Geological, physical, and chemical characteristics of seafloor hydrothermal vent fi elds*

ZENG Zhigang , CHEN Zuxing ZHANG Yuxiang LI Xiaohui

1 Seafloor Hydrothermal Activity Laboratory, CAS Key Laboratory of Marine Geology and Environment, Institute of Oceanology, Chinese Academy of Sciences, Qingdao 266071, China

2 Laboratory for Marine Mineral Resources, Qingdao National Laboratory for Marine Science and Technology, Qingdao 266061, China

3 University of Chinese Academy of Sciences, Beijing 100049, China

4 Center for Ocean Mega-Science, Chinese Academy of Sciences, Qingdao 266071, China

Abstract Seafloor hydrothermal vent fi elds (SHVFs) are located in the mid-ocean ridge (MOR), backarc basin (BAB), island arc and hot-spot environments and hosted mainly by ultramafi c, mafi c, felsic rocks, and sediments. The hydrothermal vent fluids of SHVFs have low oxygen, abnormal pH and temperature, numerous toxic compounds, and inorganic energy sources, such as sulfuric compounds, methane, and hydrogen. The geological, physical, and chemical characteristics of SHVFs provide important clues to understanding the formation and evolution of seafloor hydrothermal systems, leading to the determination of metal sources and the reconstruction of the physicochemical conditions of metallogenesis. Over the past two decades, we studied the geological settings, volcanic rocks, and hydrothermal products of SHVFs and drawn new conclusions in these areas, including: 1) the hydrothermal plumes in the Okinawa Trough are affected by the Kuroshio current; 2) S and Pb in the hydrothermal sulfi des from MOR are mainly derived from their host igneous rocks; 3) Re and Os of vent fluids are more likely to be incorporated into Fe- and Fe-Cu sulfi de mineral facies, and Os is enriched under low-temperature (＜200°C) hydrothermal conditions in global SHVFs; 4) compared with low-temperature hydrothermal sulfi des, sulfates, and opal minerals, high-temperature hydrothermal sulfi des maintain the helium (He) isotopic composition of the primary vent fluid; 5) relatively low temperature (＜116°C), oxygenated, and acidic environment conditions are favorable for forming a native sulfur chimney, and a “glue pudding” growth model can be used to understand the origin of native sulfur balls in the Kueishantao hydrothermal fi eld; and 6) boron isotope from hydrothermal plumes and fluids can be used to describe their diffusive processes. The monitoring and understanding of the physical structure, chemical composition, geological processes, and diverse organism of subseafloor hydrothermal systems will be a future hot spot and frontier of submarine hydrothermal geology.

Keyword: vent fi elds; hydrothermal products; volcanic rocks; vent organisms; seafloor hydrothermal systems

1 INTRODUCTION

Seafloor hydrothermal vent fi elds (SHVFs) contain diverse vent fluids, hydrothermal sulfi de deposits, hydrothermal plumes, metalliferous sediments, biological species. Volcanic rocks create a hydrothermal environment with alterations involving components of geological, physical, chemical, and biological variations (Hannington et al., 2005; Zeng, 2011; Humphris and Klein, 2018). Hydrothermal sulfi de deposits with vent fluids, hydrothermal plumes, metalliferous sediments, biological species, and volcanic rocks provide new windows for understanding subseafloor fluid and magma processes, as well as the impact of seafloor hydrothermal activities on seawater, sediment, and ecological environments (Von Damm, 1995; Glasby and Notsu, 2003; Hrischeva et al., 2007; Zeng et al., 2017a).

Material transport and heat budget processes of seafloor hydrothermal activities in SHVFs and the associated controlling mechanism for forming hydrothermal products (HPs) (e.g., vent fluid, hydrothermal plumes, hydrothermal sulfi des, metalliferous sediments, altered volcanic rocks, organisms) remain unclear. Subseafloor fluid circulation and its physical and chemical variations are not well recognized (Tivey, 2007; Humphris and Klein, 2018). For example, the quantitative impacts of fault structure, magmatism, fluid-rock interaction, sediment, and seawater on the formation and preservation of HPs and environments remain poorly understood and the subseafloor structure and material composition of SHVFs are unconstrained, which seriously restricts the determination of the formation mechanism and fluid conditions of HPs, ecological environment, and the potential of biology resources. Shallow (＜1 m), medium (1–10 m), and deep (＞10 m) drilling and sampling campaigns with hydrothermal cycle modeling should be carried out in SHVFs to (I) investigate global subseafloor hydrothermal mineralization (e.g., Petersen et al., 2005, 2014), (II) study subseafloor hydrothermal systems and their constraints and effect on HPs, seawater, rock, and ecological environment, (III) determine the oreforming mechanism of subseafloor hydrothermal sulfi de deposits, and (IV) provide scientifi c support for understanding subseafloor biological processes and ways to protect the seafloor hydrothermal environment. These broad-scale fi ndings are expected to provide accurate insights into subseafloor hydrothermal systems worldwide.

In this paper, we summarize major research advances concerning volcanic rocks, vent fluids, hydrothermal plumes, hydrothermal sulfi des, hydrothermal alteration, oxyhydroxide, metalliferous sediment, organic matter, hydrothermal organism, and shallow water hydrothermal activity as a reference for understanding the geological, physical, and chemical characteristics of SHVFs.

2 GEOLOGICAL SETTINGS OF THE SHVFS

SHVFs are located on mid-ocean ridges (MORs), back-arc basins (BABs), island arcs (IAs) and hot-spots, which are hosted mainly by ultramafi c, mafi c, and felsic rocks, and sediments (Zeng et al., 2010a, 2014a, 2015a, b).

2.1 East Pacifi c Rise between 12°N and 13°N

The East Pacifi c Rise (EPR) between 12°N and 13°N is a fast-spreading ridge (10–12 cm/a) located at the boundary of the Pacifi c plate and Cocos plate with axial graben, marginal high, and seamount. The bathymetry changes symmetrically and becomes gradually deeper from the middle to either sides (Hékinian et al., 1983). Fresh basaltic lava have fi lled in the faults and fi ssures near the axial grabens (Gente et al., 1986). SHVFs in the EPR are hosted by midocean ridge basalts (MORBs) and distributed in the axial grabens, marginal high, and southeast seamount with highly developed faults (Fouquet et al., 1996).

2.2 Mid-Atlantic Ridge

The slow-spreading Mid-Atlantic Ridge (MAR) is divided into the North MAR (NMAR) and South MAR (SMAR). At the Atlantic-Indian Ridge near 54°S, the SMAR turns and crosses the Crozet Plateau, which continues westwards to the Scotia Ridge and eastwards to the Southwest Indian Ridge (SWIR). SHVFs in the SMAR are hosted by MORBs, where the spreading rate is approximately 3.4 cm/a (DeMets et al., 1994).

The hydrothermal sulfi de deposits in the Logatchev hydrothermal fi eld (LHF) of the NMAR are mainly hosted by serpentinized harzburgite and dunite, gabbronorite, basalt, and pelagic sediments (Rouxel et al., 2004; Petersen et al., 2009; Zeng et al., 2014a, 2015a, b).

