Meeting Abstract

24.4  Monday, Jan. 4  Stochastic Population Dynamics of a Desert Lizard ADOLPH, S.C.**; DAVIS, A.R.; FEDEROWITZ, M.; PETERSEN, J.; Harvey Mudd College; University of California, Berkeley; Harvey Mudd College; Harvey Mudd College adolph@hmc.edu

Desert organisms experience highly variable precipitation, which can strongly affect reproduction and therefore population dynamics. In the viviparous desert night lizard (Xantusia vigilis), Zweifel and Lowe (1966) found that fecundity is highly correlated with the previous winter’s rainfall. We used Zweifel and Lowe’s demographic data to construct a rainfall-dependent population model. A simple matrix model with (a) constant, age-independent survival rate and (b) fertility rates that increase linearly with rainfall predicted the historical data very well. We combined this model with historical rainfall data to simulate long-term stochastic population dynamics. Although litter size only varies from 0 to 3 in this species, population size could fluctuate substantially due to random runs of either wet or dry years. The resulting long-term stochastic population growth rate corresponded to an average annual growth rate of λ = 0.97, not appreciably different from a stable average λ = 1.0. A simplified (non age-structured) stochastic model relating annual rainfall to λ captured the essential behavior of the age-structured model. We used our model to explore climate change scenarios by modifying the historical rainfall distribution. λ increased linearly with mean rainfall but was relatively insensitive to changes in the variance of the rainfall distribution. Finally, we revisited Zweifel and Lowe’s Mojave Desert field site in two different years — an extremely wet year and an extremely dry year — to obtain additional demographic data on these lizards. Our samples, taken more than 50 years after their pioneering study, confirmed the strong dependence of fecundity on rainfall.