Factors Controlling Anaerobic Ammonium Oxidation with Nitrite in Marine Sediments

Abstract

Factors controlling the anaerobic oxidation of ammonium with nitrate and nitrite were explored in a marine sediment from the Skagerrak in the Baltic-North Sea transition. In anoxic incubations with the addition of nitrite, approximately 65% of the nitrogen gas formation was due to anaerobic ammonium oxidation with nitrite, with the remainder being produced by denitrification. Anaerobic ammonium oxidation with nitrite exhibited a biological temperature response, with a rate optimum at 15°C and a maximum temperature of 37°C. The biological nature of the process and a 1:1 stoichiometry for the reaction between nitrite and ammonium indicated that the transformations might be attributed to the anammox process. Attempts to find other anaerobic ammonium-oxidizing processes in this sediment failed. The apparent K_m of nitrite consumption was less than 3 μM, and the relative importance of ammonium oxidation with nitrite and denitrification for the production of nitrogen gas was independent of nitrite concentration. Thus, the quantitative importance of ammonium oxidation with nitrite in the jar incubations at elevated nitrite concentrations probably represents the in situ situation. With the addition of nitrate, the production of nitrite from nitrate was four times faster than its consumption and therefore did not limit the rate of ammonium oxidation. Accordingly, the rate of this process was the same whether nitrate or nitrite was added as electron acceptor. The addition of organic matter did not stimulate denitrification, possibly because it was outcompeted by manganese reduction or because transport limitation was removed due to homogenization of the sediment.

Nitrogen in the biosphere is distributed between a pool of N₂, which is by far the larger pool and can be utilized only by prokaryotes that are able to fix N₂, and a much smaller pool of fixed nitrogen that can be utilized by almost all living organisms. The availability of fixed nitrogen is a major factor regulating the primary production of the sea (10, 30). Once nitrogen has been transferred into the fixed pool by biological and industrial nitrogen fixation (12), it can be recycled many times between organic and inorganic forms before returning to the N₂ pool. Until recently, denitrification was recognized as the only important process removing nitrogen from the fixed pool in natural environments (see, for example, reference 18). Recently, however, it was discovered that ammonium is oxidized anaerobically in sediments in the presence of nitrate and that this alternative pathway contributes significantly to benthic N₂ production (25). In continental shelf sediments, up to 67% of the N₂ formation was due to anaerobic ammonium oxidation with nitrate (or possibly nitrite) and only 33% of the N₂ formation was due to denitrification (25). The contribution of anaerobic ammonium oxidation to N₂ formation was observed to be greatest in sediment with low organic loading at a water depth of 695 m. Closer to shore, at a depth of 380 m, where the organic loading of the sediment was higher, anaerobic ammonium oxidation was responsible for 24% of the N₂ production, and in a eutrophic coastal bay, the contribution of this process to the total N₂ production was much lower (2%). It was therefore concluded that anaerobic ammonium oxidation was quantitatively most important in sediments with moderate organic loading, as is found on continental shelves. It has been suggested that nitrite accumulating through nitrate reduction is the actual oxidant for ammonium and that the reaction proceeds through the anammox pathway, which was originally described in a wastewater purification system (16). During anammox, NO₂⁻ and NH₄⁺ are combined into N₂ under anoxic conditions (16, 28) by microorganisms capable of autotrophic growth (27). The microorganisms responsible have not yet been isolated but are affiliated with the Planctomycetales order of bacteria (20, 21). The process has now also been described in other wastewater treatment systems (6, 9).

Here we report the results of further investigations of the factors controlling anaerobic ammonium oxidation in marine sediments and its relation to the anammox process. We explored the involvement of nitrite in anaerobic ammonium oxidation, quantified the kinetics of nitrite consumption, assessed the biological nature of the process, and further investigated its competitive relationship to denitrification.

MATERIALS AND METHODS

Study site and sampling.