2.3 Indian Ocean Ridge

The Indian Ocean Ridges include the Central Indian Ridge (CIR), SWIR, southeast Indian Ridge (SEIR), and northwest Indian Ridge (NWIR). For example, the SWIR located between 45°E and 70°E and 26°S and 40°S stretches from Bouvet Triple Junction to the Rodriguez Triple Junction and is a super-slow spreading ridge (＜2 cm/a) that acts as the main boundary between the African and Antarctica plates (Suo et al., 2017). The SWIR is approximately 8 000 km long with an oceanic crust thickness of 3.0 to 6.0 km (Muller et al., 1999). The volcanic rocks in the SWIR mainly consist of peridotite, gabbro, and basalt, and the axial rifts and limbs of SWIR are covered with thick sediment (＞10 m). SHVFs located in the CIR, SWIR, and NWIR. For example, the Kairei hydrothermal vent fi eld in the CIR is hosted by basalt that is adjacent to mafi c-, ultramafi c olivine-rich rocks, and the hydrothermal fluids interact with and circulate through ultramafi c rocks (Nakamura et al., 2009). Numerous hydrothermal fi elds were discovered in the SWIR, including the Duanqiao-1, Yuhuang-1, and Longqi-1 (Tao et al., 2014; Liao et al., 2019). The hydrothermal fluids in the Longqi-1 active vent fi eld in the SWIR undergo a long reaction path involving both mafi c and ultramafi c lithologies (Tao et al., 2020).

2.4 Back-arc basins

2.4.1 Okinawa Trough

The Okinawa Trough (OT) is located in the western Ryukyu arc-trench system and extends from northeast of Taiwan, China, to southwest of Kyushu. The OT developed on the eastern edge of the Eurasian continental lithosphere and is a nascent BAB in the western Pacifi c. The OT is characterized by the development of normal faulting of transitional crust (atypical crust with mantle-derived material), extremely high heat flow values (up to 400–600 mW/m2; Yamano et al., 1989) and frequent magma intrusions (includes basic to silicic magmas), which provides a favorable geological environment for the development of seafloor hydrothermal sulfi des in SHVFs (Sibuet et al., 1998; Ishibashi et al., 2015). The OT is divided by the Tokara and Kerama faults into northern (NOT), middle (MOT), southern (SOT) sections (Shinjo and Kato, 2000), and the topography changes remarkably from the MOT to SOT.

By 2016, at least 15 SHVFs have been reported in the OT based on the InterRidge data base, including the Minami-Ensei, Iheya North, Clam, Jade, Hakurei, Hatoma, Yonaguni Knoll IV, and Tangyin hydrothermal vent fi elds (Zeng et al., 2017a). The Iheya North knoll hydrothermal fi eld located in the Iheya North knoll volcanic complex in the MOT has a water depth of approximately 1 000 m and is hosted by pumiceous volcanic clasts, hemipelagic sediments, and hydrothermally altered volcanogenic breccias (Ishibashi et al., 2015). The Clam hydrothermal fi eld is located in a small depression on the northern slope along the eastern part of the Iheya Ridge in the MOT (Ishibashi et al., 2015). The Yonaguni Knoll IV hydrothermal fi eld is situated in an elongated valley with dimensions of approximately 1 000 m×500 m and mostly covered by sediment, except for the northern slope and active SHVF (Suzuki et al., 2008). The Tangyin hydrothermal fi eld is located on the top of the Yuhua Hill seamount and is hosted by a felsic volcanic basement with patches of sediment adjacent to a submarine canyon (Zeng et al., 2017a).

2.4.2 Eastern Manus Basin

In the eastern Manus Basin (EMB), volcanic ridges located within a remnant of island-arc crust. Lavas erupted along the ridges have variable composition including a complete series from basalt to rhyolite (Binns and Scott, 1993; Sinton et al., 2003; Hannington et al., 2005).

There are four SHVFs in or near the volcanic ridges in the EMB: the PACMANUS, DESMOS caldera, Susu Knolls, and Solwara 12 hydrothermal vent fi elds. The PACMANUS hydrothermal vent fi eld, located on the Paul ridge, is notable for its distinctly siliceous volcanic host rock (Binns and Scott, 1993; Zeng et al., 2012a). The DESMOS caldera hydrothermal fi eld is located on the southeast ridge in the EMB and hosted by basaltic andesite (Gamo et al., 1997; Park et al., 2010). Abundant CO2and excess F in the DESMOS caldera vent fluid indicate magma degassing. The DESMOS caldera basaltic andesite is altered by interaction with hot acidic fluid originating from the mixing of magmatic fluid and seawater (Gena et al., 2001). The Susu Knolls hydrothermal fi eld is in the eastern most part of the EMB and hosted by porphyritic dacitic volcanic rocks, forming three conical peaks informally known as North Su, South Su, and Suzette (Hrischeva et al., 2007). The Solwara 12 hydrothermal fi eld is located 25 km west-northwest of Solwara 1 on the southeastern edge of the DESMOS caldera hydrothermal fi eld. There is extensive sediment cover in the Solwara 12 hydrothermal fi eld, and the mapped hydrothermal chimney fi eld is approximate 200 m across and includes clusters of old hydrothermal sulfi de deposit.

2.4.3 North Fiji Basin

The Sonne 99 hydrothermal vent fi eld is located in the North Fiji Basin (NFB) and hosted by volcanic rocks. The volcanic rocks in the NFB include normal MORB (N-MORB) and ocean island basalt (OIB)- related enriched MORB (E-MORB) (Eissen et al., 1994; Nohara et al., 1994; Koschinsky et al., 2002; Kim et al., 2006; Zeng et al., 2017b).

2.5 Shallow-water hydrothermal vent fi eld

The Kueishantao islet is a volcanic center along the southwestern tip of the OT in the west Pacifi c, located about 10.8 km from Wushi Harbor at Toucheng Town on the Ilan plain. The last major volcanic eruption in the Kueishantao Islet occurred around 7 ka (Chen et al., 2001). The Kueishantao hydrothermal vent fi eld (KHF) (121°55′E, 24°50′N, approximately 0.5 km2) is situated southeast of the Kueishantao islet (Zeng et al., 2007, 2011) and hosted by andesite, lava flows, and pyroclastics (Chen et al., 2005a, b; Zeng et al., 2013).

3 VOLCANIC ROCKS OF SHVFS

It is well known that the basement rocks of submarine hydrothermal fi elds signifi cantly impact the chemical characteristics of hydrothermal productions, such as sulfi de deposits and vent fluids (Hannington et al., 2005; Zeng, 2011; Humphris and Klein, 2018). The study of volcanic rocks from SHVFs is therefore important for understanding the formation of submarine hydrothermal systems.

3.1 Volcanic rocks of SHVFs in the OT

Most SHVFs in the OT are host by felsic volcanic rocks, and one hydrothermal fi eld is hosted by basalt (Ishibashi et al., 2015).

