Two gravity cores (GCs; ARA07C-St 13 and ARA07C-St 14) and five GCs (ARA09C-St 04, ARA09C-St 06, ARA09C-St 07, ARA09C-St 12, and ARA09C-St 16) were collected from the ARAON Mounds during the ARA07C Expedition in 2016 and the ARA09C Expedition in 2018 onboard Ice-Breaking Research Vessel (IBRV) ARAON, respectively. In addition, one GC was sampled at Site ARA09C-St 08 as a reference site. The retrieved core length of all GCs is less than 6 m (Figure 1; Table 1).

Pore fluid was extracted by Rhizons in whole round cores or split cores in ~10-60 cm intervals at room temperature on the IBRV ARAON. Extracted pore fluid was collected in acid-prewashed syringes equipped with an in-line 0.20 µm disposable polytetrafluoroethylene filter. The bottom seawater was also collected from Site ARA09C-St 04 (Table 1) using a CTD/rosette system that holds 12-5L Niskin bottles (Seabird 911 Plus). Fluid subsamples for shipboard and anion analyses were collected in acid-prewashed high-density polyethylene (HDPE) bottles. Fluid aliquots for cation and ⁸⁷Sr/⁸⁶Sr isotope ratios analyses were transferred into acid-prewashed HDPE bottles (~2-4 ml) and acidified with 20 µl ultrapure grade HNO₃. Subsamples for stable isotopic properties of water (δD and δ¹⁸O) and dissolved inorganic carbon (δ¹³C_DIC) were collected in 2 ml septum screw-lid glass vials. Fluid samples for δ¹³C_DIC analysis were preserved with 30 μl HgCl₂. Pore fluid and seawater samples were stored at approximately 4°C in the refrigerator until the analyses.

Gas hydrates from Sites ARA07C-St 13, ARA09C-St 06, and ARA09C-St 16 were carefully scraped to minimize any contamination that may have occurred during the earlier handling and were dissociated in clean 20 ml glass beakers at room temperature on the IBRV ARAON. The aliquots of these fluids (hydrate-bound waters) for the analyses of compositional and isotopic properties were collected using the same techniques as described above for pore fluid and seawater, and then stored at approximately 4°C in the refrigerator until the analyses.

Methane-derived authigenic carbonates were distributed from ~0.2 to ~2.2 meters below the seafloor (mbsf) in the cores during the ARA07C and ARA09C Expeditions (Table 2). Some MDACs were observed with several cm-thickness in the split cores (Jin and Shipboard Scientific Party, 2017; Jin and Shipboard Scientific Party, 2019). MDACs were hand-picked in the split cores and collected in plastic bags during these expeditions (Figure 2). These samples were stored at approximately 4°C in the refrigerator until the analyses.

The chlorinity (Cl^-) and alkalinity of the pore fluid, hydrate-bound water, and seawater were measured onboard during the ARA07C and ARA09C Expeditions. The Cl^- concentration was measured via titration with 0.1 M silver nitrate (AgNO₃) and the alkalinity was determined immediately by titration with 0.02 M HCl. The reproducibilities of Cl^- and alkalinity by repeated analyses of the International Association of Physical Sciences of the Oceans (IAPSO) standard seawater were < 2% and < 0.5%, respectively. Sulfate (SO₄^2-) in pore fluid and hydrate-bound water from the ARA07C Expedition and the bottom seawater from Site ARA09C-St 04 was analyzed by ion chromatography (IC) at the Korea Basic Science Institute (KBSI; Dionex ICS-1100, Thermo Scientific). IAPSO standard seawater was repeatedly used to verify the analytical quality of the instruments, and the analytical reproducibility was better than 3%. In addition, SO₄^2- in pore fluid and hydrate-bound water collected from the ARA09C Expedition was analyzed using the IC at the Kitami Institute of Technology (KIT; 2707 plus Autosampler, 1525 Binary HPLC Pump, and 432 Conductivity Detector, Nihon Waters K.K., Japan). The reproducibility, estimated from repeated analyses of IAPSO standard seawater, was < 3%.

Major and minor cations (Na⁺, K⁺, Mg²⁺, Ca²⁺, Sr²⁺, and H₄SiO₄) were analyzed by inductively coupled plasma-optical emission spectroscopy (Perkin Elmer Optima 8300) at the KBSI. The reproducibility, estimated from repeated analyses of certified reference materials (SLRS-5 and TMDW), was < 5%.

