Methanogens limited to lower rhizosphere and to an atypical salt marsh niche along a pristine intertidal mangrove continuum

Mangroves are valuable ecosystems that facilitate primary production, carbon sequestration, and regulation of greenhouse gas (GHG) cycles in coastal sediments, with microorganisms playing key roles. Specialized bacteria and archaea compete for energy and resources in mangrove sediments to inhabit optimal ecological niches and can produce or consume methane (CH4)—a potent GHG—in the process. CH4 cycling in mangroves has gained growing attention, yet uncertainties regarding functional and spatial distributions of microorganisms remain. Here, we demonstrate that in a pristine mangrove forest, CH4 concentrations and methanogen communities are concentrated within lower or below rhizosphere depths. We also reveal atypical niches for methanogens in the upper tidal salt marsh zone where vegetation is sparse and highest methanogens abundances were detected at deepest depths (4715 reads g−1) despite relatively high redox potentials (> 250 mV). Pore water CH4 concentrations were highest at the deepest depth within the mangrove forest (max. 3.40 ± 0.21 μM) and coincided with the highest sediment CH4 fluxes (276.4 ± 54.2 μmol m−2d−1) and methanotroph abundances at the surface (1309 reads g−1). Sediment CH4 oxidation fractions between the deepest (60 cm) and shallowest (5 cm) depths were estimated between 18.8% and 64.9%. Positive correlation between crab burrows and CH4 fluxes suggests that CH4 from deeper sediment and salt marsh niches can be transported via conduits to the atmosphere. The spatial data from this study highlights the importance of investigating CH4 dynamics across estuarine ecosystem gradients to better understand the complex roles of vital coastal vegetation zones in the face of a changing climate.

Coastal vegetation zones are highly productive ecosystems and provide fundamental habitats for a variety of nutrientcycling organisms (Sippo et al. 2018;Al-Haj and Fulweiler 2020).They play an important role in the global greenhouse gas (GHG) cycle, with mangrove forests alone fixing 218 AE 72 Tg C annually from the atmosphere via primary production (Bouillon et al. 2008).Organic carbon is buried in forest sediments where it provides energy to organisms in lower trophic levels, including microbial communities dwelling in the sediment and pore water (Feller et al. 2010;Santos et al. 2019).This "blue carbon" (refers to "C" sequestered in vegetated coastal ecosystems like mangroves) represents an important global sink of GHGs (Mcleod et al. 2011;Maher et al. 2018).However, methanogenic archaea, which metabolize stored organic carbon to generate methane (CH 4 ), may offset some of the fixed blue carbon through the production and emission of CH 4 in these ecosystems (Rosentreter et al. 2018;Al-Haj and Fulweiler 2020).Previous assessments of mangrove forests have calculated global annual CH 4 emissions of 0.23-0.25 Tmol, which may offset blue carbon burial by $ 20%, but these processes are still highly uncertain (Rosentreter et al. 2018;Santos et al. 2019;Al-Haj and Fulweiler 2020).Additionally, the disruption of coastal environments via anthropogenic land-use practices and rising global temperatures are placing a growing strain on coastal ecosystem cycles.These impacts may increasingly tip the scale toward higher methanogenic activity and GHG feedback effects (Li et al. 2019;Euler et al. 2020).
While coastal ecosystem-scale carbon and CH 4 budgets are being increasingly characterized, the microbial communities that carry out GHG cycling within different parts of coastal ecosystems are less well understood (Segarra et al. 2013;Brankovits et al. 2017).Archaeal diversity plays an important role in coastal CH 4 dynamics, but remains underexplored (Taketani et al. 2010;Tahon et al. 2021).Furthermore, competition between microbial communities in different layers of mangrove sediments can play a significant role in the proliferation and productivity of CH 4 cycling communities (Hibbing et al. 2010;Sela-Adler et al. 2017).In downward sediment profiles, a vertical transition from sulfate reduction to methanogenesis can often be observed concomitant with decreasing redox potential gradients.Sulfate reducers generally dominate metabolic activity at around À 220 mV with optimal conditions for methanogens typically found at the lower end of the thermodynamic energy landscape at redox potentials below À 300 mV (Lyu et al. 2018;Shima et al. 2020).Recent discoveries of syntrophic relationships between sulfate reducers, methanogens as well as methanotrophs, indicate complex interactions within the microbiome facilitating electron transfer and CH 4 cycling (Sela-Adler et al. 2017;Chen et al. 2019a).Community-level data of mangrove sediment dwelling microorganisms coupled to high resolution biogeochemical measurements is needed to further our understanding of these important coastal ecosystems.
Microbial communities of CH 4 cyclers can constitute bacteria, archaea, and fungi.Methanogenic prokaryotes mostly belong to the Euryarchaeota phylum and to date, entail seven known orders (Methanobacteriales, Methanocellales, Methanococcales, Methanomassiliicoccales, Methanomicrobiales, Methanosarcinales, and Methanopyrales) (Liu and Whitman 2008;Borrel et al. 2013).Further, methanogenic eukaryotes in the form of saprotrophic fungi have been found to contribute to microbial CH 4 generation (Lenhart et al. 2012).Methanotrophy can be carried out by phylogenetically diverse prokaryotes.The proteobacterial type I (Methylococcales order) and type II (Methylocystaceae family) methanotrophs are both able to metabolize CH 4 as their sole energy source (Bowman 2006).New types of aerobic, acidophilic methanotrophs within the Verrucomicrobia phylum have been recently discovered (specifically the Methylacidiphilales order) (Mohammadi et al. 2017).Anaerobic methanotrophic archaea (ANME) are closely related to methanogenic archaea (within the Euryarchaeota) and can oxidize CH 4 in deeper anoxic sediments and coastal sulfate-CH 4 transition zones (Beulig et al. 2019).
CH 4 produced in the deeper layers of mangrove sediments can be heterogeneously distributed across the ecosystem (Abril and Borges 2005;Maher et al. 2018).While previous studies have found that methanogenesis increases with depth along the vertical sediment profile, indicating abundant methanogen communities below the rhizosphere, studies at depths below 50 cm are lacking (Taketani et al. 2010;Bulseco et al. 2020).CH 4 may be partially oxidized by methanotrophs before reaching the sediment surface prior to emission to the atmosphere (Santos et al. 2019;do Carmo Linhares et al. 2021).Pneumatophores, tree stems, and prop roots, as well as crab burrows can act as conduits for transporting CH 4 from deeper sediments to the atmosphere, potentially providing pathways that bypass sediment CH 4 oxidation (Jeffrey et al. 2020a;Grow et al. 2022;Zhang et al. 2022).In estuarine mangrove forests, CH 4 can also be laterally transported to estuarine waters through "tidal pumping," where CH 4 is regularly flushed from the sediment layers via tidal pore water exchange, resulting in a net transfer of CH 4 from the mangrove forest to the estuary over time (Maher et al. 2018).To understand how local concentrations of CH 4 in sediments can be modified by lateral transport of aqueous CH 4 and methanotrophy at different locations in the ecosystem, an understanding of the ecological structure on both the macroscopic as well as the microscopic scale is necessary.
To test the hypothesis that microbial CH 4 production dynamics are dominated by lower rhizosphere processes in mangrove ecosystems, we evaluate microbial communities and biogeochemistry linked to the CH 4 cycle within sediments along a hydrological-topographic estuarine intertidal mangrove continuum, spanning lower-tidal estuarine sediments, mangroves and a hypersaline upper-tidal salt marsh.In order to do this, we characterized general prokaryotic diversity and specific abundances of methanogens, methanotrophs and sulfate reducers at numerous depths in each zone.The resulting microbial assemblages were correlated with ecological and biogeochemical characterizations including CH 4 and carbon dioxide (CO 2 ) concentrations and fluxes, δ 13 C-CH 4 isotope compositions, dissolved organic carbon (DOC) and physicochemical profiles (redox and salinity).Using this approach, we aim to explore the ecosystem scale co-factors governing CH 4 generation, transport, absorption, and ultimately emission into the atmosphere in an effort to further constrain microbiologically mediated CH 4 processes, which have the potential to partially offset carbon burial in the blue carbon paradigm.

