Australia - Early Aboriginal Sites

One of the few things archaeologists seem to agree on is that the first colonists came from the north, with Arnhem Land, Cape York and the Kimberley region being the most likely entry points because of their relative accessibility from New guinea and Island southeast Asia. After thirty years of archaeological exploration, these three regions have all yielded Pleistocene human occupation sites as predicted, with some sites being in excess of 30,000 years old. A greater surprise were the Pleistocene sites found in extremely arid areas in the Pilbara, Central Australia and the Nullarbor Plain. A well-developed inland economy exploiting macropods and emu eggs apparently existed in the Pilbara by 25,000 BP, and in the Central Australian Ranges humans were present in the spinifex sand hills throughout the glacial maximum according to Michael Smith, who sees availability of drinking water as the governing factor.

At the base of Cape York Peninsula, about 100 kilometres west of Cairns, a spectacular tabletop mountain rises 400 metres above the surrounding savanna woodlands and plains. Its name is Nurrabullgin, also known as Mount Mulligan. This steep-sided mountain is 18 kilometres long and some 6 kilometres wide, with a volcanic base capped by sandstone. On its top lies Nurrabullgin Cave, a large sandstone rock-shelter with good headroom throughout and, significantly, close to deep, permanent waterholes. In this region, with little permanent ground water during the long winter dry season, permanent water sources were all-important. And in summer, the wide rock roof would have provided welcome shelter from monsoon rain and the heat of the tropical sun.

Excavations were carried out in the early 1990s by Bruno David, with the permission of the Kuku Djungan Aboriginal Corporation, who now owns the land. the occupation deposit is well stratified and contains stone artefacts, bone and large quantities of charcoal, representing a number of distinct phases of habitation separated by a series of hiatuses. Ochre fragments, including pieces showing striations from use, were present only in mid- and late Holocene levels. The cultural remains show considerable continuity, with the same raw materials - mainly basalt, chert and quartz - being used throughout. the remarkable thing about the occupation at Nurrabullgin Cafe is its age. an internally consistent sequence of radiocarbon dates on charcoal gave a date for the lowest occupation of greater than 37,170 BP. the oldest date came from a depth of only 30 centimetres, but David has shown that the deposition of sediments into the shelter would have been very slight, and the dates seem soundly based.

Nurrabullgin Cave presents important evidence of cultural continuity and great antiquity. It is the oldest human occupation site yet found in north Queensland, but some argue that hunters had penetrated the Atherton Tablelands region just south of Chillagoe about 38,000 years ago. this assertion is based on the long pollen sequence from Lynch's Crater, in which thee is a huge increase in the amount of charcoal at the same time as the vegetation changes from rainforest to fire-adapted Eucalyptus. this change can only readily be explained, according to pollen analyst Peter Kershaw, by the arrival of humans with their fire-sticks. In the limestone karst formations of Chillagoe, thee are many caves and rock-shelters, and Pleistocene occupation has been found in two of these. Fern Cave is a large cave with two high-domed chambers, where Bruno David found occupation at least 26,000 years old. A few heavily patinated pckings on the wall adjacent to the excavation consist of a series of loosely clustered pits, a star shape and three four-pronged motifs, resembling 'tridents' or bird tracks. These have been demonstrated to be similar to other patinated peckings from the Chillagoe, Mitchell-Palmer, Laura and Koolburra regions, which are thought to have considerable antiquity on the basis of degree of patination and nature of superimpositions.

Another long cultural sequence at Chillagoe has been uncovered by a major, ongoing excavation by John and Mireille Campbell, and their students from James Cook University, in Walkunder Arch Cave. Occupation goes back more than 18,000 years - there are two heavily patinated geometric engravings, and the lowest level contained a horsehoof core, a waisted tool, shells and wallaby bones. Interestingly, burnt antbed or termite mound was found throughout the deposit, indicating that this was used as fuel in both Pleistocene and more recent times. Further north in the Laura region, Michael Morwood of the University of New England has recently uncovered a 32,000-year-old occupation at Sandy Creek Shelter l. Earlier, terminal Pleistocene occupation had been found by Richard Wright of Sydney University at Mushroom Rock, and Andree Rosenfeld of the Australian National University had uncovered both occupation and patinated geometric rock engravings in excess of 13,000 years old at Early Man shelter.

