The two Monterey Bay cold seep sites investigated in this study were Clam Field and Extrovert Cliff (Fig. 1a). The Clam Field site is located on a sedimented apron of the Monterey Canyon wall and is constituted by a broad band of seepage parallel to the canyon. Dense fields of live clams are found across the main band of seepage, along with patchy microbial mats (Fig. 1b–c). Clam Field is located within the Monterey Bay fault zone (Orange et al., 1999), suggesting that fluid seepage there is tectonically influenced and likely driven by artesian flow through the highly fractured, hydrocarbon-rich shale of the Monterey Formation (Barry et al., 1996; Lorenson et al., 2002; LaBonte et al., 2007; Füri et al., 2009). Previously measured δ13C values of methane at this site (-50 ‰ to -55 ‰) are indicative of a high degree of thermogenic input – more so than at many other Monterey Bay seeps (Lorenson et al., 2002). The Extrovert Cliff site is located on the slope of a slide scar between the San Gregorio and Monterey Bay fault zones. The site is characterized by distinct, concentric rings of seepage covered by thick microbial mats and bordered by live clams (Fig. 1d–e). Fluid flow rates at Extrovert Cliff are temporally variable and tidally influenced, suggesting fluid conduits from an overpressurized aquifer (LaBonte et al., 2007; Füri et al., 2009).

Sediment push cores were collected from both seep sites in April 2019 on the R/V Western Flyer, using ROV Doc Ricketts. At each site, two push cores up to 20 cm long were collected from each of four locations: (1) the center of the cold seep, (2) the inner edge of the cold seep, (3) 5 m outside the seep boundary, and (4) 100 m outside the seep boundary (Fig. 2a). These were categorized as “Seep,” “Seep-Edge”, “Background-5m,” and “Background-100m” cores, respectively (Table S1 in the Supplement). Seep boundaries were delineated by the sudden termination of white, filamentous microbial mats and live clam beds on the sediment surface. These boundaries were validated by the sulfidic smell of the Seep and Seep-Edge cores from each site once cores were recovered.

On board, cores were kept at 4 °C until extruded from push core liners and sectioned within 8 h of collection. One core of each pair was sectioned into 2.5 cm (0–10 cm b.s.f.) or 5 cm (10+ cmbsf) horizons and subsampled for molecular and geochemical analyses, while the other was sectioned into 5 cm horizons and preserved anaerobically within Whirl-Paks sealed in Mylar bags at 4 °C for incubation experiments. Several 1 mL subsamples of each depth horizon in the molecular/geochemical cores were immediately flash-frozen in liquid nitrogen and preserved at -80 °C for later DNA and RNA extraction. Subsamples of 3 mL were transferred into 25 mL butyl-rubber-sealed vials filled with 5 mL of 5 M sodium hydroxide solution for methane analysis, 3–5 mL subsamples were scooped onto a pre-weighed sheet of aluminum foil for dehydration and porosity measurements, and 0.5 mL subsamples were transferred into 2 mL epitubes pre-loaded with 1 mL of 4 % PFA for microscopy. Porewater was squeezed from sediments using a porewater pressing bench (KC Denmark Research Equipment, Silkeborg, Denmark) under a stream of argon gas immediately after sectioning, and porewater was filtered with 0.2 µm Durapore^® PVDF membrane filters (EMD Millipore, Burlington, MA, USA). A total of 0.5 mL of porewater was fixed with 0.5 mL of 0.5 M zinc acetate and stored at 4 °C for sulfide measurements, and 2 mL of porewater was added to a 12 mL Exetainer^® (Labco Limited, Ceredigion, UK) pre-loaded with 1 mL 85 % phosphoric acid, evacuated, and N₂-flushed for δ13C analysis of dissolved inorganic carbon (DIC). The remaining porewater was stored at -20 °C.

Immediately after returning from the field, preserved sediment from all four Clam Field locations was anaerobically incubated and subsampled according to the method described in Dekas et al. (2009, 2014, 2016). Clam Field was chosen for incubations because its surface expression was similar to that of seeps sampled and characterized in previous studies (McVeigh et al., 2018; Seabrook et al., 2018; Semler et al., 2022), and thus responses of characteristic seep microbial communities to varying methane headspace concentration could be tested. The top two horizons (0–5 and 5–10 cm) from the Seep and Seep-Edge cores and the top three horizons (0–5, 5–10, and 10–15 cm) from the Background-5m and Background-100m cores (nhorizons=10) were incubated with varying concentrations (0–2 atm) of methane to evaluate community responses to methane addition; ¹⁵NH4+ was also added to all incubations at a final concentration of 100 µM to measure total anabolic activity (99 at. % ¹⁵N; Cambridge Isotopes, NLM-467-1). The methane was a 4:1 mixture of natural abundance ¹³C / ¹²C and 99 at. % ¹³C methane (Sigma-Aldrich, 490229-1L). Table S2 lists each of five methane treatments per horizon; all treatments were performed in triplicate (nincubations=150). Incubations were subsampled at 0-, 0.5-, 1-, 3-, and 6-month time points.