Sediment was sampled by using a multiple corer (temperature and kinetics experiment) or a box corer (organic-matter experiment) at a water depth of 695 m in the Skagerrak in the North Sea between Denmark and Norway (station S9 [2, 3]) where anaerobic NH₄⁺ oxidation has previously been demonstrated (25). All handling, storage, and incubation of sediment was done at bottom-water temperature (6°C) except for the incubations in the temperature experiment. Sediment from 10-cm (inside diameter) multiple-corer tubes was sectioned in a glove bag under an atmosphere of N₂, and sediment from 13 to 38 mm below the surface, i.e., including the oxic-anoxic interface and the upper part of the anoxic sediment where NO₃⁻ and NO₂⁻ are present (2, 3, 24), was pooled from several cores to obtain a sufficient volume for the experiments and was stirred to homogeneity. The sediment was then processed as described below without being removed from the glove bag. Sediment from box cores was obtained by first removing the upper portion of sediment (approximately 15 mm) by using a thin Plexiglas plate and then transferring the next portion of sediment (approximately 25 mm) into a plastic box where it was stored under 2 cm of bottom water until the experiment was started. The time periods between the sampling and the start of the experiments were 20 h, 2 days, and 8 days for the organic-matter, kinetics, and temperature experiments, respectively.

Nitrite dependence of ammonium oxidation.

Combinations of labeled and unlabeled nitrogen compounds (see below) were added to portions of sediment, and then the sediment was stirred for 1 min and transferred to 28-ml glass centrifuge tubes, which were capped with butyl rubber stoppers, leaving no headspace. Four parallel incubations were performed by adding nitrogen compounds from concentrated stock solutions to the following final concentrations: 100 μM ¹⁴NH₄⁺ plus 50 μM ¹⁵NO₃⁻ for the first incubation, 100 μM ¹⁴NH₄⁺ plus 50 μM ¹⁵NO₂⁻ for the second incubation, 25 μM ¹⁵NH₄⁺ plus 100 μM ¹⁴NO₃⁻ for the third incubation, and 100 μM ¹⁵NH₄⁺ for the fourth incubation.

Duplicate samples were taken at seven time points during the 48-h incubation by centrifugation of tubes from each treatment at 1,600 × g for 10 min. The stopper was then carefully removed, and a 2-ml sample was taken with a high-precision glass syringe with a hypodermic needle for analysis of the isotopic composition of N₂. Approximately 0.1 ml of a 50% (wt/vol) ZnCl₂ solution was drawn into the syringe after the sample had been drawn, and the content of the syringe was then slowly emptied into a 6.6-ml helium-flushed Exetainer vial (Labco, High Wycombe, United Kingdom), with care taken not to mix water and headspace. The overpressure was then removed before the needle was retracted. The remaining supernatant was filtered through a 0.45-μm-pore-size cellulose acetate filter into polypropylene vials and frozen for later analysis.

Temperature dependence.

Portions of sediment (9 ml each) were transferred to a series of 12.6-ml Exetainer glass vials (Labco). Each Exetainer vial was closed with a screw cap holding a 5-mm-thick butyl rubber septum and removed from the glove bag, and the headspace was immediately flushed with helium. The sediment was preincubated at the incubation temperatures of 6.5, 15.7, 25.6, 31.9, 35.7, 38.0, 41.5, 45.1, and 52.9°C for 12 h. The experiment was started by injection of ¹⁵NO₃⁻ and ¹⁴NH₄⁺ from concentrated stock solutions into the vials, followed by vigorous shaking for 1 min until final concentrations of 50 and 100 μM, respectively, were reached.

For sampling, the Exetainer vials were shaken vigorously for 1 min and centrifuged at 1,600 × g for 10 min. To sample the headspace for the isotopic composition of N₂, we first injected 2 ml of helium, flushed the syringe five times with the headspace, and finally retracted a 2-ml sample. The gas was transferred to an Exetainer vial (6.6 ml) prefilled with helium-gassed water. An open hypodermic needle allowed the excess water to escape as the gas was injected into the Exetainer vial. The pore-water supernatant was filtered through a 0.45-μm-pore-size cellulose acetate filter into polypropylene vials and frozen for later analysis. Sampling was performed in duplicate or triplicate three times during the incubation period at each incubation temperature. Incubation times varied from 4.3 h (35.7°C) to 14.2 h (6.7°C), which in all cases was too short for the depletion of the NO₃⁻-plus-NO₂⁻ pool, and production of nitrogen gas could therefore be recorded throughout the incubation times at all temperatures. For determination of the initial isotopic composition of N₂ for all incubations, samples were taken from three Exetainer vials preincubated at 6.5°C without added ¹⁵NO₃⁻ and ¹⁴NH₄⁺.