Sandy Creek shelter l had previously been partly excavated by Percy Trezise in 1969, yielding a deep cultural deposit, 'buried' rock engravings and many stone tools, including a ground-edge axe lying on bedrock at a depth of 3 metres. The earliest evidence of occupation was a stone-knapping floor of twenty-six small artefacts of crystalline quartz, near the base of the rubble and associated with charcoal dating to 32,000 BP. these were very few artefacts above this until the first systematic use of the shelter around 18,000 years ago.

The most important artefact found at Sandy Creek was the ground-edge axe which has been seen in 1981 at Cape York. Morwood has convincingly established that the minimum age of the base of the rubble and the axe is in the order of 32,000 years. The drawing by architect Eddie Oribin and old photographs give a good idea of its size and form. It was made from pink quartzite, 8.7 centimetres long, with a ground working edge, a slight 'waist' and a groove to aid hafting. It compares closely with the ground-edge axes from Arnhem land. Small rock fragments with grinding marks in both the Mushroom Rock deposit and a 10,000-year-old layer at Early Man Shelter had hinted at the presence of edge-grinding in the late Pleistocene in Cape York Peninsula, but this find considerably extends the time depth of ground-edge axes both in the region and continent-wide. Ground-edge artefacts have now been found in Pleistocene layers in a number of sites in north Queensland, the Top End of the Northern Territory and in highland New guinea (at the sites of Kafiavana, Kiowa, Yuku and Nombe, where a complete axe was recovered from a 14,500- to 26,000-year-old layer). In the Kimberley in Western Australia, as described below, flakes showing signs of grinding were excavated from a 27,000 BP layer in Widgingarri l and the 18,000-year-old basal deposit of Miriwun shelter.

Pleistocene ground-edge artefacts appear to be restricted to north of the Tropic of Cancer, and to the extreme north of the continent if we can judge by the present small sample. Later in the Holocene, ground-edge axes were the regular chopping tool over most of the mainland, but not in Tasmania. Morwood has carried out a valuable analysis of the distribution of various forms of axes in time and space. It is difficult to account for the limited distribution of ground-edge axes in the Pleistocene. A case has been made that the large waisted axes in New guinea wee used for ringbarking rainforest trees, but on the mainland, the ground-edge Pleistocene axes are relatively small and light, and come from sites in Eucalyptus woodlands rather than rainforest. Only further research, hopefully including analysis of use-wear and any residues remaining on the working edges of these artefacts, will help to solve this puzzle of their function and restricted distribution to the north of Carpentaria.

In northwestern Queensland on the Barkly Tableland, traces of occupation more than 17,000 years old have been found. This remote area northwest of Mount Isa contains spectacular gorges, permanent rivers and waterholes with abundant fish and shellfish, and plant food such as the nuts of pandanus and cycad palms. Along 40 kilometres of river, there is only one good rock-shelter - the deep, well-protected shelter on Colless Creek. A small excavation was carried out there by Philip Hughes and Peter Hiscock, revealing an extraordinarily rich site with an average density of 50,000 artefacts per cubic metre of deposit. The uppermost cubic metre held half a million pieces of oldest in a series of dates obtained on shell from the site.

The ancient environment was reconstructed through an analysis of sediments. conditions in the vicinity of the shelter over the last 18,000 years were considerably drier than during the preceding phase of human occupation, which probably extended back beyond 30,000 years. the high degree of weathering of the deposit, patination on the artefacts and heavy staining on bones suggest that basal occupation is at least 30,000 BP, and perhaps much older. During the last 18,000 years, Colless Creek Shelter and the surrounding well-watered awn Hills Gorge would have acted as an oasis in this arid region. the period around 17,000 years ago was a time of great aridity in iother inland regions, such as the Willandra Lakes, and it may have been this climatic stress that drove people to such an 'oasis' at that time.

