Isotopic Signature of N2O Produced by Marine Ammonia-Oxidizing Archaea

See allHide authors and affiliations

Science  02 Sep 2011:
Vol. 333, Issue 6047, pp. 1282-1285
DOI: 10.1126/science.1208239


The ocean is an important global source of nitrous oxide (N2O), a greenhouse gas that contributes to stratospheric ozone destruction. Bacterial nitrification and denitrification are thought to be the primary sources of marine N2O, but the isotopic signatures of N2O produced by these processes are not consistent with the marine contribution to the global N2O budget. Based on enrichment cultures, we report that archaeal ammonia oxidation also produces N2O. Natural-abundance stable isotope measurements indicate that the produced N2O had bulk δ15N and δ18O values higher than observed for ammonia-oxidizing bacteria but similar to the δ15N and δ18O values attributed to the oceanic N2O source to the atmosphere. Our results suggest that ammonia-oxidizing archaea may be largely responsible for the oceanic N2O source.

Tropospheric concentrations of nitrous oxide (N2O) are currently 322 parts per billion and rising at a rate of ~0.25% per year (1). Future projections for this radiatively active trace gas are uncertain, primarily because biological sources and sinks of N2O are spatially and temporally variable, and their mechanisms are not fully understood. Marine N2O sources to the atmosphere are estimated to represent ~30% of total “natural” inputs, or ~4 Tg N2O-N per year (2). Changes in the magnitudes of these sources depend on the mechanisms and controls of microbial N2O production.

Marine N2O production is thought to be carried out by ammonia-oxidizing bacteria (AOB) during the oxidation of ammonia (NH3) to nitrite (NO2) and reduction of NO2 to N2O in a process frequently termed “nitrifier-denitrification” (3). Heterotrophic denitrifying bacteria also produce N2O as an intermediate and may be an additional source of N2O under suboxic or anaerobic conditions (4). Despite geochemical support for N2O production via nitrification (5, 6), the isotopic composition of marine N2O (69) appears inconsistent with bacterial nitrification as a primary source of marine N2O (10).

Marine Group 1 (MG1) archaea are ubiquitous and abundant members of ocean bacterioplankton, approaching 35% of total microbial cells in the deep ocean (11). At least some MG1 archaea are capable of growing as autotrophic ammonia oxidizers (12), and they are now recognized as the dominant ammonia-oxidizing organisms in the coastal and open oceans (13), potentially due to their high affinity for NH3 (14). Because marine N2O production is tied geochemically to nitrification, the contributions of ammonia-oxidizing archaea (AOA) to marine nitrification may have important implications for marine N2O production. Estimates of in situ N2O production rates suggest that MG1 archaea must contribute to marine N2O production to account for the observed rates (15).

We used the archaeal enrichment cultures CN25 and CN75 from the Pacific Ocean (16) to investigate whether AOA are capable of N2O production (17). We observed N2O production in all experiments where ammonia oxidation occurred (table S1). Time-course experiments (Fig. 1A) showed that N2O production occurred from the onset of ammonia oxidation until NH4+ was exhausted. The N2O yield for archaeal ammonia oxidation ranged from 4.4 × 10−5 to 1.1 × 10−3, or between 0.04 and 1.1 nmol of N2O-N produced for every μmol of NO2 produced (table S1). This is at the low end of previous findings for cultured AOB (18) but consistent with recent results for field estimates (15) and for AOB grown under more environmentally relevant conditions (19). The N2O yield increased with increasing final [NO2] (Fig. 2), as has been observed for marine AOB (19).

Fig. 1

Nitrous oxide production by the marine archaeon CN25. (A) Time course of N2O-N production (solid circles) and NO2 production (open circles) during ammonia oxidation by CN25. Each point represents an individual experimental bottle. Average measurement precision for this experiment was 0.06 nmol for N2O and 0.02 μmol for NO2. (B) Time course δ15N of N2O (solid squares) and NO2 (open squares) produced by CN25 growing in 15N-enriched medium (δ15NNH4 ~690‰, dashed line). (C) Comparison of δ15NN2O and δ15NNO2 produced by CN25 growing in two time-course experiments and two endpoint experiments. The solid line represents 1:1 correspondence between δ15NN2O and δ15NNO2.

Fig. 2

AOA nitrous oxide yield increases with increasing nitrite concentration. Data from nine experiments and 70 individual bottles are shown with both strains CN25 (solid circles) and CN75 (open circles). N2O yield is defined as mol N2O-N produced per mol NO2 produced. Error bars denote multiple-measurement uncertainties for the N2O-N yield calculation (17).

