The Peru Trench is part of the larger Atacama Trench that is located between the continental South American Plate and the accretionary wedge of the oceanic Nazca Plate (Suess, 1981). ODP Site 1230 is located on the lower slope of the Peru Trench at 5,086 m water depth (Figure 1). Sediments were drilled to ~270 mbsf during ODP Leg 201 in 2002 (D’Hondt et al., 2003). Three boreholes (A, B, and C) were within ~20 m of one another (D’Hondt et al., 2003). Sediment temperatures are low, increasing linearly from 2°C at the seafloor to 12°C at 270 mbsf. The upper 200 m of sediment consist of clay-rich, diatomaceous mud that was largely relocated from the continental shelf throughout the Holocene and Pleistocene. At approximately 216 mbsf, the sediment column changes to Miocene diatom ooze in a stratigraphic hiatus of 4.5 million years (Shipboard Scientific Party, 1988; Meister et al., 2005). Throughout the sediment column, organic carbon contents mostly range from 2 to 4% dry sediment weight (Meister et al., 2005). DIC concentrations to 160 mM and sulfate depletion in the upper ~10 mbsf indicate active microbial remineralization of organic matter, largely by sulfate reduction (D’Hondt et al., 2003). After sulfate is depleted, methane concentrations increase rapidly and reach in situ saturation by 28 mbsf (Spivack et al., 2005). Geophysical and chemical data suggest that hydrates are first present at ~70 mbsf and occur intermittently to 278 mbsf (D’Hondt et al., 2003).

For molecular biological analyses, 5-cm whole-round intervals of cores were frozen at −80°C. Only sediment from the nearly contamination-free core interiors was used (House et al., 2003; Lever et al., 2006). For carbon isotope analyses, 5-mL subsamples were frozen in pre-combusted glass vials.

We used published depth profiles of DIC, sulfate, methane, dihydrogen (H₂), formate and acetate concentrations (Figures 2A–E; D’Hondt et al., 2003). Due to outgassing during core retrieval, measured methane concentrations below ~12 mbsf were underestimates of in situ concentrations. We calculated methane concentrations below the saturation depth at in situ temperature, pressure, and salinity, assuming a uniform pore size of 1.0 μm based on the equilibrium model for methane hydrate-seawater-porous media (Sun and Duan, 2007). Modeled methane concentrations generally agree with measured in situ methane concentrations based on pressure coring (4 depths analyzed at ODP Site 1230; Spivack et al., 2005).

δ¹³C-C₁ (~99% ¹³C-CH₄) and -DIC (Figure 3) were measured as described previously (Biddle et al., 2006). All values are shown in Supplementary Table S1.

RNA was extracted as in Biddle et al. (2006), except that the extraction buffer was supplemented with 120 mM sodium phosphate. DNA was extracted using the same protocol as for RNA, except that the pH of the extraction buffer and phenol were raised to 8.0, the bead beating time reduced to 15 s, and the bead beating speed reduced to 4.0 (Qbiogene, Carlsbad, CA). Moreover, the DNase incubation was omitted, and DNA purified with the PowerClean DNA Clean-Up Kit (MOBIO laboratories, Carlsbad, CA) instead of the RNeasy Mini Kit (Qiagen, Valencia, CA).

Two previously published general mcrA primer pairs yielded no amplification (ME1/ME2, Hales et al., 1996), or amplification at only one depth interval (mcrI, Springer et al., 1995; see Inagaki et al., 2006). The mcrIRD primer pair, a modified version of the mcrI primer pair with a reduced number of nucleotide degeneracies and consequently improved detection sensitivity, was used in conjunction with the ANME-1-specific ANME-1-mcrI primer pair [both published in Lever and Teske (2015)]. Special primers for ANME-1 detection were necessary due to the high number of nucleotide mismatches between the mcrI primer pair primer and the genetically divergent mcrA sequences of ANME-1. To confirm that detected mcrA detected belonged to active and living members of the methane-cycling community, we performed RT-PCR of mcrA-mRNA and 16S rRNA in several depth horizons using new group-specific primers for maximum amplification efficiency and hence detection sensitivity (mRNA: ODP1230, ANME-1; 16S rRNA: Msaeta 268F/927R, ANME-1 42F/898R; ANME-1-SG 35F/1038R). All primer sequences used in this study are shown in Table 1. All nucleotide sequences are publicly accessible at GenBank.

