CRS has been widely used as an indicator of authigenic pyrite content since the method was introduced and continuously advanced (Canfield et al., 1986). In this study, from 181 to 481 cmbsf and below 1,101 cmbsf, the contents of CRS were significantly higher throughout the core, and the δ³⁴S values of the CRS showed a slightly increasing trend (Figure 6). The content of CRS at Z22–3 varied greatly but ranged from 0.2 to 0.5 wt% within SMTZs, which displayed an abnormally high value. The CRS contents in SMTZs of other nearby seepage sites were 1.01–2.60 wt% (Site 2A; Lin et al., 2016), 1.04–1.65 wt% (W19-15; Gong et al., 2022), and 0.38–1.57 wt% (973-4; Liu et al., 2020). The CRS content of our study site was slightly lower than the contents of the aforementioned sites, but still within a reasonable range. Consequently, we suggest that SD-AOM was the main factor causing the increase in CRS and the sulfur isotopic composition of pyrite (Figure 6). However, we also found from our study that the value of δ³⁴S_CRS in SMTZs only showed inconspicuous positive bias with an existing extremely negative value ranging from −47.69‰ to −41.0‰. Such a negative value of δ³⁴S_CRS can also be found in other gas hydrate-bearing areas like the Gulf of Mexico, Southwestern Taiwan Basin, Qiongdongnan Basin, East Coast of India, and Pearl River Basin (Peketi et al., 2012; Formolo and Lyons, 2013; Li et al., 2016; Lin et al., 2016; Zhang M. et al., 2018; Miao et al., 2021b), which indicates that when using the sulfur isotope of authigenic pyrites as an indicator of SMTZs, more specific constraints are needed.

According to previous studies, due to strong SD-AOM, the values of δ³⁴S_CRS can gradually converge to the seawater values or even higher values. However, the extremely negative value of δ³⁴S can be attributed to environmental and depositional situations. 1) “Open” condition: Under the “open” sedimentary condition, the exchange between pore fluids and seawater sulfate is more frequent, so the ³²S consumed can be sufficiently supplemented, and the residual ³⁴S-enriched sulfate pool is not effectively utilized (Fike et al., 2015; Peketi et al., 2015). Low sedimentary rate can contribute to the “open” condition by promoting the communication between pore water and underlying seawater, thus enlarging the sulfur isotopic fractionation (Liu et al., 2019, 2021). 2) Limited iron supply: Reactive iron supply is also important to control the sulfur isotope of pyrites. If iron is available at depths in sediments, it can react with the ³⁴S-enriched sulfide and form high δ³⁴S in pyrite (Spruzen et al., 2024). However, if the reactive iron is limited, it can only react with sulfide near the surface, which is not yet rich in ³⁴S, thus forming negative δ³⁴S_py (Formolo and Lyons, 2013). 3) Disproportionation of sulfur intermediates: According to mass balance studies (Markovic et al., 2016; Yao et al., 2020), much of the sulfide produced by microbial sulfate reduction is re-oxidized, resulting in the disproportionation of sulfur intermediates. Canfield and Thamdrup (1994) suggested that the repeated sulfur reduction and disproportionation can increase the sulfur isotopic fractionation between dissolved sulfate and sulfide. The value of sulfur isotope above 181 cmbsf in Z22–3 ranged from −47.7‰ to −18.9‰, much lower than the value of seawater sulfate (≈21‰; Rees et al., 1978). Here, we consider that the open environment and the disproportionation of sulfur intermediates contribute most to the low value of sulfur isotopes in pyrites.

The existence of authigenic carbonate minerals is an important piece of evidence for methane seepage (Bayon et al., 2011; Feng et al., 2014; Bayon et al., 2015). The enhanced AOM can increase the alkalinity in pore water to facilitate the precipitation of authigenic carbonate minerals such as calcite, dolomite, and aragonite (Baker and Burns, 1985; Raiswell, 1988). They exist in the form of carbonate rocks on the sea floor or dispersed carbonate minerals in the sediment column (Han et al., 2014; Yang et al., 2014); however, it is hard to detect massive carbonate rocks because they only exist when the seepage intensity is moderate. Bayon et al. (2007) showed that Sr/Ca and Mg/Ca ratios can be used for identifying Sr-rich (aragonite) and Mg-rich (high-Mg calcite and dolomite) carbonate layers in cold seepage sediments. From 181 to 481 cmbsf and from 1,101 cmbsf to the bottom of the core, the marked increase in Sr/Ca and Mg/Ca may indicate the formation of authigenic carbonate minerals like aragonite and calcite. The simple mixing model of four end-members (aragonite, high-Mg calcite, biogenic calcite, and detrital fraction) can be used to identify the proportion of carbonate mineral components. The four-component end-member model was used to calculate the carbonate mineral composition of Z22–3 in this paper, and the samples were plotted in the model using the content of Sr/Ca-Mg/Ca. It was found to fall within the range defined by the model and was distributed on both sides of the high-Mg calcite–aragonite mixing curve. This indicated that the authigenic carbonate minerals in the sediments mainly consisted of aragonite and high-Mg calcite, which was very similar to gas hydrate-bearing sites such as the northern South China Sea (Chen et al., 2016) and Qiongdongnan Basin (Miao et al., 2021a) (Figure 7). Therefore, it was believed that the abnormal Mg/Ca-Sr/Ca in the 181–481-cm and 1,101–1,401-cm layers was related to the carbonate minerals produced by methane seepage. However, different from the value of Mg/Ca-Sr/Ca, the CaO value within the two identified layers showed a lower value in the sediment core. This may be attributed to the low level of authigenic carbonate minerals because of the alteration caused by extra H₂S. The hypothesis still needs sufficient mineralogical evidence to be validated.

