Elsevier

Quaternary Science Reviews

Volume 145, 1 August 2016, Pages 161-182
Quaternary Science Reviews

The timing and cause of megafauna mass deaths at Lancefield Swamp, south-eastern Australia

https://doi.org/10.1016/j.quascirev.2016.05.042Get rights and content

Highlights

  • Lancefield Swamp contains thousands of megafauna bones, mainly giant macropods.

  • Our multi-disciplinary analyses show how and when these bones accumulated.

  • Megafauna died at Lancefield Swamp in mass numbers during severe droughts.

  • These events started well before humans appeared in Sahul.

  • Pleistocene climatic variability was a factor in regional megafauna extinctions.

Abstract

Lancefield Swamp, south-eastern Australia, was one of the earliest sites to provoke interest in Pleistocene faunal extinctions in Sahul (Pleistocene Australia-New Guinea). The systematic investigation of the deposit in the early 1970s identified megafaunal remains dominated by the 100–200 kg kangaroo Macropus giganteus titan. Associated radiocarbon ages indicated that the species was extant until c.30,000 BP, suggesting significant overlap with human settlement of Sahul. This evidence was inconsistent with contemporary models of rapid human-driven extinctions. Instead, researchers inferred ecological tethering of fauna at Lancefield Swamp due to intense drought precipitated localised mass deaths, consistent with Late Pleistocene climatic variability. Later investigations in another part of the swamp, the Mayne Site, remote to the initial investigations, concluded that mass flow disturbed this area, and Electron Spin Resonance (ESR) analyses on megafauna teeth returned wide-ranging ages. To clarify site formation processes and dating of Lancefield Swamp, we excavated new test-pits next to previous trenches in the Classic and Mayne Sites. We compared absolute chronologies for sediments and teeth, sedimentology, palaeo-topography, taphonomy, and macropod age at death across the swamp. Luminescence dating of sediments and ESR analysis of teeth returned ages between c.80,000 and 45,000 years ago. We found no archaeological remains in the bone beds, and evidence of carnivore activity and fluvial action, in the form of reactivated spring flow. The latter disturbed limited parts of the site and substantial areas of the bone beds remained intact. The faunal assemblage is dominated by megafaunal adult Macropus, consistent with mass die-offs due to severe drought. Such droughts appear to have recurred over millennia during the climatic variability of Marine Isotope Stages 4 and 3. These events began tens of millennia before the first appearance of Aboriginal people in Sahul and only the very youngest fossil deposits could be coeval with the earliest human arrivals. Therefore, anthropogenic causes cannot be implicated in most if not all of mass deaths at the site. Climatic and environmental changes were the main factors in site formation and megafauna deaths at Lancefield Swamp.

Introduction

In the contentious debate over the causes of Late Pleistocene megafaunal extinctions, understanding the contexts of megafauna deaths is critical. One of the richest fossil megafauna sites in Sahul (Pleistocene Australia-New Guinea) is Lancefield Swamp in south-eastern Australia, which contains discrete accumulations of bones (referred to as bone beds) of several thousand giant kangaroos and other extinct taxa (Fig. 1) (Gillespie et al., 1978, Van Huet, 1999, Van Huet et al., 1998). In the 1970s, radiocarbon samples collected beneath one bone bed returned ages of 26,600 ± 650 BP and 25,200 ± 800 (uncalibrated) BP (∼31-30 ka BP). These age estimates post-dated the arrival of humans in Sahul, as it was then understood, by several millennia (Bowler and Thorne, 1976). Current evidence places the first humans in Sahul around 50–45 ka BP (Summerhayes et al., 2010, Allen and O’Connell, 2014). Taphonomic and palaeo-ecological studies revealed that the megafauna perished at Lancefield as a result of prolonged drought or environmental stress (Gillespie et al., 1978), perhaps associated with the lead-up to the Last Glacial Maximum (LGM), 30–19 ka BP (Lambeck et al., 2002).

Later investigations, in a different part of the swamp (Fig. 2), concluded that there was significant re-working and disturbance at the site (Van Huet, 1999, Van Huet et al., 1998). These conclusions were at odds with the original interpretation of a waterhole death assemblage (Gillespie et al., 1978). On the basis of Electron Spin Resonance (ESR) analyses, Van Huet, 1994, Van Huet, 1999, Van Huet et al., 1998) and Peel (2001) argued that the bone deposits were much older than originally argued, proposing an age range of 30–60 ka BP. In this interpretation, the megafaunal bone bed accumulated over a long period as a result of various site formation processes, including mass flow, and represented bones from multiple sources and death events.

