High Abundances and Expression Levels of Atypical, Non‐Denitrifier N2O Reductases Drive Strong Microbial N2O Consumption Rates in a Minimally Impacted Mangrove Stand

Knowledge of the ecological mechanisms governing N2O cycling in marine sediments lags that of water columns and terrestrial soils, leaving much to be learned about how microbial community dynamics relate to variability in sediment N2O fluxes. The present study assesses these relationships across two distinct environments by focusing on the community structure and activity of N2O reducing microorganisms. The N2O sink capacity of minimally impacted Bermudian mangrove sediments was first estimated using trace‐level microsensors and profile interpretation modeling. Molecular data obtained from these sediments were then compared with those from the Northeast Subarctic Pacific (NESAP) outer continental margin, where previous measurements suggest considerable N2O effluxes. Net N2O uptake was observed for mangrove sediments under ambient and elevated dissolved inorganic nitrogen concentrations (−0.22 ± 0.15 to −0.30 ± 0.26 μmol N2O m−2 d−1), suggesting the microbial potential for N2O consumption exceeded the potential for production via combined nitrification and denitrification. Targeting of bacterial nosZI and nosII gene clusters for quantification using qPCR indicated higher abundance and expression of non‐denitrifier nosZII genes in mangrove sediments demonstrating net N2O uptake. Net N2O production in NESAP sediments was associated with higher abundance and expression of nosZI genes associated with canonical denitrifiers. These results suggest that organisms possessing atypical nosZII genes may act as important N2O scavengers in low‐nitrogen coastal sediments.


Introduction
Nitrous oxide (N 2 O) is an increasingly abundant, ozone depleting greenhouse gas produced and consumed by microbe-mediated metabolic processes in soils, and in freshwater and marine water columns and sediments (Canadell et al., 2021).N 2 O is produced as a by-product of ammonia oxidation in the presence of oxygen, and as an obligatory intermediate of the denitrification pathway under oxygen-limiting conditions (Bange et al., 2010).Elevated rates of N 2 O production and accumulation are generally observed in suboxic waters where high N 2 O yields from both pathways coincide with the inhibition of N 2 O consumption by trace levels of dissolved oxygen (Babbin et al., 2015;Grundle et al., 2017;Ji et al., 2018).Much of the present literature on marine N 2 O cycling is focused on suboxic and anoxic water columns that contribute disproportionately to atmospheric N 2 O emissions (Freing et al., 2012;Yang et al., 2020).Redox gradients in these environments span tens to hundreds of meters, allowing for high-resolution assessments of biogeochemical rate processes, microbial community dynamics, and relationships between the two (Bertagnolli & Stewart, 2018;Ji et al., 2015;Sun, Frey, et al., 2021;Sun, Jayakumar, et al., 2021).Investigations of a similar sort are more challenging in sediment environments, where shallow oxygen penetration depths (OPDs) drive sharp redox gradients that often span just a few millimeters (Devol, 2015).
Coastal and continental shelf sediments can be considerable sources of N 2 O at local scales when dissolved inorganic nitrogen (DIN) levels are high (Allen et al., 2011;Barnes & Owens, 1999;Jameson et al., 2021).Current estimates from estuarine, intertidal, and coastal vegetated environments such as salt marshes and mangrove stands suggest a combined global efflux of 0.15-0.91Tg N 2 O-N yr 1 (Murray et al., 2015).Contributions from nitrification and denitrification to total N 2 O production have been demonstrated in mangrove sediments that act as net sources to the water column and atmosphere (Allen et al., 2007;Bauza et al., 2002;Muñoz-Hincapié et al., 2002).However, observations of N 2 O cycling processes in nearshore sediments in general are biased toward highly populated and anthropogenically disturbed regions of Europe and Asia, potentially leading to overestimations of the relative contributions of coastal and estuarine environments to global N 2 O budgets (Murray et al., 2015).
Systems with low DIN loads, such as minimally impacted coastal mangroves and temperate estuaries, can act as net N 2 O sinks (Foster & Fulweiler, 2016;Maher et al., 2016).Benthic sediments in these systems bring reducing environments near the water column and atmosphere, allowing for net N 2 O consumption when metabolic precursors are limited.Previous work suggests that some sediments can switch rapidly from N 2 O sinks to N 2 O sources when exposed to sufficient DIN enrichment (Maher et al., 2016;Muñoz-Hincapié et al., 2002).Observations of vertical N 2 O profiles in porewaters have also demonstrated increased subsurface N 2 O production following incubation under elevated NH 4 + or NO 3 concentrations (R. L. Meyer et al., 2008;Nielsen et al., 2009).
However, such experiments typically enrich DIN concentrations to millimolar ranges and thus may not accurately reflect the response of low-nitrogen systems to more modest eutrophication scenarios.Furthermore, variability in the magnitude and duration of nutrient enrichment in natural systems can be expected to differentially affect microbial community structure over a range of timescales with likely consequences for the direction and magnitude of sediment N 2 O fluxes.
Surveys of soil and sediment microbial communities reveal stark contrasts in denitrifier community composition at the functional and taxonomic levels across natural DIN and organic matter gradients (Wallenstein et al., 2006;Xie et al., 2020).Genomic surveys have also shown the denitrification pathway to be highly modular and that many organisms do not possess the full genetic machinery required to perform complete denitrification (Graf et al., 2014).High-throughput sequencing of environmental samples has delineated two primary clades of N 2 Oreductase genes (nosZ) responsible for reducing N 2 O to dinitrogen gas (N 2 ) (Jones et al., 2013;Orellana et al., 2014;Sanford et al., 2012).Organisms that possess the clade I nosZ variant (nosZI, or "typical" nosZ) are more likely to carry the full denitrification gene compliment, while clade II nosZ variants (nosZII, or "atypical" nosZ) are associated with genomes that lack upstream pathway components (Bertagnolli et al., 2020;Sanford et al., 2012).
Observations suggest that N 2 O-reducing microbial populations in marine microbiomes are dominated by atypical nosZII variants (Bertagnolli et al., 2020), and that these organisms may act as important N 2 O scavengers across a wide range of environments (Jones et al., 2013(Jones et al., , 2014;;Sun, Frey, et al., 2021;Sun, Jayakumar, et al., 2021).This is consistent with previous work in terrestrial soils that suggests denitrification N 2 O yields are affected by variability in denitrifier community structure (Cavigelli & Robertson, 2000;Jones et al., 2014).However, comparatively little is known about the factors that govern the distribution and activity of nosZI-and nosZII-type reductases in marine sediments and how these dynamics influence benthic N 2 O cycling.At present, the broader relationships between microbial community structure and N 2 O source or sink status remain understudied in marine sediments, especially in offshore and minimally impacted coastal systems.
consumption rates, as well as vertical fluxes, can be accurately estimated from porewater profiles under steadystate conditions through the application of profile interpretation models (Berg et al., 1998).These methods present an alternative means of quantifying N 2 O cycling rate processes in sediment environments in relation to porewater dissolved O 2 concentrations.Combining these approaches with modern molecular tools may prove useful for shedding light on the role of microbial community structure in regulating the direction and magnitude of N 2 O flux across the sediment surface.
Previous work leveraging microsensor profiling techniques demonstrated that continental margin sediments underlying the Northeast Subarctic Pacific oxygen minimum zone (OMZ) are a considerable source of N 2 O to the water column (Jameson et al., 2021).Vertical fluxes of up to 687 nmol m