3.1.1 Origins and evolution of felsic volcanic rocks

The OT is characterized by widespread felsic volcanic rocks with ages of less than 1 Ma (Shinjo and Kato, 2000; Huang et al., 2006; Chen et al., 2018), which is distinct from basalt-dominated intra-oceanic BABs. At present, the origin of the felsic volcanic rocks (rhyolites) in the OT remains uncertain. Type-1 rhyolites in the MOT are produced by re-melting of andesites with residue amphiboles, whereas type-2 rhyolites in the MOT are derived from the re-melting of andesites without residual amphiboles (Zhang et al., 2018a, 2020). The crustal rocks or melts with compositions analogous to those of the andesites from Kueishantao or the upper crustal rocks of southwest Japan might have contaminated the rhyolitic magma in the MOT, suggesting that the OT crust probably contain isotopically enriched crustal materials (Zhang et al., 2020). Nevertheless, both the fractional crystallization of basaltic magma and partial melting of andesites cannot generate melts with a SiO2content of 62–68 wt.%, which might shed light on the origin of the compositional gap for bimodal magmatism in the MOT (Zhang et al., 2018a). In contrast, magma mixing plays an important role in controlling the origins of silicic magmas in the southwest OT (Chen et al., 2018, 2019), and the parent magma experienced a multilayer magma chamber systems before eruption (Guo et al., 2018).

3.1.2 Origins and evolution of mafi c volcanic rocks

The OT mafi c magmatism has been influenced by subduction components (Guo et al., 2017; Shu et al., 2017; Zhang et al., 2018d), which results in the different geochemical imprint of young basaltic rocks. For example, the magma sources of the SOT basalts were principally influenced by subducted fluids and sediments. The magma sources of the MOT basalts were impacted by subducted fluids from both altered oceanic crust and sediment. The geochemical characteristics of the SOT and MOT basalts are variable owing to different Wadati-Benioff depths and tectonic formation environments (Guo et al., 2017). Sediment fluxes account for the Tl isotopic variations in the OT volcanic rocks and require sediments subducted in the range of ＜1%, 0.1%–1%, and 0.3%–2% from the depleted mantle source to account for the Tl of volcanic rocks from the northern, central, and southern portions of the Ryukyu arc and OT. Bulk sediment mixing is required for the generation of the volcanic rocks from the Ryukyu arc and OT (Shu et al., 2017).

3.2 Volcanic rocks of the SHVFs in the Mid-Ocean Ridge

3.2.1 Geochemical and isotopic analyses of MORBs from EPR 1°S–2°S

MORBs from EPR between 1°S and 2°S show a wide range of trace element and isotopic compositions. One depleted magma source and two enriched magma sources were proposed to contribute to the formation of MORBs from EPR 1°S–2°S. However, basalt samples 02 and 10 from the EPR between 1°S and 2°S were derived from a mixture of enriched and depleted magma sources, whereas basalt sample 07 originated from a depleted magma source that was not influenced by magma mixing (Zhang et al., 2018c).

3.2.2 Silicon and oxygen isotopes in basalts from the EPR

Silicon and oxygen isotopes are two important components in volcanic rocks that are used to trace diagenetic processes, study isotope fractionation, and identify the isotopic characteristics in MORBs from the EPR (Wang et al., 2013a). We found that δ30Si and δ18O correlate positively with SiO2content, which indicates that the SiO2content of MORBs affected the Si and O isotopic fractionation (Wang et al., 2013a).

3.2.3 Geochemical characteristics of abyssal peridotites from the SWIR

In the serpentinite-hosted SHVFs of slow-spreadingMORs, abyssal peridotites contain relics of mantle minerals, which contain primary information about melting and melt extraction processes beneath MORs. In our previous studies, abyssal peridotites from the SWIR near 65°E were comprised mainly of lizardite, chlorite, carbonate, and magnetite with minor amounts of talc, pyroxene phenocrysts, and sparse olivines. Olivine grains in abyssal peridotites from the SWIR near 65°E display exsolution lamellae, indicating the occurrence of talc reduction or decompression during seawater-rock interaction. The abyssal peridotites in the SWIR near 65°E were derived from a depleted mantle magma source that underwent partial melting. Additionally, elemental anomalies (e.g., Rb, Ba, U, Pb, Sr, and Li) and the Ce/Pb ratio imply that these abyssal peridotites in the SWIR near 65°E have been strongly altered by seawater (Zeng et al., 2012b).

Table 1 Physical characteristics and chemical compositions of seafloor hydrothermal vent fluids in different geological environments

4 HYDROTHERMAL PRODUCTS OF SHVFS

4.1 Vent fluids

The vent fluid of SHVFs shows obvious differences in its physical and chemical characteristics (Table 1; Zeng, 2011). The fluids are generally more acidic (although there are also more alkaline hydrothermal fluids, such as the Atlantic Lost City hydrothermal vent fi eld, with a vent fluid pH of 9.8; Hannington et al., 2005) at relatively high temperature (especially hydrothermal vent fluids from black chimneys) with diverse chemical composition, and are clearly affected by subseafloor geological processes and magmatism. However, according to the ejected state of the fluid, it can be divided into focused fluid and diffuse fluid. Compared with magma fluid, the temperature and pressure of hydrothermal fluid is substantially lower and its chemical composition is affected by seawater. The temperature range of vent fluid is large, ranging between 3 and 464°C (Von Damm, 1995). According to the temperature, seafloor hydrothermal vent fluids can be divided into three types: high temperature (＞300°C), medium temperature (100–300°C), and low temperature (＜100°C). The range of pH value (measured at 25°C) of the vent fluid is also large (1.52–10.6). Compared with seawater, the chemical composition of the end-member fluid has high concentrations of H2S, Fe, and Mn (Table 1). Furthermore, there are gas components in the vent fluid (CO2, H2, H2S, and CH4in bubbles) in shallowwater SHVF, and the CH4and H2contents in the vent fluid of the black chimney are quite high (CH4: 25–100 μmol/kg, H2: 50–1 000 μmol/kg) compared with those of the shallow water hydrothermal vent fluid (CH4: 0.007–0.200 μmol/kg, H2: 0.001–0.220 μmol/kg) (Tarasov et al., 2005). This suggests that seafloor hydrothermal activity, cold-seep and gas hydrate may have same methane source and exhibit different methane sinks. Therefore, we propose a new hypothesis of same source and different sinks of seafloor hydrothermal activity, cold-seep and gas hydrate.

4.2 Hydrothermal plumes

4.2.1 Hydrothermal plumes in the OT

The Kuroshio flows from east of Taiwan, China, northwards along the OT, with a maximum speed of 1 m/s and width of 100 km (Liang et al., 2003), which can notably impact the seawater characteristics in the northwestern Pacific. However, the effects of the Kuroshio on the hydrothermal plumes in the OT remain unclear. The results from our previous studies demonstrated that the input of the Kuroshio influences the chemical and physical properties of the hydrothermal plume in the OT, with influence decreasing from the SOT to the MOT (Zeng et al., 2018a).

In the SHVFs of the OT, major elements exhibit linear correlations in hydrothermal plumes (e.g., B3+and Sr2+), and the anomalous layers show similar element concentrations (Mg2+, SO42ˉ) and element ratios (Mg2+/Ca2+, SO42ˉ/Mn2+) to those of the OT vent fluids but lower than those of other layers in the hydrothermal plumes. This reveals that the discharge of vent fluid with high concentrations of K+, Ca2+, and B3+and low concentrations of Mg2+and SO42ˉ results in chemical variations of the OT hydrothermal plumes (Zeng et al., 2018a).

