Substantial oxygen consumption by aerobic nitrite oxidation in oceanic oxygen minimum zones

Oxygen minimum zones (OMZs; DO <20 µM) and anoxic marine zones (AMZs) are hotspots of coupled carbon, oxygen, and nitrogen cycling. In these regions, nitrite is produced via nitrate reduction (forming secondary nitrite maxima) and may be either further reduced (leading to nitrogen loss) or oxidized to nitrate by nitrite-oxidizing bacteria (NOB). Despite very low DO, nitrite oxidation has been observed at appreciable rates in OMZs and may represent a significant oxygen sink that helps maintain low-oxygen conditions. However, the quantitative contribution of nitrite oxidation to overall oxygen consumption has not been directly assessed due to the scarcity of simultaneous oxygen consumption rate (OCR) measurements. The study’s central question is whether nitrite oxidation is a significant sink for DO in OMZs and over what DO range this occurs. Addressing this is critical for understanding OMZ formation/expansion and the balance of carbon respiration and nitrogen transformations, particularly within the secondary chlorophyll maximum (SCM) where photosynthetic O2 production requires concurrent O2 consumption.

Previous work shows that nitrite oxidation is widespread in the ocean, peaking near the base of the euphotic zone but remaining high in OMZs despite low DO. NOB have high oxygen affinities (reported Km values in the sub-µM range), enabling activity at nanomolar O2. Some NOB (e.g., Nitrospina, Nitrococcus) exhibit adaptations to low oxygen and metabolic flexibility. While in situ profiles often show elevated nitrite oxidation at low DO, oxygen manipulation experiments generally show decreasing nitrite oxidation with decreasing DO, suggesting niche partitioning and differing O2 affinities among NOB. OCR measurements in OMZs are rare but indicate rapid O2 consumption in SCMs and strong declines with depth/DO. Photosynthetic O2 production by Prochlorococcus in SCMs (up to ~100 nM O2 d−1) implies significant concurrent sinks to prevent O2 buildup. The extent to which nitrite oxidation contributes to OCR, particularly relative to heterotrophic respiration and other processes, has remained poorly quantified.

Study area and sampling: Six stations in the eastern tropical North Pacific (ETNP) were sampled (three OMZ and three AMZ stations). Depth profiles targeted the upper 100 m to capture the euphotic zone and SCM, and a range of DO values from ~200 µM to anoxic SNM waters at AMZ stations. Samples were collected in April 2017 and June 2018 using a CTD rosette equipped with sensors for salinity, temperature, DO, chlorophyll fluorescence, and PAR.

Rate measurements: Parallel measurements of (i) nitrite oxidation (15NO2− tracer), (ii) oxygen consumption rates (OCR; optode sensor spots), and (iii) ammonia oxidation (15NH4+ tracer) were performed along depth profiles and in controlled oxygen manipulation experiments.

Oxygen measurements and OCR: Optical sensor spots were used in incubation bottles to quantify DO decline over 20–24 h incubations. Two systems were used: Fibox (Loligo Systems; detection limit ~100 nM) for 2017 profiles and FireSting (Pyroscience; detection limit ~10 nM) for 2018 experiments and low-level OCR. Sensors were calibrated using O2-saturated and O2-free (sodium sulfite, He-purged) water. OCR was calculated from linear decreases in DO over time; five replicate bottles per depth were used for profiles, including bottles with 15N tracers. DO was measured initially, at 10–14 h, and at 20–24 h; most time courses were linear (r2 ≈ 0.97–0.99).

Nitrite oxidation rates: 98 at.% 15NO2− was added to 5–10% of in situ NO2− (final 12–184 nM, except where in situ NO2− was undetectable in the upper 100 m). After 24 h dark incubations at in situ temperature, samples were frozen. In the lab, residual NO2− was removed with sulfamic acid, neutralized with NaOH, and the 15N enrichment of nitrate was measured via the denitrifier method. Rates were calculated from the accumulation of 15N in NO3−, accounting for initial NO2− concentration, label, and any unlabeled NO2− from ammonia oxidation.

Ammonia oxidation rates: Measured with 15NH4+ additions at tracer levels following established methods; used to correct mass balance for nitrite labeling.

