Archaeol. Oceania 28 (1993) 94-99 Research Report Humans and megafauna in a late Pleistocene environment from Cuddie Springs, north western New South Wales John Dodson, Richard Fullagar, Judith Furby, Robert Jones and Ian Prosser Abstract We report the first site in Australia with a dated association of human technology and megafauna, in a palaeoenvironmental context. There are few sites in Australia where evidence of humans and Pleistocene megafauna coincide (Flannery and Gott, 1984; Flood, 1983 151-159; Gillespie et al., 1978; Gorecki et al., 1984). Such coincidences are often poorly dated or undated due to their antiquity or stratigraphic context. Cuddie Springs contains several distinct concentrations of megafaunal bone, in addition to a scatter of bone through all the sediments examined. Most of the deposit is beyond radiocarbon dating but an upper portion of sediment has been dated between 19,000 and 30,000 B.P. Artefacts and increased charcoal appear about 30,000 B.P. and then have a continuous presence. The artefact assemblage includes grindstones with starch residues, ochre, a probable cylcon and stone artefacts with reworked edges containing blood and hair. These combine to provide evidence of plant and animal processing and cultural practices at 30,000 B.P. The bones and artefacts were deposited when Cuddie Springs was a shallow freshwater lake surrounded by a relatively arid shrubland, the lake then became ephemeral and the environment more arid as the glacial maximum approached. The unequivocal record of human antiquity in Australia extends to about 40,000 B.P. (Allen, 1989), which is near the limit of conventional radiocarbon dating and conse- quently invites debate (Bowdler, 1990; Hiscock, 1990; Roberts, et al., 1990). Whatever the occupation span of the continent may be, its antiquity necessitates con- sideration of humans in a variety of environmental contexts vastly different from today, including extreme aridity, full glacial climatic conditions and an overlap of many millenia with a number of species of extinct fauna (White and Flannery, 1991). Cuddie Springs (Figure 1), in north western New South Wales provides a window on both of these issues in a stratigraphic and hence temporal sequence, a situation rare for Australian prehistory. In addition, the site contains a Pleistocene palaeoecological record from Australia's semi-arid region where very few have been studied (Dodson, in press). E. JD, JF, IP: School of Geography, University of New South Wales, Sydney; RF, RJ: Australian Museum, Sydney. Ms receivedAugust 1992,accepted January 1993. 94 The Cuddie Springs fossil remains are found in a 2.5 ha pan surrounded by low relief Quaternary alluvial and aeolian deposits. A Eucalyptus largiflorens and micro- theca woodland marks out a larger 500 ha Cuddie Springs palaeo-lake floor, possibly fed by an ancient stream to the north, observable on air photographs. Today the pan receives surface runoff from the immediate vicinity and can have standing water for several months but is mostly dry. The regional vegetation is a Flindersia maculosa, E. populnea and Callitris columellaris woodland and a variety of floodplain forests and woodlands. The significance of the site for megafauna was established during excavations earlier this century (Anderson and Fletcher, 1934) but it was not until the present study that artefacts wereidentified in the deposits. Methods During two field periods (June and November, 1991) a pit was excavatedin five sections at the centre of the clay pan (Figure 2). The pit was initially 1m x 2m in size and excavated to a depth of 3m (SI and S2). Artefacts were first discovered in Unit 5 of S1. During the second field period the pit was extended to the east in three 1 x 0.50m sections (S3-S5) to a depth of l.6m. In S2-S5 and below 1.0m in SI sediment was removed by trowel, shovel and kanga hammer in spits (maximum depth 15cm)within or corresponding to stratigraphic units. Sievingand artefact collection procedures included bulk sampling of deposits, hand sorting on site, dry sieving (lOmm, 5mm) on site, wet sieving (lOmm, 5mm, 2mm, lmm) at UNSW and bulk removal of skeletal material in plaster jackets. Sediment from SI-S5 was described and sampled for radiocarbon, pollen, charcoal, bone identification, artefact and particle size analyses to examine the context and association with the earlier established bone layers. Radiocarbon dates were determined on acid and alkali washed charcoal and given extended counting time to reduce the statistical error. Pollen analyses were under- taken after using heavy liquid separation and acetolysis (Moore et al. 1991). Charcoal abundance was estimated using the point count method (Clark, 1982). Particle size analysis was conducted using the hydrometer technique and dry sieving in phi intervals. Sediment fractions are described using the Wentworth scale (Pettijohn, 1975). Residues on stone artefacts were examined using incident light microscopy, and analysed by Ames Hemastix and immunological dot blot tests using Staphylococcal protein A with silver enhancement (Loy and Wood, 1989; Loy and Hardy, 1992). Results Stratigraphy The basal sediments comprise a massive, uniform unit of well-sorted fine sands from 5.0m to 2.75m depth (Figure