Lizard distributions on islands: Community ecology and biogeography

The dissertation employs process-based, bioenergetic modelling to understand the distributions of Anolis lizards on Lesser Antilles Islands. Extending our current pattern-based understanding of species distributions to the underlying processes is essential to predicting distributions in unsampled regions and in response to environmental change. I present one of the first models to integrate from individual energetics to population dynamics to predict spatial distributions. The model, which is parameterized with body size, thermal physiology, and prey abundance, predicts lizard abundance as a function of environmental temperature. The model is empirically validated and tested on islands with and without species interactions and with differential topography.; The bioenergetic model successfully predicts how the abundance of a solitary lizard species declines with increasing elevation. The empirical trends for the northern and southern one-species islands have similar slopes but different intercepts at sea-level, corresponding to differential prey abundance. The empirical anole abundance trends diverge from the bioenergetic model predictions on the two-species islands, where one species is more abundant at sea level and the other more abundant at higher elevations compared to the trend for solitary anoles. Incorporating competition in the bioenergetic model accounts for how species interactions superimpose on the responses to environmental temperature.; I then test whether the bioenergetic model with competition can predict patterns of species coexistence in the Grenadines, a chain of small one- and two-species islands. The model successfully predicts coexistence on the largest islands and competitive exclusion on the smallest islands. Additional island characteristics such as habitat diversity, resource availability, and temporal disturbance patterns may prevent coexistence on the mid-sized islands.; Additional empirical research addresses how habitat partitioning interacts at multiple spatial scales and how island characteristics influence lizard density. I demonstrate that patterns of local habitat partitioning vary in response to landscape abundance trends. I document a hump-shaped density area relationship for anoles on the Grenadines, corresponding to the initiation of habitat diversity. The dissertation couples empirical and theoretical approaches to understand how species resolve environmental variation into biological habitats at multiple, interacting spatial scales.