Effects of organic matter.

Sediment was transferred from the plastic box into a glass beaker after first siphoning off the water on top of the sediment. The beaker was then placed in the anoxic glove bag, and the sediment was stirred until it was homogeneous and was then left for 2 h to allow the sediment oxygen consumption to remove any oxygen that might have entered while the sediment was handled outside the glove bag. The sediment was divided into two portions, one of which served as a control while the other received an aliquot of a highly concentrated pure culture of planktonic green algae (Tetraselmis sp.; Reed Mariculture, Inc., San Jose, Calif.) to a final concentration of 1.5 g (ash-free dry weight) per liter of sediment. After being stirred for 1 min, each of the two portions was split into two halves that received two different treatments. Each set consisted of a Tetraselmis-amended portion and an unamended portion, with one set receiving 100 μM ¹⁵NO₃⁻ and the other set receiving 100 μM ¹⁵NH₄⁺ (final concentrations). The sediment was then transferred to glass centrifuge tubes and sampled as described above for the kinetics experiment.

Analysis.

The concentration of NO₃⁻ plus NO₂⁻ was determined by using the vanadium chloride reduction method (1) (NO_x analyzer model 42c; Thermo Environmental Instruments, Inc., Franklin, Mass.). Nitrite was analyzed spectrophotometrically (7), and the concentration of NH₄⁺ was determined by using the flow injection method with conductivity detection (8). For analysis of the isotopic composition of N₂, the Exetainer vials were first shaken vigorously for 1 min in order to establish equilibrium between the N₂ in the water and the N₂ in the headspace (>96% of the N₂ was in the headspace at equilibrium). The headspace was then injected into a gas chromatographic column coupled to a triple-collector isotopic ratio mass spectrometer (RoboPrep G⁺ in line with TracerMass; Europa Scientific, Crewe, United Kingdom), and the abundance and concentrations of ¹⁴N¹⁵N and ¹⁵N¹⁵N were analyzed. The isotopic composition of the NO₃⁻ pool was estimated from the concentrations of NO₃⁻ before and after the addition of ¹⁵NO₃⁻ (99.6% ¹⁵N), and the isotopic composition of the NH₄⁺ pool was determined by using the combined microdiffusion-hypobromite method (19).

Calculations.

The rates of denitrification and NH₄⁺ oxidation with NO₂⁻ in the homogenized sediment were calculated from the ¹⁵NO₃⁻ and ¹⁵NO₂⁻ addition experiments as previously described (25). The calculations are based on the production of ²⁹N₂ and ³⁰N₂ and assume a 1:1 pairing of N from NH₄⁺ and NO₃⁻ or NO₂⁻ during anaerobic ammonium oxidation and random isotope pairing during denitrification. In the NO₃⁻ addition experiments, it is assumed that the ¹⁵N labeling rates of the NO₃⁻ and NO₂⁻ pools are identical.

Estimates of K_m values for NO₃⁻ uptake in the sediment were obtained from progress curves of NO₂⁻ concentrations versus time (5). In the temperature experiment, the rates of NH₄⁺ oxidation with NO₂⁻ and denitrification were determined from linear regressions of the amount of nitrogen gas produced by each of the processes versus time.

RESULTS

Nitrite dependence of ammonium oxidation.

With no addition of NO₃⁻ or NO₂⁻, the NH₄⁺ concentration increased as a result of mineralization by 25 μM over 30 h of anoxic incubation (Fig. 1H), while in the experiments where either NO₃⁻ or NO₂⁻ was added, the NH₄⁺ concentration remained stable during the consumption of NO₃⁻ and NO₂⁻ (Fig. 1B, D, and F). ¹⁵N labeling of the NH₄⁺ pool showed that this difference was due to the consumption of NH₄⁺ in the presence of NO₃⁻ and NO₂⁻. Thus, the ¹⁵N content of the NH₄⁺ pool was 53% at the beginning of the experiment but decreased to 16% at the end of the experiment due to dilution by unlabeled NH₄⁺ produced by mineralization. In experiments with ¹⁵NH₄⁺ and unlabeled NO₃⁻ and NO₂⁻, the consumption of NH₄⁺ was accompanied by the accumulation of ¹⁵N-labeled nitrogen gas, and all of the ¹⁵N converted to N₂ was recovered as ²⁹N₂ (Fig. 1E). No N₂ production occurred in the absence of NO₃⁻ and NO₂⁻ (Fig. 1G).

