Abstract
Zooarchaeological age-at-death profiles for domesticated ruminants can be inferred from tooth eruption, replacement and wear. These profiles contain important information on slaughter management, and have been used informally to infer the goals of past husbandry strategies. In principle, sex-specific survival curves could inform on various productivity parameters, including herd growth rates and sustainability, milk and meat yields, macronutrient and calorie yields, and feed consumed. Knowledge of these parameter values would allow identification of differences in husbandry economics in different archaeological contexts. However, archaeological age-at-death profiles are rarely sex-specific and are typically derived from small sample sizes. As such, challenges remain in inferring sex-specific survival curves using explicit models that account for sampling uncertainty. We present a Bayesian method for inferring sex-specific survival curves from un-sexed cattle zooarchaeological age-at-death profiles that can accommodate data from any combination of age class boundaries. Our model relies on the assumption that asymmetric sex-specific slaughter leads to a sigmoidal change in sex ratio, which we show is broadly consistent with slaughter practices in modern unimproved cattle herds. By combining inferred sex-specific archaeological survival curves with ethnographic data from modern unimproved cattle, we generate estimates of herd growth rate, milk and meat yields, macronutrient and calorie yields, and feed consumed per animal. We use our framework to examine zooarchaeological age-at-death profiles previously proposed to be aimed at prioritizing milk or meat production, and a set of kill-off profiles from 10 Neolithic sites located across Europe. We infer that there is considerable scope for improvement in prehistoric slaughter management.