Field Sampling
Field sampling and incubation experiments for this study took place between 28 October and 01 November 2020 in a subtidal mangrove stand in Ferry Reach, Bermuda, and between 29 September and 04 October 2019 in the NE Subarctic Pacific Ocean (NESAP) (Figure 1).Mangrove sediments were collected in Pyrex mini-core tubes (i.d. = 3 cm, length = 10 cm) at low tide from a mangrove stand, and transported in a cooler approximately 200 m to an outdoor flume mesocosm at the Bermuda Institute of Ocean sciences (Figure 1).Low tide sampling was required to access the sediments for coring.Sediment interfaces were adjusted to ∼2 mm below the edge of the core tube prior to submersion to prevent water stagnation over the sediment interface.Core tubes were then submerged in a 1,500 L mesocosm tank with continuous seawater flow drawn from Ferry Reach and acclimated for 4-8 hr before the start of each incubation period.Core tubes were positioned within 50 cm of the water-air interface to ensure that profile measurements could be made directly in the flume without disturbing the samples.

Experimental Design and Incubation Procedures
Four separate incubation experiments were conducted on each of four consecutive days using sediment cores sampled at the same location.After acclimation, freshly collected sediment cores were incubated under variable nutrient regimes.Incubations were performed for 8-10 hr during dark periods to avoid confounding effects of light and large temperature fluctuations while also assuring steady-state conditions.Indeed, previous work in coastal sediments has shown that steady state conditions are rapidly established (<2 hr) in surface layers following experimental manipulation (Berninger & Huettel, 1997;Rasmussen & Jørgensen, 1992).
Control incubations and incubations involving DIN and N 2 O amendments were performed by submerging cores in a 6 L polyethylene chamber that was partially submerged in the flume tank to create a semi-enclosed system while maintaining consistent water temperatures.For the first experiment, four replicate sediment cores were incubated under elevated N 2 O conditions by bubbling 1.5 ppmv N 2 O gas (in zero-air carrier gas) into the incubation chamber through an aquarium air stone, to a final concentration of 30-32 nmol L 1 (∼5x atmospheric enrichment).

Microelectrode Profiles
Porewater N 2 O and O 2 profiles were obtained at sub-millimeter resolutions using Clarke-type microelectrodes with manufacturer stated detection limits of <25 nmol L 1 and 300 nmol L 1 , and sensor tip diameters of 500 and 200 μm, respectively (Unisense).Sensors were calibrated prior to each measurement period by two-point calibration according to manufacturer instructions.Briefly, zero-point N 2 O and O 2 calibration standards were made fresh daily in 20 mL crimp-sealed vials by bubbling N 2 gas in filtered seawater for at least 15 min.The vials were then sealed in plastic bags and secured within the flume to ensure equilibrium with flume temperature prior to calibration.N 2 O-enriched standards were prepared in a CAL300 calibration chamber (Unisense) immediately prior to calibration by bubbling 1.5 ppmv N 2 O gas into flume seawater for 10 min.High-point O 2 calibration values were obtained by bubbling atmospheric air into the calibration chamber for 10 min to ensure saturation.
The final concentrations of air-saturated and N 2 O-enriched standards were calculated using the solubility coefficients tables of Garcia andGordon (1992, 1993) and Weiss and Price (1980).
Following calibration, microelectrodes were mounted on a Unisense MM33 motorized micromanipulator and sensor tips were manually adjusted to the sediment-water interface.SensorTrace Suite Profiling software (Unisense) was used to control the sensor tip position and log sensor output during profiling.Discrete porewater N 2 O and O 2 concentrations were estimated at 500 and 200 μm depth intervals, respectively, starting one mm above the sediment interface and concluding after two or more consecutive zero values were recorded.At each measurement depth, O 2 sensors were allowed to acclimate for 5 s while the longer response time of the N 2 O sensor necessitated a 25 s acclimation period.All profile measurements were made directly in the flume to minimize sediment disturbance and ensure continuity between incubation and measurement conditions.
Duplicate porewater N 2 O profiles were measured in each mini core yielding a total of 8 profile replicates per experimental group.Raw N 2 O profiles were corrected for baseline drift prior to calculating N 2 O concentrations from the standard curve (Supplementary Methods).Dissolved O 2 profiles were not obtained in post-incubation cores given the considerable time constraints associated with repeated N 2 O profiling.Instead, we utilized a separate O 2 profile data set obtained from experiments that took place between 24 and 27 October 2019 in the same mangrove stand.Briefly, fresh sediment cores were collected daily in quadruplicate at low tide and monitored in the flume over a daily light-dark cycle.Dissolved O 2 profiles obtained at 4 hr intervals over dark periods were used as context for understanding general porewater O 2 distributions and sediment O 2 consumption rates in the mangrove sediment system.