Study site
The study site featured a pristine estuarine continuum in the Great Barrier Reef catchment area near the towns of Seventeen Seventy and Agnes Water in Queensland, Australia (Fig. 1).A large part of the surrounding catchment (23,000 ha) has been protected through the creation of the Eurimbula National Park.The study site was characterized by distinct vegetation and elevation zonation over relatively short distances.Over a 230 m transect, there was a transition from lower-tidal estuarine sediments and Avicennia sp.pneumatophores to densely vegetated mangrove forest and upper-tidal hypersaline sparsely vegetated salt marsh system (supporting information Fig. S1), which is a zonation common in some areas of Australia.The subtropical climate in this region receives a mean annual rainfall of 1142 mm and has an annual mean minimum and maximum temperature of 18.9 C and 25.9 C, respectively (www.bom.gov.au).
The estuary site (Site 1) of the transect (Fig. 1) consisted of poorly consolidated sandy sediments.The transition from the sandy estuary to the central mangrove forest at the pneumatophores site (Site 2) consisted of only pneumatophores and crab burrows, but no mangrove tree stems.The dense mangrove site (Site 3) was dominated by mangroves of the genus Rhizosphora and featured organic-rich silty/clay marine sediments, rich in crab holes and root systems.Approximately 170 m from the estuary, the mangrove forest rapidly decreased in density and gave way to the sparse mangrove site (Site 4) with increasing porewater salinity and coarser, more sandy sediments.The salt marsh site (Site 5) was characterized by a steep rise in porewater salinity, a eukaryotic algal mat on top of sandy sediments and sparse vegetation.Moving further inland, the end of the intertidal estuarine continuum was marked by the prevalence of upland forests.Images of each site can be found in the supporting information Fig. S1.

