Tradeoffs between reproduction and survival are central to life-history theory and are expected to shape patterns of phenotypic selection, but the ecological factors structuring these tradeoffs and resultuant patterns of selection are generally unknown. We manipulated reproductive investment and predation regime in island populations of brown anole lizards (Anolis sagrei) to test (1) whether previously documented increases in the survival of experimentally non-reproductive females (OVX = ovariectomy) reflect the greater susceptibility of reproductive females (SHAM = control) to predation, and (2) whether phenotypic selection differs as a function of reproductive investment and predation regime. OVX females exceeded SHAM controls in growth, mass gain, and body condition, indicating pronounced energetic costs of reproduction. Although mortality was greatest in the presence of bird and snake predators, differences in survival between OVX and SHAM were unrealted to predation regime, as were patterns of natural selection on body size. Instead, we found that body condition at the conclusion of the experiment differed significantly across populations, suggesting that local environments varied in their ability to support mass gain and positive energy balance. As mean body condition improved across populations, the magnitude of the survival cost of reproduction increased, linear selection on body size shifted from positive to negative, and quadratic selection shifted from stabilizing to weakly disruptive. Our results suggest that reproductive tradeoffs and patterns of phenotypic selection in female brown anoles are more sensitive to inferred variation in environmental quality than to experimentally induced variation in predation.

Data were collected as records of size and survival of individual lizards in the field, then entered into Excel files. These primary data have been used to calculate additional measures such as body condition residuals, and to derive means, standard errors, least-square means, and other summary statistics for populations.

See associated ReadMe file. Methods of analysis are described in the associated manuscript.