Biodiversity and dynamics of cyanobacterial communities during blooms in temperate lake (Harsha Lake, Ohio, USA)

Cyanobacterial blooms are intensifying global ecological hazards. The fine structure and dynamics of bloom community are critical to understanding bloom development but little understood. Here, the questions whether dominant bloomers have high diversity and whether dominant OTUs (operational taxonomical units) compete with one another were addressed. 16S rENA gene amplicons from an annual bloom at five locations in Harsha Lake (Ohio, USA) showed cyanobacteria were the dominant phylum, and co-existing major bacterial phyla included Proteobacteria, Bacteroidetes, Actinoacteria, and Verrucomicrobia. On the genus level, the initial dominance by Dolichospenman in June yielded to Planktothrix in July, which were replaced by Microcystis and Cylindrospermopsis in August throughout the bloom. Based on the number of verified unique OTUs (a withingenus biodiversity metric), dominant genera tended to have high within-genus diversity. For example, Dolichospertmum had 57 unique OTUs, Planktothrix had 36, Microcystis had 12, and Cylindrospermopsis had 4 unique OTUs. Interestingly, these different OTUs showed different dynamics and association with other OTUs. First, no between-OTU competitions were observed during the bloom cycle, and dominant OTUs were abundant throughout the bloom. Such biodiversity of OTUs and their dynamics were verified in Microcystis aeruginosa with two microcystin synthetase genes (mcyA and mcyG): the relative abundance of both genes varied during the bloom based on quantitative PCR. Two Dolichospermum circinale OTUs and one P. rubescens OM were most abundant and persistently present throughout the entire bloom. Second, these OTUs differed in the OTUs they were associated with. Third, these OTUs tended to have different levels of association with the environmental factors, even they belonged to the same genera. These findings suggest the structure and dynamics of a cyanobacterial bloom community is complex, with only few OTUs dominating the bloom. Thus, high-resolution molecular characterization will be necessary to understand bloom development