|Abstract or Summary
- This dissertation focuses on the evolutionary forces of genetic drift and gene flow in frog populations. The balance of these two forces and the force of mutation largely determine the amount of neutral genetic variation within populations as well as the degree of genetic similarity among populations. The stochastic evolutionary change caused by genetic drift can be quantified through the use of the effective population size (N[subscript e]) parameter. The effective size of a population is the number of breeding individuals in a conceptual, ideal population that would evolve by genetic drift at the same rate as the real population being studied. How a population responds to mutation, selection, and gene flow depends on N[subscript e], rather than the actual census population size (N). In most natural populations, N[subscript e] is considerably smaller than N. For these reasons, N[subscript e] is a fundamental parameter in basic population genetics theory as well as in applied conservation genetics. The degree of neutral genetic similarity between populations is highly dependent upon gene flow. When gene flow between a pair of populations is low, the populations are likely to become genetically differentiated. Conversely, when gene flow between populations is high, the populations will tend to be more genetically similar.
Amphibians are good model organisms for studying genetic drift and gene flow because they tend to exhibit strong population structure at small spatial scales. This is a consequence of their generally small population sizes, natal philopatry, limited dispersal capabilities, and restricted habitat requirements. They are expected to have easily-detectable signatures of spatial genetic structure and genetic drift. Amphibians can be used as models to further our understanding of evolutionary processes and that understanding can be applied to the conservation of amphibians. Equipped with knowledge of what naturally influences genetic drift and gene flow in amphibians, we can apply the principles of population genetics to mitigate the genetic consequences of amphibian declines.
In Chapters 2 and 3, I used molecular genetic data from frog populations to investigate N[subscript e] and the related parameter N[subscript b] (the effective number of breeders). Chapter 2 is a study of a single population of the Oregon spotted frog (Rana pretiosa). My aim was to determine where in the life cycle of this species the greatest reduction in N[subscript b] occurs. I used genetic data from microsatellites to estimate N[subscript b] at two different life stages, eggs and metamorphs, and found that estimates of N[subscript b] were similar at both stages. This result suggests that inflated variance in family size due to egg mass mortality is not a primary cause of N[subscript e] reductions relative to N in this species. Chapter 3 is a comparison of N[subscript e] estimates within and among four species of frogs in the family Ranidae: R. pretiosa, R. luteiventris, R. cascadae, and Lithobates pipiens. I obtained N[subscript e] estimates for 90 populations across the four species, using microsatellite data and several different estimators. The first three species and the western populations of L. pipiens have very small effective sizes (< 50). Eastern populations of L. pipiens are much larger, with N[subscript e] estimates in the hundreds and thousands. I also found significant correlations between N[subscript e] estimates and latitude, longitude, or altitude in R. luteiventris and L. pipiens.
Chapter 4 is a study of gene flow among populations of the Cascades frog (Rana cascadae) in the Olympic Mountains of Washington. I quantified genetic differentiation among 22 R. cascadae populations with data from microsatellite markers and used a landscape genetics approach to identify environmental features that have strong influences on gene flow in this species. I used a Random Forests statistical procedure to assess which of several structural connectivity models and 15 landscape variables explained the most variation in genetic distances among populations. I found that the best-fitting Random Forests models were based on different structural connectivity models for two datasets: 'within' and 'between' genetic clusters of populations. The landscape variables identified as the most important also differed across the two datasets, suggesting that landscape influences vary across spatial scales.
The results presented in this dissertation led to an increased understanding of effective population size in ranid frogs and of the environmental factors that influence population structure in R. cascadae. These studies provide a foundation for further research on the specific factors that influence genetic drift and gene flow in these species.