaceae/Casuarinaceae ratio and the organic content of the sediment from loss-on-ignition.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 104

Chalson (1991) undertook palynological analysis on seven swamps in the Blue Mountains, New South Wales to find a trend to wetter conditions from 11 to 8 ka with maximum moisture from 8 ka to 6.5 ka. A similar pattern is detected at Lake Baraba with Restionaceae and trilete spores, likely to be wet indicators, increasing from ~8.6 ka to a maximum representation at ~6 ka.

The fungal spores of Zone 9 (6 – 5.2 ka) suggest the swamp surface periodically dried out (Pals et al., 1980; Elsik, 1996) but these dry periods were not severe enough to disrupt peat production. This zone follows the two peaks of major fire activity, which continued at lower levels through the zone. It is likely that dry spells allowed increased fire activity. During this time there was a decreased tree cover and Cyperaceae and Poaceae pollen are elevated which may reflect this postulated drier climate. Very high concentrations of charcoal in the latter part of Zone 8 (~6 ka) suggest high fire activity, and this may have also influenced the tree cover during Zone 9 (6 – 5.1 ka).

By 6 ka and especially after ~5 ka, the vegetation composition is similar to the Eucalyptus/Casuarina woodland that grows on the site presently. The 6 ka biome reconstruction by Pickett et al. (2004) also implies that the broad-scale pattern of Australian vegetation at this time was largely similar to the contemporary environment. The surface samples that include the European period have decreased concentrations of tree pollen (i.e. Casuarinaceae and Myrtaceae) perhaps caused by land clearing and logging postdating European colonisation.

5.5.3 Fire history At Lake Baraba charcoal was found in all parts of the analysed profile but varied markedly, inferring that fire has been a persistent but variable component of the environment. Kershaw et al. (2002) compiled a number of charcoal records throughout the Australasian region that cover the LGM. When the Lake Baraba record is compared with these records it has comparatively little charcoal during the LGM (Table 5.4). Fire was a more significant feature of the Holocene at the site, however, the aforementioned changes to the sedimentary environment should be considered. The initiation of the swamp in the early Holocene

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 105

means that fire may have impacted on the site itself occasionally, thereby increasing charcoal delivery.

Anthropogenic fire must also be considered since fire is often attributed to the presence of Aboriginal people. Attenbrow (2004) recently analysed archaeological data from 58 sites in

Table 5.4: The key Australasian sites that have a record of charcoal throughout the Last Glacial Maximum. The scale for abundance of charcoal was based on the ranking of charcoal in the time periods formulated by Kershaw et al. (2002: 6). Site Abundance of Charcoal measure Reference charcoal* Lombok Ridge, eastern Very high Particles/cm3 Wang et al. (1999) Indian Ocean Banda Sea, west of New Very high Particles/cm2/yr + van der Kaars et al. Guinea mg/cm2/yr (2000) ODP site 820 offshore, Very low Particles/cm3 Moss (1999); Moss and NE Queensland (QLD) Kershaw (1999) Lynchs Crater, NE QLD High Particles/cm3 Kershaw (1986) Lake George, New Very low Surface area%/ unit Singh et al. (1981); South Wales (NSW) volume of sediment Singh and Geissler southern highlands (1985) Lake Wangoon, western Very high Particles/cm3 Edney et al. (1990); Victoria Harle (1998) Egg Lagoon, NW Very high mm2/cm3 D’Costa (1997) Tasmania (TAS) Lake Selina, western Very high Not given Colhoun et al. (1999) TAS Darwin Crater, western Very high Not given Colhoun and van de TAS Geer (1988) Lake Euramoo, NE Low Part./gm/yr Haberle (in press) QLD Lake Baraba, Sydney Very low Area (mm2/cm3) Basin, NSW *Relative abundance of charcoal when compared to the rest of the sequence

Sydney and the NSW South Coast, including the hinterland, to determine the rate of Aboriginal habitation establishment and the number of habitations used in each millennium. Figure 5.6 shows that charcoal at Lake Baraba and the number of habitations used in the region both gradually increase from the LGM to the early Holocene. From ~7 to 8 ka this relationship breaks down, with the number of habitations increasing more rapidly whilst charcoal content declines steadily. The low levels of charcoal and high site usage

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 106

from during the late Holocene may represent a change in fire regime (e.g. low intensity fires) associated with Aboriginal activity. However as site usage is only a coarse measure of past population levels, at best (Attenbrow, 2004), the influence of humans on fire in this environment warrants further investigation.

Wang et al. (1999) have suggested the possibility that the expansion of Eucalyptus in the mid-Holocene “marks the development of fire as a management strategy” by Aboriginal

60 No. habitations used 55 charcoal

50

45

40

35

30

25

20

15

No. of habitations used/ charcoal concentration (mm2/cm3) concentration charcoal used/ of habitations No. 10

5

0

-2 -6 -7 -8 -9 1 2 3 4 5 6 8 2 3 4 5 0-1 2-3 3-4 4-5 5 6 7 8 1 -1 -1 -1 -1 -1 -17 -1 2 2 -2 2 1 9-10 1 2 3 5 3 10- 1 1 1 14 1 16 17 18-19 19-20 20-21 21- 22- 2 24- Time period (in thousands of years BP)

Figure 5.6. A comparison between the average charcoal values for Lake Baraba and the number of habitation sites used in the region for millennial time scales. Source of archaeological data: Attenbow (2004; Table A4/1). people associated with intensification. As the increase in Myrtaceae at Lake Baraba began from ~8.5 ka, Wang et al.’s (1999) suggestion that the “genetically plastic” eucalypts became adapted to Aboriginal fires regimes in the mid-Holocene is not supported by the Lake Baraba data.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 107

As an alternative explanation, the gradual increase in charcoal at Lake Baraba from the LGM may reflect the likely increase in biomass accompanying climatic amelioration. It is entirely plausible that other changes in fire during the Holocene may also be unrelated to human activity. Peaks in charcoal appear to be related to times of vegetation or environmental change and fires are possibly more frequent or intense at times of climatic instability (Edney et al., 1990; Haberle et al., 2001; Kershaw et al., 2002; Black and Mooney, in press). As an example, at Lake Baraba the increase in fire activity from 6.7 ka may reflect the onset of modern ENSO activity, and the decrease in fire from about 5.5 ka the cessation of the response to this continuing variability. Notably, Kershaw et al. (2002) suggested that 7 to 5 ka was a period of less fire in south-eastern Australia.

5.6. Conclusion

The analyses of the Lake Baraba sequence resulted in a vegetation, fire and sedimentary history dating to >43 ka and included the LGM and subsequent climatic amelioration. There were abrupt sedimentary changes between black organic clays and yellow/orange clays prior to the LGM and this is likely to be due to lake level fluctuations. Prior to the early Holocene the site was a lake depositing clay. Peat formation began from ~8.5 ka indicative of the change from a lake to a swamp due to the infilling of the site.

The vegetation remained the same from >43 ka to the early Holocene, most probably a Casuarina woodland/shrubland with a mixed understorey. The sclerophyllous vegetation adapted to the relatively infertile substrate of the site appears to be relatively resilient to changes in climate, or the site was a refugium. Myrtaceae expanded at the expense of Casuarinaceae from the early Holocene. This shift in vegetation preceded any increase in fire and hence it is unlikely that this was the cause.

Fire activity was low during the Pleistocene and especially low during the LGM. Fire activity increased during the Holocene although the record does not concur with the broad

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 5 108

pattern of Holocene fire activity described by Kershaw et al. (2002). There was no apparent relationship between Aboriginal site usage and fire activity and hence how Aboriginal people fire activity at Lake Baraba remains speculative. Although the low levels of charcoal during the late Holocene may be associated with an altered fire regime with less intense fires due to more Aboriginal activity in the catchment. Local biotic and abiotic factors (e.g. the change in the depositional environment from a lake to a swamp) may have been important controls of fire activity and charcoal deposition at the site.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 109

Chapter 6: A ~6,100 yr vegetation and fire history from Kings Waterhole, Wollemi National Park, New South Wales

Abstract Pollen and macroscopic charcoal has been analysed from a 6.1 ka sedimentary sequence from a site within Wollemi National Park to the northwest of Sydney, which forms a part of the Greater Blue Mountains World Heritage Area. There have been relatively minor changes in the vegetation over the last 6100 yr, which is interpreted as reflecting the resilient nature of the sclerophyllous vegetation found on Hawkesbury Sandstone throughout the Sydney Basin. Casuarinaceae declined in the late Holocene, a trend that has been detected in numerous palaeoecological studies throughout southeastern Australia. This decline was unrelated to fire, which has been a persistent feature at the site over the past 6.1 ka. Fire activity increased dramatically from 5.7 ka but decreased from 3 ka and remained low until the recent past. This is interpreted as reflecting the adjustment to the onset of increased climatic variability from the mid-Holocene, but it may also be associated with Aboriginal people and their burning practices. Aboriginal people may be managing fire more intensely in the mid-late Holocene due to the risks of more intense fires under an El Niño dominated climate. The record showed no marked changes in the vegetation composition or fire activity associated with the arrival of European people.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 110

6.1 Introduction

This study presents a ~6 100 yr record of vegetation change and one of the few contiguous fire records from the Sydney Basin region. The impact of human activity and climate on the vegetation and fire dynamics of the site are investigated in this study. The site is found within Wollemi National Park, which protects the largest wilderness area remaining in New South Wales (NPWS 2001) and forms a part of the Greater Blue Mountains World Heritage Area. The Mellong Swamp plant communities that surround the study site are poorly conserved outside the reserve and these communities support a unique assemblage of plants and animals (NPWS 2001). The ecological importance of applying appropriate fire regimes to Sydney sandstone communities has been well-documented (e.g. Clark 1988; Cary & Morrison 1995; Morrison et al. 1995; Conroy 1996).

Dodson and Thom (1992) described that our understanding of the vegetation history of the Sydney Sandstone Complex were in their infancy and to this date there are still few studies examining the Holocene environmental history of the region. The palaeoenvironmental record of the Wollemi region is also poorly understood, although there are a number of studies from the Sydney Basin that can be used for comparison: Kodela and Dodson (1988) presented a 5-6 000 yr pollen and charcoal record from Ku-ring-gai Chase National Park; Dodson and Thom (1992) interpreted a Holocene pollen and charcoal record from the nearby Hawkesbury Valley; Johnson (1994) and Martin (1994) have provided records from the Kurnell Peninsula; Chalson (1991) and Black and Mooney (in press) have provided pollen and charcoal records from the Blue Mountains region; Chalson (1983) and Mooney et al. (2001) provided palaeoenvironmental records from Jibbon Swamp in Royal National Park, south of Sydney; and Black et al. (submitted) provided a palaeoenvironmental record from Lake Baraba southwest of Sydney and Mooney et al. (submitted) have presented the palaeoecological record from Griffith Swamp. Kershaw et al. (2002) also provided a synopsis of fire patterns in south-eastern Australia and the Kings Waterhole fire record will be compared to this. In addition studies further

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 111

afield in southeastern Australia have provided a basic understanding of the climate and vegetation history of the late Quaternary (e.g. Dodson 1994; Kershaw 1995).

The transition from the Pleistocene to the Holocene was a time of significant climate change and this occurred at ~11.5 ka (Kershaw 1995). Dodson and Mooney (2002) have previously suggested that the Holocene was relatively stable when compared to the late Pleistocene. Nonetheless, some overall trends are evident during the Holocene. During the early-mid Holocene climatic amelioration continued culminating in the ‘Holocene Climatic Optimum’ (~7 – 5 ka) where temperatures, precipitation and vegetation cover were higher than present (Kershaw 1995).

After the relative stability of the Holocene Climatic Optimum, climatic instability is believed to have increased from the mid-Holocene and this may have been related to the onset of ‘modern’ El Niño Southern Oscillation (ENSO) events (McGlone et al. 1992; Shulmeister 1999; Rodbell et al. 1999; deMenocal et al. 2000; Sandweiss et al. 1999). Shulmeister (1999) has described a decoupling of the northern (tropical) and southern (temperate) climate systems of Australia at ~5 ka resulting in increased westerlies, the loss of summer monsoon rainfall and a sharp decline in effective precipitation in southern Australia (Shulmeister 1999). Sandweiss et al. (2001) suggested the onset of modern ENSO events from ~5.8 ka with an increased frequency from ~3.2 ka. Similarly Riedinger et al. (2002) described an increase in the intensity and frequency of El Niño events from 3.1 ka.

In addition to climatic trends, there are a number of general trends in the vegetation that are thought to have occurred in southeastern Australia during the Holocene. The replacement of Casuarinaceae-dominated woodland with Eucalyptus has been widely reported (Dodson 1974; Clark 1983; D’Costa et al. 1989; Ladd et al. 1992; Devoy et al. 1994; Harle 1998; Gale & Pisanu 2001) although the reason for this transition have been debated (e.g. Lloyd & Kershaw 1997). Black et al. (submitted) have recently demonstrated that the decline in Casuarinaceae during the Holocene from Lake Baraba, southwest of Sydney, was unrelated to fire. High moisture levels during the Holocene Climatic Optimum are believed to be

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 112

responsible for maximum values of woody taxa and rainforest taxon in many of southeastern Australia’s pollen diagrams during this time (Kershaw 1995).

Kershaw (1995), and more recently Pickett et al. (2004), have suggested that the general features of today’s vegetational landscape were established from ~6 ka. However subtle changes, such as the opening up of vegetation canopies and increases in sclerophyll, heath and peatlands, are thought to have occurred since this time. These changes are variously attributed to increased climatic variability or the intensification of human populations (Kershaw 1995).

The theorised intensification of Aboriginal occupations from the mid- to late-Holocene (e.g. Lourandos 1980; 1983) has been one of the most contentious debates within Australian archaeology. The theory of ‘intensification’ is thought have involved social, ideological, political and demographic changes to Aboriginal populations (Lourandos 1983) and hence increased ecological impacts from Aboriginal people, especially through their use of fire, have been suggested (e.g. Lourandos 1983; Head 1989). It became the focus of archaeological research from the 1980s and was inferred from the unprecedented increases in the discard rates of artefacts and site usage from ~4 ka (e.g. Lourandos 1980: 1983; Rowland 1983; Morwood 1984; Ross 1984; Flood et al. 1987; David 1990; Bird & Frankel 1991; Lourandos & Ross 1994) and often associated with the ‘Small Tool Tradition’ (Attenbrow 2004).

The Wollemi region of New South Wales has recently become the focus of archaeological research with the finding of significant sites of rock art within Wollemi National Park. Prior to these finds only 120 Aboriginal sites had been recorded within Wollemi National Park (NPWS 2001) and hence the area was not thought to have supported an intensive or permanent human population. It is now possible that the Wollemi region became a ‘wilderness’, void of people, only after the arrival of European people.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 113

6.2 Site description

The study is based at Kings Waterhole (33o1’S, 150o40’E, 280 m asl), which forms a part of the Mellong Swamp complex and is located on the eastern boundary of Wollemi National Park. The Park forms the northern and western edge of the Sydney Basin and is located 100 km northwest of Sydney (Figure 6.1). The Mellong Plateau is dominated by Hawkesbury Sandstone which forms the ridges and the plateaux but Narrabeen Group sandstones and shales, which underlie Hawkesbury Sandstone, form the gullies (Ryan et al. 1996). Quaternary deposits of sand, clays and peats have infilled many of these gullies and also occur along waterways. The plateau forms the drainage divide between the MacDonald River and Wollemi Creek. Kings Waterhole is one such example of a low relief valley that has been infilled with Quaternary alluvial deposits and peat, dissected by sandstone ridges. A geophysical survey of the Mellong area by Riley & Henry (1987) described recent valley development on the Mellong Plateau as one of aggradation combined with colluvial and alluvial deposition.

The nearest climate station is Richmond (33o36’S, 150o47’S), some 70km south of the site. The climate is warm and temperate with an average maximum temperature range of 29.6oC in January to 17.2oC in July, and average minima of 17.4oC and 5oC in January and July, respectively (BoM, 2005). The mean annual rainfall is 810.3 mm with a distinctly summer dominated rainfall meaning that February is the wettest month (105.6mm) and July the driest month (35.9mm) (BoM 2005). Kings Waterhole is located in a rainshadow of the Hawkesbury-Nepean catchment between the Great Dividing Range, the upper Blue Mountains and the ranges towards the coast (Ryan et al. 1996) hence average annual rainfall is likely to be less than at Richmond.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 114

Figure 6.1. Location of Kings Waterhole.

Ryan et al. (1996) described four major vegetation communities in the study area. Mellong Woodlands are found on a range of poorly- to well-drained soils on the plateau, Mellong Swamps are found in water-logged soils, Sydney Sandstone Ridgetop Woodlands occur on the exposed slopes of Hawkesbury Sandstone and the Sydney Sandstone Gully Forests are found in sheltered slopes and gullies (Ryan et al. 1996). The woodland communities are dominated by Eucalyptus (e.g. Redblood, Yellowblood, Scribbly Gum), Angophora and Casuarina. The major species on the swamp surface include Eleocharis spacelata, Persicaria decipiens, Triglochin procera, Lepryodia scaroisa and Schoenus brevifolius. Stands of Callistemon rigidus and Acacia longifolium surround the swamp.

At Mellong Swamp the fire season extends from late spring to late summer and is when low relative humidity and high temperatures occur. Although the highest rainfall occurs during summer, there is also high evaporation resulting in reduced water availability (NPWS 2001). Documentation of fires for the area surrounding Mellong Swamps began in 1943 and since this time there have been 14 bushfires that have affected the area averaging one every four years (NPWS 2004).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 115

Previous research has suggested that Aboriginal people have been using Wollemi National Park from ~12 ka with others suggesting establishment of habitation from 10 – 5 ka (Moore 1981). The remoteness and ruggedness of Wollemi National Park has, up until recently, prevented much archaeological investigation of the region. The recent discovery of rock art sites has led to speculation on the intensity and antiquity of Aboriginal occupation within the region. Attenbrow (2003) suggested that the region may have belonged to a range of bands associated with the language group Darginung. The Windradyne, Wanaruah, Darkinjung and Dharug Aboriginal Land Councils are currently involved in the management of the Aboriginal heritage within Wollemi National Park (NPWS 2001).

6.3 Method

A 5.55 m sediment core was extracted from the swamp adjacent to Kings Waterhole (KW) using a Russian d-section corer (Jowsey 1966) in April 2004. The stratigraphy of the core was described using a modified version of the Troels-Smith method (Kershaw 1997) and was photographed. Three sections of the core (147-154, 348-353 and 547-553 cm) were submitted for radiocarbon dating. The radiocarbon ages were calibrated using the SHcal04.14C data set (McCormac et al. 2004) in CALIB v5 (Stuiver et al. 2005).

