PETERSEN, J.; ADOLPH, S.C.*: Lizards in a Fluctuating Environment: a Stochastic Population Model for the Desert Night Lizard Xantusia vigilis

We constructed and analyzed a stochastic population model for the desert night lizard (Xantusia vigilis) in the western Mojave Desert of California. The model is based on demographic data of Zweifel and Lowe (1966), who found that winter rainfall strongly influences reproductive output in X. vigilis. Annual rainfall is highly variable in the Mojave Desert, and most rain falls during the winter. We used rainfall-dependent population projection matrices, combined with historical rainfall data, to simulate long-term population dynamics. We found that the stochastic population growth rate corresponded to &lambda = 0.97, and we obtained bootstrap confidence intervals for &lambda. The geometric mean &lambda was very similar to this value. A simplified stochastic model relating &lambda to annual rainfall was also quite consistent with the results of the full matrix model. These results suggest that simplified versions of matrix models may sometimes be useful for modeling population dynamics in fluctuating environments. Because Xantusia has low fecundity and high survival rates, annual fluctuations in precipitation are not likely to lead to large fluctuations in population size. We also examined the potential effects of directional climate change on stochastic population growth, by either increasing or decreasing the mean annual rainfall by a constant value (while retaining the year-to-year variation). In this analysis, stochastic growth rate varied approximately linearly with mean annual rainfall, with a moderate slope.