Stable water isotopes (δD and δ¹⁸O) of pore fluid, hydrate-bound water, and seawater from the ARA07C and ARA09C Expeditions were determined with a VG Prism stable isotope ratio mass spectrometer (SIRMS) at the KBSI and an off-axis integrated cavity output spectroscopy laser absorption spectrometer (Los Gatos Research (LGR) Liquid Water Isotope Analyzer [LWIA-24d]) at the KIT, respectively. The analytical reproducibilities by the SIRMS and LWIA were ±0.1‰ for δ¹⁸O and ±1‰ for δD, and ±0.2‰ for δ¹⁸O and ±0.6‰ for δD, respectively.

The carbon isotopic composition (δ¹³C_DIC) in the fluid was analyzed with a Finnigan DELTA-Plus mass spectrometer using a Gas-Bench II automated sampler at Oregon State University. The reproducibility was better than 0.07‰, based on the multiple standard measurements. Isotopic values are reported in the conventional δ-notation relative to Vienna Standard Mean Ocean Water (V-SMOW) for hydrogen and oxygen, and Vienna Pee Dee Belemnite (V-PDB) for carbon.Dissolved strontium in the fluid was separated for isotopic analysis using Sr-Spec columns (Eichrom-Sr resin). Strontium isotopic ratios (⁸⁷Sr/⁸⁶Sr) were measured using a Neptune multi-collector inductively coupled plasma mass spectrometer (Thermo Finnigan, Bremen, Germany) upgraded with a large dry interface pump at the KBSI. The measured ⁸⁷Sr/⁸⁶Sr ratios were normalized to ⁸⁶Sr/⁸⁸Sr = 0.1194, and repeated NBS 987 measurements yielded 0.71025 ± 0.00002 (2σ_mean, n = 24).

The pretreatment of MDACs from the ARA07C Expedition was performed at the Korea Institute of Geoscience and Mineral Resources (KIGAM). The sample was washed several times using ultrapure distilled water and sonicated for a few minutes to eliminate adhered materials. Then, it was rinsed again with ultrapure distilled water and dried at 60 °C in an oven for 12 h. For δ¹³C and δ¹⁸O analyses in MDACs, approximately 5 mg of sample was reacted with 100% H₃PO₄ at 90°C for 5 min, and the evolved CO₂ gas was automatically injected into a VG Prism SIRMS at the KBSI. The analytical reproducibility was better than ±0.1‰ for both δ¹³C and δ¹⁸O. Isotopic values are reported in the conventional δ-notation relative to V-PDB for carbon and oxygen.

The MDAC sample was ultrasonically extracted three times with solvent mixtures (dichloromethane (DCM):methanol (MeOH) (2:1 v/v)). Detailed procedures for lipid biomarker analyses have been previously described by Lee et al. (2018). In short, one-half of the total lipid extract (TLE) was dried over anhydrous Na₂SO₄ and treated with tetrabutylammonium sulfite reagent to remove elemental sulfur. The TLE was chromatographically separated into apolar and polar fractions over an Al₂O₃ column (activated for 2 h at 150°C). The apolar fraction was eluted using hexane:DCM (9:1 v/v), and 40 µL of 5α-androstane (10 µg/ml) was added as an internal standard. The polar fraction was recovered with DCM:MeOH (1:1 v/v) as an eluent and 40 µl of C₂₂ 7,16-diol (10 µg/ml) was added as an internal standard. This aliquot was derivatized through silylation, prior to quantification by gas chromatography (GC) and identification with gas chromatography-mass spectrometry (GC-MS). Molecular compounds were determined by comparing their mass spectral fragmentation patterns and retention times with previously published data (Stadnitskaia et al., 2008; Lee et al., 2018). The δ¹³C values of lipid compounds are expressed via conventional δ-notation relative to V-PDB and the analytical reproducibility is less than ±0.4‰ for all lipid compounds.

