The sediment cores (YS 4, and YS 7, ~60 m penetration depth) were taken from offshore of the ECS ( Figure 1 ) in 2017, The water depth of the boreholes were 12.43 m and 7.57 m, respectively. Fifty-six pore water and 72 headspace gas samples were sampled at 0.8~6.0 m intervals based on the visual lithologic variation. Pore water samples were collected using Rhizon samplers of a hydrophilic, porous polymer tube. The porous polymer tube was inserted into the sediments, and the opposite end was connected to a 20 mL syringe. After approximately 1~2 h, 10~19 mL pore water was collected in each syringe. The pore water was divided in half and then preserved in two 10 mL glass vials, and N₂ was used to remove the air in the headspace. 10 μL saturated HgCl₂ solution was injected into one of the vials for dissolved inorganic carbon (DIC), δ¹³C and δD analyses. While another one for the anion and cation concentration analyses was added to 0.1 mL 8 mol/L nitric acid to prevent redox and precipitation reactions. Pore water samples were preserved at 4°C until laboratory analysis. The wet sediments were taken via cut-off syringe immediately after each section was sliced and quickly added to 50 mL glass serum vials containing 10 mL saturated sodium chloride solution, which were stoppered and crimp-sealed with butyl rubber stoppers to minimize gas loss (Etiope et al., 2007; Golding et al., 2013).

The pore water samples (0.5 mL) for measurement of DIC were treated with H₃PO₄ in glass vials at 26°C. The released CO₂ was purged with He and injected into the mass spectrometer through which δ¹³C values were measured. δ¹³C were reported using the conventional delta notation per mil (‰) against Vienna Pee Dee Belemnite (VPDB) international standard. The analytical precision of this method was estimated to be < 0.5% and ± 0.1‰ for DIC concentrations and δ¹³C_DIC values, respectively. δ¹³C_DIC were measured using a Thermo MAT 253 continuous flow isotope ratio mass spectrometer attached to GasBench II. The external precision was typically better than 0.20‰.

Before the ion analyses, 0.5 mL pore water samples were filtered through 0.22μm filter membranes and diluted 1:100 or 1:20 with ultra-pure water. The concentrations of dissolved cations (K⁺, Ca²⁺, Mg²⁺, Sr²⁺and Ba²⁺) in pore water were measured using an IRIS INTREPID II XSP ICP-AES (produced by Thermo Fisher Scientific) with an analytical precision better than 2%~5%. The dissolved anions (Cl^- and SO₄ ^2-) were determined using an ICS-3000 ion chromatography (Thermo Fisher Scientific Dionex). Parallel measurements show that the analytical accuracy was better than 0.5% (n = 5).

Concentrations of methane and carbon dioxide in the headspace were measured using a Thermo Fisher Scientific Ultra Trace gas chromatograph equipped with a 30 m PLOT Q column (0.32mm ID) and a thermal conductivity detector. Air contamination during sampling was removed from gas totals. The precision of the measurements within on standard deviation was less than 3.0%. Carbon stable isotope ratios of headspace methane and carbon dioxide were measured using a Thermo MAT 253 continuous flow isotope ratio mass spectrometer (GC Isolink–IRMS) attached to a gas chromatograph (Ultra Trace GC). The external precision was typically better than 0.20‰ VPDB for δ¹³C based on long-term measurement of carbon dioxide standards (RM8562, RM 8563, and RM8564).

Mixed benthic foraminifera and shell from seven samples were picked for accelerator mass spectrometry (AMS) ¹⁴C dating at the Beta Analyses Company, USA. All radiocarbon dates were calibrated to calendar years before the present (cal. BP, 0 cal. BP = AD 1950, at 95.4% probability) using the latest version of the correction program IntCal 13 and Marine 13.