AOM caused by gas hydrate dissociation can alter the redox conditions of sediments and underlying seawater, leading to the enrichment of elements like Mo, Cu, and U (Tribovillard et al., 2006; Scott and Lyons, 2012; Smrzka et al., 2019a). These redox-sensitive elements, especially Mo and U, are commonly used to record the paleo-sedimentary environment and seawater chemistry, owing to their stability under oxidizing conditions and their susceptibility to being reduced and transferred to sediments in anoxic water. In oxic seawater, Mo and U primarily exist as MoO₄²⁻ and U⁶⁺ adsorbed onto Mn oxyhydroxides. In anoxic water, MoO₄²⁻ and U⁶⁺ are converted into thiomolybdates (MoS_(4−x)O_x²⁻) and U⁴⁺, respectively, when the concentration of H₂S reaches a critical level of approximately 10–100 μM (Morford and Emerson, 1999; Tribovillard et al., 2006). At this point, Mo and U are quickly sequestered into organic fraction or precipitated with Fe–Mo sulfide minerals and transferred to sediments (Helz et al., 1996; Tribovillard et al., 2006). As a result, Mo enrichment in marine sediments can occur in two distinct ways: 1) a transient enrichment associated with the recycling of Mn oxyhydroxides at or near the sediment–water interface and 2) a permanent enrichment in the sulfidic zone, where MoS_(4−x)O_x²⁻ forms and is sequestered (Scott and Lyons, 2012). Mo was apparently enriched in the 181–481-cmbsf and 1,101–1,401-cmbsf layers to a certain extent (Figure 6). (Mo/U)_EF mostly fell on the diagonal line at 0.3 times the present-day seawater, which indicated a suboxic–sulfidic sedimentary environment (Figure 8). Chen et al. (2016) proposed that OSR provided the release of dissolved sulfide and that AOM activated the Mo enrichment in sediments during seepage events. High methane flux induced sulfidic conditions, which allowed the capture of Mo into pyrites through the partition shuttle process (Sato et al., 2012; Bazzaro et al., 2020). These could properly illustrate the enrichment of Mo as well as the depletion of elements like Cu within the 181–481-cmbsf and 1,101–1,401-cmbsf layers (Figures 3, 9).

Trace element distribution of authigenic pyrites reflects the geochemical conditions during the growth of mineral crystals (cf. Rickard, 2012). Compared to their concentrations in seawater, most trace elements in pyrite analyzed herein showed various degrees of enrichment (Figure 3; Table 1). Co, Ni, and Mo exhibited similar patterns for core Z22-3, with higher contents and greater variation of these trace elements in pyrite and lower contents in sediments (Figure 3). Although Co, Ni, and Mo showed great enrichments in pyrites in SMTZs, the enrichment factors of bulk sediments of these elements are not elevated in the same depths. Thus, the element patterns of pyrites may be attributed to early diagenesis rather than the primary trace element contents of the host sediments (Lin et al., 2022).

Concerning the mechanism of Mo and the enrichment of other trace metals, two hypotheses have been brought forward: 1) the trace metals were carried to the surface sediments by ascending seepage fluids, resulting in trace metal enrichment in shallow pore fluid and sediments (Nath et al., 2008; Cangemi et al., 2010); 2) the trace metals were predominantly derived from seawater, scavenged by iron and manganese oxyhydroxides in the water column, and then transferred to the surface sediments (Tribovillard et al., 2013). In this study, the latter hypothesis was more consistent with the characteristics of trace metals in sediments at Z22-3, as other trace metals carried by chloride-rich ascending brine (such as Cu, Ni, and Zn) showed greater enrichment than Mo (Figure 3). Tribovillard et al. (2013) proposed that trace elements like Mo, As, and Sb were thought to be actively transferred to sediments through the particle shuttle process by Fe and Mn phases. Then, Mo could be trapped by pyrite formed during early diagenesis, and Mo was reported to have the highest degree of trace metal pyritization (DTMP; Huerta-Diaz and Morse, 1992; Burdige, 2007) among other trace metals, which means that Mo was highly prone to incorporation into pyrite during early diagenesis. In the OSR-affected zone (0 to 181 cmbsf), the contents of Mn in the sediments significantly decreased with increasing depth, while the contents of Mo increased; in the 481–1,101-cmbsf layers, the Mn contents of the sediments at the study site were low and constant, and Mo revealed much higher and more variable contents. In the SD-AOM-affected zone, the enrichment factors of Mo in sediments slightly decreased as the Mo contents in pyrite apparently increased. With the higher H₂S concentrations occurring in seawater (H₂S ≥ 0.011 mM), the chemical reaction of Mo thiolation is given as follows (Equation 3) (Erickson and Helz, 2000):