Sample collection
Sampling took place between the 5th and the 11th of August 2019 in the subtropical winter, with dry weather and only one minor rainfall event on the 7th of August (< 5 mm).The mangrove rhizosphere depth was used to spatially constrain vertical sampling depths and assumed to be around 50-60 cm, with below-ground root density typically continuously falling off below that depth (Kristensen et al. 2005).For microbiological analysis, discrete sediment samples were collected using sterile techniques from three profile excavation depths (5, 30, and 60 cm at Sites 2-5 and 5, 20, 40 cm at Site 1 where the water table limited the depth profile) per site by homogenizing about 100 g of sediment, extracting a small amount (ranging from 0.23 to 0.51 g), and then storing samples on ice during field work.For each microbiological sampling depth, redox measurements were recorded using a portable Eh meter and probe (Hach Ag/AgCl, KCl, 3 mol L À1 ) directly inserted into the sediments of an undisturbed core.Eh meter readings were corrected to report redox potentials relative to standard hydrogen electrode.Porewater for DOC and GHG analysis was collected from a vertical depth profile at each site (5, 30, and 60 cm at Sites 2-5 and 5, 20, and 40 cm at Site 1 where the estuary's water table limited the depth profile) utilizing a 1 m push point piezometer (Solinst) connected to airtight tubing and syringes, with filtration via 0.7 μm glass microfiber filters directly into borosilicate vials amended with saturated H 3 PO 4 to inhibit further microbial processing.
CH 4 sediment fluxes were measured in situ using sediment chambers (8-26 cm diameter) connected to a closed-loop system with a Cavity Ring-Down Spectrometer (CRDS, Picarro Gas Scouter, G4301) similar to the approach used by Jeffrey et al. (2020a).A total of 10 sediment flux measurements were collected within each of the five sites.Smaller volume chambers were used at Sites 4 and 5 featuring lowest soil fluxes.Each chamber was first gently placed 1-2 cm into the soil surface, immediately removed and flushed to minimize potential artificial soil CH 4 release, then carefully re-placed back into the same location for each flux measurement.Any measurements featuring non-linear trends were disregarded, low CH 4 fluxing incubation regressions were inspected visually and further confirmed by comparing the simultaneous CO 2 fluxes.All flux estimates were above our equipment detection limit (i.e., >3 μmol m À2 d À1 ) according to equations by Wassmann et al. (2018).The flux rate (ppm s À1 ) was extracted in R studio using a modified "GasFlux" package of Fuss (2019) as described in Jeffrey et al. (2020b).The incubation durations and chamber details are reported in supporting information Table S7.Crab hole diameters and pneumatophore counts were recorded for each chamber sampling area directly within the sediment flux chambers.
Individual microbial samples were extracted from the homogenized sediment at each of the five sites and three depths per site by hand using fresh sterile examination gloves to avoid cross-contamination.Microbial sampling material was stored in sterile 1.5 mL polycarbonate vials (Eppendorf) prepared with DNAgard (Sigma-Aldrich) and kept in the fridge for a week before sending to the Australian Genome Research Facility for processing.DOC porewater samples of 40 mL volume were filtered and stored in 40 mL borosilicate vials amended with saturated H 3 PO 4 (100 μL).GHG porewater sample volumes ranged from 4.9 to 6 mL and were directly transferred into airtight 10 mL borosilicate vials amended with saturated H 3 PO 4 (100 μL) and closed off with airtight rubber seals.Duplicates were taken for each of these samples.

Sample processing
Microbial DNA from sediment samples was extracted using a DNeasy PowerLyzer PowerSoil Kit (Qiagen).PCR amplification and sequencing were performed at the Australian Genome Research Facility using the primers and conditions outlined in supporting information Table S2.The primer pair 341F and 806R was used for the detection of prokaryotes (Takahashi et al. 2014;Apprill et al. 2015;Wasimuddin et al. 2020).Thermocycling was carried out with an Applied Biosystem 384 Veriti, using Platinum SuperFi mastermix (Life Technologies, Australia) for the primary PCR.The amplicons were quantified with a dsDNA fluorometry assay (Promega Quantifluor) after two rounds of PCR.Sequencing was then completed on an Illumina MiSeq (San Diego, CA, USA) with a V3, 600 cycle kit (2 Â 300 base pairs paired-end) adhering to the Illumina 16S metagenomics sequencing protocol.
DOC concentrations were determined using a total organic carbon analyzer coupled with an isotope ratio mass spectrometer (Thermo Fisher, Delta-V Plus) and a continuous flow system (Thermo Fisher, ConFLo) with a precision of 0.02 mM (Lalonde et al. 2014).Porewater dissolved GHG samples were analyzed using an isotopic CRDS gas analyzer (Picarro, G2201-i) with a small sample introduction module 2 (SSIM, Picarro, A0314) attached.A sterile syringe was used to extract 4 mL of gas from the equilibrated headspaces of the stored sampling vials by injecting 4 mL of H 2 O into the vial, pushing out the gas in the process.The injected gas samples were mixed with zero air (Coregas) at a ratio of 4 : 1 within the SSIM to reach a total volume of 20 mL.The CRDS was then used to determine CH 4 and CO 2 concentrations and δ 13 C-CH 4 (‰) for each sample.Due to the low number of CH 4 molecules in the spectrometer's cavity, δ 13 C-CH 4 (‰) measurements at low CH 4 concentrations (< 12 ppm) yielded higher statistical errors despite using the CRDS's high precision mode (Uhlig and Loose 2017).Vials were weighed before the analysis and then emptied and dried for post-analysis weighing and determination of individual sample volumes.This allowed for the calculation of partial pressures of the GHGs within the original headspace of the vials.