The compositional and isotopic properties of the seawater and pore fluids are represented in Figure 3 and Supplementary Table 1. The Cl^-, Na⁺, K⁺, δ¹⁸O, δD, and ⁸⁷Sr/⁸⁶Sr values of pore fluids from Sites ARA07C-St 13, ARA07C-St 14, ARA09C-St 04, ARA09C-St 07, ARA09C-St 08, and ARA09C-St 12 are relatively uniform throughout the sampling depth, and these values are generally similar to those of the bottom seawater from Site ARA09C-St 04 (Figure 3; Supplementary Table 1). However, the downcore profile of SO₄^2- from these sites can be classified into two groups. The first group, Site ARA09C-St 08 (reference site), shows a linear decrease from the seawater value (~30 mM) at the top of the sediment to ~24 mM at 5.20 mbsf. In addition, the δ¹³C_DIC values from this site continuously decrease, ranging from -20.1‰ to -1.3‰ (Figure 3; Supplementary Table 1). These results imply that Site ARA09C-St 08 does not reach the SMT. In contrast, the downcore profile of SO₄^2- in the second group, including Sites ARA07C-St 13, ARA07C-St 14, ARA09C-St 04, ARA09C-St 07, and ARA09C-St 12, exhibits three distinct variations with depth. This profile shows a relatively constant or gradual decrease at shallow depths from the seafloor and then abruptly decreases to the SMT. Below the SMT, it has a relatively constant or a slight decrease (Figure 3; Supplementary Table 1). The SMT depths of Sites ARA07C-St 13, ARA07C-St 14, ARA09C-St 04, ARA09C-St 07, and ARA09C-St 12 are estimated to be ~1.3 mbsf, ~0.9 mbsf, ~1.2 mbsf, ~3.3 mbsf, and ~2.1 mbsf, respectively (Figure 3; Table 1). The minimum δ¹³C_DIC value occurs around the SMT at each site (Figure 3; Supplementary Table 1).

The downcore profiles of Ca²⁺, Mg²⁺, and Sr²⁺ at Sites ARA07C-St 13, ARA07C-St 14, ARA09C-St 04, ARA09C-St 07, and ARA09C-St 12 display trends similar to those of SO₄^2-. In contrast, the downcore profile of alkalinity oppositely overlaps with that of SO₄^2- (Figure 3; Supplementary Table 1). The H₄SiO₄ concentrations (0.29 ± 0.12 mM, n=64) in all samples from Sites ARA07C-St 13, ARA07C-St 14, ARA09C-St 04, ARA09C-St 07, ARA09C-St 08, and ARA09C-St 12 are higher than that of bottom seawater from Site ARA09C-St 04 (~0.04 mM; Figure 3; Supplementary Table 1).Compared to other sites in the ARAON Mounds, chemical compositions (e.g., Cl^-, Na⁺, Mg²⁺, K⁺, Ca²⁺, and Sr²⁺) have depleted values while stable water isotopes (δ¹⁸O and δD) have enriched values at several intervals of pore fluids from Sites ARA09C-St 06 and ARA09C-St 16 (Figures 3, 4; Supplementary Table 1). The SO₄^2- concentrations at these sites are very low with a maximum value of ~2.0 mM and do not exhibit any distinct trend along the entire core length (Figure 3; Supplementary Table 1). Alkalinity also shows no trend, and its concentration is greater than ~14 mM at these sites (Figure 3; Supplementary Table 1). The δ¹³C_DIC values increase from -12.2‰ at ~0.35 mbsf to 2.9‰ at ~2.15 mbsf of Site ARA09C-St 16 (Figure 3; Supplementary Table 1), which is a higher value than those of other sites that reach the SMT in the ARAON Mounds. The ⁸⁷Sr/⁸⁶Sr ratios from Site ARA09C-St 16 are relatively constant (0.70917 ± 0.00001, n=4; Figure 3; Supplementary Table 1) and are similar to the open seawater (~0.70917; Paytan et al., 1993) and the bottom seawater from Site ARA09C-St 04 (0.70919; Supplementary Table 1).

The compositional and isotopic properties of the hydrate-bound waters are represented in Supplementary Table 1. Most dissolved chemical compositions in the hydrate-bound waters have lower concentrations than those in pore fluids at the same site. Although the values of δ¹⁸O and δD are limited as we have analyzed one sample from the bottom of Site ARA07C-St 13, they are 2.6‰ and 16.5‰, respectively, which are much higher than those of pore fluids at this site, ranging from 0.3‰ to 0.4‰ and from 0.9‰ and 1.6‰, respectively. However, the ⁸⁷Sr/⁸⁶Sr ratio in the hydrate-bound water from Site ARA09C-St 16 is 0.70923, which is similar to pore fluids from this site (0.70917 ± 0.00001, n=4; Figure 3; Supplementary Table 1).

