Invasive zebra mussels (Dreissena polymorpha) have significantly modified freshwater systems throughout North America. One unanticipated consequence of their introduction has been an increase in the toxic bloom-forming cyanobacterium, Microcystis aeruginosa, in low-nutrient lakes. A recent enclosure experiment demonstrated that mussels promote Microcystis under low nutrient conditions, yet can inhibit it at elevated nutrients. To begin to address why, we performed laboratory feeding experiments with mussels to test the hypothesis that intraspecific variation in coexisting, colonial Microcystis strains leads to differences in mussel feeding selectivity. Zebra mussels were collected and Microcystis strains were isolated from oligotrophic Gull Lake, MI, the site of the enclosure experiment. Mussels were presented with one of five different Microcystis strains a in two-species mixture with Ankistrodesmus falcatus, a high-quality food, following 24 hr acclimation on each feeding suspension. Controls for each feeding suspension lacked mussels and ruled out changes in algal biomass independent of mussel feeding over the 1 hr experiment. Mussel feeding selectivity was quantified as the ratio of the filtering rates on Microcystis and Ankistrodesmus, which were determined from algal cell counts in samples collected before and after mussel feeding.

Results/Conclusions

Biomass ratios of the two phytoplankton species were similar across treatments, with chlorophyll a concentration and Microcystis biomass within the ranges observed in Gull Lake. We found large differences in mussel selectivity for the coexisting Microcystis strains. Mean selectivity across Microcystis strains ranged from near zero (i.e., no feeding on Microcystis) to near one (equal feeding on the two species), evidence of near-maximal variation in grazing vulnerability across strains of Microcystis from the same population. Therefore, we demonstrate one behavioral mechanism by which mussels could have opposite effects on Microcystis biomass in a field experiment, given the same initial phytoplankton assemblage.