There were two peaks in the methane content profiles consistent with the absence of sulfate ( Figure 2 ). In YS 4, the first methane layer from 8 meters below seafloor (mbsf) to 11 mbsf, with the maximum value of 0.24 mmol/L. The second layer from 25 mbsf to 49 mbsf, with the maximum value of 1.73 mmol/L. In YS 7, the first methane layer from 4 mbsf to 11 mbsf, with the maximum value of 1.12 mmol/L. The second layer from 31 mbsf to 49 mbsf, with the maximum value of 1.96 mmol/L. The CO₂ content fluctuated sporadically, with a limited overall concentration change range, ranging from 0 mmol/L to 0.24 mmol/L in YS 4 and 0 mmol/L to 0.09 mmol/L in YS 7. The SO₄ ^2- concentration profile was showed in Figure 2 , which shows an opposite trend to the methane content, with low levels when methane content was high, and high levels when methane content was low.

Similar vertical variations of the stable carbon isotope ratios of CH₄ and CO₂ was presented in Figure 3 . In YS 4, δ¹³C_CH4 and δ¹³C_CO2 values increased from 25 mbsf to 38 mbsf and then decreased with the depth increase. The maximum values were -78.91‰ and -11.66‰, respectively. In YS 7, the δ¹³C_CH4 and δ¹³C_CO2 values increased from 30 mbsf to 38 mbsf and then decrease until 49 mbsf, with the maximum values of -77.81‰ and -11.02‰, respectively. The δ¹³C_CH4 increases gently below SMTZ3 and shows a parallel trend with δ¹³C_CO2, and shows low variability at individual sites. In YS 4 and YS 7, δ¹³C_CH4 at depths of 24 mbsf to 40 mbsf and 31 mbsf to 46 mbsf were very approaches to the trend of δ¹³C_CO2 values, respectively.

The effect of AOM and SR leads to the increase of HCO₃ ^- concentration in pore water. At the same time, Ca²⁺, Mg²⁺ and other cations were rapidly consumed and precipitate the carbonate minerals (Ritger et al., 1987; Rodriguez et al., 2000; Bayon et al., 2007; Nöthen and Kasten, 2011). The limited solubility of BaSO₄ causes AOM and SR to deplete sulfate, enhancing barite dissolution in sediment and resulting in a rapid rise in dissolved Ba²⁺ concentration in pore water (Torres et al., 1996; Dickens, 2001).

The Ca²⁺ concentration in YS 4 ( Figure 4 ) and YS 7 ( Figure 5 ) began to gradually increase at SMTZ3, which may be related to the dissolution of CaSO₄, carbonate, and other minerals. The change of Mg²⁺ concentration was consistent with the conservative element Cl^-, indicating that the vertical change of Mg²⁺ concentration in pore water was mainly affected by the sources of water, but was less affected by AOM. The vertical profile of Ba²⁺ concentration and SO₄ ^2- concentration exhibited a reciprocal relationship, indicating biogeochemical processes like sulfate depletion and barite dissolution due to AOM. In addition, the methane content of the first methane-bearing layer in YS 4 was very low, with the initial peak in Ba²⁺ concentration observed at ~ 2 mbsf, shallower than the current SMTZ at ~ 6 mbsf.

DIC in pore water mainly includes HCO₃ ^-, CO₃ ^2-, and CO₂ (Millero, 1995). The AOM (CH₄ + SO₄ ^2- → HCO₃ ^- + HS^- + H₂O) and OSR (2 CH₂O + SO₄ ^2- → 2 HCO₃ ^- + H₂S) process lead to an increase in DIC concentration in pore water by producing HCO₃ ^-. But the δ¹³C_DIC values of HCO₃ ^- produced by the two pathways were significantly different. Generally, the fractionation effect of carbon isotopes in the degradation process of OM was small, resulting in the δ¹³C_DIC value of OSR reaction being equivalent to the δ¹³C value of OM (about -26‰~-16‰) (Kim et al., 2011; Wu et al., 2016). ¹²CH₄ were used preferentially by microorganisms (AOM), resulting in δ¹³C_DIC values usually less than -20‰ or even lower (Alperin et al., 1988; Whiticar, 1999; Conrad, 2005).

The variation of DIC content in pore water showed an inverse relationship with SO₄ ^2- distribution from top to bottom. SMTZs with high DIC concentrations (about 8~20 mmol/L) and low values of δ¹³C_DIC (about -15‰~-22‰) indicate the presence of AOM in the study area. The higher δ¹³C_DIC value of pore water in SMTZ, such as the lowest value of -22‰ at SMTZ3 in YS 4, may be attributed to the influence of OSR and AOM, or the diffusion of ¹³C-enriched DIC from carbonate dissolution into SMTZ.

