Recent advances in understanding the ecology of marine systems have been greatly facilitated by the growing availability of metagenomic data, which provide information on the identity, diversity and functional potential of the microbial community in a particular place and time. Here we present a dataset comprising over 5 terabases of metagenomic data from 610 samples spanning diverse regions of the Atlantic and Pacific Oceans. One set of metagenomes, collected on GEOTRACES cruises, captures large geographic transects at multiple depths per station. The second set represents two years of time-series data, collected at roughly monthly intervals from 3 depths at two long-term ocean sampling sites, Station ALOHA and BATS. These metagenomes contain genomic information from a diverse range of bacteria, archaea, eukaryotes and viruses. The data’s utility is strengthened by the availability of extensive physical, chemical, and biological measurements associated with each sample. We expect that these metagenomes will facilitate a wide range of comparative studies that seek to illuminate new aspects of marine microbial ecosystems.

Mesophilic ammonia-oxidizing Archaea (AOA) are abundant in a diverse range of marine environments, including the deep ocean, as revealed by the quantification of the archaeal amoA gene encoding the alpha-subunit of the ammonia monooxygenase. Using two different amoA primer sets, two distinct ecotypes of marine Crenarchaeota Group I (MCGI) were detected in the waters of the tropical Atlantic and the coastal Arctic. The HAC-AOA ecotype (high ammonia concentration AOA) was ≍ 8000 times and 15 times more abundant in the coastal Arctic and the top 300 m layer of the open equatorial Atlantic, respectively, than the LAC-AOA (low ammonia concentration AOA) ecotype. In contrast, the LAC-AOA ecotype dominated the lower meso- and bathypelagic waters of the tropical Atlantic (≍ 50 times more abundant than the HAC-AOA) where ammonia concentrations are well below the detection limit using conventional spectrophotometric or fluorometric methods. Cluster analysis of the sequences from the clone libraries obtained by the two amoA primer sets revealed two phylogenetically distinct clusters. Taken together, our results suggest the presence of two ecotypes of archaeal ammonia oxidizers corresponding to the medium (1.24 µM on average in the coastal Arctic) and low ammonia concentration (< 0.01 µM) in the shallow and the deep waters respectively.

Viruses are an abundant, diverse and dynamic component of marine ecosystems and have a key role in the biogeochemical processes of the ocean by controlling prokaryotic and phytoplankton abundance and diversity. However, most of the studies on virus–prokaryote interactions in marine environments have been performed in nearshore waters. To assess potential variations in the relation between viruses and prokaryotes in different oceanographic provinces, we determined viral and prokaryotic abundance and production throughout the water column along a latitudinal transect in the North Atlantic. Depth-related trends in prokaryotic and viral abundance (both decreasing by one order of magnitude from epi- to abyssopelagic waters), and prokaryotic production (decreasing by three orders of magnitude) were observed along the latitudinal transect. The virus-to-prokaryote ratio (VPR) increased from ∼19 in epipelagic to ∼53 in the bathy- and abyssopelagic waters. Although the lytic viral production decreased significantly with depth, the lysogenic viral production did not vary with depth. In bathypelagic waters, pronounced differences in prokaryotic and viral abundance were found among different oceanic provinces with lower leucine incorporation rates and higher VPRs in the North Atlantic Gyre province than in the provinces further north and south. The percentage of lysogeny increased from subpolar regions toward the more oligotrophic lower latitudes. Based on the observed trends over this latitudinal transect, we conclude that the viral–host interactions significantly change among different oceanic provinces in response to changes in the biotic and abiotic variables.