The Kimberley region in the extreme north of Western Australia has long been thought to be one of the possible landfalls for early migrants or castaways swept southwards from Timor or the Indonesian archipelago. such migrants would have arrived on a broad plain, but it is uncertain whether it would have been grassland, savanna woodland or mud and mangroves, with possibly an accompanying dearth of encountered the high cliffs of the edge of the Kimberley escarpment, which forms the present rugged coastline. The continental shelf is quite wide and the offshore waters shallow in the Kimberley region, so early sites may now lie beneath the sea.

The Kimberley escarpment is broken in places by plains and rivers flowing out to sea through narrow gorges or broad river valleys like the Ord. It is in the Ord Valley that two early occupation sites have been found, one of which certainly dates to the Pleistocene period. In the Miriwun rock-shelter on the Ord River, Charles Dortch of the Western Australian Museum excavated an occupation deposit in 1971 as part of a salvage program before the area was flooded by the Ord River irrigation scheme. The upper levels of the site contained small tools, but in the dark brown, silty earth early assemblage. Tool types included relatively thick, denticulated or notched flakes, adze flakes, a few core-scrapers and small blades, and some pebble tools. Also found were some quartzite fragments, which may be parts of grindstones or anvils.

One of the most remarkable discoveries was that two flakes recovered from below the horizon dated to 18,000 years were pieces struck from tektites. Tektites, or australites as they are called in Australia, are small glassy pebbles, up to about 2.5 centimetres in diameter, black or dark green in colour, and shaped like buttons, discs, teardrops, balls or dumb-bells. Their chemical composition is different from that of the rocks where they are found or from that of any terrestrial lava.

The origin of tektites is a puzzle. some scientists believe that they are bits of terrestrial sedimentary rock excavated by the impact of meteorites crashing into the earth's surface, melted by the heat of impact, and congealed into glass as they are flung into the atmosphere to tall as a widely scattered shower. A more likely possibility is that they are the remains of gobs of lava fired at the earth by volcanic activity on the moon. A huge shower of tektites fell in the Australasian region 750,000 years ago. These australites are concentrated in a swathe across the southern half of Australia, particularly in Central Australia and in inland southern districts of Western Australia.

Analysis of one of the Miriwun tektites places it within the Indochinite group of tektites, the first tektite of this kind known from Australia. The seemingly remote possibility that this 18,000-year-old artefact (a very small flake) was brought from Southeast Asia cannot be entirely dismissed until finds are made in Australia of whole Indochinites (i.e. pieces showing no artificial modification) in places where there is no association with human occupation. This Miriwun tektite may be the first ice age Asian artefact found in Australia, if Indochnites are shown never to occur outside Indochina.

In the Kimberley region, there may be a long continuity of technological tradition, both in grooved ground-edge axes and in serrated flakes. Kimberley serrated spear points are renowned for their fine crafting and pleasing symmetry. They were made by pressure-flaking, a technique in which tiny flakes were pressed off by use of a bone, piece of wood recently bottle glass or telephone insulators have been used. The use of these bifacially trimmed leaf-shaped points goes back at least 3000 years. One particularly important feature of the Ord River sits is that organic material was well preserved in most of them. It shows that, throughout the 18,000 years of Miriwun's habitation, the human occupants exploited a wide range of aquatic and terrestrial fauna. Food from the surrounding land included wallabies, possums, bandicoots, lizards and rodents, and from the river and lagoons came shellfish, reptiles, catfish and goose eggs. the numerous eggshell fragments of the pied or semi-palmated goose (Anseranas semipalmata), which breeds only during the wet season, indicates that Miriwun was used as a wt-season camping place from late Pleistocene times until the European era.

