Seawater and hydrothermal fluid samples were diluted 20-fold with 2% HNO₃ (Optima grade) and spiked with indium (In) as an internal standard with a final concentration of 10 ppb. Concentrations of dissolved Fe, Mnand Bawere determined using an inductively coupled plasma mass spectrometer (ICP-MS, iCAP–RQ, Thermo Fisher Scientific) in collision cell-mode at the First Institute of Oceanography of China. The precision and accuracy of Ba, Fe, Mn measurements were verified using certified reference seawater CASS-5 (National Research Council^®, Canada), and the analytical results are shown in Table 1 . Since no Ba data are reported in certified values of the reference seawater, the analytical results reported here are compared with data published previously by (Mori et al., 2019).

Approximately 60 ng of Ba was extracted from each stock solution and spiked with a suitable amount of the ¹³⁰Ba-¹³⁵Ba double spike. Following the methods outlined by Lin et al. (2020), the spike-sample mixtures were dried, redissolved in 2.5 mol/L HCl, and purified via column separation using AG 50 W-X8 cation-exchange resin. Procedural blanks were measured for each batch, with Ba concentrations < 0.6 ng, which was <0.1% of the total Ba in the samples. The purified samples were then dried and redissolved in 3% HNO₃ (with Ba concentrations ~ 60 ppb), and Ba isotope ratios were analyzed using a Thermo Neptune XT MC-ICP-MS at the Key Laboratory of Surficial Geochemistry, Nanjing University.

Ba isotope compositions in this study are expressed as ‰ deviations (δ¹³⁸Ba) relative to the international Ba standard NIST SRM 3104a ( Equation 1 ):

The standard solution NIST SRM3104a and a seawater standard IAPSO were analyzed after every five samples. The Ba isotope compositions (δ¹³⁸Ba) of the NIST 3104a standard and IAPSO standard seawater were 0.00 ± 0.05‰ (2SD, n=11) and -0.04 ± 0.05‰, respectively. The δ¹³⁸Ba value for IAPSO in this study is consistent with the 0.05 ± 0.04‰ reported in a previous study (Pretet, 2013). We determined the error of the Ba isotope compositions (δ¹³⁸Ba) for all samples to be 0.05‰, based on the measurements of the standard SRM3104a for δ¹³⁸Ba (n=11).

The sediment samples were freeze-dried for 48 hours and then ground into a fine powder (200-mesh). For bulk concentration analysis, approximately 50 mg of sediments (weighed to a precision of 0.1 mg) were digested in a concentrated acid mixture (HNO₃/HF, 1:1). The major elements were analyzed using inductively coupled plasma optical emission spectrometry (ICP-OES, iCAP 6300, Thermo Fisher Scientific^®) with RSD < 5%, and trace elements and rare earth elements (REEs) were analyzed using ICP-MS. Sediment reference standards GBW07309 (China Reference Material) were analyzed to evaluate the reproducibility of this method (n=3). The measured values agreed well with certified values for all targeted elements (96 ± 0.6% – 105 ± 1%) and detailed information can be found in the supplementary material ( Supplementary Table S1 ).

Approximately 100 ng of Ba was extracted from each stock digest solution for Ba isotope analysis, following the same procedure as for seawater and hydrothermal samples (Section 2.3.1). The standard solution NIST SRM3104a and a coral standard SH-1 were analyzed after every five samples. The Ba isotope compositions (δ¹³⁸Ba) of the coral standard SH-1 was 0.29 ± 0.04‰ (2SD, n=8), which was consistent with the published value of 0.31 ± 0.04‰ (Lin et al., 2022).

The hydrothermal filter membrane was analyzed using scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM-EDX; FEI Quanta 200, Netherlands). A piece of the filter membrane (~ 0.5 cm × 0.5 cm) was attached to conductive adhesive using tweezers. The sample was then coated with gold for 25 seconds under vacuum conditions, after which it was placed under the scanning electron microscope for observation, with the aim of identifying as many different types of mineral particles as possible.