Data analysis
After assembling paired-end reads by aligning the forward and reverse reads, primers were identified and contigs were created in Mothur (Schloss et al. 2009).Trimmed sequences were cleaned and matched to the 16S region of interest.Sequences were quality filtered and full-length duplicate sequences were removed.Sequences were clustered and filtered for mismatches of at least 1 in every 100 bases to eliminate sequencing errors.This was followed by chimera filtering using the vsearch algorithm (Rognes et al. 2016).Taxonomy was assigned in Mothur using the Silva database (Yilmaz et al. 2014).Any 16S rRNA fragments from eukaryotes, chloroplasts and mitochondria that survived previous cleaning steps were removed from the dataset.To evaluate the number of reads per operational taxonomic unit (OTU), reads were mapped back to OTUs with a minimum identity of 97% (Schloss 2013).
Microbiological community distributions for each site and depth were visualized at phylum level using MEGAN software (MEtaGenome Analyzer) (Huson et al. 2016).The number of reads were normalized to 1 g of wet sample mass prior to the analysis of microbial community abundances.Functional groups of methanogens, methanotrophs, and sulfate reducers were plotted using bokeh in Python software (Zulkower and Rosser 2020).Heatmaps were generated using seaborn in Python.Principal coordinate analysis was carried out for microbiological assemblages on family level with MEGAN software using Bray-Curtis dissimilarity to quantify the compositional dissimilarity between two microbial communities detected in the samples (Roberts 2017).Shannon-Weaver diversity indices were calculated on family level with MEGAN software (Magurran and McGill 2011).
GHG partial pressures were calculated using the equations described by Magen et al. (2014).The statistical analysis and visualization of individual plots for physiochemical parameters, δ 13 C, GHG and DOC concentrations was carried out using Gnuplot software and matplotlib in Python.Errors were calculated using error propagation for relative uncertainties via statistical errors (standard deviation [SD]; standard error [SE]) and systematic errors (sensitivity of small sampler CRDS, isotope ratio mass spectrometers and sondes).For duplicate samples, SD was calculated for each set of the duplicates and then pooled for the total SD of the sampling location.Our linear regression analyses used Pearson's product moment correlation coefficient and Spearman's rho (where indicated) with an R 2 cut off value of 0.3 (i.e., statistical correlations with R 2 < 0.3 were not considered significant).
We estimated mangrove sediment CH 4 oxidation rates by calculating the fractionation factor α as a function of porewater temperature T ( C), as determined by Chanton and Liptay (2000) for "clay soils" using: We determined the oxidized fraction of CH 4 (F ox ), where δ initial (‰) represents the average δ 13 C-CH 4 isotopic value of the porewater from the deepest sample depth and δ final (‰) represents the δ 13 C-CH 4 isotopic value of the porewater from the shallowest depth assuming a closed-system model at nonsteady state from Liptay et al. (1998): and open system models assuming steady state using Tyler et al. (1997): and Happell et al. (1994):

Results
Microbial ecology and physicochemical parameters Prokaryotic community distributions varied notably along the estuarine continuum (Fig. 2).In the estuary site (Site 1) surface samples, Cyanobacteria were the most abundant phylum, but decreased in abundance further up the intertidal transect, as well as with increasing depths toward the upper tidal site (Site 5-salt marsh), communities were increasingly dominated by Chloroflexi, with samples taken from the deepest depth entailing 47,214 reads g À1 with Dehalococcoidales (26,255 reads g À1 ) being the most abundant order.The Proteobacteria phylum was abundant at all sites, with the Deltaproteobacteria class and numerous sulfate reducing OTUs increasing in prevalence at $ 30 cm depths of Mangrove Sites 2-4 (maximum in the dense mangrove site at 30 cm depth, with 93,138 reads g À1 ).Euryarchaeota were most abundant in deepest depth sites in the Mangrove sites (Sites 2-4) and in the salt marsh site (Site 5).Correspondingly, six of the seven known orders of methanogenic archaea (Methanobacteriales, Methanocellales, Methanococcales, Methanomassiliicoccales, Methanomicrobiales, and Methanosarcinales) were detected in the samples.Shannonweaver indices, representing prokaryotic diversity, ranged from a minimum of 3.16 and 3.32 in the salt marsh (Site 5; 5 cm and 60 cm depth, respectively) to a maximum of 5.18 and 5.01 in the sparse and dense mangrove surface samples, respectively (Sites 4 and 5).The index stayed within 3.90 AE 0.40 for all other samples (supporting information Fig. S3).
The porewater salinity along the transect ranged from 35.3 to 50.1 PSU in samples from Sites 1 to 4 (Fig. 3).The lowest salinity was recorded at 60 cm depth in the dense mangrove site (Site 3).The salt marsh site (Site 5) was hypersaline, with values ranging from 88.5 to 107.8 PSU.Redox potentials were depth-and zone-dependent, ranging from À 140 to 309.4 mV (Fig. 3).The lowest redox potential values were observed at the deepest depth in the pneumatophore site (Site 2) and dense mangrove site (Site 3) with À 32 and À 140 mV, respectively.The salt marsh site (Site 5) had the highest redox potentials at all depths (289.4-375.0mV).

