Ecological correlates and phylogenetic signal of host use in North American unionid mussels.

Mussels in the order Unionoidacomprise ∼75% of the world's freshwater bivalve species and are free-living apart from a brief larval stage that parasitizes fish. We investigatedthe relationships among species of North American unionid mussels and their known host fishesfrom a macroevolutionary perspective to test whether and how ecological and evolutionary factors correlate with patterns of host use. A subset of 69 mussel species was chosen based on data availability regarding their fish host repertoires, phylogenetic relationships, and ecology. Despite the brevity of their parasitic life stages, the mussels conformed to the right-skewed distribution of host specificity typical of parasitic taxa, in which most species are specialists and a few are generalists. Phylogenetic least squares regression models identified affinity for low-gradient and riffle habitats, and colonization of post-glacial watersheds as the best predictors for the number of fish host species per mussel. However, the second-best model identified citation number as a predictor of the number of hosts, implying that many mussel-host interactions still remain to be identified. A Multiple Regression Mantel test was performed to identify factors associated with the proportion of hosts shared between pairs of mussel species. Range overlap, citations, genetic distance, and similarity in host infection strategy were significantly correlated with the proportion of hosts shared, yet total variation as explained by the best model was low (R2 = 0.14). There was evidence of a topological association between mussels and their hosts (P= 0.001) and a significant phylogenetic signal of host specificity (λ= 0.81,P= 0.003), indicating closely related mussels that overlap in range are more likely to be competing for hosts. Our results provide an initial macroevolutionary framework for studying the evolution of host infection strategies in these mussels but also highlights gaps still remaining in our fundamental ecological knowledge of this endangered clade.