Australia: The Land Where Time Began

A biography of the Australian continent 

Aboriginal Occupation of Greater Australia - the pattern of colonisation beginning in the Late Pleistocene

The dynamics of the settlement of Australia by Aboriginal people, that took place prior to 30,000 BP, is not certain. Smith (Source 1) gives several factors by which the direct reconstruction of the way in which colonisation took place has been constrained. The range and quantity of data derived from archaeology are insufficient to determine distribution prior to about 30,000 BP, about half way to the first arrival at about 60,000 BP. His second factor is lack of precision of dating methods in this time period (Allen, 1994; Allen & Holdaway, 1995; Roberts, Jones & Smith, 1994). Other sources of data are often used to interpret archaeological evidence regarding the colonisation of the continent.

Smith suggests the first arrival of people in Australia and New Guinea occurred at some time before 35,000 BP (Allen, 1989; Groube, et al., 1986; Pearce & Barbetti, 1981), and probably between 50,000 and 60,000 BP (Roberts, Jones & Smith, 1990; Roberts et al., 1994). He believes to reach Australia and New Guinea it would have required people to make a crossing of the ocean of a few days duration from the nearest islands of the Indo-Malaysian Archipelago, also suggesting that the migration was equally likely to have occurred as a steady trickle as a single discrete migration. The first arrivals would have presumably been familiar with island and tropical environments, and would have been using watercraft for some time before making the crossing (Irwin, 1992).

The first settlement of the coast, as well as estuarine and riverine environment, would have been assisted by their familiarity with the continuity of useful species of plant and marine environments across Wallacia in northern Australia and New Guinea, though the terrestrial fauna would have been different and unfamiliar. People would have migrated to the east, remaining in the tropics as they colonised the Bismarck Archipelago and the Solomon Islands, both rich in marine resources, though their terrestrial environments were depauperate in both terrestrial plants and animals. When they moved to the east or the south they would have been confronted by an environment like nothing they had ever encountered before, such as the montane regions of the cordillera in New Guinea, arid central Australia and in Tasmania, the high latitude temperate environments. By 30,000 BP they were thriving in all these disparate environments (Smith & Sharp, 1993).

For much of the time earlier than 30,000 BP, when colonisation was taking place, the sea level was between 60 and 80 m lower than at present (Chappell, 1994), and as the present islands of New Guinea and Tasmania were both part of the single landmass of greater Australia, the area of the continent was substantially larger than at present, especially as there were also large areas of the continental shelves around the margins, especially in the north-western (Sahul) part of the continent, exposed by the lower sea level. Whenever the sea-level dropped by 12 m or more dry land, or at least a shallow strait, connected new Guinea to the Australian continent. When the sea level was 55 m or more below the present level the island of Tasmania was connected to the southern margin of the main landmass of Australia by the Bassian Rise. Following the revision of estimates of sea levels (Chappell, 1994), it would have been possible for people to walk from southern Australia to Tasmania any time between 60,000 BP and 10,000 BP.

Dispersal patterns and processes

According to Smith, there are 2 broad categories of proposed scenarios, those that favour early dispersal across the continent that took place rapidly, and those favouring slower, patchier dispersal, with parts of the continent being occupied up to 10s of thousands of years after the first arrivals on the shores of Greater Australia.

Several assumptions are made as the basis for scenarios of rapid dispersal and occupation that include a very flexible response to environmental conditions encountered, and a high intrinsic population growth rate, that resulted in occupation of the continent that was rapid. Some propose that the dispersal across large areas of the interior followed the drainage systems (Birdsell, 1957; Mulvaney, 1961), such as those of the Murray-Darling Basin and the Lake Eyre Basin, that between them cover a vast area of central Australia. As they moved from the early arrival points in the north of the continent there would have been a gradient in the vegetation cover of the land towards the south, allowing the people to adjust gradually to changing plant foods, as they gained ecological knowledge that eventually allowed them to settle the arid interior (Golson, 1971). Quantitative models have been constructed that show that it would be possible for the continent to have filled to saturation, with a population of about 300,000 people, (the estimated population current in 1957) in between 845 and 4134 years after initial arrival (Birdsell, 1957). The estimated pre-contact population has since been increased to about 750,000 (White and Mulvaney, 1987). Key components in these proposed dispersals include rapid demographic growth that drove dispersal, groups from the main population centres moving to new areas when carrying capacity is approached, a process in which social groups are replicated.