PROFILE Interpretations
Drift-corrected porewater N 2 O profiles were used as input to model depth-dependent production rates, consumption rates, and vertical fluxes across all replicate profiles using the PROFILE model described by Berg et al. (1998).Briefly, PROFILE utilizes the one-dimensional reaction-transport equation for solutes assuming steady state conditions: where ρ is the measured sediment porosity, D s is the effective diffusion coefficient, C is the porewater concentration, and C w is the bottom water concentration, and R is the net rate of production or consumption.The effective diffusion coefficient was calculated as: where D w is the diffusion coefficient of N 2 O in free water.Sediment porosities were calculated for the top 0.5 mm of sediment in two of four cores from each treatment according to Dalsgaard et al. (2000).
Separate models, using the mean porewater concentration profiles shown in Figure 2, were also run to visualize production and consumption profiles for each experimental grouping, and to assess consistency between (a) mean flux estimates averaged across individual profiles, and (b) flux estimates obtained from a single model run of the mean profile.A mean sediment porosity (±SD) of 0.82 ± 0.03 was determined for the top 5 mm of sediments in 2 out of 4 mini-cores from each experimental group (n = 16) according to Dalsgaard et al. (2000) and used for parameterization of the PROFILE models.Diffusivities of N 2 O in water at fixed temperature and salinities used in the model were obtained from Tamimi et al. (1994).Oxygen penetration depth was defined as the depth at which dissolved O 2 decreased below the 5 μmol L 1 suboxic threshold and was estimated in all replicate profiles (n = 67) following linear interpolation between profile points (Belley et al., 2016).Dissolved oxygen profiles showed a high degree of consistency across all replicate profiles.As a result, eight dissolved O 2 profiles were selected at random from the total pool of replicates as a representative subsample for PROFILE modeling of sediment O 2 consumption rates using the same porosity estimate listed above and O 2 diffusivities reported by Garcia andGordon (1992, 1993).

Nucleic Acids Extractions and cDNA Synthesis
Samples for nucleic acid extraction were obtained from the top 0.5 cm of sediment immediately following porewater profiling using 50 mL Falcon tubes for both the NESAP slopes (all cores) and Bermuda mangroves experiments (2/4 cores).For the NESAP slope sediments, initial samples were obtained from separate cores immediately following sampling.Sediment RNA samples were preserved in RNAlater (∼1:10 sample to preservative ratio) and then stored at 80°C.DNA samples were flash-frozen immediately following sampling and then stored at 80°C until extraction in the lab.Total genomic DNA (gDNA) was extracted from ∼700 mg of wet sediment using a DNeasy® PowerSoil® kit (Qiagen, Germany) according to manufacturer instructions.DNA extracts were purified prior to downstream analyses using QIAquick® PCR purification columns (Qiagen) to remove PCR inhibitors and other impurities.DNA concentrations in cleaned extracts were then quantified using a NanoDrop™ One Microvolume UV-Vis Spectrophotometer (Thermo Scientific, USA).In contrast, total RNA was extracted from 2.0 to 5.0 g of wet sediment using a RNeasy® PowerSoil® Total RNA kit (Qiagen, Germany) with the following modification.Thawed sediment samples were centrifuged for 15 min at 2,500 × g and RNA later was decanted.Sediments were then washed three times with 10 mL PBS buffer by briefly vortexing and then centrifuged for 10 min at 2,500 × g.RNA integrity was assessed using a Qubit™ RNA IQ Assay Kit and total RNA concentrations were determined using a Qubit™ RNA High Sensitivity Assay Kit (Invitrogen, USA).Before cDNA synthesis, removal of co-extracted genomic DNA was achieved by treating samples with ezD-Nase™ enzyme according to manufacturer instructions (Invitrogen, USA).A total of 200 ng RNA input was used for standardized cDNA synthesis using a High-Capacity cDNA Reverse Transcription Kit with RNase Inhibitor (Applied Biosystems, USA).Green and red boxes correspond to zones of N 2 O production and consumption, respectively.Solid and dashed horizontal lines denote the mean oxygen penetration depth ± SD, respectively.Gray shading highlights the section depth for sampling of nucleic acids.

