No abstract

Ethological reproductive isolation and genetic divergence across 26 protein loci were measured among populations of the salamander Desmognathus ochrophaeus in the southern Appalachian Mountains. Levels of ethological isolation varied from none to complete and were statistically significant for all but two pairings between populations inhabiting different mountain ranges. When geographic and genetic distances were treated as independent variables in multiple correlation analyses, they accounted for about halfthe variance in levels of ethological isolation. When genetic distance is held constant, the remaining relationship between ethological isolation and geographic distance is still statistically significant. When geographic distance is held constant, the remaining relationship between genetic distance and levels of ethological isolation is nonsignificant, as is the relationship between geographic distance and genetic distance when ethological isolation is held constant. Ethological isolation and genetic divergence evidently both reflect the gradual divergence of allopatric populations, but genetic distance is a poor predictor of ethological isolation in these salamanders.

We constructed a model for the evolution of sexual isolation by extending Lande's (1981) model of sexual selection. The model predicts that asymmetric sexual isolation is a transient phenomenon, characteristic of intermediate stages of divergence in sexually selected traits. Unlike the Kaneshiro (1976, 1980) proposal, our model does not depend upon drift and the loss of courtship elements to produce asymmetries in sexual isolation. According to our model, the direction of evolution cannot be predicted from asymmetry in sexual isolation. We tested some features of the model using data from an experimental study of sexual isolation in the salamander Desmognathus ochrophaeus. We tested for sexual isolation between 12 allopatric populations and found significant asymmetry in sexual isolation in about a quarter of the test cases. The highest degrees of asymmetry were associated with intermediate levels of divergence. A curvilinear relationship between isolation asymmetry and divergence was predicted by our model and was supported by statistical analysis of the salamander data.

No abstract

We constructed a model for the evolution of sexual isolation by extending Lande's (1981) model of sexual selection. The model predicts that asymmetric sexual isolation is a transient phenomenon, characteristic of intermediate stages of divergence in sexually selected traits. Unlike the Kaneshiro (1976, 1980) proposal, our model does not depend upon drift and the loss of courtship elements to produce asymmetries in sexual isolation. According to our model, the direction of evolution cannot be predicted from asymmetry in sexual isolation. We tested some features of the model using data from an experimental study of sexual isolation in the salamander Desmognathus ochrophaeus. We tested for sexual isolation between 12 allopatric populations and found significant asymmetry in sexual isolation in about a quarter of the test cases. The highest degrees of asymmetry were associated with intermediate levels of divergence. A curvilinear relationship between isolation asymmetry and divergence was predicted by our model and was supported by statistical analysis of the salamander data.