Others have questioned these assumptions. It has been suggested that the Aboriginal population may have grown slowly for the first few millennia after first arrival (White & O'Connell, 1982: 46-54). Suggestions have been made that stochastic fluctuations affect small founder populations (McArthur, Saunders & Tweedie, 1976). It has also been suggested that the population growth rate could have been reduced by the presence of endemic malaria in coastal lowlands in the northernmost third of the continent (Groube, 1993). It has also been suggested that if small groups continued arriving from Asia for several millennia after the first landing it may have offset these factors. The same results could have been achieved by the presence of several founder populations along the north coast and western New Guinea. Even when populations have growth rates that are almost stationary the population grows substantially over 5,000-10,000 years. Selective parameters appear to change when empty territory is available, rapid dispersal taking place (Kitching, 1986; Stodart & Parer, 1988; Rindos & Webb, 1992). Smith suggests small groups may have moved into new territory that had a rich terrestrial fauna before the carrying capacity of their original group was reached, possibly without replicating social formations. In the desert areas some desert groups have historically managed to maintain social and demographic units, even when the population density is as low as 1/200 km2 (e.g. Long, 1971).

An alternative view was proposed according to which some habitats, coastal or riverine zones, were preferentially occupied, because of the lack of anything more than the barest capacity to adapt to ecological conditions they had not previously encountered (Bowdler, 1977) or (Hallam, 1987; Horton, 1981). According to Smith, Bowdler proposed the most influential of these models, suggesting the focus would have been strongly on aquatic resources at the beginning of colonisation, resulting in the groups being tied to littoral, lacustral and riverine habitats, suggesting that it was not until 12,000-10,000 BP that montane and desert regions were occupied. One of the key parts of her suggestion is that after 15,000 BP lakes and rivers in semi-arid areas of south-eastern Australia failed, such as occurred in the Willandra region. According to this scenario (See Millet Harvesters), the drying up of lakes, such as the Willandra lakes, forced the people to rely on terrestrial food sources such as grass seeds and acacia seeds, both of which are present in the ethnography of central Australia, widespread settlement of inland arid areas becoming possible following the glacial period as a result of the shift to terrestrial food sources.

According to Smith the arid zone is singled out in a number of other models as a particularly problematic environment for occupation (Horton, 1981, Veth, 1989), and it has been suggested that woodland that was well watered would have been the preferred environment for settlement, over both aquatic and desert environments, as a result of a more flexible response towards the terrestrial environment than was allowed in the model of Bowdler (Horton, 1981; Hallam, 1987). Potable water availability was seen as a key factor in early occupation, especially by Horton, who also suggests that the distribution of the large megafauna were a guide to woodland that was well-watered in the late Pleistocene, assuming at least some of the large herbivorous megafauna were extant at the time of the first arrivals in Australia. Rock art in the northwest of the continent and elsewhere have been found that are believed to be of species of the megafauna. See Megafauna and the Dreamtime, Diprotodon optatum, Giant Kangaroos of the Dreamtime (Sthenurus), Palorchestes azeal, Quinkana, Marsupial Lions. Smith says that the context in which all  these models were proposed has been changed radically by archaeological work that has been carried out since 1977 (Allen et al., 1988; Allen, 1989; Bowdler, 1990a; Kiernan et al., 1983; Smith, 1987; Cosgrove 1989). Models for the occupation of the major continental dunefields that occurred late, about 5,000 BP, have been proposed that are somewhat similar (Veth, 1989, 1995). A common factor in these models is the special difficulties associated with these regions that have drainage that is uncoordinated, hummock grasslands and dunefields, the suggestion being made that before these dunefields could be occupied it was necessary to make adaptations, both technological and social, in adjacent regions.

Smith suggests the timeframe is a major problem for all these proposed models, as he believed it was unlikely the people took 10s of thousands of years to adapt to ecological conditions that were either difficult of unfamiliar, especially in the light of archaeological evidence indicating that habitats such as montane (Gillieson & Mountain, 1983; Mountain, 1993; White, Crook & Ruxton, 1970) and arid (Maynard, 1980; Smith, 1987, 1989) where exploited long before 14,000 BP, and it is now apparent that terrestrial-based economies were operating prior to this time in a number of places across Australia (e.g. Bowdler, 1990b; Cosgrove, 1989; Kiernan, Jones & Ranson, 1983).

Smith says these arguments risk confusing dispersal with optimisation strategies (see Rindos & Webb, 1992), suggesting that optimal adaptation to the local environment, and population densities that are comparable to ethnohistoric levels, need not occur at the time of the initial settlement. Smith believes the case for the initial occupation of these regions taking place in the Holocene because of the requirement for adaptations, economic and social,  before they could be occupied, is overstated by Veth, (1989) (Smith, 1993). This model was later recast (Veth, 1995), shifting emphasis more on to areas where the drainage systems were uncoordinated, though according to Smith it is 'clearly at odds with archaeological evidence' here (Cane, 1995; Gould, 1977; Martin, 1973; Wright, 1971; Smith, 1987, 1989). According to Smith, the picture of the settlement of deserts has changed rapidly following the finding of clear evidence of the use of sandy deserts during the Pleistocene, indicating that the chronological framework at least needs to be recalibrated (Smith et al., 1991; Veth et al., 1990; Veth, 1995; Veth & O'Connor, 1996).

