Reintroduction of a declining amphibian: Determining an ecologically feasible approach for the foothill yellow-legged frog (Rana boylii) through analysis of decline factors, genetic structure, and habitat associations

Species reintroductions provide a model for integrating practical and theoretical aspects of conservation biology. However, we currently lack a clear understanding of the factors that determine the outcome of reintroduction programs for many taxa. Combining pre-reintroduction research on ecology, genetics, and causative factors of declines with experimental reintroductions and rigorous monitoring, offers an approach that should increase reintroduction success rates. Amphibians present challenges as reintroduction subjects due to characteristics such as biphasic life cycles, low mobility, and patchy distributions. This study focused on a stream-dwelling, declining amphibian in California and Oregon (the foothill yellow-legged frog, Rana boylii ) and included three components: (1) determining primary causes of decline, (2) describing range-wide genetic variation, and (3) quantifying habitat associations. For decline factors, the analysis approach was to spatially relate the current status of R. boylii (present or absent) at historic localities to: geographic characteristics, land uses, wind-borne toxins, climatic variables, and proximity and size of dams. Climatic variables showed strong influence in multivariate models. There was also evidence for interactions, especially that negative effects of dams appeared to be exacerbated in areas with low precipitation. For genetic analyses, 1525 total base pairs from sequences of two mtDNA fragments (Cytochrome B and ND2) for 77 individuals from 34 localities were used. Phylogenetic analyses recovered several well-supported, geographically congruent clades within R. boylii. Genetic variation was low among populations in the largest, most inclusive clade, but individuals from several localities showed substantial divergence. Hydrologic regions, which represent likely dispersal corridors for R. boylii , show promise in explaining patterns of genetic variation. The habitat associations component focused on microhabitat scale oviposition site selection coupled with larger scale evaluations of occurrence and relative abundance at breeding areas. Oviposition microhabitat characteristics such as water depth, water velocity, and stream substrate showed narrow ranges among study localities. These results suggest that habitat selection results in population stability for R. boylii even within the substantial temporal and spatial variability of stream environments. I discuss the application of these results to potential reintroductions of R. boylii and propose a conceptual model for integrating this and other information into reintroduction programs.