Nitrification, the microbial oxidation of ammonia to nitrite and nitrate, occurs in a wide variety of environments and plays a central role in the global nitrogen cycle. Catalyzed by the enzyme ammonia monooxygenase, the ability to oxidize ammonia was previously thought to be restricted to a few groups within the ␤-and ␥-Proteobacteria. However, recent metagenomic studies have revealed the existence of unique ammonia monooxygenase ␣-subunit (amoA) genes derived from uncultivated, nonextremophilic Crenarchaeota. Here, we report molecular evidence for the widespread presence of ammonia-oxidizing archaea (AOA) in marine water columns and sediments. Using PCR primers designed to specifically target archaeal amoA, we find AOA to be pervasive in areas of the ocean that are critical for the global nitrogen cycle, including the base of the euphotic zone, suboxic water columns, and estuarine and coastal sediments. Diverse and distinct AOA communities are associated with each of these habitats, with little overlap between water columns and sediments. Within marine sediments, most AOA sequences are unique to individual sampling locations, whereas a small number of sequences are evidently cosmopolitan in distribution. Considering the abundance of nonextremophilic archaea in the ocean, our results suggest that AOA may play a significant, but previously unrecognized, role in the global nitrogen cycle.Crenarchaeota ͉ nitrification ͉ ammonia monooxygenase

A combination of stable isotope and molecular biological approaches was used to determine the activity, abundance and diversity of nitrifying organisms in the central California Current. Using (15)NH(4)(+) incubations, nitrification was detectable in the upper water column down to 500 m; maximal rates were observed just below the euphotic zone. Crenarchaeal and betaproteobacterial ammonia monooxygenase subunit A genes (amoA), and 16S ribosomal RNA (rRNA) genes of Marine Group I Crenarchaeota and a putative nitrite-oxidizing genus, Nitrospina, were quantified using quantitative PCR. Crenarchaeal amoA abundance ranged from three to six genes ml(-1) in oligotrophic surface waters to > 8.7 x 10(4) genes ml(-1) just below the core of the California Current at 200 m depth. Bacterial amoA abundance was lower than archaeal amoA and ranged from below detection levels to 400 genes ml(-1). Nitrification rates were not directly correlated to bacterial or archaeal amoA abundance. Archaeal amoA and Marine Group I crenarchaeal 16S rRNA gene abundances were correlated with Nitrospina 16S rRNA gene abundance at all stations, indicating that similar factors may control the distribution of these two groups. Putatively shallow water-associated archaeal amoA types ('Cluster A') decreased in relative abundance with depth, while a deep water-associated amoA type ('Cluster B') increased with depth. Although some Cluster B amoA sequences were found in surface waters, expressed amoA gene sequences were predominantly from Cluster A. Cluster B amoA transcripts were detected between 100 and 500 m depths, suggesting an active role in ammonia oxidation in the mesopelagic. Expression of marine Nitrosospira-like bacterial amoA genes was detected throughout the euphotic zone down to 200 m. Natural abundance stable isotope ratios (delta(15)N and delta(18)O) in nitrate (NO(3)(-)) and nitrous oxide (N(2)O) were used to evaluate the importance of nitrification over longer time scales. Using an isotope mass balance model, we calculate that nitrification could produce between 0.45 and 2.93 micromol m(-2) day(-1) N(2)O in the central California Current, or approximately 1.5-4 times the local N(2)O flux from deep water.

The ocean is an important global source of nitrous oxide (N(2)O), a greenhouse gas that contributes to stratospheric ozone destruction. Bacterial nitrification and denitrification are thought to be the primary sources of marine N(2)O, but the isotopic signatures of N(2)O produced by these processes are not consistent with the marine contribution to the global N(2)O budget. Based on enrichment cultures, we report that archaeal ammonia oxidation also produces N(2)O. Natural-abundance stable isotope measurements indicate that the produced N(2)O had bulk δ(15)N and δ(18)O values higher than observed for ammonia-oxidizing bacteria but similar to the δ(15)N and δ(18)O values attributed to the oceanic N(2)O source to the atmosphere. Our results suggest that ammonia-oxidizing archaea may be largely responsible for the oceanic N(2)O source.

Submarine groundwater discharge to coastal waters can be a significant source of both contaminants and biologically limiting nutrients. Nitrogen cycling across steep gradients in salinity, oxygen and dissolved inorganic nitrogen in sandy 'subterranean estuaries' controls both the amount and form of nitrogen discharged to the coastal ocean. We determined the effect of these gradients on betaproteobacterial ammonia-oxidizing bacteria (beta-AOB) and ammonia-oxidizing archaea (AOA) in a subterranean estuary using the functional gene encoding ammonia monooxygenase subunit A (amoA). The abundance of beta-AOB was dramatically lower in the freshwater stations compared with saline stations, while AOA abundance remained nearly constant across the study site. This differing response to salinity altered the ratio of beta-AOB to AOA such that bacterial amoA was 30 times more abundant than crenarchaeal amoA at the oxic marine station, but nearly 10 times less abundant at the low-oxygen fresh and brackish stations. As the location of the brackish mixing zone within the aquifer shifted from landward in winter to oceanward in summer, the location of the transition from a beta-AOB-dominated to an AOA-dominated community also shifted, demonstrating the intimate link between microbial communities and coastal hydrology. Analysis of ammonia-oxidizing enrichment cultures at a range of salinities revealed that AOA persisted solely in the freshwater enrichments where they actively express amoA. Diversity (as measured by total richness) of crenarchaeal amoA was high at all stations and time points, in sharp contrast to betaproteobacterial amoA for which only two sequence types were found. These results offer new insights into the ecology of AOA and beta-AOB by elucidating conditions that may favour the numerical dominance of beta-AOB over AOA in coastal sediments.