Interpreting Geographic Variation in Life-History Traits

Abstract

The geographic variation in the length of the larval period and the size at metamorphosis of the wood frog, Rana sylvatica, is examined for populations in the tundra of Canada, the mountains of Virginia, and the lowlands of Maryland. We argue that the observed differences in developmental plasticity, heriisbilities and genetic covariances of traits among localities result from differential selection pressures in each environment, and are related to the physiological constraints inherent in development and to the degree of compromise between the timing and size at metamorphosis allowed in each environment. In Maryland populations fitness has been maximized by evolutionary changes in size alone; body size in this population is canalized, has low heritability and is highly correlated with juvenile survival relative to developmental time. In Canada, minimum developmental time yields maximum fitness; the length of the larval period in this population is canalized and genetically monomorphic relative to body size. In contrast, fitness in the Virginia populations has been determined by correlated and pleiotropic effects of genes on both developmental time and larval body size, and both traits are equally canalized, affect juvenile survivorship equally and display moderate heritabilities. These results stress the importance of interpreting variation in life-history traits relative to constraints inherent in development and those imposed by the environment. Heritability and survivorship data support the general notion that fitness traits should have low levels of additive genetic variation, but also suggest that antagonistic pleiotropy may act to preserve genetic variation in fitness traits under simultaneous selection, and caution against inferring evolutionary importance of individual traits without considering the possible presence of pleiotropy .