GHGs, isotopes, and carbon
CH 4 concentrations in porewater varied widely along the intertidal continuum with concentrations ranging from 0.06 AE 0.01 μM to 3.40 AE 0.17 μM (Fig. 4).Maximum CH 4 concentrations were observed in the dense and sparse mangrove sites (Sites 3 and 4; 3.40 AE 0.17 μM and 2.02 AE 0.10 μM, respectively).CH 4 concentrations were also relatively high in the estuary site, where they ranged from 0.96 AE 0.05 μM to 1.66 AE 0.08 μM.At the salt marsh site (Site 5), CH 4 porewater concentrations were low at all depths (between 0.06 AE 0.01 μM and 0.07 AE 0.01 μM).CO 2 porewater concentrations were highest in the sparse mangrove site (Site 4; ranging from 2.99 AE 0.15 mM to 3.44 AE 0.17 mM).Concentrations gradually decreased toward lower intertidal areas and the pneumatophore site (Site 2; 1.20 AE 0.06 mM at the surface) before a slight increase in the estuary site (2.46 AE 0.12 mM at 20 cm depth).At the salt marsh site, low CO 2 porewater concentrations were observed at all depths (< 0.5 mM; Fig. 4).CH 4 porewater concentrations showed a trend of increasing concentrations with increasing sampling depths at all sites apart from the salt marsh site (Site 5) (Fig. 6a).The linear regression was fitted to Sites 1-4 data (R 2 = 0.347, p = 0.024; see Fig. 6).
DOC concentrations ranged from 0.23 AE 0.02 mM to 1.09 AE 0.02 mM along the transect (Fig. 4).DOC was highest in the salt marsh (Site 5), reaching 1.09 AE 0.02 mM and 1.05 AE 0.02 mM at the medium and deepest depth, respectively.From the salt marsh site, DOC concentrations decreased gradually toward the estuary with the exception of the estuary site (Site 1) surface samples (0.44 AE 0.02 mM).In the dense mangrove site (Site 3), DOC concentrations increased slightly toward lower depths (0.52 AE 0.02 mM at 60 cm depth).
δ 13 C-CH 4 values ranged from À 15.7 AE 8.2‰ at the deepest depth of the salt marsh site (Site 5) to À 81.1 AE 0.3‰ at the deepest depth of the sparse mangrove site (Site 4) (Fig. 4d).δ 13 C-CH 4 -enrichment was observed from the deepest to shallowest depths at the mangrove forest sites (Sites 2-4).
Here, the highest depth-dependent change in δ 13 C-CH 4 values was measured at the sparse mangrove site (Site 4) with À 81.1 AE 0.3‰ at the deepest depth and À 55.7 AE 4.1‰ at the surface.δ 13 C-CH 4 values were relatively enriched at all depths of the salt marsh site (Site 5) with the highest δ 13 C-CH 4 -enrichment measured at the surface (À 20.9 AE 3.5‰).The evaluation of oxidation fractions revealed that between 18.8% and 64.9% of CH 4 was being oxidized between deeper and shallower depths at Sites 2-4 (supporting information Table S6), using the Liptay et al. (1998) open system model and variable α values (see α 1 and α 2 in supporting information Table S6) (Thottathil et al. 2022).
A total of 47 sediment flux measurements were used in the interpretation of the linear CH 4 fluxes (3 sediment flux measurements were excluded due to nonlinear trends).CH 4 fluxes from the sediment varied across sites and were highest in the dense mangrove site (Site 3; 276.4 AE 54.2 μmol m À2 d À1 ; Fig. 5).At the sparse mangrove site (Site 4) and salt marsh site (Site 5), lower fluxes were observed (9.5 AE 2.3 μmol m À2 d À1 and 9.3 AE 0.5 μmol m À2 d À1 , respectively).When present (Sites 2 and 3), crab holes and pneumatophore counts were measured in the same area of the sediment flux chambers to allow for correlation of the data points.The average sum of crab hole surface area was 80.2 AE 25.3 cm 2 per m 2 (n = 6).Mangrove pneumatophores were abundant at the sites and averaged 326.5 AE 79.3 per m 2 (n = 6).Minimal crab holes and pneumatophores were observed at Sites 1 and 4 and none were observed at Site 5. Linear regressions revealed a significant positive correlation between CH 4 sediment fluxes and crab hole surface area (R 2 = 0.962, p < 0.01) at Sites 2 and 3, as well as a significant positive correlation between CH 4 sediment fluxes and pneumatophore count (R 2 = 0.556, p = 0.044) (Fig. 6b).

Environmental OTUs
Methanogenic archaea abundance varied along the ecosystem continuum and increased toward deeper depths in the mangrove-dominated sites (Sites 2-4; Fig. 7).Surprisingly, the salt marsh site (Site 5) also had considerable abundances of methanogens at both medium and deeper depths (3001 and 4715 reads g À1 , respectively), despite having relatively high redox potentials and low CH 4 concentrations in the porewater (Fig. 7).In the mangrove dominated sites (Sites 2-4), the highest abundance was reached at the deepest sampling depths beneath the pneumatophores (Site 2) and dense mangrove site (Site 3) with 3026 and 3135 reads g À1 , respectively.Across all samples, the relative methanogen abundances of the total prokaryotic communities ranged from 0% to 2.9%.The most abundant methanogenic archaea order was the Methanomassiliicoccales, representing an average of 54.4 AE 9.3%.A slight taxonomic shift toward the Methanobacteriales in the sparse mangrove site (Site 4) and salt marsh site (Site 5) in the upper tidal areas was observed (representing 24.5 AE 11.5% of total methanogenic archaea communities on average at Sites 4 and 5; Fig. 7).There was a significant negative correlation between methanogen abundances and redox potentials for Sites 1-4 (supporting information Fig. S4; R s 2 = 0.624; p < 0.01; Spearman's rho statistics used).The methanogen niche at the Site 5 is noted as atypical and does not fit in the normal redox range.Conversely, the principal coordinate analysis of microbiological data on family level is showing separate clustering for samples collected at Site 5 (Fig. 8).Sulfate reducer communities decreased in abundance at deepest depths and reveal peak values at medium sample depth within the dense mangrove site (Site 3; 70,728 reads g À1 ).At the sparse mangrove site (Site 4) and salt marsh site (Site 5), sulfate reducers occurred less abundantly, ranging from 44 to 8558 reads g À1 .Overall, sulfate reducer communities exhibited high relative abundances within the total community distributions at the family level, ranging from 3.2% to 23.6% in Sites 1-3.Notably, the Desulfobacteraceae family made up an average of 75.1 AE 1.2% of Sites 1-3 sulfate reducer communities.Methanotroph abundance was highest (1309 reads g À1 ) in the dense mangrove site (Site 3) surface samples where CH 4 sediment fluxes were also highest (Fig. 5), with the Methylocystaceae family alone representing 1094 reads g À1 .In all other samples, methanotroph abundances ranged from 0 to 310 reads g À1 and at the shallowest depth at Site 3, where aerobic conditions were prevalent, represented only 0.5% of the total community.The principal coordinate analysis also shows clustering of aerobic surface samples from the mangrove sites (Sites 2-4; Fig. 8).Methanotroph communities were exclusively bacterial and no ANME were detected across all sites or depths.

