Microcystin concentrations can be predicted with phytoplankton biomass and watershed morphology

Anthropogenically derived increases in nutrient loads and climate change are considered the primary drivers of harmful cyanobacterial bloom expansion in freshwaters; however, watershed and within lake physical variables, as well as community interactions, mediate the response of the phytoplankton community. Thus, predicting when cyanobacteria are abundant and when their toxins have elevated concentrations is potentially dependent on a large suite of variables. We used a large spatial and temporal scale study to identify the best environmental predictors of microcystin concentrations in lakes and reservoirs in Ohio (Midwest USA). Additionally, we explored predictive models for a censored dataset that included phytoplankton community composition and abiotic characteristics of lakes and their watersheds. We found that lakes had elevated microcystin concentrations when phytoplankton biomass was high and when lakes had small watershed areas relative to lake surface areas, and that biovolume of potential toxin producers and phytoplankton biomass (chlorophyll a) both predicted microcystin concentrations. Seasonal patterns in microcystin concentrations suggested that a single annual sampling may miss microcystin maxima and, contrary to expectations, a late summer maximum in microcystin was not universal.