The SMTZ was a biogeochemical reaction zone where methane from seafloor sediments reacts with sulfate in the pore water due to microbial activity (Reeburgh, 2007). This process leads to the consumption of both methane and sulfate. In a “typical” SMTZ, methane diffuses upwards while sulfate diffuses downwards, whereas in an inverse SMTZ, the opposite occurs (Reeburgh, 1976; Iversen and Jorgensen, 1985; Blair and Aller, 1995; Borowski et al., 1996; Jørgensen et al., 2006; Zindorf et al., 2019). Analysis of pore water data in the study area revealed distinct vertical profiles showing both typical and inverse SMTZ distributions. The consistent levels of Cl^- and K⁺ in the pore water suggest that differences in sulfate content between sites were primarily influenced by organic matter sulfate reduction and varying levels of AOM.

The shallow gas in the study area was mainly found in late Pleistocene to Holocene sediments, as indicated by AMS ¹⁴C dating. Sea levels fluctuated significantly from late Pleistocene to Middle Holocene, but have been relatively stable since Middle Holocene, which less influence above 16 mbsf (corresponding to ~7.2 ka to present). The sedimentary environment transitioned from continental to tidal flat and then to marine environment after the last deglaciation (Liu et al., 2020; 2023). As the sea level rose, the grain size gradually decreased, showing clear organic matter grain size control characteristics.

The migration rate of the SMTZ was influenced by sedimentation rate, with higher rates promoting methane formation. Low sedimentation rates can trap SMTZ at a specific depth, leading to H₂S production through sulfate-driven AOM. Conversely, very high sedimentation rates may push SMTZ deeper, resulting in H₂S production primarily through OSR (Liu et al., 2020). During the middle and late Holocene, the East Asian winter monsoon regulated the sedimentation rate of the shaly sedimentation center. During the 15~9 ka period, the core sedimentation rate increased significantly, and this high sedimentation rate was also found in other core sediments (such as the MZ02 core) of the ECS inland shelf (Liu et al., 2011).

The dating materials from boreholes YS 4 and YS 7, along with the stratigraphic ages of the methane-bearing layers, were illustrated in Figure 6 . By integrating these results with the dating data from borehole XZK169 near the study area (Wang et al., 2006), it evident that the first methane-bearing layer of each borehole dates back to the Holocene sedimentary layer around ~6 ka, while the subsequent methane-bearing layer was situated within the late Pleistocene strata. In the Late Pleistocene and early Holocene sea levels were lower, while the water depths of the two borehole were 12.43 m and 7.57 m. By comparing the sea level and water depth, it was inferred that the study area did not transition into a tidal flat environment until the early Holocene, following which, with the stabilization and rise of sea level, the study area evolved into a shallow sea environment.

According to the correlation between the methane distribution in each borehole and the median particle size (ϕ50) of the sediments, methane mainly occurs in silt and clay sediments ( Figure 7 ). This was attributed to the high clay content in sediments, which was rich in OM and serves as a sufficient carbon source for methane production. It was important to note that the stable isotope composition changes indicate no gas migration occurred, as the gas generated was insufficient to exceed pore pressure. Furthermore, the varying grain sizes of YS 4 and YS 7 reflect significant fluctuations in hydrodynamic conditions during the deposition period.

Previous study through two SMTZ below 200 m found that sulfate reduction coupled to H₂ oxidation was likely the predominant metabolic reaction at depths with increased sulfate levels from seawater recirculated through the oceanic basement (Cox et al., 2019). Recent researches suggested that SR and AOM activities can also occur below the SMTZ, indicating either inefficient AOM or a deep-seated sulfate source, possibly linked to the re-oxidation of sulfides to sulfate (Pohlman et al., 2008; Holmkvist et al., 2011; Mazumdar et al., 2012; Monien et al., 2014; Treude et al., 2014). In the study area, the predominant oxidation process was SAOM with sulfate serving as the electron acceptor. Based on the vertical concentration profile of SO₄ ^2- in YS 7, it was segmented into three different sedimentary layers from top to bottom: