Applying Conservation Genomic Techniques to Guide Management of the Reticulated Flatwoods Salamander (Ambystoma bishopi)

The Reticulated flatwoods salamander (Ambystoma bishopi) is a federally endangered amphibian endemic to the longleaf-pine ecosystem of the southeastern U.S. This study used analyses of single-nucleotide polymorphism (SNP) data, collected from 2,255 unique individuals across 5 breeding seasons, spread across the known extant range of A. bishopi, to characterize the genetic diversity and demographics of populations, genetic relationships among populations, and patterns and spatial extents of gene flow, and to evaluate potential effects of management on A. bishopi’s resiliency. Population structure was strongly hierarchical, with individual breeding ponds (n = 38) acting as semi-connected subpopulations within five regional metapopulations (Mayhaw in Georgia; Oglesby, Eastbay, Garcon, and Escribano in Florida). Likewise, gene flow among populations was scale-dependent: negligible genetic differentiation, indicative of high gene flow, was observed only between pairs of ponds separated by \u3c 0.5 km, whereas between 0.5 and 5 km I observed steep genetic isolation by distance, and beyond 5 km genetic differentiation was generally high and only weakly related to distance. Across several breeding seasons, the effective number of breeders (Nb) per pond per year averaged 26 individuals (range 4 to 104). Larger-area, slower-drying ponds located closer to other occupied ponds exhibited larger Nb and greater genetic diversity. Based on genetically-reconstructed pedigrees, the ongoing headstarting program at Escribano successfully captured 97.9% of the estimated total number of alleles, but only 63% of the total number families, in each cohort. Based on these results, I recommend the following: 1) Given its genetic distinctiveness, Georgia populations merit elevated priority for protection and restoration. 2) Resiliency and redundancy (a la the species’ recovery plan) should be assessed at the spatial grain of individual breeding ponds. 3) Attempts to restore habitat connectivity should consider dispersal over distances \u3e 500 m to be relatively unlikely. 4) Finally, to the extent that headstarted individuals are used to augment existing or introduce new populations, managers should consider the potential risks of founder effects, and reduce these risks by creating genetically and demographically diverse headstart samples, for example by maximizing the diversity of egg/larva collections over time and space within ponds