| Abstract: | Evolutionary and ecological hypotheses of the freshwater mussel subfamily Ambleminae are intensely geographically biased—a consequence of the complete exclusion of Mesoamerican taxa in phylogenetic reconstructions of the clade. We set out to integrate a portion of the Mesoamerican freshwater mussel assemblage into existing hypotheses of amblemine classification and evolution by generating a molecular phylogeny that includes four previously unsampled Mesoamerican genera and nine species endemic to that region. Given the traditionally hypothesized affinity to Nearctic mussels and the understanding that classification should reflect common ancestry, we predicted that (a) Mesoamerican genera would be recovered as members of the recognized tribes of the Ambleminae, and (b) genera would be supported as monophyletic. The mutilocus phylogeny (COI + 28S + 16S) reported herein does not fully support either of those hypotheses. Neither Cyrtonaias nor Psorula were supported as monophyletic and we predict several other Mesoamerica genera are also non‐monophyletic. The reconstructed phylogeny recovered four independent lineages of Mesoamerican freshwater mussels and these clades are distributed across the phylogeny of the Ambleminae, including the tribe Quadrulini (Megalonaias), Lampsilini (two lineages: Cyrtonaias explicata/Sphenonaias microdon, and Pachynaias), and a previously unrecognized, exclusively Mesoamerican and Rio Grande clade consisting of the genera Psoronaias, Psorula and Popenaias. The latter clade possesses several morphological characteristics that distinguish it from its sister taxon, tribe Lampsilini, and we recognize this newly identified Mesoamerican clade as a fifth tribe of the Ambleminae attributable to the Popenaiadini Heard & Guckert, 1970. This revised classification more completely recognizes the suprageneric diversity of the Ambleminae. |
