Rapid environmental change is a catalyst for human evolution, driving dietary innovations, habitat diversification, and dispersal. However, there is a dearth of information to assess hominin adaptions to changing physiography during key evolutionary stages such as the early Pleistocene. Here we report a multiproxy dataset from Ewass Oldupa, in the Western Plio-Pleistocene rift basin of Olduvai Gorge (now Oldupai), Tanzania, to address this lacuna and offer an ecological perspective on human adaptability two million years ago. Oldupai’s earliest hominins sequentially inhabited the floodplains of sinuous channels, then river-influenced contexts, which now comprises the oldest palaeolake setting documented regionally. Early Oldowan tools reveal a homogenous technology to utilise diverse, rapidly changing environments that ranged from fern meadows to woodland mosaics, naturally burned landscapes, to lakeside woodland/palm groves as well as hyper-xeric steppes. Hominins periodically used emerging landscapes and disturbance biomes multiple times over 235,000 years, thus predating by more than 180,000 years the earliest known hominins and Oldowan industries from the Eastern side of the basin.

Knowing to what extent lithic cores have been reduced through knapping is an important step toward understanding the technological variability of lithic assemblages and disentangling the formation processes of archaeological assemblages. In addition, it is a good complement to more developed studies of reduction intensity in retouched tools, and can provide information on raw material management or site occupation dynamics. This paper presents a new methodology for estimating the intensity of reduction in cores and tools on cobbles, the Volumetric Reconstruction Method (VRM). This method is based on a correction of the dimensions (length, width, and thickness) of each core from an assemblage. The mean values of thickness and platform thickness of the assemblage's flakes are used as corrections for the cores' original dimensions, after its diacritic analysis. Then, based on these new dimensions, the volume or mass of the original blank are reconstructed using the ellipsoid volume formula. The accuracy of this method was experimentally tested, reproducing a variety of possible archaeological scenarios. The experimental results demonstrate a high inferential potential of the VRM, both in estimating the original volume or mass of the original blanks, and in inferring the individual percentage of reduction for each core. The results of random resampling demonstrate the applicability of VRM to non size-biased archaeological contexts.

No abstract

No abstract