Headspace methane concentrations were measured from butyl-rubber-sealed vials (see above) using gas chromatography coupled to flame-ionization detection (Shimadzu GC-2014, Stanford University, Stanford, CA, USA) and back-calculated to porewater concentrations using porosity values determined via weight loss after dehydration (and assuming a sediment density of 2.65 g cm⁻³). Sulfide concentrations were measured in triplicate and determined colorimetrically from the zinc acetate-preserved porewater samples using the methylene blue method (Cline, 1969), with a detection limit of 0.02 mM. All other assays were performed without replication due to limitations on porewater volume.

DNA and RNA were extracted from flash-frozen sediments using the RNeasy PowerSoil Total RNA Isolation Kit (RNA) and the RNA PowerSoil DNA Elution Accessory Kit (DNA; Mo Bio Laboratories, Carlsbad, CA, USA). The protocol was modified from the manufacturer's instructions to include a bead-beating step of 60 s at a speed of 5.5 m s⁻¹ on a FastPrep instrument (MP Biomedicals, Santa Ana, CA, USA) to facilitate archaeal cell lysis as in Semler et al. (2022). RNA extracts were cleaned with the Ambion TURBO DNA-free Kit (Thermo Fisher Scientific, Waltham, MA, USA), and reverse transcription of RNA to cDNA was completed using Superscript III First Strand Synthesis Supermix (Thermo Fisher Scientific, Waltham, MA, USA).

DNA and cDNA were concentration normalized and amplified using a two-step PCR plan for Illumina amplicon sequencing. In the first step, universal primers 515F-Y/926R (Parada et al., 2016) were used to target the V4-V5 region of the 16S rRNA gene sequence, and mcrA_F/mcrA_R (Luton et al., 2002; Dekas et al., 2016) was used to target mcrA genes and transcripts. Both sets of primers included an extension complementary to the primers used in the second PCR. The gene-targeting regions of the primer sequences are listed in Table S3; 25 µL PCR reactions were performed containing 0.5 µL of forward and 0.5 µL of reverse primers (10 µM concentration), 10 µL 5PRIME HotMasterMix (2.5x, Quanta-Bio, Beverly, MA, USA), 13 µL DNase-free water, and 1 µL DNA or cDNA template. For 515F-Y/926R primers, the thermal cycling conditions were as follows: initial denaturing at 95 °C for 180 s; 28 cycles of 95 °C for 45 s, 50 °C for 45 s, and 68 °C for 90 s; a final elongation step at 68 °C for 300 s; and refrigeration at 4 °C until removal and storage. For mcrA_F/mcrA_R primers, the thermal cycling conditions were as follows: initial denaturing at 95 °C for 120 s; 35 cycles of 95 °C for 60 s, 50 °C for 60 s, and 72 °C for 60 s; a final elongation step at 72 °C for 300 s; and refrigeration at 4 °C until removal and storage. The 16S rRNA genes and mcrA genes were successfully amplified from all 45 sediment horizons, while 16S rRNA was successfully amplified from 42 of 45 sediment horizons, and mcrA transcripts were successfully amplified from 11 of 45 sediment horizons (Table S4).

In the second step, Illumina adaptors, barcodes, and indices were added to the amplicons. The same PCR reaction mix was used with custom primers targeting the primer extension in the first PCR. The thermal cycling conditions were as follows: initial denaturing at 95 °C for 180 s; 8 cycles of 95 °C for 30 s, 55 °C for 30 s, and 72 °C for 30 s; a final elongation step at 72 °C for 300 s; and refrigeration at 4 °C until removal and storage. Amplicons were cleaned with 0.7x AMPure XP magnetic beads (Beckman Coulter, Brea, CA, USA), pooled, and quantified before being sent to the UC Davis DNA Technologies Core Facility (Davis, CA, USA) for Illumina MiSeq 2×250 base pair (bp) (16S rRNA) or 2x300 bp (mcrA) sequencing. Ten 16S rRNA and four mcrA samples were randomly chosen for duplicate amplification. The average weighted UniFrac distance between duplicate 16S rRNA samples was 0.067. Negative (molecular grade water) and positive (mock communities of known composition) controls were processed and sequenced in parallel with the samples. Lack of DNA contamination in the RNA extracts was confirmed by processing RNA extracts (without reverse transcription) in parallel and seeing no visible amplification on a gel after the second PCR.

