Cyanobacteria blooms have recurred in western Lake Erie. Recent blooms have been dominated by Microcystis aeruginosa, a non-nitrogen fixing species that produces hepatotoxins. Bloom dominance by non-nitrogen fixing species is significant because it has happened after declines in external sources of phosphorus loading led to significant declines in cyanobacteria blooms dominated by nitrogen-fixing species in the 1970's. Recent blooms coincide with the introduction of non-indigenous mussels, Dreissena polymorpha and D. bugensis, to Lake Erie. In ecologically similar lake systems invaded by dreissenids, bloom formations dominated by M. aeruginosa have also occurred. My examinations of historical and present trends in cyanophyte abundance and species composition suggest an interconnection between high nitrate concentration, selective filtration, and excretions of ammonia and phosphate by dreissenids, and extended periods of summer temperatures in the range of 23 to 26 degrees C.
Further, concern about M. aeruginosa blooms in western Lake Erie is founded in the production of cyclic heptapeptide hepatotoxins (microcystins) by this species. We examined the fate of microcystin produced by M. aeruginosa in western Lake Erie by testing whether microcystin toxin is passed through and/or accumulated by three non-indigenous, interconnected species that have become established in Lake Erie. Dreissenid mussels, amphipods (Echinogammarus ischnus), and round gobies (Neogobius melanostomus), served as a model trophic cycle in which to follow the fate of microcystin. Our surveys of Lake Erie water showed that microcystin was present in water and cell material at low levels during Microcystis blooms in 1995, 1996 and 1998. Mussels, amphipods and several fish species collected during blooms accumulated the toxin in guts and/or liver tissue. We determined that a novel mechanism making microcystin available to benthic organisms such as E. ischnus is the deposition of microcystin-laden pseudofeces + feces by Dreissena species.
Further indirect effects caused by the introduction of dreissenids may be negative feedback effects associated with toxic Microcystis blooms that they play a part in encouraging. To determine the potential for feedback effects on mussels, I developed a dynamic programming model of D. polymorpha filtration behavior to determine if fitness is influenced by blooms of M. aeruginosa. In the model, bioenergetic parameters of mussels were influenced by assumed ecological conditions and probabilities. Model outcomes suggest small mussels experiencing bloom conditions will have reduced fitness. An additional implication of this outcome is that size-frequency of populations may be shifted toward dominance by larger sizes having potentially greater algal consumption and pseudofeces production rates. Thus, population-level feedback effects on dreissenids induced by toxic Microcystis blooms may further aggravate their ecological impacts.