LAUSR

Abstract Mortality curves are a valuable tool to study taphonomy and to infer population structure in ancient communities; however, they rely on the precise determination of ontogenetic ages of individual animals, which is often uncertain for fragmentary and isolated fossil remains. In this paper, we develop a new protocol to construct mortality curves for fossil rhinocerotids using both tooth rows and isolated teeth, based on dental wear. We tested and calibrated this protocol using 29 skulls and 18 jaws of extant rhinoceroses (Diceros bicornis, Rhinoceros sondaicus, and R. unicornis). We then applied it to four extinct rhinocerotid species from the late early Miocene locality of Beon 1 of Montreal-du-Gers, SW France: the teleoceratines Brachypotherium brachypus and Prosantorhinus douvillei, the basal rhinocerotine Plesiaceratherium mirallesi, and the early elasmotheriine Hispanotherium beonense. Findings allow us to estimate population structures of these rhinocerotids, based on the percentage of juveniles, subadults and adults. The mortality curves of Beon 1 rhinocerotids reveal several major periods of vulnerability: birth, weaning, sexual maturity, and middle-late adulthood. When compared to extant populations (D. bicornis, Ceratotherium simum, and R. unicornis) and other ancient case studies (Miocene of USA, Pleistocene of Southeast Asia and France), mortality rate curves from Beon 1 are similar in early and late life; however, they lack the increased socially-driven mortality that is observed between 25 and 50% of the potential lifespan in some populations of D. bicornis and in the Miocene teleoceratine Teleoceras proterum. The population structures of Pr. douvillei, Pl. mirallesi, and H. beonense are close to those of extant populations of D. bicornis. In contrast, adult specimens of B. brachypus are overrepresented at Beon 1, similarly to Aphelops malacorhinus at Love Bone Bed (USA), probably due to taphonomical bias. It is important to note that considering isolated or associated teeth alone results in significant differences in mortality curves and population structures at Beon 1, hence both samples should be studied.