No abstract

The source water of a drinking water treatment plant prone to blooms, dominated by potential microcystin-producing cyanobacteria, was monitored for two seasons in 2007-2008. In the 2008 season, the median value for potential microcystin-producing cyanobacterial biovolume was 87% of the total phytoplankton biovolume in the untreated water of the plant. Depth profiles taken above the plant's intake identified three sampling days at high risk for the contamination of the plant's raw water with potentially toxic cyanobacteria. Chlorophyceae and Bacillariophyceae caused false positive values to be generated by the phycocyanin probe when cyanobacteria represented a small fraction of the total phytoplanktonic biovolume present. However, there was little interference with the phycocyanin probe readings by other algal species when potential microcystin-producing cyanobacteria dominated the phytoplankton of the plant's untreated water. A two-tiered method for source water monitoring, using in vivo phycocyanin fluorescence, is proposed based on (1) a significant relationship between in vivo phycocyanin fluorescence and cyanobacterial biovolume and (2) the calculated maximum potential microcystin concentration produced by dominant Microcystis sp. biovolume. This method monitors locally-generated threshold values for cyanobacterial biovolume and microcystin concentrations using in vivo phycocyanin fluorescence.

The applications of in vivo probes that can detect the fluorescence of cyanobacterial phycocyanin are emerging and widely used for cyanobacterial detection in source waters. The objectives of this project were to study the sources of interferences involved with the readings of five probes (three commercially available probes and two prototype probes) using laboratory cultures and field samples. To compare the direct readings of different probes, the probe readings were presented in the form of a biovolume equivalent of cyanobacteria. Inorganic turbidity and the presence of algal biomass interfered with probe readings. A correction factor was developed for the cyanobacteria probes using simultaneous chlorophyll a measurements. The field data demonstrate that the potential underestimation of cyanobacterial biomass that corresponds to alert levels is a major issue with the application of in vivo probes. These alert levels are used to trigger monitoring and management actions. This study shows that the correlation between a probe's reading and cell count is almost meaningless, and that the correlation to biovolume is a relevant option for management purposes. Results show that probe users should be fully aware of the sources of interferences when applying and interpreting the results. In addition, the authors offer a novel procedure that corrects for chlorophyll a interference.

Toxic cyanobacteria threaten the water quality of drinking water sources across the globe. Two such water bodies in Canada (a reservoir on the Yamaska River and a bay of Lake Champlain in Québec) were monitored using a YSI 6600 V2-4 (YSI, Yellow Springs, Ohio, USA) submersible multi-probe measuring in vivo phycocyanin (PC) and chlorophyll-a (Chl-a) fluorescence, pH, dissolved oxygen, conductivity, temperature, and turbidity in parallel. The linearity of the in vivo fluorescence PC and Chl-a probe measurements were validated in the laboratory with Microcystis aeruginosa (r(2) = 0.96 and r(2) = 0.82 respectively). Under environmental conditions, in vivo PC fluorescence was strongly correlated with extracted PC (r = 0.79) while in vivo Chl-a fluorescence had a weaker relationship with extracted Chl-a (r = 0.23). Multiple regression analysis revealed significant correlations between extracted Chl-a, extracted PC and cyanobacterial biovolume and in vivo fluorescence parameters measured by the sensors (i.e. turbidity and pH). This information will help water authorities select the in vivo parameters that are the most useful indicators for monitoring cyanobacteria. Despite highly toxic cyanobacterial bloom development 10 m from the drinking water treatment plant's (DWTP) intake on several sampling dates, low in vivo PC fluorescence, cyanobacterial biovolume, and microcystin concentrations were detected in the plant's untreated water. The reservoir's hydrodynamics appear to have prevented the transport of toxins and cells into the DWTP which would have deteriorated the water quality. The multi-probe readings and toxin analyses provided critical evidence that the DWTP's untreated water was unaffected by the toxic cyanobacterial blooms present in its source water.

No abstract