Macroscopic charcoal, which is thought to represent local or catchment fire events (Whitlock & Millspaugh 1996), was analysed using a modified version of the ‘Oregon sieving method’ (Long et al. 1998) and image analysis (Mooney & Black 2003). Volumetric sub-samples from 2.5 cm contiguous intervals were taken from the core and dispersed for 24 hr in dilute (~6%) sodium hypochlorite (bleach) to remove the pigment from organic matter and hence aid in the identification of charcoal. This was then washed through a 250 μm sieve and a digital image of the collected material was obtained using a 4 mega-pixel Nikon Coolpix 4500 digital camera. The area of charcoal was calculated using image analysis software (Scion Image Beta 4.02 for Windows).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 116

Pollen samples were prepared using standard palynological techniques (Faegri & Iverson, 1975). Volumetric samples were taken from 1cm depth increments every 10 cm along the core giving a total of 56 samples. The samples were deflocculated with hot 10 % NaOH and then sieved through a 150 μm mesh. Silicates were removed using heavy liquid (i.e.

ZnBr2(aq.)) separation and organic matter with acetolysis. Samples were mounted in silicon oil and the palynomorphs were counted at 400 X magnification until 200 grains were identified. Nineteen different palynomorphs were quantified, chosen to represent the major taxa currently growing on or around the site. Maleleuca spp. was separated out from other members of the family Myrtaceae. The pollen counts were expressed as percentages, with all grains contributing to the pollen sum.

6.4 Results

The results of the radiocarbon dating are shown in Table 6.1. The calibrated 14C ages were used to formulate a linear age-depth curve (calibrated age in yr BP = 10.799x + 130 where x = depth in cm). The basal age of the analysed core was ~6 100 cal. yr BP based on the age-depth curve.

Table 6.1. Radiocarbon dates and calibration for Kings Waterhole. Calibration results from CALIB v5 (Stuiver et al., 2005). The mid-points of calibrated year ranges are used in age-depth model calculations.

Depth (cm) 14C date BP with Cal. yrs BP* (2 σ Lab code 1σ error error) 147 – 154 2220 ± 60 2003 – 2327 β 186146 348 – 353 3280 ± 70 3269 – 3635 β 186147 547 – 553 5560 ± 90 6014 – 6491 β 186148

The stratigraphy of the core revealed that peat and clays have been deposited for majority of the past ~6 100 yrs although occasional bands of sand suggest periods of higher energy. The upper 70 cm of the core consisted of peat with fibrous roots. Pinus pollen was found to a depth of 20 cm and this was interpreted as approximately the time of European arrival to the region.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 117

Figure 6.2 shows the results of the palynological analysis of Kings Waterhole. The pollen spectra does not change significantly throughout the sequence with all major pollen taxa represented although there were changes in the relative proportions of the groups.

Figure 6.2. Results of the palynological analysis of Kings Waterhole

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 118

Myrtaceae formed the largest component of the pollen spectra throughout the core with values averaging 33% of all pollen counted and reaching a maximum of 70% at 80cm (~1 ka). Myrtaceae values decreased in the top 70 cm (~0.9 ka) of the core and reached a minimum of 8% for the surface sample. Higher concentrations of Maleleuca spp. were found in the upper 140 cm (~1.6 ka) of the core. Casuarinaceae was also well-represented in the woody taxa, especially between 555 (~6.1 ka) and 190 cm (~2.2 ka) where it averaged 24% of total pollen count. Casuarinaceae also revealed a marked decline to an average of 6% for the upper 190 cm (~2.2 ka) of the core.

Restionaceae was the best represented of the swamp taxa and displayed a similar trend to Casuarinaceae, being well-represented in the lower 365cm (~4.1 ka) of the core but with a marked drop in the upper 190 cm (~2.2 ka). Halagoraceae, represented mostly by the freshwater aquatic Myriophyllum spp., averaged ~ 4% and was more abundant from the base (~6.1 ka) to about 365 cm (~4.1 ka). Ferns and mosses, Epacridaceae and Cyperaceae, other swamp taxa, were not very well represented throughout the core although there were very high concentrations of Cyperaceae (45%) for the surface sample. Poaceae showed an increasing trend throughout the core with values averaging 27% for the upper 70cm (~0.9 ka). Pteridium was better represented in the upper 160cm (~1.9 ka) of the core.

Banksia, Asteraceae and Chenopodiaceae were found throughout the sequence although at low concentrations. Dodonaea was well represented for the lower ~3m (~6.1 – 3.4 ka) of the core. There were very small quantities of pollen from Acacia, Juncaginaceae and Asteraceae (fennestrate).

Macroscopic charcoal was present throughout the entire sedimentary sequence of Kings Waterhole although the concentrations of charcoal did vary considerably (Figure 6.3). Charcoal was initially low between 555 and 515 cm (~6.1 and 5.7 ka) but then increased dramatically. Charcoal remained high and variable between 515 and 265 cm (~5.7 ka and 3 ka) with the exception of a period of reduced charcoal deposition between 470 and 430 cm (~5.2 and 4.8 ka). Charcoal decreased rapidly at 265 cm (~3 ka) and remained generally

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 119

low from this time to the present. Charcoal deposition for the top 80 cm (the past ~1000 yr) was especially low.

500

450

400

350

300

250

200 Macroscopic charcoal

150

100

50

0 0 1000 2000 3000 4000 5000 6000 Age (cal. yr BP)

Figure 6.3. Results of the charcoal analysis from Kings Waterhole, detailing the area (mm2/cm3) of charcoal fragments > 250μm in size and is expressed in area

6.5 Discussion

There were no major shifts in the flora over the past ~6 ka and this suggests that there was little climate change or that the sclerophyllous vegetation that occupies the site is relatively resilient to minor environmental changes. However there appears to have been a slight drying of climate over the past 2000 years based on a decrease in swamp taxa and an increase in Poaceae. The mixed Eucalyptus/ Casuarina woodland, with a sclerophyll shrubby understorey, which currently surrounds the swamp is believed to have existed for the entire analysed sequence. The lack of any significant change in the vegetation composition at Kings Waterhole from ~6.1 ka is supportive of Kershaw’s (1995) and

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 120

Pickett et al.’s (2004) suggestion that most of the characteristics of southeastern Australia’s vegetation were established by ~6 ka. Furthermore Dodson and Thom (1992) have suggested there have been no significant responses within the Sydney sandstone flora since the climatic shift of the Holocene Climatic Optimum.

Kodela & Dodson (1989) similarly recorded no significant changes over the past ~6 000 years in the vegetation from Ku-ring-gai Chase National Park, some 80 km southeast of Kings Waterhole. Notably their site was also surrounded by Sydney Sandstone Complex and dominated by Hawkesbury Sandstone. Kodela & Dodson (1989) suggested that the sclerophyllous vegetation that dominates Hawkesbury Sandstone was resilient to fire, nutrient-poor sandy soils, exposure to high insolation and drought, and probably insensitive to minor environmental changes. Black et al. (in prep.) also suggested that the sandstone vegetation surrounding Lake Baraba was resilient to climatic change based on little change in the palynology over the last ~43 ka. There was also little change during the Holocene in the Pillaga Sandstone sclerophyllous vegetation, from the northwest slope of New South Wales (Dodson and Wright 1989).

Kershaw (1995) suggested that during the late Holocene there was an opening up of forest canopies throughout southeastern Australia and attributed this to climatic variability or the intensification of human populations. At Kings Waterhole a grassy understorey, represented by high concentrations of Poaceae, increased from ~2.2 ka and there was a decrease in woody taxa (Myrtaceae) from ~1.1 ka. This shift is suggestive of this opening up of the canopy although attributing a cause is difficult. Chalson (1983) and Kodela & Dodson (1989) also revealed a decline in Eucalyptus in the late Holocene.

The representation of swamp indicators in the vegetation was highest between 5 and 2.2 ka and this may reflect a period of increased moisture and perhaps size of the swamp. Chalson (1991) identified an approximately coeval wet period from 5.5 to 2.5 ka from the nearby Blue Mountains. The increase in Poaceae from 2.2 ka may also be representative of grasses encroaching on the swamp surface during times of periodic drought. Distinct changes to the

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 121

vegetation of the swamp surface have been witnessed during the 2001- 2005 drought and this has included increased colonisation by grass species. Swamp taxa, such as Restionaceae and Halagoraceae, have a reduced representation during the past ~2.2 ka which is further supportive of a relatively drier period. Chalson (1991), Kodela & Dodson (1989) and Dodson (1986) identified the past 2 000 yrs as a period of generally drier climate. Kershaw et al. (2002), however, has suggested that the climate of southeastern Australia was drier and cooler between 4 and 2 ka and saw the return to wetter conditions in the past 2000 years. This contradiction could perhaps result from a wetter southern part of southeastern Australia during the past 2 000 years at the same time the northern parts of southeastern Australia experienced drying. This hypothesis could be associated with the southward movement of the cold fronts that are associated with precipitation in southeastern Australia.

At Kings Waterhole there is a general decline in Casuarinaceae and this is especially evident during from the past ~2.2 ka. The decline in Casuarinaceae during the Holocene in southeastern Australia has been widely reported and this is often associated with an increase in Eucalyptus (e.g. Dodson 1974; Clark 1983; D’Costa et al. 1989; Ladd et al. 1992; Harle 1998; Gale and Pisanu 2001). At Kurnell Peninsula, a coastal site just south of Sydney, Casuarina woodland was established by 8 ka, but Eucalyptus increased and by 5.3 ka exceeded Casuarina as the dominant vegetation at the site (Martin 1994). Dodson & Thom (1992) described a general trend of increasing Eucalyptus and decreasing Casuarina throughout the Holocene from Mill Creek on the Hawkesbury River, approximately 40 km southeast of Kings Waterhole, and attributed this change to a decline in upper saltmarsh communities and increasing sclerophyll and woodland cover.

At Kings Waterhole Myrtaceae remained relatively constant and hence Eucalyptus is unlikely to be associated with the decline in Casuarinaceae at this site. The substantial decline is coincident with the dry period inferred from the reduced swamp cover and possible encroachment of grass onto the swamp surface. This suggests the decline in Casuarinaceae at Kings Waterhole reflects hydrological changes at the site.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 122

The analysis of charcoal at Kings Waterhole suggests that fire was a persistent feature of the local landscape throughout the past 6.1 ka from Kings Waterhole. Despite this persistence fire activity appears to have changed significantly over this period.

Fire activity was initially low between ~6.1 and 5.7 ka and then increased markedly at ~5.7 ka to remain high and variable until 3 ka after which fire activity decreased rapidly. There was also a slight decrease in activity between 5.2 and 4.8 ka. It is likely that climates in southeastern Australia became more variable from the mid-Holocene as numerous studies have dated the onset of ‘modern’ ENSO events at this time (e.g. McGlone et al. 1992; Shulmeister 1999; Rodbell et al. 1999; deMenocal et al. 2000; Sandweiss et al. 1999). The onset of climatic variability and times of vegetational change have been linked to increases in charcoal due to more frequent or intense fire activity (e.g. Edney et al. 1990; Haberle et al. 2001; Kershaw et al. 2002; Black and Mooney in press).

Black and Mooney (in press) also found a rapid increase in fire activity at ~5.7 ka from Gooches Swamp in the nearby Blue Mountains, and they suggested increased climatic variability associated with the onset of ‘modern’ ENSO events as a cause. Kings Waterhole displayed a rapid decline in fire activity at ~3 ka whilst Black and Mooney (in press) found that high and variable charcoal persisted from the mid-Holocene to the present. Perhaps climate is the dominant control on fire activity at Kings Waterhole between ~5.7 and 3 ka. Kershaw et al. (2002) stated that there was a significant increase in burning after 5 ka in all vegetation types across Australia (except the wet forests) and this was most likely associated with climate.

Hopbush (Dodonaea spp.) responds positively to fire and has been observed to become more abundant during periods of high fire activity (Fairley & Moore 2000). This vegetational response is evident at Kings Waterhole with higher concentrations of Dodonaea spp. between 5.7 and 3 ka when fire activity was high, and very low concentrations corresponding with ~3 ka to present when fire activity was low to very low. In this setting Dodonaea spp. appears to be a good indicator of frequent fire activity. The representation of bracken (Pteridium esculentum) at Kings Waterhole, which also thrives

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 123

on the disturbance following a fire (Fairley & Moore 2000), does not appear to be related to fire activity. In fact during periods of high fire activity there are very low concentrations of bracken whilst during the period of low fire activity, in the upper part of the record there were higher levels of bracken. Dodson (1986) had previously noted little association between pollen and charcoal curves and suggested that either: 1) fire had no impact at the level of pollen identification; 2) there were significant differences between the spatial sources of the pollen and charcoal; or 3) the fire-sensitive species had already been eliminated from the environment by the time the site started recording the environmental history and that a fire-vegetation equilibrium had already been established.

At Kings Waterhole there is a marked decline in charcoal from ~3 ka, elevated charcoal between ~2.4 to 2.2 ka, and a subsequent decline from 2 ka with no apparent increase in the European period. This change in charcoal is likely to reflect a major alteration to the fire regime.

Attenbrow (2003) examined the past habitation and landuse patterns of Aboriginal people in the Upper Mangrove Creek catchment, which is approximately 30 km east of Kings Waterhole. This study analysed thirty excavated archaeological deposits from rockshelters in the region and presented several indices including rate of habitation establishment and number of habitations used per millennium to quantify “how use of the catchment changed over time” (Attenbrow 2003: 21 – 22). The results of this study suggested that habitation patterns and subsistence organisation changed over the past 11 ka with an unprecedented level of change occurring from 3 ka with a substantial increase in base camps. Attenbrow (2003) theorised that the changing ‘habitation’ distribution patterns reflected “a re- organisation of mobility patterns…relating to camp life and subsistence activities took place in the catchment at frequent intervals and even more frequently and dynamically in the last four thousand years”. Similarly Flood et al. 1987) suggested that Aboriginal occupation of the Sydney region increased from ~3 ka. A comparison between the archaeological data from the Upper Mangrove Creek and the Kings Waterhole charcoal curve is shown in Figure 6.4.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 124

The altered fire regime from 3 ka at Kings Waterhole is likely to be associated with increased human activity in the region. The fire regime of Aboriginal people is generally thought as being frequent but of low intensity (Ryan et al., 1995) associated with resource manipulation (e.g. Jones, 1969). A change from high intensity natural fire to low intensity anthropogenic fire from ~3 ka may have resulted in less charcoal being delivered to the sediments of Kings Waterhole. Furthermore large intense fires are more likely to encroach onto the swamp surface and deposit charcoal in situ. Other previous studies have attributed changes in charcoal to the fire intensity (e.g. Singh et al., 1981; Hope, 1999). In their study in the nearby Hawkesbury Valley Dodson & Thom (1992) have suggested that lower values of charcoal input to the sediment could have resulted from a more frequent but a lower intensity burning regime.

Aboriginal people may have prevented the swamp surface being burnt to protect food sources living in the swamp such as yabbies, turtles and the plant species Triglochin procera, which grows at Kings Waterhole, and has been used as an Aboriginal food in southeastern Australia (Gott 1982; 2005). However it should be noted that Lewis (1992) found that Aboriginal people in the Northern Territory applied frequent low intensity fire to swamp surfaces every 8 – 10 yr to improve productivity.

There are very low levels of charcoal found in the European period which is surprising considering the relatively high frequency of intense bushfires that have occurred in the landscape in recent times. NPWS (2001) suggested that the high frequency of fire within Wollemi National Park is likely impact on the long-term biodiversity of the plant communities.

The unexpected lack of charcoal may be attributed to the very high levels of organic matter, such as living roots from the swamp vegetation, diluting the charcoal concentrations. Because of this living organic matter growing on the swamp and infiltrating into the sediments there may exist a lag between a fire and the actual deposition of charcoal into the sediments of Kings Waterhole. A detailed chronology (e.g. using 137Cs, 210Pb) could resolve the influx of charcoal during this European period. Previous studies in the Sydney

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 125

Basin have shown various changes to charcoal concentrations during the European period with some showing increases (e.g. Johnson, 2000; Mooney et al., 2001) and others showing a decrease (e.g. Kodela and Dodson, 1988).

s al p m cts ca fa e se c charco ba i f o locations er of art er ty vi umb Macroscop N Numb Acti 0

500

1000

1500

2000

2500

3000

3500 Age (yr BP)

4000

4500

5000

5500

6000

200 400 600 100 200 300 51010 20 30 mm2/cm3 Figure 6.4. A comparison of the Upper Mangrove archaeological data (Attenbrow, 2003; 2004) and the Kings Waterhole charcoal curve.

6.6 Conclusions

There were no major changes in the vegetation composition during the past ~6,100 yr from Kings Waterhole. The mixed Eucalyptus/ Casuarina woodland, with a sclerophyll shrubby understorey, which currently surrounds the swamp is believed to have existed for the entire analysed sequence. Kershaw’s (1995) and Pickett et al.’s (2004) have previously suggested that most of the characteristics of southeastern Australia’s vegetation were established by ~6 ka. Furthermore the resilient nature of the Sydney Sandstone flora (Dodson and Thom, 1992) may mean that it is relatively insensitive to minor climatic changes. The record did display a decreased representation of Casuarinaceae throughout the mid-late Holocene and

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 6 126

this decrease has been described from several studies in south-eastern Australia (e.g. Dodson 1974; Clark 1983; D’Costa et al. 1989; Ladd et al. 1992; Devoy et al. 1994; Harle 1998; Gale & Pisanu 2001).

Charcoal increased abruptly at 5.7 ka at KW and this is likely to be associated with the onset of ‘modern’ ENSO. Under an ENSO dominated climate it is suggested that more intense natural fires became a feature of the landscape. At 3 ka there was a major decrease in charcoal which corresponded with increased Aboriginal activity in the region. It is therefore likely that from 3 ka low intensity anthropogenic fires were more prevalent perhaps to protect or promote local food resources. There were very low levels of charcoal in the European period but this may be related to biotic influences or taphonomic processes of the site.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 127

Chapter 7: A Late Quaternary palaeoenvironmental investigation of the fire, climate and human nexus from the Sydney Basin, Australia

Abstract It is widely believed that Australian Aboriginals utilised fire to manage various landscapes however to what extent this use of fire impacted on Australia’s ecosystems remains uncertain. The late Pleistocene/Holocene fire history from three sites within the Sydney Basin, Gooches Swamp, Lake Baraba and Kings Waterhole were compared with archaeological and palaeoclimatic data. The Gooches Swamp record appeared to be most influenced by climate and there was an abrupt increase in fire activity from the mid- Holocene perhaps associated with the onset of modern El Niño dominated conditions. The Kings Waterhole site also displayed an abrupt increase at this time however there was a marked decrease in charcoal from ~3 ka. Similarly Lake Baraba displayed low levels of charcoal in the late Holocene. At both Kings Waterhole and Lake Baraba archaeological evidence suggests intensified human activity in the late Holocene during this period of lower and less variable charcoal. It is hence possible that Aboriginal people dominated fire activity in the late Holocene perhaps in response to the increased risk of large intense fires under an ENSO-dominated climate became more prevalent. The fire history of the Sydney Basin varies both temporally and spatially and therefore it is not possible to use a single type of fire regime as a management objective. There are also implications for fire activity in the future with climatic variability under an enhanced Greenhouse effect.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 128

7.1. Introduction

This study presents contiguous charcoal records from three sites, Gooches Swamp, Lake Baraba and Kings Waterhole which are all located within the Greater Blue Mountains World Heritage Area in south-eastern Australia. The aim of the study was to untangle the influence of climate and people on the late Quaternary fire activity of the region. It is assumed that the three sites have experienced a similar climatic forcing throughout this time and therefore any differences in fire activity may be attributed to either the different anthropogenic influences or local biotic or abiotic factors at the various sites. The individual fire and vegetation histories of the three sites have previously been discussed separately (i.e. Black and Mooney, in press; Black et al. submitted; Black and Mooney, in prep.) however this paper aims to compare the three records and to identify trends across the region.