Recent research on the west Kimberley coast and the islands of the buccaneer Archipelago has revealed occupation about 28,000 years ago. The rock-shelters of Widgingarri 1 and 2 northeast of Derby on the Kimberley coast seem to have been inhabited from 28,000 BP, when they would have been more than 100 kilometres from the sea, until about 18,500 BP, the height of aridity at the last glacial maximum, when the sea had retreated to its maximum extent, and they were abandoned until about 7500 BP. O'Connor considers that increasing aridity rather than the retreating coastline was the prime cause of the sites' abandonment. The fragments of baler shell (a large gastropod, Melo sp.) and pearl shell found in the Pleistocene layer are now interpreted by O'Connor not as reflecting a coastal economy, but as prized items traded from the coast. In the ethnographic present, large baler shells are valued for their usefulness as containers for water and so on, and the pearl shell is valued and traded for its aesthetic qualities. O'Connor infers that this was also the case between 30,000 and 18,000 years ago. If she is correct, this is another example of the great continuity of Australian material culture and practices.

Likewise, on a small offshore island, earliest occupation in Koolan Shelter 2 dates to a minimum age of 27,300 BP. By extrapolation, O'Connor estimated the age of first occupation as about 30,000 BP, a time of relatively higher sea level, when the sea would have been close to the shelter. The date of 27,300 BP came from the mangrove shellfish Geloina coaxans, which is plentiful in the site, indicating a heavy reliance on marine resources by the site's earliest as well as its most recent occupants. Koolan Shelter 2 was vacated about 24,000 BP, when conditions became more arid and the drop in sea level caused the shore to retreat some 220 kilometres, leaving the island as a peak of an inland range in the arid west Kimberley. The inhabitants seem to have followe3d the rising and retreating seas, reoccupying the shelter about 10,400 BP, when the rising sea had again separated Koolan Island from the mainland.

Of the western extremity of the Australian arid zone lies Northwest Cape, where a small excavation in Mandu Mandu Creek rock-shelter has uncovered human occupation going back 34,000 years. This sizeable limestone rock-shelter in Cape Range national Park faces west over a 1-kilometre wide coastal plain to Ningaloo Reef. The initial 1-metre square test pit yielded over 500 stone artefacts, marine lollusc shells and marine and terrestrial bone fragments. While the only faunal remains preserved in the lower, Pleistocene layer (below a date of 19,590 BP) were fish teeth and the thickest, most durable shell fragments (such as chiton valves and robust fragments of baler shell), it is clear 'that during this early phase of occupation Aboriginal people had the knowledge and skills to exploit a variety of marine foods'. The continental shelf is narrower here than anywhere else around Australia, and it seems that the sea was about 6 kilometres from the site when it was first occupied. Morse suggests that 'there may be little real difference, in the range of food types exploited, between the marine economy of the Pleistocene occupants and that practised during Holocene times'.

Further work by Morse since 1988 has produced the earlier radiocarbon dates, and uncovered a unique find in Pleistocene Australia, twenty-two shell beads (colour plate 7). they are made from small marine cone shells, and are firmly associated with the baler shell which gave the date of 34,200 years. comparison of notches on these cone shells with similarly threaded recent shell artefacts from northwestern Australia shows analogous use-wear patterns. This is an exciting addition to the 12,000- to 19,000-year-old bone beads found at Devil's Lair, the only other Pleistocene ornaments yet recovered from an archaeological site. Necklaces of shell beads were common in more recent Aboriginal Australia, and are known particularly from Tasmania, but this is another example of the incredibly long continuities discernible in Aboriginal decorative traditions.

Pleistocene occupation from a time of low sea level has rec4ntly been found by Veth on the Monte Bello Islands, now located 120 kilometres off the present Pilbara coastline. Three limestone caves with cultural material have been excavated on Campbell Island. the age of 27,220 BP was given by marine shell at the base of the deposit in Noala Cave, reflecting a time when it was adjacent to the Plaeistocene coast. The deposit shows full use of marine resources, kangaroos and other mammals from the now submerged plains that then joined the Monte Bello and Barron islands to the mainland.