FIG. 1.

Concentrations of ²⁹N₂, ³⁰N₂, NO₃⁻, NO₂⁻, and NH₄⁺ as functions of time in anoxic sediment incubations with addition of ¹⁵NO₃⁻ plus ¹⁴NH₄⁺ (A and B), ¹⁵NO₂⁻ plus ¹⁴NH₄⁺ (C and D), ¹⁵NH₄⁺ plus ¹⁴NO₃⁻ (E and F), or ¹⁵NH₄⁺ (G and H). In panels A and B, ¹⁵NO₃⁻ accounted for 98% of the NO₃⁻ pool; in panels C and D, ¹⁵NO₂⁻ accounted for 98% of the NO₂⁻ pool; in panels E and F, ¹⁵NH₄⁺ initially accounted for 53% of the NH₄⁺ pool, decreasing to 16% at the end of the experiment; and in panels G and H, the ¹⁵N content of the NH₄⁺ pool decreased from 77 to 39% between the start and the end of the experiment. Error bars indicate standard errors.

In the experiments in which NO₃⁻ was added, there was a rapid decrease in the concentration of NO₃⁻ that was instantly accompanied by a transient appearance of NO₂⁻ (Fig. 1B and F). Production of ¹⁵N-labeled N₂ was recorded immediately after addition of either ¹⁵NO₃⁻ or ¹⁵NO₂⁻ to the anoxic sediment (Fig. 1A and C), and the concentrations of both ²⁹N₂ and ³⁰N₂ increased linearly with time until all of the NO₂⁻ was consumed (Fig. 1B and D). Control experiments showed that less than 2% of the added ¹⁵NO₃⁻ was reduced to NH₄⁺ in this sediment. Thus, dissimilatory NO₃⁻ reduction to NH₄⁺ did not affect our results. The initial ¹⁵N content in the NO₃⁻ and NO₂⁻ pools of the two experiments was 98% and was assumed not to change with time. When NO₂⁻ was added, the very small amount of endogenous NO₃⁻ was rapidly consumed, after which the concentration of NO₂⁻ decreased linearly almost to zero.

The production rates of ²⁹N₂ and ³⁰N₂ differed only slightly between the experiments with the additions of ¹⁵NO₃⁻ (Fig. 1A) and ¹⁵NO₂⁻ (Fig. 1C). The production of ²⁹N₂ was 0.84 μM h⁻¹ with NO₃⁻ and 0.82 μM h⁻¹ with NO₂⁻, and the production of ³⁰N₂ was 0.29 μM h⁻¹ with NO₃⁻ and 0.34 μM h⁻¹ with NO₂⁻. Accordingly, the relative importance of anaerobic NH₄⁺ oxidation in N₂ production was almost the same whether NO₃⁻ or NO₂⁻ was added (74 and 71%, respectively). Calculations based on the changes between subsequent samplings during the incubations showed that there was no systematic variation in the relative importance of NH₄⁺ oxidation and denitrification as a function of NO₂⁻ concentrations between 1 and 50 μM in either type of experiment (data not shown), which is in agreement with findings from previous experiments involving NO₃⁻ additions (25).

Kinetics of nitrite uptake.

The theoretical time courses of NO₂⁻ concentration that were calculated assuming Michaelis-Menten kinetics with different K_m values were compared with our measurements (Fig. 2). The parameters used for the ¹⁵NO₃⁻ addition experiment were as follows: maximum substrate consumption rate (V_max) of −1.53 μM h⁻¹ and initial substrate concentration (S₀) of 61.1 μM. For the ¹⁵NO₂⁻ addition experiment, the parameters were as follows: V_max of −2.09 μM h⁻¹ and S₀ of 65.4 μM. The best agreement between the theoretical and measured concentrations of NO₂⁻ was obtained for a K_m value of 0.1 μM; however, acceptable agreement was found for K_m values up to 3 μM. The K_m value was hence estimated to be below 3 μM (Fig. 2A and B).