Validation of nosZI and nosZII Assays
Published primer sets by Henry et al. (2006) and Jones et al. (2013) were used for quantitative real-time polymerase chain reaction (qPCR) analysis with SYBR Green I chemistry to quantify nosZI and nosZII gene (gDNA) and transcript (cDNA) copy numbers, respectively (Table S1 in Supporting Information S1).First, databases of 399 nosZI and 315 nosZII sequences from cultured organisms were downloaded from the FunGene Repository (Fish et al., 2013).In silico validation of the assays was performed by mapping primer sets to consolidated nosZI and nosZII sequences in Geneious Prime v.2023.0 to assess the coverage and specificity (https://www.geneious.com).Laboratory assay validation was achieved following the modified workflow proposed by Langlois et al. (2021).Primer specificity was assessed using cleaned DNA extracts from a broad range of environmental samples (soil, freshwater, marine) and Pseudomonas aeruginosa quantitative gDNA (strain PAO1-LAC; American Type Culture Collection) as a positive control for the nosZI primer set.Quantitative gDNA was not available for cultured organisms that possess the nosZII variant.The fragment size (in bp) of the resulting PCR amplicons was assessed through electrophoresis in 1% agarose gel, and sent for Sanger sequencing to confirm amplification of target genes.Lastly, gBlocks® synthetic DNA fragments (Integrated DNA Technologies, USA) dilutions of the target amplicons were used to construct standard curves and assess sensitivity.Amplification efficiencies for nosZI and nosZII assays were assessed based on the slope of the constructed standard curve (Figure S2 and Table S2 in Supporting Information S1).
A baseline amplification signal was detected at low copy numbers for nosZI and nosZII assays, and was indicated by lack of signal attenuation below one copy per reaction in the five-fold dilution series.Threshold C q values were thus established based on the continuous limit of detection (LOD cont ) to ensure that the signal detected in environmental samples was above background noise (Table S2 in Supporting Information S1).A certain degree of background noise is expected for qPCR assays targeting bacterial genes, as contamination of common laboratory reagents by prokaryote DNA is often unavoidable (Salter et al., 2014).Three of 39 samples exhibited C q values above the baseline threshold for the nosZII assay and were removed prior to downstream analyses.These samples corresponded to cDNA samples from the NESAP continental margin and demonstrated anomalously low 16S rRNA and nosZI transcript copy numbers.Furthermore, all samples that failed to meet the threshold C q criteria were from different experimental groupings.As such, it is assumed that these outliers reflect complications in the RNA extraction or cDNA conversion processes.

qPCR Analyses of Environmental Samples
The integrity of extracted gDNA and cDNA was assessed with the IntegritE-DNA™ assay targeting a conserved region of the plant chloroplast genome (Hobbs et al., 2019;Veldhoen et al., 2016) and with a TaqMan probebased assay targeting a conserved region of the bacterial 16S rRNA gene (Table S2 in Supporting Information S1; Ritalahti et al., 2006).Probe-based qPCR reactions were carried out in 15 μL reaction volumes using two μL of purified gDNA or cDNA, 700 nM forward and reverse primer, 100 nM TaqMan probe, and 1X of QIAcuity Probe Master Mix (QIAGEN).All thermal cycling steps were performed in a Bio-Rad CFX96 (Bio-Rad) thermocycler.The following TaqMan thermocycler profile was used for all probe-based assays: initial denaturation at 95°C for 9 min followed by 50 cycles of denaturation at 95°C for 15 s, annealing at 64°C for 30 s, and extension at 72°C for 30 s. Successful amplification of endogenous plant chloroplast and bacteria DNA contained within the samples confirms that the recovered nucleic acids are viable and inhibitory compounds have been sufficiently removed.
Quantifications of nosZI and nosZII copy numbers were carried out in 15 μL reaction volumes consisting of two μL of purified gDNA or cDNA, 700 nM forward and reverse primer, and 2X of SensiFAST SYBR® No-ROX mix (Bioline).Eight replicates each of two μL UltraPure™ DNase/RNase-Free water (Invitrogen) and two μL synthetic target DNA (20 copies/reaction) of the appropriate DNA sequence (Integrated DNA Technologies) were used as non-template and positive controls, respectively.The SYBR Green I thermal cycling profiles for nosZI assays were performed with initial denaturation at 95°C for 9 min, followed by 50 cycles of denaturation at 95°C for 15 s, annealing at 64°C for 30 s, and extension at 72°C for 30 s. Thermal cycling conditions for the nosZII assays were modified such that annealing and extension were performed at 54°C and 80°C, respectively.Concentrations of nosZ gene and transcript copies (copies g 1 ) were extrapolated from C q values using the previously generated standard curves.Absolute nosZI and nosZII copy numbers were normalized to 16S rRNA Journal of Geophysical Research: Biogeosciences 10.1029/2023JG007805 copy numbers (gDNA or cDNA) and nosZII ratios were expressed as the ratio of nosZII copies to total nosZ copies (nosZI + nosZII).

High-Throughput Sequencing and Sequence Data Processing
Cleaned raw gDNA extracts were sent to the Integrated Microbiome Resource (Dalhousie University, Halifax, Canada) for sequencing of bacterial and archeal 16S rRNA genes on an Illumina MiSeq using 2 × 300 bp pairedend V3 chemistry and universal primers targeting the V6-V8 variable regions (Table S1 in Supporting Information S1) (https://imr.bio/protocols.html).Demultiplexed forward and reverse reads were trimmed of primer binding sequences using Cutadapt (Martin, 2011) and processed according to standardized protocols in QIIME2 (v2022.8)(https://www.qiime2.org).Reads were merged with VSEARCH (Rognes et al., 2016) using a minimum overlap length of 50 bp and then filtered using a quality score threshold of 20.Denoising of filtered reads into amplicon sequence variants (ASVs) was achieved using QIIME-Deblur (Amir et al., 2017;Bolyen et al., 2019).Bacterial and archeal ASV counts were tabulated and ASV sequences were assigned taxonomic IDs using a naïve Bayesian classifier trained against the Silva reference database (version 138) (Quast et al., 2013).Inverse Simpson diversity indices were calculated separately for bacterial and archeal communities in each sample using the "qiime diversity alpha" command.