The Sr2+/Ca2+and Ca2+/Cl–ratios in hydrothermal plumes are similar to those of average seawater, suggesting that the chemical properties of local seawater can be inferred from the Sr2+/Ca2+and Ca2+/Cl–of hydrothermal plumes (Zeng et al., 2018a). Calculated salinity values of hydrothermal vent water and hydrothermal plume water are consistent with measured salinity values (34.3–34.4) of hydrothermal plumes in the MOT and SOT (Fig.1). However, the flux of hydrothermal B3+, Mn2+, Ca2+, and K+to seawater in the OT are approximately 0.293–34.7, 1.30–76.4, 1.04–326, and (2.62–873)×106kg/a, respectively, and the heat flux is approximately (0.159–1 973)×105W, which implies that approximately 0.000 6% of ocean heat is supplied from seafloor hydrothermal plumes (Zeng et al., 2018a).

4.2.2 Hydrothermal plumes in the EMB

Hydrothermal plumes are potential tools for locating, characterizing, and quantifying seafloor hydrothermal fluid discharge. A neutrally buoyant hydrothermal plume can extend many kilometers (Charlou et al., 1991). Our previous results showed strong positive correlations of arsenic (As) and antimony (Sb) with Mn ( R2＞0.8) in the hydrothermal plumes of the EMB, and the As/Mn-Mn and Sb/Mn-Mn relationships are exponential without substantial deviation (Fig.2). This demonstrates that As and Sb in EMB hydrothermal plumes can be used to identify the hydrothermal plume as a source and trace the hydrothermal plume spreading movement. However, Cl is depleted relative to ambient seawater in the anomaly layers of hydrothermal plumes at Station 18G and 18K whereas Mn, As, and Sb are slightly enriched, which reflects the contribution of Mn-As-Sb-rich and Cl-poor vent fluid. The hydrothermal plume at Station 18B is slightly enriched in Cl and signifi cantly enriched in Mn, As, and Sb, reflecting the contribution of a Mn-As-Sb-Cl-rich vent fluid (Zeng et al., 2018b).

4.3 Hydrothermal sulfi des

4.3.1 Re-Os abundance and isotopic compositions of hydrothermal sulfi des

Little is presently known about the Re-Os isotopic composition of hydrothermal sulfi des from SHVFs in MORs and BABs owing to their low concentrations in seafloor hydrothermal sulfi des and the difficulty in obtaining pure hydrothermal sulfi de samples. We found a limited range of most187Os/188Os radios (1.004–1.209) of hydrothermal sulfi des in the SHVFs from MORs and BABs. This indicates that Os in hydrothermal sulfi de is mainly from seawater and therefore clarifi ed as a seawater-derived component. The initial187Os/188Os isotopic compositions of ancient seafloor hydrothermal sulfi des might thus be a useful proxy for understanding the Os components of ancient seawater because ancient seafloor hydrothermal sulfi des were also produced by the mixing of seawater and vent fluid (Zeng et al., 2014a). The Re and Os of vent fluids in SHVFs are more likely to be incorporated into Fe- and Fe-Cu sulfi de mineral facies and Os enriched under low-temperature (＜200°C) hydrothermal conditions. Moreover,187Os/188Os values of LHF sulfi de samples are lower than those of ambient seawater, which might be affected by seawater Os and MORBs and/or ultramafi c rocks (Fig.3). The Re-Os data of seafloor hydrothermal sulfi des have also been used to estimate that SHVFs contain roughly 0.6 to 44 t of Re, and 1–48 kg of Os, and the Os flux of hydrothermal fluids to vents is about 11 kg/a in SHVFs worldwide (Zeng et al., 2014a).

4.3.2 Noble gases in hydrothermal sulfi des

Fig.1 Salinity and temperature in seawater columns and annual water budget in the OT, East China Sea, and Yellow Sea

Fig.2 Variations in As/Mn (a) and Sb/Mn (b) ratios with Mn in the EMB plume

Fig.3 γOs s versus 187 Re/ 188 Os (a) and γOs s versus 187 Os/ 188 Os (b)

Vent fluid temporal variability in SHVFs can be reconstructed by studying noble gases in seafloor hydrothermal sulfi des, which can also extend our knowledge of the historical helium (He)/heat ratio of the seafloor hydrothermal geological record (e.g., Zeng et al., 2001, 2004). Noble gas composition data for fluid inclusions in seafloor hydrothermal sulfi des, sulfates, and opal samples from seafloor hydrothermal sulfi des in MORs and BABs settings remain scarce. The results of our previous study showed that the He isotopic ratios and concentrations in hydrothermal sulfi de samples are variable (3He/4He=0.6–10.4 R a;4He=(0.12–22)×10-8cm3STP/g). Low-temperature vent fluids lose their mantle He in SHVFs during cooling, which leads to higher He concentrations in most sulfi de samples than in opal (4He=(0.017–0.028)×10-8cm3STP/g). The distinct3He/4He ratios of hydrothermal sulfi des in SHVFs originated from different He sources. Specifi cally, sulfi de samples with high3He/4He ratio (＞7 Ra) mainly stem from mantle source (MORB or OIB) by magma degassing, whereas the sulfi de samples with intermediate (1–7 Ra) and low (～1 Ra)3He/4He ratios are derived from the mixing of fluid and seawater and ambient seawater, respectively. The3He/4He ratios of sulfi des reveal that low-temperature sulfi des, sulfates, and opal minerals do not retain the He isotopic compositions of the primary high-temperature vent fluid, whereas high-temperature sulfi des do in global SHVFs (Zeng et al., 2015a). However, the concentrations of other noble gases (e.g., Ne, Ar, Kr, Xe), in seafloor hydrothermal sulfi des are signifi cantly lower than in sulfate and opal mineral samples in global SHVFs. Barite and opal minerals are characteristic of low-temperature (＜200°C) hydrothermal paragenetic associations and the Kr concentrations in our samples show positive correlations with Ne and Ar concentrations (Fig.4). This indicates that heavier noble gases are enriched under low-temperature hydrothermal conditions, which is most easily explained by the dominance of a seawater-derived component in SHVFs (Zeng et al., 2015a). Additionally, global He and heat fluxes to high-temperature fluid vents obtained from He/heat ratios are about (0.05–6)×104kg/a and 0.1–12×1012W, respectively, implying that high-temperature hydrothermal activity in global SHVFs supplies approximately 0.3% of the ocean heat (Zeng et al., 2015a).

4.3.3 Rare earth element (REE) compositions of hydrothermal sulfi des

The study of REEs in seafloor hydrothermal sulfi des is the key to evaluating the sources of hydrothermal fluid constituents, mixing processes, hydrothermal fluid evolution, and physicochemical conditions of hydrothermal fluids (Zeng et al., 2009). Limited REE composition data are presently available for seafloor hydrothermal sulfi des from various SHVFs in MORs and BABs. Our previous results showed that the majority of REE distribution patterns in the global seafloor sulfi des from the MORs and BAB exhibit light REEs (LREEs) enrichment, which is similar to that of fluids in SHVFs. However, the seafloor sulfi des in global SHVFs have variable REE concentrations, Eu anomalies, and fractionation between LREEs and heavy REEs (HREEs), which are related to the REE compositions of the sulfi deforming fluids and chemical compositions of the sulfi de minerals. Furthermore, REE substitution into seafloor Fe-, Cu-, and Zn-rich sulfi des appears to be strongly influenced by crystallographic control (Mills and Elderfi eld, 1995) and the total REE concentrations and variation range of seafloor Fe-rich sulfi des are all larger than those of Cu- and Zn-rich sulfi des, which suggests that REEs of hydrothermal fluids are more easily incorporated into Fe-rich sulfi des during seafloor hydrothermal sulfi de mineral precipitation in SHVFs (Fig.5).