FIG. 2.

Concentrations of NO₂⁻ as functions of time in two experiments with either NO₃⁻ added (A) or NO₂⁻ added (B). The solid lines indicate the theoretical decrease in NO₂⁻ concentration assuming Michaelis-Menten kinetics and a K_m value of 0.1 μM, a maximum NO₂⁻ reduction rate calculated as the slope of the linear portion of the curve, and an initial concentration calculated as the intersection of this linear regression with the y axis. The other lines depict the decrease in NO₂⁻ concentration assuming higher K_m values. Error bars indicate standard errors.

Temperature dependence.

Anaerobic NH₄⁺ oxidation and denitrification responded differently to changes in temperature. The rate of anaerobic NH₄⁺ oxidation was highest at 15°C and decreased sharply above 25°C, reaching zero around 37°C (Fig. 3A). Denitrification had a wider optimum range (from 15 to 32°C), decreased less abruptly towards higher temperatures, and was detectable at temperatures up to 45°C. The relative importance of NH₄⁺ oxidation with NO₂⁻ for N₂ production decreased with increasing temperature (Fig. 3B). Ammonium oxidation with NO₂⁻ was responsible for approximately 80% of the total nitrogen gas production at the in situ temperature of 6°C, whereas more than 98% of the nitrogen gas was produced by denitrification at 37°C. The activation energy was 61 kJ mol⁻¹.

FIG. 3.

Rates of dinitrogen production by anaerobic oxidation of NH₄⁺ with NO₂⁻ and by denitrification as a function of temperature (A) and the production of dinitrogen by the oxidation of NH₄⁺ with NO₂⁻ relative to the total dinitrogen production in the sediment (B). Error bars indicate standard errors.

Effects of organic matter on anaerobic ammonium oxidation and denitrification.

The addition of organic matter resulted in a statistically significant (P < 0.05) stimulation of the rate of mineralization, as indicated by a 102% increase in the rate of NH₄⁺ production (from 1.5 to 3.1 μM N h⁻¹). Denitrification was only slightly stimulated (from 0.58 to 0.62 μM N h⁻¹ [8% increase]; not statistically significant), and NH₄⁺ oxidation with NO₂⁻ was only slightly reduced (from 0.57 to 0.53 μM N h⁻¹ [8% decrease]; not statistically significant) by the addition of organic matter. This resulted in a decrease in the relative importance of NH₄⁺ oxidation with NO₂⁻ for the total N₂ production (from 49.7% in the unamended experiment to 45.9% in the amended experiment). As in the other experiments, NO₂⁻ accumulated transiently. With the addition of ¹⁵NH₄⁺ alone, no transfer of ¹⁵N from NH₄⁺ to N₂ could be detected in either the amended or the unamended experiment (data not shown).

DISCUSSION

Nitrite dependence of ammonium oxidation.

Previous experiments that demonstrated anaerobic NH₄⁺ oxidation in sediments were all performed with the addition of NO₃⁻, and although NO₂⁻ was suggested to be the oxidant for NH₄⁺ based on patterns of isotope pairing, a direct involvement of NO₃⁻ could not be excluded (25). The results presented here directly demonstrate that NO₂⁻ alone may serve as an oxidant for NH₄⁺ and that NH₄⁺ oxidation is not directly dependent on NO₃⁻. Thus, anaerobic NH₄⁺ oxidation proceeded in the absence of NO₃⁻ both after the depletion of added NO₃⁻ and when NO₂⁻ was added (Fig. 1). Furthermore, the process proceeded at similar rates both in the presence of NO₃⁻ and NO₂⁻ and in the presence of NO₂⁻ alone, which indicates that NO₃⁻ was not involved in pathways of NH₄⁺ oxidation that bypass NO₂⁻. As in the previous sediment studies and studies of anammox in wastewater reactors, the patterns of isotope pairing were also consistent with N₂ formation through a 1:1 pairing of nitrogen from NO₂⁻ and NH₄⁺. Thus, if NO₃⁻ were the direct oxidant of NH₄⁺, as was originally suggested for the wastewater systems (16), then the oxidation states of the nitrogen atoms in NO₃⁻ and NH₄⁺ would require that 5 NH₄⁺ react with 3 NO₃⁻ to form 4 N₂. Consequently, at least some ³⁰N₂ should be formed in the experiments with the addition of ¹⁵NH₄⁺ and ¹⁴NO₃⁻, but that was clearly not the case (Fig. 1E). By contrast, the simpler stoichiometry of NH₄⁺ oxidation with NO₂⁻ is consistent with a 1:1 pairing of their nitrogen atoms: NO₂⁻ + NH₄⁺ → N₂ + 2H₂O. We therefore infer that the anaerobic NH₄⁺ oxidation in the sediment occurred as a 1:1 reaction between NO₂⁻ and NH₄⁺.

