Prehistoric human diet composition in Northwest Patagonia: Complementing isotopic analysis with zooarchaeological information

Abstract

In this work, we study diet composition of prehistoric human populations of Northwest Patagonia by exploring the combination of two different approaches frequently used in bioarchaeology, Bayesian isotope mixing models, and zooarchaeological analysis. For this purpose, we compiled a large dataset of previously published δ¹³C and δ¹⁵N human and resource values, as well as zooarchaeological data, corresponding to the Middle–Late Holocene and distributed throughout Northwest Patagonia. We first propose a replicable approach to divide the region into different areas comprising human individuals that shared the same available resources and perform isotope mixing models at individual level using default (i.e., uninformative) prior distributions in the Bayesian mixing models. Then, we explore a potential complementation of isotopic and zooarchaeological evidence by introducing the frequency of zooarchaeological assemblages with the different resources, as well as NISP, as priors in the Bayesian mixing models to guide diet estimations. Additionally, we use these frequencies to analyze the possible absence of some important resources in previous analyses. Based on the species distribution and geographical location of bioarchaeological sites, we divided the region into five areas (Northwest and Northeast Neuquén, South Neuquén/Rio Negro, and Southwest and Central-east Mendoza) that differ in the resources available for potential human consumption. The results obtained show high diet variability among these areas, with individuals from Northwest and Northeast Neuquén consuming mainly large animals (i.e., guanaco and rhea, respectively) and South Neuquén/Rio Negro and Center-east and Southwest Mendoza consuming a larger diversity of resources. Our results show that considering zooarchaeological priors in the analyses results in diet composition estimations more in line with the ecological diversity present in the region than previous estimations. We propose that approaches like this, which are common in ecological studies, should be considered in bioarchaeology to make more robust estimates of diet composition.