Understanding Australia’s cultural history through archaeological geophysics

Abstract

The aim of this thesis is to develop and apply geophysical methods for Australian archaeology. The methods focus on magnetic susceptibility and ground-penetrating radar (GPR). The techniques are contextualised through application to the following four key archaeological questions: 1) Can magnetic susceptibility assist in resolving questions surrounding the potential downward movement of stone artefacts in rockshelter deposits? 2) Is human occupation persistent through the changing climatic regime associated with the last glacial maximum (LGM) at a Pleistocene-aged rockshelter in interior Australia? 3) How might we identify burials in a geologically complex rockshelter deposit? 4) How might magnetic susceptibility contribute to knowledge about the formation of ‘archaeologically instantaneous’ shell matrix sites?In exploring these questions, research was conducted at two rockshelters in northern Australia and on three shell mounds in the Gulf of Carpentaria, Australia. Magnetic susceptibility studies were undertaken at Gledswood Shelter 1 (GS1), a rockshelter occupied at ca 38,000 BP, to understand its history and formation processes. An experimental burning program using off-site samples was conducted to confirm that magnetically enhanced sediments in the cultural deposits were the direct result of anthropogenic burning rather than natural fires, pedogenesis or weathering. This change coincides with the level at which stone artefacts appear in the sedimentary sequence, indicating that they are in situ and have not moved down from higher layers above. Demonstrating that an increase in magnetic susceptibility is associated with human occupation is a crucial development in Australian archaeology. This will provide an opportunity to link sediments and artefacts—and this is critical to comprehending the timing of initial occupation of the continent.Magnetic susceptibility data combined with micromorphology and geoarchaeological data also revealed that occupation was continuous through the LGM at GS1, without any abandonment of the site. GS1 is situated in a region that has been characterised as a potential corridor for early colonists moving into the arid interior. The appearance of stone artefacts in the deposits corresponding with an increase in magnetic susceptibility as well as clay and charcoal coatings on quartz grains in the Pleistocene units in thin section, indicate that the site was occupied through this period, thereby implying that water was at least locally available. Despite the absence of any obvious permanent water sources, water availability at the site is  reliant on summer rainfall. This suggests that the monsoons driven by the Coral Sea off the northeast Australian coastline may have been active during this time. This has important implications for understanding climatic conditions during that period, and allows one to infer that water must have been available regionally for people to have maintained their use of the site.GPR carried out in advance of archaeological excavations at Madjedbebe, a sandstone rockshelter in western Arnhem Land, identified numerous subsurface rocks (large cobbles); excavation subsequently revealed these were associated with human burials. Post-excavation, geographical information systems (GIS) and statistical analysis clarified that a relationship between rocks and human burials exists. Graves were dug within the shelter and rocks were placed on the individuals before being covered. The rocks were the source of the strong GPR reflections and insights into burial practices derived from ethnographic sources further assisted with the geophysical interpretation. Application of this methodology provides an opportunity to test a way to identify unmarked burials at other rockshelter sites, and a useful management tool for Indigenous communities and heritage practitioners since it is noninvasive and non-destructive.The third group of sites is the shell mounds located in the Gulf of Carpentaria. Despite archaeological evidence including radiocarbon dates suggesting a single episode of deposition at these sites, the magnetic susceptibility combined with a range of sedimentary and archaeological analyses revealed that these shell mounds were repeatedly occupied. Results also demonstrated that magnetic signatures were related to cultural formation processes most likely from anthropogenic burning, rather than natural processes. These correlations between geophysical indicators and artefactual material suggest that the site retain a high degree of stratigraphic integrity. This has important implications for studies of other shell mounds sites, especially where the limitations of radiocarbon dating may mask multiple depositional events.In summary, this thesis demonstrates that both magnetic susceptibility and GPR studies can be valuable tools in deciphering key archaeological questions in the Australian landscape. The most important findings relate to the ability of magnetic susceptibility signals to clearly define levels at which humans first appear in the archaeological record. This will allow a major progress in determining the timing and dispersion of human settlements for Australian sites.