Seasonality of marine prokaryotes using taxonomic and functional diversity approaches

Abstract

The oceans are ecosystems dominated by microbes, in which bacteria and archaea play key roles in biogeochemical cycling. In temperate oceans, seasonal changes in environmental conditions deeply influence the marine microbiome. In this thesis I analyse the seasonality of the marine microbiome of a coastal ocean site, using the long-term time series of the Blanes Bay Microbial Observatory (BBMO) to understand the seasonal changes through several molecular approaches. Using amplicons of the 16S rRNA gene, I evaluate the dynamics of the main bacterial groups in this coastal oligotrophic station during 11 years and test how similar the temporal niches of closely related taxa are, and what are the environmental parameters modulating their patterns of seasonality. I further explore how conserved the niche is at higher taxonomic levels. The community presented recurrent seasonality for 297 out of 6825 amplicon sequence variants (ASVs), which constituted almost half of the total relative abundance (47%). For certain genera, niche similarity decreased as nucleotide divergence in the 16S rRNA gene increased, a pattern compatible with the selection of similar taxa through environmental filtering. Additionally, I observed evidence of seasonal differentiation within various genera as seen by the distinct seasonal patterns of closely related taxa. I then switch the focus to the seasonal patterns of a specific functional group. Using the pufM gene as a marker gene for the aerobic anoxygenic phototrophic bacteria (AAPs) −a relevant photohete-rotrophic functional group in the marine microbial food web− I evaluated their long-term temporal dynamics through multivariate and co-occurrence analyses. Phylogroup K (Gammaproteobacteria) was the greatest contributor to community structure over all seasons, with phylogroups E and G (Alphaproteobacteria) being prevalent in spring. Diversity indices showed a clear seasonal trend, with maximum values in winter, which was inverse to that of AAP abundance. I afterwards extend these analyses to 21 biogeochemical relevant functions through 7 years of metagenomic data from the BBMO. Most genes presented a seasonal abundance trend: photoheterotrophic processes were enriched during spring, phosphorous-related genes were dominant during summer coinciding with phosphate limitation conditions, and assimilatory nitrate reductases correlated negatively with nitrate availability. Additionally, I identified the main taxa driving each function in each season and showed that, for some groups, the seasonality of bacterial families is different than that of their gene repertoire, so that different taxa within the same group present different functional specialization. Finally, I complement this descriptive view of the temporal changes with manipulation experiments to test how bottom-up and top-down processes exert selection on specific bacterial genomic species over the seasons. I experimentally modified the presence of predators, viruses, nutrient limitation (by diluting the samples with filtered seawater) and light availability in seawater from the BBMO in different seasons and assessed the growth of different organisms defined by metagenome assembled genomes (MAGs) under the manipulated conditions. Overall, I recovered 262 MAGs mainly from the Rhodobacterales, Flavobacteriales and Alteromonadales classes. Season and treatment greatly influenced community composition, with 26% of the MAGs indicative of the control treatments, 24% of both the control and predator-reduced treatments, 12.8% indicators of both the virus-reduced and the diluted treatments, and 7.3% of the predator-reduced treatment only. Flavobacteriaceae MAGs developed mostly in the predator-reduced treatment with distinct species in each season, whereas Alteromonadaceae and Sphingomonadaceae taxa developed preferably in the virus-reduced and diluted treatments indistinctively of season. Overall, this dissertation provides new insights into the seasonal patterns of key taxonomic and functional groups in the coastal surface ocean through the integration of information obtained using several molecular techniques and experimental approaches applied to a long-term time series