The MDACs were observed at Sites ARA07C-St 13 and ARA07C-St 14 during the ARA07C Expedition (Figure 2; Table 2) and at Sites ARA09C-St 04, ARA09C-St 06, ARA09C-St 07, ARA09C-St 12, and ARA09C-St 16 during the ARA09C Expedition (Jin and Shipboard Scientific Party, 2017; Jin and Shipboard Scientific Party, 2019; Kim et al., 2020). We analyze the δ¹³C and δ¹⁸O of MDACs from the ARA07C Expedition. δ¹³C_MDAC and δ¹⁸O_MDAC values vary from -34.4‰ to -25.2‰, and from 4.6‰ to 6.2‰, respectively (Figure 5; Table 2).

The MDAC sample includes various lipid components of anaerobic methanotrophs, but there is a lack of biomarkers derived from aerobic methanotrophs. Among the apolar components, the irregular, tail-to-tail linked isoprenoid acyclic C₂₀ (2, 6, 11, 15-tetramethylhexadecane or crocetane) and C₂₅ (2, 6, 10, 15, 19-pentame-thylicosane, PM) hydrocarbons are present in the MDAC sample. The δ¹³C values of the isoprenoid hydrocarbons (i.e., crocetane, PMI) are -119.1‰ and -105.5‰, respectively (Figure 6). Similarly, the isoprenoid glycerol diethers archaeol and sn-2-hydroxyarchaeol are detected in the polar fractions of MDAC samples with δ¹³C values of -104.5‰ to -101.2‰. Another distinct group of detected compounds is non-isoprenoid glycerol diethers (DGDs), tentatively inferred previously as a marker of uncharacterized sulfate reducing bacteria (Werne et al., 2002; Pancost et al., 2011). These compounds (e.g., DGD (If), DGD (IIa), and DGD (IId)) have low δ¹³C values in the range of -78.2‰ to -66.1‰ (Figure 6).

The dissolved chemical species concentrations and water isotopes (δ¹⁸O, δD, and ⁸⁷Sr/⁸⁶Sr) values in pore fluids from Sites ARA07C-St 13, ARA07C-St 14, ARA09C-St 04, ARA09C-St 07, ARA09C-St 08, and ARA09C-St 12 are quite similar to those in the bottom seawater from Site ARA09C-St 04 (Figure 3; Supplementary Table 1). Moreover, the measured δ¹⁸O and δD values in pore fluids do not have a distinct relationship with each other or with Cl^- concentrations (Figure 4). Consequently, the pore fluids at these sites predominantly originate from the overlying ambient seawater in the ARAON Mounds. As H₄SiO₄ and Ca²⁺ concentrations in pore fluid samples from these sites have enriched and depleted values, respectively, compared to the bottom seawater (Figure 3; Supplementary Table 1), pore fluid properties are likely to be altered within the sediment column by water-rock interactions and by mineral precipitation (Kim et al., 2016; Kim et al., 2022).

In contrast, the downcore profiles of dissolved chemical species and stable water isotopes from Sites ARA09C-St 06 and ARA09C-St 16 show distinctly depleted and enriched excursions at several intervals, respectively, relative to those in pore fluids and seawater from other sites in the ARAON Mounds (Figures 3 , 4; Supplementary Table 1). In addition, the measured δ¹⁸O and δD values in pore fluids from these sites have a good positive correlation (R² > 0.98) whereas they show a good negative correlation with Cl^- concentrations (R² > 0.83) (Figure 4). These characteristics are typical when the fluid derived from GH dissociation affects the pore fluid. Indeed, many GHs found in situ at Sites ARA09C-St 06 and ARA09C-St 16 during the ARA09C Expedition (Jin and Shipboard Scientific Party, 2019). As the pressure decreases and temperature increases when the cores are retrieved from the sediment column to the deck, GHs in the sediment of Sites ARA09C-St 06 and ARA09C-St 16 are dissociated and release freshwater into the sediment (e.g., Hesse and Harrison, 1981; Matsumoto and Borowski, 2000; Ussler and Paull, 2001; Hesse, 2003; Torres et al., 2008; Torres et al., 2011). Therefore, the fluid from GH dissociation results in pore fluid freshening with enriched δD and δ¹⁸O values. In addition, because GH dissociation causes the sediment expansion in the core liner during the core retrieving (Kim et al., 2020), ambient seawater can flow sediment in Sites ARA09C-St 06 and ARA09C-St 16. As a result, pore fluid chemistry can be contaminated by the ambient seawater, which interrupts the estimation of the SMT depth of these sites based on the only SO₄^2- profile of pore fluids. Overall, the compositional and isotopic properties of pore fluids from Sites ARA09C-St 06 and ARA09C-St 16 have been severely altered by the GH decomposition. We also found an in situ GH at the bottom of Site ARA07C-St 13 (core catcher; > 2.35 mbsf), however, the fluid from the GH dissociation cannot significantly influence the pore fluid properties at several intervals of this site, as it does at Sites ARA09C-St 06 and ARA09C-St 16.

