Quaternary Science Reviews 37 (2012) 38e47

Contents lists available at SciVerse ScienceDirect

Quaternary Science Reviews

journal homepage: www.elsevier.com/locate/quascirev

Man and megafauna in Tasmania: closing the gap

Richard Gillespie a,b,*, Aaron B. Camens c, Trevor H. Worthy d, Nicolas J. Rawlence e,f, Craig Reid g, Fiona Bertuch h, Vladimir Levchenko h, Alan Cooper e a Centre for Archaeological Science, School of Earth and Environmental Sciences, University of Wollongong, Wollongong, NSW 2522, Australia b Division of Archaeology and Natural History, School of Culture, History and Language, Australian National University, Canberra, ACT 0200, Australia c School of Earth and Environmental Sciences, University of Adelaide, Adelaide, South Australia 5005, Australia d Department of Biological, Earth and Environmental Sciences, University of New South Wales, Sydney, NSW 2052, Australia e Australian Centre for Ancient DNA, School of Earth and Environmental Sciences, University of Adelaide, Adelaide, South Australia 5005, Australia f Department of Earth and Ocean Sciences, University of Waikato, Private Bag 3105, Hamilton 3240, New Zealand g Queen Victoria Museum and Art Gallery, PO Box 403, Launceston, Tasmania 7250, Australia h Institute for Environmental Science, Australian Nuclear Science and Technology Organisation, Locked Bag 2001, Kirrawee DC, NSW 2232, Australia article info abstract

Article history: Recent discussion on the late Pleistocene extinction of the Australian megafauna has revolved around Received 9 September 2011 interpretation of several key fossil sites in Tasmania. It has been suggested that humans did not arrive in Received in revised form Tasmania until after the megafauna became extinct, or did not hunt now extinct megafauna, and 27 November 2011 therefore that humans cannot be implicated in the extinctions. Radiocarbon results from these sites Accepted 11 January 2012 indicate that the youngest extinct megafauna are close to charcoal ages from the oldest archaeological Available online 10 February 2012 deposits, although difficulties have arisen in establishing chronologies because most relevant sites have ages near the limit for radiocarbon analysis. Keywords: d13 d15 Megafauna We report a series of new radiocarbon ages, C, N and C:N ratios on collagen and dentine fractions Radiocarbon from skeletal remains in the Mount Cripps karst area and the Mowbray Swamp, both in northwestern Collagen Tasmania, and discuss the reliability of ages from these and other sites. We also report the discovery of an Extinction articulated Simosthenurus occidentalis skeleton at Mt Cripps, that represents only the second directly- Australia dated extinct megafaunal taxon with a reliable age <50 ka cal BP from Tasmania. Our results suggest that C:N ratios measured on collagen or dentine are not an infallible guide to radiocarbon age reliability. We confirm previous reports of a temporal overlap between the megafaunal and archaeological records in Tasmania, but the presence of archaeological evidence and megafauna with the same age at the same site has not yet been demonstrated. At least two megafaunal taxadthe now- extinct Protemnodon anak and a giant Pleistocene form of the extant giganteusdwere still present in Tasmania after 43 ka, when human crossing of the Bassian landbridge from mainland Australia first became sustainable. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction mainland Australia’s large-bodied Pleistocene fauna became extinct (e.g. Roberts et al., 2001; Miller et al., 2005; Grün et al., Early human occupation of mainland Australia in the 50e40 ka 2008; Prideaux et al., 2010). This temporal coincidence has stimu- age range is well-documented, often with multi-method dating; lated hotly-debated, if largely inconclusive, argument concerning notable examples include Lake Mungo in the southeast (Bowler the sequence, timing and causes of the Australian megafaunal et al., 2003), Devil’s Lair in the southwest (Turney et al., 2001), extinctions (e.g. Gillespie et al., 2006; Brook et al., 2007; Field et al., and in the northwest at Carpenter’s Gap (O’Connor, 1995) and Riwi 2008). (Balme, 2000). Around this same 50e40 ka interval, the majority of Several aspects of the Australian debate are similar to those concerning the North American extinctions (e.g. Grayson and Meltzer, 2003; Fiedel and Haynes, 2004), with one significant difference: among the large number of sites bridging the Rancho- * Corresponding author. Centre for Archaeological Science, School of Earth and labrean Termination (Haynes, 2008), there are at least 14 well- Environmental Sciences, University of Wollongong, Wollongong, NSW 2522, Aus- “ ” tralia. Tel.: þ61 2 6677 9500. dated elephant kill sites in the North American record where E-mail address: [email protected] (R. Gillespie). both directly-dated extinct taxa and associated subsistence

0277-3791/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.quascirev.2012.01.013 R. Gillespie et al. / Quaternary Science Reviews 37 (2012) 38e47 39 archaeology of the same age occur in the same stratigraphic layers megafauna. But despite the analysis of many thousands of bone (e.g. Grayson and Meltzer, 2002; Surovell and Waguespack, 2008). fragments, no skeletal remains of any extinct taxon have been Australia has no known sites with these attributes, and conse- found in Tasmanian archaeological sites (Cosgrove et al., 2010), an quently the debate is reduced to polarised views based on the observation that we discuss below. Reliable dating of Tasmanian contents of a very small number of either fossil or archaeological archaeological and palaeontological sites has been limited due to sites in the appropriate age range. The island state of Tasmania, these sites being at or near the limit of radiocarbon dating, adding periodically connected to mainland Australia during sea level to the difficulties in establishing the provenance of sediments used lowstands, potentially offers an interesting test for late Pleistocene for OSL dating of key sites (Cosgrove et al., 2010). extinction hypotheses. For this study, we measured 14C ages, stable C and N isotopes, Lambeck and Chappell (2001) determined that the Bass Strait and C:N ratios on collagen and dentine fractions isolated from the between mainland Australia and Tasmania was bridged w43 ka, skeletal remains of three extant and four extinct taxa allowing pedestrian human crossing for the first time and recovered from the Mt Cripps and Mowbray Swamp fossil sites in suggesting a significantly shorter overlap period for humans and northwestern Tasmania. We compare our directly-dated marsupial megafauna in Tasmania than on mainland Australia. This accords ages with previous results from Turney et al. (2008) and Cosgrove well with the oldest Tasmanian archaeological sites so far docu- et al. (2010), assessing their 14C age reliability and any temporal mented, Warreen Cave (Allen, 1996) and Parmerpar Meethaner overlap with the existing Tasmanian archaeological record. The (Cosgrove, 1995), which have maximum calibrated radiocarbon locations of fossil and archaeological sites in Tasmania discussed (14C) ages measured on charcoal ca. 40e38 ka cal BP. Compared here are shown in the Fig. 1 map. with mainland Australia, the extinct megafauna record from Tas- mania is depauperate, probably including only Zygomaturus trilo- 2. Materials and methods bus, Palorchestes azael, Metasthenurus newtonae, Simosthenurus occidentalis, Protemnodon anak, Thylacoleo carnifex and Mega- Table 1 shows details of bones and teeth sampled for this study libgwilia sp. (Turney et al., 2008) plus a giant form of the extant from Mt Cripps and Mowbray Swamp, which are housed at the Macropus giganteus. Queen Victoria Museum and Art Gallery (QVMAG) in Launceston With the first defendable direct 14C age measurements on and the Australian Centre for Ancient DNA (ACAD) in Adelaide. extinct Australian marsupial bones, ca. 43e41 ka cal BP on Collagen content of the bones was estimated using a ninhydrin Protemnodon from Mt Cripps, Turney et al. (2008) suggested that method on decalcified and hydrolysed bone samples, compared Tasmania may have been the last refuge for the Australian with similarly-prepared aliquots from a modern Wallabia bicolor

Melbourne 38°S

Landbridge ~43 ka Bass Strait 40°S

Mowbray Swamp

Mt Cripps Launceston

Parmerpar Meethaner 42°S ORS 7 Pallawa Trounta JF154, JF155, Warreen Cave Ultimate Cave, Titan’s Shelter Nunamira Hobart Bone Cave 0 100 km

