Lake Eyasi Basin of northern Tanzania plays a pivotal role in the study of human-environment interactions and in understanding human flexibility and adaptability through technological innovations over time and space. In this study, phytoliths from ancient soils and fossil pollen proxies from radiocarbon-dated sequences from Kisimangeda on the north-eastern edge of the Lake Eyasi Basin, are used to interpret trends in climatic changes recorded since the Last Glacial Maximum (LGM) to the present. We used pollen and phytolith abundances from a core that was recovered on the northern margin of saline Lake Eyasi at the depth of five metres. The application of principal component and cluster analysis, together with linear regression provides insight into dataset structure and grouping with reference to the modern comparative datasets that in turn allow us to classify the various palaeoenvironments and paleohabitats occupied by the late Later Stone Age, Pastoral Neolithic, and Iron Age inhabitants of Kisimangeda. The chronological order, pollen and phytolith records in the studied part of the basin signify palaeoenvironments analogous to the Somalia-Maasai bushland and grassland ecosystems of today.
Keywords: Palaeoenvironments; Late Pleistocene; Holocene; pollen; phytoliths; human adaptation