African climate changed considerably throughout the Pleistocene (2.588 million (Ma) to 12 thousand years ago (ka)). The timing, rate, and magnitude of past climate change across the continent impacted the evolutionary and migratory history of many mammalian species, including hominins. Investigating paleoclimatic variability through time at local and regional scales allows for an assessment of the extent to which climate change affected hominin evolution in Africa. This dissertation presents three approaches for increasing the understanding of past climate change in Africa. One method is to critically synthesize the existing literature of African climate (n=85) and hominid demography (n=64) over a restricted time frame (150 ka to 30 ka) and specific spatial scale (regional). Results from this study are two-fold: 1) climate change in Africa during this period was variable by region, responding to different climate-forcing mechanisms, and 2) changes in population and climate were asynchronous and likely created alternating opportunities for migration into adjacent regions, including hominin migrations out of Africa (~140-80 ka). The second approach is to evaluate modern ecological relationships between species and their environment to better quantify interpretations of paleoecological records. A modern distribution study of 33 ostracode species from 104 sites in the southwest arm of Lake Malawi suggest that depth-dependent variables likely define species niches. Relationships between ostracodes, fish and the green algae Botryococcus, were used to inform the paleoecological interpretations in the third study of this dissertation. Additional results suggest that macrocharcoal is likely delivered to the lake basin via river rather than wind-borne methods. The third approach involves primary analysis of climatic indicators from the sedimentary record to chronicle paleoecological and paleoenvironmental change at the basin scale through time. Results from a 380.7 meter-long sediment core recovered from Lake Malawi indicate a change of state likely caused by local tectonism, which affected ostracode assemblages, but had little effect on lake level history through time. Furthermore, the local hydroclimate of Lake Malawi alternately covaried with global glacial/interglacial cycles and local insolation maxima over the past 1.25 Ma. The magnitude and frequency of hydroclimatic variability in the watershed will be further assessed in future research.

This thesis tests models on the dispersal of hominin populations in the Upper Pleistocene, specifically in relation to the Arabian Peninsula. It does so by conducting a quantified comparison of lithic assemblages from northeast Africa and southwest Asia. Lithic data from new excavations at the Jubbah Palaeolake in northern Saudi Arabia is compared to assemblages from Ethiopia, Sudan, Egypt, Jordan, Iran, Oman and other localities in Saudi Arabia. This is the first detailed inter-regional study of this area for Marine Isotope Stages 5 to 3, a critical spatial and temporal context in debates on both early modern human and Neanderthal demography and dispersal. The spatial and temporal character of the rich Arabian archaeological record correlate with emerging evidence for environmental change in Arabia; in particular the repeated dating of archaeological contexts to periods of climatic amelioration suggests that demographic growth was associated with periods of increased precipitation. The various factors influencing lithic variability and the methodologies by which they can be elucidated are reviewed. In particular this highlights the need for quantified and comparative analyses. A variety of analytical approaches are applied in this thesis, including the use of Correspondence and Principal Components Analyses to develop a nuanced view of lithic variability. Variability in cores is shown to largely reflect the related factors of size and reduction intensity. With analyses of debitage and retouched lithics a broadly similar picture emerges: assemblages which are heavily reduced have small cores and blanks and higher levels of retouch, and elements of shape variation also change in relation to reduction intensity. Elements of residual variability may reflect cultural differences. While it is felt that the evidence presented broadly orientates the Upper Pleistocene Middle Palaeolithic of Arabia to dispersals from Africa, this suggestion is subsumed with a problematization of using lithic evidence to understand dispersals. Aside from the need for further dated archaeological material from Arabia and surrounding countries, lithic analyses need to more thoroughly consider factors such as reduction intensity if we are to make robust inferences on population dispersals.