Pleistocene seed-grinding implements from the Australian arid zone.

Fullagar, Richard ; Field, Judith

Seed-grinding technologies in prehistoric societies have been linked to the development of agriculture and complex societies (e.g. Smith 1995) and the increased incidence of seed-grinding implements to the earliest evidence of sedentism around 12,800 b.p. (although grinding slabs for processing tubers have been identified at Wadi Kubbaniya in Egypt as early as 19,000 b.p.: Hillman 1989; Wright 1994). Macroscopic study indicates these early grinding-stones are multifunctional plant processing implements, part of a broad spectrum revolution following the peak of the last glaciation (Wright 1994; Edwards & O'Connell 1995). Edwards & O'Connell see Australia as providing a venue for testing the hypothesis that the terminal Pleistocene climatic change was a catalyst for this broad spectrum revolution.

This paper reports new evidence from Cuddie Springs for the antiquity of seed-grinding in Australia, and discusses implications for the role of technology and plant-food processing in the settlement of arid environments. Limiting factors on resource exploitation include climate and resource abundance which can be evaluated from palaeoecological and archaeological evidence; and social processes, more difficult to assess from archaeological data (Edwards & O'Connell 1995). The Cuddie Springs site, in the arid zone of eastern Australia, provides a unique climatic, faunal and archaeological record for the region. It has been known for over a century as a fossil megafaunal location. Excavations in 1991 and 1994 established the presence of an archaeological record overlapping with fossil megafaunal remains; 33 grinding-stone fragments reported here were recovered from a 2x2-m excavation. Of these, 26 were recovered from stratified sediments dating from the present to more than 30,000 b.p. including 21 from levels where fossil megafauna were also found (units 1 to 4, [ILLUSTRATION FOR FIGURE 1 OMITTED], TABLE 1).

Cuddie Springs

Cuddie Springs is an ephemeral freshwater lake, approximately 2 km in diameter, in central northern New South Wales [ILLUSTRATION FOR FIGURE 2 OMITTED]. Within the semi-arid zone, it receives approximately 400 mm/year precipitation (Dodson et al. 1993). A claypan in the centre of the lake floor fills after local rain to a depth of approximately 30 cm. The vegetation on the grey soils of the lake floor comprises a eucalypt woodland (Coolabah and Blackbox) interspersed with Acacia stenophylla and Chenopodiaceae with lignum in the areas where semi-permanent swamps are found. The red soil plains surrounding the lake support different plant communities that include Callitris, Casuarina, Eremophila and Flindersia species (Furby 1995).

The archaeological record comprises stone, bone, ochre and other artefacts of human occupation beginning at approximately 1.7 m depth and continuing to the present land surface. Recent excavations revealed sequential occupation units with evidence of butchering megafauna and activities associated with a domestic campsite (Furby 1995). The association of megafaunal bones and stone artefacts is in a sealed unit of stratified lacustrine clays from 1-7 m to approximately I m depth below surface [ILLUSTRATION FOR FIGURE 1 OMITTED]. Fourteen radiocarbon determinations have been obtained for the archaeological levels (1-7 m to surface), and the lowest levels date to [greater than]30,000 b.p (Furby 1995) (TABLE 1). The faunal record extends to at least 10 m depth, beyond the limits of the radiocarbon technique. Megafauna are not found in the upper 1 m of deposit.

Six archaeological units have been identified [ILLUSTRATION FOR FIGURE 1 OMITTED]. Unit 1 has been identified as a butchering location, formed when Cuddie Springs experienced extended lake full conditions. Units 2 and 3 represent periods of increasingly irregular inundation with extended dry conditions and a corresponding increase in the accumulation of stone, bone and other artefacts of human occupation. Unit 4, a dense concentration of stone, bones and charcoal, represents a deflation surface. Unit 5 is composed of undifferentiated clays with stone artefacts [TABULAR DATA FOR TABLE 1 OMITTED] and bone (Dodson et al. 1993; Furby 1995). A surface collection of stone artefacts from the Cuddie Springs lake floor comprises unit 6, a Late Holocene sample. Grinding-stones, first appearing at the interface between units 1 and 2, are present in subsequent archaeological units. Charcoal from within the deflation surface at 1.05-1 m (unit 4) has returned an age estimation of approximately 28,000 b.p. for this, the highest level in the deposit where megafauna are found. Unit 5 has been dated to [less than]19,000 b.p.