Fig.4 Variations of Ne and Kr concentrations (a), Ar and Kr concentrations (b) in seafloor hydrothermal sulfi des, sulfates, and opal mineral aggregate samples

Based on the seafloor sulfi de REE data, we estimate that SHVFs hold approximately 280 t of REEs. According to the flux and mean REE concentration (3 ng/g) of vent fluids at MORs, vent fluids in SHVFs alone transport up to 360 t of REEs to the oceans over a two-year period, which is higher than the total quantity of REEs in seafloor sulfi des. Excess REEs may be transported away from the SHVFs and become bound in seafloor sulfate deposits, metalliferous sediments, Fe-Mn crusts, and nodules distal to the SHVFs (Zeng et al., 2015b).

4.3.4 S and Pb isotopic compositions of hydrothermal sulfi des

Sulfur (S) and lead (Pb) isotopes are powerful tracers for exploring seafloor hydrothermal processes, fluid-rock interaction, magmatic activity, and S and Pb sources in SHVFs (Zeng et al., 2010b). Our previous results indicated that the S isotopic compositions of the seafloor sulfi des in SHVFs from MORs and BABs are variable (δ34S from 0.0 to +9.6‰), and S in seafloor sulfi des is derived from associated volcanic rocks (e.g., ～0‰ for basalt) and seawater. Compared with the S isotopic compositions of seafloor hydrothermal sulfates from sedimenthosted MORs, the variation range of S isotopic compositions of hydrothermal sulfates from sedimentstarved MORs is also smaller in SHVFs (Fig.6), and S in the sulfate samples is derived mainly from seawater S. However, owing to the lower degree of fluid-rock interaction and fluid-seawater mixing, the δ34S variation range of seafloor sulfi de minerals from super-fast and fast spreading MORs is limited in contrast to the wider δ34S range of sulfi des from superslow and slow spreading MORs.

In contrast to a mixed origin for the source of S, the majority of the Pb isotopic compositions (206Pb/204Pb=17.541±0.004 to 19.268±0.001,207Pb/204Pb=15.451±0.001 to 15.684±0.001,208Pb/204Pb=37.557±0.008 to 38.988±0.002) from seafloor sulfi des in SHVFs from MORs and BABs are similar to those of local volcanic rocks (e.g., basalt), which reveals that Pb in sulfi de from sediment-free MORs and mature BABs is mainly leached from host volcanic rocks. However, Pb isotope ratios of hydrothermal sulfi des on sediment-hosted MORs (e.g., Middle Valley) show a larger range than those of hydrothermal sulfi des from sediment-starved MORs (e.g., EPR 1°S–2°S) (Fig.7). Additionally, we demonstrate that variable S and Pb isotopic compositions of seafloor hydrothermal sulfi des exhibit a relationship with the S and Pb sources, fluidrock and/or sediment interaction, and fluid-seawater mixing in SHVFs (Zeng et al., 2017b).

Fig.5 K D versus ( r REE– r element) 2 in Fe-rich sulfi des (a), Zn-rich sulfi des (b), and Cu-rich sulfi de (c)

Fig.6 Sulfur isotope values of sulfates from various SHSs

4.3.5 Mineralogical, and chemical characteristics of sulfi des from the EPR

Studies of the structure, mineral, and chemical compositions of seafloor hydrothermal sulfi des can help us understand the hydrothermal fluid evolution and elucidate the interaction between subseafloor hydrothermal fluid and rock, as well as the material contributions of seawater (Zeng et al., 2009). However, studies on REEs and rare and dispersive elements of seafloor hydrothermal sulfi des from the EPR near 13°N are scarce, and variations in seafloor hydrothermal sulfi des under the influence of seawater remain poorly constrained (Zeng et al., 2010b). Seafloor hydrothermal sulfi des from the EPR near 13°N include Zn-enriched sulfi des, which are composed mainly of sphalerite, chalcopyrite, and pyrite. Fe contents and δ34S values increase progressively from high- to low-temperature sulfi de mineral assemblages, whereas Zn contents and Pb isotopic ratios progressively decrease. The phenomenon suggests that the effects of seawater on element distributions (Fe, Zn) and isotopic compositions (S, Pb) are enhanced during seafloor sulfi de formation. Furthermore, seafloor weathering accounts for the enrichment of V, Mn, and REEs in the sulfi de-oxidation layer, which results in identical REE patterns for the oxidation layer and seawater. Weathering also distinctly affects the correlations between element ratios of seafloor sulfi des (Zeng et al., 2010b).

4.3.6 Geochemical and U-series isotopic characteristics of sulfi des in the OT

The characteristics of rare and dispersive elements of seafloor hydrothermal sulfi des from the Jade hydrothermal fi eld in the OT and REEs composition of seafloor hydrothermal sulfi des bearing sulfate remain unclear and their chronological ages are poorly constrained. Our previous results demonstrated that LREEs are relatively enriched in the sulfatebearing hydrothermal sulfi de samples from the Jade hydrothermal fi eld in the OT, and all the fresh seafloor hydrothermal sulfi de samples belong to Zn-rich hydrothermal sulfi des. However, the Au and Ag contents in the hydrothermal sulfi des from the Jade hydrothermal fi eld in the OT are related to Fe-sulfi de, because low temperature promotes Au and Ag enrichment in seafloor hydrothermal sulfi des. Based on the210Pb/Pb ratios of the hydrothermal sulfi de samples, their U isotopic composition, and232Th and230Th concentrations are at base level and the formation age of the seafloor sulfi de from the Jade fi eld in the OT is between 200 and 2000 years (Zeng et al., 2009).

4.4 Hydrothermal alteration

4.4.1 Hydrothermal altered pillow basalts from the EPR

Seafloor hydrothermal fluid-basalt interaction at MORs is known to play an important role in chemical exchange between seawater and oceanic crust. However, previous studies of basalt alteration mainly focused on subseafloor samples, whereas the alteration of basalts exposed on the seafloor is less known. We found several types of hydrothermal alteration in pillow basalts from the EPR near 13°N. Variations of Al, Si, and Fe concentrations at the edges of plagioclase micro-phenocrysts in the hydrothermal altered basalt are 17.59%, 10.69%, and 109%, respectively. Analogously, variations of Al, Si, and Fe concentrations at the edges of basaltic glass are 16.30%, 9.79%, and 37.83%, respectively, owing to interaction between fluids and pillow basalt (Zeng et al., 2014b).

