Meta-analysis of genetic diversity patterns

Quantitative reviews and meta-analyses are of increasing utility in ecology, evolution, and conservation. In order to answer evolution- and conservation-related questions, I led a ten person team in the review of ~5000 primary literature articles, resulting in a database of genetic diversity estimates for nearly 2000 vertebrate populations. The resulting database has been used in two projects to date.

Genetic diversity in threatened vertebrates

Genetic diversity is necessary for evolutionary response to changing environmental conditions such as those facing many threatened and endangered species. Using our database of genetic diversity estimates, my coauthors and I showed that genetic diversity is  reduced in threatened species, suggesting that inbreeding and drift are both effective at removing genetic diversity in endangered populations. Additionally, we showed that the criteria typically used to rank species of conservation concern (declining population size, species range extent, the number of mature individuals) do not systematically identify populations with low genetic diversity. Finally, we suggested a novel approach for identifying species of conservation need by estimating the expected loss of genetic diversity and show that our approach performs significantly better than the existing methods. We hope these methods will be used to identify species that merit conservation concern in part because of reduced genetic diversity.

Proposed IUCN criterion (a) and example of usage (b). We proposed estimating census size and effective population size (Ne) in order to determine the number of generations (t) until the species loses heterozygosity below the 25% of values (Ht) gathered in our database of microsatellite studies. The number of generations can then be used as an indicator of conservation need, based on the loss and projected future loss of genetic diversity.Proposed IUCN criterion (a) and example of usage (b). We proposed estimating census size and effective population size (Ne) in order to determine the number of generations (t) until the species loses heterozygosity below the 25% of values (Ht) gathered in our database of microsatellite studies. The number of generations can then be used as an indicator of conservation need, based on the loss and projected future loss of genetic diversity.

Migratory behavior and biome influence genetic diversity

Genetic diversity is largely determined by effective population size, which may vary dramatically for species that differ in migratory behavior or in the biome they inhabit. My coauthors and I examined relationships among genetic diversity, taxonomic class, biome, and migratory behavior (see figure below for methodological outline). We found that migratory mammals, reptiles, amphibians, and fishes had less genetic diversity compared to nonmigratory species, whereas migratory birds had more genetic diversity than their nonmigratory counterparts. We also found that the biome a species inhabits influences the genetic diversity of migratory and nonmigratory species differently. These differences are likely due to differences in vagility and  philopatry among the classes and perhaps differential selection between terrestrial and aquatic species.

Heuristic overview of data collection, analysis and interpretation. We utilized four sources of data (i.e. genetic diversity, taxonomy, migratory behaviour and biome) to evaluate differences in genetic diversity between migratory and nonmigratory species (Analysis A) and the interaction between migratory behaviour and species biome (Analysis B). In both the analyses, we assessed significance by bootstrapping the linear model coefficient estimates and comparing the coefficient estimates.

Heuristic overview of data collection, analysis and interpretation. We utilized four sources of data (i.e. genetic diversity, taxonomy, migratory behaviour and biome) to evaluate differences in genetic diversity between migratory and nonmigratory species (Analysis A) and the interaction between migratory behaviour and species biome (Analysis B). In both the analyses, we assessed significance by bootstrapping the linear model coefficient estimates and comparing the coefficient estimates.