Methods

Our methodology is based on four lines of argument: the theoretical argument (Edwards & O'Connell 1995) that the development of seed-grinding technologies is related to environmental stress, such as increasing aridity during the last Glacial; artefact morphology, on which previous researchers have mostly depended; use-wear, the distinctive wear patterns arising from processing starchy and siliceous plant materials; residues, particularly those with distinct microscopic structures, on the utilized surfaces of artefacts.

Evidence for climatic and environmental change (Dodson et al. 1993) depends on the pollen record supplemented by sedimentological and palaeontological evidence from the site itself.

Artefact morphology is presented through Smith's typology of implement categories which includes millstones, mullers, mortars and pestles, and amorphous grinding-stones in which particular morphological types - e.g. millstones and mullers - are distinctly related to seed processing (Smith 1985; 1986). The overall implement morphology for very small grinding-stone fragments from Cuddie Springs could not be reconstructed, although it was possible to determine if surfaces were convex or concave, indicating use as upper or lower grinding implements, respectively.

Use-wear is described here in terms of the main recognised forms - scarring, striations, rounding and polish - focussing particularly on polish. Microscopic study of experimental and ethnographic grinding-stones indicates that polish development on weakly cemented sandstone is limited in extent by the constant abrasion of quartz grains (R. Fullagar personal observations). Many small areas of developed polish with numerous striations commonly occur on upper and lower grinding-stones. With more strongly cemented sandstone, as for quartzite, developed polish is extensive and similar to polish on flint sickle blades. Distinctive features of polish on stones used to process siliceous plant materials include a very flat, highly reflective surface, often featureless except for [TABULAR DATA FOR TABLE 2 OMITTED] micro-pitting (Kamminga 1979; Keeley 1980; Fullagar 1991).

Clearly visible residues associated with plant processing include cellulose fibres, resin, starch grains, phytoliths and oxalate crystals (raphides). Microscopically distinct structures on grinding stone surfaces were examined with brightfield/darkfield incident light at magnifications up to x1000. The presence of phytoliths and starch grains is indicative of siliceous and starchy plants respectively.

Ethnographic studies of plant-food processing indicate that several species of grass, acacia, fern and tubers were prepared with ground stone artefacts in Central Australia (Gregory 1887; Home & Aiston 1924; Gould 1969; Tindale 1977; O'Connell et al. 1983; Devitt 1988; Cane 1989). Siliceous and starchy plant products known to be processed by grinding are acacia, grass and fern (e.g. Nardoo sporocarps). These plants are distributed widely throughout the arid and semi-arid regions of Australia. Typha was processed with pounding stones.

Results

Twenty-five grinding-stones have definite microscopic traces of use including evidence of plant tissues and/or distinctive use-polish from processing siliceous plants (TABLE 2). Grinding-stone cross-sections suggest a range of upper and lower milling stones including mullers, pestles and mortars. By Smith's typology (1985; 1986), the range of grinding-stones includes pestles, mullers and millstones, as well as fragments with a possible seed-grinding function. The Cuddie Springs grinding-stone assemblage includes a range of morphological types consistent with plant processing as it has been witnessed ethnographically. Functional analysis demonstrates that the plants being processed were starchy and siliceous. Some of the grinding-stones have use-polish consistent with wet milling of seeds.

Two grinding-stones (one of which is shown in [ILLUSTRATION FOR FIGURE 3A OMITTED]) were found in unit 1 at the interface with unit 2. Four grinding-stones (two of which are shown in [ILLUSTRATION FOR FIGURES 3B, 3C OMITTED]) were recovered from unit 2. The specimen shown in FIGURE 3b is morphologically similar to complete and fragmented millstones found on the Cuddie Springs lake surface which also have similar use-wear and residues [ILLUSTRATION FOR FIGURE 4 OMITTED]. These stones are thin towards the centre of the implement, a characteristic more consistent with a grinding rather than a pounding action. The number of grinding-stone fragments from unit 2 represents less than 2% of the excavated stone assemblage from that level (Furby 1995). The Cuddie Springs grinding-stones [ILLUSTRATION FOR FIGURE 3 OMITTED] compare with Australian seed-grinding assemblages from late Holocene and ethnographic sites (Spurling & Hayden 1984; Cane 1984; Smith 1985; 1986; Nicholson & Cane 1991).

Preliminary phytolith studies (in collaboration with Dr Lisa Kealhofer, William & Mary College, Virginia) indicate the presence of grasses and a distinctive phytolith type of unknown origin; phytolith analysis is in the initial stages and Typha phytoliths have not been identified. Preliminary residue analysis indicates the extensive presence of starch with grain sizes generally small (with a range of 2-18 microns in diameter). Such starch grains and phytoliths in association with polished surfaces on formal artefact types strongly argue against processing of any of the roots or tubers known to current research.