Variability and depth-dependence of microbial communities and GHGs
Prokaryotic community distributions and GHG concentrations varied considerably between zones and depths along the estuarine intertidal continuum.Methanogens were most abundant in the deeper sediments of the lower tidal mangrove sites (Sites 2 and 3), and coincided with low redox potentials, high CH 4 concentrations and elevated CH 4 sediment fluxes (Fig. 9).There, methanogens find ideal conditions for CH 4 production and face lower competition with microorganisms that prefer energetically more favorable electron acceptors and redox potentials (Hibbing et al. 2010;Lyu et al. 2018).While it has been suggested that the deep burial of organic blue carbon may produce CH 4 emissions from vegetated ecosystems, the microbially mediated CH 4 flux and vertical shifts in methanogenic archaea community distributions remain underexplored (Mcleod et al. 2011;Rosentreter et al. 2018).This study shows a methanogenesis niche that appears to be limited to deeper mangrove sediment depths, with a relative abundance of methanogenic archaea of up to 2.1% at the mangrove-dominated sites (Sites 2-4) to a depth of 60 cm (Fig. 7).
Sulfate reducing bacteria were most abundant at medium depths within the rhizosphere of the lower intertidal mangrove sites (Sites 2 and 3; Fig. 7c).The highest abundance of sulfate reducers (Site 3; Fig. 7c) coincided with a redox potential of À 4.8 mV (30 cm depth).Sulfate reducing bacteria thrive at higher redox potentials than methanogens and have a relatively high affinity for hydrogen and acetate, outcompeting methanogens for these common electron donors where communities co-exist (Purdy et al. 2003;Sela-Adler et al. 2017).In the mangrove sites studied here, the location of the sulfate methane transition zone appeared to start at the lower end of the rhizosphere (50-60 cm depth) where sulfate reducers and methanogens still readily co-exist but conditions are increasingly beneficial for methanogenic communities (Sela-Adler et al. 2017).Previous studies have shown the coexistence and competition between sulfate reducers and methanogens, which can facilitate shifts to methanogenesis at deeper depths in healthy coastal mangrove forests (Taketani et al. 2010;Sela-Adler et al. 2017).
Our findings suggest that the accumulation of CH 4 is confined to deeper rhizosphere depths in the dense mangrove site (Site 3) and partially within the sparse mangrove site (Site 4) (see Fig. 9).Positive correlations between porewater CH 4 concentrations and depths (Fig. 6a), as well as negative correlations between methanogen abundances and redox potentials along the vertical depth gradient (Sites 1-4; Figs.7a, 9) support this spatial distribution model of our study sites.Notably, our study lacks data below 60 cm, where more expansive CH 4 accumulation and methanogenesis niches may occur.It has been suggested that a significant fraction of CH 4 produced in mangrove-dominated ecosystems is transported to the sulfate reduction zone from deeper sediment (Chuang et al. 2016) and previous studies have shown methanogenesis increasing along the vertical depth profile, but studies at deeper depths are lacking (Taketani et al. 2010;Bulseco et al. 2020).A multiparameter study with a focus on spatial measurements in below-rhizosphere depths could shed further light on uncertainties regarding methanogenesis niches and CH 4 accumulation in deeper mangrove forest sediments.