Statistical Analyses
All statistical analyses and data visualizations were conducted in the R Statistical Environment (R Core Team, 2019).N 2 O flux and nosZ gene/transcript abundance data sets were tested for normality and homogeneity of variance using Shapiro-Wilks and Levine's tests, respectively.Variability in N 2 O fluxes, and nosZ gene copy and transcript abundances between sample groupings were then tested for statistical significance using one-way ANOVA and Tukey-Kramer post-hoc tests.Variability in inverse Simpson diversity group means was assessed for significance using a Kruskal-Wallis rank sum test and pairwise comparisons were evaluated using Wilcoxon rank sum tests with a Benjamini-Hochberg correction for multiple comparisons.Finally, non-metric multidimensional scaling ordinations of bacterial communities in NESAP continental slopes samples were performed to assess patterns of community dissimilarity across slope depth and between pre-and post-incubation sediments.

PROFILE Interpretations
OPDs were highly consistent across replicate sediment cores and consecutive sampling days, ranging from 1.2 to 2.8 mm and corresponding to a mean (±SD, n = 67) oxygen consumption rate of 15.4 ± 3.9 mmol-O 2 m 2 d 1 (Figure S3 in Supporting Information S1).Mean seawater N 2 O concentrations for the control and DIN-amended incubations was 15 ± 2 nmol L 1 (Figure 2).N 2 O concentrations decreased with depth in all control, N 2 O amended, and NH 4 Cl-amended sediments, and generally reached zero values between 4 and 6 mm (Figure 2; Figure S4 in Supporting Information S1).
Depth-stratified N 2 O consumption rates ranged from 0.10 to 0.18 μmol-N 2 O cm 3 d 1 for control and NH 4 Clamended cores, and increased to a maximum of 0.59 μmol-N 2 O cm 3 d 1 in N 2 O-amended cores (Figure 3).Subsurface N 2 O maxima ranging from 17.6 to 32.7 nmol L 1 were detected below oxycline depths across all cores amended with NO 3 following an initial decrease in concentrations between the sediment surface and the base of the oxycline (Figure 2; Figure S3 in Supporting Information S1).Subsurface N 2 O maxima were associated with zones of net N 2 O production (∼1.49-1.76μmol-N 2 O cm 3 d 1 ) positioned between net N 2 O consumption zones ( 0.13 to 1.02 μmol-N 2 O cm 3 d 1 ).Zones of slight net N 2 O production were estimated in the top few mm for several profiles from the control, N 2 O, and NH 4 + treatments, although porewater concentrations remained lower than those measured in the overlying water column (Figure 2).
Mean N 2 O fluxes calculated across replicate profiles were negative under all incubation conditions, indicating net consumption.Maximum N 2 O consumption rates of 0.94 ± 0.28 μmol-N 2 O m 2 d 1 were detected in N 2 Oamended cores and were significantly higher than all other treatment conditions (p < 0.0001) (Figure 3).In contrast, N 2 O consumption rates did not vary significantly across control and DIN-amended profiles ( 0.22 ± 0.15 to 0.30 ± 0.26 μmol-N 2 O m 2 d 1 ).Mean flux estimates calculated across individual profiles were in good agreement with those derived from the mean porewater profile.Small sediment N 2 O effluxes (0.02-0.19 μmol-N 2 O m 2 d 1 ) were also estimated for two control replicates and three replicates from the NO 3 -amended treatments.However, for all cases in which net N 2 O efflux was estimated from a single profile, the results were not reproduced across both profile duplicates within the same sediment core.).Normalized nosZ gene and transcript copy abundances and nosZII gene and transcript ratios estimated in NESAP initial cores (pre-incubation) were consistent with those estimated in post-incubation cores, and are included in downstream analyses (Figure S4 in Supporting Information S1).

Variability in nosZ
Proportional abundances of nosZII gene copies relative to total nosZ copies (nosZI + nosZII) ranged from 82.5% to 95.4% (91.8 ± 4.3%) across all mangrove samples and from 75.0% to 90.6% (84.5 ± 5.6%) across all NESAP samples (Figure S5a in Supporting Information S1).Contributions of nosZII copies to the total nosZ gene pool were significantly higher in mangrove sediments compared to 200 m ambient and simulated upwelling cores (p = 0.004-0.009).However, differences in nosZII copy ratios between mangrove sediments and deeper slopes sediments were not statistically significant at the 95% confidence level.Normalized nosZI gene copy abundances were relatively consistent across sampling depths in the NESAP and were roughly twice as abundant in NESAP sediments than in the mangrove sediments (Figure S5a in Supporting Information S1).Conversely, mean normalized nosZII gene copy abundances were comparable across mangrove and NESAP sediments (Figure S5b in Supporting Information S1).Mangrove sediment nosZ transcript pools were overwhelmingly dominated by nosZII transcripts (84.8 ± 7.0%), while contributions from nosZI transcripts were significantly higher in NESAP sediments (40.5 ± 10.0%) (Figure S5b in Supporting Information S1).This was ostensibly driven by low mangrove nosZI expression levels relative to the NESAP sediments and not by substantially elevated mangrove sediment community nosZII expression (Figure S5b in Supporting Information S1).