Jones (1969) suggested that Aboriginal people used ‘fire-stick farming’ to increase or manipulate biotic resources. This has contributed to a popular notion that Aboriginal people controlled a fire regime consisting of frequent low intensity fire used within a well defined season but over small spatial scales (mosaic or patch burning). There remains much controversy over this generalisation and also whether this anthropogenic fire regime impacted on the species composition, structure and distribution of Australia’s vegetation (Bowman, 1998). There are those that believe Australian Aboriginals had minimal or no impact on vegetation (e.g. Horton, 1982) and others that have suggested that the use of fire has profoundly and irreversibly altered vegetation patterns (e.g. Singh et al., 1981; Flannery, 1994). Clark (1983) and MacPhail (1980) suggested that Aboriginal people may not have altered vegetation patterns but their use of fire was responsible for the continuation of vegetation zones and that they may have affected the rate of vegetation change. Head (1989) argued that Aboriginal people were trying to maintain a balance between the need to fire a landscape for resource manipulation and the need to protect other areas where particular plants grew such as rainforests. Head (1989) hence argued that this anthropogenic fire contributed to a mosaic of vegetation associations.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 129

Palaeoecological studies can provide information on past vegetation and fire activity and have resulted in a better understanding of the pre-history of fire (Wasson and Clark, 1987). In Australia there have been ambiguities associated with previous charcoal analyses, leading Bowman (1998) to suggest that palaeoecological studies do not objectively shed light on the issue of Aboriginal use of fire. Despite taphonomic and other concerns the analysis of charcoal “remains the best palaeoecological tool for reconstructing fire regimes on millennial time scales” (Hallett and Walker, 2000: 403).

The pre-history of fire is of relevance to the management of fire in the contemporary environment. As an example, the argument that Aboriginal people applied a low intensity/ high frequency fire regime to Australia’s vegetation is often used as a justification for hazard reduction burning (Gill, 1977). Gill (1977) has warned against the adoption of simplistic concepts of Aboriginal use of fire and generalisations over the entire continent. This is demonstrated, as described by Benson and Redpath (1997) that many vegetation types (e.g. rainforest, wet sclerophyll forest, alpine shrublands and inland chenopod shrublands) would now be rare if they had been burnt frequently by Aboriginal people.

The argument for the application of hazard reduction burning in national parks is often more strongly asserted following severe bushfire seasons (e.g. after the 1994 Sydney and the 2003 Canberra bushfires). Keith (1996) found that frequent widespread burning of the Sydney Sandstone flora could be responsible for local extinctions of plant species and that re-colonisation from unaffected sites occurs extremely slowly. The application of frequent, low intensity fire has been listed as a ‘key threatening process’ to biodiversity under the Threatened Species Conservation Act 1995 (NSW).

It is impossible to separate whether charcoal in sedimentary sequences originated from anthropogenic or natural fire events caused, for example, by lightning. However one possible method to resolve this is by comparing charcoal records with archaeological data and palaeoclimatic proxies in an attempt to separate the human and other influences. Bowman (1998), for example, has highlighted the importance of comparing palaeoecological and archaeological data to better understand fire history. Here the three

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 130

charcoal records, reflecting local fire activity, will be compared with each other and nearby archaeological information, used as an index of human activity through time.

The Holocene period, from ~11 500 cal. yr BP is generally described as a stable and relatively warm interglacial. Relatively stable sea levels, atmospheric CO2 concentrations, global ice cover and the Earth’s orbital parameters within 5% of present are thought to have contributed to a relatively stable period from the mid-Holocene (Dodson and Mooney, 2002). Kershaw et al. (2002) have identified the period of 7-5 ka as the Holocene precipitation peak (the Holocene Climatic Optimum), the period between 4-2 ka as a drier and perhaps cooler period, and the last 2 000 years as a return to wetter conditions in south- eastern Australia. Nonetheless there is increasing evidence that the Holocene climate contained some variability including abrupt changes (e.g. Bond and Lotti, 1995; deMenocal et al. 2000; Maasch et al., 2003).

The mid-Holocene has been identified as a period of climate change in Australasia (Bowler et al., 1977; Shulmeister, 1999) and further afield (Rodbell et al., 1999; deMenocal et al., 2000; Sandweiss et al., 1999). Lees (1992) linked climatic variability in northern Australia with the stabilisation of sea levels at about 5.5 ka. Steig (1999: 1485) described the mid- Holocene as “a period of particularly profound change” and Hodell et al. (2001) noted an abrupt cooling of sea-surface temperatures, expansion of sea ice and increased ice-rafted detritus accumulation in the Southern Ocean, between 5.5 to 5ka. deMenocal et al. (2000) identified large scale changes in the sub-tropical North African climate at 5.5 ka with the cessation of the ‘African Humid Period’ and linked to the Asian Monsoon..

Clement et al. (2000) suggested that ENSO variability was present throughout the Holocene but underwent a steady increase from the mid-Holocene to the present. Haug et al. (2001) suggested that there was a southward shift in the Inter-tropical Convergence Zone throughout the Holocene perhaps associated with precessional insolation changes. The resultant changes in insolation throughout the Holocene may be linked to the increased the prevalence of ENSO throughout the Holocene (Haug et al., 2001).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 131

Modern El Niño-Southern Oscillation (ENSO) events are believed to have occurred from the mid to late Holocene (e.g. McGlone et al., 1992; Shulmeister and Lees, 1995; Rodbell et al., 1999; Shulmeister, 1999; Sandweiss et al., 2001; Moy et al., 2002; Riedinger et al., 2002; Gagan et al., 2004). Shulmeister and Lees (1995) suggested that ENSO was initiated from ~4 ka. Shulmeister (1999: 81) identified a ‘flipping’ of climates in the Southern Hemisphere Pacific Basin from ‘Early Holocene mode’ to a ‘Late Holocene mode’ at ~ 5 ka and associated this with increased seasonality and the possible onset of modern ENSO. Rodbell et al. (1999) argued that ENSO progressively achieved modern characteristics by ~5 000 cal. yr BP although Moy et al. (2002) placed this earlier, at ~7 000 cal. yr BP. Sandweiss et al. (1996; 2001) also described ENSO events from ~5 800 yr BP, albeit at a low frequency until about 3 200 yr BP. Riedinger et al. (2002) also described an increase in the intensity and frequency of El Niño events from 3 100 cal. yr BP.

Climatic variability associated with ENSO has been linked with increased fire activity (e.g. Swetnam and Betancourt, 1990; Goldammer, 1999; Siegert et al., 2001; Kitzberger, 2002) and variability in charcoal abundance has been shown correspond with Holocene ENSO activity (e.g. Haberle et al., 2001; Haberle and Ledru, 2001; Kershaw et al., 2002).

Westerling and Swetnam (2003) suggested that the history of fire in the Western United States contained climatic signals similar to palaeo-fire reconstructions. Furthermore, Haberle and Ledru (2001) suggested that periods of rapid climate change or climatic variability leads to increased fire activity.

Haberle and Ledru (2001) identified an alignment in the fire activity from a number of sites in Indonesia and Papua New Guinea, and Central and South America from the mid- Holocene and attributed this to the intensification of ENSO variability. There is a relationship between ENSO and severe fire seasons in Australia, since El Niño events are generally associated with drier and warmer than average conditions, and La Niña events with conditions that are wetter and cooler than average (Lindsay et al., 2004). Edwards (2002) found a particularly strong link between the Southern Oscillation Index and fire season severity in southeastern Australia. Therefore it can be expected that fire activity in

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 132

the Sydney Basin should increase with the onset of ‘modern’ ENSO from the mid-late Holocene.

7.2. Site description and selection The study sites used in this study are located within the Sydney Basin, on the central east coast of New South Wales, Australia. The Sydney Basin is a geological province of approximately 3.7 million ha and consists of a number of discrete geomorphic units that include the Blue Mountains Plateau to the west, the Wollemi-Colo and Hawkesbury Plateau to the north, the Woronora Plateau and Southern Highlands to the south and the central Cumberland Plain (Herbert, 1983) (Figure 7.1).

The Sydney Basin is dominated by a temperate climate characterised by warm summers and no dry season, although a sub-humid climate can be found in the northern parts of the Basin and small areas of Montane climate are found in the Blue Mountains (NPWS, 2005). Average annual rainfall is variable across the Basin and resultant from altitudinal changes and the distance from coast. Shallow skeletal sands are found on the sandstone plateaus and these soils have a poor water-holding potential, are very acidic and infertile (Herbert, 1983). Elsewhere soils may be metres deep and enriched by silt and organic matter. The considerable variation in geology, soils, climate and topography has resulted in one of the most species diverse botanical divisions in Australia (NPWS, 2004) and includes communities of dry and wet sclerophyllous forests and woodlands, warm temperate rainforests, heath, mangroves and swamps (Benson, 1986; 1992).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 133

Figure 7.1. The major geomorphic units of the Sydney Basin and the location of the three sites. Adapted from Herbert (1983).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 134

The species composition and structural formation of the vegetation communities that occupy the sandstone plateaus of the Sydney Basin vary due to altitude, soil depth and rainfall (Benson, 1986; 1992). Heath and low open woodland are commonly found on the rocky platforms and ridgetops whereas taller open forests occur on the deeper plateau soils and on the slopes (Benson, 1986; 1992). The Sydney Sandstone complex vegetation communities that are found on the sandstone plateaus surrounding Sydney are dominated by Eucalyptus, Angophora, Corymbia and Syncarpia (Benson, 1986; 1992).

The study sites have been confined to swamp deposits on sandstone locations. These communities are found as isolated communities throughout the Sydney Basin and include swamps found on poorly drained Quaternary deposits and hanging swamps on the sandstone plateaus (Fairley and Moore, 2000). Floristic composition varies locally in relation to soil moisture gradients and the vegetation on these swamps can form monocultures of Common Reed or complex communities of Prickly-leaved Tea-tree (Melaleuca stypheloides) and Paperbark (Melaleuca quinquenervia) associations, with Swamp Mahogany (Eucalyptus robusta), Swamp Oak, Sedges, Tall Spike Rush (Elaeocharis sphacelata) and Juncus (Juncus sp.) (Keith and Benson 1988; Benson and Keith 1990).

The locations of the three sites are shown in Figure 7.1. Gooches Swamp (GS) (~33o27’S, 150o16’E, 960 m asl) is a narrow, elongated swamp in a low headwater valley and is located within the Blue Mountains World Heritage Area, in the western part of Sydney Basin. Kings Waterhole (KW) (33o1’S, 150o40’E, 280 m asl) is a low relief valley infilled with Quaternary sand and peat and is located within Wollemi National Park, northwest of Sydney. Lake Baraba (LB) (34o13’S, 150o13’E, 305 m asl), is an infilled lake within an entrenched meander and is found within Thirlmere Lakes National Park, in the southwest part of the Sydney Basin. More detailed descriptions of the three sites have been described previously (in Black and Mooney, in press; Black et al., in prep; Black and Mooney, in prep).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 135

Kings Waterhole and Lake Baraba share similar environmental settings. Average annual rainfall and temperatures, altitude and the swamp and surrounding vegetation communities are broadly similar. Gooches Swamp is located in the upper Blue Mountains at an elevation of 960 m and hence experiences a slightly different climatic regime to the other sites with greater variations in temperature throughout the year and a higher average annual rainfall. The vegetation community that grows at GS has recently been listed as a Endangered Ecological Community in the New South Wales’ Threatened Species Conservation Act 1995 (NPWS, 2004). Gooches Swamp is dominated by shrubs (e.g. Baeckea linifolia, Grevillea acanthifolia subsp. acanthifolia, Epacris paludosa and Leptospermum spp.) and sedges (e.g. Restio australis, Baloskion australe, Empodisma minus, Lepyrodia scariosa and Lepidosperma limicola) and occurs on a low slope headwater valley with impeded drainage (Keith and Benson, 1988). The surface vegetation at Lake Baraba and Kings Waterhole, however, is dominated by sedges and rushes (e.g. Lepironia articulata, Eleocharis sphacelata), Philydrum lanuginosum, Brasenia schrebi).

The three sites were inhabited by different Aboriginal language groups in pre-European times (Figure 7.2). The Newnes Plateau, on which Gooches Swamp is located, may have been a place of interaction or a transport corridor for various Aboriginal groups including the Dharug (or Daruk) and Gundungurra people (Stockton and Holland, 1974; Horton, 1994). It has been suggested that Aboriginal occupation of the Blue Mountains was sensitive to climatic variations due to the altitude, rugged topography and limited resources (Flood, 1980; Stockton and Holland, 1974). Bowdler (1981) suggested that there was sporadic occupation of the Blue Mountains between 14 000 and 12 000 yr BP followed by a hiatus and then an intensification of occupation from the mid-late Holocene. Stockton (1970) described human occupation in the upper Blue Mountains as ‘spasmodic’ with seasonal hunting trips during the milder periods of the Holocene. More recently Stockton (2005) suggested that there would have been enough resources,

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 136

Figure 7.2. The location of the territory of the major Aboriginal language Groups and the key archaeological sites in the Sydney Basin relevant to this study.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 137

particularly in the swamp areas to sustain Aboriginal throughout winter. Furthermore Stockton (2005) suggested that during times of aridity, associated with the glacial period and transition, the Blue Mountains would have been relatively moister than the regions of the east and west and may therefore have served as a haven for plant and animals, including human beings. However Hesse et al. (2003) suggested that sand dunes were active, with a lack of vegetation, in the upper Blue Mountains during the Last Glacial Maximum therefore making the area inhospitable at this time.

The region was also thought to be used for ceremonial practices and it is possible that it was also neutral or inter-tribal (Stockton and Holland, 1974). There is evidence of Aboriginal occupation on the main ridge and close-by spurs of the Blue Mountains, and down the slopes and into the shallow headvalleys with only deep gullies apparently void of occupation (Stockton and Holland, 1974). The earliest evidence of Aboriginal occupation in the Blue Mountains is from the Kings Tableland and this has been dated to ~22.4 ka (Stockton and Holland, 1974). The archaeological history of the Blue Mountains remains relatively poorly understood with the only limited sites of detailed study.

The territory of the Darginung (also written Darkinjung, Darkinjang and Darkinung) covered an area extending south from the Hunter River, including portions of the Macdonald and Colo Rivers and lower Hawkesbury River at Wisemans Ferry (Mathews, 1897, p1; Tindale, 1974, pp193; Dharug and Lower Hawkesbury Historical Society, 1988, p4; Attenbrow, 2003; 2004) and Kings Waterhole is situated within this territory.

The Darginung were believed to be relatively mobile hunter-gatherers, who used several base camps as well as many activity locations (e.g. sites for hunting and gathering, tool maintenance, etc) and transit or short-term camps within their country (Attenbrow, 2003; 2004).

The traditional custodians of Lake Baraba were the D’harawal and Gundangarra people. The lakes and wetlands of the Thirlmere Lakes were likely to represent a plentiful supply of food and ethnographic evidence suggests that the Aboriginal people of the region

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 138

frequently applied fire to the landscape (NPWS, 1995). Attenbrow (2004) has suggested that the establishment of Aboriginal sites in the region increased from ~8 ka with the habitation rates of these sites generally increasing until the arrival of European people.

7.3 Method

Sediment cores were extracted from Gooches Swamp, Lake Baraba and Kings Waterhole using a Russian d-section corer (Jowsey, 1966) and the analysed cores were 3.55 m, 6.35 m and 5.55 m in depth, respectively. The stratigraphy of the cores were described using a modified version of the Troels-Smith method (Kershaw, 1997) and were photographed.

The three sedimentary sequences were dated using radiocarbon dating and these dates were calibrated with CALIB v5 (Stuiver et al., 2005) using the IntCal04.14c (Reimer et al., 2004) and ShCal04.14c (McCormac et al., 2004) data sets.

Macroscopic charcoal, which best represents local or catchment fire events (Clark, 1988; Whitlock and Millspaugh, 1996), was analysed using a modified version of the ‘Oregon sieving method’ (Long et al., 1998) and image analysis (Mooney and Black, 2003). Volumetric sub-samples were taken from contiguous 1cm sections for the Gooches Swamp core, and at contiguous 2.5cm sections for the Lake Baraba and Kings Waterhole sedimentary sequences. The samples were placed in 8% sodium hypochlorite (bleach) for 24 hrs to remove the pigment from organic matter and, hence, aid in the identification of charcoal. This was then washed through a 250 µm sieve and the collected material was photographed in a petrii dish using a digital camera (Nikon Coolpix 4 mega-pixel). The area of charcoal was calculated using image analysis software (Scion Image Beta 4.02 for Windows). Charcoal is expressed as an area in mm2/cm3. The sedimentation rates at Gooches Swamp, King Waterhole and the peat section of Lake Baraba were near linear and hence were not re-expressed as an influx of charcoal.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 139

The charcoal curves of the three sites were statistically analysed using psimpoll v 4.25 (Bennett, 1998; 2005). The skewness and kurtosis were tested and a Runs test was used to assess whether the data was showing a trend or just random or normally distributed values with q values > 0.05 indicating random data. Three types of correlations (Linear, Spearman’s and Kendall’s) were carried out to test whether there was increasing charcoal with younger age. It has been suggested that a negative correlation (i.e. increased charcoal with younger age) could reflect anthropogenic or taphonomic influences. Fourier transformation was used in order to detect peaks in spectral density, indicative of cyclical changes in the charcoal data and to identify any cycles that could be associated with known climatic cycles,

The three charcoal records were compared to local archaeological sequences. The Gooches Crater record was compared to the nearby (i.e. ~35 km away) Capertee 3 archaeological sequence (McCarthy, 1964; Hiscock and Attenbrow, 1998: 2004; Attenbrow, 2004: 2005), the Kings Waterhole record was compared to the nearby (~30 km away) Upper Mangrove Creek archaeological record (Attenbrow, 2003; 2004) and the Lake Baraba record will be compared to regional archaeological from the New South Wales South Coast and hinterland summarised by Attenbrow (2004).