The aim of this paper is to present the results of more recent investigations involving sedimentological, geomorphic and taphonomic analyses, coupled with radiocarbon, Optically Stimulated Luminescence (OSL) and ESR studies. The paper assesses the various interpretations of the Lancefield Swamp fossil assemblages, particularly the timing and mode of bone deposition at Lancefield Swamp (Dortch, 2004, Lee, 2009, Ngo, 2004). Our findings are discussed in the context of current debates concerning the decline and disappearance of the Late Pleistocene megafauna suite in Sahul and globally (Barnosky et al., 2004, Wroe and Field, 2006, Field et al., 2008, Field et al., 2013, Brook et al., 2013, Wroe et al., 2004, Wroe et al., 2013, Stuart, 2014, Saltré et al., 2016).

Lancefield Swamp (37° 17′ S, 144° 43′ E) is a spring-fed swamp adjacent to the small town of Lancefield, in southern-central Victoria, south-eastern Australia (Fig. 1, Fig. 2). The swamp is located on a 10 km wide shallow basalt plain in the upper catchment of the Maribyrnong Basin at 500 m above sea level (ASL) (Fig. 1B). The present-day swamp features two depressions, one each at the western and eastern ends, and both featuring Pleistocene and Holocene deposits. In the terminology of Van Huet, 1994, Van Huet, 1999, the western depression is known as the Classic Site and the eastern depression as the Mayne Site (Fig. 1C). The positions of the western and eastern depressions correspond to two springs mapped in the late 1800s and to the positions of 19th and 20th century wells (Fig. 1, Fig. 2). Historic and recent dam construction has effectively removed parts of the swamp between the two depressions.

Surrounding the swamp are basaltic clays capped by laterite. Gillespie et al. (1978) attributed the swamp’s formation to rapid spring flow collapsing the laterite capping and creating a depression in the land surface. The depression subsequently filled with clay and organic remains. Gillespie et al. (1978) argued that the first erosional phase of spring flow, due to relatively high spring discharge, occurred well before the LGM. Faunal remains accumulated on the basaltic clay surface at the base of the depression.

Channel-cutting and filling within and below the bone bed in one part of the Classic Site, revealed in Square H, indicate that erosion occurred before and continued after initial bone deposition. The results of sediment, pollen and taphonomic analyses indicated that large parts of the swamp deposit formed by sub-aqueous low-energy deposition of fine sediments around the bones of the large animals (Ladd, 1976, Gillespie et al., 1978, Macumber, 1991). As environmental conditions improved during the terminal Pleistocene and Holocene, vegetation growth increased across the swamp and black clay formed, capping the Pleistocene deposit. The most recent phase of rapid sediment filling was generated by land clearing around the swamp in the 19th and 20th centuries (Ladd, 1976).

Fossil megafauna bones were first discovered in a well excavated in 1843 by J.P. Mayne (Hobson, 1846). The high water table and consequent inundation at that time discouraged further investigations (Orchiston et al., 1977). The first systematic excavations took place in the 1970s at the Classic Site and the excavators used pumps to remove water from the trenches (Horton, 1976, Ladd, 1976, Gillespie et al., 1978, Horton and Samuel, 1978, Horton and Wright, 1981). Later work focussed on the South Site (Fig. 1; investigated 1983–1984, 1991) (Munro, 1983, Van Huet, 1999), followed by the Mayne Site (investigated in 1991–1994, 2001: Peel, 2001, Van Huet, 1994, Van Huet, 1999). Van Huet (1999) reported evidence for 19th century well-digging in the Mayne Site and argued that it was the location of the 1843 discovery. Both the Classic Site and the Mayne Site are close to historic wells, with the latter closest to former and extant house sites from the 1800s (Fig. 1C).