Microbial Community Structure
Amplicon sequencing of prokaryote 16S rRNA genes in the top 5 mm of sediment produced 1.8 million pairedend reads from eight mangrove samples and 19 NESAP continental margin samples.A total of 888,582 bacterial and 388,521 archeal reads were retained following merging and quality filtering.Denoising of filtered reads obtained in mangrove sediments generated 412 archeal ASVs and 1210 bacterial ASVs from 164 Families while 385 archeal ASVs and 4042 bacterial ASVs from 222 Families were detected in samples from the NESAP continental margin.Bacterial community diversity was relatively consistent across mangroves and NESAP sediments while archeal diversity was significantly higher in NESAP sediments (Figure S6a in Supporting Information S1).Simulated upwelling conditions did not have an appreciable effect on microbial community structure, and pre-incubation samples also agreed well with post-incubation samples at discrete slope sampling depths (Figure 4a; Figure S6b in Supporting Information S1).Similarly, microbial community structure was highly consistent across all mangrove sediment manipulation experiments.Approximately one third of mangrove sediment bacterial communities was constituted by ASVs mapped to the Cytophagales (8.8 ± 1.9%), Thiotrichales (8.0 ± 2.1%), Rhodobacterales (7.7 ± 2.1%), and Steroidobacterales (7.1 ± 1.1%) (Figure 5a).A further third of the mangrove communities consisted of ASVs from the Flavobacteriales (6.3 ± 1.4%), Ectothiorhodospirales (6.2 ± 2.9%), Polyangiales (5.6 ± 1.3%), Desulfobulbales (5.1 ± 1.1%), Desulfobacterales (4.1 ± 0.8%), and Bacteroidales (3.6 ± 1.0%).In contrast, NESAP sediments were dominated by ASVs from the uncultured Deltaproteobacteria order NB1-j (12.8 ± 1.6% to 21.5 ± 2.3%), which increased in relative abundance with slope depth along with members of the Steroidobacterales (4.4 ± 0.1%) (Figure 5a).Increases in relative abundances of putative ammonia oxidizing bacteria (AOB) ASVs from the Nitrosococcales order were also observed along the NESAP slope depth gradient.Substantial contributions to community structure were also observed for members of the Thiohalorhabdales, Flavobacterales, Polyangiales, and Cytophagales despite no clear depth-related trends.ASVs belonging to putative sulfatereducing groups, including the Desulfobacterales, Desulfobulbales, and Ectothiorhodospirales, accounted for less than 5% of all NESAP sequence reads combined.