PCR assays of mcrA were performed using the Takara SpeedSTAR HS DNA polymerase kit (TaKaRa Bio USA, Madison, WI) using (1) 1 × 2 min denaturation (98°C), (2) 40 × (a) 10s denaturation (98°C), (b) 30s annealing (Table 1 for temperatures), (c) 1 min extension (72°C), and (3) 1 × 5 min extension (72°C). Negative controls and reaction blanks were included.

RT-PCR assays were carried out using TaKaRa RNA PCR Kits (AMV) Version 3.0 (TaKaRa Bio USA, Madison, WI) and (1) 1 × 15 min reverse transcription, (2) 5 min denaturation (98°C), (3) 40 × (a) 30s denaturation (98°C), (b) 30s annealing (Table 1 for temperatures), (c) 1 min extension (72°C), and (4) 1 × 5 min extension (72°C). Negative controls and reaction blanks were included. Absence of DNA was confirmed by DNA-PCR with the same treatments but omitting the reverse transcription step.

PCR products were purified in a 2.5% low-melting point agarose gel using 1 × Tris acetate - EDTA buffer (TAE). Gel slices containing PCR fragments of the correct length were excised and purified using a S.N.A.P. Mini Kit (Invitrogen, Carlsbad, USA). Purified PCR fragments were cloned using the Topo TA Kit (Invitrogen, Carlsbad, USA) and transformed into TOP10 electrocompetent cells following the manufacturer’s instructions. Plasmid extraction and purification was done using the GeneJET Plasmid Miniprep Kit (ThermoFisher Scientific) and cycle sequencing was performed on an ABI 3730 Sequencer with M13 universal primers (SP010-SP030) at the Josephine Bay Paul Center at MBL (Woods Hole, MA). Sequences were BLAST analyzed using the nucleotide collection in GenBank.¹ Phylogenetic trees were created and bootstrap analyses (1,000 replicates) performed in ARB² using manually optimized SILVA 16S rRNA gene alignments, and a custom-built, publicly accessible mcrA database (name: mcrA4All)³ with >2,400 high-quality, aligned mcrA amplicon and genome sequences.

Gibbs energy yields (ΔG_r) of methanogenesis reactions from H₂ + CO₂ (2 HCO₃⁻ + 4 H₂ + H⁺ ➔ CH₄ + 3 H₂O) and acetate (CH₃COO⁻ + H₂O ➔ CH₄ + HCO₃⁻), and the methanogenic conversion of formate to methane [4 HCOO⁻ + H₂O + H⁺ ➔ CH₄ + 3 HCO₃⁻; note: this reaction presumably involves the initial oxidation of formate to H₂ and HCO₃⁻, which is not known to conserve energy except in certain hyperthermophiles (Schink et al., 2017)] were calculated based on the equation: Δ G r = Δ G r 0 + R T ln Q r where ΔG _r ⁰ is the Gibbs energy (kJ mol⁻¹ of reaction) at standard concentrations (1 M for reactants and products, pH 7.0), corrected for in situ temperature T (K) and pressure p (bar) based on standard enthalpies and molar volumes as outlined in Stumm and Morgan (1996), R is the universal gas constant (0.008314 kJ mol⁻¹ K⁻¹), and Q_r the quotient of product and reactant activities. Calculations were done for measured pH and concentrations of DIC (HCO₃⁻), H₂, and acetate. Measured methane concentrations were used for the upper 12 mbsf, while modeled concentrations were used below. Activities of all chemical species were calculated by multiplying concentrations by their activity coefficients. These were γ_HCO32− = 0.532 (Millero and Schreiber, 1983), and γ_CH4 = 1.24 (Millero, 2000). The activity coefficients of H₂, acetate, and formate were approximated with those of CH₄ (H₂) and HCO₃⁻ (acetate, formate). Standard Gibbs energies (∆G _f °), standard enthalpies (∆H _f °), and standard molal volumes (∆V _f °) of formation are shown in Supplementary Table S2.