Although CN25 and CN75 are not pure cultures, all evidence suggests that the ammonia-oxidizing archaea in the enrichments are responsible for the observed N2O production. Bacterial cells make up <5% of cells during exponential growth and at most 14% during the stationary phase (16). We used the polymerase chain reaction to screen for genetic evidence of other N2O-producing organisms, including those from the β- and γ-proteobacterial AOB (using amoA, encoding ammonia monooxygenase subunit α, and nirK, encoding the copper-containing nitrite reductase) and denitrifying bacteria (using nirK and nirS, encoding cytochrome cd1 nitrite reductase), and found no evidence of these organisms in the enrichments (17). The addition of allylthiourea, a specific nitrification inhibitor, at concentrations sufficient to inhibit AOA also resulted in complete inhibition of N2O production (fig. S1) despite the presence of NO2, indicating that heterotrophic denitrification was not responsible for the N2O production.

Experiments with added 15NH4+15NNH4 = 690 per mil (‰) versus atmospheric N2 (AIR)] showed that the produced N2O was enriched in 15N (δ15NN2O = 579 to 752‰), compared with NO215NNO2 = 471 to 592‰) at every time point (Fig. 1B), indicating that at least some of the N2O was produced directly from the 15N-enriched NH3 pool (or an intermediate stemming from NH3) without going through the extracellular NO2 pool. This pattern was also observed in endpoint experiments with 15NH4+, where the δ15NN2O produced was usually greater than the δ15NNO2 (Fig. 1C).

Further evidence for a nitrification source of N2O in the enrichments comes from analysis of the intramolecular distribution of 15N in the linear N2O molecule (15N site preference). Relative ratios of the two 15N-substituted isotopomers (15Rα = 14N15NO/14N14NO and 15Rβ = 15N14NO/14N14NO) and site preference [the difference between them (SP = δ15Nα – δ15Nβ)] can be used to determine whether the produced N2O derives mainly from nitrification, or nitrite reduction via denitrification or nitrifier-denitrification pathways (20). Enrichment of 15N at the alpha, or central, site is thought to be consistent with a nitrification source, whereas denitrification produces N2O with much lower site preference (SP = ~0‰). We found a mean site preference of 30.8 ± 4.4‰ for N2O produced by CN25 (Table 1), which can be used to rule out denitrification as the main source of N2O in the culture. Rather, N2O produced in enrichment cultures of CN25 is likely a mixture of N2O produced from both nitrification and nitrifier-denitrification pathways. Marine AOB produce N2O with a site preference ranging from 36.3‰ for NH2OH oxidation to –10.7‰ for nitrifier-denitrification (19). Further experiments will be needed to determine whether AOA have more than one pathway for N2O production and the “end-member” site preference values for each.

Table 1

Isotopic characteristics of N2O produced by the marine ammonia-oxidizing archaeon CN25. SP (site preference) = δ15Nα – δ15Nβ. SD, standard deviation. δ15N values are reported in ‰ versus AIR, and δ18O values are reported in ‰ versus VSMOW. All experiments used air-equilibrated medium with a presumed δ18OO2 ~ 23.5‰ versus VSMOW and an initial [NH4+] of 79 μM. There was complete NH4+ consumption in each experiment; thus, initial δ15NNH4 is approximately equal to the final produced δ15NNO2. Isotope values for NO2 and N2O are the isotopic composition of the NO2 and N2O produced after correcting for preexisting material (17).

View this table:

To further elucidate the pathways by which AOA produce N2O, we conducted experiments with varying amounts of H218O to determine the sources of the oxygen atoms to NO2 and N2O during ammonia oxidation. AOB incorporate one oxygen atom from dissolved O2 during the first step of NH3 oxidation to NH2OH and a second oxygen atom from H2O during the oxidation of NH2OH to NO2 (21, 22), leading to a 50% dependence of δ18ONO2 on δ18OH2O if no further oxygen atom exchange occurs (23). After accounting for abiotic equilibration (17), we found that the δ18ONO2 produced by AOA increased linearly with the δ18OH2O of the culture medium with a slope of 0.5 (Fig. 3A), as would be predicted for the incorporation of one and only one oxygen atom from H2O. Therefore, although the biochemical pathways may differ, the ratio of oxygen atom sources incorporated into NO2 (1:1 H2O to O2) appears to hold for both AOA and AOB.

Fig. 3

Dependence of δ18ONO2 and δ18ON2O produced by CN25 on δ18OH2O. (A) δ18ONO2 produced by CN25 grown in different δ18OH2O media. δ18ONO2 data have been corrected for preexisting NO2 and abiotic isotopic exchange between NO2 and H2O (17). Correspondence of δ18ONO2 to a theoretical slope of 0.5 (dashed line) indicates that one and only one oxygen atom from water is incorporated into NO2 during archaeal ammonia oxidation. (B) δ18ON2O produced by CN25 grown in different δ18OH2O media. If N2O were produced solely from NO2, the slope would be 0.5 (dashed line), the same as δ18ONO2 versus δ18OH2O. If N2O were produced solely from the first step of ammonia oxidation, no oxygen atoms from H2O would be incorporated into N2O and the resulting slope would be zero. The regression slope of 0.18 ± 0.06 shows that the N2O produced by CN25 most likely comes from both nitrification and nitrifier-denitrification pathways. The intercept of this line (27.9 ± 5.7‰) is the δ18ON2O that would be produced in average seawater with δ18OH2O = 0‰.