The 200 yr averaged charcoal values for each of the three sites were compared to the changes in ENSO activity throughout the Holocene (e.g. Moy et al., 2002), seasonality and also a general summary of the past climates of southeastern Australia (e.g. Kershaw et al., 2002). Seasonality was calculated by calculating the difference between January and July insolation levels at 30oS throughout the Holocene using the data from Berger (1992).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 140

7.4 Results

7.4.1 Stratigraphy and chronology The results of the radiocarbon dating for Gooches Swamp (GS), Lake Baraba (LB) and Kings Waterhole (KW) are given in Table 7.1, and the basal dates of the three sedimentary sequences at 14.2, >43 ka and 6.1 ka, respectively. The sediments of GS and KW both displayed a relatively constant rate of accumulation and were composed of humified peat interspersed with clay, charcoal and sand. At Lake Baraba, peat was found above 172 cm, a transition layer of peat and clay from 172 – 410 cm, which became more clayey with depth, and clay below 410 cm.

Table 7.1. Radiocarbon dates and calibration for Gooches Swamp, Lake Baraba and Kings Waterhole sediments. Calibration results from CALIB v5 (Stuiver and Reimer, 2004). The mid-point of the entire calibrated year range is used in age-depth model calculations.

Gooches Swamp 14C date BP with 1σ Cal. yrs BP* (2 σ error) Lab code Sample depth (cm) error 48-53 1 760 ± 60 1419 –– 1811 β-169992 80-90 2 450 ± 60 2333– 2708 β-192605 150-156 4 950 ± 130 5322 –5912 β-169993 295-307 10 360 ± 140 11646 –– 12737 β-169994 Lake Baraba 14C date BP with 1σ Cal. yrs BP* (2 σ error) Lab code Sample depth (cm) error 147 – 153 4 130 ± 70 4421 – 4821 β-186144 275 – 285 5 950 ± 60 6549 – 6887 β-192607 347 – 353 6 750 ± 80 7433 – 7675 β-186145 464 – 472 19 411 ± 196 22541 – 23716 NZA-192608 595 – 601 > 43 630 N/A β-192608 Kings Waterhole 14C date BP with 1σ Cal. yrs BP* (2 σ error) Lab code Sample depth (cm) error 147 – 154 2220 ± 60 2003 – 2327 β -186146 348 – 353 3280 ± 70 3269 – 3635 β -186147 547 – 553 5560 ± 90 6014 – 6491 β -186148

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 141

The deposition rate of the clays at Lake Baraba is much lower (i.e. 0.04 mm/yr) than within the peat (~0.67 mm/yr). Based on the chronology the sample resolution for GS is ~40 yr, LB is ~77 yr (between 0 and 410 cm) and ~630 yr (>410 cm) and KW is ~27 yr.

7.4.2 Charcoal analysis The charcoal records of the three sites are presented in Figure 7.3. For the purpose of comparison the charcoal record of LB has been truncated at 14.2 ka (430 cm) i.e. the basal date of GS. At GS macroscopic charcoal was relatively high between 250-281 cm (~9.8 – 11.1 ka), 232-244 cm (~9.1 – 9.6 ka), 97–150 cm (~3.5 – 5.7 ka), 67-87 cm (~2.3 – 3.1 ka) and 0–6 cm (the late European period). There were low levels of charcoal between 325-353 cm (~12.9 -14 ka), 287-315 cm (~11.3 – 12.5 ka), 150-232 cm (~5.7 – 9 ka), 87-97 cm (~3.1 – 3.5 ka), and from 40 to 25 cm (~1.1 – 0.5 ka) and 6-13 cm (early European occupation).

Figure 7.3. The results of the macroscopic charcoal analysis for Gooches Swamp, Lake Baraba and Kings Waterhole.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 142

Most of the analysed sediment profile from Lake Baraba had low levels of charcoal, with peaks at 240 (~6 ka), 270 (~6.4 ka) and 400 cm (~8.4 ka), a series of higher peaks at 230 – 200 cm (~5.9 – 5.4 ka) and minor peaks at 170 – 130 cm (~5 – 4 ka). Macroscopic charcoal concentrations were very low between 430 and 405 cm (~13.6 – 8.5 ka), relatively low but variable between 405 and 275 cm (~8.5 – 6.5 ka) but then increases abruptly and remains very high and variable until 200 cm (~5.4 ka). Between 185 and 55 cm (~5.2 – 1.7 ka) there is a decreasing trend in charcoal concentrations, from being relatively high to moderately low, but it remains variable throughout this interval. The upper samples (50 to 0 cm, past ~1.5 ka) have very low charcoal concentrations with some samples almost void of charcoal. There was much less charcoal found in the clays when compared to the peat sediment.

Macroscopic charcoal was present throughout the entire sedimentary sequence of Kings Waterhole although the concentrations of charcoal did vary considerably. Charcoal was initially low between 555 and 515 cm (~6.1 and 5.7 ka) but then increased dramatically to remain high and variable between 515 and 265 cm (~5.7 ka and 3 ka) with the exception of a period of reduced charcoal deposition between 470 and 430 cm (~5.2 and 4.8 ka). Charcoal decreased rapidly at 265 cm (~3 ka) and remained generally low from this time to the present. Charcoal deposition for the top 80 cm (the past ~1000 yr) was especially low.

7.4.3 Statistical analysis The results of the Skewness, Kurtosis and Runs test are presented in Table 7.2. The Skewness and Kurtosis results of all three sites are positive and significant and this indicates a common, non-normal shape or distribution. The positive Kurtosis results reflect the ‘peaky’ nature of all three records. The q-values at all three sites are < 0.05 and hence the data is not random but is showing some kind of trend.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 143

Table 7.2. Statistical analysis results for the three sites.

Site Mean Standard Deviation Skewness Kurtosis Runs (q) Gooches Swamp 119.3582 144.6212 2.1352* 5.5980* 9.572E-03 Lake Baraba 46.9872 79.7630 2.5422* 7.0829* 1.96E-11 Kings Waterhole 44.6821 66.6278 2.6908* 9.4131* 1.03E-11 * = significant

The three correlation co-efficients (i.e. Linear, Spearman’s and Kendall’s) for the sites are given in Table 7.3. GS and LB display a negative value for all three correlation co- efficients and this suggests that there are increasing levels of charcoal with younger age. GS is the only site that shows a significant trend.

Table 7.3. Results of the correlation for the three sites.

Site Linear Correlation Spearman’s Correlation Kendall’s Correlation Gooches Swamp -0.3442* -0.4384* -0.294* Lake Baraba -0.0654 -0.0283 -0.0159 Kings Waterhole 0.3820 0.5914 0.3915 * = significant

A number of significant cycles were identified from the spectral analysis however caution needs to be taken with interpreting the significance of cycles identified in this analysis as sample resolution and record length can influence the outcome (Figure 7.4). At GS 41 yr, 1 300 yr, 2 600 yr and 5 900 yr cycles were significant, at LB a 73 yr, 3 700 yr and 8 000 yr cycle were identified and Kings Waterhole had a 54 yr and a 6 100 yr significant cycle. The 41, 54 and 71 yr cycles found at GS, KW and LB, respectively, are associated with the sampling resolution of the records and hence are not important. Similarly the 8 000 and the 6 100 yr cycles from LB and KW are a function of the record length and hence also should be ignored. This means the only significant cycles are the 3 700 yr for Lake Baraba, and the 1 300 yr and 2 600 yr cycles from Gooches Swamp (although the 2 600 yr cycle at GS is likely to be linked to the 1 300 yr cycle).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 144

Figure 7.4. The results of the spectral analysis (Fourier’s analysis). Results are significant at the 0.05 level.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 145

7.4.4 Comparison to archaeological records

The comparisons of the GS charcoal record with the Capertee 3 archaeological sequence, the LB record with the summary of the southern Sydney Basin archaeological data, and the KW with the Upper Mangrove Creek archaeological sequence, are shown in Figures 7.5, 7.6 and 7.7, respectively.

al cts rco a s efa t rt ch fac ic e ed a op sc ack ro b ac on- M Backed art N 0

1000

2000

3000

4000

5000

6000

7000

8000 Age (yr BP)

9000

10000

11000

12000

13000

14000 250 500 750 1000 100 200 300 50 100 150 mm2/cm3 # per 1000 years # per 1000 years Figure 7.5. A comparison of the Gooches Swamp macroscopic charcoal curve with the number of backed and non-backed artefacts from Capertee 3 (Hiscock and Attenbrow, 1998; 2004).

At Gooches Swamp discard rates and average charcoal concentrations were both relatively low between ~9.7 and 6 ka however average charcoal concentrations were high between ~6 and 3.6 ka whilst discard rates remained low. Discard rates were highest between ~3.6 and 1.7 ka with the layer ~3-2.3 ka having the highest discard rates. The latter layer corresponds

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 146

with the very high levels of charcoal. The period ~1.7 ka to present has very low artefact discard rates corresponding with relatively high levels of charcoal accumulation.

t en m lish l b esta rcoa d use ation bit tions a of h Habita ate Macroscopic cha # R 0

2000

4000

6000

8000

10000

12000

Age (yr BP) (yr Age 14000

16000

18000

20000

22000

24000 100 200 300 20 40 60 50 100 150 200 mm2/cm3 # per 1000 years # per 1000 years

Figure 7.6. A comparison of the Lake Baraba charcoal curve with a summary of the archaeological data from the southern Sydney Basin compiled by Attenbrow (2004).

Charcoal at Lake Baraba and the number of habitations used in the region both gradually increase from the LGM to the early Holocene (Figure 7.6). From ~7 to 8 ka this relationship breaks down, with the number of habitations increasing more rapidly whilst charcoal content declines steadily. Very high levels of habitation use and establishment in the past 3 – 4 000 yr corresponds with very low levels of macroscopic charcoal.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 147

l oa s ct rc camps ha fa e ons c rte bas ic f a f op o sc er er o ty locati ro b i mb ac um u M N N Activ 0

500

1000

1500

2000

2500

3000

3500 Age (yrBP)

4000

4500

5000

5500

6000

200 400 600 100 200 300 51010 20 30 mm2/cm3 # per 1000 years # per 1000 years # per 1000 years

Figure 7.7. A comparison of the Kings Waterhole charcoal curve with the archaeological data from the Upper Mangrove Creek catchment (Attenbrow, 2004).

Macroscopic charcoal levels at Kings Waterhole are relatively high and variable between 6.1 and 3 ka and this corresponds with low numbers of artefacts, base camps and activity locations (Figure 7.7). At 3 ka there is a dramatic increase in the number of artefacts and base camps and this is associated with a marked drop in charcoal. Charcoal remains very low to the present whilst the three indices of the archaeological data remain elevated.

7.4.5 Comparison with climatic data The three charcoal records were compared to the frequency of ENSO events (Moy et al., 2002), seasonality based on changes in insolation associated with Milankovitch mechanism and a climatic summary of south-eastern Australia (Figure 7.8).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 148

Figure 7.8 a) Seasonality at 30oS based on the difference in insolation between summer and winter (Berger, 1992) b) The frequency of ENSO events per 100 yrs based on Moy et al. (2002) c) The climatic summary of south-eastern Australia (Kershaw et al., 2002; Lees, 1995; Shulmeister, 1999 etc) d) The smoothed Gooches Swamp charcoal record constructed by summing the 200 yr values e) The smoothed Kings Waterhole charcoal record constructed by summing the 200 yr values f) The smoothed Lake Baraba constructed by summing the 200 yr values.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 149

The GS charcoal record displayed increases in charcoal with the onset of climatic amelioration and the Pleistocene/Holocene transition. There were very low levels of charcoal during the Holocene Climatic Optimum (~9-6 ka) and there was a dramatic increase in charcoal from the mid-Holocene when ENSO frequencies increased. Charcoal remained high and variable from this time to present under an ENSO dominated climate. The seasonality increased throughout the Holocene and began to plateau from the mid- Holocene. The GS record follows this general trend although the changes at GS are much more abrupt.

KW shows an increase in charcoal associated with the onset of an ENSO dominated climate however charcoal decreases dramatically at 3 ka whilst ENSO events remained relatively frequent. Similarly there is no obvious relationship with the degree of seasonality and the KW curve.

The increase in charcoal from ~8 ka at LB may be associated with the slight increase in ENSO frequency suggested by Moy et al. (2002). There was a large decrease in charcoal from the mid-Holocene at LB and this occurs at the same time that ENSO began to really dominate southeastern Australia’s climate. There is no clear association with the seasonality record and the LB charcoal curve.

7.5 Discussion

There are major differences in the three Sydney Basin fire records presented in this study and this suggests that climate, anthropogenic influences and the depositional history had varying influences throughout time. However there are also a number of generalisations that can be made across the region.

The fire history of Gooches Swamp appears to be most greatly influenced by climate. The Gooches Swamp record displayed an increase in fire activity associated with the late

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 150

glacial/ Holocene transition followed by a decrease associated with the relatively stable Holocene Climatic Optimum. When the frequency of ENSO events provided by Moy et al. (2002) is compared to the GS charcoal record there is not a very strong relationship however it should be noted that the data provided by Moy et al. (2002) is but one site of several (e.g. Shulmeister, 1999; Rodbell et al., 1999; Clement et al., 2000; Riedinger et al., 2002) and as yet there is no consensus of the actual frequency of ENSO throughout the Holocene. However it is generally agreed upon that from the mid-Holocene (~5 ka) there was a shift to an ENSO dominated climate and this has been included in the climatic summary of southeastern Australia (Figure 7.8). There is a dramatic change in fire activity at Gooches Swamp from the mid-Holocene suggesting that this change was caused by the shift to an ENSO dominated climate. Furthermore from this time more intense fires were a feature of the landscape. The large, intense fires that occur approximately every decade in the Blue Mountains during modern times (see Cunningham, 1984) are likely to have occurred from the mid-Holocene based on the similarities in the charcoal curve from the European period and the preceding ~5 000 yr.

Haberle et al. (2001) constructed a regional cumulative charcoal curve, based on ten sites throughout Indonesia and Papua New Guinea, and when this is compared to the Gooches Swamp charcoal record there are strong similarities (Figure 7.9). The GS charcoal record was smoothed using a 41 point running average. Haberle et al. (2001) linked the changes in fire activity in this region with the onset of climatic variability during the post-glacial transition and with the onset of modern ENSO from the mid-late Holocene. The close association between these records suggests that the fire activity in the catchment of Gooches Swamp may be responding to climatic controls and charcoal record displayed similarities to the seasonality curve, with an increase in charcoal occurring when seasonality increased, although the changes at GS were much more abrupt. Here fire activity may undergo a non-linear change in response to the linear forcing in a manner similar to that proposed for the Holocene African climate by deMenocal et al. (2000).

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 151

300 400 GCR charcoal curve* Haberle et al. 2001 curve

350 250

300 ) 3 200 /cm 2 250

150 200

150 charcoal cumulative corrected . (2001) 100 et al et Gooches Crater charcoal (mm charcoal Crater Gooches

100 Haberle Haberle

50 50

0 0 250 1500 2750 4000 5250 6500 7750 9000 10250 11500 Age (cal. yr BP)

Figure 7.9. A comparison of the composite charcoal results from several sites in tropical Sahul, from Haberle et al. (2001), and the charcoal curve from Gooches Swamp. The Gooches Swamp charcoal curve has been smoothed using a 41 point running average.

Turney et al. (2004) identified a 1 490 yr cycle based on spectral analysis of peat humification data from Lynchs Crater in northern Queensland. They interpreted this cycle as being associated with changes in precipitation associated with long-term changes in ENSO and also correlated with Dansgaard-Oeschger events in the North Atlantic. Similar cyclicity of these palaeo-ENSO cycles have also been identified in southern Ecuador (Moy et al., 2002) and North America (Wang et al., 2000). The 1 300 yr cycle identified in the Gooches Swamp charcoal record may possibly be related to these palaeo-ENSO cycles further suggesting the climatic control of fire activity at Gooches Swamp.

Based on the comparison of the Capertee 3 archaeological sequence and the GS charcoal curve there does not appear to be any clear association between human activity and the fire history at GS. However, as previously discussed, the archaeological history of the Blue Mountains is not well understood and it is possible that future archaeological investigations

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 152

may reveal a stronger relationship. As an example archaeological evidence from the Sydney Region suggests that there were increases in Aboriginal activity from the mid-late Holocene (Attenbrow, 2004). However with the current evidence at hand it appears that the GS fire history is responding predominately to climatic controls.

The variability in charcoal at GS since the mid-Holocene may also be related to changes in the biotic and abiotic dynamics of the swamp. It is plausible that as the sediment has built up through time the aerial vegetation on the swamp surface has burnt more frequently. This scenario, however, is not supported by the palynology at the site (Black and Mooney, in press) which demonstrated little vegetation change during the Holocene and thereby does not support the development of the wet-heath vegetation from this time. The water-table is likely to have risen concurrently with the accumulation of sediment since the sediment is accumulating as a result of a rock-fall dammed constriction.

Charcoal at Lake Baraba was very low during the late Pleistocene/early Holocene and increased dramatically at ~8.5 ka. During this time the number of habitations used in the region by Aboriginal people was also increasing and hence the changes in fire activity could potentially be attributed to humans. Alternatively the increase in charcoal may be related to increases in biomass accompanying climatic amelioration.

Abiotic influences are likely to have affected the charcoal record at Lake Baraba since there is a change in the depositional environment as the site went from a lake to a swamp at ~8.5 ka. The change meant that fire may have impacted the vegetation of the swamp surface occasionally, thereby increasing charcoal delivery to the sediments and explaining why more charcoal was found in the peat sediments compared to the lacustrine clays of Lake Baraba. There is no clear relationship between ENSO frequency and the fire history of Lake Baraba. There is a marked decrease in charcoal over the past 3 000 yr and during this time the rate of established and number of Aboriginal sites used in the region increased dramatically. The changed fire regime at 3 ka may therefore be caused by Aboriginal people.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 153

Charcoal increased abruptly at 5.7 ka at KW and this is coeval with the increase at Gooches Swamp and hence is also likely to be associated with the onset of ‘modern’ ENSO. Charcoal remains high and variable from this time until 3 ka and this probably reflects intense fires under an ENSO dominated climate. At 3 ka charcoal decreased dramatically and remained low for the remainder of the record, a trend that is also detected at Lake Baraba. This marked change is temporally associated with changes in the archaeological record of the Upper Mangrove Creek catchment and hence fire activity is likely to be associated with human activity. There is no evidence of any major biotic changes, indicated by the palynology (Black and Mooney, submitted, or see Chapter 6), or any depositional changes and hence this is unlikely to have impacted on the charcoal record at Kings Waterhole.