Therefore, by combining these results with those of previous studies, we suggest that Mo was most probably adsorbed and carried by iron and manganese oxyhydroxides. Then, in the OSR-affected zone, Mo was dissolved and turned into MoS₄²⁻, entering pyrite during early diagenesis under the effect of SD-AOM.

Trace elements such as Ni, Co, As, and Se are commonly lattice-bound within pyrite (Vaughan and Rosso, 2006). Co and Ni commonly substitute for ferrous in pyrite due to their similar ionic radii and charge (Huerta-Diaz et al., 1998; Tribovillard et al., 2006). The ratio between Co and Ni (Co/Ni) in pyrite has been effectively used to investigate the evolution of geochemical conditions during pyrite formation (e.g., Dill and Kemper, 1990; Maslennikov et al., 2009). The Co and Ni contents of the two discussed zones in our study area are plotted in Figure 10, and the slope of the linear trends represents the ratio between Co and Ni (Co/Ni). The OSR-derived pyrite (with a slope of up to 0.44) revealed a higher Co/Ni ratio than the SD-AOM-derived pyrite (small slope of 0.17), but both were higher than the ratio of seawater (ca. 0.004; Bruland et al., 2014). The elevated Co/Ni ratio of pyrite from the OSR-affected zone of the study site indicated stronger enrichment of Co compared to Ni in pyrite at this depth. Ni could be more easily incorporated into pyrite than Co under highly sulfidic conditions (Smrzka et al., 2019a, 2020), in accord with enhanced SD-AOM as a cause for the decreased Co/Ni ratio for the AOM-affected zone. With respect to the Ni and Co contents, both revealed detectable enrichment in SD-AOM-derived pyrites compared with OSR-derived pyrites. A convincing hypothesis indicates that Ni and Co are dispensable metals of enzymes required in the reverse methanogenesis pathway (Hallam et al., 2003; Scheller et al., 2010; Glass and Orphan, 2012); thus, Ni and Co may be taken up through reverse methanogenesis and then accumulate in pyrites.

In oxic marine environments, Mn typically coexists with metastable Fe–Mn oxyhydroxides (Bayon et al., 2004; Smrzka et al., 2019a, 2020). Due to the strong sorption capacity, Fe–Mn oxyhydroxides can adsorb most ions in the water column or the pore water, such as Mn, Pb, Zn, Cu, and V (Bayon et al., 2004; Lin et al., 2022; Miao et al., 2022). However, the adsorbed trace elements are released into pore water when iron and manganese oxyhydroxides are reductively dissolved by hydrogen sulfide in the sediment (Equations 4, 5) (Berner, 1981; Burdige and Nealson, 1986; Aller and Rude, 1988; Canfield and Thamdrup, 2009):

In this study, we used CRS and sulfur isotopic composition to constrain the depths of SMTZs, and we used (Mo/U)_EF to indicate the sulfide conditions, which favor the reductive dissolution of iron and manganese oxyhydroxides and the release of trace elements. No enrichment of elements Mn, Pb, Zn, Cu, and V was observed for bulk sediments (Figure 9), and reduced contents of Mn, Zn, Cu, and V in pyrite from both the OSR- and SD-AOM-affected zones were identified (Figure 3). Some studies have suggested that the water exchange reaction kinetics of metals influence mineral precipitation (Huerta-Diaz and Morse, 1992; Morse and Luther, 1999). Trace elements, such as Cu, Pb, and Zn, which have faster water exchange kinetics rates than Fe, tend to precipitate as independent sulfide rather than incorporate into pyrite. It is likely that at Z22-3, Mn, Cu, and Zn were released and accumulated when SD-AOM rates were high and then precipitated as independent metal sulfides. It is suggested that the weakening of the OSR can limit the release of Zn and Cu in the methane seepage environment, which also contributes to the depletion of pyrites (Yu et al., 2025). However, the enhancement of Pb content in pyrite may indicate that the metal sulfide of Pb has further incorporated into pyrite. V is usually not present in sulfide minerals such as pyrite, and dissolved V in pore water is reduced and precipitates in the insoluble V phase in the presence of hydrogen sulfide (Emerson and Huested, 1991; Wanty and Goldhaber, 1992). In the OSR-affected zone (481 to 1,101 cmbsf), the contents of Mn, Zn, Cu, and V in the pyrites were also very depleted (Figure 3). This is because OSR also leads to a substantial solution of iron and manganese oxyhydroxides, making it difficult to use these depleted elements as proxies to identify SD-AOM and OSR.