Ecogenomics and biogeochemical impacts of uncultivated globally abundant ocean viruses

Abstract Ocean microbes drive global-scale biogeochemical cycling 1 , but do so under constraints imposed by viruses on host community composition, metabolism, and evolutionary trajectories 2–5 . Due to sampling and cultivation challenges, genome-level viral diversity remains poorly described and grossly understudied in nature such that <1% of observed surface ocean viruses, even those that are abundant and ubiquitous, are ‘known’ 5 . Here we analyze a global map of abundant, double stranded DNA (dsDNA) viruses and viral-encoded auxiliary metabolic genes (AMGs) with genomic and ecological contexts through the Global Ocean Viromes (GOV) dataset, which includes complete genomes and large genomic fragments from both surface and deep ocean viruses sampled during the Tara Oceans and Malaspina research expeditions 6,7 . A total of 15,222 epi- and mesopelagic viral populations were identified that comprised 867 viral clusters (VCs, approximately genus-level groups 8,9 ). This roughly triples known ocean viral populations 10 , doubles known candidate bacterial and archaeal virus genera 9 , and near-completely samples epipelagic communities at both the population and VC level. Thirty-eight of the 867 VCs were identified as the most impactful dsDNA viral groups in the oceans, as these were locally or globally abundant and accounted together for nearly half of the viral populations in any GOV sample. Most of these were predicted in silico to infect dominant, ecologically relevant microbes, while two thirds of them represent newly described viruses that lacked any cultivated representative. Beyond these taxon-specific ecological observations, we identified 243 viral-encoded AMGs in GOV, only 95 of which were known. Deeper analyses of 4 of these AMGs revealed that abundant viruses directly manipulate sulfur and nitrogen cycling, and do so throughout the epipelagic ocean. Together these data provide a critically-needed organismal catalog and functional context to begin meaningfully integrating viruses into ecosystem models as key players in nutrient cycling and trophic networks.