144°E 146°E 148°E

Fig. 1. Map showing the location of Tasmanian fossil sites (red filled circles) and archaeological sites (white filled circles) discussed in the text, with an approximate contour for the Bassian landbridge w43 ka (modified from Cosgrove et al., 2010; Lambeck and Chappell, 2001). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 40 R. Gillespie et al. / Quaternary Science Reviews 37 (2012) 38e47

Table 1 Lastly, all fractions were freeze dried before combustion, graphi- Bones and teeth analysed in this study, showing sample location, museum accession tisation and AMS radiocarbon analysis using standard methods at numbers, body part and sample ID used in subsequent tables; taxa marked with an ANSTO (Fink et al., 2004) or the ORAU (Bronk Ramsey et al., 2004). asterisk are extinct megafauna. Separate sub-samples of each fraction were combusted in a coupled Location Species Body part QVM No. Sample ID elemental analyser-isotope ratio mass spectrometer system for Mt Cripps Simosthenurus upper I1 frag. 2009 GFV:0002 MCSt-1 measurement of d13C, d15N and atomic C:N ratios. (CP222) occidentalis* s fi * We used the 2 sample signi cance and pooled mean age tools Mt Cripps Protemnodon anak L tibia 2001 GFV:2 MCP-1 14 (CP213) in CALIB 6.0 (Stuiver et al., 2010) to compare conventional C ages L femur 2001 GFV:40 MCP-2 from the different fractions prepared in this study. Those ages L lower I1 2011:GFV:1 MCP-3 considered acceptable were calibrated with OxCal 4.1 (Bronk Thylogale billardierii L tibia 2001 GFV:68 MCT-1 Ramsay, 2009) using the IntCal09 atmospheric data in Reimer Vombatus ursinus R femur 2001 GFV:30 MCV-1 et al. (2009). upper P3 2011:GFV:2 MCV-2 Sarcophilus harrisii R femur 2001 GFV:39 MCS-1 Mowbray Zygomaturus trilobus* rib frag. 1992 GFV:148 MSZ-1 3. Results and discussion Swamp femur frag. 1992 GFV:181 MSZ-2 Conventional radiocarbon ages, d13C, d15N and C:N ratios Palorchestes azael* femur frag. 2007 GFV:22 MSPal-1 measured on our Tasmanian marsupial samples, and on the radiocarbon bone standard VIRI sample E (a mammoth bone from tibia (Gillespie and Brook, 2006). Examined under 10 binocular Quartz Creek, Yukon; Scott et al., 2009) processed with our samples magnification, the decalcified bone samples from Mt Cripps and at ANSTO, are presented in Table 2. We note that C:N ratios for Mowbray Swamp exhibited acceptable collagen pseudomorphs collagen and gelatin fractions from these bones and teeth are all (Stafford et al., 1987), with collagen content ranging from 3 to 17% very similar, and all are within the acceptable range of 2.9e3.5 by weight of original bone; these were analysed at the Australian (DeNiro, 1985; Brock et al., 2007). Nuclear Science and Technology Organisation (ANSTO) in Sydney. The teeth, which yielded 1.9e4% dentine, were analysed at the 3.1. Standard bone Oxford Radiocarbon Accelerator Unit (ORAU), UK. Bones and teeth were processed using methods described in Stafford et al. (1991) Radiocarbon ages for our three independent preparations of and Brock et al. (2007) to prepare a subset of the following ultrafiltered gelatin from the VIRI E standard bone are statistically fractions: identical. Their pooled mean age of 39,430 310 BP is not signif- icantly different from the VIRI consensus age of 39,305 121 BP, rawC e cold decalcification: 4 C for several days, daily acid suggesting that the methods employed here produce results change consistent with best practice for late Pleistocene bone. coldFG e cold decalcification, gelatinisation, 0.45 m PTFE filtration 3.2. Mt Cripps stdUG e Oxford protocol: room temp. ABA, gelatinisation, Eezi- filter, Vivaspin-15 ultrafilter Collagen and gelatin fractions from Mt Cripps bones are pale coldUG e same except cold decalcification of bone fragments straw colour, similar to those from the well-preserved standard

Table 2 Analytical results for Mt Cripps and Mowbray Swamp skeletal remains, and the VIRI E standard bone, showing chemical fraction (see text for details), percent Modern C, conventional 14C age, stable C and N isotope ratios, and atomic C:N ratio; taxa marked with an asterisk are extinct megafauna. Data from the AMS facilities at ANSTO, Sydney, and Oxford, UK; n/a ¼ not available.

Site Taxon Sample ID Fraction Lab No. % Mod. C Conv. 14C age d13C d15N Atomic C:N measured (1s BP) (0.4&) (0.4&) ratio (0.2) Mt Cripps CP222 Simosthenurus MCSt-1 stdUG OxA-17143 0.39 0.05 44,500 1000 22.9 n/a 3.2 occidentalis * Mt Cripps CP213 Protemnodon anak * MCP-1 stdUG OZM081 1.08 0.04 36,350 300 24.3 2.7 3.4 MCP-1/2 coldUG OZM739 1.00 0.05 37,010 410 24.0 2.7 3.4 Protemnodon anak * MCP-2 stdUG OZM082 0.67 0.03 40,170 370 24.2 2.6 3.3 MCP-2/2 coldUG OZM744 0.72 0.06 39,660 680 23.8 2.7 3.3 MCP-2/4A rawC OZN266U1 0.92 0.04 37,670 360 23.9 2.9 3.3 MCP-2/4B coldFG OZN266U2 0.83 0.03 38,520 300 23.6 2.5 3.3 Protemnodon anak * MCP-3 stdUG OxA-22701 0.59 0.10 41,200 1400 22.1 2.5 3.3 Thylogale billardierii MCT-1 stdUG OZM083 0.99 0.04 37,080 330 24.0 2.3 3.5 Vombatus ursinus MCV-1 stdUG OZM736 0.70 0.04 39,900 470 23.9 3.8 3.3 Vombatus ursinus MCV-2 stdUG OxA-22702 0.53 0.10 42,000 1600 22.2 3.7 3.1 Sarcophilus harrisii MCS-1 stdUG OZM084 0.65 0.03 40,510 380 22.6 2.0 3.3

Mowbray Swamp Zygomaturus trilobus * MSZ-1 stdUG OZM085 1.54 0.04 33,510 210 21.9 5.0 3.3 MSZ-1/2 coldUG OZM740 1.05 0.06 36,570 460 21.7 5.0 3.3 MSZ-1/4 coldFG OZN265 1.38 0.05 34,410 300 21.5 4.4 3.3 Zygomaturus trilobus * MSZ-2 coldUG OZM737 0.29 0.02 46,810 560 20.6 6.2 3.3 MSZ-2/4 coldFG OZN264 0.50 0.03 42,500 480 20.4 6.3 3.4 Palorchestes azael * MSPal-1 coldUG OZM738 1.05 0.04 36,580 310 22.3 3.2 3.4 MSPal-1/2 coldFG OZN263 1.14 0.04 35,920 290 22.3 3.2 3.3

