Sievers, Christine2013-07-302013-07-302013-07-30http://hdl.handle.net/10539/12922A thesis submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy 2013Cyperaceae (sedge) nutlets dominate the archaeobotanical assemblage of fruits and seeds recovered from the Middle Stone Age deposits at the rock shelter Sibudu, KwaZulu-Natal, South Africa (Sievers 2006). My aim is to investigate the implications of the nutlet presence in terms of human behaviour and to demonstrate that the nutlets were likely brought into the shelter on sedge culms (stems) deliberately harvested by people and informally placed on the shelter floor to provide “bedding”, a surface for working, resting or sleeping. I use various empirical and experimental approaches to confirm the use of sedges for bedding at Sibudu as early as ~77 000 years ago, almost 50 000 years earlier than any previously identified archaeological bedding. The bedding consists of the sedges Cladium mariscus subsp. jamaicense, Scleria natalensis, S. melanomphala, Cyperus sp. and a panicoid grass, identified through Scanning Electron Microscopy To investigate repeated and deliberate burning of bedding at Sibudu, I use experimental micromorphology and I compare the signatures of the Sibudu sediments with burned fresh sedge and grass bedding. I undertake further fire experiments, also in open air situations, to answer questions about burning sedge beds and the taphonomic implications. Experimental sedge bedding fires are hot and brief. The matrix beneath the fires affects the temperatures achieved both on the surface directly under the fire, and at depths of 2 cm and 5 cm below the surface; an ash matrix conducts heat more effectively than a matrix of 1–2 mm sized particles and allows for carbonisation of buried nutlets. The burning of dry and green bedding indicates that once the bedding is burning, the temperatures are sufficient to carbonise sedge nutlets below both dry and moist bedding. The methodological innovations I introduce are the use of experimental micromorphology to address an archaeobotanical question and the use of GIS-based coexistence analysis of southern African archaeobotanical data to make interpretations about past climate. The analysis develops previous palaeovegetation research in the area (Sievers 2006; Wadley et al. 2008) and provides an environmental context for people/plant activities at Sibudu.enMesolithic period - South Africa - Sibudu.Cyperaceae.Thesis