H218O-labeling experiments also showed that the δ18ON2O produced by AOA is dependent on δ18OH2O, but that only ~18% of the oxygen atoms in N2O originate from H2O under these culture conditions (Fig. 3B). A dependence of δ18ON2O on δ18OH2O could arise in one of three ways: (i) biochemical incorporation of H2O into NO2, followed by reduction of NO2 to N2O (nitrifier-denitrification), (ii) oxygen isotope exchange between H2O and NO2 before NO2 reduction to N2O, or (iii) equilibration of an intermediate product of ammonia oxidation (NH2OH, HNO, or NO) with H2O before decomposition to N2O. Experiments that determined the O atom source in NH2OH from AOB, however, have demonstrated that little exchange occurs between NH2OH and water (21). Moreover, the 15N-labeling experiments and δ15N site preference measurements presented above suggest that a combination of oxidative and reductive processes most likely contribute to N2O production in AOA. We therefore suggest that, although most N2O produced in the enrichment cultures appears to originate from an oxidative process, the δ18OH2O dependence of δ18ON2O most likely arises from a process akin to nitrifier-denitrification in AOA.

The intercept of a linear regression of δ18ON2O against δ18OH2O in CN25 cultures (Fig. 3B) gives an estimate of the δ18O of N2O produced by archaeal ammonia oxidation in average seawater (δ18OH2O = 0‰) of 27.9 ± 5.7‰ versus Vienna standard mean ocean water (VSMOW), consistent with direct measurements of δ18ON2O produced in natural seawater medium without added H218O (Table 1). These values are higher than potential oxygen atom sources [O218O ~ 23.5‰), H2O (δ18O ~ –0.5‰), and NO218O = 9 to 11‰)] but in line with estimates of δ18O for near-surface sources of oceanic N2O (24) and the consistently elevated δ18O values of marine N2O relative to O2 (3, 8). The δ18O enrichment of this source may reflect the influence of oxygen isotope branching effects during N2O production, as has been described for both denitrifiers (25) and AOB (19). If N2O is formed from NO, HNO, or NO2, 16O oxygen atoms may be lost preferentially, resulting in N2O with δ18O greater than its source molecules (25).

Global N2O isotope budgets can be closed by incorporating an ocean source with δ15NN2O = 5 to 10‰ versus AIR and δ18ON2O = 38.5 to 53.5‰ versus VSMOW (26). Oceanographic measurements of N2O isotopologues support these isotopic values of marine N2O sources (7, 9). Although it had previously been assumed that marine N2O was the product of nitrification and nitrifier-denitrification by AOB (3, 8), it has been difficult to reconcile these moderate δ15N and δ18O values with N2O produced from AOB, which tends to have low δ15N (–68 to –10‰ versus AIR, or –68 to –4 ‰ versus supplied NH4+) and δ18O values (10 to 25‰ versus VSMOW) (10, 19). The δ15N values measured here for ammonia-oxidizing archaea are about 8.7‰ versus AIR (Table 1), or 6.2‰ versus supplied NH4+. These values are much higher than for AOB and are in line with δ15NN2O values reported for the shallow subsurface source of N2O in the North Pacific (8) and the global ocean (24, 26), assuming a reduced organic N source similar to suspended particulate nitrogen [–1 to 5‰; (27)]. δ18ON2O values also fall within the range necessary to close the global N2O budget and supply the subsurface source of N2O to the North Pacific (8). Site preferences measured here for AOA, however, are higher than modeled in situ production values (SP = –0.6 to 7.2‰) estimated by (8). This suggests that N2O production by AOA in situ may be associated with low-oxygen microenvironments producing more N2O by NO2 reduction with a lower mean site preference than measured in aerobic batch culture.

With the demonstration of high abundances of AOA coincident with high nitrification rates (15, 28), and now the ability of AOA to produce N2O with the appropriate isotopic signatures, we suggest that AOA likely play an important role in N2O production in the near-surface ocean.

Supporting Online Material

Materials and Methods

Fig. S1

Table S1

References (2947)

References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. Acknowledgments: We thank C. Wuchter, S. Sievert, and M. Johnson for assistance in the laboratory. We also thank C. Frame, L. Codispoti, and three anonymous reviewers for helpful discussions and feedback. Funding for this work was provided by the Woods Hole Oceanographic Institution Ocean Life Institute, a Woods Hole Oceanographic Institution Postdoctoral Scholar Fellowship to A.E.S., and U.S. National Science Foundation grants OCE-0526277 and OCE-0961098 to K.L.C. The 16S ribosomal RNA gene and amoA gene sequences for CN25 and CN75 have been deposited in GenBank under the accession nos. HQ338108, JF521547, and JF521548. The CN25 and CN75 strains are available by request from A.E.S.
View Abstract

Navigate This Article