Discussion
Net N 2 O consumption was consistently observed across all mangrove sediment treatment conditions, with significantly higher rates of uptake detected in N 2 O-amended cores.Mean seawater N 2 O concentrations were approximately double the expected values for control and DIN-amended incubations, assuming equilibrium with the atmosphere and an atmospheric dry air mole fraction of 333 ppb (Dlugokencky et al., 2020).This discrepancy is unlikely attributable to issues with calibration curves, as evidenced by the high accuracy of measurements obtained above the sediment interface in N 2 O-amended seawater.Previous reports of surface water N 2 O supersaturation in the Sargasso Sea suggest that local conditions on the Bermuda platform might reflect slightly supersaturated inshore waters (A. C. S. Meyer et al., 2022), although the absence of discrete surface water N 2 O measurements precludes a definitive evaluation of this possibility.Nonetheless, while N 2 O fluxes estimated for the control and DIN-amended cores may be biased slightly high, the overall trends reported herein appear robust.Additionally, the N 2 O consumption rates reported in this study align well with those documented in other lownutrient estuarine, coastal, and terrestrial sediments (Foster & Fulweiler, 2016;Murray et al., 2015;Syakila et al., 2010).
Mean N 2 O consumption rates in Bermudian mangrove sediments remained consistent across control and DINamended incubations, despite observable subsurface production in cores amended with NO 3 .This contradicts previous reports in other coastal, estuarine, and mangrove systems, where experimental DIN enrichment has been shown to stimulate sediment N 2 O efflux (Kreuzwieser et al., 2003;R. L. Meyer et al., 2008;Muñoz-Hincapié et al., 2002).However, such experiments have been typically conducted in environments with relatively high DIN availability and often involve nutrient amendments that exceed concentrations observed in even the most hypernutrified estuaries (Robinson et al., 1998).Results of this study suggest that minimally impacted systems may be capable of mitigating the acute effects of moderate nutrient enrichments over short time frames.Moving forward, further research is needed to delineate concentration thresholds and temporal scales governing changes in the magnitude and direction of N 2 O flux from marine sediments in response to nutrient enrichment.
It is important to note here that these results do not capture variability N 2 O fluxes over tidal or diel light-dark cycles.Daily fluctuations resulting from tidal flushing, temperature changes, and faunal activities impacts sediment biogeochemistry with likely implications for benthic N 2 O cycling (Glud, 2008).For example, previous work documented maximum sediment N 2 O effluxes during dark periods, potentially resulting from increased NH 4 + production and availability fueling N 2 O production via nitrification (Bauza et al., 2002).Variability in sediment OPDs driven by physical or biological processes can also affect N 2 O fluxes by modifying diffusion distances between the sediment interface and zones of N 2 O production or consumption (Rysgaard et al., 1994).In systems with low DIN availability and low N 2 O production potentials, shoaling of OPDs during dark periods may stimulate additional N 2 O consumption by bringing anoxic sediments in closer proximity to the water column (Berg et al., 2013).
The robustness of the mangrove N 2 O sink in the face of elevated DIN concentrations emphasizes the predominance of microbial N 2 O reduction over N 2 O production potentials from nitrification and denitrification.This is further supported by the considerable increase in N 2 O drawdown observed in N 2 O-amended sediments, implicating N 2 O as an important substrate fueling the anaerobic oxidation organic matter in carbon-rich environments.Enhanced sediment N 2 O consumption in undisturbed systems has been consistently attributed to environmental factors such as low DIN concentrations, high C:N ratios, and elevated inputs of allochthonous organic carbon (Erler et al., 2015;Maher et al., 2016;Murray et al., 2018).Indeed, the shallow OPDs reported in this study, along with the high rates of NH 4 + regeneration documented previously, suggest rapid organic matter remineralization (Hines & Lyons, 1982).Moreover, the ambient NO 3 concentrations measured in Bermuda surface waters are consistent with those observed in other mangrove N 2 O sinks, indicating limited substrate availability for denitrifying organisms (Maher et al., 2016;Sims et al., 2020).
These findings contrast with to those reported for NESAP continental margin sediments, where downslope increases in net N 2 O production and sediment efflux coincided with increasing bottom water NO 3 and decreasing bottom water O 2 concentrations (Jameson et al., 2021).Additionally, upper slope sediments exposed to simulated upwelling conditions demonstrated stimulated N 2 O effluxes, emphasizing the denitrifying community's rapid response to elevated NO 3 inputs.In contrast, sediment OPDs measured in Bermudian mangroves were within the range of those reported in NESAP sediments, and DIN amendments to mangrove sediments exceeded the maximum concentrations found in NESAP bottom waters (Capelle & Tortell, 2016).These discrepancies collectively highlight the role of biogeochemistry in structuring microbial communities and highlight the importance of microbial community dynamics in modulating responses to short-term environmental variability.
Comparing the abundance and expression of nosZI versus nosZII genes between Bermudian mangroves and the NESAP outer continental margin provides further insights the role of microbial community structure in regulating N 2 O flux from marine sediments.Normalized nosZII gene copy abundances were roughly an order of magnitude higher than that of nosZI variants across all samples, supporting previous molecular surveys demonstrating nosZII dominance in other marine microbiomes (Bertagnolli et al., 2020).However, NESAP sediments demonstrated higher nosZI gene copy ratios and significantly higher nosZI gene expression levels compared to the mangrove sediments.Relative expression levels of the nosZI variant increased along the depth gradient following increases in bottom water NO 3 concentrations and sediment N 2 O effluxes (Jameson et al., 2021).The typical nosZI variant is commonly associated with organisms that possess the upstream components required for complete denitrification including respiratory NO 3 and NO 2 reductases (Bertagnolli et al., 2020;Sanford et al., 2012).This indicates an increase in the relative contributions of denitrification to sediment N 2 O production along the depth gradient, and suggests that organisms possessing typical nosZI variants are favored in environments with high DIN availability.
Mean normalized nosZII gene and transcript copy abundances, on the other hand, were highest in mangrove sediments.Atypical nosZII variants accounted for over 90% of total nosZ genes and over 80% of total nosZ transcripts, supporting previous reports of correlations between microbial community N 2 O consumption potential and atypical nosZII in terrestrial soils and estuarine sediments (Hu et al., 2023;Jones et al., 2014;Xiang et al., 2023).The atypical nosZII variant is phylogenetically distinct from the typical nosZI variant found in canonical denitrifiers, demonstrates a higher affinity for N 2 O (Yoon et al., 2016), and has been found in the genomes of a wide array of non-denitrifying organisms (Bertagnolli et al., 2020;Sanford et al., 2012).Nitrogen limitation in minimally impacted mangrove systems has been previously linked to elevated rates from nitrogen conservation pathways such as dissimilatory nitrate reduction to ammonium (DNRA) (Fernandes et al., 2012).Interestingly, organisms that perform DNRA generally possess atypical nosZII variants and have been implicated as mediators of sediment N 2 O consumption through direct N 2 O consumption and competition with typical denitrifiers for nitrogen substrates (Maher et al., 2016;Sanford et al., 2012).
Mangrove bacterial communities reported in this study also revealed high abundances of ASVs from the Thiotrichales, an order that contains known sediment-associated taxa capable of DNRA coupled to sulfide oxidation (Otte et al., 1999;Preisler et al., 2007) and has been recently linked to high nosZII gene abundances and elevated N 2 O reduction potentials in sediments of the Pearl River Estuary (Xiang et al., 2023).Abundances of putative AOA and AOB were also comparatively low in mangrove sediments and N 2 O fluxes were unaffected by shortterm exposure to elevated NH 4 + concentrations over 8-10-hr time frames.This contrasts with data from the NESAP, where archeal communities dominated by putative AOA belonging to the Nitrosopumilales and Nitrososphaerales coincided with net N 2 O production across the oxycline and small NH 4 + effluxes (Alves et al., 2018;Belley et al., 2016;Jameson et al., 2021).
Given the high porewater NH 4 + concentrations and organic matter remineralization rates reported previously in Bermudian mangrove sediments, it is unlikely that the nitrifying community is substrate limited.It has been previously suggested that inhibitory compounds exuded by mangrove roots that restrict nitrifying processes may act as a chemical defense against nitrogen loss (Maher et al., 2016;Subbarao et al., 2009;Subrahmanyam et al., 1999).This may present an additional mechanism for mitigating N 2 O emissions by limiting N 2 O production by AOA near the sediment interface.However, it should be emphasized inferences regarding functional capacities based on 16S rRNA sequences are speculative and further work is needed to elucidate specific functional groups that drive variability in benthic N 2 O cycling.

