Changes in biodiversity are predicted to have a major impact on ecosystem processes.  A fundamental challenge for ecologists is to determine the influence of species on ecosystem processes prior to declines or losses, and to determine whether there are species that may compensate for the loss of other species. The southern Appalachian Mountains are global hotspot for stream plethodontids.  Plethodontids are the numerically dominant vertebrate predators of high-elevation first order streams, and as such are hypothesized to be influential in the capture and retention of nutrients and the flow of energy to higher trophic levels.  Our objectives were to determine the potential consequences of a reduction of stream salamander diversity to nutrient capture and retention within first and second order streams at the Coweeta Hydrological Laboratory.  To meet this objective, we used ecological stoichiometry to estimate the amount of nutrients captured and exported by a stream salamander community, generated predictive models of species loss under climate change scenarios, and used field removal experiments to determine whether predicted species losses alter nutrient capture and retention or whether there is compensation by other salamander species or stream fauna.  We found stream plethodontids are significant nutrient reservoirs when compared to other stream taxa and are influential to stream nutrient cycling; however, models project a decline in suitable habitat associated with climate change scenarios for many species within the Coweeta basin.  Further, our results suggest that while subordinate plethodontid species respond positively to the loss of the dominant species, there is still a net loss of salamander biomass produced within streams.  These results suggest that the potential loss of some salamander species from the region that includes Coweeta could have significant implications for stream function including the capture, retention and ultimately export of nutrients.