Vertical distribution of dissolved Ba concentrations ([Ba]_diss) at stations S1, S2 and S3 are shown in Figure 3 and Table 2 . The dissolved Ba concentrations range from approximately 37 to 107 nM, with a typical increasing pattern from surface to bottom. The concentration varied slightly within the surface 1000 m (37 – 50 nM), and increased sharply to over 80 nM at 2000 m. S3, the station located above the Tiancheng hydrothermal vent, had a higher [Ba]_diss (~ 90 nM) compared to the ones further south (S1 and S2). Barium is considered a bio-intermediate element, with its dissolved concentrations ([Ba]_diss) in the water column typically following a nutrient-like depth profile. In surface waters, [Ba]_diss is reduced due to its removal via association with marine particles. However, [Ba]_diss becomes enriched in deep water, primarily as a result of the decomposition and remineralization of settling particles (Lea and Boyle, 1991; Hsieh and Henderson, 2017; Carter et al., 2020; Yu et al., 2022).

In contrast, δ¹³⁸Ba values decrease from 0.65‰ near the surface to about 0.20‰ at greater depths ( Figure 3 ). This inverse relationship between Ba concentration and isotope composition mirrors the vertical distributions that have also been observed in the Atlantic and Pacific Oceans ( Figure 4 ) (Horner et al., 2015; Hsieh and Henderson, 2017). The observed profiles are attributed to isotopic fractionation that occurs during pelagic barite precipitation while its dissolution process shows negligible isotopic fractionation. Different from [Ba]_diss in which all three water profiles shown similar trends and values with depth, the vertical profiles of δ¹³⁸Ba for stations S2 and S3 are quite similar, and δ¹³⁸Ba for stations S1 and S2 are also exactly similar (S1: 0.68‰;S2: 0.67‰) while the ones for S1 are heavier by ~ 0.1―0.2 ‰ within the top 2000 m. At 2500 m, the δ¹³⁸Ba at S1 (0.18 ± 0.05‰) is marginally lower than at S2 (0.31± 0.05‰) and S3 (0.27 ± 0.05‰), though the difference is within analytical uncertainty ( Figure 3 ). The relationship between dissolved Ba concentrations and Ba isotope compositions (δ¹³⁸Ba) is displayed in Supplementary Figure S1 , revealing a statistically significant negative correlation (p < 0.01). In addition, the temperature and salinity of the seawater samples are plotted in the T-S diagram ( Figure 5 ). With the exception of a few surface seawater samples, all data points form distinct curves or clusters, and no extreme temperature or salinity characteristics are observed. These results indicate that the analyzed seawater samples do not exhibit significant influence from hydrothermal input.

The comparison of the vertical profile of station S2 with the ones in other ocean basins is shown in Figure 4 . Except for the vertical distribution in the Southern Ocean, [Ba]_diss in S2 in this study always lie between those from the Pacific Ocean and Atlantic Ocean, both in surface waters and deep waters. In surface waters, the [Ba]_diss are relatively higher (~ 45 nM) in the Atlantic Ocean and lower (~ 30 nM) in the Pacific Ocean, while the opposite is true in deep waters. The [Ba]_diss in deep waters are relatively higher (~130 nM) in the Pacific Ocean and lower in the Atlantic Ocean (~ 70 nM), with the ones found in this study similar to the deep water in the Southern Ocean (~ 90 nM).

While the study by Jacquet et al. (2005) on barium in the Crozet–Kerguelen Basin demonstrated that barite precipitation can induce surface water alterations in the Southern Ocean, recent investigations of dissolved Ba concentrations in water column profiles from the high-latitude Southern Ocean (Weddell Sea) have revealed that the Southern Ocean exhibits relatively high [Ba]_diss in surface waters, with minimal variation between surface and deep waters (Yu et al., 2022). This limited variation is attributed, in part, to the region’s classification as a high nutrient, low chlorophyll (HNLC) zone, where restricted barite formation leads to minor changes in surface water [Ba]_diss (Fu and Wang, 2022; Yu et al., 2022). Additionally, the strong upwelling of Circumpolar Deep Water (CDW) from below further reduces the extent of [Ba]_diss variation throughout the water column (Hsieh and Henderson, 2017; Yu et al., 2022).

