Neanderthals in the desert? Population history best explains Middle Palaeolithic technological heterogeneity in Arabia

Abstract

Northeast Africa and Southwest Asia present a rich record of hominin occupations between ca. 100 and 50 thousand years ago (ka) at a critical biogeographic crossroads, with differences in physical landscapes, environmental responses to climate change, and varied population histories contributing to archaeological diversity. Research in Arabia has identified previously unexplained technological heterogeneity amongst stone tool assemblages. To investigate this pattern, we compiled a database of Levallois blank attributes to quantify patterns of variability observed across sites in Northeast Africa, Arabia, and the Levant. A range of alternate explanations for variability within this comparative dataset are explored using Simple Mantel tests and Multiple Matrix regressions including assemblage age, landscape setting, palaeoenvironmental conditions, assemblage size, composition and raw material use, site type, and population history. The results highlight that shape variability in Levallois blanks across the region is best explained by differences in population history. We then train logistical regression models to differentiate the population history associated with Levallois blanks with a >93% agreement rate and apply this model to key Arabian assemblages dating to ca. 55 ka. The results strongly support a Neanderthal attribution. We interpret these findings in the context of wider debates surrounding Neanderthal expansions and a wider role for Arabia in South-West Asian hominin demography.