Fine-scale adaptive divergence of amphibian populations in response to habitat-mediated selection

Background/Question/Methods

It is widely appreciated in ecology that movement of individuals among habitats can increase the size and genetic variation of the recipient populations. However, evolutionary theory suggests that the gene flow associated with this movement may negatively affect population persistence by preventing adaptation to localized environmental conditions. With increasing evidence that adaptive divergence among populations can occur over short time scales, merging the ecological and evolutionary perspectives will become increasingly important in ecological studies. This includes the use of molecular genetic tools to quantify gene flow and population connectedness. In this study, I estimated the phenotypic divergence among populations of the wood frog (Rana sylvatica) that inhabit seasonal ponds differing in key habitat attributes. Specifically, I performed a field-based reciprocal transplant experiment between eight pairs of neighboring open and closed-canopy ponds – pond types that have been shown to present different selection pressures for larval amphibians. I evaluated the response of larval wood frogs (including survival, growth, and developmental rates) to both their natal and introduced habitats. Additionally, I used microsatellite-based genetic methods to estimate rates of gene flow and the genetic divergence among the focal populations. I also estimated traits related to egg quality in order to determine what role maternal effects might play in any observed divergence.     

Results/Conclusions

I found strong evidence of phenotypic divergence between ponds of dissimilar canopy cover. Wood frog larvae originating from closed-canopy ponds were 27% heavier and 10% longer than those originating from open-canopy ponds. Closed-canopy individuals also developed faster (Gosner staging method), but this difference was only significant when analyzing the pond pairs independently. There were no differences in survival between the two pond types, particularly at the early stages of development. These experimental field results indicate that wood frog populations can diverge phenotypically over very small geographic scales. However, analyses of the microsatellite data show very little genetic differentiation between populations within each pair, indicating high gene flow. These genetic insights could indicate that selection is intense in this system, as populations are able to diverge phenotypically in spite of high gene flow. Alternatively, this may indicate that adaptive divergence can occur quite rapidly in this system as canopy coverage changes over time. Lastly, preliminary results related to egg quality do not support maternal effects as a primary driver of these phenotypic differences among populations.