Overall, deep-water [Ba]_diss follows the sequence: North Pacific > South Pacific > Southern Ocean > Southwest Indian Ocean > South Atlantic > North Atlantic. This trend correlates positively with the age of the water masses, i.e., the older the water mass, the higher the [Ba]_diss, which again is the accumulation result of the dissolution of barite in deep water.

The comparison of the δ¹³⁸Ba values of the vertical profile of station S2 with the ones in other ocean basins is shown in Figure 4 . Similar to the distribution of [Ba]_diss, the δ¹³⁸Ba found in S2 in this study is between those in the Pacific and Atlantic Oceans in deep waters. The δ¹³⁸Ba values in deep waters are relatively higher (~ 0.45‰) in Atlantic Oceans and lower in Pacific Oceans (~ 0.25‰), and the ones found in this study is ~ 0.30‰. Likewise, The Southern Ocean exhibits relatively light Ba isotope compositions in deep water, with minimal variation between surface and deep waters. The δ¹³⁸Ba values show an inverse relationship to water mass age, indicating that lighter Ba isotopes are released into seawater during barite particle dissolution.

Deep-sea hydrothermal vent fluids are in a reduced state and highly enriched in dissolved trace metals such as Fe, Mn, Cu, and Zn, with concentrations several orders of magnitude higher than those in surrounding seawater (Rouxel et al., 2008; Findlay et al., 2019; Gartman et al., 2019). When these high-temperature, reducing hydrothermal fluids mix with cooler, oxygenated seawater, most trace metals precipitate as sulfides or oxides near the vent (Rouxel et al., 2016; Yuan et al., 2018). Recently, GEOTRACES studies have shown that a portion of these metals (Fe, Zn, Cu) released from the hydrothermal vents can be stabilized as organic complexes or nanoparticles, remaining in the dissolved phase, and transported over thousands of miles (Bennett et al., 2009; Hawkes et al., 2013; Li et al., 2014; Fitzsimmons et al., 2017; Lough et al., 2019).

Dissolved Mg, Fe, Mn, Ba concentrations and Ba isotope compositions in samples collected in and near the hydrothermal vents in this study are reported in Table 3 . Dissolved Mg concentrations are commonly used to indicate the extent of seawater mixing with hydrothermal fluids. During hydrothermal fluid-rock interactions, Mg is typically removed as magnesium silicates, so that Mg concentration in hydrothermal fluids that is not mixed with seawater (termed unmixed hydrothermal fluids hereafter) should approach zero theoretically (Alt, 1995; Mottl et al., 1978). This inference is supported by Ji et al. (2017) who reported that Mg concentration in unmixed hydrothermal fluids from the Longqi hydrothermal field is 1.41 mM. Conversely, the background seawater adjacent to Longqi hydrothermal vent has a concentration of Mg of 52.5 mM (Gallant and Von Damm, 2006), significantly higher than that in unmixed hydrothermal fluids.

In this study, the Mg concentrations were found to range from 40.9 to 54.2 mM. One sample, referred to as “hydrothermal fluid” (LQ-LT), collected within the vent during the DY-78 expedition, had a Mg concentration of 40.9 mM, suggesting that mixing with background seawater had already occurred within the vent. In addition, the plume samples (LQ-1, LQ-2, LQ-3, TC-1, TC-2, TC-3, TC-4), collected near hydrothermal vents, exhibited Mg concentrations between 51.8 and 54.2 mM, indicating that these samples were highly mixed with seawater.

Despite the high degrees of mixing with background seawater, dissolved Fe and Mn concentrations in these samples were significantly higher than those reported in background seawater (Fang and Wang, 2021), indicating a notable contribution from hydrothermal sources. The relationship between Fe and Mn concentrations and their distance from hydrothermal vents is shown in Figure 6 . The results reveal that among the hydrothermal samples from the Longqi vent, the fluid sample LQ-LT collected from the vent interior exhibits the highest dissolved Fe and Mn concentrations, with concentrations rapidly declining in the remaining plume samples. All hydrothermal samples from the Tiancheng vent belong to plume samples, showing significantly lower Fe and Mn concentrations compared to the fluid sample (LQ-LT). Furthermore, Fe and Mn data from both vents indicate that only two samples (LQ-LT and TC-3) collected in close proximity to the hydrothermal vents display higher Fe concentrations than Mn concentrations. In all other plume samples, Mn concentrations decline at a slower rate relative to Fe. Previous studies have reported that during the mixing of hydrothermal fluids and seawater, dissolved Mn tends to be more stabilized in dissolved phase compared to Fe, the latter of which is removed as particles quite quickly during this process (Resing et al., 2015; Fitzsimmons et al., 2017). The distribution characteristics of these metals provide critical fluid mixing context for understanding barium isotope fractionation in hydrothermal systems.

