Description: Publication date: 15 January 2015 Source: Palaeogeography, Palaeoclimatology, Palaeoecology, Volume 418 Author(s): Travis Z. Wicks , Kaustubh Thirumalai , Timothy M. Shanahan , Christopher J. Bell δ 13 C records from vertebrate teeth offer an opportunity to gain insight into changes in past vegetation. Increasingly, teeth from small mammals are used for such purposes, but because their teeth grow rapidly, seasonal changes in vegetation are potentially a large source of variability in their carbon isotope composition, complicating interpretations of isotope data. To investigate the influence of seasonality on the stable isotope composition of fossil teeth, we constructed a Monte-Carlo-based model to simulate the effects of changes in the seasonal pattern of diet in leporid lagomorphs (rabbits and hares) on the distribution of δ 13 C values in random populations of leporid teeth from the Edwards Plateau in central Texas. Changes in mean-state, seasonal vegetation range, and relative season length manifest themselves in predictable ways through changes in the median, standard deviation, and skewness of simulated tooth δ 13 C populations, provided sufficient numbers of teeth are analyzed. This Monte Carlo model was applied to the interpretation of a 20,000 year record of leporid tooth δ 13 C values from Hall's Cave on the Edwards Plateau in central Texas. Populations of teeth from different climate intervals (e.g., the late Glacial, Younger Dryas, and the Holocene) display changes in tooth δ 13 C distributions that may reflect changes in seasonal vegetation. The inferred changes in seasonality of vegetation based upon our modeling are consistent with hypothesized climatic changes. These results indicate that small mammal teeth can potentially provide unique insights into climate and vegetation on seasonal and longer timescales that complement other approaches, but should be interpreted with a careful consideration of local conditions, taxon ecology and physiology, and the dominant timescales of isotope variability in a rigorous statistical framework.