Demultiplexed sequences were trimmed with cutadapt (v. 2.10; Martin, 2011) and then filtered and processed using the R (v. 4.2.1) package DADA2 (v. 1.26.0; Callahan et al., 2016). Reads were trimmed to 216 (16S rRNA) or 260 and 230 (mcrA; forward and reverse reads, respectively) base pairs, with those containing more than two expected sequencing errors removed. Amplicon sequence variants (ASVs) were then inferred from filtered and merged reads. The majority of paired ends were merged for both genes (an average of 76 % for 16S rRNA reads in non-incubated samples and 95 % of mcrA reads), and the overlap length was roughly 30 bp for 16S rRNA and 50 bp for mcrA. Phylogenetic classification of 16S rRNA ASVs was based on the SILVA SSU database (v. 132; Quast et al., 2013). Classification of mcrA ASVs was determined manually based on placement on a reference tree (detailed below). On average, 937 16S rRNA reads were recovered per blank sample, while 36 372 16S rRNA genes and 4644 16S rRNA reads were recovered per in situ sample. The minimum number of reads recovered per 16S rRNA gene sample (DNA) was 7377, while the minimum number of reads recovered per 16S rRNA sample (RNA) was 71. On average, 4 mcrA reads were recovered per blank sample, while 46 941 mcrA gene and 11 066 mcrA transcript reads were recovered per in situ sample.

To classify mcrA sequences, the tool EPA-ng (v 0.3.8) was used to place the ASVs onto a manually compiled reference tree. Reference tree sequences included published sequences from cultured methanogens or ANME archaea, as well as six of the ASVs themselves (those that represented > 10 % of any sample's total sequences and had no cultured match above 90 % similarity in NCBI and would therefore be represented poorly on the reference tree). Reference sequences were aligned with MAFFT (v. 7.490) and incorporated into a RAxML (v. 8.2.12) best-scoring maximum likelihood (ML) reference tree with the GTR+G+I substitution model and 100 bootstraps. After ASV placement with EPA-ng, relative abundances of those ASVs in each sample were displayed in a heatmap using ggtree (v. 3.6.2). Relative abundances of ASVs placed on internal tree nodes were divided among tip nodes associated with that internal node.

Non-metric multidimensional scaling (NMDS) of microbial communities was carried out based on the weighted UniFrac distance metric (Lozupone et al., 2007) using the R package vegan (v. 2.5-7; Oksanen et al., 2020). Analysis of similarity (ANOSIM) was used to determine the significance of microbial community differences between groups of samples, also based on the weighted UniFrac distance metric. The R package DESeq2 (v. 1.38.3; Love et al., 2014) was used to test whether ASVs were significantly enriched in abundance with time (0 months vs. 6 months) across Clam Field seep incubations. Raw data from four US Atlantic Margin (USAM) seep sites (characterized in Semler et al. 2022) were simultaneously processed using the same packages for comparison with Monterey Bay sites.

Droplet digital PCR (ddPCR) was used to quantify abundances of mcrA genes and transcripts in Seep and Background-5m cores from Clam Field and Extrovert Cliff using the primer pair mcrA_F/mcrA_R (Luton et al., 2002; Dekas et al., 2016). For comparison, mcrA genes and transcripts were also quantified in a seep and a background core from an alternative seep site (New England seep) on the northern US Atlantic Margin; 25 µL PCR reactions were performed containing 1 µL of forward and 1 µL of reverse primers (5 µM concentration), 12.5 µL EvaGreen Supermix (2x, Bio-Rad Laboratories, Hercules, CA, USA), 9 µL DNase-free water, 0.5 µL bovine serum albumin (2.5 µg µL⁻¹), and 1 µL diluted DNA or cDNA template. Droplets were generated on a QX200 Droplet Generator (Bio-Rad) at the Stanford Functional Genomics Facility (Stanford, CA, USA) using droplet generation oil for EvaGreen (Bio-Rad). Thermal cycling was performed immediately afterward on a C1000 Touch Thermal Cycler, with thermal cycling conditions as follows: initial denaturing at 95 °C for 300 s; 45 cycles of 95 °C for 60 s, 52 °C for 90 s, and 72 °C for 75 s; signal stabilization steps at 4 °C for 300 s and 90 °C for 300 s; and a final 10 °C hold overnight. The overall ramp rate was set at 1 °C s⁻¹.