The hydrothermal samples exhibited dissolved barium concentrations ([Ba]_diss) ranging from 93 to 869 nM ( Figure 6 ). Notably, two samples showed exceptionally high values: the hydrothermal fluid LQ-LT (869 nM) and the plume sample TC-3 (738 nM). Excluding these, the remaining of plume samples displayed a narrower concentration range (93-198 nM). Given the limited reported [Ba]diss data for the Indian Ocean, we calculated a background seawater value (~89 nM) by averaging measurements from station S2 (located away from hydrothermal vents) at depths≥2000 m. All hydrothermal samples exceeded this background value, demonstrating a clear influence of hydrothermal input on Ba enrichment.

The δ¹³⁸Ba in hydrothermal samples ranged from -0.10 to 0.27‰ and it exhibits a positive correlation with the distance from the hydrothermal vent ( Figure 6 ). The two samples with the nearest distance and highest [Ba]_diss, LQ-LT and TC-3, had lightest isotope composition -0.01‰ and -0.10‰, respectively, significantly lower than δ¹³⁸Ba values found in background seawater (0.29 ‰) in this study. The station with the third highest Ba concentration, TC-4, also had a light Ba isotope composition, being 0.05‰. The Ba isotopic compositions (δ¹³⁸Ba) of samples from the Tiancheng (TC) hydrothermal vent exhibit relatively lower values compared to those from the Longqi (LQ) vent, albeit within analytical uncertainty. Furthermore, the δ¹³⁸Ba values of TC samples display a slower rate of change with increasing distance from the vent, whereas LQ samples show a rapid increase in δ¹³⁸Ba values as distance from the vent increases. This contrast likely reflects differences in Ba isotope behavior under distinct geological settings. The Longqi Hydrothermal Field is an active high-temperature system (>300°C) hosted in mafic rocks (Tao et al., 2012). In contrast, the Tiancheng Hydrothermal Field is characterized by low-temperature hydrothermal activity (fluid temperatures ≈13.2°C; Tao et al., 2014), limited magmatic supply, and intense tectonic activity, with fractured basalt dominating the substrate (Chen et al., 2018). These geological disparities—specifically, variations in thermal regimes, host rock composition, and tectonic stress—may govern the observed differences in Ba isotope fractionation and transport dynamics between the two vent fields. Overall, the hydrothermal samples closer to the vents with high [Ba]_diss (LQ-LT, TC-3, TC-4) have a mean δ¹³⁸Ba value of -0.02‰(SD=0.07‰), significantly lower than that of background seawater (0.29‰, SD=0.04‰), which suggests that the hydrothermal vent introduce lighter Ba into surrounding seawater. The Ba isotope compositions of hydrothermal fluids from the Longqi vent field (δ¹³⁸Ba = −0.01‰) and plume samples near the Tiancheng vent field (δ¹³⁸Ba = −0.10‰ to +0.05‰) on the Southwest Indian Ridge (SWIR) align closely with the range of initial endmember vent fluids (−0.17‰ to +0.09‰) reported by Hsieh et al. (2021) for mid-ocean ridge (MOR) systems. This consistency suggests that the Ba isotopic signatures of the SWIR fluids are primarily controlled by water-rock interactions with oceanic crustal lithologies, similar to other MOR hydrothermal systems. The Longqi fluid value (−0.01‰) approaches the range of mid-ocean ridge basalts (MORBs: +0.02‰ to +0.15‰; Nielsen et al., 2018) and falls within the range of pelagic sediments (−0.21‰ to +0.11‰; Bridgestock et al., 2018), supporting Hsieh et al.’s conclusion that Ba isotopes are not significantly fractionated during high-temperature fluid-rock reactions. Notably, the SWIR data lack the extremely heavy δ¹³⁸Ba values (up to +0.91‰) observed in sediment-influenced systems like the Main Endeavour Field (MEF) (Hsieh et al., 2021). This absence reinforces Hsieh et al.’s finding that sediment interaction can elevate fluid Ba isotope ratios, a process unlikely in the ultramafic- or basalt-hosted SWIR vents. Instead, the SWIR results align more closely with basalt-dominated systems (e.g., East Pacific Rise), further emphasizing the role of host rock composition in shaping initial fluid signatures. These comparisons highlight the utility of Ba isotopes in tracing hydrothermal contributions to oceanic Ba cycling, while underscoring the need to account for local geological and physicochemical conditions when interpreting isotopic variability.