Conclusions and Outlook
In contrast with open ocean OMZs, coastal sediments bring anoxic, reducing environments in close proximity to the atmosphere, allowing for appreciable atmospheric N 2 O drawdown in systems with low DIN concentrations.
Our results suggest that minimally impacted mangrove environments can drive net sediment N 2 O uptake mediated by a diverse community of atypical N 2 O scavengers, and that the N 2 O sink capacity of these communities is resilient against modest increases in DIN and seawater N 2 O concentrations.Whether or not these environments contribute meaningfully to global N 2 O budgets will require a complete inventory of the geographic extent of natural N 2 O sinks, including information about the seasonal and inter-annual variability in N 2 O consumption rates.Note also that these results represent the functional capacity of N 2 O-reducing microorganisms to compensate for short-term increases in sedimentary N 2 O production.Sustained chronic eutrophication of minimally impacted systems will also lead to structural shifts in sediment microbial community composition, and thus affect the balance between N 2 O production and consumption processes over longer timescales.These results provide further scientific justification for the conservation and restoration of coastal wetland ecosystems in areas with substantial anthropogenic influence.

Figure 1 .
Figure 1.Map of Bermuda.Map depicts the location of the Bermuda Institute of Ocean Sciences mesocosm facility (green circle) and the Ferry Reach mangrove stand used as a sediment sampling site (red circle).Sediment core sampling and incubation experiments took place between 28 October and 01 November 2021.

Figure 2 .
Figure2.Depth-stratified N 2 O production and consumption rates.N 2 O production and consumption rates (μmol-N 2 O cm 3 d 1 ) were modeled from mean porewater N 2 O concentration profiles in post-incubation cores.Solid red lines represent the modeled concentration profiles and individual points correspond to mean porewater N 2 O concentrations (nmol L 1 ± SD). Green and red boxes correspond to zones of N 2 O production and consumption, respectively.Solid and dashed horizontal lines denote the mean oxygen penetration depth ± SD, respectively.Gray shading highlights the section depth for sampling of nucleic acids.

Figure 3 .
Figure 3. Modeled sediment N 2 O fluxes.Open circles represent fluxes (μmol-N 2 O m 2 d 1 ) modeled from individual porewater concentration profiles in mangroves sediments.Control grouping contains pooled estimates across all DIN-amendment experiments.Black triangles represent fluxes modeled from the mean profiles shown in this figure.Fluxes N 2 Oamended cores were significantly greater (***) than control and DINamended cores ( p < 0.0001).Negative values correspond to net sediment N 2 O uptake.

Figure 4 .
Figure 4. Relative abundances of nosZII genes and transcripts.NosZII gene (a) and transcript (b) copies as the ratio of nosZII copies to the total sum of nosZ copies (nosZI + nosZII).Mangrove sediment samples were pooled across control, DINamended, and N 2 O-amended cores.Slope sediments were sampled from 200 m ambient and 200 m simulated upwelling, 475 m ambient, and 850 m ambient incubations.Initial samples were obtained in separate cores at each station prior to incubation and are indicated in orange.Asterisks denote significant differences ( p < 0.05* or p < 0.01**).

Figure 5 .
Figure 5. Sediment bacterial and archeal community structure.Community structure heatmap depicts proportional abundances (log 1o -transformed) of order-level taxonomic groupings for (a) the top 20 bacterial taxa and (b) all archeal taxa.Mangrove sediment samples representing sediment N 2 O sinks are highlighted in red on the left side of the figure.NESAP slope sediment samples are grouped left to right in order of increasing depth and N 2 O efflux.Mangrove sediment samples were pooled across control, DIN-amended, and N 2 O-amended cores.Slope sediments were sampled from 200 m ambient and 200 m simulated upwelling, 475 m ambient, and 850 m ambient incubations.Initial samples were obtained in separate cores at each station prior to incubation.
2 d 1 were detected in sediments contacting the OMZ core, driven by low bottom water O 2 concentrations, shallow OPDs, and a high supply of bottom water NO 3 .Building on this research, the present study reports on the N 2 O sink capacity of oligotrophic mangrove sediments under ambient conditions and modest increases in seawater N 2 O and DIN concentrations.Molecular data obtained from mangrove sediments were then contrasted with unpublished data from the Northeast Subarctic Pacific (NESAP) outer continental slope to better understand relationships between microbial community dynamics and N 2 O fluxes in two model systems with contrasting biogeochemical regimes and roles in benthic N 2 O cycling.

Gene Abundance and Activity Between N 2 O Sources and Sinks Logistical
(Jameson et al., 2021)replicate sampling of nucleic acids from mangrove sediments, except for the N 2 O-amendment experiments in which duplicate DNA and RNA samples were obtained.However, 16S-normalized nosZI and nosZII gene and transcript copy abundances were relatively consistent across all mangrove samples aside from a few statistical outliers (FigureS5in Supporting Information S1).To examine variability in microbial community dynamics between N 2 O source and sink sediments, we contrasted mangrove microbial communities with those surveyed in sediments from the NESAP continental margin.Corresponding flux estimates from NESAP sediments indicate a net source of N 2 O to the overlying water column(Jameson et al., 2021).Mean sediment N 2 O effluxes increased with slope depth and corresponding decreases in bottom water O 2 concentrations in the NESAP, ranging from 0.11 ± 0.03 μmol-N 2 O m 2 d 1 at 200 m depth to 0.69 ± 0.41 μmol-N 2 O m 2 d 1 at 850 m depth.Maximum sediment N 2 O effluxes were observed in outer shelf sediments (200 m) exposed to simulated upwelling conditions (0.81 ± 0.61 μmol-N 2 O m 2 d 1