The decreased level of charcoal in the late Holocene at both Lake Baraba and Kings Waterhole are hence thought to indicate an anthropogenic change to the fire regime. As to why less charcoal was deposited at these sites under intensified human activity is perhaps best related to a frequent low intensity fire regime associated with the management of natural resources (e.g. Nicholson, 1981; Gott, 2005). Regular low intensity fires are likely to consume less biomass and hence the production, and deposition, of charcoal would be much lower especially if fires within the catchment were relatively small. Furthermore large intense fires are more likely to encroach onto the swamp surface and deposit charcoal in situ. In this study it is suggested that low levels of charcoal are interpreted as frequent but low intensity fire.

Previous studies have attributed changes in charcoal to the intensity of fire. Singh et al. (1981: 43) argued that the “greater amounts of charcoal particles and the large fluctuations… may be consistent with a pattern of intermittent high intensity fires with considerable accumulation of litter between them” from Lashmars Lagoon. They also suggested that relatively small amounts of charcoal with lower variability may reflect more frequent but less intense fires but with less accumulation of fuel. Hope (1994) also interpreted increases in charcoal as reflecting a change in fire regime from frequent burning

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 154

to periodic, destructive and intense fires ignited by lightning or the occasional human visitation.

This suggests that during the late Holocene at Lake Baraba and Kings Waterhole Aboriginal people may have intensively managed the landscape. The data also implies that the swamp surface was not burnt which is somewhat at odds with previous general descriptions (e.g. Gott, 2005). Preventing the burning of the swamp surface may have protected animal food sources, such as turtles and yabbies, which survive within these swamps. Perhaps the risk of more intense fires under an ENSO dominated climate meant that there was an increased need for Aboriginal people to manage in some landscapes (e.g. Lake Baraba and Kings Waterhole) to protect food sources.

7.6 Conclusion

The three fire records from the Sydney Basin displayed some major differences. In south- eastern Australia pre-European fire activity is popularly associated with Aboriginal people, however at Gooches Swamp climate appears to be the dominant control of fire activity over the last ~14 200 years. The Gooches Swamp record was most greatly influenced by climate with periods of climatic instability, such late glacial transition and the onset of ENSO dominated climates from the mid-Holocene, associated with higher levels of charcoal. The elevated levels of charcoal are likely to be due to an increased intensity and frequency of fire. It appears that the GS record is responding to a regional climatic signal resulting in strong similarities with Haberle et al.’s (2001) regional charcoal curve from the Sahul region. At Gooches Swamp anthropogenic influences are not clear based on the current archaeological data from the Blue Mountains.

At Kings Waterhole an increase in intense fires occurred at the same time as Gooches Swamp, also likely reflecting ENSO-related climatic variability. However at Kings Waterhole charcoal levels decreased dramatically at the same time Aboriginal activity intensified in the region. This decrease was also found at Lake Baraba and an altered

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 7 155

regime to small scale less intense but more frequent fires caused by anthropogenic activity is suggested as a cause. Although humans are likely to have manipulated fire, it is probable that this occurred within a framework dictated by the prevailing climate of the time. For example the risk of more intense fires under a strengthened ENSO dominated climate may have increased the need for Aboriginal people to tightly manage fire especially in the fire prone Sydney Sandstone vegetation communities. This potentially describes a complex nexus between climate and humans resulting in significant variability in fire through time. The longer term charcoal record of Lake Baraba was complicated since there was a change in the depositional environment from a lake to a swamp at ~8.5 ka and this had taphonomic implications.

The spatial and temporal variations in fire activity within the small geographical region of the Sydney Basin mean that there is not a single pre-European fire regime that can be presented as a management objective. The conclusions of this study mirror those of Head (1989, p. 41) who noted that there is a common assumption that Aborigines “had a single ongoing impact”: this erroneous assumption ignores climatic change and population and cultural change. Furthermore, the predominance of climate as a control on past fire activity has been described by Kohen (1996) and Kershaw et al. (2002). Kohen (1996; p. 20-21 ) concluded “perhaps what we are observing in the last few thousand years is the struggle between anthropogenic fires and climate, with climate seeming to come out in front in most situations”. Similarly Kershaw et al. (2002; p. 19) observed that “(the) relative importance of climate and human influence is difficult to assess but evidence… suggests that climate was the major driving force.”

This study also highlights the important influences of climate change and ENSO on fire activity. At GS fire was not necessarily linked with a drier climate and there was an apparent increase in fire activity during periods of climate change. The longer temporal perspective afforded by this study demonstrates that fire appears to be related to ENSO through time however the nature of this relationship requires further study. With projected rapid anthropogenic climate change in our near future, fire in this landscape is likely to be of considerable concern.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 8 156

Chapter 8: Summary of Conclusions

8.1. New Method for analysing charcoal From this study a novel methodology for analysing macroscopic charcoal has been published in Quaternary Australasia. The new method utilises digital photography and image analysis software to calculate the area of macroscopic charcoal per sample. This method is much faster than the traditional optical methods and hence it enables higher sampling resolutions. The method also utilises readily available equipment and free software. It was found that there were strong correlations between this new method and the traditional method of optically quantifying macroscopic charcoal. It is hoped that this method will become the standard method for quantifying charcoal in Australia allowing greater inter-site comparisons in future studies.

8.2. The Late Quaternary vegetation history of the Sydney Basin The three sites studied here are all found within the Greater Blue Mountains World Heritage Area and were surrounded by Sydney Sandstone Complex vegetation. There were no major changes in the composition of the flora at all sites throughout late Pleistocene/Holocene although there were some changes in the relative abundance of different taxa. It is suggested that the Sydney Sandstone flora is relatively resistant to environmental changes. This may be because the sclerophyllous nature of the vegetation means that it has mechanisms allowing it to cope with drought and climatic change. However subtle changes with pollen groups may be not be able to be detected because of the level of identification.

At Lake Baraba the vegetation record extended to >43 ka, most probably to ~60 ka thereby incorporating several climatic oscillations including the Last Glacial Maximum (LGM). During the LGM Casuarinaceae was present at the site and therefore Lake Baraba may have acted as a potential refugium for more mesic communities. There was a decrease in Casuarinaceae and increase in Eucalyptus from the early Holocene and this is attributed to

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 8 157

localised hydrological changes. This shift in vegetation has been described from a number of sites in south-eastern Australia and was also apparent from the mid-Holocene at Kings Waterhole.

A number of wet and dry periods have been identified based on the palynology of the sites. At Lake Baraba it appears that the lake dried out between 5.1 and 6 ka. At Kings Waterhole it appears that there was a slight drying from 2 ka to present. At Gooches Swamp there was an increase in fern representation between 5.5 and 3.5 ka and this may either suggest a wetter climate or may be associated with fire activity at the site.

8.3. The Late Quaternary fire history of the Sydney Basin The three fire records from the Sydney Basin displayed some major differences. In south- eastern Australia pre-European fire activity is popularly associated with Aboriginal people, however at Gooches Swamp climate appears to be the dominant control of fire activity over the last ~14 200 years. The Gooches Swamp record was influenced by climate to a greater extent than the other studied sites, with periods of climatic instability, such late glacial transition and the onset of ENSO dominated climates from the mid-Holocene, associated with higher levels of charcoal. The elevated levels of charcoal are likely to be due to an increased intensity and frequency of fire. It appears that the Gooches Swamp record is responding to a regional climatic signal resulting in strong similarities with Haberle et al.’s (2001) regional charcoal curve from the Sahul region. At Gooches Swamp anthropogenic influences are not obvious based on the current archaeological data from the Blue Mountains.

At Kings Waterhole an increase in intense fires occurred at the same time as Gooches Swamp, also likely reflecting ENSO-related climatic variability. However at Kings Waterhole charcoal levels decreased dramatically at the same time Aboriginal activity intensified in the region. This decrease was also found at Lake Baraba and an altered regime to small scale (mosaic or patch), less intense but more frequent fires caused by

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black Chapter 8 158

anthropogenic activity is suggested as a cause. Although humans are likely to have manipulated fire, it is probable that this occurred within a framework dictated by the prevailing climate of the time. For example the risk of intense hazardous fires under a strengthened ENSO dominated climate may have increased the need for Aboriginal people to tightly manage fire especially in these fire prone Sydney Sandstone vegetation communities. This potentially describes a complex nexus between climate and humans resulting in significant variability in fire through time. The longer term charcoal record of Lake Baraba was complicated since there was a change in the depositional environment from a lake to a swamp at ~8.5 ka and this had taphonomic implications.

The spatial and temporal variations in fire activity within the small geographical region of the Sydney Basin mean that there is not a single pre-European fire regime that can be presented as a management objective. The conclusions of this study mirror those of Head (1989, p. 41) who noted that there is a common assumption that Aborigines “had a single ongoing impact”: this erroneous assumption ignores climatic change and population and cultural change. Furthermore, the predominance of climate as a control on past fire activity has been described by Kohen (1996) and Kershaw et al. (2002). Kohen (1996; p. 20-21 ) concluded “perhaps what we are observing in the last few thousand years is the struggle between anthropogenic fires and climate, with climate seeming to come out in front in most situations”. Similarly Kershaw et al. (2002; p. 19) observed that “(the) relative importance of climate and human influence is difficult to assess but evidence… suggests that climate was the major driving force.”

This study also highlights the important influences of climate change and ENSO on fire activity. At GS fire was not necessarily linked with a drier climate but there was an apparent increase in fire activity during periods of climate change. The longer temporal perspective afforded by this study demonstrates that fire appears to be related to ENSO through time however the nature of this relationship requires further study. With projected rapid anthropogenic climate change in our near future, fire in this landscape is likely to be of considerable concern.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 159

References

Aaby, B. & Tauber, H. (1975). Rates of peat formation in relation to degree of humification and local environment, as shown by studies of a raised bog in Denmark. Boreas 4, 1-14.

Aaby, B.(1986).Palaeoecological studies of mires in Handbook of Holocene Palaeoecology and Palaeohydrology (B.E. Berglund Ed.) John Wiley and Sons Ltd.

Adamson, D.A. & Fox, M.D. (1982).‘Changes in Australasian vegetation since European settlement’, A history of Australasian vegetation, (J.M.B Smith, Ed.), pp.109-139. McGraw-Hill, Sydney.

Aitken, D. & Kershaw, A.P. (1993). Holocene vegetation and history of Cranbourne Botanic Garden. Proceedings of the Royal Society of Victoria 105, 67 – 80.

Allan, R. & Lindsay, J. (1998). Past climates of Australasia. In Climates of the southern continents ( J. Hobbs, J.A. Lindesay and H.A. Bridgman Eds.) John Wiley and Sons Ltd.

Allen, J.A. & Holdaway, S. (1995). The contamination of Pleistocene radiocarbon determinations in Australia. Antiquity 69,101 – 112.

Alley, R.B., Meese, D.A., Shuman, C.A., Gow, A.J., Taylor, K.C., Grootes, P.M., White, J.W.C., Ram, M., Waddington, E.D., Mayewski, P.A. & Zielinski, G.A., (1993). Abrupt increase in Greenland snow accumulation at the end of the Younger Dryas event. Nature 362, 527-529.

Anderson, A., (1994). Comment on J.Peter White’s paper “Site 820 and the evidence for early occupation in Australia Quaternary Australasia 12(2), 30 – 31.

Andres, M.S., Bernasconi, S.M., McKenzie, J.A. & Rohl, U., (2003). Southern Ocean deglacial record supports global Younger Dryas Earth and Planetary Science Letters 216, 515-524.

Attenbrow V.J. (1982) The Archaeology of Upper Mangrove Creek Catchment - Research in Progress. Coastal Archaeology of Eastern Australia (S. Bowdler, S. Ed.), pp. 63-78. Department of Prehistory, Research School of Pacific Studies, Australian National University, Canberra.

Attenbrow, V. (2002). Sydney’s Aboriginal Past: investigating the archaeological and historic records. UNSW Press, Sydney.

Attenbrow, V. (2003). Habitation and land use patterns in the Upper Mangrove Creek catchment, New South Wales central coast, Australia. Australian Archaeology 57, 20-31.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 160

Attenbrow, V. (2004). What's changing : population size or land use patterns? : the archaeology of Upper Mangrove Creek, Sydney Basin Pandanus Books, Canberra.

Baker, M. (1997). Fire and the pre-European cultural landscapes of the Hawkesbury- Nepean catchment. Science and Technology in the environmental management of the Hawkesbury-Nepean catchment 14 (Geographic Society of NSW).

Banks, J.C.G. (1989). ‘A history of forest fire in the Australian Alps’. In: The Scientific Significance of the Australian Alps. (R.B. Good, Ed.), pp265-280. Australian Alps National Parks Liaison Committee.

Bannerman, S.M. & Hazelton, P.A. (1990). Soil landscapes of the Penrith 1:100000 map sheet Soil Conservation Service of NSW, Sydney.

Barrows, T.T., Stone, J.O., Fifield, L.K. & Cresswell, R.G. (2002). The timing of the Last Glacial Maximum in Australia Quaternary Science Reviews 21, 159 – 173.

Beaton, J.M. (1983) Does intensification account for changes in the Australian Holocene archaeological record? Archaeology in Oceania, 18(2), 94-97.

Bengtsson, L. & Enell, M. (1986). Chemical analysis. In: Handbook of Holocene Palaeoecology and Palaeohydrology. (B.E. Berglund, Ed.).Wiley, Chichester.

Bennett, K.D. (2005). Documentation for psimpoll 4.25 and psccomb 1.03: C programs for plotting and analysing pollen data. Quaternary Geology, Department of Earth Sciences, Uppsala Universitet Villavġen, Sweden.

Benson, J.S. (1986). The vegetation of the Gosford and Lake Macquarie 1:100000 vegetation map sheet Cunninghamia 1(1), 35-57.

Benson, J.S. (1991) The effect of 200 years of European settlement on the vegetation and flora of New South Wales. Cunninghamia 2(3),343-501.

Benson, J.S. (1992). The natural vegetation of the Penrith 1:100000 map sheet Cunninghamia 2(4), 541 – 596.

Benson, D.H. & Keith, D.A. (1990). The natural vegetation of the Wallerawang 1:100,000 map sheet. Cunninghamia, 2(2), 305-335.

Benson, D.H. & Howell, J. (1994). The natural vegetation of Sydney. Cunninghamia 3(4), 677 – 1004.

Benson, J.S. & Redpath, P.A. (1997). The nature of pre-European native vegetation in south-eastern Australia: a critique of Ryan, D.G., Ryan, J.R. and Starr, B.J. (1995) “The

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 161

Australian landscape – observations of explorers and early settlers” Cunninghamia 5(2), 285-328.

Benson, J.S. (1998) Fire and vegetation National Parks Journal August, 2004.

Berger, A. (1992) IGBP PAGES/World Data Center-A for Paleoclimatology Data Contribution Series # 92-007.

Bird, C.F.M. & Frankel, D. (1991). Chronology and explanation in western Victoria and south-east South Australia. Archaeology in Oceania 26(1),1-16.

Birdsell, J.B. (1953). Some environmental and cultural factors influencing the structuring of Australian Aboriginal populations. The American Naturalist 87,171-207.

Birks, H.J.B. (1993). Quaternary palaeoecology and vegetation science: current contributions and possible future developments. Review of Palaeobotany and Palynology 79, 153-177.

Birmingham, A. (1966). Victoria natural radiocarbon measurements. Radiocarbon, 8, 507- 521.

Black, M.P. & Mooney, S.D. (in press) Holocene fire history from the Greater Blue Mountains World Heritage Area, New South Wales, Australia: the climate, humans and fire nexus Regional Environmental Change.

Black, M.P. & Mooney, S.D. (submitted) The archaeological and climatic implications of a 14 200 yr contiguous fire record from Gooches Crater, Blue Mountains, Australia Quaternary Science Reviews.

Black, M.P., Mooney, S.D. & Martin, H. (submitted) A >43 000 year vegetation and fire history from Lake Baraba, New South Wales, Australia Quaternary Science Reviews.

Black, M.P. & Mooney, S.D. (submitted) A ~6,100 yr vegetation and fire history from Kings Waterhole, Wollemi National Park, New South Wales Australian Geographer.

Blunier, T., Schwander, J., Stauffer, B., Stocker, T., Dällenbach, A., Indermühle, A. & Tschumi, J. (1997). Timing of the Antarctic Cold Reversal and the atmospheric CO2 increase with respect to the Younger Dryas event. Geophysical Research Letters, 24(21), 2683-2686.

Blunier, T., Chappellaz, J., Schwander, A., Dällenbach, B.,Stauffer, T., Stocker, D., Raynaud, J.,Jouzel, H.B.,Clausen, C.U., Hammer & Johnsen, S.J. (1998) Asynchrony of Antarctic and Greenland climate change during the last glacial period, Nature 394: 739- 743.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 162

Bodkin, F. (2004). Aboriginal fire management: background, applications and challenges, paper presented at the Bushfire in a Changing Environment, Nature Conservation Council of NSW Conference, Sydney, NSW, 24-25 June, 2004.

BoM (2003). Commonwealth Bureau of Meteorology Website (http://www.bom.gov.au). Department of Environment, Federal Government of Australia. Accessed December, 2003.

BoM (2005). Commonwealth Bureau of Meteorology Website (http://www.bom.gov.au). Department of Environment, Federal Government of Australia. Accessed February, 2005.

Bond, G.C., Broecker, W., Johnsen, S., MacManus, J., Labeyrie, L., Jouzel, J. & Bonani, G. (1993). Correlations between climate records from North Atlantic sediments and Greenland ice Nature 365, 143 –147.

Bond, G.C. & Lotti, R. (1995). Iceberg discharges into the North Atlantic on millenial time scales during the last glaciation Science 267: 1005-1010.

Bowdler, S. (1977). “The coastal colonisation of Australia” In J. Allen, J. Golson and R. Jones (eds) Sunda and Sahul: Prehistoric Studies in Southeast Asia, Malenesia and Australia, pp.205-46. Academic Press, London.

Bowdler, S. (1981). Hunters in the Highlands: Aboriginal Adaptations in the Eastern Australian Uplands Archaeology in Oceania 16: 99-111.

Bowler, J.M. (1976). Recent developments in reconstructing late Quaternary environments in Australia. In: The Origin of the Australians, (R.L.Kirk and A.G. Thorne, Eds.), pp. 55 – 77. Human Biology Series No. 6, AIAS, Canberra, and Humanities Press, New Jersey.

Bowler, J.M. (1981). Australian Salt Lakes: a palaeohydrological approach. Hydrobiologia 82, 431-444.

Bowler, J.M. (1986). Spatial variability and hydrologic evolution of Australian lake basins: Analogue for Pleistocene hydrologic change and evaporite production. Palaeography, Palaeoclimatology, Palaeoecology 54, 21 – 42.

Bowler, J.M., Johnston, H., Olley, J.M., Prescott, J.R., Roberts, R.G., Shawcross, W. & Spooner, N.A. (2003). New ages for human occupation and climatic change at Lake Mungo, Australia. Nature 421,837 – 840.

