Scienze biologiche

Heterotrophic nanoflagellate (HNF) grazing is one of the major source of prokaryotic mortality in marine ecosystems, acting as a strong selection pressure on communities. Protozoans may thus affect prokaryotic abundance and alter the diversity and the taxonomic composition of the prey community, as individual prokaryotes can develop distinct grazing-resistant mechanisms. Moreover, the microbial loop is well known to regulate carbon fluxes in surface marine environments but few studies have quantified the impact of HNF predation on prokaryotes in the dark ocean. The present work was aimed to: (1) quantify the impact of HNF predation on the deep prokaryotes biomass; (2) investigate if and how prey diversity varies in response to different predation pressure; (3) define taxonomic community composition in studied areas and identify most affected prokaryotic phylotypes by HNF grazing (4) evaluate the effects of small HNF (<3 µm), which are known to dominate nano-sized compartment and represent the main bacterivores in aquatic ecosystems, being an important link between bacteria and larger protists; (5) evidence differences in community sensitivity to grazing between surface and mesopelagic ecosystems (6) identify the main environmental drivers shaping microbial community diversity. Predation experiments were performed with surface and mesopelagic water samples collected from the Southern Adriatic and Northern Ionian basins. An additional predation experiment was set up in the North-eastern Adriatic Sea. We coupled the traditional ‘dilution method’ with high-throughput molecular analysis (ARISA and Ion Torrent/454 sequencing) to provide a quantitatively and qualitatively evaluation of the grazing process occurring in marine microbial communities. The present work is structured by four manuscripts in preparation and one manuscript already submitted. 1. Heterotrophic nanoflagellate grazing on picoplankton in deep waters (manuscript in preparation) 2. Effects of heterotrophic flagellate predation on bacterial community diversity (manuscript in preparation) 3. HNF grazing impact on taxonomic composition of marine prokaryotic community (manuscript in preparation) 4. Environmental drivers structuring surface and deep bacterial communities in Adriatic and Ionian Seas (manuscript in preparation) 5. Biodiversity changes of bacterial community under predation pressure analyzed by 16S rRNA pyrosequencing (manuscript submitted) My PhD research led to important progresses in the comprehension of microbial dynamics regulating carbon cycles and bacterial diversity in the Adriatic and Ionian basins. Prokaryotic abundance and biomass were one order of magnitude higher in the photic than in the aphotic layers of Southern Adriatic and Ionian Seas (surface biomass 1.68 ± 1.76 µC L-1, deep biomass 9.00 ± 2.11 µC L-1). The Northern Adriatic community presented the highest biomass value (57.46 µC L-1), according to its richer trophic status. All in situ communities displayed the same evenness, being dominated by rare phylotypes. Rare taxa were confirmed to represent the major contributors of microbial communities, with only a few phylotypes dominant. Mesopelagic bacterial communities were as rich and variable as surface assemblages, despite the significant biomass decrease along the water column. Natural archaeal assemblages were characterized by very low richness as we recovered only two genera (Cenarchaeum and Nitrosopumilus), while in situ bacterial communities were composed by the six major marine phyla (Proteobacteria, Cyanobacteria, Bacteroidetes, Actinobacteria, Firmicutes and Deinococcus-Thermus), whose contribution varied according to sampling depth. Flagellates were demonstrated to efficiently control their preys (ingestion rates: 7.86-22.26 µg C L-1 in surface experiments, 0.53-10.61 µg C L-1 in deep experiments), causing important losses in the potentially produced prokaryotic biomass. Despite picoplankton and HNF abundance reduction with depth contrasts with the hypothesis that at least 108 picoplanktonic cells L-1 are necessary to sustain HNF community, our data confirm that also in mesopelagic waters prey and predator concentrations are sufficient to sustain efficient microbial food webs. HNF grazing modified bacterial community diversity in both surface and deep marine systems but with different strength. Mesopelagic communities were more sensitive to grazing impact, evidencing a bell-shaped response to the increasing ingestion rates. Moderate-high top-down control preserved or enhanced bacterial diversity, that fell at low predation. In upper communities grazing did not induce wide variations of bacterial richness and evenness, revealing to be more stable. Small HNF (<3 µm) were the dominant size fraction within flagellate communities and likely constituted the main bacterivores. After the removal of large HNF, a higher fraction of prokaryotic phylotypes was affected. Larger protists partially reduced small flagellate impact on their preys. Larger HNF had a more important role in photic systems compared to mesopelagic waters. The fraction of bacterial taxa favored or affected by predation when small HNF were the only predators more markedly varied in surface experiments, while few phylotypes changes their behavior between the two size treatments in deep experiments. Some taxa were consumed mainly by larger HNF (3-10 µm), while others were grazed by smaller ones (<3 µm). Over 50% of the predated phylotypes belonged to the rare biosphere, mainly in the surface experiments. Rare bacteria are thus not only a dormant ‘seed bank’ but constitute a fundamental component of microbial food webs and actively vector the carbon transfer toward higher trophic levels, being as important as dominant organisms. Although general patterns applicable to all communities were not found, trends of selectivity over different phylotypes were highlighted within sampling layer along the water column and between different systems. While the majority of predator-prey interactions were characteristic to specific environments, some can be considered common to different systems (e.g. Burkholderiaceae and Pseudomonadaceae were exclusively selected in all mesopelagic sites, Bacterivoracaceae were subjected to small HNF predation independently from sampling site or depth). The Southern Adriatic and Ionian basins were significantly distinguished by both the physicochemical water characteristics and the prokaryotes and protists abundance distributions. Cluster analysis based on Jaccard and Bray-Curtis metrics evidenced that depth and geographical location of sampling sites influenced bacterial community similarity. The Southern Adriatic Sea was clearly distinguished from the Ionian Sea. The Northern Adriatic samples were always separated from the others, coherently with different biotic and abiotic characteristics of the sub-basin. Additionally, temperature, chl a and O2 concentration represented important environmental drivers shaping biodiversity of bacterial communities that inhabit Adriatic and Ionian basins. In conclusion, we evidenced that heterotrophic flagellates control bacterial biomass and select certain taxa among all possible preys, grazing also on the rare ones. HNF predation thus shapes bacterial community structures, which in turn influence the ecosystem functioning. Despite the cell abundance decrease of both predators and preys reduces encounter probabilities, the dark ocean hosts complex microbial food webs, structured around three trophic levels (i.e. prokaryotes, small and large heterotrophic flagellates).