Genetic and Morphological Analyses of Three Freshwater Mussel Species in Isolated River Drainages Across Appalachia

The Upper Tennessee River drainage of Virginia and Tennessee, Green River drainage of Kentucky, and Cumberland River drainage of Kentucky and Tennessee are known for their freshwater mussel species diversity. These river systems harbor dense populations of freshwater mussels that have significant impacts on surrounding ecosystems; however, due to their sedentary lifestyles, freshwater mussels are particularly susceptible to many biotic and abiotic stressors. Managers strive to preserve the fragile ecosystems that include freshwater mussels and, hence, study the life-history strategies of endangered and common freshwater mussel species. The goals of this project were to inform management decision-making based on whether Leaunio ortmanni is endemic to the Green River drainage and a species distinct from Leaunio vanuxemensis based on molecular identification, estimation of genetic diversity, and morphometric analysis, and to screen for cryptic biodiversity of populations of the Cambarunio iris species complex in the Upper Tennessee, Cumberland, and Green River drainages. I used the mitochondrial DNA (mtDNA) gene from the first subunit of the NADH dehydrogenase (ND1) as a genetic marker for species-level assessment of L. ortmanni and L. vanuxemensis from the Green and Cumberland River drainages. Additional mtDNA sequences of L. ortmanni and L. vanuxemensis were added to increase sample sizes and coverage of historical distribution. A Bayesian phylogenetic analysis of mtDNA sequences did not result in monophyletic lineages for either species based on the ND1 marker. Haplotype networks of mtDNA sequences demonstrated that haplotype sharing is occurring between the two focal taxa, and also among additional taxa, all of which previously belonged to the genus Villosa. A total of eight nuclear DNA microsatellites were successfully genotyped for the two focal taxa. The nuclear DNA microsatellites showed a strong phylogeographic signal between L. ortmanni of the Green River drainage and L. vanuxemensis of the Cumberland River drainage. An assignment-test based analysis in program STRUCTURE and a phylogenetic tree constructed using Nei's D genetic distance indicated well-differentiated populations across the two drainages. Additionally, the nuclear DNA microsatellite analysis showed a recent loss of genetic diversity across all populations, including when populations were pooled together at the sub-basin level. Further delineation of the focal taxa was assessed using geometric morphometrics and decision tree and random forest analyses. Decision tree and random forest analyses identified periostracum color, nacre color, overall shape, and sex to be distinguishing factors for field identification of L. ortmanni and L. vanuxemensis. Geometric morphometrics comparing species, shape, and drainage showed clear differentiation in shell shape between L. ortmanni and L. vanuxemensis. This study was able to delineate these two taxa, showing that L. ortmanni and L. vanuxemensis are separate species, and that L. ortmanni warrants listing under the Endangered Species Act. Management actions should focus on broodstock collection and propagation strategies to increase genetic diversity within established populations. Additionally, propagation and augmentation should look to reintroduce populations of L. ortmanni into its historical geographic range in unoccupied sections of the mainstem Green River. In addition, I assessed genetic diversity and differentiation again using ND1 and eight nuclear DNA microsatellite loci, and morphological differences among different shell forms of C. iris in the Upper Tennessee, Cumberland, and Green river basins and of the sister species C. taeniatus in the Cumberland, and Green river basins. Additional mitochondrial DNA sequences of C. iris and C. taeniatus were added to increase sample sizes and coverage of historical distribution. Mitochondrial DNA analysis demonstrated haplotype sharing between taxa, but with many unique haplotypes occurring in the upper Tennessee River basin. Nuclear DNA microsatellite loci revealed low levels of genetic diversity within populations of C. iris within the Upper Tennessee River basin, but high levels of divergence from C. iris and C. taeniatus of the Green and Cumberland River basins. The nuclear DNA analysis showed high admixture within and among sampled populations of C. iris throughout the Upper Tennessee River Basin with minimal geographic structuring among sub-basins. Further, phenotypic diversity was assessed using geometric morphometrics and decision tree and random forest analysis. Decision-tree and random forest analysis identified maximum height from the umbo to the ventral margin, periostracum color, shell width, and ray coverage to be defining characteristics for field identification of the focal taxa. Geometric morphometrics showed high overlap of shell shape for the focal taxa regardless of locality. While cryptic biodiversity was not detected in the Upper Tennessee River Basin, on a larger geographic level, cryptic biodiversity was detected using the combination of the mtDNA, nuclear DNA, and morphological data, which showed that C. taeniatus and C. iris from the Green River drainage were distinct from populations of C. iris in the upper Tennessee River basin.