Genome‐resolved viral ecology in a marine oxygen minimum zone

Abstract

Oxygen minimum zones (OMZs) are critical to marine nitrogen cycling and global climate change. While OMZ microbial communities are relatively well-studied, little is known about their viruses. Here, we assess the viral community ecology of 22 deeply sequenced viral metagenomes along a gradient of oxygenated to anoxic waters (<0.02 μmol/l O 2 ) in the Eastern Tropical South Pacific (ETSP) OMZ. We identified 46 127 viral populations (≥5 kb), which augments the known viruses from ETSP by 10-fold. Viral communities clustered into six groups that correspond to oceanographic features. Oxygen concentration was the predominant environmental feature driving viral community structure. Alpha and beta diversity of viral communities in the anoxic zone were lower than in surface waters, which parallels the low microbial diversity seen in other studies. ETSP viruses were largely endemic, with the majority of shared viruses (87%) also present in other OMZ samples. We detected 543 putative viral-encoded auxiliary metabolic genes (AMGs), of which some have a distribution that reflects physico-chemical characteristics across depth. Together these findings provide an ecological baseline for viral community structure, drivers and population variability in OMZs that will help future studies assess the role of viruses in these climatecritical environments.

Microbial Ecology of Oxygen Minimum Zones Amidst Ocean Deoxygenation

“…In each OMZ, viral community composition was appreciably different in anoxic samples compared to oxic samples. In the case of the ETSP (Vik et al, 2020) and Cariaco Basin (Mara et al, 2020), some identified viral populations were detected in other OMZs contained within the Global Ocean Virome 2.0 dataset (Gregory et al, 2019), suggesting that some viral populations specialized for low-oxygen environments may be distributed across global OMZs. However, this represented a small fraction of the total virus populations which hints at a majority of virus populations in OMZs being endemic to their OMZ of origin.…”

“…Further, microeukaryotes such as those belonging to Protozoa are known to be distributed along oxygen gradients from fully oxic to nearly anoxic (Fenchel, 2014). Likewise, viruses have unique communities in OMZs compared to over-and underlying oxic waters (Cassman et al, 2012;Mara et al, 2020;Vik et al, 2020). Thus, physiological and community composition differences are important in assessing how future deoxygenation events will expand or contract microbial niches within OMZs.…”

“…To date, the total virus community composition has only been assessed in the ETSP (Cassman et al, 2012;Vik et al, 2020), the Cariaco Basin (Mara et al, 2020), and Saanich Inlet (Chow et al, 2015) and the putative archaeal virus community has been assessed in the ETNP (Vik et al, 2017). In each OMZ, viral community composition was appreciably different in anoxic samples compared to oxic samples.…”

“…However, the relative abundance of psbA genes in T4-like cyanophages decreases with depth in the ETNP, which may be reflective of its low-light adapted Prochlorococcus host in anoxic chlorophyll maxima and suggests other AMGs are more important in OMZs (Fuchsman et al, 2021). For instance, several putative AMGs have been identified from the Cariaco Basin, the ETNP, the ETSP, and Saanich Inlet (Roux et al, 2014;Ahlgren et al, 2019;Mara et al, 2020;Vik et al, 2020;Gazitúa et al, 2021). Some of the putative AMGs found within OMZs and their overlying oxic waters potentially contribute to important processes in the carbon, nitrogen, and sulfur cycles (Figure 2).…”

Deep sea sediments associated with cold seeps are a subsurface reservoir of viral diversity

“…This process excluded 389 vOTUs owning to low viral scores based on VirSorter2, with 2,496 vOTUs being retained for AMG analysis. Putative AMGs with both auxiliary scores <4 and gene descriptions were selected according to the distillation output of DRAM-v annotations using default parameters [64,66]. Following the method reported by Horst et al [67], and to be conservative, we performed manual curation of the annotation output to improve the confidence in AMG identification, mainly by removing categories of genes related to nucleotide metabolism, organic nitrogen, glycosyl transferases and ribosomal proteins.…”