¹³C-isotopic signatures of porewater methane and DIC provide insights into the zones of biological methane production and oxidation (Figure 3). Throughout the sediment column, methane was ¹³C-depleted relative to DIC. δ¹³C-DIC-values (only determined in borehole A) decreased slightly from −10.4‰ in the upper meter to −13.3‰ by 7.65 mbsf, then increased sharply in the uppermost methanogenic layer to +6‰ at 20 mbsf. The steepest increase in δ¹³C-DIC occurred within the interval from 7.65 mbsf (δ¹³C-DIC: −13.2‰) to 9.15 mbsf (δ¹³C-DIC: −8.4‰), and suggests onset of methanogenesis by CO₂ reduction in this interval. Below 20 mbsf, δ¹³C-DIC-values continued to gradually increase to reach a maximum of +20‰ at 123 mbsf, below which values slightly fell off to +15‰ at 246 mbsf (Figure 3). The ¹³C-isotopic compositions of methane, determined in the methanogenesis zones of boreholes A (>25 mbsf) and B (0–12 mbsf), were in a range typical of biological methanogenesis (Whiticar et al., 1986; Whiticar, 1999). δ¹³C-CH₄ was ~ −65‰ in the upper 6 mbsf, and then decreased to −75‰ at 12 mbsf. This increase in δ¹³C-CH₄ upward through the SMTZ is consistent with isotopic discrimination of AOM against ¹³C-CH₄. Below, values gradually increased from to ~ −65‰ at 246 mbsf (Figure 3). The difference in δ¹³C-CH₄ relative to δ¹³C-DIC was ~ −53‰ in the upper 6 mbsf, decreased across the SMTZ reaching −71‰ in the upper methanogenesis zone at 12 mbsf, and stabilized at −80‰ to −85‰ below 25 mbsf (Figure 3).

We detected mcrA sequences of three phylogenetic groups (Figure 4): (1) putatively anaerobic methanotrophic ANME-1a-b Archaea, (2) a sister group of ANME-1, which we here refer to as ANME-1d, and (3) sequences of Methanotrichales that cluster with a genus-level group that includes the known aceticlastic methanogens Methanothrix harundinacea and Methanothrix pelagica.

The three groups were vertically zonated (Figure 5). ANME-1a-b mcrA sequences were found in horizons near the upper (7.8 mbsf) and lower limit (9.7 mbsf) of the STMZ in Borehole B and four horizons in the upper methanogenesis zone (10.25–20.6 mbsf; Table 2). Sequences of ANME-1d were detected with ANME-1a-b sequences at one depth in the upper methanogenesis zone (20.6 mbsf) and in three horizons below (to 189.0 mbsf). Methanothrix mcrA showed a distribution similar to ANME-1d, but was detected in more sediment horizons and to greater depth (to 227 mbsf; Table 2).

We detected mcrA-mRNA of ANME-1-d in one of the 9 samples examined using the ANME-1-mcrA primer pair (core 6H-2, 44 mbsf; Table 2). In 11 replicate RT-PCRs of RNA extracts, controls (PCR negative, extraction blank, DNA controls) always tested negative, whereas 8/11 RNA extracts tested positive. By comparison, four other samples (from 7.8, 9.7, 10.2, 10.8 mbsf) yielded RT-PCR detection with the same primers, but DNA controls were positive (albeit weaker than cDNA bands), indicating that traces of DNA had resisted the DNAse treatment. We cloned cDNA of mcrA-mRNA from core 6H-2 and confirmed the presence of mcrA of ANME-1d.

In addition to mcrA-mRNA, we examined 16S rRNA sequences. RT-PCRs with new Methanothrix-specific 16S rRNA gene primers (Table 1) yielded Methanothrix-like sequences in two additional depth horizons (Table 2; Supplementary Figure S3). This primer pair also generated 16S rRNA sequences of a sister group of Methanosarcinales, previously detected in methane seep and mud volcano environments, with unknown metabolism at two depths (15H-6, 30H-1; Supplementary Figure S3). Interestingly, despite detecting mRNA with ANME-1-mcrA primers, we were unable to detect 16S rRNA of ANME-1a-b or ANME-1d with newly designed group-specific 16S rRNA gene primers (Table 1).

The changes in δ¹³C-DIC and –methane provide insights into the sources of DIC and pathways of methane production at ODP Site 1230. The δ¹³C-DIC isotopic values in the upper ~9 mbsf (−10.4 to −13.2‰) are consistent with organic matter mineralization becoming the main DIC source with increasing sediment depth. Most of this organic matter is likely to be phytoplankton-derived organic matter (δ¹³C-total organic carbon: ~22–23‰; Biddle et al., 2006) that was initially deposited under the upwelling regime of the Peru Margin, and subsequently reworked and laterally transported downslope to the Peru Trench. Toward the sediment surface, the δ¹³C-DIC increases, most likely due to an increasing contribution of ¹³C-enriched DIC from deep sea bottom water, which typically bears a ¹³C-composition of ~0 to +1.2‰ (Lynch-Stieglitz et al., 1995).