4.4.2 Pumice affected by hydrothermal fluids

Hydrothermally altered pumice can record information about the variation or evolution of seafloor hydrothermal systems. The results in our previous studies showed that fluids have at least twostage effects on T3-3 pumice samples near the Iheya North fi eld in the MOT. In the fi rst stage, amorphous silica precipitated from fluid into the vesicles of the pumice owing to conductive cooling and fluidseawater mixing. In the second stage, the pumice suffered low-temperature alteration while precipitated amorphous silica re-dissolved, leading to Si and Fe defi cits and Mg, Zn, Pb, and Cu enhancements in the altered pumice. The change from silica precipitation to re-dissolution in the altered pumice might imply increasing temperature and/or decreasing silica concentrations in the fluids, suggesting a change of hydrothermal environment. Furthermore, ferruginous fi lamentous silica, which might be related to Feoxidizing bacteria, also formed in the hydrothermal altered pumice (Zhang et al., 2018b).

4.5 Hydrothermal Fe-Si-Mn-oxyhydroxides

4.5.1 Hydrothermal Fe-Si-Mn-oxyhydroxides in the EPR

Fe-oxyhydroxides have been discovered at many SHVFs, which occur either as chimneys, mounds, interstitial precipitates fi lling cracks between lava flows, or as irregularly shaped edifi ces. In our published studies, amorphous Fe-oxyhydroxide samples from the EPR near 13°N with a few sphalerite microlites were formed by secondary oxidation in a low temperature, oxygenated hydrothermal environment (Zeng et al., 2008). However, the Feoxyhydroxide samples have similar trace (As, Co, Ni, Cu, Zn) and major element (Fe, Ca, Al, Mg) concentrations to those of sulfi des, suggesting that the Fe-oxyhydroxide represents a secondary oxidation product of seafloor sulfi des. Furthermore, Feoxyhydroxide samples have lower ΣREE contents with a notably negative Ce anomaly (0.12–0.28), and their chondrite-normalized REE patterns are similar to those of seawater, which are distinct from the REE compositions of hydrothermal plume particles and vent fluids. These results suggest that the REEs of the Fe-oxyhydroxide are derived mainly from seawater and the Fe-oxyhydroxides might be a sink of REEs from seawater in the SHVFs. Furthermore, the quick settling of hydrothermal plume particles resulted in lower REE contents and higher Mn contents in these Fe-oxyhydroxides, which are captured in part by V and P from the seawater through adsorption (Zeng et al., 2008). Recognizing the mineralogy, geochemistry, and generation of Fe-Si-Mn-oxyhydroxides in seafloor hydrothermal geological environments is an important component for understanding ancient volcanogenic massive sulfi de deposits (Zeng et al., 2012c).

Altered seafloor basalt samples from the EPR near 13°N are analyzed (Zeng et al., 2014b) to obtain a clearer understanding of the role of hydrothermal and hydrogenetic processes in the formation of Fe-Si-Mnoxyhydroxide encrustations on MORBs. The results show that these encrustations are mainly composed of 1–2 mm thick amorphous Fe-Si-Mn-oxyhydroxides that are characterized by laminated, spherical, porous aggregates with several bio-detritus, anhydrite, nontronite, and feldspar particles. However, the Fe-Si-Mn-oxyhydroxide encrustations contain anhydrite particles and nontronite crystals, which indicate that these encrustations may have formed under relatively low- to high-temperature hydrothermal conditions. Their growth rate suggests that they are unlikely to have resulted from hydrogenic deposition alone but have a hydrothermal and hydrogenic origin, and formed during several stages of seafloor hydrothermal activity. During the initial formation stage of the Fe-Si-Mn-oxyhydroxide encrustations, dense and Mnpoor Fe-Si-oxyhydroxides were deposited from a relatively reducing fluid and loose Fe-Si-Mnoxyhydroxides are subsequently deposited on them. Furthermore, Si-oxide is inhibited and Mn-oxide will precipitate with Fe-oxyhydroxides owing to the increasing oxidation state of the seafloor fluid in the SHVFs (Zeng et al., 2014b).

4.5.2 Hydrothermal Fe-Si-Mn oxyhydroxides in the EMB

The geochemistry and temperature gradients produced by the mixing of oxidized seawater and reduced fluids deliver an appropriate environment and energy sources (CH4, H2S, CO2, Fe2+, and Mn2+) for microbial growth in the SHVFs, which further affect the formation and microstructure of HPs. Previous studies have proposed that neutrophilic Feoxidizing bacteria play a key role in the generation of Fe-Si oxides in SHVFs (Emerson and Moyer, 2002; Edwards et al., 2011). However, the competitive relationship between abiotic and biotic oxidation reactions and abiotic and biotic kinetic mechanisms in Fe oxidation remain unclear. Our previous results show that Fe-Si oxide samples from the PACMANUS hydrothermal vent fi eld in the EMB had abundant rod-like or twisted fi lamentous and granular structures composed mainly of Fe and Si. However, the amount of Fe oxides around the hydrothermal vent was larger than the amount determined by strict abiotic kinetic calculation in the EMB (Yang et al., 2015).

In the PACMANUS hydrothermal vent fi eld, the Fe-Si-Mn-oxyhydroxides constituted by Fe- and Mnoxyhydroxides with opal-A and nontronite, have extremely low contents of trace elements (exclusive of Ba, Mo, V, and U) and REEs, and show REE distribution patterns with positive Eu anomalies and slight LREEs enrichments. The differences in REE distribution patterns between the Fe-oxyhydroxide fraction and Mn-oxyhydroxide fraction originate from diagenetic processes in the EMB. Furthermore, there are various fi lamentous micro-textures that are similar to unique microbial populations, implying that microbially-mediated mineralization occurred during the formation of Fe-Si-Mn-oxyhydroxides (Zeng et al., 2012a). We proposed an original model for the formation of Fe-Si-Mn-oxyhydroxides in the PACMANUS hydrothermal vent fi eld (Zeng et al., 2012a).

Furthermore, there are micro-textures in the Fe-Si-Mn-oxyhydroxide samples from the PACMANUS hydrothermal vent fi eld in the EMB that resemble fossil microbes such as fi lamentous silica and hollow pipes. Our previous results showed that flakes of nontronite crystals precipitated from low-temperature fluids and microbes may have affected their formation. The nontronite crystals either developed a honeycomb texture or dispersed on the surface of the hollow pipes. Moreover, we found that Si-Fe-Mnoxyhydroxides from the PACMANUS hydrothermal vent fi eld in the EMB have two types of nuclei: Si-Mn nuclei and Si nuclei, both of which are encircled by similar Si-Fe outer layers in the rod-shaped oxyhydroxide and spheroidal oxyhydroxide, respectively. The formation of Si-Mn nucleus is closely related to microbes, whereas Si nucleus is of inorganic origin (Zeng et al., 2012c).