Droplets were read with the QX200 Droplet Reader (Bio-Rad). Threshold fluorescence values were initially inspected using QuantaSoft software (Bio-Rad), but the values were later adjusted using the minimum density clustering method, which better separated droplet clusters upon manual inspection. Amplicon copy numbers per well were then converted to copies per gram of dry sediment. Technical replicates were run for a randomly selected half of the samples to confirm the precision of our assay. Results of the replicate runs are shown in Fig. S1 in the Supplement.

To visualize putative ANME aggregates under the microscope, cells were fixed on board in 4 % paraformaldehyde according to the protocol of Dekas et al. (2009). Fixed sediment was diluted 1:20 in phosphate-buffered saline (PBS), sonicated 3×10 s at an amplitude of 30 on a Q500 sonicator (Qsonica, Newtown, CT, USA), and floated on top of a preestablished Percoll–PBS gradient (protocol described in Orphan et al., 2002, and Dekas and Orphan, 2011). In total, a 1 mL sonicated sample was added to 9 mL Percoll–PBS. After floating the sample, the mixture was centrifuged at 4780 rpm for 15 min at 4 °C. The supernatant was then filtered through a 25 mm 3 µm polycarbonate filter (Millipore, no. TSTP02500) backed with a glass microfiber filter (Cytiva Whatman, GF/F, no. 1825-025) under low (< 5 psi) vacuum and washed with 1 mL PBS and then 1 mL 100 % EtOH. Each filter was sectioned with a razor blade and stained with 4^′,6-diamidino-2-phenylindole (DAPI; no. D9542, Sigma-Aldrich, Darmstadt, Germany) before visualization on an inverted fluorescence microscope (Nikon Eclipse Ti). The number of aggregates in 100 random fields of view at 400× magnification was counted, corresponding to a detection limit (< 1 aggregate per 100 fields of view) of 8.58×104 aggregates (g⁻¹ dry sediment).

The amount of methane potentially diffusing out of sediments in the absence of biological consumption was estimated according to Fick's laws of diffusion (Boudreau, 1997). Equation (1) calculates the diffusion rate (J) in mmol m⁻² yr⁻¹: 1J=-φDsdCdz, where ϕ is the sediment porosity, Ds is the sediment diffusion coefficient (in m² yr⁻¹), and dC/dz is the methane concentration gradient over sediment depth (in mmol m⁻³ m⁻¹). The sediment diffusion coefficient, Ds, can be calculated via Eq. (2): 2Ds=D01+n(1-φ), where D0=1.4×105 cm² s⁻¹ (the initial diffusion coefficient of methane at 20 °C; Boudreau, 1997), n=3 (the lithology factor of silty clay), and ϕ is the sediment porosity.

Both cold seep sites contained moderate to high concentrations of porewater methane in all cores collected within the putative seep boundaries (i.e., both the Seep and Seep-Edge cores; Fig. 2b). Methane concentrations were 2 orders of magnitude higher at Extrovert Cliff (ranging from ∼ 3000 to ∼6400 µM) than at Clam Field (ranging from 25 to 183 µM), but even Clam Field contained methane concentrations roughly 4 times higher than the surrounding background sediment. At Clam Field, methane concentrations increased with depth in the sediment; methane concentrations at Extrovert Cliff, in contrast, showed no trend with depth but were consistently high throughout the sediment core. At both sites, methane concentrations in background sediments were measurable and ranged from 21 to 56 µM.

Sulfide concentrations at both cold seep sites were also elevated within the seep boundaries (Fig. 2b). At Extrovert Cliff, sulfide concentrations were elevated at all sediment depths (ranging from 2.7 to 12.7 mM), with no depth trend. At Clam Field, sulfide concentrations were below detection at the surface and increased with sediment depth (up to 11.2 mM).

With sediment depth, δ13C–DIC decreased at Clam Field (Fig. 2b) from -3.1 ‰ to -17.7 ‰, with the lowest values found in the deepest depths (Fig. 2). At Extrovert Cliff, δ13C–DIC was consistent with sediment depth (from -9.8 ‰ to -11 ‰), though at both sites, δ13C–DIC was more negative in cores collected within the seep than in cores collected in background sediments (from -0.1 ‰ to -4.9 ‰).