Notably, despite the strong geochemical evidence for AOM in the SMTZ, which might be expected to produce highly ¹³C-depleted DIC from the oxidation of methane, we do not observe a strong downward swing in δ¹³C-DIC within the SMTZ. This phenomenon has been observed previously in SMTZs and has been explained with concomitant AOM and methane production (Beulig et al., 2019), microbially mediated isotope exchange between methane and DIC (Yoshinaga et al., 2014), and reversibility of intracellular methane-cycling reactions at low sulfate concentrations (Wegener et al., 2021). In our case, the mcrA data argue against the first scenario, if ANME-1a-b are assumed to only perform methanotrophy. Yet, if – as proposed previously - ANME-1a-b are facultative methanogens, which matches the detection of this group throughout the upper ~12 m of the methanogenesis zone, then the first scenario is also plausible. Notably, porewater dissolved barium concentrations increase sharply throughout the AOM and upper methanogenesis zone (e.g., from 2.7 μM at 6.15 mbsf to 290 μM at 23.15 mbsf at ODP Site 1230A; D’Hondt et al., 2003), consistent with (slow) release of sulfate through chemical dissolution of barite (BaSO₄). This sulfate could fuel low rates of AOM, and thus also support concomitant AOM and methane production throughout the upper methanogenesis zone. AOM coupled to iron or manganese reduction could also support low rates of AOM, as was recently proposed for subsurface sediments of the South China Sea, where ANME-1 were detected meters below the SMTZ (Zhang et al., 2023). Yet, the low porewater concentrations of Fe²⁺ (0.6 to 3.9 μM) and Mn²⁺(0 to 0.3 μM) in the upper methanogenic sediment layer where we detected ANME-1a-b at ODP Site 1230 (D’Hondt et al., 2003) do not support an important role of AOM coupled to metal reduction.

Below 9 mbsf, the δ¹³C-DIC increased, consistent with a strong isotopic imprint of methanogenesis by CO₂ reduction. Strong isotopic discrimination against ¹³C-CO₂ is the norm in methanogenesis from H₂/CO₂ (Whiticar, 1999; Penning et al., 2005) and can result in significant ¹³C-enrichment of the residual DIC pool (Alperin and Hoehler, 2009; House et al., 2009). Based on measured porewater geochemical data, methanogenesis from H₂/CO₂ is, however, not thermodynamically favorable (Figure 6), with in situ Gibbs energies in the positive (i.e., endergonic) range (ΔG _r ’ > 0 kJ mol⁻¹) throughout the sediment column of ODP Site 1230. Since methanogenesis from formate follows the same biochemical route as hydrogenotrophic methanogenesis after the initial oxidation of formate to CO₂ and H₂ by formate dehydrogenase (Sparling and Daniels, 1986), a similar isotopic fractionation can be expected. Indeed, the complete conversion reaction of formate to methane is thermodynamically favorable, and based on that alone formate a potential methanogenic substrate at ODP Site 1230 (Figure 6). Yet, assuming that energy is not conserved during the initial formate oxidation step, but only in the second step involving methanogenic CO2 reduction with H₂ (Schink et al., 2017), then formate conversion to methane appears less plausible. This is because intracellular H₂ concentrations can be expected to be close to equilibrium with H₂ concentrations in the surrounding sediment due to H₂ leakage out of methanogenic cells (Finke et al., 2007). As stated above, however, measured H₂ concentrations in the surrounding sediment are too low to energetically support hydrogenotrophic methanogenesis. A more recently documented form of methanogenic CO₂ reduction involves interspecies electron transfer (IET). This form of methanogenesis, which was first discovered in Methanothrix harundinaceae (Rotaru et al., 2014), involves cellular structures, e.g., cytochromes, that attach to conductive mineral surfaces or syntrophic partner organisms (Gao and Lu, 2021). The isotopic fractionations of these reactions are not known but most likely also cause δ¹³C-enrichment of residual DIC. In principle, the conversion of formate to methane could also operate via a direct electron transfer mechanism, e.g., from syntrophic bacteria to methanogens. This mechanism could bypass H₂ as a catabolic intermediate and even render formate catabolism a potential source of methane. Thus, based on the available geochemical data, the dominance of methanogenic CO₂ reduction at Site 1230, which was inferred from the δ¹³C-DIC profile below 9 mbsf, is most plausibly explained with electron transfer from syntrophic partner organisms or mineral surfaces to methanogens.

