M. A. Halverson

We examined the effects of inbreeding on the performance of wood frog (Rana sylvatica) larvae in the field and in the laboratory. We used microsatellite analysis to establish the parentage and degree of inbreeding of the larvae. Two different estimators of inbreeding were used. The first was based on average multilocus heterozygosity, and the second was based on a molecular relatedness estimator. The estimators were highly correlated, and both showed a significant negative relationship between inbreeding and survival in the wild. However, there was no evidence that inbreeding influenced growth or development in the wild. Neither was there any evidence that inbreeding affected survival, growth, or development in the laboratory. These results suggest that, for wood frogs, inbreeding has a bigger effect on fitness in the wild than in captivity and that measurements of survival are more sensitive than measures of growth or development.

According to kin selection theory, the location of an individual with respect to its relatives can have important ramifications for its fitness. Perhaps more than any other vertebrate group, anuran amphibian larvae have been the subject of many experiments on this topic. Some anuran species have been shown in the laboratory to recognize and associate with their siblings and half-siblings. However, due to the difficulty of identifying sibships, no kinship studies with anuran larvae have been conducted in the wild. Here, we use microsatellite analysis to show that wood frog (Rana sylvatica) tadpoles were nonrandomly distributed in two ponds with respect to their relatives. In one pond, the tadpoles were significantly clumped with their siblings or half-siblings as expected from other published laboratory studies on this species. However, in another pond, the tadpoles were significantly nonrandomly dispersed from their siblings or half-siblings. This is the first example of kin repulsion of nonreproductive animals in the wild and the first time a species has been shown to display both aggregation and repulsion under different circumstances. These results suggest that kin distribution is context dependent and demonstrate the importance of testing kin selection hypotheses under natural conditions.