Previous studies on Ba isotopes in hydrothermal systems remain limited (Hsieh et al., 2021; Zhang et al., 2024). Hsieh et al. (2021) examined Ba isotopes in hydrothermal fluids and particulates from multiple vents. After correcting for particulate data, they reported Ba isotope compositions in endmember hydrothermal fluids ranging from -0.17‰ to 0.09‰, which are lighter than background seawater. They calculated an isotope fractionation factor (Δ¹³⁸/¹³⁴Ba _Dregs-fluid) of -0.35‰ during the mixing of hydrothermal fluids with seawater, suggesting preferential removal of lighter Ba isotopes. Modeling with this fractionation value, they further inferred that the Ba isotope composition would increase to +1.7 ± 0.7‰ when barite becomes undersaturated, indicating a potential contribution of heavier Ba isotopes from hydrothermal systems to the ocean. In contrast, Zhang et al. (2024) performed direct measurements on non-buoyant plume samples from the Rainbow hydrothermal field and found the δ¹³⁸Ba values of hydrothermally affected deep seawater exhibit a lighter isotopic signature (~ 0.3‰) compared to those found at similar depths at distant locations in the Atlantic Ocean (~ 0.45‰). Interestingly, they found that dissolved Ba undergoes conservative mixing between Rainbow hydrothermal endmember fluids and the non-buoyant plume, as evidenced by a significant linear correlation between dBa and ³He. However, the Rainbow is an ultramafic system, distinct from most mafic-hosted vents. The linear regression used for Rainbow is based on a single sample and whether the conservative Ba behavior observed in the Rainbow field also occurs in other hydrothermal systems is unknown. To better constrain the global hydrothermal Ba inputs, more observations in hydrothermal plumes are required.

In this study, we also found the hydrothermal fluid and plume samples with high concentration of Ba exhibited lighter Ba isotope compositions, ranged from -0.10‰ to 0.05 ‰, which is consistent with the result reported in the previous study (Zhang et al., 2024). The relationship between dissolved Ba concentrations and Ba isotope compositions (δ¹³⁸Ba) is displayed in Supplementary Figure S2 . One station in our samples (TC-4) exhibited a dissolved Ba concentration of 198 nM, closely matching the effective input Ba concentration of 200 nM calculated in Hsieh et al. (2021). However, the Ba isotope composition at TC-4 was 0.05‰, substantially lighter than the model-based value (+1.7 ± 0.7‰) reported by Hsieh et al. (2021). This significant discrepancy between the observed isotopic composition and the model-predicted endmember values raises a critical question: what mechanisms drive the marked deviation, particularly given the opposing trends?

It is critical to consider that Ba may be sequestered from hydrothermal fluids via coprecipitation with iron-manganese oxides and sulfide minerals, as well as adsorption onto authigenic particulates, processes that can significantly influence its isotopic fractionation (Rouxel et al., 2016; Yuan et al., 2018). Previous studies indicate that in high-temperature hydrothermal environments, anhydrite has a higher saturation index than barite, which suggests that anhydrite may precipitate firstly, thereby limiting barite precipitation (Jamieson et al., 2016). Zhang et al. (2024) also highlighted the complexity of barite precipitation and Ba removal in hydrothermal systems. Therefore, assuming the Ba removal is solely due to barite formation, as proposed by Hsieh et al. (2021), may not be correct. As reported by previous studies, trace elements are often precipitated with particles released from hydrothermal vents (Rouxel et al., 2016; Yuan et al., 2018). Analysis of filter membranes from Longqi and Tiancheng hydrothermal vents samples using Scanning Electron Microscopy and Energy-Dispersive X-ray Spectroscopy (SEM-EDX; FEI Quanta 200) revealed that, various mineral species exist in addition to barite, such as Fe oxides, FeS, FeS₂, Fe_(1-x) Sx, CeO₂, ZnS, and SiO₂, etc. ( Figures 7A ; 7B ). Therefore, barium may co-precipitate with iron-manganese oxides and metal sulfide particles in addition to forming barite; these various processes could result in different isotopic fractionation. Further studies should be conducted to investigate the Ba isotopic fractionation associated with different removal processes.