Unique ecosystem niches along the estuarine intertidal continuum
In contrast to the mangrove sites (Sites 2-4), the salt marsh site (Site 5) featured more extreme physicochemical conditions (i.e., hypersaline), sparse vegetation, and a stark shift in prokaryotic communities.Bray-Curtis dissimilarities between sample clusters display community compositions (family level) in the salt marsh site as different to other sites at all depths, with samples from the mangrove sites (Sites 2-4) revealing a more depth-dependent clustering (Fig. 8).Notably, the sediments of the salt marsh site revealed atypical niches for methanogenic archaea (Fig. 9) despite hypersaline conditions and high redox potentials.The abundance of a number of methanogenic archaea orders (Fig. 7) was an unusual find in both medium and deeper depths.While the Euryarchaeota phylum has been detected in hypersaline conditions, most of its members belonged to the order Holobacteriales (Hollister et al. 2010;Vera-Gargallo et al. 2019).Previous studies have shown the role of seasonality in salt marsh carbon and CH 4 dynamics, with drier and more saline conditions linked to a reduction in both CH 4 concentrations and methanogenesis (Wilson et al. 2018;Zhu et al. 2021).The scope of our study was limited by only measuring during the dry subtropical winter, which suggests that further research focusing on microbial CH4 dynamics in different seasons is needed.
High DOC and low CO 2 concentrations in the salt marsh site (Site 5) suggest lower microbial activity in this challenging environment.As DOC is abundant here (potentially supplied by the surface algal mat; Fig. 9), this suggests limited DOC uptake by other microorganisms, which may therefore favor the prevalent methanogens observed at medium and deeper depths, by providing a substantial nutrient source (Torres-Alvarado et al. 2013;Lyu et al. 2018).This could also give rise to the low prokaryotic diversity found at the salt marsh site (supporting information Fig. S2).Future studies could further address this by including loss of ignition and/or additional sedimentary organic carbon content measurements in mangrove continuums.A lack of CO 2 in the porewater of the site indicates low microbial respiration taking place, with some CO 2 potentially being utilized by methanogens and/or possibly laterally transported to the adjacent sparse mangrove site.Relatively high redox potentials at all depths of the salt marsh site (Fig. 3) emphasize this shift toward algal primary production and aerobic processes.However, previous studies have shown that some methanogens can produce CH 4 in environments with redox potentials of up to + 420 mV, despite optimal hydrogenotrophic methanogenesis conditions typically found around À 340 mV (Fetzer and Conrad 1993;Shima et al. 2020).In addition, bulk sediment Eh measurements may not reflect the potential for the existence of sediment microniches with lower redox potential more favorable for methanogens.
The atypical findings in the salt marsh site, where relatively high methanogen abundances coincide with low CH 4 porewater concentrations and sediment fluxes, give rise to a number of hypotheses.First, CH 4 could be laterally transported via tidal pumping and/or groundwater flow as well as transported and outgassed via crab burrows and forest prop roots in the nearby fringing upper tidal mangrove forest acting as gas conduits (further discussed in section 3.3).Second, microorganisms could simply be less active overall in this extreme environment due to physicochemical constraints such as high salinity and low DO.Seasonality could also play a significant role in this, with methanogens being more active in the warmer, wetter summer months.In order to test this hypothesis, microbiological analyses including discrete rate measurements and the expression of methanogenesis related RNA with different "omics"-approaches like transcriptomics could be employed (Cai et al. 2022).Lastly, methanogenesis niches within the continuum are likely to expand into even deeper depths and could reveal larger relative abundances of methanogens and CH 4 at the mangrove forest sites.
A prominent algal mat (supporting information Fig. S5) was also observed on the surface of the salt marsh site.As no notable distribution of Cyanobacteria was present in the salt marsh site's surface samples, we suggest that the algal mat is mostly eukaryotic in nature.Different algae species have been shown to be more abundant in the colder subtropical months and can fix CO 2 to offset emissions (Chen et al. 2019b).Algae can be large contributors to marine blue carbon concentrations and have been shown to make up $ 33% of the eDNA spool in respective habitats (Ortega et al. 2020;Brown et al. 2021).Eukaryotic saprotrophic fungi have also been reported to produce CH 4 and could contribute to GHG budgets in ecosystems like intertidal mangrove forests (Lenhart et al. 2012).This could potentially give rise to additional GHG sinks and sources and could explain discrepancies in prokaryotic communities within the atypical niches of the site.Subsequent limnological studies in estuarine intertidal continuums could overcome these uncertainties by including additional genetic targets into their analysis pipeline, accounting for relevant eukaryotic organisms.

