Australasian Science: Australia's authority on science since 1938

Out of Asia

Credit: Jamie Tufrey

Credit: Jamie Tufrey

By Sue O’Connor

The discovery of ancient fish hooks and the bones of offshore fish species reveals that the people living to the north of Australia more than 50,000 years ago had the maritime skills and equipment necessary to reach Australia.

Some time prior to 50,000 years ago, modern humans left mainland Asia (Sunda) and began the first of a series of maritime voyages that was to culminate in the colonisation of Sahul (Australia and New Guinea). This was a remarkable accomplishment at this early date, and demonstrates the capacity for complex planning and technological innovation that has become the hallmark of our species.

Firstly, it required the successful navigation of many water crossings between islands. Secondly, the colonists arriving on new island shores had to adapt to unfamiliar faunas, floras and landscapes. Finally, on an archaeological timescale, the journey from one continental region to the other appears to have been accomplished very rapidly, and more than 30,000 years before the colonisation of the Americas.

Obviously the crossing from Sunda to Sahul was not accomplished by a single founder group or within a few generations. We do not know how many millennia or generations passed or how many attempts at colonisation of islands en route may have ended in failure or extinction.

However, we do know that people were moving through the northern savannah regions of Australia by 50,000 years ago, and that by 45,000 years ago they had migrated east as far as the Bismarck Archipelago. By 35,000 years ago some of the smallest and most remote of the South-East Asian islands to the north of Australia, such as the Talaud-Sangihe group, had been settled.

What were the unique cultural adaptations and behaviours that were needed to make this journey?

Boats and Biogeography
During times of global low sea levels, the southernmost extension of the Eurasian continent comprised Sumatra, Java, Bali, Borneo and the now-drowned shelf in between (Sunda; Fig.1). New Guinea, Australia and the Aru islands were joined as a separate continent, Sahul. Between these two continental landmasses are the 17,000 islands of Wallacea. During the time span covered by human evolution, Sunda, Sahul and the islands of Wallacea have never been connected by a land bridge, although some of the Wallacean islands have been combined as larger islands during low sea stands.

At the time of initial human colonisation of Australia, global sea level was 60–70 metres lower than it is today. While this would have slightly reduced the distances between island hops, at least eight separate crossings and one leg of about 70 km would have had to be made, regardless of the route taken. Archaeologists have been perplexed by this voyaging accomplishment as the earliest sites display no hint of the level of maritime capacity that must have been required to make these open water crossings.

The observations made by early European explorers in Australia indicate that the watercraft in use were simple canoes and rafts. The dugout canoes of northwest Australia and those with outriggers that were in use in Torres Strait and along the Queensland coast seem to have been relatively recent introductions from outside Australia – appearing in only the last few thousand years – rather than the continuation or development of pre-existing maritime technologies.

Likewise, the shell and bone fish hooks found in archaeological coastal sites in eastern Australian first appear in the north-east about 1200 years ago. The distribution and chronology of the hooks indicates that they were introduced from outside Australia, possibly from across Torres Strait or even from the east by early Polynesian voyagers.

The islands around the Australian coastline were mostly formed 8000–6000 years ago as global sea level rose. Few of these islands, with their abundant supplies of sea birds, seals and fish, have any evidence of human visitation prior to about 3500 years ago.

If we look at the earliest records of occupation from sites in Australia and New Guinea, which date to between 55,000 and 50,000 years ago, they do not suggest that the early migrants had high-level maritime skills. The scant evidence for the exploitation of marine resources in the few Australian archaeological sites that were close to the coastline at the time of first settlement, and the lack of sophistication in watercraft recorded ethnographically, have long puzzled researchers and led some to ponder whether the crossing to Australia might not have been fortuitous, resulting from coastal raft-using communities swept south-east following a storm or other natural event. However, it should be remembered that the earliest sites on the coastline of Australia at the time of human arrival would have been drowned as global sea level rose 18,000 years ago.

The sites that we find today tell us about early savannah occupation, but are too distant from the coast to inform us about the maritime abilities of those who made the sea journeys to get here. If we want to find out about the nature of early maritime use we need to look at regions where the offshore profile drops very steeply into deep water, as only in these regions will the rise and fall of sea level have had little impact on the shape of the coastline.

The Maritime Skills of the First Settlers in East Timor
On the north coast of East Timor an unusual combination of geological factors have preserved cave sites that contain evidence of coastal exploitation dating back to 42,000 years ago (Fig. 2). The archaeological discoveries from these sites provide us with a unique window on the maritime skills of the early voyagers in the islands to our north, and perhaps some indication of what the earliest Australian coastal sites might look like if they were preserved.