Major time lags between the coastal region occupation and that of the continental interior are believed by Smith to be unlikely (Source 1). It is difficult to model the growth of population because of the linear configuration of the coastal zone without concluding that some movement into the interior took place long before the populating of the coastal zone of the continent was complete (White & O'Connell, 1982, Fig. 3.7). Westward movement around the margin of the continent would be expected to have required adaptation at an early stage to arid conditions, as the continental western coast is arid, as well as less emphasis on littoral resources (Nicholson & Cane, 1994; White & O’Connell, 1982: 52). If a suggestion about the presence of malaria in northern and swamp regions is correct (Groube, 1993), groups moving inland would have been free to grow and disperse while those in the malaria infested areas would have been slowed down. It has also been suggested that larger territories would be needed, as a result of reduced carrying capacity in arid areas, higher dispersal rates also resulting in these areas (Birdsell, 1957). This has been confirmed by studies of rabbits, which have spread throughout arid and semi-arid regions at 100 km/y, those in coastal and forest habitats spreading at 10-15 km/y (Stodart & Pared, 1988).

Continental colonisation - archaeological evidence

Smith suggests that because of the remoteness of the time of first arrival there are problems with the archaeology and the chronological resolution of this event. These problems make it difficult to investigate the pattern and rate of settlement, only in the very broadest terms being possible. According to Smith, current data [1998] suggest that all parts of the continent had been occupied by 15,000-10,000 BP, though in places, such as south-western Tasmania, the New Guinea Highlands and parts of the arid interior, did not have a record of continuous occupation at the time of writing, 1998.

High latitude regions

In Tasmania, the southwest region that is densely forested, has produced a number of finds such as Beginners Luck Cave (Murray et al., 1980), Kutikina Cave (Fraser Cave) (Kiernan et al., 1983) and Nunamira Cave (Bluff Cave) (Cosgrove, 1989) that demonstrate that the region has one of the richest records in Australia of occupation during the Pleistocene that begins about 35,000 BP. The occupation sites were being used at a time when the region, on the extreme southern margin of the continent, was vegetated by exposed alpine grasslands.  The oldest date of 35,000 BP was obtained from Warreen Cave in south-western Tasmania. These early dates for occupation in the Tasmanian southwest indicated that humans were exploiting highland and extremely high-latitude regions much earlier than expected, causing problems for the previous models of the occupation of Australia. The archaeological evidence apparently shows that the region was occupied continuously throughout the last glacial period, when the largest Tasmanian ice sheet existed. These sites also provided evidence for a much richer fauna, dominated by small macropods, than had been expected, prompting Bowdler (1990a) to reassess her coastal colonisation model, conceding that the evidence showed that adaptation to the exploitation of terrestrial resources had occurred much earlier than she had believed.

The occupation sites of Tasmania, 1 of only a few parts of Australia to provide an opportunity to test the proposals of speed of colonisation of the continent. At Parmerpar Meethaner a sterile layer is present that has been dated to about 40,000 BP, the earliest occupation layer having been dated to about 34,000 BP (Cosgrove, 1995). Cosgrove suggested that this has put the time of arrival of people in Tasmania at about 34,000 BP. If the excavator is correct it would suggest that there was a large time lag, of about 20,000 years between the time of the earliest arrival in northern Australia and the arrival of the first people in Tasmania. As it is not known how long the sterile layer took to accumulate it is not certain if this actually brackets the earliest arrival in the region, if it took a relatively short time to form it would suggest that the earliest time of arrival has not been effectively bracketed.


Archaeological evidence for occupation of arid areas was interpreted as opportunistic use of the upper reaches of coastal catchments or the landward section of coastal territories, when the dates obtained were of Late Pleistocene age (Bowdler, 1977). This interpretation was based on evidence from such sites as the Newman Site in the Pilbara dated to 21,000 BP (Maynard, 1980), and on the Nullarbor Plain, Koonalda Cave, 22,000 BP, (Wright, 1971and Allens Cave (N145), more than 20,000 BP (Martin, 1973). Puritjarra Rockshelter (Smith, 1987, 1989; Smith et al., in press in 1998) that has been dated to about 35,000 BP provided archaeological evidence of the early occupation of the arid central regions of the Australian continent. Later finds in the arid zone are such sites as the JSN site (14,400 BP (Smith et al., 1991) in the dunefields of the Strzelecki Desert (see Australian Archaeological Sites), Cuckadoo 1 Rockshelter, 15,000 BP (Davidson et al., 1993), Katumpul, about 22,000 BP (Veth, 1995: 36) in the Laverton region and at Serpent's Glen, about 24,000 BP (Veth & O'Connor, 1996). According to Smith the remaining evidence, though sparse, indicates that the arid interior of the continent was already being exploited over widespread areas in the Late Pleistocene, from at least 30,000 BP. The Little Sandy Desert (Veth & O'Connor, 1996) has provided more recent evidence, dates that the Pleistocene occupation pattern included some of the more arid areas of Australia that included the major continental dunefields. It has been suggested that this represents a form of opportunistic use of the arid zone (Bowdler, 1990a; Veth, 1995) and not a fully operating system (c.f. Smith, 1989). As more recent evidence accumulates Smith says it has become more difficult to maintain the interpretation of the evidence as possibly indicating opportunistic use of the arid zone.