As discussed in the previous section, some elements (e.g. Fe, Mn, Cu, Zn) enriched in hydrothermal fluids typically precipitate with particles, and eventually settle to the seafloor near the vents (Liao et al., 2018). Even after alteration during early diagenesis processes, the hydrothermal elements that are enriched in sediments in general remain stable (Feely et al., 1987; Li et al., 2015; Agarwal et al., 2020). These sediments preserve geochemical information related to hydrothermal activities, offering valuable insights into complex hydrothermal systems (Dias and Barriga, 2006; Liao et al., 2018).

The measured major and minor element contents in selected sediment samples are reported in Table 4A . Visual observation on the selected surface sediment samples revealed that samples P1 to P9 are light gray, while P10 to P15 are reddish-brown, likely due to higher metal content from hydrothermal influence. The analytical results show that sediments P10–P15 have Fe contents ranging from 8.8% to 32.3%, with an average of 20.8%, significantly higher than those in sediments P1–P9 (0.4% – 3.2%, with an average of 1.21%). The Cu and Zn contents in sediments P10–P15 range from 182 μg/g to 130 mg/g (averaged to be 47 mg/g) and 74 μg/g to 33 mg/g (averaged to be 8.8 mg/g), respectively. These values are substantially higher than those in sediments P1–P9, where Cu and Zn contents range from 19 to 100 μg/g (averaged to be 52.8 μg/g) and 15 to 66 μg/g (averaged to be 40.6 μg/g), respectively.

The principal component analysis of the geochemistry data set was performed using IBM SPSS Statistics 25.0 software to further constrain the factors influencing the geochemistry data for the surface sediments in the study area ( Supplementary Table S2 ). Three major factors with eigen values greater than 1 that together explain approximately 80% of the variance of the data set were extracted. And the PCA scores and loading Plot of surface sediment geochemistry is showed in Supplementary Figure S3 . The dominant factor (Factor 1 (F1)) explains approximately 42% of the total variance and is loaded primarily with lithogenic elements, e.g., Fe, Cu, Pb, S, V, Zn, which frequently reflect the hydrothermal component. Factor 2 (F2), explaining a significant portion of the remaining variance, was predominantly associated with elevated Al and Ti concentrations, which are commonly used as lithogenic tracers. Factor 3 (F3) highlighted Ba enrichment, likely linked to marine authigenic processes, such as biogenic barite precipitation or scavenging by sinking organic matter. This tripartite structure underscores the interplay of hydrothermal activity, terrestrial sediment supply, and pelagic marine influences in shaping the geochemical signatures of the studied sediments.

The elemental ratios serve as a robust proxy for assessing hydrothermal input intensity (Qiu et al., 2023). To evaluate hydrothermal influence on sediments along the Southwest Indian Ridge, two geochemical indices were employed: (1) the metal enrichment index Al/(Al + Fe + Mn) and (2) the Fe/Ti ratio. These indices effectively differentiate hydrothermal contributions from detrital components in marine sediments.

The Al/(Al + Fe + Mn) index distinguishes metal-rich hydrothermal sediments from background sediments. Previous studies (e.g., Fisher et al., 1969; Dias and Barriga, 2006) established that background deep-sea sediments, devoid of hydrothermal input, typically exhibit values >40% due to higher detrital aluminosilicate content. In contrast, hydrothermal sediments near vent systems are characterized by lower values, reflecting Fe-Mn oxide dominance. Concurrently, the Fe/Ti ratio quantifies hydrothermal metal enrichment relative to detrital inputs, with ratios >20 indicating significant hydrothermal influence (Zhou et al., 2020). By cross-plotting these indices ( Figure 8A ), hydrothermal and detrital sediment sources can be systematically discriminated (Liao et al., 2018).