To evaluate methane cycling both on and off each cold seep and to compare those results with methane and sulfide concentrations, we quantified mcrA genes and transcripts with droplet digital PCR (ddPCR). mcrA encodes the alpha subunit of methyl coenzyme-M reductase and is a marker gene for both methanogens and anaerobic methanotrophs (Luton et al., 2002; Hallam et al., 2003; Krüger et al., 2003); therefore, quantification of the gene cannot distinguish between the two metabolisms. mcrA gene copies (g⁻¹ dry sediment) were highest in Extrovert Cliff seep sediments and peaked between 2.5–5 cmbsf at 5.3×107 copies (g⁻¹ dry sediment) – roughly the same sediment depth where methane concentrations began to decrease up the core (Fig. 3). Gene copies of mcrA were an order of magnitude less abundant at Clam Field seep – peaking in the uppermost sediment horizons at concentrations of 4.2×106 copies (g⁻¹ dry sediment). However, gene copy numbers at Clam Field seep were still above the average of 7.4×105 mcrA gene copies (g⁻¹ dry sediment) in background sediments from either site. Transcript copies of mcrA (per g dry sediment) were elevated in Extrovert Cliff seep sediments compared to background sediments (by a factor of 2) but were equivalent between Clam Field seep and background sediments.

To characterize the microbial communities at these sites, we performed 16S rRNA gene and 16S rRNA amplicon sequencing in the seep (Seep and Seep-Edge) and background samples (5 and 100 m from seepage) from both sites. In the 21 seep samples, we recovered 6402 unique 16S rRNA ASVs. The most abundant of these were members of the phyla Bacteroidota, Campylobacterota, Desulfobacterota, Chloroflexi, Proteobacteria (mainly Gammaproteobacteria), and Verrucomicrobiota. The relative abundances of Campylobacteria and Chloroflexi tended to increase with sediment depth in a core (by an average of 141 % and 422 %, respectively), while the relative abundance of Bacteroidota decreased (by an average of 53.4 %) with sediment depth (Fig. 4a). Across 24 background samples, we recovered 12 731 unique 16S rRNA ASVs. Microbial communities in the background sediments from both sites were similar to one another at the phylum level (ANOSIM: p value = 0.066) with Bacteroidota, Chloroflexi, Desulfobacterota, and Proteobacteria as the most abundant groups (Fig. 4a). There was no significant difference between microbial communities in Background-5m (5 m from seepage) and Background-100m cores (100 m from seepage) – neither at the phylum level (ANOSIM: p value = 0.102) nor at the ASV level (ANOSIM: p value = 0.084). Consistent with the typical “seep microbiome” (Ruff et al., 2015; Semler et al., 2022), Caldatribacteriota, Campylobacterota, and Verrucomicrobiota were enriched in seep cores in comparison to background cores (> 3 times greater) at both sites, while Desulfobacterota and Proteobacteria (mainly Gammaproteobacteria) had particularly high relative abundances in background cores in comparison to seep cores (roughly 2 times greater; Fig. 4a). While relative potential activity of many groups (as inferred by the presence of 16S rRNA (cDNA)) mirrored many of the trends seen in the DNA, reads from several phyla, including Halobacterota and Desulfobacterota, were more than twice as abundant in the cDNA fraction of the seep core at Extrovert Cliff; Halobacterota was 40 times more abundant in the cDNA analysis of the seep core at Extrovert Cliff.

Halobacterota – the phylum containing the majority of methanogenic and anaerobic methanotrophic taxa (including ANME archaea) – were present in very low relative abundances in the 16S rRNA gene dataset. At Clam Field, these organisms were not detected in seep or background sediments, with the exception of extremely low (< 0.1 %) abundances of ANME-3 in the two upper sediment horizons of the Seep-Edge core (Fig. 4b). No Halobacterota reads were detected in cDNA at Clam Field. At Extrovert Cliff, all seep sediment horizons contained low levels of Halobacterota (< 0.3 % of the community), with the majority assigned to ANME-2c or ANME-3. A small number of methanogenic Methanomicrobia were also present in the lowest sediment depths of the Extrovert Cliff Seep core. While Halobacterota were not highly abundant at Extrovert Cliff, their relative potential activity was high – up to ∼ 18 % of total community cDNA in some cores. This transcriptional activity was almost entirely associated with ANME-2c and was highest in the 5–7.5 cmbsf depth horizons, consistent with the location of the peak in mcrA transcripts in the ddPCR data. Relative abundance of rRNA is a not precise proxy for relative activity between taxa, but it can indicate which taxa are likely translationally active (Blazewicz et al., 2013). We refer to the detection of rRNA as reflecting “potential” activity to emphasize the limitations of this proxy.

