Freshwater mollusks are disproportionately represented among recent extinctions. In North America, estimates of historical extinctions of freshwater bivalves (Unionidae) range from 21-37 species, or about 10% of the pre-settlement fauna. Many more extinctions are impending. Reasons for this decline are complex and mainly involve anthropogenic habitat destruction and fragmentation. The dependence of unionids on particular species of fish as hosts for parasitic larva development is an unusual example of âhabitatâ specialization that may contribute to vulnerability and that complicates efforts to define diversity and conservation priorities. Recent studies of host specificity indicate that some morphologically defined unionid species consist of multiple species or at least host races, which are differentiated by their adaptation to sympatric host fish species and populations. Such differentiation is expected to be most pronounced in mussels that utilize geographically fragmented and genetically diverse host populations. In the unionid morphospecies Cyprogenia aberti, for example, mussel populations in different drainages are able to utilize local populations of several species of darters (Etheostoma, Percina) but generally not populations or species from other drainages. Both host specificity and genetic evidence indicate that C. aberti consists of 3-4 species. Reliance on particular host species or populations probably increases extinction risks for unionids because they share vulnerabilities of the host species as well as their own. However, the hypothesis that host abundance, host genetics, or particular aspects of the host-parasite relationship are responsible for unionid declines is generally untested.