Bowman, D.M.J.S. & Brown, M.J. (1986) Bushfires in Tasmania: a botanical approach to anthropological questions. Archaeology in Oceania 21, 166-171.

Bowman, D.M.J.S. (1998). Tansley Review No. 101: The impact of Aboriginal landscape burning on the Australian biota. New Phytologist 140, 385-410.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 163

Blue Mountains Bushfire Management Committee (2000). Bushfire risk management plan 14th December, 2000, Blue Mountains Bushfire Management Committee, Katoomba.

Bradstock, R.A. (1993). ‘Fire management for conservation of native plant diversity’. In: The burning Question: Fire Management in NSW. (J.Ross Ed.), pp. 115-119. University of New England, Armidale.

Branagan, D.F. (1979) An outline of the geology and geomorphology of the Sydney Basin. Science Press, University of Sydney, Sydney.

Brandes, I. (2003). Burning questions: Australia’s bushfire policy.Water Research Laboratory, Canberra.

Breckell, M. (1993). Shades of grey: Aboriginal contact in the Blue Mountains. Blue Mountain Dreaming: the Aboriginal Heritage (E. Stockton, Ed.), pp. 114-121. Three Sisters Production, Winmallee.

Cary, G.J., & Morrison, D.A. (1995). ‘Effects of fire frequency on plant species composition of sandstone communities in the Sydney region: Combinations of inter-fire intervals’, Australian Journal of Ecology 20, 418-426.

Chalson, J.M. (1983). Palynology and palaeoecology of Jibbon Swamp, Royal National Park, New South Wales, Hons Thesis, University of New South Wales, Sydney.

Chalson, J.M. (1991). The late Quaternary vegetation and climatic history of the Blue Mountains, New South Wales, Australia Unpublished PhD thesis, University of New South Wales, Sydney.

Chappell, J. (1991). Late Quaternary environmental changes in eastern and central Austalia and their climatic interpretation. Quaternary Science Reviews 10: 377-390.

Chen, Y. (1986). Early Holocene vegetation dynamics of Lake Warrine Basin northeastern Queensland, Australia Unpublished PhD thesis, Australian National University, Canberra.

Cheney, N.P. (1981). Fire behaviour. In: Fire and the Australian Biota.(A.M.Gill, R.H.Groves, and I.R. Noble, Eds.), pp. 151-175. Australian Academy of Science, Canberra.

Christensen, P.E. & Burrows, N.D. (1986). ‘Fire: an old tool with a new use’. In: Ecology of biological invasions: an Australian perspective.(R.H.Groves and J.J.Burdon, Eds.), pp.97-105. Australian Academy of Science, Canberra.

Clark, J.S. (1988). Stratigraphic charcoal analysis on petrographic thin sections: application to fire history in northwestern Minnesota Quaternary Research 30: 81 – 91.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 164

Clark, J.S. & Robinson, J. (1993). Paleoecology of fire. In: Fire in the environment: the ecological, atmospheric, and climatic importance of vegetation fires (P.J. Crutzen, and J.G.Goldammer, Eds.) John Wiley and Sons Ltd., Germany.

Clark, J.S. & Royall, P.D. (1995). Particle-size evidence for source areas of charcoal accumulation in Late Holocene sediments of Eastern North American lakes Quaternary Research 43.

Clark, J. S. & Hussey, T. C. (1996). Estimating the mass flux of charcoal from sedimentary records: effects of particle size, morphology and orientation. The Holocene 6(2), 129-44.

Clark, J.S. & Robinson, J. (1993).‘Paleoecology of Fire’. In: Fire in the Environment – The Ecological, Atmospheric, and Climatic Importance of Vegetation Fires. (P.J.Crutzen and J.G. Goldammer, Eds.),pp.193-214. John Wiley & Sons Ltd, Chichester, England.

Clark, J.S., Cachier, H. & Goldammer, J.G. (1997). Sediment records of biomass burning and global change. Springer, Berlin.

Clark, J.S., Gill, A.M. & Kershaw, A.P. (2002). ‘Spatial variability in fire regimes: its effects on recent and past vegetation’. In: Flammable Australia – the fire regimes and biodiversity of a continent. (R.A.Bradstock, J.E.Williams and A.M. Gill, Eds.), pp.125- 141.Cambridge University Press.

Clark, R.L. (1982). Point count estimation of charcoal in pollen preparations and thin sections of sediments. Pollen et spores 24, 523-535.

Clark, R.L. (1983). Pollen and charcoal evidence for the effects of Aboriginal burning on the vegetation of Australia Archaeology in Oceania 18, 32-37.

Clark, R. L. (1984). Effects on charcoal of pollen preparation procedures. Pollen et Spores 26, 559-76.

Clark, R.L. (1990). ‘Ecological history for environmental management, Proceedings of the Ecological Society of Australia 16, 1-21.

Clark, S.S. & McLoughlin, L.C. (1986). Historical and biological evidence for fire regimes in the Sydney region prior to the arrival of Europeans: implications for future bushland management Australian Geographer 17, 101-112.

Clark, S.S. (1988). Effects of hazard-reduction burning on populations of understorey plant species on Hawkesbury sandstone, Australian Journal of Ecology 13, 473-484.

Clement, A.C., Seager, R. & Cane, M.A. (2000). Suppression of El Niño during the mid- Holocene by changes in the Earth’s orbit Paleoceanography 15: 731–737.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 165

COHMAP Project Members. (1988). Climate change of the last 18000 years: observations and model simulations Science 241: 1043 – 1052.

Cofer III, W.R., Koutzenogii, K.P., Kokorin, A. & Ezcurra, A. (1997). Biomass burning emissions and the atmosphere. In: Sediment records of biomass burning and global change. (J.S.Clark, H, Cachier, J.G.Goldammer, and B.J. Stocks. Eds.), pp. 189-206. NATO ASI Series 1: Global Environmental Change 51 Springer Verlag, Berlin.

Colhoun, E.A. & van de Geer, G.(1988). Darwin Crater, the King and Linda Valleys. In: Cainzoic Vegetation of Tasmania. Department of Geography Special Paper, pp. 30 -71, University of Newcastle, Newcastle.

Colhoun, E.A., Pola, J.S., Barton, C.E. & Heijnis, H. (1999). Late Pleistocene vegetation and climate history of Lake Selina, western Tasmania. Quaternary International 57/58, 5 – 23.

Colhoun, E.A. (2000). Vegetation change during the last interglacial - glacial cycle in western Tasmania. Palaeogeography, Palaeoclimatology, Palaeoecology, 155: 195-209.

Conroy, R.J. (1996). To burn or not to burn? A description of the history, nature and management of bushfires within Ku-Ring-Gai National Park. Proceedings of the Linnean Society of New South Wales 116, 79-95.

Conroy, R.J. & Sanders, J.M. (1993). Recommendations for sustainable fire management in natural areas. Poster Presentation – NPWS Biodiversity Conference. National Parks and Wildlife Service, Hurstville.

Cope, M.J. & Chaloner, W.G. (1985). ‘Wildfire: an interaction of biological and physical processes’. In: Geological Factors and the Evolution of Plants . (B.H.Tiffney, Ed.), pp.257- 269. Yale University Press, London.

Cosgrove, R. (1995). Late Pleistocene behavioural variation and time trends: the case from Tasmania. Archaeology in Oceania 30(3), 83 – 104.

Crowley, G.M. (1994). Groundwater rise, soil salinisation and the decline of Casuarina in south-eastern Australia during the Late Quaternary. Australian Journal of Ecology 19, 417 – 424.

Cunningham, C.J. (1984). Recurring natural fire hazards: a case study of the Blue Mountains, New South Wales, Australia. Applied Geography, 4, 5-27.

Cupper, M.L., Drinnan, A.N. & Thomas, I. (2000). Holocene palaeoenvironments of salt lakes in the Darling Anabranch region, south-western New South Wales, Australia. Journal of Biogeography 27, 1079–1094.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 166

D’Costa, D.M., Edney, P.A., Kershaw, A.P. & De Deckker, P. (1989). Late Quaternary palaeoecology of Tower Hill, Victoria, Australia. Journal of Biogeography 16, 461-482.

D’Costa, D.M. (1997). The reconstruction of Quaternary vegetation and climate on King Island, Bass Strait, Australia. PhD thesis, Monash University, Melbourne.

Daansgard, W., White, J.W.C. & Johnsen, S.J. (1989). The abrupt termination of the Younger Dryas event. Nature 339, 532 – 534.

Davis, M.B. (1994). Ecology and palaeoecology begin to merge. TREE (Trends in Evolution and Ecology) 9(10), 357-358.

De Deckker, P., Kershaw, A.P. & Williams, M.A.J. (1988). ‘Past environmental analogues’ In: Greenhouse (G.I. Pearman, Ed), pp.473-488. CSIRO, Melbourne.

De Deckker, P. (2001). Late Quaternary cyclic aridity in tropical Australia. Palaeogeography Palaeoclimatology Palaeoecology 170,1 – 9.

DeMenocal, P., Ortiz, J., Guilderson, T., Adkins, J., Sarnthein, M., Baker, L. & Yarusinsky, M. (2000). Abrupt onset and termination of the African Humid Period: rapid climate responses to gradual insolation forcing. Quaternary Science Reviews 19, 347-361.

Denton, G.H. & Hendy, C.H. (1994). Younger Dryas age advance of Franz Josef Glacier in the Southern Alps of New Zealand Science 264, 1434 – 1437.

Devoy, R.J., Dodson, J.R., Thom, B.G. & Nichol, S. (1994). Holocene environments in the Hawkesbury Valley, New South Wales: A comparison of terrestrial and marine records. Quaternary Science Reviews 13, 241 – 256.

Dharug and Lower Hawkesbury Historical Society. (1988). The Ferry, the Branch, the Creek: Aspects of Hawkesbury History. Southwood Press, Sydney.

Dodson, J.R. (1974). Vegetation and climatic history near Lake Keilambete, Western Victoria. Australian Journal of Botany 22: 709 – 717.

Dodson, J.R. (1975). Vegetation history and water fluctuations at Lake Leake, South- eastern South Australia. II* 50,000 BP to 10000 BP Australian Journal of Botany 22(719), 815 – 31.

Dodson, J.R. (1977). Late Quaternary Palaeoecology of Wyrie Swamp, southeastern South Australia. Quaternary Research 8, 97 – 114.

Dodson, J.R. (1983). Modern pollen rain in southeastern New South Wales, Australia. Review of Palaeobotany and Palynology 38, 249 – 268.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 167

Dodson, J.R. (1986). Holocene vegetation and environments near Goulburn, New South Wales. Australian Journal of Botany 34, 231–249.

Dodson, J.R. (1994). Quaternary vegetation history. In: Australian vegetation. (R.H. Groves, Ed).Cambridge University Press, Cambridge.

Dodson, J.R., Greenwood, P.W. & Jones, R.L. (1986). Holocene forest and wetland vegetation dynamics at Barrington Tops, New South Wales. Journal of Biogeography 13: 561-585.

Dodson, J.R., & Wright, R.V.S. (1989). ‘Humid to arid to subhumid vegetation shift on Pillaga Sandstone, Ulungra Springs, New South Wales.’ Quaternary Research 32, 182– 192.

Dodson, J.R. & Thom, B.G. (1992). Holocene vegetation history from the Hawkesbury Valley, New South Wales Proceedings of the Linnean Society if New South Wales 113(2), 121-134.

Dodson, J.R., Frank, K., Fromme, M., Hickson, D., McRae, V., Mooney, S. & Smith, J.D. (1994) Environmental systems and human impact at Cobrico Crater, South-western Victoria. Australian Geographical Studies 32(1), 27-40.

Dodson, J.R. & Mooney, S.D. (2002). An assessment of historic human impact on south- eastern Australian environmental systems using late Holocene rates of environmental change. Australian Journal of Botany 50,455-464.

Dortch, C.E. (1979) 33,000 year old stone and bone artefacts from Devil’s Lair, Western Australia Records of the Western Australian Museum 7(4), 329 – 67.

Dortch, C.E. & Merrilees, D. (1973). Human occupation of Devil’s Lair, Western Australia during the Pleistocene Archaeology and Physical Anthropology in Oceania 8(2), 89 – 115.

Edney, P.A., Kershaw, A.P. & De Deckker, P. (1990). A late Pleistocene and Holocene vegetation and environmental record from Lake Wangoom, Western Plains of Victoria, Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 80, 325-343.

Edwards, K. (2002). Atmospheric factors affecting fire season severity in south-eastern Australia, Unpublished Honours Thesis, School of Resources, Environment and Society, The Australian National University.

Elkin, A.P. (1954). The Australian Aborigines: how to understand them. Angus and Robertson, Sydney.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 168

Elsik, W.C. (1996). Chapter 10 Fungi. In: Palynology: principles and applications (J. Jansonius and D.C. McGregor, Eds.).American Association of Stratigraphic Palynologists Foundation 1, 293 – 305.

Faegri, K. & Iverson, J. (1975). Textbook of Pollen Analysis. Blackwell, Oxford.

Fairley, A. (1978). A field guide to the National Parks of New South Wales. Rigby Publishing, Sydney.

Fairley, A. & Moore, P. (2000). Native Plants of the Sydney District: An identification guide. Revised edn. Kangaroo Press, Sydney.

Fanning, P. (1988). The origin of Thirlmere Lakes, New South Wales. Unpublished report for the New South Wales National Parks and Wildlife Service, Sydney.

Field, J. & Dodson, J. (1999). Late Pleistocene megafauna and archaeology from Cuddie Springs, south eastern Australia. Proceedings of the Prehistoric Society 65, 275–301.

Fifield, L.K., Bird, M.I., Turney, C.S.M., Hausladen, P.A. & Santos, G.M. ( 2001). Radiocarbon dating of the human occupation of Australia prior to 40 ka BP – successes and pitfalls Radiocarbon 43(2), 1139-1145.

Flannery, T. (1994). The future eaters. Reed Books, Melbourne.

Flood, J. (1980). The moth hunters: Aboriginal prehistory of the Australian Alps. Australian Institute of Aboriginal Studies, Canberra.

Flood P.G., David, B., Magee, J. & English, B. (1987). Birrigai: a Pleistocene site in the south-eastern highlands Archaeology in Oceania 22(1),9-26.

Florence, R.G. (1994). The ecological basis of forest fire management in New South Wales. In The burning continent, forest ecosystems and fire management in Australia: current issues (P. Attiwill, R.G. Florence, W.E. Hurditch and W.J. Hurditch, Eds), pp. 15- 33, Institute of Public Affairs.

Forster, G.R., Campbell, D. & Moore, R.M. (1977). Vegetation and soils of the Western Region of Sydney CSIRO Division of Land Use Research. Technical Memorandum No. 77/10.

Fox, M .D. & Fox, B. J. (1986). The effect of fire frequency on the structure and floristic composition of a woodland understorey Australian Journal of Ecology 11.

Furby, J.H. (1995). Megafauna under the microscope, archaeology and palaeoenvironment at Cuddie Springs. Unpublished Ph.D. Thesis, School of Geography, UNSW.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 169

Gagan, M.K., Hendy, E.J., Haberle, S.G. & Hantoro, W.S. (2004). Post-glacial evoluation of the Indo-Pacific Warm Pool and El Nino-Southern Oscillation Quaternary International 118-119, 127-143.

Gale, S.J. & Pisanu, P.C. (2001). The late-Holocene decline of Casuarinaceae in southeast Australia. The Holocene 11, 485 – 490.

Galloway, R.W. (1973). Late Quaternary glaciation and periglacial phenomena in Australia and New Guinea. Palaeoecology Africa and Antarctica 8, 125 – 138.

Gardner, J. J. & Whitlock, C. (2001). Charcoal accumulation following a recent fire in the Cascade Range, northwestern USW, and its relevance for fire-history studies. The Holocene 11(5), 541-549.

Genever, M., Grindrod, J. & Barker, B. (2003). Holocene palynology of Whitehaven Swamp, Whitsunday Island, Queensland, and implications for the regional archaeological record Palaeogeography, Palaeoclimatology, Palaeoecology 201, 141 – 156.

Gentilli, J. (1972). Australian climate patterns. Thomas Nelson, Melbourne.

Gill, A.M. (1975) Fire and the Australian flora: a review. Australian Forestry 38, 4-25.

Gill, A.M. (1977). Management of fire prone vegetation for plant species conservation in Australia Search 8(1-2), 20-25.

Gill, A.M., Groves, R.H. & Noble, I.R. (1981). Fire and the Australian Biota Australian Academy of Science, Canberra.

Gill, M.A. (1994). How fires affect biodiversity. In: Fire and biodiversity: the effects and effectiveness of fire management: Proceedings of the Conference held 8-9 October 1994, Footscray, Melbourne,” Biodiversity Series 8, 47 – 55.

Gill, A.M. & Bradstock, R.A. (1995) Extinctions of biota by fires. Conserving Biodiversity: Threats and Solutions (R.A. Bradstock, T.D. Auld, D.A. Keith, R. Kingsford, D. Lunney and D. Sivertsen Eds), pp. 309-322. Surrey Beatty and Sons, Sydney.

Gill, A.M. (1996). ‘How fires affect biodiversity’. In: Fire and Biodiversity – The effects and effectiveness of fire management. (J.R.Merrick, Ed.), pp. 47-55. Biodiversity Series, Paper No. 8, Biodiversity Unit, Commonwealth Department of Environment Sport and Territories, Canberra.

Gillson, L. (2004) Testing non-equilibrium theories in savannas: 1400 years of vegetation change in Tsavo National Park, Kenya. Ecological Complexity 1, 281-298.

Goede, A., McDermott, F., Hawkesworth, C., Webb, J. & Finlayson, B. (1996). Evidence of Youner Dryas and neo-glacial cooling in a late Quaternary paleaotemperature

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 170

record from a speleotherm in eastern Victoria, Australia. Journal of Quaternary Science 11(1), 1 – 7.

Goldammer, J.G. (1999). Forests on fire Science 284, 1782 – 1783.

Gott, B. (1982). ‘Ecology of root use by the Aborigines if Southern Australia’, Archaeology in Oceania 17(1), 59–67.

Gott, B. (2005). Aboriginal fire management in south-eastern Australia: aims and frequency. Journal of Biogeography 32, 1203-1208.

Green, D., Singh, G., Polach, H., Moss, D., Banks, J. & Geissler, E.A. (1988). A fine- resolution palaeoclimatology from south-eastern Australia. Journal of Ecology 76,790-806.

Grimm, E.C. (1992). TILIA Software. Illinois State Museum, Springfield, Illinois.

GRIP Members. (1993). Climate instability during the last interglacial period recorded in the GRIP ice core. Nature 364: 203 – 207.