Desulfobacterota – the phylum comprising known ANME syntrophs as well as other free-living sulfate reducers common at cold seeps – comprised roughly 20 % of the Clam Field microbial community and roughly 10 % of the Extrovert Cliff microbial community (Fig. 4b). The most common subgroups within the Desulfobacterota were Desulfobacteria (including Seep-SRB1), Seep-SRB2 (at Extrovert Cliff only), and Seep-SRB4 (particularly at Clam Field). Seep-SRB1, the group containing many known obligate ANME syntrophs (Knittel et al., 2003; Schreiber et al., 2010; Skennerton et al., 2017; Metcalfe et al., 2021), were found in low abundances at both seep sites, ranging from 0.70 %–2.04 % of sequences in Clam Field seep cores and from 0.73 %–1.26 % of sequences at Extrovert Cliff. Also present – particularly at Clam Field – were Desulfomonadia, a group associated with sulfur- and iron-reducing capabilities (Ravenschlag et al., 1999; Wunder et al., 2021). At Clam Field, the cDNA profiles of Desulfobacterota generally mirrored the DNA profiles, with Desulfobacteria dominating the reads overall, and Desulfobulbia, Seep-SRB4, and Seep-SRB1 showing relative potential activity peaks in the top, middle, and bottom portions of the cores, respectively. At Extrovert Cliff, Desulfobacterota comprised roughly 3 times more of the cDNA reads than they did for the DNA, mirroring the high relative potential activity observed for the Halobacteria there. The increase was largely due to Seep-SRB2, which alone comprised up to 30 % of the cDNA reads in the 5–7.5 cmbsf depth horizon in both cores. This is consistent with the previously described association of members of ANME-2c and Seep-SRB2 (Krukenberg et al., 2018; Kleindienst et al., 2012) and suggests active anaerobic methanotrophy by these groups at this depth.

To provide greater insight into the diversity of methane-cycling microorganisms at these sites, and specifically increase our ability to detect low-abundance methanotrophs, if present, we sequenced mcrA genes at both sites. Sequencing the mcrA gene specifically increased our detection limit of methane-cycling organisms by nearly 3 orders of magnitude, as we recovered on average 46 941 reads per sample using mcrA gene sequencing versus 61 reads of putative methane-cycling organisms per sample with 16S rRNA gene sequencing. Within Extrovert Cliff seep samples, the mcrA gene results were generally consistent with the 16S rRNA results, with mcrA sequences affiliated with ANME-3 detected in shallow-depth horizons and ANME-2c and methanogenic archaea appearing deeper in the cores. Interestingly, within the two seep cores at Extrovert Cliff (both Seep and Seep-Edge), a single ASV affiliated with the ANME-2c group comprised ∼ 50 % of the total mcrA reads, peaking between 5 and 10 cmbsf (Fig. 5). Similar observations of local dominance of individual ANME ASVs had been made at other seep sites previously (Semler et al., 2022).

The mcrA sequences from Clam Field revealed a more nuanced perspective of methane cycling than the 16S rRNA sequences, with seep samples containing ANME-1 as well as a variety of putatively methanogenic archaea. Notably, most of these, including the ANME-1 reads, were more relatively abundant in the background samples at Clam Field than the seep samples, indicating they were not enriched by increasing methane concentrations. The one exception was ASV.2, which was the most abundant ASV at Clam Field seeps (alone comprising roughly 45 % of seep ASVs) and was almost absent from background sediments (0.1 % of background ASVs). ASV.2 has no cultured relative above 90 % similarity in NCBI but clusters with sequences from Methanohalophilus halophilus and Methanomethylovorans hollandica, anaerobes involved in methylotrophic methanogenesis (Lomans et al., 1999). The most highly abundant ASV (ASV.3) in Clam Field and Extrovert Cliff background sediments is most closely related to strain MO-MCD, also a methylotrophic methanogen, belonging to the genus Methanococcoides (Singh et al., 2005). Although this organism is also present within the Clam Field seep, the clear difference in distribution between ASV.2 and ASV.3 across the seep boundary highlights niche separation between putatively similar methylotrophic methanogens at these sites. In general, while ANME-affiliated sequences dominated the mcrA dataset from seep samples at Extrovert Cliff (85.6 % of mcrA reads), reads associated with methylotrophic methanogens dominated both Clam Field (> 80 % of mcrA reads; only 5.6 % were affiliated with ANME archaea) and the background sediments from each site (> 75 % and > 60 % of reads from background cores at Extrovert Cliff and Clam Field, respectively). We also attempted to sequence mcrA transcripts at Clam Field (Table S4), but mcrA expression was below the limit of detection (no visible amplification of the cDNA on a gel).