Quartz Ck., Yukon Mammuthus sp. * VIRI E-1 stdUG OZI812U1 0.75 0.07 39,350 740 20.4 n/a n/a (Reference Standard) VIRI E-2 stdUG OZI812U2 0.72 0.04 39,620 450 20.8 n/a n/a VIRI E-3 stdUG OZI812U3 0.76 0.05 39,220 530 20.7 n/a n/a R. Gillespie et al. / Quaternary Science Reviews 37 (2012) 38e47 41 bone. We compared four fractions prepared from Protemnodon Our oldest acceptable radiocarbon age of 44,500 1000 BP was bone MCP-2, finding that the ages on standard and cold-decalcified measured on an upper incisor fragment (MCSt-1) of an adult S. ultrafiltered gelatin were statistically identical (pooled mean age occidentalis skull from cave CP222 (see photo in Fig. 2B), part of an 40,050 325 BP), while the ages on raw collagen and 0.45 m PTFE articulated skeleton to be described in detail elsewhere (Camens filtered gelatin fractions from the same bone were significantly and colleagues, manuscript in prep.), and the second directly-dated younger and consequently rejected. The juvenile Protemnodon tibia extinct marsupial species from Mt Cripps. (MCP-1, see photo in Fig. 2A) also produced identical ages for standard and cold-decalcified ultrafiltered gelatin, with pooled 3.3. Mowbray swamp mean age 36,580 240 BP. We accept ages on all standard or cold- decalcified ultrafiltered gelatin fractions from samples MCP-1, Given previous unsuccessful attempts to find collagen in MCP-2, MCP-3, MCT-1, MCS-1, MCV-1 and MCV-2 (cave CP213), megafaunal bones and teeth from Lancefield Swamp and several and MCSt-1 (cave CP222). other mainland Australian fossil sites (Gillespie et al., 1978; Measurements on Protemnodon femur MCP-2 and the isolated Gillespie and Brook, 2006; Gillespie, unpubl. data), the presence of lower incisor MCP-3 are identical, and their pooled mean age of collagen in extinct marsupial bones from Mowbray Swamp was 40,110 320 BP is significantly older than the MCP-1 tibia mean unexpected; however, we found 2e4% collagen in samples MSZ-1, age. Ages on bone MCT-1 from the extant taxon Thylogale has the MSZ-2 and MSPal-1. These Z. trilobus and P. azael bones were same age as Protemnodon bone MCP-1, and samples from the extant described by Gill and Banks (1956) as “.various shades of taxa Sarcophilus (MCS-1) and Vombatus (MCV-1, MCV-2) have the brownish grey and grey, being stained presumably by the peat and same age as Protemnodon bone MCP-2. Thus we have shown that probably such iron as is presently being chemically reduced by the both extant and extinct species from Mt Cripps CP213 have a similar decomposing vegetable matter.” Viewed under 50 magnification, age range ca. 36e42,000 BP. thin brown-black filaments are visible within the translucent cold-decalcified collagen pseudomorph from rib bone MSZ-1 (see photo in Fig. 2C), and the gelatin fractions we prepared from these bones are also darkly stained. Radiocarbon ages on the gelatin preparations are scattered. Our comparison of three fractions from bone MSZ-1 found all ages to be significantly different, with the oldest age of 36,570 460 BP on cold-decalcified ultrafiltered gelatin. Cold-decalcified 0.45 m PTFE filtered gelatin yielded a younger age than ultrafiltered gelatin for bone MSZ-2, whereas these fractions from MSPal-1 have statisti- cally identical ages (pooled mean age 36,230 210 BP). Ages on wood and peat from the Mowbray Swamp (Gill and Banks, 1956; van de Geer et al., 1986) exhibit the same asymp- totic pattern as those from the nearby Pulbeena Swamp (Colhoun et al., 1982). As discussed by Chappell et al. (1996), a radiocarbon ‘event horizon’ exists at the limits of available decontamination and measurement technology, and when a set of 14C dates cluster near an asymptote, as they do for Mowbray Swamp, there is a high probability that the ages are unreliable. We interpret these wood and peat 14C results to mean that the sediments overlying the layers containing the extinct sampled here are >52,000 BP and could be much older. This does not necessarily imply that the bones are the same age as the peat layers in which they were found, nor that the three bones necessarily have the same true age. These low-lying peat swamps were drained by farmers early in the 20th century, and when sampled by van de Geer et al. (1986) the thickness of peat at Mowbray Swamp had decreased from more than 4 m to w2 m. While the expectation would be that humic and fulvic acids from the peat would contribute older carbon, the contamination we observe appears to be young. Even though C:N ratios of gelatin fractions from these Zygo- maturus and Palorchestes bones are all within the acceptable range, regardless of chemistry variations, we reject all our Mowbray Swamp results. The bones contain browneblack organic contami- nation that so far has proven impossible to completely remove using the usual acid and alkali solutions, several combinations of organic solvents, or the Oxford ultrafiltration protocol. Even our oldest result of 46,810 560 BP on bone MSZ-2 may seriously underestimate the true age.

3.4. Stable C and N isotopes Fig. 2. (A) Protemnodon anak juvenile tibia MCP-1 (QVM:2001 GFV: 2), scale bar equals 100 mm; (B) Simosthenurus occidentalis adult skull MCSt-1 (QVM:2006 GFV: 002), The d13C and d15N ratios measured on bone collagen/gelatin scale bar equals 100 mm; (C) decalcified collagen from Zygomaturus trilobus rib MSZ-1 (QVM:1992 GFV:148), note internal browneblack filaments of possible organic fractions prepared in this study (see Table 2 and Fig. 3) exhibit contamination, scale bar equals 1 mm. several interesting features: 42 R. Gillespie et al. / Quaternary Science Reviews 37 (2012) 38e47

AB Zygomaturus trilobus 7 Palorchestes azael Mowbray Swamp 6 Protemnodon anak Simosthenurus occidentalis 5 dentine Mt Cripps Thylogale billardierii 4 Vombatus ursinus

N (‰) 3 Sarcophilus harrisii 15 M. rufogriseus δ 2 M. giganteus (large morph) 1 Titan’s Shelter, JF155, Simosthenurus sp. JF154, Ultimate Cave 0 -25 -24 -23 -22 -21 -20 -25 -24 -23 -22 -21 -20 δ13C (‰) δ13C (‰)

Fig. 3. (A) d13C(0.4& PDB) values for bone gelatin or tooth dentine on taxa with ages >35 ka cal BP, diagonal dashed line separates northern (Mt Cripps, Mowbray Swamp) from southern sites (Titan’s Shelter, JF154, JF155, Ultimate Cave, Warreen Cave). Rectangles are Mt Cripps samples (the carnivore Sarcophilus harrisii has star shape), circles are Mowbray Swamp samples, triangles are from the southern sites; red filled shapes are extinct taxa, dark green filled shapes are extant taxa, light green shapes are M. giganteus (large morph). (B) Plot of available d13Cv.d15N(0.4& AIR) values for bone gelatin from herbivorous taxa with ages >40 ka cal BP; symbols as above. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