Gunston, N. (1974) Australian Reminiscences and Papers of L.E. Threlkeld Vol. II. Australian Aboriginal Studies No. 40. Australian Institute of Aboriginal Studies, Canberra.

Haberle, S.G., Hope, G.S. & Van der Kaars, S. (2001) Biomass burning in Indonesia and Papua New Guinea: natural and human induced fire events in the fossil record. Palaeogeography, Palaeoclimatology, Palaeoecology 171, 256-268.

Haberle, S.G., Atkinson, A., Heijnis, H., Zoppi, U. & Lawson, E. ( 2001) Fire, El Niño, and Rain Forest Dynamics in Northeast Queensland. Prog. Abstr. Australian Quaternary Association Biennial Conference. Port Fairy, Australia. 5th - 9th Feb, 2001.

Haberle, S.G. & David, B. (2004). Climates of change: human dimensions of Holocene environmental change in low latitudes of the PEPII transect. Quaternary International, 118-119, 165-179.

Haberle, S.G. & Ledru, M.P. (2001). Correlations among charcoal records of fires from the past 16,000 years in Indonesia, Papua New Guinea, and Central and South America. Quaternary Research, 55, 97-104.

Haberle, S.G. (in press) A 22ka pollen record from Lake Euramoo, Wet Tropics of NE Queensland, Australia Quaternary Research

Hallam, S.J. (1975) Fire and Hearth – a study of Aboriginal usage and European usurpation in south-western Australia. Australian Institute of Aboriginal Studies, Canberra.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 171

Hallett, D.J. & Walker, R.C. (2000). Paleoecology and its application to fire and vegetation management in Kootenay National Park, British Colombia Journal of Paleolimnology 24: 401-414.

Harle, K.J. (1997). Late Quaternary vegetation and climate change in southeastern Australia: palynological evidence from marine core E55-6. Palaeogeography, Palaeoclimatology, Palaeoecology 131, 465 – 483.

Harle, K.J. (1998). Patterns of vegetation and climate change in southwest Victoria over the last 200,000 years, PhD thesis. Monash University, Melbourne.

Harrison, S.P. (1993). Late Quaternary lake-level changes and climates of Australia. Quaternary Science Review 12, 211-232.

Harrison, S.P. & Dodson, J. (1993). Climates of Australia and New Guinea since 18,000 yr B.P., in H.E. Wright et al (Eds) Global Climates since the Last Glacial Maximum University of Minnesota Press: USA.

Harvard, O. (1944). Mrs Felton Mathew’s journal. Journal of the Royal Australian Historical Society 29, 88-128.

Haug, G.H., Hughen, K.A., Sigman, D.M., Peterson, L.C. & Röhl, U. (2001) Southward migration of the Intertropical Convergence Zone through the Holocene. Science 293, 1304 – 1308.

Head, L. (1989). Prehistoric aboriginal impacts on Australian vegetation: an assessment of the evidence Australian Geographer 20(1), 36-45.

Head, L. (1996). Australian Aborigines and a changing environment- views of the past and implications for the future. Aboriginal Involvement in Parks and Protected Areas ( J. Birkhead, T. DeLacy and L. Smith, Eds.), pp. 47-56. Aboriginal Studies Press, Canberra.

Herbert, C. (1983). Sydney Basin stratigraphy in Geology of the Sydney 1:100000 sheet 9130 (C. Herbert ed.) Geological survey of New South Wales Department of Mineral Resources, Sydney.

Hesse, P.P. (1994). The record of continental dust from Australia in Tasman Sea Sediments Quaternary Science Reviews 13(3), 257 – 272.

Hesse, P.P., Humphreys, G.S., Selkirk, P.M., Adamson, D.A., Gore, D.B., Nobes, D.C., Price, D.M., Schwenninger, J.L., Smith, B., Tulau, M. & Hemmings, F. (2003). Late Quaternary aeolian dunes on the presently humid Blue Mountains, eastern Australia. Quaternary International 108, 13 – 32.

Hiscock, P. & Attenbrow, V. (1998). Early Holocene backed artefacts from Australia Archaeology in Oceania 33(2), 49-62.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 172

Hiscock, P. & V. Attenbrow. (2004). A revised sequence of backed artefact production at Capertee 3 Archaeology in Oceania 39, 94-99.

Hodell, D.A., Kanfoush, S.L., Shemesh, A., Crosta, X., Charles, C., & Guilderson, T.P. (2001). Abrupt cooling of Antarctic surface waters and sea ice expansion in the South Atlantic sector of the Southern Ocean at 5000 cal yr B.P. Quaternary Research 56, 191 – 198.

Hooley, A.D., Southern, W. & Kershaw, A.P. (1980). Holocene vegetation and environments of Sperm Whale Head, Victoria, Australia. Journal of Biogeography 7, 349 – 362.

Hope G.S. (1994) Quaternary Vegetation. In R.Hill (Ed.) History of Australian Vegetation. Cretaceous to Recent. pp 368-389. Cambridge University Press, Cambridge.

Hope, G.S. (1999) Vegetation and fire response to late Holocene human occupation in island and mainland north west Tasmania. Quaternary International 59, 47-60.

Horsfall, L., Jelinek, A. & Timms, B. (1988). The influence of recreation, mainly power boating on the ecology of Thirlmere Lakes, NSW. Australian Verh. Internat. Limnol 23, 580 – 587.

Horton, D.R. (1982). The burning question: Aborigines, fire and Australian ecosystems. Mankind 13, 237-251.

Horton, D.R. (1994). The encyclopedia of Aboriginal Australia Aboriginal Studies Press, Canberra.

Horton, D.R. (2001). The Pure State of Nature: Sacred cows, destructive myths and the environment Allen and Unwin, St Leonards, New South Wales.

Hughes, P.J. & Sullivan, M.E. (1981). Aboriginal burning and late Holocene geomorphic events in eastern NSW Search 12, 277 – 278.

Hughes, P.J. & Lampert, R.J. (1982). Prehistoric population change in southern coastal New South Wales. Coastal Archaeology in Eastern Australia (S. Bowdler, Ed.), pp. 16-28. Department of Prehistory, Research School of Pacific Studies, Australian National University, Canberra.

Huntley, B. (1996) ‘Quaternary palaeoecology and ecology’, Quaternary Science Reviews 15, 591-606.

Hutchinson, G. (1957). A treatise on limnology: Volume 1 Geography, Physics and Chemistry John Wiley and Sons Inc., USA.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 173

Jacobson, G.L. & Bradshaw, R.H.W. (1981). The selection of sites for paleovegetational studies Quaternary Research 16, 80-96.

Jackson, W. D. (1968). Fire, air, water and earth- an elemental ecology of Tasmania Proceedings of the ecological society of Australia 3, 9-16.

Johnsen, S.J., Clausen, H.B., Dansgaard, W., Fuhrer, K., Gundestrup, N., Hammer, C.U., Iversen, P., Jouzel, J., Stauffer, B. & Steffensen, J.P. (1992). Irregular glacial interstadials recorded in a new Greenland ice core. Nature 359, 311–313.

Johnson, I. (1979). The getting of data: a case study from the recent industries of Australia Unpublished PhD thesis, Australian National University, Canberra.

Johnson, A.G. (1994). Late Holocene environmental changes on Kurnell Peninsula, NSW. Proceedings of the Linnean Society of New South Wales 114(3), 119–132.

Johnson, A.G. (2000) Fine resolution palaeoecology confirms anthropogenic impact during the late Holocene in the lower Hawkesbury Valley, NSW. Australian Geographer 31(2), 209–235.

Jones, R. (1969). Fire-Stick Farming Australian Natural History, 16, 224-228.

Jones, R. (1977). Sunda and Sahul: An introduction. In Sunda and Sahul: Prehistoric Studies in South-East Asia, Melanesia and Australia ( J. Allen, J. Golson and R. Jones Eds.), pp1-9. London: Academic Press.

Jones, R. (1995). Mindjongork: Legacy of the firestick. In Country in Flames: Proceedings of the 1994 Symposium on Biodiversity and Fire in North Australia, (D.B. Rose, Ed.), pp.11-17. Biodiversity Series 3. Canberra: Biodiversity Unit, Department of the Environment, Sport and Territories and North Australia Research Unit, Australian National University, Canberra.

Jouzel, J., Masson, V., Cattani, O., Falourd, S., Stievenard, M., Stenni, B., Longinelli, A., Johnsen, S.J., Ste¡enssen, J.P., Petit, J.-R., Schwander, J., Souchez, R. & Barkov, N.I. (2001). A new 27 ky high resolution East Antarctic climate record, Geophysical Research Letters 28, 3199-3202.

Jowsey, P.C. (1966). An improved peat sampler. New Phytologist 65, 245 – 248.

Keith, D.A. (1992) Fire and the conservation of native bushland plants. National Parks Journal October, 20-22.

Keith, D.A. (1996). Fire-driven extinction of plant populations: a synthesis of theory and review of evidence from Australian vegetation Proceedings of the Linnean Society of New South Wales 116, 37 – 78.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 174

Keith, D.A. & Benson, D.H. (1988). The natural vegetation of the Katoomba 1:100 000 map sheet. Cunninghamia 2, 107-143.

Keith, D.A. & Myerscough, P.J. (1993). Floristics and soil relations of upland swamp vegetation near Sydney. Australian Journal of Ecology 18, 325-344.

Keith, D.A, Williams, J.E. & Woinarski, C.Z. (2002). ‘Fire management and biodiversity conservation: key approaches and principles’. In Flammable Australia – the fire regimes and biodiversity of a continent. (R.A.Bradstock, J.E. Williams and A.M. Gills Eds.), pp.401-425.Cambridge University Press, Cambridge.

Kerr, R.A. (1993). How ice age got the shakes Science 260, 890-2.

Kershaw, A.P. (1976). A late Pleistocene and Holocene pollen diagram from Lynch’s Crater, north-eastern Queensland, Australia New Phytologist 77, 469-498.

Kershaw, A.P. (1978). Record of last interglacial-glacial cycle from northeastern Queensland Nature 272,159-161.

Kershaw, A.P. (1983). A Holocene pollen diagram from Lynch’s Crater, north-eastern Queensland, Australia New Phytologist 94, 669-82.

Kershaw, A.P. (1985). An extended late Quaternary vegetation record from north-eastern Queensland and its implications for the seasonal tropics of Australia Proceedings of the Ecological Society of Australia 13,179-189.

Kershaw, A.P. (1989) Was there a ‘Great Australian Arid Period’? Search 20,89-92.

Kershaw, A.P. (1995). Environmental change in Greater Australia. Antiquity 69, 656 – 675.

Kershaw, A.P. (1997). A modification of the Troels-Smith system of sediment description and portrayal. Quaternary Australasia 15(2), 63-68.

Kershaw A.P., D'Costa D.M., McEwan Mason J.R.C. & Wagstaff B.E. (1991). Palynological evidence for Quaternary vegetation and environments of mainland southeastern Australia Quaternary Science Reviews 10, 391-404.

Kershaw, A.P., McKenzie, G.M. & McMinn, A. (1993). A Quaternary vegetation history of northeastern Queensland from pollen analysis of ODP site 820. In McKenzie, J.A., Davies, P.J., Palmer-Julson, A. (Eds) Proceedings of the Ocean Drilling Program, Scientific Results, 133: 107 – 114

Kershaw, A.P., Clark, J.S. & Gill, A.M. (2002). A history of fire in Australia. In Bradstock, R., Williams, J. and Gill, A.M. (Eds.) Flammable Australia: the Fire Regimes and Biodiversity of a Continent pp. 3-25, Cambridge University Press, Cambridge.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 175

Kershaw, A.P., Van der Kaars, S. & Moss, P.T. (2003). Late Quaternary Milankovitch- scale climatic change and variability and its impact on monsoonal Australasia Marine Geology 201, 81 – 95.

Kirkpatrick, J. (1999). A continent transformed-Human impact on the natural vegetation of Australia 2nd edition. Oxford University Press, South Melbourne, Victoria, Australia.

Kitzberger, T. (2002). ENSO as a forewarning tool of regional fire occurrence in northern Patagonia, Argentina. International Journal of Wildland Fire 11, 33–39.

Kodela, P.G. & Dodson, J.R. (1988). A late Holocene vegetation and fire record from Ku- ring-gai Chase National Park, New South Wales. Proceedings of the Linnean Society of New South Wales. 110 (4), 317 – 326.

Kohen, J.L. (1993). The Darug and their neighbours – The traditional Aboriginal owners of the Sydney region. Darug Link and the Blacktown and District Historical Society.

Kohen, J.L. (1996). Aboriginal use of fire in southeastern Australia. Proceedings of the Linnean Society of New South Wales 116, 19-26.

Ladd, P.G. (1979). Past and present vegetation on the Delegate River in the highlands of eastern Victoria. Vegetation and climatic history from 12,000 B.P. to present Australian. Journal of Botany 27, 185 – 202.

Ladd, P.G. (1988). The status of Casuarinaceae in Australian Forests. In Australia’s Ever Change Forests: Proceedings of the First National Conference on Australian Forest History, Canberra, 9-11 May, 1988 pp. 63 – 99.

Ladd, P.G., Orchiston, D. W. & Joyce, E. B. (1992). Holocene vegetation history of Flinders Island. New Phytologist 122 (4), 757 – 767.

Lambeck, K. & Chappell, J. (2001). Sea Level Change through the last Glacial Cycles. Science 292, 679-686.

Latz, P.K. & Griffin, G.F. (1978).‘Changes in Aboriginal land management in relation to fire and to food plants in Central Australia’. In The nutrition of Aborigines in relation to the ecosystem of Central Australia (Hetzel, B.S. & Frith, H.J. Eds.) Papers presented at a Symposium CSIRO 23-26 October 1976, Canberra. pp.77-85.

Lees, B. (1992). Geomorphological evidence for late Holocene climatic change in northern Australia. Australian Geographer 23 (1), 1 – 10.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 176

Lewis, H.T. (1992). ‘The technology and ecology of nature’s custodians: anthropological perspectives on Aborigines and national parks’ in Aboriginal involvement in Parks and Protected Areas (J. Birkhead, T. De Lacy and L. Smith Eds.), Australian Institute of Aboriginal and Torres Strait Islander Studies Report Series, Aboriginal Studies Press, Canberra, pp. 15–28.

Lindesay, J., Colls, K., Gill, M. & Mason, I. (2004) Historical Climate and Fire Weather in the ACT Region: a Review, A Report to the ACT Department of Urban Services, Canberra.

Long, C.J., Whitlock, C., Bartlein, P.J. & Millspaugh, S.H. (1998) A 9000-year fire history from the Oregon Coast Range, based on a high-resolution charcoal study. Canadian Journal of Forestry Research, 28, 774-787.

Lourandos, H. (1980) Change or stability? Hydraulics, hunter gatherers and population in temperate Australia. World Archaeology, 11, 245-266.

Lourandos, H. (1983) Intensification: a late Pleistocene-Holocene archaeological sequence from southwestern Victoria. Archaeology in Oceania, 18(1), 81-94.

Lourandos, H. (1997) Continent of Hunter-Gatherers. Cambridge University Press, Cambridge.

Lloyd, P.J. & Kershaw, A.P. (1997) Late Quaternary vegetation and early Holocene quantitative climate estimates from Morwell Swamp, Latrobe Valley, South-eastern Australia. Australian Journal of Botany 45, 549-563.

Maasch, K.A., Mayewski, P.A., Karlen, W., Rohling, E.J., Stager, J.C. & Steig, E.J. (2003) Holocene climate variability. EOS: Transactions American Geophysical Union, 84 (46, Supplement), p.F908.

McCarthy, F. D. (1964) The archaeology of the Capartee Valley, New South Wales. Records of the Australian Museum, 26, 197-246.

McCormac, F.G., Hogg, A.G., Blackwell, P.G., Buck, C.E., Higham, T.F.G. & Reimer, P.J. (2004) ShCal04 Southern Hemisphere Calibration, 0–11.0 Cal Kyr BP Radiocarbon 46(3), 1087-1092.

MacDonald, G. M., Larsen, C. P. S., Szeicz, J. M. & Moser, K. A. (1991) The reconstruction of boreal forest fire history from lake sediments: a comparison of charcoal, pollen, sedimentological and geochemical indices. Quaternary Science Review 10, 102- 114.

McGlone, M., Hope, G., Chappell, J. & Barrett, P. (1996) Past climate change in Oceania and Antarctica. In Greenhouse: coping with climate change (WJ Bouma, GI Pearman, MR Manning Eds.) pp. 81-90 CSIRO Publishing, Melbourne

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 177

Mackaness, G. (1965) Fourteen journeys over the Blue Mountains of New South Wales, 1813-1814. Horwitz-Grahame, Sydney.

Mackereth, F. (1965) Chemical investigation of lake sediments and their interpretation Proceedings of the Royal Society 250: 165-213.

MacPhail, M.K. (1980) Regeneration processes in Tasmanian forests Search 11, 184 – 190.

Macphail, M.K. & Hope, G.S. (1985) Late Holocene mire development in montane southeastern Australia: a sensitive climatic indicator Search 15(11/12), 344 – 349.

Markgraf, V., Dodson, J.R., Kershaw, A.P., McGlone, M.S. & Nichols, N. (1992) Evolution of late Pleistocene and Holocene climates in the circum-south Pacific land areas Climate Dynamics 6, 193 – 211

Martin, A.R.H. (1994) Kurnell Fen: an eastern Australian coastal wetland, its Holocene vegetation, relevant to sea-level change and aboriginal land use Review of Palaeobotnay and Palynology, 80, 311-332

Mathews, R.H. (1897) The Burbung of the Darkinung Tribes. Proceedings of the Royal Society of Victoria 10 (n.s.)(1), 1–12.

McCormac, F.G., Hogg, A.G., Blackwell, P.G., Buck, C.E., Higham, T.F.G. & Reimer, P.J. (2004) ShCal04 Southern Hemisphere Calibration, 0–11.0 Cal Kyr BP Radiocarbon 46(3), 1087-1092.

McGlone, M.S., Kershaw, A.P.& Markgraf, V. (1992) El Niño/ Southern Oscillation climatic variability in Australasian and South American palaeoenvironmental records In El Niño: Historical and paleoclimatic aspects of the Southern Oscillation (H.F. Diaz and V. Markgraf, Eds.), pp. 435 – 462, Cambridge University Press, Cambridge.

McLoughlin, L.C. (1998) Season of burning in the Sydney region: The historical records compared with recent prescribed burning. Australian Journal of Ecology 23, 393-404.

Miller, G., Mangan, J., Pollard, D., Thompson, S., Felzer, B. & Magee, J. (2005) Sensitivity of the Australian Monsoon to insolation and vegetation: Implications for human impact on continental moisture balance Geology 33(1), 65–68.

Millspaugh, S. H. & Whitlock, C. (1995) A 750-year fire history based on lake sediment records in central Yellowstone National Park, USA. The Holocene 5(3), 283-92.