Based on low Cl^- values that are affected by GH dissociation and background Cl^- value in pore fluid that is not affected by GH dissociation, GH saturation (S_h, % of pore space) can be estimated using the following equation (Matsumoto and Borowski, 2000; Ussler and Paull, 2001; Hesse, 2003; Torres et al., 2008; Torres et al., 2011):

where C_b represents the in situ background dissolved Cl^- of the water, which in this study is the pore fluid value at 0.35 mbsf from Site ARA09C-St 16 (~563 mM). C_s is the Cl^- measured in the sample after GH dissociation and β is a dimensionless constant that accounts for the density change due to GH dissociation and equals 1.257 (Ussler and Paull, 2001; Malinverno et al., 2008; Torres et al., 2011; Kim et al., 2013).

The estimated S_h values at Sites ARA09C-St 06 and ARA09C-St 16 have ranges of approximately 9-32% (n=8; average=20%) and 2-46% (n=4; average=16%), respectively, which vary widely within < 2.5 m-length core at each site (Table 3). These results indicate that the GH is likely to be heterogeneous within the sediment of Sites ARA09C-St 06 and ARA09C-St 16.

The downcore profile of SO₄^2- concentration in pore fluids from Site ARA09C-St 08 indicates that this site does not reach the SMT under steady state condition (Figure 3). In contrast, these profiles at Sites ARA07C-St 13, ARA07C-St 14, ARA09C-St 04, ARA09C-St 07, and ARA09C-St 12 are concave-up and/or kink type with a minimum SO₄^2- concentration of ~0 mM. Since this downcore profile of SO₄^2- usually suggests the non-steady state of pore fluid chemistry (Hensen et al., 2003; Henkel et al., 2011), the observed SO₄^2- profiles imply that Sites ARA07C-St 13, ARA07C-St 14, ARA09C-St 04, ARA09C-St 07, and ARA09C-St 12 penetrate the SMT under the non-steady state condition. The non-steady state downcore profiles of SO₄^2- have already been documented in many regions such as the Argentine Basin, South China Sea, offshore Namibia, and offshore Svalbard (Niewöhner et al., 1998; Fossing et al., 2000; Hong et al., 2017; Hu et al., 2019), which is attributed to either bioirrigation of macrofauna, seawater intrusion during methane ebullition, mass-transport deposits, or increasing upward methane and fluid flux. Because the bioirrigation of macrofauna has rarely been observed in split sediment cores from Sites ARA07C-St 13, ARA07C-St 14, ARA09C-St 06, and ARA09C-St 16, and methane flares have not been observed at all of ARAON Mounds sites (Jin and Shipboard Scientific Party, 2017; Jin and Shipboard Scientific Party, 2019), bioirrigation of macrofauna and the seawater intrusion associated with methane ebullition have not likely occurred at these mounds. In addition, sediment structures related to mass transport deposits have not been found in the sediment facies and SBP images of ARAON Mounds (Jin and Shipboard Scientific Party, 2017; Jin and Shipboard Scientific Party, 2019; Kim et al., 2021b), implying that large-scale sediment reworking cannot account for the non-steady state of SO₄^2-. In contrast, during the ARA07C and ARA09C Expeditions, we found many in situ GHs at shallow depths at Sites ARA07C-St 13, ARA09C-St 06, and ARA09C-St 16 (core length < 3 m) (Jin and Shipboard Scientific Party, 2017; Jin and Shipboard Scientific Party, 2019; Kim et al., 2020; Choi et al., 2021). To form the GH near the seafloor in the marine environment, the CH₄ flux from deep-seated sediment to the seafloor through the conduit should be high. Thus, we postulate that the non-steady state shown in the downcore profile of SO₄^2- of the ARAON Mounds is linked to the variation in upward CH₄ flux from the deep sediment column. This result is partly consistent with the gas chemistry in the ARAON Mounds (Kim et al., 2020).