(1) the small amount of organic carbon contamination that occidentalis; one age on bone and one age on a tooth of the extant reduces the age of contaminated bone gelatin appears not to Vombatus ursinus; and one age each on bones of the extant T. bill- significantly affect the stable isotope or C:N ratios; ardierii and S. harrisii. These new 14C ages, which cluster in the 2s (2) the observed difference found between bone from the only calibrated range 40.9e49.8 ka cal BP, are listed in Table 3 and carnivorous marsupial sampled here (Sarcophilus harrisii, graphed as probability plots with median age in Fig. 4. d13C ¼22.6&) and the similarly-aged herbivorous marsupial d13 & bones from Mt Cripps ( C range 23.8 to 24.2 )isasex- 3.6. Comparison with previous Tasmanian marsupial results pected given their trophic level; (3) where measurements are available on bones and teeth from Turney et al. (2008) reported inter-laboratory comparisons on the same taxon, e.g. Protemnodon samples MCP-3 (tooth) and two Protemnodon bones from different individuals found in cave MCP-2 (bone) and Vombatus MCV-2 (tooth) and MCV-1 (bone), CP213 at Mt Cripps. They accepted the ages on bones processed d13 dentine gelatin has less negative C ratios than bone gelatin; using the standard ultrafiltration protocol at the ORAU, and rejected (4) the stable isotope signatures of our herbivore bones, regardless results on the same bones processed at the University of Wollon- of chemistry variations, cluster in two distinct geographic gong using a step-combusted gelatin method because they yielded groups; a third cluster is apparent when previous results from significantly younger ages. We concur, and also reject their result ’ the southern fossil sites Titan s Shelter, JF154, JF155, Ultimate on a Protemnodon bone from Titan’s Shelter processed with the Cave (Cosgrove et al., 2010) are included, shown in Fig. 3B. step-combusted gelatin procedure because the C:N ratio of 2.8 is below the acceptable range (Brock et al., 2007). Our two ages on the The stable isotope variations in these Tasmanian herbivores are Protemnodon left tibia sample MCP-1 are statistically the same as likely correlated with latitude, regional climate, water availability, the younger Protemnodon right tibia age accepted by Turney et al. elevation and temperature (e.g. DeNiro and Epstein, 1978; (2008) from cave CP213. The pooled mean age for these three Bocherens et al., 1995; Richards et al., 2002; Murphy et al., 2007). measurements on one individual at 36,430 190 BP Marsupial lifestyle may also be important, for example the dietary (41.5 0.4 ka cal BP) is the youngest for any Tasmanian extinct fl d13 changes re ected in C values of bone collagen from pouch young megafauna. to adult (Witt and Ayliffe, 2001). Adult body mass range Acceptable ages on our femur sample MCP-2 and tooth MCP-3, for the herbivores sampled here (e.g. Murray, 1991; Helgen et al., both Protemnodon from cave CP213, with pooled mean age 2006; van Dyck and Strahan, 2008) extends from the very large 40,110 320 BP (44.1 0.6 ka cal BP), are significantly older than e w 500 700 kg Z. trilobus and 300 kg P. azael at the coastal lowland our ages on MCP-1 and both tibia analysed by Turney et al. (2008), e Mowbray Swamp (10 15 m above sea level, van de Geer et al., indicating a third individual. The combined results from four taxa e e 1986), through the 97 136 kg S. occidentalis,122 153 kg P. anak suggests that the deposit spans 41e44 ka (see Table 3 and Fig. 4), e and the small 4 7kgThylogale billardierii at the upland Mt Cripps and we agree with Turney et al. (2008) that there is no conflict karst (355 masl, Turney et al., 2008). Our observation that samples between these calibrated 14C ages and the 36 3 ka OSL age on from sites cluster together regardless of taxa suggests that site- sand from the nasal cavity of one of the Mt Cripps Protemnodon fi speci c characteristics affect the stable isotope composition more skullsdthe OSL age has to be younger because the sand was fi than intrinsic species-speci c factors. Exploring the relationships necessarily deposited some time after that died. between the stable isotope composition of marsupial skeletal Cosgrove et al. (2010) reported inter-laboratory comparisons on remains with body mass, vegetation and climatic variables is bones from cave sites in south central Tasmania, and we agree with beyond the scope of this paper and will not be discussed further. their acceptance of ca. 41e44 ka cal BP ages on M. giganteus titan bones from Ultimate Cave and Titan’s Shelter square F, which have 3.5. Summary of results C:N ratios within the guidelines (Brock et al., 2007). However, Macropus titan Owen, 1838 inhabited a similar niche (Poole, 1982), We report acceptable radiocarbon ages on two extinct and three was synonymised with the extant taxon M. giganteus by Dawson extant marsupial taxa from Mt. Cripps: five ages on two bones and and Flannery (1985), and has since been treated as a synonymous one tooth of the extinct P. anak; one age on a tooth of the extinct S. giant Pleistocene morph (e.g. Roberts et al., 2001; Easton, 2006; R. Gillespie et al. / Quaternary Science Reviews 37 (2012) 38e47 43

Table 3 Radiocarbon ages on extant and extinct fauna, and archaeological charcoal from Tasmania. Conventional 14C ages were calibrated with OxCal 4.1 (Bronk Ramsay, 2009) using the IntCal09 atmospheric dataset in Reimer et al. (2009), showing the 2s calibrated age range and median calibrated age.

Location Taxon or material dated Lab No. Conv. 14C age Cal. age range Median age Ref. (1s BP) (2s cal BP) (ka cal BP) Fossil Sites Mt Cripps CP222 Simosthenurus occidentalis * OxA-17143 44,500 1000 49,840e46,000 47.8 1 Mt Cripps CP213 Protemnodon anak * OZM081 36,350 300 41,990e40,930 41.5 1 OZM739 37,010 410 42,490e41,240 41.9 1 OZM082 40,170 370 44,760e43,370 44.1 1 OZM744 39,660 680 44,750e42,680 43.7 1 OxA-22,701 41,200 1400 47,170e42,560 44.8 1 OxA-16417 36,200 300 41,900e40,740 41.3 2 OxA-16418 37,920 340 43,010e41,920 42.5 2 Vombatus ursinus OZM736 39,900 330 44,650e43,080 43.9 1 OxA-22702 42,000 1600 48,800e41,620 45.5 1 Thylogale billardierii OZM083 37,080 330 42,450e41,380 41.9 1 Sarcophilus harissii OZM084 40,510 380 45,060e43,650 44.4 1 Titan’s Shelter M. giganteus (large morph) CAMS-128558 37,650 470 42,990e41,620 42.3 3 CAMS-96368 36,210 390 41,980e40,570 41.3 3 CAMS-96369 35,610 360 41,590e39,700 40.8 3 Ultimate Cave CAMS-96367 39,530 580 44,540e42,730 43.6 3 Archaeological Sites Warreen Cave M. rufogriseus OZF415 31,660 250 36,670e35,280 36.1 3 Vombatus ursinus OZF416 30,610 350 36,270e34,600 35.1 3 charcoal Beta-46873/ETH-8509 31,610 370 36,720e35,130 35.9 4 charcoal Beta-46872/ETH-8508 27,160 250 31,790e31,080 31.4 4 charcoal Beta-42059/ETH-8501 30,210 300 35,330e34,080 34.8 4 charcoal Beta-42122B/ETH-7665B 34,790 510 41,030e38,740 39.9 4 charcoal ANUA-11710/143B 33,170 670 39,430e36,470 37.9 3 Parmerpar Meethaner charcoal Beta-68158/CAMS-10270 33,850 450 40,090e37,280 38.7 5 charcoal Beta-68159/CAMS-10271 33,260 420 38,860e36,810 37.9 5 charcoal Beta-61608/CAMS-6050 27,780 230 32,690e31,400 32.0 5 Nunamira Cave charcoal Beta-25880 27,770 420 33,050e31,270 32.0 6 charcoal Beta-28323 28,000 720 34,430e31,160 32.4 6 charcoal Beta-25881 30,420 690 36,480e33,490 35.0 6 ORS 7 charcoal Beta-23404/ETH-3724 30,840 480 36,420e34,660 35.5 6 Bone Cave charcoal Beta-29987 29,000 520 34,680e32,080 33.6 7 charcoal Beta-37547 28,330 720 34,450e31,430 32.7 8 Pallawa Trounta charcoal Beta-44081 29,800 720 36,270e32,550 34.3 9 charcoal Beta-62285 27,250 530 32,910e30,880 31.6 9

References: 1 This work; 2 Turney et al. (2008);3Cosgrove et al. (2010);4Allen (1996);5Cosgrove (1995);6Cosgrove (1989);7Allen (1989);8Holdaway and Porch (1996);9 Stern and Allen (1996).

