Spatiotemporal molecular analysis of cyanobacteria blooms reveals Microcystis--Aphanizomenon interactions.PLoS One. 2013; 8(9):e74933.Plos
Spatial and temporal variability in cyanobacterial community composition (CCC) within and between eutrophic lakes is not well-described using culture independent molecular methods. We analyzed CCC across twelve locations in four eutrophic lakes and within-lake locations in the Yahara Watershed, WI, on a weekly basis, for 5 months. Taxa were discriminated by length of MspI-digested cpcB/A intergenic spacer gene sequences and identified by comparison to a PCR-based clone library. CCC across all stations was spatially segregated by depth of sampling locations (ANOSIM R = 0.23, p < 0.001). Accordingly, CCC was correlated with thermal stratification, nitrate and soluble reactive phosphorus (SRP, R = 0.2-0.3). Spatial variability in CCC and temporal trends in taxa abundances were rarely correlative between sampling locations in the same lake indicating significant within lake spatiotemporal heterogeneity. Across all stations, a total of 37 bloom events were observed based on distinct increases in phycocyanin. Out of 97 taxa, a single Microcystis, and two different Aphanizomenon taxa were the dominant cyanobacteria detected during bloom events. The Microcystis and Aphanizomenon taxa rarely bloomed together and were significantly anti-correlated with each other at 9 of 12 stations with Pearson R values of -0.6 to -0.9 (p < 0.001). Of all environmental variables measured, nutrients, especially nitrate were significantly greater during periods of Aphanizomenon dominance while the nitrate+nitrite:SRP ratio was lower. This study shows significant spatial variability in CCC within and between lakes structured by depth of the sampling location. Furthermore, our study reveals specific genotypes involved in bloom formation. More in-depth characterization of these genotypes should lead to a better understanding of factors promoting bloom events in these lakes and more reliable bloom prediction models.