Urbanization and conversion of scrub to citrus groves have reduced the number and size of scrub habitat patches, increased isolation of patches, and inhibited natural dynamics of fire. Because scrub is a pyrogenic ecosystem, habitat quality may decline for many scrub species if scrub is not burned or otherwise restored. As a result of these impacts, more than 50 scrub species are recognized as endangered or of conservation concern including the Florida scrub lizard (Sceloporus woodi). We examine the landscape genetics of S. woodi and explicitly test for associations between landscape variables and genetic structure. If the matrix habitat between populations has no effect on gene flow then simple Euclidean distance can be used to characterize landscape connectivity. Alternatively, if matrix habitat affects scrub lizard movements then ecological metrics of connectivity may better describe movements of scrub lizards. We use microsatellite loci and a landscape approach to examine the genetic structure of S. woodi and test for effects of landscape variables at the scale of neighboring patches. We also evaluate alternative ecological metrics of connectivity using genetic data.
Results/Conclusions
Expected heterozygosity and allelic richness were associated negatively with the isolation of a patch as theory would predict for populations receiving a low number of migrants. Furthermore, both least-cost distance and pairwise isolation (a metric used in ecological studies that includes patch size, quality and an isolation index) were better predictors of genetic distance than Euclidean distance, indicating that mesic and hydric habitats surrounding scrub patches influence spatial patterns in genetic variation. The genetic differentiation observed at such small geographic distances indicates that lizard movement is limited and restricts the possibility for metapopulations to groups of scrub patches separated by at most a few hundred meters. Patches separated by more than a few hundred meters of mesic habitat may contain isolated populations with little opportunity for recolonization if local populations go extinct. Our results support the need for focusing on spatial distribution of scrub habitat at the scale of neighboring patches in conservation management and restoration. Also, our study points to the value of integrating landscape ecology metrics into landscape genetics.