4.6 Hydrothermal metalliferous sediments

4.6.1 Smectite minerals from the EPR near 13°N

The formation mechanism of authigenic smectite and its material source in metalliferous sediments can reflect the interactions between seafloor hydrothermal activity and non-hydrothermal mineral phases. However, authigenic smectite in global marine sediments has different origins, and the origin of smectite minerals in the SHVFs from EPR remains unclear. We reported new data on smectite minerals from the EPR near 13°N. The reaction of Feoxyhydroxide with silica, as well as seawater in metalliferous sediments, is responsible for the generation of the smectite minerals. The Si in the smectite minerals may originate from siliceous microfossils (diatoms or radiolarians), detrital mineral phases, or vent fluids. In contrast to authigenic smectites, these smectites have higher δ30Si owing to selective absorption of light Si isotopes onto Feoxyhydroxides during the formation of hydrothermal smectite. The large ionic radii of REEs likely prevent substitution in either the tetrahedral or octahedral lattice sites in the structure of hydrothermal smectite. Thus, REEs are lost and scavenged by Feoxyhydroxides during the formation of hydrothermal smectite, which reduces the value of metalliferous sediments as a latent resource for REEs in the SHVFs (Rong et al., 2018).

4.6.2 Major and trace elements in SMAR sediments

Sulfi des, Fe-Mn oxides, and oxyhydroxides precipitated from hydrothermal plumes may scavenge metal elements from seawater and settled into the sediment surrounding hydrothermal vents. Our previous studies revealed high elemental contents (e.g., Fe, Mn, Cu, Zn, V, and Co) in samples from the metalliferous sediments near SHVFs in the SMAR, whereas other element concentrations (e.g., Sr, Ca, and Ba) in the metalliferous sediment samples displayed reverse trends, and positive correlations between Fe and Zn, Cu, Ni, Co, Pb, and V contents were observed. These results are consistent with the chemical evolution of dispersing hydrothermal plumes from SHVFs (Huang et al., 2017a).

4.7 Organic matter in the hydrothermal vent fi elds

4.7.1 Hydrocarbons in sediments from the NOT

The discovery of seafloor hydrothermal activity offers a new motivation for understanding the nature of organic matters (OM) in SHVFs (Lein et al., 2003; Simoneit et al., 2004). We measured the abundance and distribution of hydrocarbons in the sediment core from the NOT. The data demonstrate that n-alkanes in this sediment core exhibit a bimodal distribution and an odd-to-even predominance of high molecular weights compared with an even-to-odd predominance in low molecular weight n-alkanes. Moreover, the distribution and composition of hydrocarbons in this sediment core indicate that one or several unobserved SHVFs may exist in the NOT (Huang et al., 2017b).

4.7.2 Abundance and distribution of polyaromatic hydrocarbons in SMAR sediments

Polyaromatic hydrocarbons (PAHs) generally have two to seven or more conjugated aromatic rings and are stable under high-temperature hydrothermal conditions. We measured the abundance and distribution of PAHs from SMAR sediment samples and compared with PAHs values from sediments of different distances from the SHVFs (Huang et al., 2014). The previous results showed that ΣPAHs is higher in sediment samples near the SHVFs and lowest in sediment samples farthest away from the SHVFs, implying a plausible hydrothermal origin for ΣPAHs. Moreover, sample 22V-TVG10 showed a maximum ratio between the parent methylphenanthrene and phenanthrene, which likely reflects the degree of seafloor hydrothermal alteration and indicates that the PAHs of SMAR sediments mainly originated from hydrothermal alteration (Huang et al., 2014).

4.7.3 Organic constituents of hydrothermal barnacles and sediments from the SWIR

Previous studies have investigated the OM of hydrothermal fluids, sulfi des, rocks, and sediments, and the organic components of tubeworms, bivalves, gastropods, shrimp, crabs, and fi sh from SHVFs. The results showed that high concentrations of aromatic compounds in hydrothermal barnacle and sediment samples from the SHVFs in the SWIR might result from macromolecular hydrothermal alteration. Microorganism, especially those associated with sulfur metabolism in the SHVFs, might be the source of high concentrations of fatty acids detected in the hydrothermal barnacle and sediment samples from the SWIR. Moreover, n-alkanes might originate from the hydrothermal alteration of carboxylic acids and other lipid compounds in the high temperature and pressure hydrothermal environments of SHVFs in the SWIR (Huang et al., 2013).

4.8 Organisms in hydrothermal vent fi elds

4.8.1 Chemical compositions of mussels and clams

Studies of the chemical characteristics of mussels and clams in SHVFs are important for understanding the concentrations, transport, and biological effects of chemicals in mussels and clams, as well as the mass fluxes and elemental partitioning from seafloor hydrothermal vents into the oceanic biosphere, metal bioaccumulation of seafloor hydrothermal systems, and the sources and sinks of biogeochemical and fluid cycles. This information helps assess the organisms’ biosorption capacity for metals, the transmission of elements between animals and fluids and/or rocks, the roles of metals in the metabolism of the hydrothermal animals, and the mechanism of metal toxicity in the SHVFs (Zeng et al., 2017a).

However, the influence of hydrothermal environment on the behavior of vent mussels and clams in the OT remains unclear. We analyzed the concentrations of major elements, trace elements, and REEs, as well as the carbon and oxygen isotope compositions in the tissues and shells of mussels and clams from the Tangyin and Yonaguni Knoll IV fi elds in the SOT (Zeng et al., 2017a). The data show linear correlations between metal elements in the shells and tissues of the mussels and clams. The Zn, Mo, and Pb contents in clam tissues vary by tissue type, suggesting that not all positive correlations of the elements in the tissues are inherited by the shells in the SHVFs. Moreover, the element ratios (V/As, Ca/Sr, and Fe/Cr) in the mussels and clams are similar to those of the seawater, implying that the element ratios of seawater might be inherited by organisms from seafloor hydrothermal fi eld, which suggests that the V/As and Fe/Cr ratios of the mussel and clam shells can be used to trace local seawater composition in the SHVFs (Zeng et al., 2017a). However, the mussel and clam tissue samples have high total LREE concentrations, LREE enrichment, and no or only slightly negative Eu anomalies, indicating that the mussels and clams in SHVFs are a sink of LREEs from fluids.

Furthermore, the δ13C values of the mussel shells are heavier than those of the clam shells, implying that more than one carbon source is required for explaining the δ13C compositions of the shells. However, the δ18O values of clam shells are similar to those of the mussel shells and fluid, indicating that the δ18O values of mussel and clam shell carbonate are affected by fluids (Zeng et al., 2017a).

4.8.2 Chemical compositions of crab and snail

Crabs, clams, mussels, shrimps, tube worms, limpets, cyclopoid copepods, and snails are known to exist in SHVFs. A depiction of the chemical compositions of benthic animals that inhabit SHVFs is crucial for understanding their biomineralization processes, bioaccumulation of metals, chemical transport, and variations under the physic-chemical conditions of the seafloor hydrothermal environment. However, very little is known about the ecology of the Kueishantao hydrothermal fi eld (KHF) and roles of host-rock, fluid and/or plumes in the life history of crabs and snails, as well as the biological and chemical characteristics of snails.

We analyzed the element compositions of crab and snail shells from the KHF (Zeng et al., 2018c) and showed that the element contents (e.g., Mn, Hg, and K) in the male crab shells are higher than those in female crab shells, whereas the reverse is true for the accumulation of boron, which suggests that Mn, Hg, K, and B accumulation in the crab shells in the KHF is sensitive to gender. However, the Li, Mg, and Co concentrations of crab and snail shells range between Kueishantao andesite and vent fluid concentrations, suggesting that Co-enrichment in snail is affected by the Kueishantao andesite (Fig.8). The majority of LREE distribution patterns in the crab and snail shells resemble those of the fluids, with LREEs enrichment, indicating that the LREEs in the crab and snail shells originate from fluids in the KHF (Zeng et al., 2018c).

