50.5 Thursday, Jan. 5 Genetic and demographic patterns of populations of Ambystoma opacum NUNZIATA, Schyler O*; LANCE, Stacey L; SCOTT, David E; University of Georgia, Savannah River Ecology Lab; University of Georgia, Savannah River Ecology Lab; University of Georgia, Savannah River Ecology Lab email@example.com
Population genetic studies often focus on conserving genetic diversity, which is affected by gene flow and effective population size (Ne). Although the importance of Ne is well accepted in maintaining genetic diversity, there is still a lack of understanding on how to measure Ne and factors impacting it. In this study, we used both demographic and genetic methods to estimate Ne in two populations of marbled salamanders, Ambystoma opacum. For genetic analyses we subsampled two isolated wetland populations, Rainbow Bay (RB) and Ginger’s Bay (GB), on the Savannah River Site (SC) over a 20-yr period. The numbers of breeding adults (N) and their offspring are known for 30 yrs for RB and 15 yrs for GB. A. opacum colonized the RB wetland in 1980 and the population has steadily expanded; the GB population has remained relatively large and stable over this same time period. To calculate Ne with demographic models we used field estimates of longevity, sex ratio, generation time, and breeding success. For genetic estimates we screened samples across 12 microsatellite loci to determine Ne and other classical population genetic parameters (Ho, allelic richness). Genetic methods of determining Ne gave markedly lower estimates than demographic methods, with Ne/N ranging from 0.01-0.11. Both the RB and GB populations incur complete reproductive failure in some years due to early pond drying; failure is more frequent at RB. Zero recruitment years may cause a bias in both demographic and genetic estimates of Ne. In particular, the demographic estimate of Ne does not account for high juvenile mortality, which may partly explain the departure from genetic estimates of Ne. Our results provide important information on what factors should be incorporated into models for estimating Ne.