Discussion

As with many issues in Pleistocene Australia, study of seed-grinding origins has been hindered by lack of evidence and small artefact samples from sites of Pleistocene age: grinding-stones comprise only 1 or 2% of the total assemblage - if they are present in excavated samples. Without representative grinding-stone assemblages from Pleistocene archaeological sites, there have been only individual specimens recovered from poor contextual situations. The presence of a range of morphological types known to be associated with seed processing at Cuddie Springs provides strong evidence for a seed-grinding economy around 30,000 b.p.

In Australia, seed-grinding is associated ethnographically with arid-zone sites. Similarly, these seed-grinding implements recognized archaeologically are generally not from the temperate margins of the country where resources are more abundant and predictable. The origin of seed-grinding in Australia has been argued to be in the semi-arid margins, before its move into the arid zone (Bowler 1976; Allen, 1972), with the appearance of grindstones in Last Glacial Maximum contexts related to broad economic changes towards the intensive use of seeds (White & O'Connell 1982). The energy expended in the gathering and processing of seeds is much higher, relative to the calorific returns, than for other resources (O'Connell & Hawkes 1981; O'Connell et al. 1983; Cane 1989), and optimal foraging theory predicts those other resources would be preferred (Winterhalder 1981). So the timing and origins of seed-processing have been linked to periods of stress in the arid and semi-arid zones. The palaeoecological record at Cuddie Springs (Dodson et al. 1993; Furby 1995) indicates a period of increasing aridity with expanding grasslands from about 30,000 years ago; a period also associated with the decline in range of megafaunal species.

The evidence from Cuddie Springs is significant in a world context as it relates to the settlement of desert/arid regions and the adoption of new plant resources, in particular seeds. In testing the alternative explanations for the 'broad spectrum' diet change in the Late Pleistocene (Edwards & O'Connell 1995), the Cuddie Springs evidence supports a combination of local and long-term climatic change where lower ranked resources, in this case seeds, were exploited as higher ranked resources (megafauna) declined or disappeared. Populations in the arid zone margins may have used seeds on an intermittent basis, when other higher-ranked resources were unavailable. Rather than 'gearing up' to move into the arid zone, people may have expanded their resource base to maintain a presence in regions incorporated into the expanding arid zone of the last Glacial period. The notion of climate-driven abundance of grasses is consistent with the environmental data, but it is insufficient as an explanation for the adoption of a broad spectrum diet at Cuddie Springs. Social processes, on the other hand, cannot be assessed on the available information, although Cuddie Springs is part of a dreaming track and a highly significant place for Aboriginal people to the present day.

Seed-grinding in arid eastern Australia as early as 30,000 b.p. suggests that the 'broad spectrum revolution' was not a synchronous event, either in Australia or globally. Rather, a combination of local and long-term climatic changes may have provided a context for the development of seed-grinding at radically different times in different parts of the world and in different social contexts.

Conclusion

The evidence from Cuddie Springs suggests a broad spectrum diet prior to the Last Glacial Maximum. Evidence from morphological typology, use-wear and residues each support the hypothesis that some form of plant-processing economy, not very dissimilar to that documented ethnographically, has been around for about 30,000 years. Taken together, these distinct lines of evidence support the hypothesis that these plant-processing practices, throughout this long time-period, included the culinary preparation of seeds and sporocarps. We suggest that these conclusions fit a theoretical model relating seed-processing to increasing climatic and dietary stress beginning locally at Cuddie Springs some 30,000 years ago.

Acknowledgements. For funding we thank the Australian Research Council, the Australian Institute of Aboriginal and Torres Strait Islander Studies, the Australian Museum, National Geographic and the Department of Employment, Education and Training. We are grateful for the contributions of the Johnstone family, Jan Currey, Doug Green, John Dodson, Leanne Brass, Georgia Britton and members of the Brewarrina Local Aboriginal Land Council, particularly Garry Lord. Jim Allen and Mike Smith provided cogent advice.

References

ALLEN, H. 1972. Where the crow flies backwards: man and land in the Darling Basin. Unpublished Ph.D thesis, Department of Prehistory, Research School of Pacific Studies, Australian National University.

BOWLER, J. 1976. Recent developments in reconstructing late Quaternary environments in Australia, in R.L. Kirk & A.G. Thorne (ed.), The origin of the Australians: 55-77. Canberra: The Australian Institute of Aboriginal Studies.

CANE, S. 1984. Desert camps: a case study of stone artefacts and human behaviour in the western desert of Australia. Ph.D thesis, Australian National University.

