Mussel assemblages upstream from three Kansas reservoirs.

Abstract

Reservoir construction by damming of rivers has contributed to dramatic declines in species richness and abundance of freshwater mussel assemblages throughout North America. The effects of reservoirs on mussel assemblages within and downstream from the reservoir pool are relatively well known, but few studies have examined effects on upstream mussel assemblages. During summers 1999 and 2000, I surveyed 40 sites in the Marais des Cygnes (n=15), Fall (n=13), and Elk (n=12) rivers in eastern Kansas, upstream from three reservoirs, to examine effects of reservoir inundation on upstream mussel assemblages. I predicted that the present mussel assemblage would be composed of fewer species than the historic assemblage, that the percent of species missing from the historic assemblage would increase nearer the reservoirs, that mussel species richness and abundance would decrease nearer the reservoirs, and that substrate embeddedness and silt in the substrate would increase downstream. I recorded present and historically-occurring species plus 10 habitat variables at each site, then used Student's t-test, linear regression, and canonical correspondence analysis to examine decline in species richness in each river, to elucidate trends in species richness, mussel abundance, and habitat values in relation to frequency of reservoir inundation, and to model environmental correlates of assemblage structure. I collected 1367 live mussels of 18 species, and 29 species as weathered valves. In all three rivers, significantly fewer species were present II alive than were present as weathered valves. Live species richness and abundance decreased nearer the reservoir in the three rivers, whereas historic species richness was not significantly related to flood frequency in any river. Percent of species missing from the historic assemblage increased nearer the reservoirs, but this trend was significant only in the Marais des Cygnes. Substrate embeddedness and percent of silt in the substrate were not related to flood frequency in any river. Canonical correspondence indicated that Marais des Cygnes sites had a higher percentage of fine substrates than Fall and Elk river sites, and that this river's mussel assemblage was different from those of the Fall and Elk. Siltation caused by reservoir inundation might be an episodic event that does impact species richness and abundance nearer the reservoir, but that is difficult to detect except during inundation events.