Noala and two other adjacent caves, Morgan's and Haynes caves, all show intensive occupation around 8000 BP, with shellfish (predominantly mangrove species), fish and land mammals being exploited. Retouched, utilised stone artefacts were made from exotic materials, such as metamorphic rock, not visible on the present islands. Between about 8000 and 7500 BP the islands became part of the mainland. soon after 7500 BP they were apparently abandoned; by 6500 BP they were 50 kilometres offshore and uninhabited.

Surprisingly, a number of Pleistocene sites have been discovered on the Hamersley Plateau in the Pilbara. This area is part of Australia's arid zone and would have lain 500 kilometres inland and been even drier at the height of the last glacial period. the first was the Mount New man rock-shelter Orebody XXIX (PO187), which overlooks the headwaters of the Fortescue River. Ash, charcoal and ochre were found throughout he 1-metre deposit excavated, but no bone. Eleven hearths were found, one of which was typical of fire-pits used by modern Aborigines for baking animals. Most of the 400 artefacts found were simple flaked or retouched pieces, but two diagnostic implement types were found: steep-edged scrapers and notched scrapers. Radiocarbon dates revealed that the 1-metre deep deposit is more than 20,750 years old. Eighteen kilometres to the northeast, and close to the east side of the Fortescue River, another site at Ethel Gorge (PO255.2) gave a date on near-basal occupation of 26,300 BP. these are conservative dates for initial occupation of the sites, as in neither case were the excavations taken to bedrock or culturally sterile units.

Both rock-shelters were occupied occasionally before some 20,000 years ago. Nowadays, the bed of the Fortescue River only flows after heavy rains, and normally only a few [pools of water are to be found on its upper section and in the gorges of what is now the Hamersley National Park. Veth interprets the Hamersley Ranges as a 'refuge' area, but also points out that 'there is currently no unequivocal evidence for continuity of occupation during the height of the last glacial maximum from approximately 18,000 and 15,000 BP. Another Pleistocene site has recently been identified in a coastal but arid area at shark Bay, on Peron Peninsula - the most westerly part of the Australian continent (450 kilometres south of Northwest Cape). this is an open site called Silver dollar, excavated by Bowdler. the lower occupational layer contained stone artefacts associated with large amounts of emu eggshell and teeth of kangaroos and wallabies, and some fragments of baler shell. Radiocarbon dating of this baler shell and emu eggshell gave an age range between 18,000 and and 6000 years ago, after which marine remains became abundant.

I am forced by my own data to concede that well-developed inland economies might well have been in place in northwest Australia by 25,000 BP or earlier. In this case, we see exploitation of emu eggs and macropods some 100 km from the coast over a period of some 7,000 years or more. This does not disagree with my original and fundamental premise that Australia was colonised by coastally adapted people whose colonisng routes were around the coasts and up the major river systems, but it certainly suggests a much earlier adaptation to peculiarly Australian interior environments than I have previously been prepared to concede.

Not until 1987 did the first proof come that the arid heart was inhabited in the Pleistocene, with the discovery by Michael Smith of 22,000-yqr-old occupation in Puritjarra rock-shelter, almost in the dead centre of Australia. Puritjarra lies close to the only permanent water in the Cleland Hills, near the eastern boundary of the Western Desert, some 320 kilometres west of Alice Springs in the Northern Territory. It is a region of spinifex grasslands and mulga woodland broken up by the complex topography of the central ranges. Rainfall averages less than 350 millimetres a year, but the ranges have some permanent springs, waterholes, deep rock 'reservoirs' and soakages in creek beds. All the rivers in the region, such as the Finke, are intermittent, but their beds usually contain some waterholes and soakages.

