Population genomics reveal low differentiation and complex demographic histories in a highly fragmented and endangered freshwater mussel

Abstract

Abstract Freshwater mussels are an important element of freshwater biodiversity and provide essential ecosystem services. However, mussels are among the most imperilled groups of organisms in the world. Although research has increased in recent years, information about range‐wide genetic diversity and historical demography of most species is lacking. One such species is Cumberlandian combshell Epioblasma brevidens, which is listed as endangered under the US Endangered Species Act. Genetic diversity of E. brevidens was analysed using a high‐resolution RADseq approach and included the previously overlooked Bear Creek population. Hypotheses were tested about population decline, comparative genetic diversity and population structure with model‐based approaches enabled by a genome‐scale dataset. Estimates of genetic differentiation among populations of E. brevidens were lower than past analyses, suggesting higher historical population connectivity than previously known. Demographic analyses indicate relatively recent splits among E. brevidens populations in the late Pleistocene to early Holocene, with clear founder effects in two populations. The Clinch River population has the highest genetic diversity and effective population size, despite demographic analyses revealing decline of this population since the Pleistocene. Analyses of both population structure and migration show evidence of past gene flow, but all populations are currently isolated by artificial barriers. Analyses indicate that populations began to decline before industrialization, but fragmentation and population extirpation has been exacerbated by modern habitat destruction. Relatively high genetic diversity in the Bear Creek population indicates that water quality improvements in the last 20 years have had a positive impact on population viability, offering promise for targeted management actions. In contrast, the Big South Fork population that has been presumed stable showed low genetic diversity and effective population size. Furthermore, genetic structure among sampled populations indicates that reintroduction efforts should use broodstock from as close to the reintroduction site as possible.