Early Fishing Practices
Archaeological cave sites in this region have yielded significant findings over the past decade, but the recent archaeological investigations at Jerimalai shelter (Fig. 3) produced some surprises. Among the bones left behind from ancient meals, the number and volume of fish bones far exceed other fauna. Significantly, fast-swimming open ocean species such as tuna comprise almost 50% of the total fish assemblage in the earliest levels dated between 42,000 and 38,000 years ago. The remains of marine turtles are also abundant.

The bones found at the site indicate that when humans first began to use this site there was a limited range of vertebrates available as food. Aside from marine animals, only small quantities of bone from rodents, bats, birds and reptiles (mainly pythons and smaller lizards) were found.

While the land-based fauna in Jerimalai looks very unappetising, the marine fauna was evidently abundant and diverse. More than 38,000 fish bones were recovered from a single 1 m2 test excavation (Square B) at Jerimalai shelter (Fig. 4). From these bones, 23 different taxa were identified. Parrotfish, trevally, triggerfish and groupers were also common. Other major fish families identified are emperors, snappers, unicornfish, wrasse and puffers. Rays and sharks were also caught.

Jerimalai shelter appears to have little evidence of occupation between 38,000 and 24,000 years ago. This may be because falling sea levels made the shelter less accessible to the coast during this time.

Between 18,000 and 9000 years ago sea levels rose again, and by 9000 ago Jerimalai would have been almost as close to the coast as it is today. The overall ratio of offshore fish exploited remained high between 9000 and 7000 years ago (46%), but there were a few changes at this time. Trevally numbers caught dramatically increased during this phase while the exploitation of tuna remained similar. In terms of inshore species, the number of groupers and triggerfish increased during this phase.

The last 6000 years of fishing at Jerimalai presents a striking contrast with the older horizons. There has been a dramatic increase in the proportion of inshore reef species caught (66–76%) compared with tuna and other fast-swimming offshore species (34–24%). The increase in inshore reef species at this time correlates well with sea level stabilisation, which began about 6000 years ago, and the warmer sea surface temperatures of this time. The stabilisation of sea levels would have enabled coral reefs to establish along the coast, providing ideal habitats for inshore reef fish.

The World’s Oldest Fish Hooks and Other Maritime Technology
The fish bones from Jerimalai clearly demonstrate that the level of maritime capacity and fishing technology of the early modern humans who colonised our region was much higher than previously believed. Fast-moving fish such as tuna are more difficult to capture than inshore reef fish, which are more territorial and can be caught using baited stationary hooks.

Since no hooks and lures were found from this earliest period, it is uncertain what methods were used to catch the tuna. Ethno-archaeological and ethnological studies of contemporary fishing in South-East Asia and Oceania indicate angling or trolling, although the use of nets is also a possibility. Simple fish-

aggragating devices such as tethered log floats could have been used to attract tuna, which could have then been caught by angling ot trolling. An ancient method of catching tuna used historically in Andalusia and Sicily employed set maze nets that lead to a central pool.

The size of the tuna captured at Jerimalai indicates catches of immature individuals, with most estimated at 50–70 cm or less based on comparison with modern specimens. While tuna are usually caught in the open ocean, immature individuals sometimes come in close to the shore. It is possible that they were caught in the deep channel that runs between the eastern end of Timor and Jaco Island. Even if the tuna were caught close to the shore, these powerful fish would have required high-level fishing skills to capture.

Two partial fish hooks were recovered at Jerimalai. The oldest is dated to between 23,000 and 16,000 years ago, and is the earliest example of such technology anywhere in the world (Fig. 5).

This finding is only a fragment, but an almost complete hook was recovered from nearby Lene Hara Cave dated to ~11,000 years ago (Fig. 6). Based on what we know about similar hooks in use in Melanesia in historic times, these are single-piece jabbing hooks used for angling. They would have had a fibre line tied to the shank, and the hook would have been baited. This type of hook would have been effective for inshore reef fishing.

No artefacts related to netting and trapping have yet been recovered, but the hooks would have been attached to strong fibre lines.

The recent findings from East Timor demonstrate that maritime skills and complex technologies to target tuna and other fast-swimming open water species were in place by 42,000 years ago. It seems likely that early modern humans moving out of mainland Asia would already have possessed a fair degree of maritime capacity that would have assisted them to begin the journey through the islands of South-East Asia.

Settlement on the islands would have presented many challenges to the first colonists, not least of which would have been the restricted range of land animals available as game. A focus on marine resources would have been a successful strategy for overcoming these challenges.

Sue O’Connor is Professor of Archaeology and Natural History at The Australian National University.