Oxygen manipulation experiments: Seven experiments were conducted (OMZ edge, SCM, and SNM at selected stations). For each experiment, 24 serum bottles (500 mL) with FireSting spots were filled from a single depth. Bottles were sealed with deoxygenated stoppers, He-bubbled to reduce DO to nanomolar levels, and small air additions established a gradient of starting DO (from ~10 nM to hundreds of nM, and in some OMZ-edge treatments into the µM range). Eight bottles received 15NO2−, eight 15NH4+, and eight were unlabeled. DO was monitored continuously in 4–8 bottles per experiment to assess linearity and estimate community O2 affinity (Km) at very low DO when nonlinearity occurred (<~235 nM DO). Incubations ran 16–24 h in the dark at ambient temperature.

Calculations and comparisons: The contribution of nitrite oxidation to OCR was computed by stoichiometry assuming 0.5 mol O2 consumed per mol NO2− oxidized. Relationships between rates and DO were evaluated using power-law fits. Michaelis–Menten parameters (Km, Vmax) for OCR and nitrite oxidation versus DO were estimated across experimental DO gradients (overall Km) and from low-DO time courses (low-level Km). Reported Km values included: OCR low-level Km 53–127 nM; overall OCR Km ~1.6–6.0 µM; nitrite oxidation overall Km as low as ~34 nM (SNM) and typically <~350 nM in SCM/SNM, higher at OMZ edge.

Isotopes and ‘omics: Natural abundance nitrate δ15N–δ18O pairs were measured (denitrifier method) to compute Δ(15,18) deviations from 1:1 denitrification slopes, diagnosing nitrite oxidation ‘overprinting’ under low O2. 16S rRNA gene and rRNA sequencing identified Nitrospina ASVs and activity. Metagenomes from OMZ edge, SCM, and SNM were screened for Nitrospina nitrite oxidoreductase (nxr) genes and high-affinity terminal oxidases (cytochrome c oxidase; bd-type), as well as genes suggestive of anaerobic flexibility (chlorite dismutase, formate dehydrogenase, nitrate reductase) and Prochlorococcus abundance.

Quality control and replication: Multiple independent bottle replicates enabled statistical robustness and sensitivity to low OCR. DO contamination was minimized and quantified via optodes in every bottle; rates measured below the SCM (where in situ DO is typically absent) are reported as potential rates given unavoidable O2 introduction during handling.

• Nitrite oxidation rates were highest at AMZ stations (1–3), with pronounced elevations within the secondary chlorophyll maximum (SCM) and into the secondary nitrite maximum (SNM). Peak nitrite oxidation rates at the base of the euphotic zone and in the OMZ were on the order of ~69–96 nmol L−1 d−1, with values exceeding 100 nmol L−1 d−1 where SCM and SNM overlapped.
• OCR profiles declined exponentially with depth and decreasing DO, with SCM OCR typically 160–1380 nmol O2 L−1 d−1 and below-SCM potential OCR 120–390 nmol O2 L−1 d−1, consistent with prior OMZ studies.
• The fraction of OCR attributable to nitrite oxidation increased systematically as DO declined, particularly below 2 µM DO. From profiles, nitrite oxidation contributed typically 10–40% of OCR (up to 69% at Station 1), whereas ammonia oxidation contributed <5%.
• Oxygen manipulation experiments corroborated profiles: OCR decreased strongly with decreasing DO, especially below ~1–2 µM, while nitrite oxidation was less sensitive to DO decreases. Consequently, nitrite oxidation accounted for progressively larger fractions of OCR at lower DO. In the SCM at Station 2, nitrite oxidation explained up to 97% of OCR at DO just below 393 nM, indicating it can sustain nearly all DO consumption at sub-µM levels. Across experiments and profiles, the nitrite oxidation share of OCR followed significant power-law relationships with DO (profiles r2 ≈ 0.51; experiments r2 ≈ 0.40).
• Kinetic parameters indicated high community O2 affinity: low-level OCR Km 53–127 nM; overall OCR Km ~1.6–6.0 µM. Nitrite oxidation overall Km values were as low as 34 nM (SNM) and generally <~350 nM in SCM/SNM, consistent with high-affinity NOB.
• Isotopic evidence (nonlinear δ15N–δ18O relationships; negative Δ(15,18) anomalies) demonstrated isotopic ‘overprinting’ by nitrite oxidation within and around the SCM across AMZ stations, consistent with active nitrite reoxidation coupled to nitrate reduction under low DO.
• ‘Omics support: Nitrospina ASVs were abundant in DNA and rRNA libraries in SCM/SNM layers, and metagenomes contained abundant Nitrospina nxr and high-affinity cytochrome oxidase genes, with higher representation in SCM/SNM than at OMZ edges. Genes indicative of anaerobic flexibility (chlorite dismutase, formate dehydrogenase, nitrate reductase) were also detected, especially in SNM samples. Prochlorococcus genes were abundant in SCMs, supporting cryptic oxygen cycling.
• Overall, nitrite oxidation emerges as a substantial oxygen sink across OMZ gradients, becoming dominant only at very low DO (<~0.4 µM), thereby contributing to the maintenance of ultra-low O2 conditions.