The initial occupation of these sites had been believed to be controlled by their adaptation to new plant resources, mainly grass seeds and acacia seeds. The archaeological evidence is not consistent with this being the case, seed-grinding implements that are identifiable not being found until about 4,000-3,000 BP, suggesting that these plant foods were not the first resources to be exploited in the arid regions (Smith, 1986). Though direct evidence of prehistoric subsistence is rare, the faunal assemblages at the Silver Dollar Site, 25,000 BP, in which macropods and emu egg shells predominate, suggest the early adjustments were to these resources (Bowdler, 1990b). Small macropods also dominate the rich faunal assemblage found in Allens Cave (Cane, 1995) on the Nullarbor Plain where the occupation extends to about 39,000 BP. An early adaptation to very arid regions is demonstrated by the sites on the Nullarbor Plain, occupation at Allens Cave apparently continuing throughout the last glacial maximum, at which time the site was on a vast arid inland plain, situated on a saltbush steppe.

Situated in the south-eastern sector of the arid zone, Karolta has been reported to have been dated to about 30,000 BP, based on cation ratio assays and AMS radiocarbon dating of charcoal that was embedded in rock varnish (Dorn et al., 1988; Dorn & Nobbs, 1992). If these dates are confirmed it would strengthen the argument that fully operating regional systems were established early in the colonisation of the arid zone.

Islands See Archaeology of Sahul or Greater Australia - Melanesian Lowlands


Smith's conclusion is that the initial settlement of the coastal and riverine environments of northern Sahul probably took a few thousand years, and included the movement of people to the large islands to northern Melanesia, possibly involving deliberate as opposed to accidental voyaging. He also concludes that the northern Australian savannahs and those of the Sahul Shelf were probably widely occupied before the colonisation of the coastal environments was complete. He suggests that settlement began slowly, becoming more rapid over time, the drier climates of the interior leading to better health of the people in those regions that allowed more rapid dispersal, and the larger territories that were necessary to support each of these groups in the interior. Smith suggests similar factors may have operated in the colonisation of the central Cordillera of New Guinea, the richest terrestrial fauna being found in the montane forests and alpine habitats, with the added benefit of relief from malaria in the cooler climates.

He suggests that the northern and eastern coastlines of Australia, with their rich coastal habitats, probably promoted rapid selective settlement. On the west and northwest coasts the arid littoral zone probably promoted a move to exploitation of terrestrial resources at an early stage of colonisation, as part of a broad spectrum mixed economy. He suggests settlement of the arid zone may have taken place within a few thousand years of the first arrivals in Australia, though populations in the deserts have probably always been low and widely dispersed, and subject to the vagaries of the erratic climate of these regions. The last parts of the continent to be occupied may have been the major Australian deserts, both sand ridge and stony deserts, or at least the last places to have an established fully operating regional system, though he says it is uncertain if the colonisation of these arid areas lagged behind the colonisation of the northern savannahs by 10s of thousands of years.

In Tasmania, the high latitude temperate habitats were occupied at an early stage, based on the available archaeological evidence it appears that Tasmania was first occupied about 35,000 BP, though he considered the dates of 35,000 BP for both Tasmania and New Ireland to be difficult to reconcile with other archaeological evidence that indicates that Greater Australia was first occupied about 50,000-60,000 BP. He believes lags of 10s of thousands of years are unlikely to be correct.

See Stone Tools

Sources & Further reading

  1. M.A. Smith in Murray, Tim, 1998, Archaeology of Aboriginal Australia, Allen & Unwin.


  1. The Adventures of Archaeology Wordsmith

Author: M. H. Monroe
Last Updated 22/12/2011


Stone Tools

Deep Sea Fishing


Journey Back Through Time
Experience Australia
Aboriginal Australia
National Parks
Photo Galleries
Site Map
                                                                                           Author: M.H.Monroe  Email:     Sources & Further reading