A continental narrative: Human settlement patterns and Australian climate change over the last 35,000 years

by Alan N. Williams, Peter Veth, Will Steffen, Sean Ulm, Chris S. M. Turney, Jessica M. Reeves, Steven J. Phipps, Mike Smith

Read the paper here. Below is the abstract:

Drawing on the recent synthesis of Australian palaeoclimate by the OZ-INTIMATE group (Reeves et al., 2013a), we consider the effects of climate systems on past human settlement patterns and inferred demography. We use 5044 radiocarbon dates from ∼1750 archaeological sites to develop regional time-series curves for different regions defined in the OZ-INTIMATE compilation as the temperate, tropics, interior and Southern Ocean sectors to explore human–climate relationships in Australia over the last 35,000 years. Correlations undertaken with improved palaeoclimatic data and archaeological records indicate that the regional time-series curves are robust, and can be used as a proxy for human behaviour. However, interrogation of the datasets is essential with artificial peaks and taphonomic over-correction being critical considerations. The time-series curves are interpreted as reflecting population growth, stasis and even decline in phase with terminal Pleistocene/early Holocene climatic fluctuations. This coupling, however, decreases during the last 5000 years, most likely due to increased population levels, greater territoriality, technological solutions to stress, and social and ideational innovation. Curves from all sectors show exponential population growth over the last 5000 years. We identify future research priorities, highlighting the paucity of archaeological records across several parts of Australia (<1 dated site/4,000 km2), especially around the fringes of the arid zone, and the need for improved taphonomic correction techniques. Finally, we discuss how these time-series curves represent a first-order framework, not dissimilar to global climate models, which researchers can continue to test and refine with local, regional and continental records.