Kim et al. (2020) revealed that the CH₄ source originates from thermogenic and microbial, which remarkably changes around the SMT on the basis of gas chemistry in the ARAON Mounds. The thermogenic CH₄ that migrated from deep-seated sediments (> 1 km) toward the seafloor through the faults/fractures alters the shallow gas compositional and isotopic properties below/around the SMT in this mound while the microbial CH₄ mainly impacts the area above/around the SMT (Kim et al., 2020). Since the SO₄^2- gradient of pore fluid can be mainly controlled by the upward gas flux in the marine sediment (Borowski et al., 1996; Borowski et al., 1997), its sharp decreasing gradient above the SMT at Sites ARA07C-St 13, ARA07C-St 14, ARA09C-St 04, ARA09C-St 07, and ARA09C-St 12 (Figure 3) can be directly attributed to the upward thermogenic CH₄ flux around the SMT. However, this mechanism does not explain a gradually decreasing or non-observed SO₄^2- gradient in the shallow sediments of these sites (Figure 3). It is likely another process that modifies the fluid chemistry associated with GHs and MDACs within the sediment column.

The downcore profiles of Ca²⁺, Mg²⁺, and Sr²⁺ at Sites ARA07C-St 13, ARA07C-St 14, ARA09C-St 04, ARA09C-St 07, and ARA09C-St 12 generally follow the trend of SO₄^2-, with rapidly lower concentrations in the zone showing a linear decrease (Figure 3; see Compositional and Isotopic Properties of Fluids). These results suggest that intense carbonate precipitation occurs in this zone at each site, as evidenced by the many MDACs found during the ARA07C and ARA09C Expeditions (Jin and Shipboard Scientific Party, 2017; Jin and Shipboard Scientific Party, 2019; Kim et al., 2020). In contrast, an MDAC does not occur above the SMT. Therefore, the precipitation of carbonates is likely to be triggered by the increase in alkalinity associated with the AOM.

The carbon isotopes (δ¹³C_MDAC) of MDACs reflect the CH₄ source involved in the AOM reaction. In general, the δ¹³C_MDAC in association with the thermogenic CH₄ usually ranges from ~-40‰ to -20‰ while it is less than -40‰ by the AOM using microbial CH₄ (Naehr et al., 2007; Crémière et al., 2016; Crémière et al., 2018). The measured δ¹³C_MDAC values at the ARAON Mounds range from -36.4‰ to -21.9‰ (Figure 5), indicating thermogenic CH₄ as a predominant source of MDACs in the ARAON Mounds. This result is consistent with the CH₄ source around/below the present SMT in these mounds (Kim et al., 2020). Under such conditions, specific lipid molecules (e.g., archaeol and sn-2-hydroxyarchaeol) preserved within MDAC, can identify the predominance of anaerobic methanotrophs (ANMEs), as well as paleoenvironmental conditions (e.g., methane sources and fluctuation) in association with the AOM reaction during the formation of MDAC (Blumenberg et al., 2004; Elvert and Niemann, 2008; Niemann and Elvert, 2008; Himmler et al., 2015). Typically, microbial communities dominated by ANME-2 at the cold seeps of the northwestern Black Sea contain higher amounts of sn-2-hydroxyarchaeol relative to archaeol, whereas the reverse is observed in microbial mats dominated by ANME-1 (Blumenberg et al., 2004). Indeed, the ratio of isotopically depleted sn-2-hydroxyarchaeol relative to archaeol can be used to distinguish ANME-1 (0–0.8) from ANME-2 (1.1–5.5), with ANME-3 (2.4) falling within the range of ANME-2 (Niemann et al., 2005; Niemann and Elvert, 2008). The compound ratios (i.e., sn-2-hydroxyarchaeol/archaeol) in MDAC from the ARAON Mounds are ~1 (Figure 6), implying that ANME-1 is predominantly involved in the AOM reaction to form MDACs using the deep CH₄ sources. The diagnostic archaeal (crocetane, PMI, archaeol, and, sn-2-hydroxyarchaeol) and bacterial lipids (DGDs) with strongly depleted ¹³C in MDAC (Figure 6) also support this interpretation.