Helgen et al., 2006; Webb, 2008, 2009). Given that the taxonomic Although the C:N ratio is within the guidelines, we reject this result distinction of M. titan has not yet been established through rigorous because the uncertainty is greater than usual for comparable taxonomic or biomolecular analysis, we regard it as a junior material (such as our results) and the pooled mean age of synonym of M. giganteus and not as an extinct taxon (species or 47,930 2220 BP is beyond the IntCal09 calibration range. All the subspecies) until such an analysis is undertaken. finite and non-finite 14C ages on bones from sites JF154 and JF155 The two youngest M. giganteus bones analysed by Cosgrove are likely to be unreliable because they cluster on an asymptote et al. (2010) from Titan’s Shelter square F are statistically the near the limit of the 14C method (Chappell et al., 1996); their true same age as the youngest Protemnodon bones from Mt Cripps age is almost certainly >50 ka cal BP. analysed in this work and by Turney et al. (2008). In our chemistry Table 3 lists the acceptable direct 14C age determinations on the variations on Protemnodon bone MCP-2, we found that the ultra- skeletal remains of extinct and extant marsupials from Tasmanian filtered gelatin ages were significantly older than those measured fossil sites, and 14C ages (mostly on charcoal) from Tasmanian on 0.45 m PTFE filtered gelatin and raw collagen fractions, the archaeological sites, showing calibrated ages in the range opposite pattern to that observed for M. giganteus bones by 30e50 ka cal BP. Fig. 4 shows the individual calibration plots and Cosgrove et al. (2010). This might be explained by improved median ages for all samples listed in Table 3, generated with OxCal quality control for bone dating at the ORAU and ANSTO since the 4.1 (Bronk Ramsay, 2009) using the IntCal09 atmospheric dataset in Cosgrove samples were processed; the Oxford ultrafiltration Reimer et al. (2009). protocol is time-consuming, precleaning the ultrafilters can be tricky, and degraded bones sometimes have little or no high- 4. Significance for Australian megafauna extinction molecular weight fraction (e.g. Higham et al., 2006; Brock et al., chronologies 2007; Hüls et al., 2007). Cosgrove et al. (2010) accept the older of two significantly Accepting multi-method chronologies supporting the first different ages on paired samples of M. giganteus bone from Titan’s human colonisation of greater Australia w50 ka (e.g. Brook et al., Shelter, following the standard radiocarbon judgement that the 2007), and that a sustainable Bassian landbridge w43 ka first older of two dates likely represents a more complete removal of gave humans pedestrian access to Tasmania (Lambeck and contamination. But they unaccountably accept the younger of two Chappell, 2001), we now address the nested set of questions con- statistically identical ages (50,600 3000 BP and cerning the evidence for direct interaction between humans and 44,700 3300 BP) on a Simosthenurus sp. bone from cave JF155. megafauna. 44 R. Gillespie et al. / Quaternary Science Reviews 37 (2012) 38e47 Archaeological Sites Archaeological Bone Cave Beta-37547 Beta-29987 Bassian Pallawa Trounta Landbridge Beta-62285 Beta-44081 Numamira Beta-25880 Beta-28323 Beta-25881 ORS7 Beta-23404/ETH-3724 Parmerpar Meethaner Beta-61608/CAMS-6050 Mt Cripps Beta-68159/CAMS-10271 OxA-17143 Beta-68158/CAMS-10270 OxA-22701 Wareen Cave OZM082 Beta-46872/ETH-8508 OZM744 Beta-42059/ETH-8501 OxA-16418 OZF416 Extinct OZM739 Beta-46873/ETH-8509 OZF415

Fossil Sites OZM081 OxA-16417 ANUA-11710/143B OZM083 Beta-42122B/ETH-7665B OZM736 OZM084

OxA-22702 Extant Titan’s Shelter CAMS-96369 CAMS-96368 CAMS-128558

CAMS-96367 Extant

50 45 40 35 30 Calibrated Radiocarbon Age (ka cal BP)

Fig. 4. Diagram showing probability plots for calibrated ages of samples from Tasmanian fossil and archaeological sites, with w43 ka age for the Bassian landbridge emergence (Lambeck and Chappell, 2001); colours for fauna as in Fig. 3, grey filled shapes are archaeological charcoal (see Table 3). Plots generated with OxCal 4.1 (Bronk Ramsay, 2009) using the IntCal09 atmospheric dataset (Reimer et al., 2009), short black vertical bars are median probability. Graph is truncated at 50 ka cal BP by the calibration limit, and only calibrated ages >30 ka cal BP are included; all plots have the same vertical scale, baselines were shifted to show locations.

4.1. Fossil site megafauna Our direct 14C dating confirms that Protemnodon survived at Mt Cripps until w41 ka cal BP, and adds a second extinct taxon, S. The number of Australian sites with acceptable ages from occidentalis, to the megafauna of Tasmania with reliable ages directly-dated extinct fauna is still very small compared with those <50 ka cal BP. Many of the Mt. Cripps karst area cave deposits, documenting the North American and Eurasian extinctions (e.g. including CP213 and CP222, contain in situ articulated or associated Haynes, 2008; Mellars and French, 2011), and issues of 14C dating skeletons, indicating that little or no post-depositional reworking reliability for late Pleistocene megafaunal assemblages in Australia has occurred, and when taken as a whole, record faunal change have attracted recent attention (e.g. Miller et al., 1999; Roberts across the megafaunal extinction boundary, with caves in the area et al., 2001; Gillespie et al., 2006; Gillespie, 2008; Turney et al., still acting as pitfall traps for the modern fauna (Camens, unpubl. 2008). However, unlike the North American extinction period data). For cave CP213, where there is essentially no stratigraphy, our ca. 14e12 ka, ages in the 50e40 ka cal BP range are pushing the directly-dated surface finds show that three extant species died in limits of the 14C method, and the glaciofluvial stratigraphic record the same ca. 45e41 ka cal BP period as two extinct Protemnodon available to constrain the age of late Pleistocene events is better individuals. developed in Europe than it is in Australia. Roberts et al. (2001) suggested, based on a statistical analysis of 4.2. Archaeological sites seven sites containing articulated skeletal remains with reliable OSL or U/Th ages <55 ka, that megafaunal extinction took place on The reliability of radiocarbon dating for many mainland the Australian mainland ca. 51e40 ka. This extinction window has Australian archaeological sites has also been disputed (e.g. Allen been supported by several recent studies, including continent-wide and Holdaway, 1995; Chappell et al., 1996; Bowler et al., 2003; extinction of the giant flightless bird Genyornis newtoni at 50 5ka Gillespie et al., 2006; Field et al., 2008; Grün et al., 2010). Some of (Miller et al., 1999), and reports that Zygomaturus, Protemnodon, the “chronometric hygiene” principles proposed by Spriggs (1989) sthenurine kangaroos and T. carnifex were among the last mega- have been applied to the 14C record from Tasmanian rock shelter faunal taxa to disappear (ca. 45 ka cal BP) from the fossil record at archaeological sites by Holdaway and Porch (1996), eliminating the Naracoorte Caves in southeastern South Australia (Pate et al., results with larger than usual error bars, non-finite ages, and ages 2002), and that Protemnodon and sthenurine kangaroos were also clearly inconsistent with others in the same site sequence; the the last to disappear from the fossil record in southwestern acceptable screened ages are listed in Table 3 and graphed in Fig. 4. Western Australia (Prideaux et al., 2010). Direct uranium series- The oldest published Tasmanian archaeological sites have cali- electron spin resonance (US-ESR) dating on megafauna tooth brated 14C ages on charcoal of 39.9 1.2 ka cal BP from Warreen enamel from several sites in South Australia (Grün et al., 2008), Cave (Allen, 1996) and 38.7 1.4 ka cal BP from Parmerpar Meet- from Cuddie Springs in New South Wales (Grün et al., 2010), and haner (Cosgrove, 1995), while charcoal from archaeologically sterile from Titan’s Shelter in Tasmania (Cosgrove et al., 2010), also found layers has been dated to 39,970 950 BP at Parmerpar Meethaner no extinct taxa younger than the Roberts et al. (2001) extinction and at Parawee to >42,000 BP (Cosgrove et al., 2010). These authors window, but a precise timing for the extinctions remains elusive. suggest that those dates provide evidence that humans were not in R. Gillespie et al. / Quaternary Science Reviews 37 (2012) 38e47 45