Puritjarra is a huge rock-shelter, 45 metres long and about 20 metres high, in a cliff of hard red sandstone. The extensive array of rock art includes stencils, paintings and some Panaramitee-style engravings, well known at the nearby Thomas Reservoir site. Eleven square metres of the 400 square metres of level, shaded earth floor were excavated in 1986 and 1988. Twelve radiocarbon dates were obtained on charcoal, and six thermoluminescence (TL) dates on the sediments as an independent check. The results have been startling. The very base of the lower layer has a preliminary TL date of about 30,000 years old. No details are yet available, but the internal consistency of both series of dates indicates that the site has 'good stratigraphic integrity despite the low rate of sediment accumulation', and Smith believes the earliest occupation is 'in the order of 30,000 years'. The shelter was first occupied, albeit fleetingly, well before 22,000 years ago. The first substantial use began at about 22,000 BP - marked by charcoal, some ten pieces of red and purple high-grade ochre, sixty stone flake artefacts including 'a single large steep-edged implement', and some 200 small pieces of flaking debris. Between 22,000 and 13,000 years ago, the shelter was used only occasionally, and no more than a few artefacts were deposited each millennium. the uppermost stratigraphic layer is a loose, gritty sand containing intact cooking hearths, charcoal, flaked stone tools, many grindstones (absent from the Pleistocene layer), ochere and emu eggshell. This layer spans the last 6000 years, but also attests to a major increase in occupation of the region during the last 1000 years.

The 22,000 -year-old occupation coincides with the onset of major aridity. This 'presumably reflects the beginning of the pattern of land use tethered to reliable water resources'. Between 22,000 and 13,000 BP, the period of full glacial aridity, thee is evidence for repeated, if slight, usage of the shelter. The repeated used of Puritjarra, together with its location away from any natural corridor, indicates the presence of a resident local population in this 'refuge'. Smith suggests that visits may have been short affairs, without much need for the replacement or maintenance of implements (hence the low numbers of artefacts), by small, highly mobile groups resident in the main ranges to the east.

Very early occupation has now been found in underground limestone caves on the Nullarbor Plain. At least two caves in the far southwest of south Australia were in use before 30,000 years ago: Koonalda Cave and Allen's Cave. radiocarbon dates on charcoal have shown occupation at Koonalda by about 24,000 BP and at Allen's Cave (N145) by 25,000 BP. Now thermoluminescence (TL) dates on lower occupation levels where charcoal did not survive have given dates in the order of 34,000 years. At Allen's Cave, an artefact lies 1 metre below a preliminary optically stimulated luminescence (OSL) date of 34,000 +_ 7000 BP, and OSL dating attributes a similar antiquity to the earliest occupation at Koonalda. Koonalda Cave is a crater-like doline (limestone sinkhole) in the karst scenery of the extensive, flat, and Nullarbor Plain (below). The cave was a flint mine. Quarrying was done underground, at times with no natural light, and the quarried nodules were taken elsewhere to be made into tools. Hearths, charcoal and the residue of the quarrying process were found inside the cave, mainly in the first, dimly lit chamber some 100 metres down from the entrance and 76 metres below the surface of the plain. Initial pioneering exploration by Gallus was followed by major excavation by Wright. His 60metre-deep pit produced evidence that flint miners had been visiting the cave between about 24,000 and 14,000 years ago.

The cave was thoroughly explored and found to contain a series of lakes, an invaluable source of drinking water, and, most remarkably, Pleistocene 'rock art', comprising finger markings on the walls in total darkness, some 300 metres inside the cave entrance. The sinkhole had two major attractions for early humans. It was a reliable source of water on an arid plain and its walls held nodules of flint, the best raw material for stone tool manufacture available in the continent. At Allen's Cave, some 80 kilometres west of Koonalda near eucla, there is a culturally sterile deposit bracketed between dates of 20,200 +- 1000 BP and 11,950 +- 250 BP after the early occupation. this has been taken to indicate abandonment of the site, at least between 17,500 and 15,000 years ago, a period of intense aridity. About 15,000 to 18,000 years ago, when the sea reached its lowest level of the last glaciation, the coast was some 160 kilometres further south and the eucla-Koonalda region became treeless plains, with an estimated average annual rainfall of only 160 to 180 millimetres. Allen's Cave was virtually deserted at this time and Koonalda Cave experienced only intermittent occupation between 22,000 and 15,000 BP; no doubt the Aboriginal groups had moved south onto the wide coastal plain that emerged as a result of the fall in sea level. The sea level began to rise again about 12,000 years ago, bringing the coastline closer and thus again increasing the rainfall, water resources and mallee scrub.

