LAUSR

The toxic cyanobacterium Microcystis is one of the most pervasive harmful algal bloom (HAB) genera and naturally occurs in large colonies known to harbor diverse heterotrophic bacterial assemblages. While colony-associated microbiomes may influence Microcystis blooms, there remains a limited understanding of the structure and functional potential of these communities and how they may be shaped by changing environmental conditions. To address this gap, we compared the dynamics of Microcystis-attached (MCA), free-living (FL), and whole water (W) microbiomes during Microcystis blooms using next-generation amplicon sequencing (16S rRNA), a predictive metagenome software, and other bioinformatic approaches. Microbiomes were monitored through high resolution spatial-temporal surveys across two North American lakes, Lake Erie (LE) and Lake Agawam (LA; Long Island, NY, United States) in 2017, providing the largest dataset of these fractions to date. Sequencing of 126 samples generated 7,922,628 sequences that clustered into 7,447 amplicon sequence variants (ASVs) with 100% sequence identity. Across lakes, the MCA microbiomes were significantly different than the FL and W fractions being significantly enriched in Gemmatimonadetes, Burkholderiaceae, Rhizobiales, and Cytophagales and depleted of Actinobacteria. Further, although MCA communities harbored > 900 unique ASVs, they were significantly less diverse than the other fractions with diversity inversely related to bloom intensity, suggesting increased selection pressure on microbial communities as blooms intensified. Despite taxonomic differences between lakes, predicted metagenomes revealed conserved functional potential among MCA microbiomes. MCA communities were significantly enriched in pathways involved in N and P cycling and microcystin-degradation. Taxa potentially capable of N2-fixation were significantly enriched (p < 0.05) and up to four-fold more abundant within the MCA faction relative to other fractions, potentially aiding in the proliferation of Microcystis blooms during low N conditions. The MCA predicted metagenomes were conserved over 8 months of seasonal changes in temperature and N availability despite strong temporal succession in microbiome composition. Collectively, these findings indicate that Microcystis colonies harbor a statistically distinct microbiome with a conserved functional potential that may help facilitate bloom persistence under environmentally unfavorable conditions.