Tasmania before 40 ka, in fact they show only that humans were not analysis of published 14C ages on charcoal found that all samples in at those sites before w44 ka cal BP. It is also possible that older the same two units had an error-weighted mean age of archaeological sites became submerged when the Bassian land- 35.9 2.2 ka cal BP with no ageedepth relationship (Gillespie and bridge was flooded w14 ka (Lambeck and Chappell, 2001), as Brook, 2006), and (c) detailed geochemistry and direct US-ESR previously suggested for mainland Australia (e.g. Chappell, 1993; dating (Grün et al., 2010) found that the sand, charcoal and Oppenheimer, 2009). extinct megafauna remains cannot be in primary context with their Cosgrove et al. (2010) reported that “analysis of over 950,000 ages represented by the OSL ages. We support the geochronology bones from the eight well dated and stratigraphically controlled rather than archaeological speculation, contra Field et al. (in press), late Pleistocene archaeological sites from western, central and and reject Cuddie Springs as an extinct megafauna-human overlap northern Tasmania have not identified megafauna as part of human site. Nombe Rockshelter in Papua New Guinea (Gillieson and prey.” However, this large number is for recovered bone fragments, Mountain, 1983), Cloggs Cave in Victoria (Flood, 1973), Seton not whole bones, and since many of the bones were apparently Rockshelter on Island, South Australia (Hope et al., 1977), smashed to extract the nutritious marrow (Garvey, 2011), the and the Tedford Locality at Lake Menindee in western New South number of fragments from any one skeleton is also large. Very few Wales (Cupper and Duncan, 2006) are also removed from conten- of these bone fragments are from the deepest layers of the oldest tion, because they have no directly-dated extinct fauna and the sites, where the presence of extinct taxa might perhaps be antici- inclusion of older fossils in the dated archaeological horizons pated. From Warreen Cave, for example, Cosgrove and Allen (2001) cannot be precluded. estimated the minimum number of individual marsupials in the oldest layers, ca. 35e40 ka cal BP, at three Macropus rufogriseus and 4.3. Potential bias in sample selection and site conditions one V. ursinus. Allen (1996) notes that his excavation was “little more than a test pit” with all the limitations implied by “telephone The small number of well-dated sites may suggest a tapho- booth” archaeology, and the directly-dated extant fauna (Cosgrove nomic or sampling bias in the Australian fossil and archaeological et al., 2010) are significantly younger than the oldest charcoal. records. Recent modelling studies support the idea that younger Among the six archaeological sites with ages in excess of ages tend to be more common in datelists because “the longer it is 30 ka cal BP included here, mostly excavated and dated more than since something happened, the greater is the probability that 15 years ago, only one charcoal sample from Warreen Cave and one evidence of that event will have disappeared” (Johnson and Brook, from Parmerpar Meethaner have 2s age ranges extending older 2011). This bias in some measure represents the combined effects than 40 ka cal BP. Similarly, only two Protemnodon bones from Mt of site erosion, taphonomy, diagenesis, site destruction and Cripps and two M. giganteus bones from Titan’s Shelter have 2s age opportunistic sampling, which may be detected (and potentially ranges extending younger than 41 ka cal BP (see Fig. 4). This is corrected for) by comparing archaeological and geological dating a very small sample of both archaeological and fossil cave sites with results (e.g. Surovell et al., 2009). All of the Tasmanian archaeo- ages relevant to the contentious arguments about megafauna logical and fossil sites with acceptable ages listed in Table 3 and extinction. Cosgrove et al. (2010) conclude that the archaeological graphed in Fig. 4 are caves or rockshelters; the fossil sites are record implies that the Tasmanian megafauna were rare, or that pitfall traps, the archaeological sites presumably not. From their humans did not hunt them, or that the extinctions happened prior analysis of 822 Holocene 14C ages from 267 Australian rockshelter to human arrival in Tasmania. We contend that the third possibility sites, Johnson and Brook (2011) suggest that “.if the evidence has now been ruled out. dated by archaeologists is more easily erased than the material We dismiss the oft-repeated claims for late-surviving mega- used to date geological substrates, the taphonomic effect on fauna at several archaeological sites in greater Australia. Cuddie archaeology will be underestimated.” For the much older sites Springs in northern New South Wales, for example, has been widely relevant to the Australian late Pleistocene extinctions, contra promoted as the only site on mainland Australian where extinct Cosgrove et al. (2010), the absence of extinct taxa in known fauna and archaeology occur in the same stratigraphic layers (e.g. Tasmanian archaeological deposits may only suggest that sites Field et al., 2008; Fillios et al., 2010), but the large scale excavations from the earliest human occupation of Tasmania have yet to be and multiple dating methods applied do not support the conclusion identified or adequately sampled. We also note that large uncer- that bones and stones are the same age. Cuddie Springs has tainties are given for the ages and sill elevations used by Lambeck previously been rejected because (a) single-grain OSL dating of and Chappell (2001) to derive the w43 ka Bassian landbridge sand found multiple age distributions with maxima in two strati- emergence time, which may change with new data (J. Chappell, graphic units ca. 30 and 36 ka (Roberts et al., 2001), (b) statistical personal communication 2011).

Bassian Landbridge

Mt Cripps & Warreen Cave Extant (n=6)

Titan’s Shelter M. giganteus (large morph, n=4)

Mt Cripps Extinct (n=8) Archaeological charcoal (n=16)

50 45 40 35 30 Calibrated Radiocarbon Age (ka cal BP)

Fig. 5. Summed probability distributions for calibrated ages measured on Tasmanian extant and extinct marsupials, and on archaeological charcoal, colours as for Fig. 4, with w43 ka age for the Bassian landbridge emergence (Lambeck and Chappell, 2001). Fossil data from this work, Turney et al. (2008) and Cosgrove et al. (2010), archaeological data as in Table 3, generated with OxCal 4.1 (Bronk Ramsay, 2009) using the IntCal09 atmospheric dataset (Reimer et al., 2009); all plots have the same vertical scale, baselines were shifted for clarity. 46 R. Gillespie et al. / Quaternary Science Reviews 37 (2012) 38e47