By multiplying the mcrA copy number in samples (as determined by ddPCR) by the proportion of ANME-affiliated mcrA reads in that sample (as determined by the amplicon sequencing analysis using the same primer set) – and by assuming a single mcrA gene copy per cell – ANME cell numbers in each sample were estimated (Table S5). In total, ANME archaea were not enriched in Clam Field seep sediments (4.5×102 to 8.2×104 ANME cells g⁻¹ dry sediment) relative to background sediments (8.3×103 to 6.9×105 ANME cells g⁻¹ dry sediment), and ANME mcrA genes at Clam Field were attributed primarily to ANME-1 (Fig. 5). ANME cell numbers were estimated to be roughly 4 orders of magnitude higher at Extrovert Cliff seep (8.2×106 to 4.7×107 ANME cells g⁻¹ dry sediment) than at Clam Field seep (Table S5), with ANME mcrA genes attributed primarily to ANME-2c and ANME-3 (Fig. 5). The assumption of one mcrA copy per cell is imperfect, but only one or two copies of mcrA have been found in sequenced methanogen genomes (Alvarado et al., 2014), including that of Methanosarcina mazei – a close relative of ANME-2 (Deppenmeier et al., 2002; Nunoura et al., 2006). The ANME-1 genome also contains a single operon for MCR (Krukenberg et al., 2018; Chadwick et al., 2022; Laso-Pérez et al., 2023). As a result, the difference of roughly 4 orders of magnitude in ANME cell numbers between Clam Field and Extrovert Cliff would not be significantly affected by likely variation in ANME mcrA copy number.

To visualize and quantify potential aggregates of ANME archaea and sulfate-reducing bacteria from these seeps, we examined DAPI-stained cells from methane and sulfate replete sediments at Clam Field (7.5–10 cmbsf) and Extrovert Cliff (5–7.5 cmbsf) seeps. ANME-2 and ANME-3 typically form tight associations with their syntrophic partners, resulting in cellular aggregates of characteristic morphologies (Boetius et al., 2000; Orphan et al., 2002). While not a taxa-specific assay, quantifying DAPI-stained cell aggregates typical of the ANME–SRB morphology provides independent support for the presence or absence of aggregate-forming ANME-2 and ANME-3 archaea in the molecular data and has been used previously to approximate potential ANME–SRB aggregate abundances (Dekas et al., 2009; Zhang et al., 2011). In sediments from Extrovert Cliff, where ANME-2 and ANME-3 were found in the DNA (0.2 % of 16S rRNA gene reads) and cDNA analyses (10.1 % of 16S rRNA reads), we detected putative ANME cell aggregates at a concentration of 4.35×106 aggregates (g⁻¹ dry sediment) (Fig. S2). Considering ANME aggregates can contain anywhere from tens to thousands of cells, this value is roughly consistent with the estimate of ANME density derived from the molecular data. Cell aggregates were not found at Clam Field, consistent with the lack of ANME-2 and dearth of ANME-3 sequences detected at this site. Our aggregate detection limit corresponded to 8.58×104 aggregates (g⁻¹ dry sediment).

To investigate whether sulfate-coupled methane oxidation was occurring at Clam Field and determine whether canonical ANME archaea could be enriched there, Clam Field sediments were incubated under varying concentrations of methane. Sulfide concentrations increased continuously over 6 months in sediment from the seep center, up to 10 mM, indicating active sulfate reduction (Fig. S3). A smaller increase (∼ 1–2 mM) was observed in sediments from the Seep-Edge core. However, there was not a statistically significant difference in sulfide production with or without methane in either core, indicating that sulfate reduction was not methane-dependent. We therefore did not see evidence of sulfate-coupled methane oxidation in these sediments, in contrast to previous observations at other seeps (e.g., Hydrate Ridge (Nauhaus et al., 2002, 2007), Eel River basin (Dekas et al., 2009), and the Costa Rica margin (Dekas et al., 2014)).

