The hunters or the hunters: Human and hyena prey choice divergence in the Late Pleistocene Levant

Hunting preferences reveal a great deal about the life of Paleolithic humans, and may reflect changes in human demography, technology, and adaptations to changing environments. However, the effects of hunting preferences and environmental availability are often conflated, stressing the need for comparisons to other predators that exploited the same environment. Manot Cave (Israel), preserved rich Early Upper Paleolithic (46-33 ka) human occupations, along with repeated spotted hyena (Crocuta crocuta) occupations, allowing us to compare anthropogenic and biogenic bone assemblages within the same space and time frame. We focused on the faunal remains retrieved in the middle of the cave (Area D), and conducted detailed taphonomic and zooarchaeological analyses. The Area D archaeofauna was dominated by Mesopotamian fallow deer (Dama mesopotamica) and mountain gazelle (Gazella gazella), which showed high abundance of carnivore damage. This and the carnivore-to-ungulate ratio, the presence of juvenile hyena and numerous coprolites match the criteria of a hyena den, confirming that the bone assemblage was created mainly by hyenas. Manot Area D thus reveals hyena prey choice in the Upper Paleolithic Galilee, which we then compared with human prey choice. Our results showed that hyena prey assemblages in Manot and elsewhere in the Levant were Dama-dominated while human assemblages were dominated by Gazella, demonstrating that hyenas and humans hunted different animal size groups, possibly in different habitats. We interpret this phenomenon as resulting from two possible scenarios: the emergence of projectile technology which may favor hunting in open environments, and niche partitioning derived by human-hyena competition. Hyenas were abundant and hunted unselectively while the anthropogenic record presents population turnovers and some dietary diversification. Whereas both scenarios are difficult to test directly with the available evidence, we argue that either one explains quite well the late Pleistocene archaeofaunal patterns in the Levantine record.