The study directly quantifies the role of nitrite oxidation as an oxygen sink in OMZ waters. By pairing 15NO2−-based nitrite oxidation rates with simultaneous OCR across natural gradients and controlled DO manipulations, the authors show that nitrite oxidation becomes an increasingly important fraction of oxygen consumption as DO falls below ~2 µM. At sub-µM DO, nitrite oxidation can, under certain conditions (e.g., SCM at Station 2), account for nearly all measured OCR, indicating that NOB can scavenge scarce O2 effectively. This addresses the central question by demonstrating that nitrite oxidation is a significant DO sink over a broad low-oxygen range and can dominate at nanomolar DO.

The findings imply strong coupling between nitrate reduction (supplying nitrite) and Nitrospina-mediated nitrite oxidation in OMZ interiors, sustaining ‘cryptic’ nitrogen-oxygen cycling. Isotopic anomalies and Nitrospina abundance/activity corroborate this coupling. As DO declines, heterotrophic aerobic respiration rates decrease and facultative heterotrophs may switch to nitrate reduction, further fueling nitrite oxidation. The high O2 affinity and presence of high-affinity terminal oxidases in Nitrospina enable efficient DO scavenging, which can help maintain DO at levels that favor anaerobic nitrogen transformations, reinforcing OMZ conditions.

These results refine our understanding of OMZ biogeochemistry, quantitatively constraining the share of oxygen consumption attributable to nitrite oxidation, reconciling previous contrasting observations (high in situ nitrite oxidation at low DO vs. declining rates in manipulations), and highlighting depth- and assemblage-specific kinetics and ecotypes.

This work establishes quantitative bounds on the contribution of aerobic nitrite oxidation to oxygen consumption in the ETNP OMZ. Nitrite oxidation typically accounts for 10–40% of OCR below 2 µM DO and can dominate (>50%, up to ~97%) at sub-µM DO, especially within SCM waters. Coupled evidence from rate measurements, oxygen manipulations, nitrate isotopes, and microbial community genomics/transcriptomics identifies Nitrospina as a key low-oxygen nitrite oxidizer with high O2 affinity that helps maintain ultra-low oxygen conditions and recycles nitrite to nitrate, limiting further nitrogen loss.

Future directions include in situ incubations minimizing O2 perturbations, higher-resolution temporal studies to capture variability and mixing events, ecotype-resolved kinetics for Nitrospina and co-occurring functional groups across DO gradients, and integration of these kinetics into biogeochemical and ecosystem models to project OMZ dynamics under deoxygenation.

• Potential rates below the SCM: Because in situ DO is often undetectable there, any O2 introduced during sampling/handling can artifactually enable aerobic processes; rates below SCM are therefore reported as potential.
• Experimental DO ranges: Some SCM/SNM treatments included DO levels higher than typical in situ conditions to span responses; OCR values at these higher DOs represent potential rates.
• Bubbling effects: He bubbling and air additions used to set DO could alter particulate/DOM availability and stimulate OCR (noted for OMZ edge experiments).
• Mixed-assemblage kinetics: Derived Km and Vmax reflect community-level aggregates across multiple O2-consuming processes; disentangling process-specific affinities remains challenging.
• Spatial/temporal coverage: Hurricanes limited experimentation at some stations; results may vary with time due to physical mixing, primary production, and substrate variability.
• Unavoidable O2 contamination: Despite optode monitoring, complete avoidance of O2 exposure is impossible; however, per-bottle DO was measured to quantify and interpret rates as functions of DO.