Consistent with a lack of methane oxidation, ANME archaea were not enriched during the 6-month incubation, as assessed by 16S rRNA sequencing (DNA and cDNA; Fig. S4). Taxa that were significantly enriched after 6 months with methane included several ASVs within the Desulfobacterota, specifically Desulfovibrio, Desulfobulbus, Desulfuromonas, and Seep-SRB4 (Fig. S5a). A putatively methylotrophic methanogen in the genus Methanococcoides was also significantly enriched relative to the pre-incubation community. Aerobic sulfide-oxidizing groups, including Candidatus Maribeggiatoa, Colwellia, and Thiomargarita, significantly decreased in abundance over 6 months with methane. Without methane, none of the above taxa were significantly enriched or unenriched after 6 months of incubation time (Fig. S5b). A single Verrucomicrobia ASV had significantly increased in abundance, and a single Proteobacteria ASV (from the family Beggiatoaceae) had significantly decreased in abundance. Furthermore, there were no ASVs that were significantly enriched or under-enriched when comparing incubations with and without methane after 6 months.

To directly compare Monterey Bay seep microbial communities with canonical seep communities, we compared them to those of four seeps along the US Atlantic Margin (USAM), which were sampled and sequenced with the same methodologies. First, we compared the 16S rRNA gene profiles of the Monterey Bay seeps to those of the USAM seeps using non-metric multidimensional scaling (NMDS). Monterey Bay seep samples formed a separate cluster from the Atlantic samples (Fig. 6a), and communities from Extrovert Cliff were nested within those from Clam Field. While the primary axis of the NMDS plot was defined by sediment depth (Fig. 6b), the secondary axis was defined by geographic region. When including background samples from all sites, background samples from Monterey Bay clustered together within background sediment communities sequenced from the USAM (Fig. 6c). As in the seep-only community comparison, the primary axis of the plot was defined by sediment depth (Fig. 6d), though the secondary axis was instead defined by environment type (seep vs. background), rather than sampling region.

We also measured mcrA concentrations via ddPCR at New England seep – one of the four USAM seeps – to better contextualize the trends in mcrA abundance observed within Monterey Bay. At the New England seep, where ANME archaea had represented 13.4 % of the microbial community in 16S rRNA gene surveys on average (Semler et al., 2022), we found that mcrA gene copy numbers reached 2.5×108 copies (g⁻¹ dry sediment) at the deepest sediment depths, outnumbering those at both Extrovert Cliff and Clam Field by 1 and 2 orders of magnitude, respectively (Fig. S1). However, mcrA transcript copy numbers at the New England seep were comparable to those at Extrovert Cliff, though they peaked at lower sediment depths (15+ cmbsf).

We also chose two USAM seeps (New England and Shallop Canyon East) at which to visualize putative ANME aggregates. Again, fixed sediments from methane and sulfate replete depths (0–3 cmbsf) at these two sites were examined under an epifluorescence microscope with a DAPI stain. In the chosen horizons, ANME archaea had comprised 4.8 % (New England) and 0.5 % (Shallop Canyon East) of the microbial community in 16S rRNA gene surveys (Semler et al., 2022). Consistent with their higher relative abundances in the molecular data compared to the Monterey seeps, we discovered higher concentrations of putative aggregates in these sediments by microscopy: 1.15×107 and 1.34×106 aggregates (g⁻¹ dry sediment), respectively.

The porewater methane concentrations at Clam Field decreased approximately linearly up the core, indicative of diffusive flow between the methane-rich sediments and the methane-poor overlying water column. This is in contrast to a concave uptrend, which is indicative of biological methane consumption and therefore AOM (Reeburgh, 1976; Martens and Berner, 1977; Ward et al., 1987). The lack of methane-dependent sulfide production in sediments from this site, as well as the scarcity of ANME archaea, further supports a lack of significant biological methane consumption. Although we cannot exclude the possibility of low levels of biological oxidation, if we assume diffusion is the only mechanism removing methane from seep sediment, we can calculate the diffusive flux to determine the potential amount of methane released to the water column at this site. The methane concentration gradient over sediment depth (dC/dz) was determined from the average slope of the linear relationship between methane concentration and sediment depth (Fig. S6). Calculated fluxes were 17.8 and 18.5 mmol methane m⁻² yr⁻¹ for the Clam Field Seep and Seep-Edge cores, respectively (Table S6).