No abstract

The increasing incidence of toxic cyanobacteria blooms worldwide has created a need for practical and efficient monitoring in order to protect public health. We developed a monitoring and alert framework based on World Health Organization (WHO) recommendations and applied it on Lake Champlain during the summers of 2002-2004. The protocol began with collection of net samples of phytoplankton in order to maximize the chance of finding potential toxin-producing cyanobacteria. Samples were collected lake-wide in partnership with ongoing monitoring efforts, but because open water sample sites did not capture conditions along the shoreline, we added near-shore and shoreline stations in problem areas. Samples were examined qualitatively until potential toxin-producing taxa were found. Then quantitative analyses began, using a rapid screening method to estimate cell density based on colony size. A final cell density of 4000 cells/mL triggered toxin analyses. Primary analysis was for microcystins using ELISA methods. Cell densities, locations of colonies, and toxin concentrations were reported weekly to public health officials. We found that screening for potential toxin-producing cyanobacteria and then measuring toxin concentrations when cell densities reached critical levels worked well to identify problem locations. Although the WHO recommends using chlorophyll a concentration, it was not a good indicator of problem densities of potential toxin-producing cyanobacteria. Our cell density screening method missed no developing blooms but produced less precise density estimates at high cell counts. Overall, our framework appears to provide an efficient and effective method for monitoring cyanotoxin risks.

FlowCAM combines flow cytometry and imaging to rapidly enumerate, classify and measure particles. The instrument potentially increases processing speed of phytoplankton samples. FlowCAM, however, requires extensive comparison to microscopy before incorporation into monitoring and research. Past studies have compared FlowCAM and microscopy results for mostly marine rather than freshwater phytoplankton communities. We compared phytoplankton biovolume, density and taxonomic classifications between FlowCAM and microscopy for 113 samples from Lake Champlain, USA—a large freshwater system with diverse phytoplankton. Total biovolume estimates from FlowCAM were higher than microscope biovolumes due to higher individual particle biovolumes. Biovolume relationships, however, were closely correlated between the two methods. Shape-specific biovolumes from FlowCAM images slightly improved estimates compared to area-based biovolumes. Diatoms and filamentous cyanobacteria showed the strongest relationships between FlowCAM and microscope biovolumes. Microscope natural unit counts were generally higher than FlowCAM counts. Genus richness was weakly related between FlowCAM and microscopy, demonstrating a potential tradeoff between finer taxonomic resolutions with a microscope versus the higher number of particles processed with FlowCAM. Both methods produced reproducible biovolumes with replicate samples. We conclude that microscopy is more reliable when fine taxonomic resolution is needed and FlowCAM is suitable for rapid processing of major phytoplankton groups.