Systematic investigations led by R.V.S. Wright and D. Horton at the Classic Site revealed a horizontal 10–20 cm thick bed of dense, interlocking bones, covering an area at least 30 m in diameter and containing more than 1000 bones per square metre (Fig. 2a and b; Gillespie et al., 1978). Radiocarbon age estimates on bone were rejected due to suspected exchange between the material dated and groundwater carbonate (Table 1). Fill from a channel underlying the bone bed in Square H and Square H extension yielded charcoal fragments that were radiocarbon-dated to ∼31–30 ka BP, suggesting a maximum age for the formation of the bone bed (Gillespie et al., 1978). More than 80% of the taxonomically identifiable bones were from very large adult kangaroos attributed to Macropus giganteus titan (Marsupialia, Mammalia). M. giganteus titan is considered to be the Pleistocene ancestor of the modern eastern grey kangaroo (M. giganteus) (Dawson and Flannery, 1985, Wroe et al., 2013). Macropus giganteus titan is estimated to have been c.40% larger and three to four times heavier than M. giganteus (Helgen et al., 2006). Previous estimates of the number of M. giganteus titan that died in or near the swamp range up to 10,000 individuals. Taxonomic distribution, age profile and bone pathologies indicated that severe drought coupled with predation by large carnivores such as Thylacoleo carnifex were the most likely factors causing kangaroo deaths on such a scale (Gillespie et al., 1978, Horton and Samuel, 1978, Horton and Wright, 1981). Two flaked stone artefacts were reported, one a 15 cm blade embedded in the bone bed, the other a 3 cm2 broken flake in the channel underlying the bone bed. Although these artefacts documented a human presence around the time of bone accumulation, there was no evidence for a human role in bone bed formation. The presence of pollen from aquatic plants in the bone bed indicated either standing water or marshy conditions, at least seasonally, during this period (Ladd, 1976).

In the 1990s Van Huet excavated at the South Site and the Mayne Site, finding additional bone beds, and also sediment rafts, poorly sorted gravels, heavily abraded bone fragments and upwards fining of sediments. These results were interpreted as evidence for re-working – possibly mass flow or overland stream flow (Munro, 1983, Peel, 2001, Van Huet, 1999).

ESR and radiocarbon analyses of diprotodontid teeth from the Mayne Site returned ages between c.30,000 and 60,000 years (Van Huet et al., 1998, Table 1). Van Huet, 1994, Van Huet, 1999 argued that the bones were removed from (as-yet-unidentified) locations scattered across the surrounding landscape and transported by mass flow to be deposited in the swamp. Further excavation and subsurface survey with an auger indicated that the Mayne Site bone bed was situated at the bottom of a small gully aligned with the present-day surface depression (Peel, 2001). The ‘mass flow’ event proposed by Van Huet slowed on entering the swamp, subsequently filling the gully with sediment and bones (Peel, 2001).

As the only place where evidence for sediment ‘reworking’ was found is within this excavated channel, and given that the Lancefield site is located in a broad shallow valley, it appears unnecessary to invoke ‘mass flow’ as an explanation for the formation of the deposit in this channel in this part of the site. An alternative interpretation, assessed here, is that the channel excavated by Van Huet was formed during a period of high spring discharge. In this scenario, spring flow caused the displacement of sediment, and the abrasion and displacement of bone within the bone bed.

The research preceding our investigation had returned widely ranging age estimates, and indicated quite different bone accumulation scenarios. Direct dating had been problematic. The Mayne and South Site bones contained insufficient bone collagen for radiocarbon dating (as also reported for Classic Site bones; Gillespie et al., 1978). It was clear that radiocarbon determinations using bone apatite were imprecise. Furthermore, ESR dating methods were compromised because the high water table prevented measurement of background soil radiation for calculation of results (Van Huet et al., 1998).

The bone beds at Lancefield formed either by deposition of fresh bones in situ (i.e. in the locations from which they were recovered), and/or reworking from other, older deposits. Based on previous assessments of the site and the processes that operate at open sites, five site formation models (Table 2) may have operated either in isolation or in combination: (1) waterhole drought death assemblage (Horton, 1976, Horton and Samuel, 1978, Gillespie et al., 1978); (2) culturally-mediated accumulation (Gillespie et al., 1978); (3) carnivore-mediated accumulation (Horton and Wright, 1981); (4) mass flow (Van Huet, 1994, Peel, 2001); and (5) fluvial transport (not discussed for Lancefield but see Behrensmeyer, 1988). Mass flow and fluvial transport are geomorphological processes that can re-work bones across, onto, or off sites. The other three processes imply in situ bone deposition within a site.

The taphonomic criteria assessed at Lancefield are summarised in Table 2. With respect to fluvial transport, two taphonomic modes of bone deposition in channels are important: channel-lag and channel-fill (Table 3; Behrensmeyer, 1988: 192). Bones from channel-lag are usually allochthonous deriving from various parts of a drainage basin, while bones from channel-fill are mostly autochthonous – derived from deposition of animal carcases at or near the channel deposit.

Given the above considerations, understanding the context within which the animals died and the bones accumulated was critical. The approach used here hinged on examination of geomorphic processes, direct dating of bones, (dating of associated sediments), taphonomy (including population age profiles), and consideration of regional environmental changes.