With NO₂⁻ as the direct oxidant of NH₄⁺, the reduction of NO₃⁻ to NO₂⁻ is a prerequisite for NH₄⁺ oxidation with NO₂⁻ to gain access to the pool of NO₃⁻, which is generally much larger than the pool of NO₂⁻ in the marine environment (12). In the S9 sediment, however, the reduction rate of NO₃⁻ during the first 10 h of the experiment was four times faster than the reduction rate of NO₂⁻ during the last 30 h of the experiment (Fig. 1B), and thus the reduction of NO₃⁻ to NO₂⁻ did not limit NH₄⁺ oxidation with NO₂⁻. This is further supported by the fact that the production rates of ²⁹N₂ were the same in the experiments with added ¹⁵NO₃⁻ and ¹⁵NO₂⁻. The reduction of NO₃⁻ to NO₂⁻ can be carried out by a number of different microorganisms, including denitrifying and other anaerobic-respiring bacteria as well as fermenting bacteria, and there is generally a high capacity for this process in sediments (4, 13). It is thus generally expected that the NH₄⁺ oxidation with NO₂⁻ in marine sediments is not limited by the reduction of NO₃⁻ to NO₂⁻. In intact sediment, NO₂⁻ may also be produced in the oxic surface layer by nitrification and then be transported into the anoxic zone, where it may support an anaerobic NH₄⁺ oxidation.

It has been suggested that oxidation of NH₄⁺ with MnO₂ could convert nitrogen from NH₄⁺ to N₂ under anoxic conditions (11, 15). However, in agreement with our previous findings (24), NH₄⁺ was not oxidized to N₂ in the absence of NO₃⁻ or NO₂⁻ (Fig. 1G).

Temperature dependence.

The typical biological temperature spectrum provides good evidence that anaerobic NH₄⁺ oxidation in the sediment is microbially catalyzed (Fig. 3A). The temperature optimum for NH₄⁺ oxidation with NO₂⁻ of approximately 15°C found in this sediment is much lower than the optimum for the anammox process of 37°C found in wastewater treatment systems (14, 22). The activation energies of the anammox process in wastewater reactors (22) and the NH₄⁺ oxidation with NO₂⁻ in marine sediment were similar (70 and 61 kJ mol⁻¹, respectively). The bottom-water temperature at station S9 is between 4 and 6°C throughout the year (23), and a special adaptation to this low temperature can be expected. Therefore, even though the data shown in Fig. 3 indicate that denitrification would be favored over NH₄⁺ oxidation with NO₂⁻ at higher temperatures, this does not necessarily imply that NH₄⁺ oxidation with NO₂⁻ would be of lesser importance in warmer environments. Instead, the two rather different temperature optima found in these two environments may indicate that the NH₄⁺ oxidation with NO₂⁻ is rather flexible with respect to temperature, and different temperature characteristics may be found in other environments. Similar relatively narrow temperature ranges have also been reported for nitrification in sediments, with the optimum temperature varying as a function of the temperatures in situ (26).

Kinetics.

The relative quantitative importance of N₂ production by NH₄⁺ oxidation with NO₂⁻ and denitrification was found to be independent of the NO₂⁻ concentration. We therefore argue that even though the NO₂⁻ and NO₃⁻ concentrations applied in these experiments were higher than those found in situ (typically ranging from 0 to 30 μM for NO₃⁻ and from 0 to 5 μM for NO₂⁻ [B. Thamdrup and T. Dalsgaard, unpublished data]), the balance between NH₄⁺ oxidation with NO₂⁻ and denitrification found in the anoxic jar experiments represents the balance under in situ conditions. The importance of NH₄⁺ oxidation with NO₂⁻ for nitrogen cycling in this sediment was underlined by the fact that the consumption of NH₄⁺ by the process was able to keep pace with the concomitant NH₄⁺ production by mineralization (Fig. 1B and D).

