Additive Partitioning of Shannon Diversity in Palaeolithic Zooarchaeology: A Case Study from Qesem Cave

Abstract Understanding how faunal diversity is internally organized remains a central but still underdeveloped issue in Palaeolithic zooarchaeology. This paper proposes additive partitioning of Shannon diversity to the published faunal datasets from Qesem Cave (Israel) to determine whether assemblage diversity is structured mainly by contrasts among predefined analytical groups or by heterogeneity retained within them. In practical terms, total Shannon diversity is separated into a between-group (interclass) component and a within-group (intraclass) component: the former describes how remains are distributed among analytical classes, whereas the latter describes the diversity preserved inside those classes. The results show that taxonomic diversity at Qesem is consistently dominated by the interclass component, indicating that contrasts among prey-size classes structure the assemblages more strongly than taxonomic heterogeneity within those classes. However, the intraclass contribution varies among horizons: SCW represents the most strongly interclass-structured assemblage, whereas SW Yab shows the highest intraclass contribution. This pattern is maintained under both NISP and MNI, although MNI makes intraclass taxonomic heterogeneity more visible and reduces the sharpness of some contrasts. Fragment integration increases total Shannon diversity unevenly across the sequence, showing that the fragmented but size-class-attributed fraction affects not only the magnitude of diversity but also its internal partitioning among prey-size classes. Anatomical diversity follows a different pattern: when each assemblage is considered a whole, anatomical profiles are comparatively homogeneous and slightly intraclass-dominated, while clearer contrasts emerge within size classes, especially among medium-sized prey, where diversity is strongly structured around proximal limb elements rather than evenly distributed across anatomical regions. Taken together, these results suggest that additive partitioning captures differences in how faunal assemblages were structured, accumulated, or segmented through time, rather than simply recording changes in prey spectrum. The approach is therefore best understood as a reproducible descriptive framework that formalizes assemblage structure and gains archaeological meaning when read alongside independent taphonomic, anatomical, and contextual evidence.