Transportation, consumption, and outgassing of CH 4
Our study revealed a healthy mangrove forest situated in an intertidal continuum with complex below-ground root systems and crab burrows.These below-ground structures likely acted as conduits for GHGs which is demonstrated by the significant correlation between CH 4 sediment fluxes and average crab burrow diameters and pneumatophore counts in our sampling areas at Sites 2 and 3 (Fig. 6b).Estuarine intertidal continuums have been characterized as dynamic systems with a multitude of lateral and vertical processes distributing compounds across different ecological zones (Maher et al. 2018;Al-Haj and Fulweiler 2020;Jeffrey et al. 2020a).Tree stems and roots of mangroves and various wetland tree species, as well as crab burrows have previously been shown to act as conduits for the vertical transport of CH 4 to the atmosphere (Jeffrey et al. 2020a;Grow et al. 2022;Zhang et al. 2022).Lateral transport can also deliver dissolved carbon in coastal ecosystems through processes such as tidal pumping, porewater exchange and groundwater flow, which has been shown to partially offset the carbon burial in these ecosystems (Tait et al. 2016;Maher et al. 2018;Santos et al. 2019).Porewater CH 4 and CO 2 may be subject to some degree of lateral translocation from the salt marsh site and upper tidal area (Sites 4 and 5) toward the lower intertidal forest sites (Sites 2 and 3), where they may accumulate (e.g., the low rhizosphere CH 4 observed at Site 3) and contribute to elevated sediment fluxes (see Fig. 9).However, the potential for and magnitude of any lateral porewater translocation at this site requires further investigation.
At locations in the ecosystem where CH 4 is outgassed, methanotrophs can thrive and partly mitigate emissions.In this study, methanotrophs were most abundant in the dense mangrove site (Fig. 9) where CH 4 concentrations were high in the deep sediment and a high pneumatophore density was recorded at the surface (Fig. 6b).Correspondingly, the methanotroph abundance at this site coincided with the highest CH 4 sediment fluxes, showing the spatial affinity of aerobic methanotrophs for CH 4 near the top layer of the sediment where oxygen is most abundant.Consumption of CH 4 via microbially driven methanotrophy has previously been shown to take place in various ecosystem locations of other forest types, including deeper anoxic sediment layers, oxic surface layers and in microhabitats around the tree roots, stems and bark, although this remains poorly understood (Jeffrey et al. 2019(Jeffrey et al. , 2021;;do Carmo Linhares et al. 2021).Competition in deeper mangrove sediments, as well as extreme conditions found in the salt marsh site could explain the discrete methanotrophy niche on the mangrove forest floor.The most prevalent methanotrophs in our samples belong to the family Methylocystaceae (Fig. 7), which are widespread across tropical and subtropical ecosystems and versatile in adapting to different conditions (Finn et al. 2020;Cheng et al. 2022).
The vertical transport and oxidation of CH 4 within the estuarine continuum is further illuminated by δ 13 C analysis.Previous studies show that δ 13 C-CH 4 fractionation can reveal the oxidation pathways of CH 4 in the sediment (Fry 2006;Uhlig and Loose 2017).In our data, the greatest vertical δ 13 C-CH 4 enrichments occur within the mangrove forest sites (Sites 2-4; Fig. 4d), and at the same locations where methanotrophs were most abundant near the sediment surface (Fig. 7).At these sites, the calculated oxidation fraction suggests that between 18.8% and 59.0% of CH 4 was oxidized between deeper to shallower depths (see supporting information Table S6; it is noted that the oxidation fraction can vary depending on the temperature and temperature calculated constant α).Therefore, δ 13 C suggests that the methanotrophic microorganisms are actively mitigating CH 4 within the mangrove forest rhizosphere.At the salt marsh site (Site 5), δ 13 C-CH 4 is significantly more enriched than at the mangrovedominated sites, thus pointing to differences in the organic carbon sources (e.g., from the algal mat).Oxidation up the profile did not readily occur at the salt marsh site (Site 5) and estuary site (Site 1).Notably, although the δ 13 C-CH 4 and oxidation fraction data may be limited in significance due to high statistical errors for some of the low CH 4 samples (see SD bars in Fig. 4), it accords with the CH 4 transport pathways elucidated earlier.
Overall, this study revealed a highly variable estuarine intertidal continuum with distinct ecological niches for microorganisms linked to GHG cycling.In the hypersaline salt marsh, where vegetation was sparse, an atypical and shallower depth methanogenesis niche was detected alongside abundant DOC; yet, there was little apparent CH 4 accumulation.The discrepancy between high relative abundances of methanogenic archaea communities and low porewater CH 4 concentrations at this site suggests the possibility of either low overall activity of the methanogen communities at the time of sampling due to seasonality and physicochemistry; CH 4 translocation elsewhere; and/or limited data from deeper sediment skewing relative abundances.The findings also underpin the importance of coastal vegetation such as mangrove forests (Sippo et al. 2018;Arai et al. 2021).Here, we show that these forests are confining methanogenesis to deeper layers within and potentially below the rhizosphere, likely via vertical nutrient distributions and competition with other microorganisms, such as an abundance of sulfate reducers within the rhizosphere.The forest ecosystem also facilitates the microbially mediated offset of CH 4 emissions via upper sediment CH 4 oxidation.Uncertainties regarding lateral and vertical transport of GHGs, seasonal differences in fluxes and communities, as well as below rhizosphere deep sediment CH 4 dynamics remain.Our study further demonstrates the diversity of niches CH 4 cycling microorganisms can assume in complex vegetated ecosystems and shows the importance of high-resolution spatial measurements employing metagenomics together with GHG measurements and other approaches around coastal ecosystems, establishing that further research into these processes is needed to facilitate the development of remediation and mitigation measures in the face of accelerating climate change.

Fig. 1 .
Fig. 1.Map of the estuarine intertidal continuum study site situated between Agnes Water and Seventeen Seventy, Australia.

Fig. 2 .
Fig. 2. Microbial and ecological parameters depicted as: (a) Community distribution bar plot showing relative abundances of prokaryotic phyla; ecosystem continuum indicated on top; legend of phyla on the right; phyla with an abundance of < 0.1% were discarded; (b) Total number of reads per site normalized to 1 g (w/w) of sampling material.

Fig. 3 .
Fig. 3. Physicochemical parameters plot of: (a) redox potential, and (b) Salinity for all sites along the intertidal continuum at depths between 5 and 60 cm.

Fig. 6 .
Fig. 6.Correlation plots showing: (a) CH 4 porewater concentrations per depths for all sites and (b) CH 4 sediment fluxes, per crabhole surface area (x-axis scale displayed on the bottom) and pneumatophore count (x-axis scale displayed on the top) at Sites 2 and 3 within the measurement area of the CH 4 sediment flux chamber (26 cm diameter).

Fig. 7 .
Fig. 7. Normalized abundances of individual prokaryotic communities along the intertidal continuum for all study sites and depths.Functional groups of (a) methanogens, (b) methanotrophs, and (c) sulfate reducers were combined in the stacked bar plots, with individual OTUs denoted below the graphs.

Fig. 8 .
Fig. 8. Principal coordinate analysis of microbiological data on family level, using Bray-Curtis dissimilarity.Sites and depths are depicted via colors and shapes as denoted in the box on the top right.Sample clusters have been circled and labeled, respectively.Bacterial and archaeal abundance maxima with the highest impact on clustering have been depicted in the supporting information Fig. S8.

Fig. 9 .
Fig. 9. Ecosystem heatmap of relative prokaryotic abundances and CH 4 , CO 2 , and DOC concentrations at the different sites and depths.Relative values are shown (see scale on the right side) with the sum normalized to 1 for each of the six parameters.