Modelling early modern human dispersal patterns through the Wallacean Archipelago

Archaeological records from Australia provide the earliest, indirect evidence for maritime crossings by early modern humans anywhere in the world. As the island archipelago of Wallacea has never been connected to the continental landmasses of Sunda in the west (mainland Southeast Asia) or Sahul in the east (Australia and New Guinea), water crossings through this region were essential for the human colonisation of Australia. The archaeological record for human movement in Wallacea is still a work in progress, however, and none of the evidence pre-dates the archaeological record from Sahul. In 1977 the American anthropologist Joseph B. Birdsell suggested two main initial routes from Sunda to Sahul, a northern route with a landing in New Guinea, and a southern route landing in northern Australia. Models of this significant colonisation event have since been suggested, however, there has been little consensus about which model describes the most parsimonious pathway for first landfall in Sahul. Likewise, later movements between islands is still in question. The dearth of sampled islands and sites in Wallacea has made testing movement models against archaeological evidence difficult, if not impossible. This PhD attempted to examine inter-island movements using a three-pronged approach: 1) archaeological survey and excavation of Wallacean sites, combined with review and revision of the archaeological literature, to identify the timings of early human occupation in the region and enlarge the number of archaeological sample points; 2) reconstructing the palaeogeography of the Wallacean archipelago for the likely periods of initial human colonisation, and through this, model island connectivity and develop a model for human movement through the region; and 3) identify proxy records of human movement between islands to provide information on networks of connection between prehistoric island communities, and in doing so test the reliability of the model outcomes. This study provides a much needed up date to Birdsell's model from over 40 years ago, with an innovative use of geographic information systems to digitise and expand upon the reconstructions of palaeo-Wallacea and the differing variables which would have impacted early modern human movements through the region. The resulting reconstructions and models identified a multitude of now-submerged islands throughout Wallacea which would have significantly increased the connectivity between islands in the prehistoric archipelago, produced the first reconstructions of palaeo-hydrology for the region, developed two measures of intervisibility between islands and in applying these calculations to the Wallacean archipelago, identified ca. 65 ka and ca. 45 ka as the periods of likely human arrival with the greatest intervisibility. This study is also the first to apply least-cost modelling techniques to sea travel, enabling the model to be expanded across the Wallacean archipelago for the first time. Models and palaeogeographic reconstructions constructed herein provide overwhelming support for an initial northern route through Wallacea, with a landing on the modern island of Misool, near the Papuan Bird's Head. Extensive archaeological research on islands of the southern route has so far failed to recover any dates which pre-date those for Sahul, and thus support this inference. Human translocations of cuscus (Phalangeridae) throughout Wallacea found good support for a humanly-assisted introduction from Papua to Timor ca. 3ka, supporting movement between these islands by at least the mid-Holocene