Quantification of climate and vegetation from southern African Middle Stone Age sites – an application using Late Pleistocene plant material from Sibudu, South Africa

In southern Africa numerous Middle Stone Age (MSA) sites document important steps in technological and behavioural development leading to significant changes in the lifeways of modern humans. To assess whether these cultural changes and developments may be related to environmental changes we need to ascertain past environments. To do this we apply a new quantitative method, the GIS-based Coexistence Approach (CAGIS), on fossil plant material from the MSA site Sibudu, KwaZulu-Natal, South Africa. Previous qualitative environmental interpretations of the fossil fauna and flora of the site remain ambiguous. Because much of the material is anthropogenically introduced, it is difficult to distinguish between the effects of natural changes in the local vegetation and behavioural changes of the people that inhabited the shelter. CAGIS can be applied to such biased assemblages and seems to be an adequate method to directly quantify palaeoclimate and vegetation parameters at an archaeological site. The CAGIS analysis shows that during the Howiesons Poort (HP) Industry winters were slightly colder and drier than present, whereas during summer, temperatures and precipitation were similar to today. Post-HP winters were drier and colder than present, presumably colder than during the HP. Summer temperatures remained the same, but summer precipitation decreased from the HP to post-HP. Vegetation cover was less than today, may be even less than during the HP. The late MSA was observably warmer than the older periods, especially during winter. At the same time summer precipitation slightly increased and vegetation became more dense, but still remained generally open similar to today's anthropogenic landscape. Generally, climatic changes are most pronouncedly reflected in winter temperature parameters, especially in minimum winter temperatures, and to a lesser extent by changes in summer precipitation. The observed ecological trends seem to be affected mainly by variations through time in winter temperatures. This refinement of interpretation was not discernible using previous methods for analysing the Sibudu data.