1989. Australian Aboriginal seed grinding and its archaeological record: a case study from the Western Desert, in Harris & Hillman (ed.): 99-119.

DEVITT, J. 1988. Contemporary Aboriginal women and subsistence in remote, arid Australia. Unpublished Ph.D thesis. Department of Anthropology and Sociology, University of Queensland.

DODSON, J., R. FULLAGAR, J. FURBY & I. PROSSER. 1993. Humans and megafauna in a Late Pleistocene environment from Cuddie Springs, northwestern New South Wales, Archaeology in Oceania 28: 93-9.

EDWARDS, D.A. & J.F. O'CONNELL. 1995. Broad spectrum diets in arid Australia, Antiquity 69: 769-83.

FULLAGAR, R. 1985. Use-polish on a muller from Ooraminna rock hole, central Australia. Unpublished report prepared for M. Smith, Research School of Pacific Studies, Australian National University.

1991. The role of silica in polish formation, Journal of Archaeological Science 18: 1-24.

FURBY, J. 1995. Megafauna under the microscope: archaeology and palaeoenvironment at Cuddie Springs. Unpublished Ph.D thesis, University of New South Wales.

GOULD, R.A. 1969. Subsistence behaviour among the Western Desert Aborigines of Australia. Oceania 39(4): 253-74.

GREGORY, A.C. 1887. Memoranda on the Aborigines of Australia. Journal of the Anthropological Institute 26: 1313.

HARRIS, D.R. & G.C. HILLMAN (ed.). 1989. Foraging and farming: the evolution of plant exploitation. London: Unwin Hyman.

HILLMAN, G.C. 1989. Late Palaeolithic plant foods from Wadi Kubbaniya in Upper Egypt: dietary diversity, infant weaning, and seasonality in a riverine environment, in Harris & Hillman (ed.): 207-39.

HORNE, G. & G. AiSTON. 1924. Savage life in Central Australia. London: Macmillan.

KAMMINGA, J. 1979. The nature of use-polish and abrasive smoothing on stone tools, in B. Hayden (ed.), Lithic use-wear analysis: 143-57. London: Academic Press.

KEELEY, L. 1980. Experimental determination of tool uses. Chicago (IL): Chicago University Press.

NICHOLSON, A. & S. CANE. 1991. Desert camps: analysis of Australian Aboriginal proto-historic campsites, in C.S. Gamble & W.A. Boismier (ed.), Ethnoarchaeological approaches to mobile campsites: hunter-gatherer and pastoralist case studies: 263-354. Ann Arbor (MI): International Mono-graphs in Prehistory. Ethnoarchaeological series.

O'CONNELL, J.F. & K. HAWKES. 1981. Alywarra plant use and optimal foraging theory, in Winterhalder & Smith (ed.): 99-125.

O'CONNELL, J.F., P.K. LATZ & P. BARNETT. 1983. Traditional and modern plant use among the Alyawara of central Australia, Economic Botany 37(1): 80-109.

SMITH, B.D. 1995. The emergence of agriculture. New York (NY): Freeman.

SMITH, M.A. 1985. A morphological comparison of central Australian seed-grinding implements and Australian Pleistocene-age grindstones, The Beagle 2: 23-38.

1986. The antiquity of seed grinding in Australia, Archaeology in Oceania 21: 29-39.

SPURLING, G.B. & HAYDEN, B. 1984. Ethnoarchaeology and intrasite spatial analysis: a case study from the Australian Western Desert, in H.J. Hietala (ed.), Intrasite spatial analysis in archaeology: 224-41. Cambridge: Cambridge University Press.

TINDALE, N.B. 1977. Adaptive significance of the Panara or grass seed culture of Australia, in R.V.S. Wright (ed.) Stone tools as cultural markers: change, evolution and complexity: 345-9. Canberra: Australian Institute of Aboriginal Studies. Prehistory and Material Culture series 12.

WHITE, J.P. & O'CONNELL, J.F. 1982. A prehistory of Australia, New Guinea and Sahul. Sydney: Academic Press.

WINTERHALDER, B. 1981. Optimal foraging strategies and hunter-gatherer research in anthropology: theory and models, in Winterhalder & Smith (ed.): 13-35.

WINTERHALDER, B. & E. SMITH (ed.). 1981. Hunter-gatherer foraging strategies: ethnographic and archaeological analysis. Chicago (IL): University of Chicago Press.

WRIGHT, K. 1994. Ground stone-tools and hunter-gatherer subsistence in southwest Asia: implications for the transition to farming, American Antiquity 59: 238-62.