Freshwater mollusk shell morphology exhibits clinal variation along a stream continuum that has been termed the Law of Stream Distribution. We analyzed phylogenetic relationships and morphological similarity of two freshwater mussels (Bivalvia: Unionidae), Obovaria jacksoniana and Villosa arkansasensis, throughout their ranges. The objectives were to investigate phylogenetic structure and evolutionary divergence of O. jacksoniana and V. arkansasensis and morphological similarity between the two species. Our analyses were the first explicit tests of phenotypic plasticity in shell morphologies using a combination of genetics and morphometrics. We conducted phylogenetic analyses of mitochondrial DNA (1416 bp; two genes) and morphometric analyses for 135 individuals of O. jacksoniana and V. arkansasensis from 12 streams. We examined correlations among genetic, morphological, and spatial distances using Mantel tests. Molecular phylogenetic analyses revealed a monophyletic relationship between O. jacksoniana and V. arkansasensis. Within this O. jacksoniana/V. arkansasensis complex, five distinct clades corresponding to drainage patterns showed high genetic divergence. Morphometric analysis revealed relative differences in shell morphologies between the two currently recognized species. We conclude that morphological differences between the two species are caused by ecophenotypic plasticity. A series of Mantel tests showed regional and local genetic isolation by distance. We observed clear positive correlations between morphological and geographic distances within a single drainage. We did not observe correlations between genetic and morphological distances. Phylogenetic analyses suggest O. jacksoniana and V. arkansasensis are synonomous and most closely related to a clade composed of O. retusa, O. subrotunda, and O. unicolor. Therefore, the synonomous O. jacksoniana and V. arkansasensis should be recognized as Obovaria arkansasensis (Lea 1862) n. comb. Phylogenetic analyses also showed relative genetic isolation among drainages, suggesting no current gene flow. Further investigation of in-progress speciation and/or cryptic species within O. arkansasensis is warranted followed by appropriate revision of conservation management designations.

Many authors have described a pattern of morphological variation in freshwater bivalves where shells taken from lentic and lotic environments, or headwaters and main stem reaches, appear to exhibit phenotypic gradients in size and shape. For example, mussels taken from headwater reaches tend to possess smooth, less inflated shells compared to the more obese, sculptured individuals downstream. Others observed similar relationships in certain freshwater gastropods, but this variation has not been quantified nor its existence explained in an ecological or evolutionary context. Geometric morphometrics indicated freshwater snails shells from the pleurocerid genus Lithasia from the Duck River, Tennessee, USA, show phenotypic gradients similar to those in freshwater mussels. Shells from upstream areas were narrow and less sculptured on the posterior portions of their body whorls, while downstream shells were more inflated and possessed significantly more sculpture. This phenotypic variation may reduce predation or damage due to dislodging. The nature of the observed plasticity suggests an unidirectional environment similar to that proposed by the river continuum concept.

The Pleurobemini (Bivalvia: Unionida) represent approximately one-third of freshwater mussel diversity in North America. Species identification within this group is challenging due to morphological convergence and phenotypic plasticity. Accurate species identification, including characterisation of currently unrecognised taxa, is required to develop effective conservation strategies because many species in the group are imperiled. We examined 575 cox1 sequences from 110 currently recognised species (including 13 Fusconaia and 21 Pleurobema species) to understand phylogenetic relationships among pleurobemine species (mainly Fusconaia and Pleurobema) and to delineate species boundaries. The results of phylogenetic analyses showed no geographic structure within widespread species and illustrated a close relationship between Elliptio lanceolata and Parvaspina collina. Constraint tests supported monophyly of the genera Fusconaia and Pleurobema, including the subgenus P. (Sintoxia). Furthermore, results revealed multiple conspecifics, including P. hanleyianum and P. troschelianum, P. chattanoogaense and P. decisum, P. clava and P. oviforme, P. rubrum and P. sintoxia, F. askewi and F. lananensis, and F. cerina and F. flava. Species delimitation analyses identified three currently unrecognised taxa (two in Fusconaia and one in Pleurobema). Further investigation using additional genetic markers and other lines of evidence (e.g. morphology, life history, ecology) are necessary before any taxonomic changes are formalised.