Reconstructing Early Iron Age subsistence: analysis of northeastern South African faunal assemblages using Zooarchaeology by Mass Spectrometry (ZooMS)

The introduction of agropastoralism in northeastern South Africa coincides with the arrival of the Bantu-speaking people in the first millennium CE. This transition is defined in archaeology as the “southern African Iron Age”, with pottery serving as the primary diagnostic artefact. Despite the widespread recognition of pottery as a key cultural indicator, faunal remains offer a more direct means of reconstructing past foodways and human-animal interactions at this key transition. Unfortunately, in many northeastern Early Iron Age sites, the fragmentation and paucity of specimens have limited the possibilities for dietary reconstructions. The fundamental issue of taxonomic identification of fragmented bone specimens is the first major hurdle, as morphology-based zooarchaeological studies can only achieve broad classification, potentially leading to misinterpretations in the composition of the faunal assemblage and thus the subsistence patterns of Early Farming Communities. Zooarchaeology by Mass Spectrometry (ZooMS), or peptide mass fingerprint of bone collagen, is a powerful tool for accurate identification of morphologically unidentifiable bone fragments. This study represents the first application of proteomics to investigate animal exploitation within Early Farming Communities contexts in the region. By analyzing 68 bone fragments from five archaeological sites – Lydenburg Heads, Riverside, Doornkop, Langdraai and Lebalelo – this research aimed to evaluate the suitability of bone remains for taxonomic identification using ZooMS and to gain new insights into faunal assemblage composition. Good collagen preservation enabled the identification of 60% of the analysed specimens. The presence of wild and domestic species (Bos taurus and Ovis aries) indicates the persistence of hunting during the introduction of agropastoral practices. Furthermore, the absence of browsing species suggests a selection towards grazers as a meat source, despite early farmers’ settlement preferences in river valleys and wooded environments. Beyond subsistence patterns, this study contributes to broader ethnoarchaeological debates on the cultural significance of animals. Finally, the identification of a canid specimen at the Langdraai site highlights the necessity for further proteomic research, which could enable differentiation between wild and domestic canid species. This project demonstrates the potential of biomolecular approaches to complement traditional zooarchaeological and ethnographic studies and highlights the broader applicability of ZooMS to other Early Iron Age sites, offering deeper insights into past human-animal interactions and Early Farming Communities’ subsistence strategies.