Millspaugh, S.H., Whitlock, C. & Bartlein, P.J. (2000) Variations in fire frequency and climate over the past 17 000 yr in central Yellowstone National Park Geology 28(3), 211- 214.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 178

Mooney, S. & Radford, K. (2001). A simple and fast method for the quantification of macroscopic charcoal in sediments. Quaternary Australasia 19(1), 43-46.

Mooney, S. D., Radford, K. L. & Hancock, G. (2001) Clues to the 'burning question': pre-European fire in the Sydney coastal region from sedimentary charcoal and palynology. Ecological Management and Restoration 2(3), 203-12.

Mooney , S. & Black, M. (2003) A simple and fast method for calculating the area of macroscopic charcoal isolated from sediments. Quaternary Australasia 21(1), 18-21.

Moore, D.R. (1981) Results of an archaeological survey of the Hunter River Valley, NSW, Australia. Part II: Problems of the lower Hunter and contacts with the Hawkesbury Valley Records of the Australian Museum 33(9), 388 – 442.

Morrill, C. (2003) Timing and spatial extent of Mid-Holocene abrupt climate change. EOS Transactions of the American Geophysical Union 84, F911.

Morrison, K. D. (1994) Monitoring regional fire history through size-specific analysis of microscopic charcoal: the last 600 years in south India. Journal of Archaeological Science 21, 675-685.

Morrison, D.A., Cary, G.C., Pengelly, S.M., Ross, D.G., Mullins, B.J., Thomas, C.R. & Anderson, T.S. (1995) Effects of fire frequency on plant species composition of sandstone communities in the Sydney region: Inter-fire interval and time-since-fire. Australian Journal of Ecology 20, 239-247.

Moss, P.T. (1999) Late Quaternary environments of the humid tropics of northeastern Australia. PhD thesis, Monash Unviersity, Melbourne.

Moss, P.T. & Kershaw, A.P. (1999) Evidence from marine ODP Site 820 of fire/vegetation/climate patterns in the humid tropics of Australia over the last 250,000 years. In Bushfire 99, Proceedings of the Australian Bushfire Conference, Albury, Australia, July 1999, 269 – 279.

Moy, C.M., Seltzer, G.O., Rodbell, D.T. & Anderson, D.M. (2002) Variability of El Niño/Southern Oscillation activity at millennial timescales during the Holocene epoch. Nature 420, 162 - 165

Mulvaney, D.J. (1971) Aboriginal social evolution: a retrospective view. Aboriginal Man and Environment in Australia (D.J. Mulvaney and J. Golson Eds.). Australian National University Press, Canberra.

Mulvaney, D.J. (1975) The Prehistory of Australia. Penguin, Harmondsworth.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 179

Nanson, G.C., Young, R.W. & Stockton, E.D. (1987) Chronology and palaeoenvironment of the Cranebrook Terrace (near Sydney) containing artefacts more than 40,000 years old. Archaeology in Oceania 22(2), 72 – 78.

Nanson, G.C., Price, D.M. & Short, S. (1992) Wetting and drying of Australia over the past 300 ka. Geology 20, 791–794

National Parks and Wildlife Services (1995) Thirlmere Lakes National Park New Plan of Management. NSW National Parks and Wildlife Service, Sydney.

National Parks and Wildlife Service (2001) Wollemi National Park Plan of Management, New South Wales National Parks and Wildlife Service, Sydney.

National Parks and Wildlife Service (2004) Sydney Basin Bio-region. http://www.nationalparks.nsw.gov.au/npws.nsf/Content/Sydney+Basin+Bioregion Website Accessed: October 2005

Nieuwenhuis, A. (1987) The effects of fire frequency on the sclerophyll vegetation of the West Head, New South Wales. Australian Journal of Ecology 12, 373-385.

Nicholson, P.H. (1981) ‘Fire and the Australian Aborigine’. In Fire and the Australian Biota (A.M. Gill, R.H. Groves and I.R. Noble Eds.) Australian Academy of Science, Canberra..

Noakes, A.J. (1998) A study of the land use change and fire record of the Thirlmere Lakes area since European settlement as provided by the sediment record. Unpublished thesis (Hons.), University of Wollongong, Wollongong.

Ogden, J., Basher, L. & McGlone, M. (1998) Fire, forest regeneration and links with early human habitation: evidence from New Zealand. Annals of Botany 81: 687-696.

Pals, J.P., Van Geel, B. & Delfos, A. (1980) Paleoecological studies in the Klokkeweel Bog near Hoogkarspel (Prov. Of Noord-Holland). Review of Palaeobotany and Palynology 30, 371 – 418.

Patterson, W. A., Edwards, K. J. & Maguire, D. J. (1987) Microscopic charcoal as a fossil indicator of fire. Quaternary Science Reviews 6, 3-23.

Petit, J.R., Basile, I., Leruyuet, A., Raynaud, D., Lorius, C., Jouzel, J., Stievenard, M., Lipenkov, V.Y., Barkov, N.I., Kudryashov, B.B., Davis, M., Saltzman, E. & Kotlyakov, V. (1997) Four climate cycles in Vostok ice core. Nature 387, 359-360.

Petit, J.R., Jouzel, J., Raynaud, D., Barkov, N.I., Barnola, J.M., Basile, I.,Benders, M., Chappellaz, J., Davis, M., Delayque, G., Delmotte, M., Kotlyakov, V.M., Legrand, M., Lipenkov, V.Y., Lorius, C., Pépin, L., Ritz, C., Saltzman, E. & Stievenard, M. (1999)

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 180

Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature 399, 429-436.

Pyne, S.J. (1991) Burning bush: a fire history of Australia, Allen and Unwin, Sydney.

O'Connell, J. F. & Allen, F.J. (1998) When did humans first arrive in Greater Australia, and why is it important to know? Evolutionary Anthropology 6, 132-146.

Pickett, E.J., Harrison, S.P., Hope, G., Harle, K., Dodson, J.R., Kershaw, A.P., Prentice, I.C., Backhouse, J., Colhoun, E.A., D’Costa, D., Flenly, J., Grindrod, J., Haberle, S., Hassell, C., Kenyon, C., MacPhail, M., Martin, H., Martin, A.H., McKenzie, M., Newsome, J.C., Penny, D., Powell, J., Raine, J.I., Southern, W., Stevenson, J., Sutra, J-P., Thomas, I., van der Kaars, S. & Ward, J. (2004) Pollen- based reconstructions of biome distributions for Australia, Southeast Asia and the Pacific (SEAPAC region) at 0, 6000 and 18,000 14C yr BP. Journal of Biogeography 31, 1381 – 1444.

Prosser, I. (1990) ‘Fire, humans and denudation at Wangrah Creek, Southern Tablelands, N.S.W.’, Australian Geographic Studies 28, 77-95.

Purdie, R.W. & Slatyer, R.O. (1976) Vegetation succession after fire in sclerophyll woodland communities in south-eastern Australia. Australian Journal of Ecology 1, 223- 236.

Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Bertrand, C., Blackwell, P.G., Buck, C.E., Burr, G., Cutler, K.B., Damon, P.E., Edwards, R.L., Fairbanks, R.G., Friedrich, M., Guilderson, T.P., Hughen, K.A., Kromer, B., McCormac, F.G., Manning, S., Ramsey, C.B., Reimer, R.W., Remmele, S., Southon, J.R., Stuiver, M., Talamo, S., Riedinger, M.A., Steinitz-Kannan, M, Last, W.M. & Brenner, M. (2002) A ~6100 14C yr record of El Niño activity from the Galapagos Islands. Journal of Paleolimnology 27, 1-7.

Riley, S.J. & Henry, H.M. (1987) A geophysical survey of Culoul and Mellong Creek valley fills: implications for valley development in sandstone terrain. Journal and Proceedings of the Royal Society of New South Wales 120, 101–115.

Roberts, R.G., Jones, R. & Smith, M.A. (1990) Thermoluminescence dating of a 50000 year old human occupation in northern Australia. Nature 345, 153 – 156.

Roberts, R.G., Jones, R., Spooner, N.A., Head, M.J., Murray, A.S. & Smith, M.A., (1994) The human colonisation of Australia: optical dates of 53000 and 60000 years bracket human arrival at Deaf Adder Gorge, Northern Territory. Quaternary Science Reviews 13, 573 – 84.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 181

Rodbell, D.T., Seltzer, G.O., Anderson, D.M., Abbott, M.B., Enfield, D.B. & Newman, J.H. (1999) An ~15,000 year record of El Niño-driven alleviation in southwestern Ecuador. Science 283, 516-520.

Rolls, E. (1981) A million wild acres Penguin Books, Melbourne.

Rose, D. B. (1998) Totemism, Regions, and Co-management in Aboriginal Australia Presented at "Crossing Boundaries", the seventh annual conference of the International Association for the Study of Common Property, Vancouver, British Columbia, Canada, June 10-14, 1998.

Ross, A. (1985) Archaeological evidence for population change in the middle to late Holocene in southeastern Australia. Archaeology in Oceania, 10, 81-89.

Rowland, M.J. (1983) Aborigines and environment in Holocene Australia: changing paradigms. Australian Aboriginal Studies 2, 62-77.

Rowland, M.J. (1999) Holocene environmental variability: have its impacts been underestimated in Australian prehistory? The Artefact, 22, 11-48.

Roy, P.S. (1994) Holocene estuary evolution – stratigraphic studies from south-eastern Australia in Incised-valley systems: Origin and sedimentary sequences SEPM Special publication, Society of Sedimentary Geology 51, 241-63.

Roy, P.S. (1998) 25: Cainozoic geology of the coast and shelf. In Geology of New South Wales – Synthesis. Vol. 2 – Geological Evolution (H. Basden Ed.), pp 361 – 85 Geological Survey of New South Wales, Memoir Geology 13(2). Department of Mineral Resources, Sydney.

Ryan, D.G., Ryan, J.E. & Starr, B.J. (1995) The Australian landscape – observations of explorers and early settlers Murrumbidgee Catchment Management Committee: Wagga Wagga.

Ryan, K., Fisher, M. & Schaeper, L. (1996) The natural vegetation of the St Albans 1:100000 map sheet. Cunninghamia 4(3), 433–482.

Sandweiss, D.H., Maasch, K.A. & Anderson, D.G. (1999) Transitions in the mid- Holocene. Science, 283, 499-500.

Sandweiss, D. H., Maasch, K.A., Burger, R.L., Rischardson, J.B., Rollins, H.B. & Clement, A. (2001) Variation in Holocene El Nino frequencies: Climate records and cultural consequences in ancient Peru. Geology, 29(7), 603-606.

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.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 182

Shulmeister, J. & Lees, B.G. (1995) Pollen evidence from tropical Australia for the onset of an ENSO dominated climate at circa 4000 BP. The Holocene 5, 10-18.

Siddiqi, M. Y., Carolin, R. C. & Myerscough, P. J. (1975) Studies in the ecology if coastal heath in NSW. III. Regrowth of vegetation after fire. Proceedings of the Linnean Society of New South Wales 101(1), 12-16.

Siegert, F., Ruecker, G., Hinrichs, A. & Hoffman, A.A. (2001) Increased damage from fires in logged forests during droughts caused by El Niño. Nature 414, 437 – 440.

Singh, G., Kershaw, A.P. & Clark, R. (1981) Quaternary vegetation and fire history in Australia. In Fire and the Australian Biota (A.M. Gill, R.H. Groves and I.R. Noble Eds.) pp. 23 – 76, Australian Academy of Science, Canberra.

Singh, G. & E.A. Geissler (1985) Late Cainozoic history of vegetation, fire, lake levels and climate, at Lake George, New South Wales, Australia. Philosophical Transactions of the Royal Society of London B 311, 379-477.

Smith, A.J., Donovan, J.J., Ito, E., Engstrom, D.R. & Panek, V. (2002) Climate-driven hydrologic transients in lake sediment records: multiproxy record of mid-Holocene drought. Quaternary Science Reviews 21, 625-631.

Sousa, W.P. (1984) The role of disturbance in natural communities. Ann. Rev. Ecol. Syst. 15, 353-91.

Steig, E.J. (1999) Paleoclimate: Mid-Holocene Climate Change. Science 286 (5444), 1485- 1487.

Stenni, B., Masson-Delmotte, V., Johnsen, S., Jouzel, J., Longinelli, A., Monnin, E., Rothlisberger, R. & Selmo, E. (2001) An oceanic cold reversal during the last deglaciation. Science 293, 2074-2077.

Stevenson, J., Sutra, J-P., Thomas, I., van der Kaars, S. & Ward, J. (2004) Pollen- based reconstructions of biome distributions for Australia, Southeast Asia and the Pacific (SEAPAC region) at 0, 6000 and 18,000 14C yr BP. Journal of Biogeography 31, 1381 – 1444.

Stockton, E.D. (1970) An archaeological survey of the Blue Mountains. Mankind 7, 295- 301.

Stockton, E.D. & Holland, W. (1974) Cultural sites and their environment in the Blue Mountains. Archaeology and Physical Anthropology in Oceania 9, 36-65.

Stockton, E.D. (2004) Cranebrook Terrace revisited. Archaeology in Oceania April, 2004.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 183

Stockton, E.D. (2005) The First Inhabitants: The Aborigines. A Hand Shake Through Time.http://www.midmountainshistory.org.au/original.html Web site accessed: October 2005.

Stuiver, M., Reimer, P. J. & Reimer, R. W. (2005) CALIB 5.0. [WWW program and documentation]. .

Sweller, S. & Martin, H.A. (2001) A 40,000 year vegetation history and climatic interpretations of Burraga Swamp, Barrington Tops, New South Wales Quaternary International 83-85, 233-244.

Swetnam, T.W. & Betancourt, J.L. (1990) Fire-Southern Oscillation relations in the southwestern United States. Science 249, 1017-1020.

Taylor, F.W., van der Plicht, J. & Weyhenmeyer, C.E. (2004) IntCal04 Terrestrial Radiocarbon Age Calibration, 0–26 Cal Kyr BP. Radiocarbon 46, 1029-1058.

Timms, B. (1992) Lake geomorphology. Gleneagles Publishing, Adelaide

Tinner, W., Conedera, M., Ammann, B., Gaggeler, H. W., Gedye, S., Jones, R. & Sagesser, B. (1998) Pollen and charcoal in lake sediments with historically documented forest fires in southern Switzerland since AD 1920. The Holocene 8(1), 31-42.

Tindale, N.B. (1974) Aboriginal Tribes of Australia. Australian National University Press, Canberra.

Turbet, P. (2001) The Aborigines of the Sydney district before 1788. Kangaroo Press, Sydney.

Turney, C.S.M., Kershaw, A.P., Moss, P., Bird, M.I., Fifield, L.K., Cresswell, R.G., Santos, G.M., di Tada, M.L., Hausladen, P.A. & Zhou, Y. (2001) Redating the onset of burning at Lynch's Crater (North Queensland): Implications for human settlement in Australia. Journal of Quaternary Science 16, 767-771.

Turney, C.S.M., McGlone, M.S. & Wilmshurst, J.M. (2003) Asynchronous climate change between New Zealand and the North Atlantic during the last deglaciation Geology March 2003, 223-226.

Turney, C.S.M., Kershaw, A.P., Clemens, S.C., Branch, N., Moss, P.T. & Fifield, L.K. (2004) Millenial and orbital variations of El Niño/ Southern Oscillation and high-latitude climate in the glacial period. Nature 428, 306-310. van der Kaars, S., Wang, X., Kershaw, A.P. and Guichard, F. (2000) A late Quaternary palaeoecological record from the Banda Sea, Indonesia: patterns of vegetation, climate and biomass burning in Indonesia and northern Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 155, 135 – 153.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 184

Vigilante, T. (2001) Analysis of explorers’ records of Aboriginal burning in the Kimberley region of Western. Australia Journal of the Institute of Australian Geographers 39(2).

Wang, H., Follmer, L. R. & Liu, J. C. (2000) Isotope evidence of paleo-El Niño-Southern Oscillation cycles in loess-paleosol record in the central United States. Geology 28, 771– 774.

Wang, X., van der Kaars, S., Kershaw, A.P., Bird, M.& Jansen, F. (1999) A record of fire, vegetation and climate through the last three glacial cycles from Lombok Ridge core G6-4, eastern Indian Ocean, Indonesia. Palaeogeography, Palaeoclimatology, Palaeoecology 147, 241 – 256.

Wasson, R.J. & Clark, R.L. (1988) The Quaternary in Australia- Past, present and future. Quaternary Australasia 6, 17 – 22.

Wasson, R.J., Fitchett, K., Mackey, B. & Hyde, R. (1988) Large-scale patterns of dune type, spacing and orientation in the Australian continental dunefield. Australian Geography 19, 89 – 104.

Weaver, A.J, Saenko, O.A., Clark, P.U. & Mitrovica, J.X. (2003) Meltwater Pulse 1A from Antarctica as a Trigger of the Bølling-Allerød Warm Interval Science 299(5613), 1709-1713.

Westerley, A.L. & Swetnam, T.W. (2003) Interannual to decadal drought and wildfire in the Western United States. Eos 84(49), 553 – 555.

Whelan, R.J. (1995) The ecology of fire. Cambridge University Press.

Whelan, R.J., Rodgerson, L., Dickman, C.R. & Sutherland, E.F. (2002) ‘Critical life cycles of plants and animals: developing a process-based understanding of population changes in fire-prone landscapes’. In Flammable Australia – the fire regimes and biodiversity of a continent (R.A. Bradstock, J.E. Williams and A.M. Gill, Eds.). pp.94-124. Cambridge University Press, Cambridge.

Whitlock, C. & Millspaugh, S.H. (1996) Testing the assumptions of fire-history studies: an examination of modern charcoal accumulation in Yellowstone National Park, USA. The Holocene 6(1), 7-15.

White, P.J. (1994) Site 820 and the evidence for early occupation in Australia. Quaternary Australasia 12(2), 21 – 22.

Williams, M.A.J., Dunkerley, D., DeDeckker, P., Kershaw, P. & Chappell, J. (1993) Quaternary Environments Routledge, USA.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black References 185

Williams, J.E., Whelan, R.J. & Gill, A.M. (1994) Fire and environmental heterogeneity in southern temperate forest ecosystems: Implications for management. Australian Journal of Botany 42, 125-137.

Williams, J.E. & Gill, A.M. (1995) The impact of fire regimes on native forests on eastern New South Wales. Environmental Heritage Monograph Series No. 2. NSW National Parks and Wildlife Service, Hurstville.

Yokoyama, Y., Purcell, A., Lambeck, K. & Johnston, P. (2001) Shoreline reconstruction around Australia during the Last Glacial maximum and Lake Glacial States. Quaternary International 83-85, 9-18.

A Late Quaternary palaeoenvironmental investigation of the fire, climate, human and vegetation nexus from the Sydney Basin, Australia Manu P. Black