The formation of MDACs in the ARAON Mounds can drastically alter the physical properties of sediment and flow pathway because it reduces the permeability of sediments and potentially prevents direct flow or retards gas and fluid migration within the sediments (Hovland, 2002; Bahr et al., 2007; Bayon et al., 2009). Similarly, the formation and growth of GHs in the sediment by a high CH₄ flux can modify physical properties and flow pathways (Sassen et al., 1999; Römer et al., 2012; Sultan et al., 2014). Therefore, we postulate that MDACs and GHs observed widespread in the ARAON Mounds are likely to prevent or interrupt the upward migration of gas and fluid in the sediment column (Figure 7). Given that the fluid coupled with gas continuously migrates through the sediment column, the downcore profile of SO₄^2- reaches steady state condition and the CH₄ source should be steadily changed from the seafloor to deep-seated sediment with the CH₄ flux (Figures 7A–C). Otherwise, the upward gas and fluid are blocked or redirected by MDACs and GHs within the sediment column, and then sediment is continuously deposited over them within a short time scale. The gas and pore fluid chemistry above these deposits are mainly controlled by ambient seawater and by (bio)geochemical reactions (e.g., organic matter degradation by sulfate reduction, microbial methanogenesis) associated with the microbial activity to degrade the organic matter rather than by the upward CH₄ flux (Figure 7D). As a result, the chemical zone is entirely separated between the below and above deposition of the MDACs and GHs (Figure 7D).

The δ¹⁸O signature of carbonates can inform the oxygen isotopic composition of the fluid when the carbonate precipitates (e.g., Greinert et al., 2001; Naehr et al., 2007). Assuming that aragonite and carbonated formed in isotopic equilibrium with ambient bottom seawater at the temperature and δ¹⁸O value of the ARAON Mounds, respectively, the theoretical δ¹⁸O_aragonite and δ¹⁸O_carbonate values can be estimated by these equations (Kim and O’Neil, 1997; Kim et al., 2007);

The temperature and δ¹⁸O value of the present bottom seawater from Site ARA09C-St 04, as measured during the ARA09C Expedition, are 0.8°C (Jin and Shipboard Scientific Party, 2019) and 0.3‰ (V-SMOW; Supplementary Table 1), respectively. Thus, the theoretical δ¹⁸O of aragonite and carbonate is ~4.0‰ (V-PDB) and ~3.3‰ (V-PDB), respectively. All analyzed δ¹⁸O_MDAC values from Sites ARA07C-St 13 and ARA07C-St 14 are higher than these calculated equilibrium values (Figure 5; Table 2), suggesting the incorporation of enriched ¹⁸O fluid during MDAC precipitation. The enriched ¹⁸O fluid may originate from either clay mineral dehydration, opal diagenesis, or GH dissociation (Hesse and Harrison, 1981; Kastner et al., 1991; Ussler and Paull, 1995; Hesse, 2003; Kim et al., 2013; Kim et al., 2021a). In general, the ⁸⁷Sr/⁸⁶Sr ratio tends to be lower than that of the ambient present seawater by clay dehydration, and the δ¹⁸O value increases with a relatively constant δD value by opal diagenesis (Kastner et al., 1991; Kim et al., 2013; Kim et al., 2021a). Since most ⁸⁷Sr/⁸⁶Sr ratios in pore fluids are similar to the present bottom seawater of ARA09C-St 04 and the values of δ¹⁸O and δD are relatively constant in the non-GH-bearing intervals, irrespective of sampling depth (Figures 3, 4; Supplementary Table 1), we excluded clay mineral dehydration and opal diagenesis as the primary sources for enriched ¹⁸O fluid in the ARAON Mounds. Instead, it is reasonable that the ¹⁸O-enrichment of MDACs can be attributed to fluids derived from GH dissociation in relation to climate change through geological time (Greinert et al., 2001; Naehr et al., 2007), as indicated by the occurrences of shallow GHs around MDACs at Sites ARA07-St 13, ARA09C-St 06, and ARA09C-St 16 (Figures 1, 2; Tables 1, 2; Supplementary Table 1).