4.4. Summary of the Tasmanian evidence Cripps and 14C dating at Oxford was provided by ARC Discovery grant 20101413 to AC (Australian Centre for Ancient DNA). There are clear parallels between the few well-dated fossil and archaeological records from the earliest period of human occupa- References tion on mainland Australia with the even smaller number of Tas- s manian records discussed here. The graph in Fig. 5 shows the 2 Allen, J., 1989. Excavations at bone cave, south central Tasmania, JanuaryeFebruary summed probability distributions for calibrated ages measured on 1989. Australian Archaeology 28, 105e106. Tasmanian extant fauna (n ¼ 6), large morph M. giganteus (n ¼ 4), Allen, J., 1996. Warreen cave. In: Allen, J. (Ed.), Report of the Southern Forests extinct megafauna (n ¼ 8) and archaeological charcoal (n ¼ 16), Archaeological Project. Site Descriptions, Stratigraphies and Chronologies, vol. 1. La Trobe University, Melbourne, pp. 135e167. generated with OxCal 4.1 (Bronk Ramsay, 2009); the vertical scale is Allen, J., Holdaway, S., 1995. The contamination of Pleistocene radiocarbon deter- identical for each plot, with baselines adjusted for clarity. Although minations in Australia. Antiquity 69, 101e112. they may not quite touch spatially, there is temporal overlap of Balme, J., 2000. Excavations revealing 40,000 years of occupation at Mimbi Caves in south central Kimberley of Western Australia. Australian Archaeology 51, 1e5. humans with the extinct P. anak and a large Pleistocene morph of Bocherens, H., Fogel, M.L., Tuross, N., Zeder, M., 1995. Trophic structure and climatic the extant M. giganteus. information from isotopic signatures in Pleistocene cave fauna of southern England. Journal of Archaeological Science 22, 327e340. Bowler, J.M., Johnston, H., Olley, J.M., Prescott, J.R., Roberts, R.G., Shawcross, W., 5. Conclusions Spooner, N.A., 2003. New ages for human occupation and climatic change at Lake Mungo, Australia. Nature 421, 837e840. The work reported here on fractions prepared from marsupial Brock, F., Bronk Ramsay, C., Higham, T., 2007. Quality assurance of ultrafiltered bone dating. Radiocarbon 49, 187e192. bones and teeth from Mt Cripps, Tasmania, suggests that variations Bronk Ramsay, C., 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51, of the Oxford ultrafiltration protocol yield acceptable radiocarbon 337e360. ages on well-preserved skeletal remains from Tasmanian sites near Bronk Ramsey, C., Higham, T., Leach, P., 2004. Towards high-precision AMS: prog- ress and limitations. Radiocarbon 46, 17e24. the limits of the radiocarbon method. Our attempts to obtain reli- Brook, B.W., Bowman, D.M.J.S., Burney, D.A., Flannery, T.F., Gagan, M.K., Gillespie, R., able ages using the same methods on contaminated bones from Johnson, C.N., Kershaw, P., Magee, J.W., Martin, P.S., Miller, G.H., Peiser, B., Mowbray Swamp were unsuccessful, and we conclude that an Roberts, R.G.,2007.Would theAustralian megafauna have become extinct if humans e atomic C:N ratio in the guideline 2.9e3.5 range is a necessary but had never colonised the continent? Quaternary Science Reviews 26, 560 564. Chappell, J., 1993. Late Pleistocene coasts and human migrations in the Austral not sufficient quality control criterion. region. In: Spriggs, M., Yen, D.E., Ambrose, W., Jones, R., Thorne, A., Andrews, A. No sites containing the coeval remains of both human occupa- (Eds.), A Community of Culture, the People and Prehistory of the Pacific. e tion and directly-dated extinct megafauna have been documented, Department of Prehistory, Australian National University, Canberra, pp. 43 48. Chappell, J., Head, J., Magee, J., 1996. Beyond the radiocarbon limit in Australian and no human-extinct taxon interaction has been unequivocally archaeology and Quaternary research. Antiquity 70, 543e552. demonstrated, anywhere in greater Australia. However, the data Colhoun, E.A., van de Geer, G., Mook, W.G., 1982. Stratigraphy, pollen analysis, and shows that some of the Mt Cripps now-extinct P. anak individuals paleoclimatic interpretation of Pulbeena Swamp, northwestern Tasmania. Quaternary Research 18, 108e126. dated in this study and by Turney et al. (2008), as well as some of Cosgrove, R., 1989. Thirty thousand years of human colonization in Tasmania: new the Titan’s Shelter extant M. giganteus (large morph) individuals Pleistocene dates. Science 243, 1706e1708. dated by Cosgrove et al. (2010), were alive w41 ka cal BPdafter Cosgrove, R., 1995. Late Pleistocene behavioural variation and time trends: the case e fi from Tasmania. Archaeology in Oceania 30, 83 104. humans could rst have walked across the emerging Bassian Cosgrove, R., Allen, J., 2001. Prey choice and hunting strategies in the Late Pleisto- landbridge from mainland Australia to Tasmaniadand therefore cene: evidence from southwest Tasmania. In: Anderson, A., O’Conner, S., Lilley, I. the potential for human-extinct taxon interaction did exist. (Eds.), Histories of Old Ages: Essays in Honour of Rhys Jones. Coombs Academic fi Publishing, Australian National University, Canberra, pp. 397e430. Furthermore, we con rm that a second extinct megafaunal taxon, S. Cosgrove, R., Field, J., Garvey, J., Brenner-Coltrain, J., Goede, A., Charles, B., Wroe, S., occidentalis, was present w48 ka cal BP at Mt Cripps, and given that Pike-Tay, A., Grün, R., Aubert, M., Lees, W., O’Connell, J., 2010. Overdone overkill: there is no evidence of this species being in decline prior to human the archaeological perspective on Tasmanian megafaunal extinctions. Journal of e arrival in Tasmania, it likely can also be counted among potential Archaeological Science 37, 2486 2503. Cupper, M.L., Duncan, J., 2006. Last glacial megafaunal death assemblage at Lake interactees. Menindee, southeastern Australia. Quaternary Research 66, 332e341. These geochronological results are closing the temporal gap Dawson, L., Flannery, T., 1985. Taxonomic and phylogenetic status of living and fossil separating the archaeological record from the natural record of kangaroos and wallabies of the genus Macropus Shaw (: Marsu- pialia), with a new subgeneric name for the larger wallabies. Australian Journal megafauna in Tasmania, and we speculate that such investigations of Zoology 33, 473e498. will reveal the first convincing evidence of human interaction with DeNiro, M.J., 1985. Postmortem preservation and alteration of in vivo bone collagen megafauna in Australia in the not too distant future. isotope ratios in relation to palaeodietary reconstruction. Nature 317, 806e809. DeNiro, M.J., Epstein, S., 1978. Influence of diet on the distribution of carbon isotopes in . Geochimica et Cosmochimica Acta 42, 495e506. Acknowledgements Easton, L.C., 2006. Pleistocene grey kangaroos from the Fossil Cave of Victoria Fossil Cave, Naracoorte, South Australia. Transactions of the Royal Society of South Australia 130, 17e28. We thank P. Darby and L. Gray for guiding us through the Mt. Fiedel, S., Haynes, G., 2004. A premature burial: comments on Grayson and Melt- Cripps karst, for assistance in the field we thank J. Metcalf, B. zer’s “Requiem for overkill”. Journal of Archaeological Science 31, 121e131. Llamas, J. Wood, F. Salt, H. Geraldene and I. Houshold, and for Field, J.H., Fillios, M., Wroe, S., 2008. Chronological overlap between humans and megafauna in Sahul (Pleistocene AustraliaeNew Guinea): a review of the assistance in identifying specimens we thank G. Medlin, G. Pri- evidence. Earth Science Reviews 89, 97e115. deaux and W. Boles. For access to reference specimens we thank M. Field, J., Wroe, S., Trueman, C.N., Garvey, J., Wyatt-Spratt, S., Looking for the Binnie and B. McHenry at the South Australian Museum, and L. archaeological signature in Australian Megafaunal extinctions. Quaternary Gershwin at QVMAG. We appreciate the assistance provided by A. International, in press Fillios, M., Field, J., Charles, B., 2010. Investigating human and megafauna co- Vains in sampling bones from the QVMAG collections, G. Jacobsen occurrence in Australian prehistory: mode and causality in fossil accumula- for access to laboratory facilities at ANSTO, and T. Higham for the tions at Cuddie Springs. Quaternary International 211, 123e143. Oxford data. For useful discussions we thank R. Cosgrove, E. Col- Fink, D.M., Hotchkis, Q., Hua, G., Jacobsen, A.M., Smith, U., Zoppi, D., Child, C., Mifsud, H., van der Gaast, Williams, A., Williams, M., 2004. The ANTARES AMS houn, D. Bowman, B. Murphy, C. Johnson, R. Roberts and J. Chappell. facility at ANSTO. Nuclear Instruments and Methods B 223e224, 109e115. Funding for RG’s travel to Launceston was provided by the School of Flood, J., 1973. Pleistocene human occupation and extinct fauna at Cloggs Cave, Earth and Environmental Sciences, University of Wollongong, and Buchan, southeast Australia. Nature 246, 303. ’ 14 Garvey, J., 2011. Bennett s wallaby (Macropus rufogriseus) marrow quality versus C dating at ANSTO was funded by AINSE grants 09/068 and 10/143 quantity: evaluating human decision-making and seasonal occupation in late to RG (University of Wollongong). Funding for field work at Mt Pleistocene Tasmania. Journal of Archaeological Science 38, 763e783. R. Gillespie et al. / Quaternary Science Reviews 37 (2012) 38e47 47