By contrast, chemoautotrophy, acetogenesis or other methanogenic pathways are unlikely drivers of the observed δ¹³C-DIC increase in the methanogenesis zone. Although (certain) ANME-1a-b are chemoautotrophs (Kellermann et al., 2012), past studies indicate that AOM of isotopically highly depleted methane (~ − 75 per mil) to CO₂ occurs at much (≥40-fold) higher rates than C-assimilation by chemoautotrophy (e.g., Nauhaus et al., 2007; Wegener et al., 2008). Consequently, AOM would be expected to overprint any C-isotopic enrichment of DIC by chemoautotrophy. Acetogenesis from H₂/CO₂, which also strongly discriminates against ¹³C (Gelwicks et al., 1989), is unlikely based on δ¹³C-acetate values of −12 to −18‰ at ODP Site 1230 that indicate fermentation as the main acetate source (Heuer et al., 2006). The other widespread methanogenesis pathways from acetate (aceticlastic methanogenesis) and methylated substrates (e.g., methanol, dimethyl sulfide and methyl amines; methylotrophic methanogenesis), produce rather than consume CO₂ (Whitman et al., 2014). In aceticlastic methanogenesis, this CO₂ has the same ¹³C-depleted isotopic composition as the methane produced (Gelwicks et al., 1994) and would thus lower (rather than increase) the δ¹³C-DIC. Despite high concentrations of acetate, our calculations, moreover, indicate that aceticlastic methanogenesis is close to thermodynamic equilibrium throughout most of the methanogenesis zone (Figure 6), with Gibbs energies not reaching the theoretical minimum required for biological energy conservation by proton translocation (ΔG _r ’ ≅ 10 kJ mol⁻¹; Hoehler et al., 2001; Lever et al., 2015). ¹³C-depletion of DIC is also expected for methylotrophic methanogenesis, even though this pathway produces methane with similar isotopic fractionations as CO₂ reduction (Conrad, 2005). The reason for ¹³C-depletion of DIC is that the main isotopic fractionation of methylotrophic methanogenesis is produced by the first enzymes in the reaction chain (methyl transferase I and/or II; Krzycki et al., 1987) and hence upstream of where C fractions enter separate enzymatic pathways to produce CO₂ and methane (Thauer, 1998). Notably, another form of methylotrophic methanogenesis, which involves methylated substrates and hydrogen, e.g., methanol + H₂ does not produce CO₂ (Dridi et al., 2012; Whitman et al., 2014). A fourth methanogenic pathway that involves the conversion of methoxy-groups from lignin monomers to methane with CO₂ as a co-substrate (Mayumi et al., 2016) is in theory also possible. Yet, this pathway is unlikely to be important given the primarily phytoplanktonic origin of organic matter at ODP Site 1230 (Shipboard Scientific Party, 1988; D’Hondt et al., 2003) and recent evidence suggesting minimal long-term degradation of lignin in anoxic sediment (Han et al., 2022).

The increase in δ¹³C-DIC with depth is steepest from the lower SMTZ to ~20 mbsf (Figures 3, 5), consistent with rates of methanogenic CO₂ reduction being highest in this interval. The subsequent decrease in the slope of δ¹³C-DIC with depth can be explained with an increase in the DIC pool size and decline in the rates of CO₂ reduction. In addition, it is possible that the relative contributions of other methanogenic pathways, e.g., aceticlastic methanogenesis, increase below this depth. Nonetheless, the ¹³C-isotopic depletions of −80 to −86‰ of δ¹³C-methane relative to δ¹³C-DIC that were consistently measured below 18 mbsf (Figure 3) indicate that CO₂ reduction accounts for most of the methane that has accumulated throughout the methanogenesis zone of ODP Site 1230.

When examined in the geochemical context, the distribution of mcrA genes within the methanogenesis zone of ODP Site 1230 may be surprising. Isotopic compositions of DIC and methane suggest predominance of methanogenesis by CO₂ reduction, whereas detected mcrA sequences belong to phylotypes of putatively anaerobic methane-oxidizing ANME-1a-b, its catabolically uncharacterized sister group ANME-1d, and Methanothrix, a genus that was traditionally believed to consist uniformly of obligately aceticlastic methanogens.

