Extracellular microcystin prediction based on toxigenic Microcystis detection in a eutrophic lake.

Existing models for predicting microcystin concentration in water body generally use chlorophyll or cyanobacteria concentration as input variables, although microcystins only originate from toxigenic strains of a few species. Moreover, the nonconcurrency between harmful algal growth and toxin release has yet to be quantified. Therefore, this study explored a new prediction method that considers these toxin production mechanisms for the eutrophic Yangcheng Lake, a large-scale drinking water source in China. The Lake was monitored weekly at six sampling sites from July to October in 2012, including the detection of toxigenic Microcystis (expressed as mcyA copy number) by qPCR. Compared with chlorophyll a, cyanobacteria, and total Microcystis abundance, toxigenic Microcystis concentration was more significant in predicting extracellular microcystin. Site-specific nonlinear regression models that link mcyA to microcystins were established. Parameters for toxin release delay (i.e., one or two weeks) were embedded in these models. Further analysis ascribed the different release timescale to NH3-N:TN and TN:TP ratios of approximately 0.015 and 9.2, respectively, which may decrease the delay in microcystin release. Model applications in determining mcyA monitoring frequency and its warning thresholds were discussed.