Section snippets

Excavation

The late summer excavation seasons in 2004 and 2005 followed several drought years which resulted in a lower water table, dry trenches and easy access to the bone bed. Systematic excavation involved extending the Classic and Mayne Site trenches (Fig. 3, Fig. 4). No attempt was made to revisit the South Site to the north of the Peel trench, as consultation with previous investigators (Van Huet and Peel), and local residents involved in its discovery indicated that little remained of the fossil

Stratigraphy and sediments

The 2004–2005 investigations confirmed the original stratigraphic sequences (Gillespie et al., 1978, Van Huet, 1999, Fig. 5, Fig. 6). Description of the Classic Site stratigraphy draws on Gillespie et al. (1978) and Macumber (1991), and that of the Mayne Site on Van Huet (1994, 1998; Van Huet et al., 1998) and Peel (2001).

Unit 1, 0–60 cm depth (0–50 cm below surface at the Mayne Site): brown humic sediment, matted roots, and contains historical artefacts and European tree pollen (Ladd, 1976).

Bone accumulation scenarios

Re-excavation combined with re-analysis of samples excavated from across the entire Lancefield Swamp fossil deposit indicates that the bone assemblages are dominated by mature adults of a single taxon, Macropus giganteus titan, the Pleistocene ancestor of the modern eastern grey kangaroo, Macropus giganteus. At two of the three sample locations, bones are damaged and disarticulated, but not re-worked. The third location – the Van Huet trench–contains bones, the characteristics of which are

Conclusion

The re-excavation and redating of the Lancefield sequence confirms previous interpretations that megafauna deaths occurred during a period of climatic variability and drying. The original dating study indicated that these events unfolded during the Last Glacial Maximum (30–19 ka BP), but the OSL, ESR and 14C analyses undertaken here suggest that they occurred at a much earlier time. Severe droughts resulted in localised mass deaths as early as 90 ka BP and as late as 40 ka BP, which are

Contributions

JD: led project, collected and collated data, drafted and edited ms.

MC: OSL and sedimentological studies.

RG: ESR studies.

BH: taphonomic analyses.

KL: megafauna age profiles.

JF: proposed project; supervised BH; edited ms.

Acknowledgments

We thank members of the Wurundjeri Tribe Land and Heritage Council for the endorsement of our research, which was funded by the Australian Research Council (ARC DP0342843), a University of Sydney Sesquicentenary R&D Grant held by JD, and a Carlyle Greenwell Bequest to BN. We are grateful for the support of the Lancefield Park Management Committee, Office of Aboriginal Affairs Victoria, Department of Environment, Land, Water and Planning (Vic.) and Shire of Macedon Ranges. We are indebted to

References (124)

  • S. Eggins et al.

    238U, 232Th profiling and U-series isotope analysis of fossil teeth by laser ablation ICPMS

    Quat. Sci. Rev.

    (2003)
  • S.M. Eggins et al.

    In situ U-series dating by laser-ablation multi-collector ICPMS: new prospects for Quaternary geochronology

    Quat. Sci. Rev.

    (2005)
  • J. Field et al.

    Chronological overlap between humans and megafauna in Sahul (Pleistocene Australia–New Guinea): a review of the evidence

    Earth-Sci. Rev.

    (2008)
  • J. Field et al.

    Looking for the archaeological signature in Australian megafaunal extinctions

    Quat. Int.

    (2013)
  • M. Fillios et al.

    Investigating human and megafauna co-occurrence in Australian prehistory: mode and causality in fossil accumulations at Cuddie Springs

    Quat. Int.

    (2010)
  • D. Fink et al.

    The ANTARES AMS facility at ANSTO

    NIM B

    (2004)
  • J. Garvey

    Bennett’s wallaby (Macropus rufogriseus) bone marrow quality vs quantity: evaluating human decision making and seasonal occupation in late Pleistocene Tasmania

    J. Archaeol. Sci.

    (2011)
  • D.K. Grayson et al.

    Clovis hunting—revisited

    J. Archaeol. Sci.

    (2015)
  • R. Grün

    Electron Spin Resonance (ESR) dating

    Quat. Int.

    (1989)
  • R. Grün et al.

    ESR and U-series analyses of faunal material from Cuddie Springs, NSW, Australia: implications for the timing of the extinction of the Australian megafauna

    Quat. Sci. Rev.

    (2010)
  • R. Grün et al.

    Laser ablation U-series analysis of fossil bones and teeth

    Palaeogeogr. Palaeoclimatol. Palaeoecol.

    (2014)
  • G. Haynes

    Mass deaths and serial predation: comparative taphonomic studies of modern large mammal death sites

    J. Archaeol. Sci.

    (1988)
  • A. Lenoble et al.