Application of this dual-index approach to 15 surface sediment samples revealed distinct geochemical signatures. Six samples (P2, P10–P13, P15) exhibited Fe/Ti >20 coupled with Al/(Al + Fe + Mn) <40%, unambiguously identifying them as hydrothermal-impacted sediments ( Figure 8A ). These results align with established thresholds and underscore the utility of combining elemental ratios to disentangle complex sedimentary provenance.

Another way to identify the significance of hydrothermal activity on surface sediments is by looking into the rare earth element (REE) composition in sediments. The hydrothermal fluids from mid-ocean ridges are characterized by Light REE (LREE) enrichment and positive Eu anomalies (Douville et al., 1999; Allen and Seyfried, 2005; Cao et al., 2012). Consequently, sediments affected by hydrothermal activity often exhibit similar characteristics (Gillis and Thompson, 1993; Liao et al., 2018; Zhou et al., 2020). The Chondrite-normalized REE patterns of the sediment samples are shown in Figure 8B . In general, all sediment samples showed a trend of LREE enrichment. Whilst, several samples (from P10 to P15), to different extents, showed apparent positive Eu anomaly ((Eu/Eu^*)_N = 0.90–3.39) compared to the value (0.71) reported by Liao et al. (2018) for the background sediments.

It is specifically noted that although the metal ratios of P14 exceed the predicted range, they fall near the threshold boundaries. Additionally, P14 exhibits elevated Fe, Cu, and Zn concentrations compared to samples P1–P9, while its Ca content is significantly lower. Visually, the sample displays a reddish-brown coloration—a characteristic indicative of hydrothermal activity. Based on this integrated evidence, we have classified P14 as hydrothermal-impacted sediment. The sediment types are explicitly detailed in Table 4B . Overall, by taking into consideration of the results, we considered that, among the 15 surface sediment samples, P10 to P15 can be identified as hydrothermal-impacted sediments, and that P1 to P9 represent background sediments.

Following sediment classification, we measured the Ba contents across all samples and selected eight representative samples for Ba isotope analysis. This subset included three background sediments (P3, P5, P7) and five hydrothermal sediments (P11 to P15). The Ba contents and its isotopic compositions in sediment samples are reported in Table 4B . The Ba content ranged from 110–1109 µg/g, and there is no apparent elevation in Ba content in hydrothermal influenced samples. Paytan and Griffith (2007) reported that the content of Ba in marine sediments can be significantly affected by dilution effects due to the accumulation of other sedimentary components.

To the contrary, the δ¹³⁸Ba values in background samples (P3,P5,P7) ranged from 0.01‰ to 0.14‰, similar to the deep-sea sediments reported in previous study (0 – 0.10‰) (Bridgestock et al., 2018). In hydrothermal-impacted sediments (P11 to P15), δ¹³⁸Ba values ranged from -0.16‰ – 0.01‰. The average δ¹³⁸Ba of hydrothermal sediments (-0.08‰, SD=0.07‰, n=5) is notably lower than that of the background sediments (0.09‰, SD=0.07‰, n=3). The relationship of Ba contents and δ¹³⁸Ba is shown in Supplementary Figure S4 . The results show that the data points do not follow a distinct fractionation trend (i.e., lighter Ba isotopes correlating with higher Ba concentrations). Instead, samples with lighter Ba isotopes exhibit both high and low Ba concentrations. We therefore propose that the Ba concentrations in hydrothermal-impacted sediments are influenced by dilution effects from co-occurring hydrothermal metal elements.

As we discussed in section 3.2.3, during the mixing process between hydrothermal fluid and seawater, lighter Ba isotopes are preferentially removed, resulting the settling particles compose lighter Ba isotope. This is consistent with our observation in the hydrothermally influenced sediment samples in this study. As conclusion, while the Ba content of hydrothermal sediments cannot be distinguished from the background sediments, the Ba isotope retains lighter isotopic signature, potentially can serve as an indicator of hydrothermal influence on sediment of the study region.