In these experiments, the other substrate for NH₄⁺ oxidation with NO₂⁻, NH₄⁺, did not reach a level where it limited the rate of the process. For example, in the experiments with addition of organic matter, the average NH₄⁺ concentrations were 52 μM without and 250 μM with addition of the algal culture (data not shown). This fivefold increase in NH₄⁺ concentration did not stimulate NH₄⁺ oxidation with NO₂⁻, and we conclude that the K_m value for NH₄⁺ uptake by this process must be well below 50 μM.

The K_m value for NO₂⁻ uptake in the sediment was estimated by use of the progress curve technique (5) to be below 3 μM (Fig. 2). Since the balance between N₂ production through NH₄⁺ oxidation with NO₂⁻ and through denitrification was independent of NO₂⁻ concentration, we assumed that the balance between the NO₂⁻ uptakes by these two processes was also independent of NO₂⁻ concentration and that the two processes had a similar affinity for NO₂⁻ within the range of concentrations examined here. For anammox sludge, the K_m value was found to be less than 7 μM (22), and the K_m for NO₃⁻ uptake in denitrifying bacterial communities was between 1.8 and 13.7 μM (17). The K_m for anaerobic ammonium oxidation in marine sediments thus seems to be in the same range as that of other nitrate- or nitrite-reducing bacteria.

A change in the controlling factors for denitrification was expected to affect the balance between denitrification and NH₄⁺ oxidation with NO₂⁻. One such factor was expected to be the availability of organic matter, which should favor denitrification. However, the addition of organic matter to the sediment did not have a significant effect on NH₄⁺ oxidation with NO₂⁻ or on denitrification. The addition of the algal debris clearly stimulated the mineralization in the sediment, as indicated by the doubling of the NH₄⁺ production. The insignificant stimulation of denitrification by the addition of organic matter indicates either that the denitrifying bacteria were already saturated with electron donors or that they were outcompeted by other pathways of carbon oxidation. The first scenario could be a result of the homogenization of the sediment, which destroys the gradients that exist in the intact sediment, allowing the bacteria to utilize electron donors from sediment strata from which they were spatially separated in the intact sediment. The incubation may have been too short for the denitrifying community to respond to the increased substrate availability through growth. Alternatively, a major fraction of the added organic matter could be oxidized through dissimilatory Mn oxide reduction. A dominance of Mn-reducing bacteria over denitrifiers in the competition for organic matter is supported by the relatively low rate of denitrification relative to NH₄⁺ accumulation in the sediment without added organic matter. Assuming a typical ratio of carbon oxidation to NH₄⁺ accumulation (the molar CO₂ production to NH₄⁺ accumulation ratios reported for Skagerrak sediments range from 5 to 12 [2, 3]) and a ratio of 5:4 for CO₂ production and NO₃⁻ consumption during organotrophic denitrification, we estimate that denitrification supported 2 to 9% of the carbon oxidation in the incubations. So, we attribute the lack of an effect of carbon addition on the rate of denitrification to either the short incubation time or the unusually high Mn oxide content in sediment at station S9. For more typical marine sediments, where Mn reduction is insignificant in anaerobic carbon oxidation, we expect that the addition of reactive organic matter, given a sufficient response time, would significantly stimulate denitrification at the expense of NH₄⁺ oxidation with NO₂⁻.

Together with our previous findings (25), our demonstration here of the NO₂⁻ dependence of ammonium oxidation, the biological catalysis of the process, and the 1:1 stoichiometry between NO₂⁻ and NH₄⁺ provides a strong indication that anaerobic NH₄⁺ oxidation with NO₂⁻ in sediments occurs through the anammox pathway. Further verification studies should include the detection and possibly the cultivation of the organisms that carry out the process and an investigation of the intermediates in the reaction. The anammox process has two characteristic intermediates, hydroxylamine and hydrazine (29), and identification of these in the marine process would further verify that the biochemistry of the anammox process is also responsible for the anaerobic oxidation of NH₄⁺ in marine sediments.