    Fabric of Paleolithic levels: methods and implications for site formation processes

    J. Archaeol. Sci.

    (2004)
  • J. Lomax et al.

    The onset of dune formation in the Strzelecki Desert, South Australia

    Quat. Sci. Rev.

    (2003)
  • B.G. Markey et al.

    A new flexible system for measuring thermally and optically stimulated luminescence

    Radiat. Meas.

    (1997)
  • S.D. Mooney et al.

    Late quaternary fire regimes of Australasia

    Quat. Sci. Rev.

    (2011)
  • A.S. Murray et al.

    Measurement of the equivalent dose in quartz using a regenerative-dose single-aliquot protocol

    Radiat. Meas.

    (1998)
  • A.S. Murray et al.

    Luminescence dating of quartz using an improved single-aliquot regenerative-dose protocol

    Radiat. Meas.

    (2000)
  • G.C. Nanson et al.

    Alluvial evidence for major climate and flow regime changes during the middle and late Quaternary in eastern central Australia. Geomorphology

  • J.R. Prescott et al.

    Cosmic ray contributions to dose rates for luminescence and ESR dating: large depths and long-term time variations

    Radiat. Meas.

    (1994)
  • M.J. Aitken

    An Introduction to Optical Dating: the Dating of Quaternary Sediments by the Use of Photon-stimulated Luminescence

    (1998)
  • J. Allen et al.

    Both half right: updating the evidence for dating first human arrivals in Sahul

    Aust. Archaeol.

    (2014)
  • A.D. Barnosky et al.

    Assessing the causes of Late Pleistocene extinctions on the continents

    Science

    (2004)
  • A.K. Behrensmeyer

    Terrestrial vertebrate accumulations

  • L.R. Binford

    Bones − Ancient Men and Modern Myths

    (1980)
  • S. Boggs

    Principles of Sedimentology and Stratigraphy

    (2001)
  • S. Bowdler

    Comment: Pleistocene faunal loss: implications of the aftershock for Australia’s past and future

    Archaeol. Ocean.

    (1990)
  • J.M. Bowler et al.

    Human remains from lake Mungo: discovery and excavation lake Mungo III

  • J.M. Bowler et al.

    New ages for human occupation and climatic change at Lake Mungo, Australia

    Nature

    (2003)
  • B.W. Brook et al.

    Lack of chronological support for stepwise prehuman extinctions of Australian megafauna

    Proc. Natl. Acad. Sci.

    (2013)
  • T.J. Cohen et al.

    Hydrological transformation coincided with megafaunal extinction in central Australia

    Geology

    (2015)
  • T.J. Cohen et al.

    Continental aridification and the vanishing of Australia’s megalakes

    Geology

    (2011)
  • A. Cooper et al.

    Abrupt warming events drove Late Pleistocene Holarctic megafaunal turnover

    Science

    (2015)
  • M.-A. Courty et al.

    Soils and Micromorphology in Archaeology

    (1989)
  • M.L. Cupper

    Last glacial to Holocene evolution of semi-arid rangelands in southeastern Australia

    Holocene

    (2005)
  • J.C. Davis

    Statistics and Data Analysis in Geology

    (2002)
  • T.J. Dawson

    Kangaroos: Biology of the Largest Marsupials

    (2002)
  • L. Dawson et al.

    Taxonomic and phylogenetic status of living and fossil kangaroos and wallabies of the genus Macropus Shaw (Macropodidae: Marsupialia), with a new subgeneric name for the larger wallabies

    Aust. J. Zool.

    (1985)
  • C. Denys

    Taphonomy and experimentation

    Archaeometry

    (2002)
  • J. Dortch

    Archaeology at Lancefield Swamp: Report of the February 2004 Excavations (Community Report)

    (2004)
  • Cited by (0)

    1

    Present address: Centre for Rock Art Research + Management, M257, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia.

    2

    TrEnD Lab, Curtin University, Bentley, WA 6101, Australia.

    3

    Present address: Research Centre of Human Evolution, Environmental Futures Research Institute, Griffith University, 170 Kessels Road, Nathan, QLD 4111, Australia.

    4

    Present address: Department of Archaeology and History, School of Humanities and Social Sciences, La Trobe University, Bundoora, VIC 3086, Australia.

    5

    Present address: 35 Bon Accord Avenue, Bondi Junction, NSW 2022, Australia.

    6

    Present address: School of Biological, Earth and Environmental Sciences, University of New South Wales, Kensington, NSW 2052, Australia.

    View full text