Gill, E.D., Banks, M.R., 1956. Cainozoic history of Mowbray Swamp and other areas Oppenheimer, S., 2009. The great arc of dispersal of modern humans: Africa to of north-western Tasmania. Records of the Queen Victoria Museum Launceston Australia. Quaternary International 202, 2e13. 6, 1e42. Pate, F.D., McDowell, M., Wells, R.T., Smith, A.M., 2002. Last recorded evidence for Gillespie, R., 2008. Updating Martin’s global extinction model. Quaternary Science megafauna at Wet Cave, Naracoorte, South Australia 45,000 years ago. Reviews 27, 2522e2529. Australian Archaeology 54, 53e55. Gillespie, R., Brook, B.W., 2006. Is there a Pleistocene archaeological site at Cuddie Poole, W.E., 1982. Macropus giganteus. Mammalian Species 187, 1e8. Springs? Archaeology in Oceania 41, 1e11. Prideaux, G.J., Gully, G.A., Couzens, A.M.C., Ayliffe, L.K., Jankowski, N.R., Jacobs, Z., Gillespie, R., Brook, B.W., Baynes, A., 2006. Short overlap of people and megafauna 2010. Timing and dynamics of Late Pleistocene extinctions in south- in Pleistocene Australia. Alcheringa Special Issue 1, 163e185. western Australia. Proceedings of the National Academy of Sciences USA 107, Gillespie, R., Horton, D.R., Ladd, P., Macumber, P.M., Rich, T.H., Thorne, A.R., 22157e22162. Wright, R.V.S., 1978. Lancefield Swamp and the extinction of the Australian Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk megafauna. Science 200, 1044e1048. Ramsey, C., Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Grootes, P.M., Gillieson, D., Mountain, M.J., 1983. Environmental history of Nombe rockshelter, Guilderson, T.P., Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Papua New Guinea. Archaeology in Oceania 18, 53e62. Kromer, B., McCormac, F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Grayson, D.K., Meltzer, D.J., 2002. Clovis hunting and large mammal extinction: Southon, J.R., Talamo, S., Turney, C.S.M., van der Plicht, J., Weyhenmeyer, C.E., a critical review of the evidence. Journal of World Prehistory 16, 313e359. 2009. IntCal09 and Marine09 radiocarbon age calibration curves, 0e50,000 Grayson, D.K., Meltzer, D.J., 2003. A requiem for North American overkill. Journal of years cal BP. Radiocarbon 51, 1111e1150. Archaeological Science 30, 585e593. Richards, M.P., Mays, S., Fuller, B.T., 2002. Stable carbon and nitrogen isotope values Grün, R., Wells, R., Eggins, S., Spooner, N., Aubert, M., Brown, L., Rhodes, E., 2008. of bone and teeth reflect weaning age at the medieval Wharram Percy site, Electron spin resonance dating of South Australian megafauna sites. Australian Yorkshire, UK. American Journal of Physical Anthropology 119, 205e210. Journal of Earth Sciences 55, 917e935. Roberts, R.G., Flannery, T.F., Ayliffe, L.K., Yoshida, H., Olley, J.M., Prideaux, G.J., Grün, R., Eggins, S., Aubert, M., Spooner, N., Pike, A.W.G., Muller, W., 2010. ESR and Laslett, G.M., Baynes, A., Smith, M.A., Jones, R., Smith, B.L., 2001. New ages for U-series analyses of faunal material from Cuddie Springs, NSW, Australia: the last Australian megafauna: continent-wide extinction about 46,000 years implications for the timing of the extinction of the Australian megafauna. ago. Science 292, 1888e1892. Quaternary Science Reviews 29, 596e610. Scott, E.M., Cook, G.T., Naysmith, P., 2009. A report on phase 2 of the fifth inter- Haynes Jr., C.V., 2008. Younger Dryas “black mats” and the Rancholabrean termi- national radiocarbon intercomparison (VIRI). Radiocarbon 52, 846e858. nation in North America. Proceedings of the National Academy of Sciences USA Spriggs, M., 1989. The dating of the island Southeast Asian Neolithic: an attempt at 105, 6520e6525. chronometric hygiene and linguistic correlation. Antiquity 63, 587e613. Helgen, K.M., Wells, R.T., Kear, B.J., Gerdtz, W.G., Flannery, T.F., 2006. Ecological and Stafford Jr., T.W., Brendel, K., Duhamel, R., Donohue, D., 1987. Study of bone dating evolutionary significance of sizes of giant extinct kangaroos. Australian Journal accuracy at the University of Arizona NSF accelerator facility for radioisotope of Zoology 54, 293e303. analysis. Radiocarbon 29, 24e44. Higham, T.F.G., Jacobi, R.M., Bronk Ramsay, C., 2006. AMS radiocarbon dating of Stafford Jr., T.W., Hare, P.E., Currie, L., Jull, A.J.T., Donohue, D.J., 1991. Accelerator ancient bone using ultrafiltration. Radiocarbon 48, 179e195. radiocarbon dating at the molecular level. Journal of Archaeological Science 18, Holdaway, S., Porch, N., 1996. Dates as data: an alternative approach to the 35e72. construction of chronologies for Pleistocene sites in southwest Tasmania. In: Stern, N., Allen, J., 1996. Pallawa Trounta shelter. In: Allen, J. (Ed.), Report of the Allen, J. (Ed.), Report of the Southern Forests Archaeological Project. Site Southern Forests Archaeological Project. Site Descriptions, Stratigraphies and Descriptions, Stratigraphies and Chronologies, vol. 1. La Trobe University, Mel- Chronologies, vol. 1. La Trobe University, Melbourne, pp. 169e193. bourne, pp. 251e277. Stuiver, M., Reimer, P.J., Reimer, R., 2010. CALIB Radiocarbon Calibration (Version Hope, J.H., Lampert, R.J., Edmondson, E., Smith, M.J., van Tets, G.F., 1977. Late 6.0) Downloaded from. http://calib.qub.ac.uk/calib/. Pleistocene faunal remains from Seton rock shelter, Kangaroo Island, South Surovell, T., Waguespack, N., 2008. How many elephant kills are 14? Clovis Australia. Journal of Biogeography 4, 363e385. mammoth and mastodon kills in context. Quaternary International 191, 82.e97 Hüls, M.C., Grootes, P.M., Nadeau, M.J., 2007. How clean is ultrafiltration cleaning of Surovell, T.A., Finley, J.B., Smith, G.M., Brantingham, P.J., Kelly, R., 2009. Correcting bone collagen? Radiocarbon 49, 193e200. temporal frequency distributions for taphonomic bias. Journal of Archaeological Johnson, C.N., Brook, B.W., 2011. Reconstructing the dynamics of ancient human Science 36, 1715e1724. populations from radiocarbon dates: 10,000 years of population growth in Turney, C.S.M., Bird, M.I., Fifield, L.K., Roberts, R.G., Smith, M.A., Dortch, C.E., Australia. Proceedings of the Royal Society B 278, 3748e3754. Grun, R., Lawson, E., Ayliffe, L.K., Miller, G.H., Dortch, J., Cresswell, R.G., 2001. Lambeck, K., Chappell, J., 2001. Sea level change through the last glacial cycle. Early human occupation at Devil’s Lair, southwestern Australia 50,000 years Science 292, 679e686. ago. Quaternary Research 55, 3e13. Mellars, P., French, J.C., 2011. Tenfold population increase in Western Europe at the Turney, C.S.M., Flannery, T.F., Roberts, R.G., Reid, C., Fifield, L.K., Higham, T.H.G., Neandertal-to-modern human transition. Science 333, 623e627. Jacobs, Z., Kemp, N., Colhoun, E.A., Kalin, R.M., Ogle, N., 2008. Late-surviving Miller, G.H., Magee, J.W., Johnson, B.J., Fogel, M.L., Spooner, N.A., McCulloch, M.T., megafauna in Tasmania, Australia, implicate human involvement in their Ayliffe, L.K., 1999. Pleistocene extinction of Genyornis newtoni: human impact extinction. Proceedings of the National Academy of Sciences USA 105, on Australian megafauna. Science 283, 205e208. 12150e12153. Miller, G.H., Fogel, M.L., Magee, J.W., Gagan, M.K., Clarke, S.J., Johnson, B.J., 2005. van de Geer, G., Colhoun, E.A., Mook, W.G., 1986. Stratigraphy, pollen analysis and Ecosystem collapse in Pleistocene Australia and a human role in megafaunal palaeoclimatic interpretation of Mowbray and Broadmeadows Swamps, north extinction. Science 309, 287e290. western Tasmania. Australian Geographer 17, 121e133. Murphy, B.P., Bowman, D.M.J.S., Gagan, M.K., 2007. Sources of carbon isotope van Dyck, S., Strahan, R., 2008. of Australia. Reed New Holland, Sydney. variation in kangaroo bone collagen and tooth enamel. Geochimica et Cosmo- Webb, S., 2008. Megafauna demography and late Quaternary climatic change in chimica Acta 71, 3847e3858. Australia: a predisposition to extinction. Boreas 37, 329e345. Murray, P.F., 1991. The Pleistocene megafauna of Australia. In: Vickers-Rich, P., Webb, S., 2009. Late Quaternary distribution and biogeography of the southern Lake Monaghan, J.M., Baird, R.F., Rich, T.H. (Eds.), Vertebrate Palaeontology of Aus- Eyre basin (SLEB) megafauna, South Australia. Boreas 38, 25e38. tralasia. Pioneer Design Studio, Lilydale, pp. 1071e1164. Witt, G.B., Ayliffe, L.K., 2001. Carbon isotope variability in the bone collagen of red O’Connor, S., 1995. Carpenter’s Gap Rockshelter 1: 40,000 years of Aboriginal occu- kangaroos (Macropus rufus) is age dependent: implications for palaeodietary pation in the Napier Ranges, Kimberley, WA. Australian Archaeology 40, 58e59. studies. Journal of Archaeological Science 28, 247e252.