Archaeological evidence has now established that all major geographical regions, coastal and inland, were occupied relatively early in the colonisation of Australia (that is, before 30,000 years ago). During this 'lacustral phase', when Lake Mungo and other lakes of the interior were full of fresh water, there was occupation in some extremely arid regions. Peter Veth has explained this in biogeographical terms, suggesting that there was early occupation of less marginal habitats such as 'refuges' (with reliable water), plus intermittent occupation during more climatically favourable periods of 'corridots between 'refuges' and the extensive 'barriers' of the sandridge deserts (below).

The onset of the full glacial climate, which was at its peak between about 18,000 and 15,000 BP, brought these hunters and gatherers the problems of intense aridity and extremely cold, frosty winters. From around 25,000 BP, it was increasingly dry and windy, with widespread drying of lakes, extensive dune-building, expansion of the arid zone and a major drop in sea level. At this stressful time, some siters in 'corridors' were vacated, according to Verth, but other sites such as Puritjarra lay in 'refuge' areas by permanent water and continued to be in use, at least occasionally. Smith has recently seriously questioned Verth's 'barrier desert' theory, raising doubts whether the major sandridge deserts do form a biogeographic unit, whether the division into 'corridors', 'refuges' and 'barriers' can be sustained, and whether the 'barrier deserts' would indeed have represented a new challenge to humans. He emphasises the strong links between Veth's 'barrier deserts' and adjacent 'refugia' and 'corridors', some of which combine dunefields with hummock grassland and unco-ordinated drainage in the same way as do sandridge deserts. smith also points out that the evidence of Bowdler's silver dollar site suggests that well-developed inland economies based on macropods and emu eggs may have been in place on the arid coast of Western Australia by 25,000 BP or earlier. Throughout the late Pleistocene and early Holocene, the occupants of Puritjarra Shelter would have been dependent on the food resources of the surrounding spinifex sandhill habitat.

At the climate improved, new sites were occupied within the 'corridors', such as Cuckadoo 1 Shelter near Cloncurry in semiarid Queensland. A hearth with fragments of mussel shells and charcoal, the JSN site, was dated to 13,850 + 190 BP, showing the penetration of the core of the Strzelecki dunefields by that time. there was penetration of the relatively well watered Flinders Ranges at Hawker Lagoon by 15,000 BP, and two hearths associated with stone artefacts in dune cores on the Lower cooper Creek in the central Lake Eyre Basin date to some 11,500 years. The shores of Lake Frome in the arid belt became popular in the generally moister conditions between about 95500 and 4000 BP, and the widespread occupation of the Strzelecki and other dunefields came later, within the last 5000 years. Much archaeological exploitation of 'barrier deserts' and adjacent dunefields (such as the Rudall River and Balgo region, Simpson Desert, Lake Eyre basin, Coongie Lakes and Cooper Basin by outstanding field workers such as Scott Cane, Michael Smith, Ron Lampert, Phillip Hughes, Peter Veth and Elizabeth Williams) has revealed many hundreds of sites belonging to the last 5000 years, but a general absence of Pleistocene cultural evidence, even in the most favourable habitats.

Smith and Veth disagree on details of the timing, nature and explanation of this Holocene move into the dunefields, with Smith seeing access to potable water as the critical key, whereas Veth emphasises other factors. Nevertheless, the widespread exploitation of grass seed and the other resources of the dunefields in the mid-Holocene appears the last step in the colonisers' successful adaptation to the world's driest continent.

Pacific Islands Radio

Subscribe!

Please enter your email address for your free Pacific Islands Radio Newsletter!

COUNTRIES	LINKS	MYTHOLOGY	RESOURCES	POSTCARDS	RADIO	HOME
MELANESIA	MICRONESIA	POLYNESIA	JANE	TRIBAL ART	TATTOOS	CHATROOM