Species‐Specific Genetic Patterns in Sympatric Freshwater Turtles Challenge a Generalized Multi‐Species Conservation Approach

ABSTRACT Habitat loss reduces genetic connectivity, population size, and dispersal among wildlife populations, increasing extinction risk. In long‐lived species, spatial genetic structure reflects historic genetic connectivity and can inform restoration efforts to connect anthropogenically isolated populations. Comparative spatial genetics of sympatric species with varying life history traits can test how well spatial genetic structure is predicted by conservation genetics theory, and how it can be generalized among species to enable multi‐species landscape restoration. Here, we used microsatellite genotypes to compare spatial genetic structure and diversity in the snapping turtle ( Chelydra serpentina ), Blanding's turtle ( Emydoidea blandingii ), and spotted turtle ( Clemmys guttata ) sampled in areas of co‐occurrence across ~49,160 km 2 . We hypothesized that genetic structure would increase as species' dispersal ability decreased, that inferred barriers to gene flow would be geographically congruent, and that genetic diversity would increase with species' relative abundance, potential fecundity, and dispersal ability. Our results supported the first hypothesis; we identified four genetic clusters in the more vagile snapping turtle and Blanding's turtle, and 10 clusters in the less vagile spotted turtle. The second hypothesis was not supported, as inferred barriers to gene flow for each species were not geographically congruent. The third hypothesis was also not supported as we observed the lowest genetic diversity (heterozygosity and allelic richness) in the snapping turtle, which had the highest relative abundance, clutch size, and dispersal ability of the three species. We speculate that this result may reflect interspecific variation in reproductive variance, including different mating systems. Unexpectedly low genetic diversity in the snapping turtle suggests vulnerability to environmental change despite its apparent abundance. Overall, our results reinforce the risk of generalizing spatial genetic patterns among species and highlight the challenge of developing multi‐species conservation actions to maintain landscape connectivity.