The presence of ANME-1a-b in the SMTZ and in underlying sediment that is net methanogenic can be explained with different scenarios. The first one is that ANME-1a-b are indeed facultatively methanogenic, as proposed previously based on strong heterogeneity in δ¹³C of ANME-1-biomass in seep sediments (House et al., 2009), detection of ANME-1a-b mcrA transcripts in methanogenic sediment (Lloyd et al., 2011), and combined methanogenesis rate measurements and mcrA analyses across anaerobic methane-oxidizing and methanogenic sediment (Beulig et al., 2019). This environmental evidence has been supported by recent genomic detections of hydrogenase genes, that are potentially involved in hydrogenotrophic methanogenesis, across multiple ANME-1a-b and ANME-1c taxa (Laso-Pérez et al., 2023). Alternatively, AOM by ANME-1a-b may continue as a cryptic process in the presence of methanogenesis throughout the uppermost part of the methanogenesis zone. Our thermodynamic calculations indicate that a reversal of methanogenic CO₂ reduction with H₂ is thermodynamically favorable throughout bulk sediments of ODP Site 1230 (Figure 6), though the electron acceptor is unclear. As discussed earlier, the increase in dissolved barium indicates barite (BaSO₄) dissolution in this part of the sediment column as a potential source of sulfate, whereas the very low Fe²⁺ and Mn²⁺ concentrations argue against a significant role of AOM coupled to metal reduction. Under this scenario, the organisms that were responsible for the production of the measured methane are unknown. While neither possibility can be ruled out, the available evidence from this and past studies supports ANME-1a-b contributing to the production of methane by CO₂ reduction in the upper methanogenesis zone of ODP 1230.

The metabolism of ANME-1d, which replaces ANME-1a-b in deeper sediment layers of the methanogenesis zone, is even less understood than that of ANME-1a-b. This group, which was previously also referred to as “ANME-1-related group” and represents a poorly studied, sister family, or even sister order, of ANME-1a-b (Lever and Teske, 2015), has been found across a range of anoxic environments. These include hydrothermal vents in ultramafic settings (Kelley et al., 2005), deeply buried terrestrial coalbeds (Fry et al., 2009), marine gas hydrate sediments (Kormas et al., 2005), and tidal creek sediments (Edmonds et al., 2008). Given the sole detection of ANME-1d DNA and mRNA deep in the methanogenesis zone, a methanogenic lifestyle seems likely. This group could reduce CO₂ to methane using electrons from IET and thus contribute to the observed strong δ¹³C-depletion of methane relative to DIC.

The predominant detection of Methanothrix-mcrA sequences below 20 mbsf, despite δ¹³C-DIC compositions that indicate mainly methanogenesis by CO₂ reduction, and Gibbs energies of aceticlastic methanogenesis near thermodynamic equilibrium, is perplexing, given that members of Methanothrix are traditionally considered to be obligate aceticlasts. One explanation is that these sequences belong to inactive or dead cells. Yet, this explanation does not match the detection of rRNA of Methanothrix (Table 2), and is at odds with research suggesting that the vast majority of DNA from dead microorganisms is degraded over time scales of centuries in subsurface sediments (Torti et al., 2018). Instead, the Methanothrix mcrA and 16S rRNA sequences may not belong to (obligate) aceticlasts. While genomic data of Methanothrix thermophila indicate potential for hydrogenotrophic metabolism in this group (Smith and Ingram-Smith, 2007), methanogenesis involving H₂ has never been shown for Methanothricaceae. Yet, more recent experiments with pure cultures have demonstrated that members of Methanothrix - including Methanothrix harundinacea, which mcrA sequences from ODP Site 1230 cluster with (Figure 5) - are capable of methanogenic growth by CO₂ reduction using electrons received directly or through mineral intermediates from syntrophic partner organisms (Rotaru et al., 2014; Yang et al., 2019; Gao and Lu, 2021). Experiments involving rice paddy soils and lake sediments have provided additional evidence for CO₂ reduction by Methanothrix in the environment (Holmes et al., 2017; Rotaru et al., 2019). Consequently, the observed δ¹³C-DIC and δ¹³C-methane compositions and dominance of mcrA sequences of Methanothrix may not be a contradiction, but instead match revised knowledge on the metabolic capabilities of Methanothrix.