In 1995, the fiftieth anniversary of the end of the Second World War occasioned many reflections on the place of science and technology in the conflict. That the war ended with Allied victory in the Pacific theatre, inevitably focussed attention upon the Pacific region, and particularly upon the Manhattan project and its outcome. It was in the Pacific that Western physics and engineering gave birth to the Atomic Age. However, the Pacific war had also proved a testing time, and a testing space, for other disciplines and institutions. Extreme environments and opemtional distances, and the fundamental demands of logistics, required the Allies and the Japanese to innovate many scientific and technological practices. Just as medicine and botany were called upon to fight tropical diseases and insect pests, so engineers, anthropol ogists and geographers were called upon to understand local conditions and cli mates, and to work with local peoples whose traditional lives were changed forever by the experience. At the same time, the war played midwife to a host of new de velopments, not least in scientific intelligence and in chemical and biological weapons, which were to acquire far greater importance after 1945.

The research presented here is primarily concerned with human-environment interactions on the tropical coast of northern Australia during the late Holocene. Based on the suggestion that significant change can occur within short time-frames as a direct result of interactive processes, the archaeological evidence from the Point Blane Peninsula, Blue Mud Bay, is used to address the issue of how much change and variability occurred in hunter-gatherer economic and social structures during the late Holocene in coastal northeastern Arnhem Land. The suggestion proposed here is that processes of environmental and climatic change resulted in changes in resource distribution and abundance, which in turn affected patterns of settlement and resource exploitation strategies, levels of mobility and, potentially, the size of foraging groups on the coast. The question of human behavioural variability over the last 3000 years in Blue Mud Bay has been addressed by examining issues of scale and resolution in archaeological interpretation, specifically the differential chronological and spatial patterning of shell midden and mound sites on the peninsula in conjunction with variability in molluscan resource exploitation. To this end, the biological and ecological characteristics of the dominant molluscan species is considered in detail, in combination with assessing the potential for human impact through predation. Investigating pre-contact coastal foraging behaviour via the archaeological record provides an opportunity for change to recognised in a number of ways. For example, a differential focus on resources, variations in group size and levels of mobility can all be identified. It has also been shown that human-environment interactions are non-linear or progressive, and that human behaviour during the late Holocene was both flexible and dynamic.