5 SHALLOW-WATER HYDROTHERMAL ACTIVITY IN THE KHF

The vents in the KHF can be classifi ed into two types: yellow spring and white spring. The temperature of the yellow-spring fluids (78–116°C) is higher than that of the white-spring fluids (30–65°C) (Chen et al., 2005a, b; Zeng et al., 2013), and the temperature variation of the vent fluids is associated with diurnal tides, reaching a maximum 2–4 h after each high tide (i.e., high pressure) (Kuo, 2001; Chen et al., 2005a, b). The yellow-spring fluids are characterized by extremely low pH (≥1.52) and variable chemical compositions (Chen et al., 2005a, b). The white- spring fluids have relatively low CH4, Fe, and Cu concentrations (Chen et al., 2005a, b; Zeng et al., 2013). Moreover, native sulfur deposits in the KHF are present in a number of different forms: sand, chimneys, and balls (Zeng et al., 2007, 2011).

5.1 Native sulfur chimney

We previously reported sulfur isotopic compositions of 14 native sulfur samples from a chimney in the KHF. The element compositions of the native sulfur samples suggests that the sulfur of the native sulfur chimney in the KHF (H2S and SO2) originated by magmatic degassing, REEs and trace elements in the native sulfur chimneys are mostly derived from Kueishantao andesite and partly from seawater, and relatively low temperature (＜116°C), oxygenated, and acidic environment are favorable for the formation of native sulfur chimneys in the KHF (Zeng et al., 2007).

5.2 Native sulfur ball

Native sulfur ball samples have high S contents (up to 99.96 wt.%), similar to native sulfur chimney in the KHF (Zeng et al., 2011). The sulfur contents, REE, and trace element compositions of the native sulfur balls and local environment in the KHF indicate that the slower growth of native sulfur balls results in relatively higher REE and trace element concentrations in native sulfur balls than in native sulfur chimneys. We propose a growth model called “glue pudding” for understanding the origin of native sulfur balls in the KHF, which developed by mixing oxygenated seawater with an acidic, low-temperature fluid containing SO2and H2S gases, which were then shaped by tidal and/or bottom currents (Zeng et al., 2011).

5.3 REE in vent fluids

Fig.8 Ternary diagram of Li-Mg-Co in crab and snail shell samples from the KHF in the southwestern tip of the OT

Numerous investigations of REEs geochemistry of hydrothermal vent systems have been reported, which are critical for understanding sub-seafloor hydrothermal processes, whereas few investigations have been conducted on the shallow-water hydrothermal systems of the KHF. Our previous results showed that the total REE concentrations of yellow-spring fluids are notably higher than those of ambient seawater but similar to those of white-spring fluids. The chondrite-normalized REE distribution patterns of the yellow-spring fluids show slight convex-downward curvatures at Eu in contrast to those of the white-spring fluids that show no Eu anomalies, which is attributed to the more oxidizing and low-temperature conditions. Compared with HREEs, LREEs are slightly enriched in KHF fluids and the behavior and patterns of REEs in both yellow- and white-spring fluids are affected by the short water-rock interaction time, exceptionally low pH (2.81 and 2.29), fluid boiling, and precipitation of native sulfur (Wang et al., 2013b).

5.4 Boron in vent fluids and hydrothermal plumes

Seafloor hydrothermal circulation can trigger extensive boron isotopic and chemical exchange, which is controlled by fluid temperatures and subseafloor water-rock interactions. Accordingly, the sub-seafloor water/rock ratios are recorded by the boron isotopic compositions in fluids, which provides an effective method for studying the origin of fluid and hydrothermal processes. However, very little is known about the boron isotope compositions of the shallow-water vents in the KHF, which hinders the understanding of sub-seafloor water-rock interaction and causes bias in the flux quantifi cation of global hydrothermal boron to the oceans (Zeng et al., 2013).

We measured boron concentrations and isotope compositions of seawater, andesite, and fluid and plume samples from the KHF. Fluids and plumes from the yellow and white springs display a regular array of data points, which suggests that the boron in the fluids and plumes is mainly from seawater, rather than from KHF andesite, which further implies that the duration of fluid-andesite interaction is short in the KHF. However, the pH, boron concentrations, and isotopic compositions from fluids to hydrothermal plumes indicate a close relationship with one another, which suggests that the δ11B/B and pH/B ratios of the hydrothermal plumes have constant values within a small distance (approximately 15 m). This further suggests that the diffusive processes controlling the chemical compositions of hydrothermal plumes in the seawater can be depicted by boron data from the hydrothermal plumes (Zeng et al., 2013). The subseafloor water/rock ratios are between 1.96 and 3.63 in the KHF, and the hydrothermal flux of boron from the fluids into the oceans is between 6.69×104and 1.32×105mol/a (Zeng et al., 2013).

6 CONCLUSION

This paper describes the geological settings and volcanic rocks of SHVFs from MORs and BABs, systematically revealed the material sources and controlling factors of the HPs on global SHVFs, introduces the HP formation model (e.g., hydrothermal sulfi de, native sulfur chimney and balls), and indicates that the new means for revealing the formation mechanism of HPs is to expand the research of seafloor hydrothermal vent organisms.

(1) Systematic Re-Os content and isotopic composition of global seafloor hydrothermal sulfi des have been studied. Seawater is a signifi cant source of Re and Os in seafloor hydrothermal sulfi des and Os is enriched under low-temperature conditions. The187Os/188Os ratios of seafloor hydrothermal sulfi des are not controlled by the sulfi de mineral facies, and the initial187Os/188Os of ancient sulfi des can trace the ancient seawater component. Helium in seafloor hydrothermal sulfi des mainly comes from the mantle, whereas Ne, Ar, Kr, and Xe mainly from seawater are enriched in fluid inclusions of low-temperature sulfate and opal. Furthermore, hydrothermal vent fluid provides REEs for global sulfi de formation in different tectonic environments, which results in similar LREEs/HREEs ratios between sulfi de and hydrothermal fluid. The REEs content and distribution model of seafloor hydrothermal sulfi des are restricted by the characteristics of mineral chemical composition, physical and chemical conditions of fluid when sulfi de forms, REEs content and distribution model of hydrothermal fluid, mixing degree of hydrothermal fluid and seawater, and interaction between subseafloor fluid and rock. Moreover, the As and Sb of seawater columns can be used to identify the hydrothermal plume as a source and trace hydrothermal plume spreading.

(2) The published results have revealed the material source and formation conditions of native sulfur chimneys and native sulfur balls, established their formation modal, and constrained the relationship with subseafloor geological processes. Compared with deep sea hydrothermal systems, shallow-water hydrothermal systems have their own particularity and complexity. They differ from volcanic hot springs on land and draw attention to the impact of atmospheric precipitation on submarine hydrothermal fluid, impact of tides and typhoons on the growth of HPs such as chimneys, native sulfur balls, and the relationship between earthquakes and hydrothermal activities. The δ11B values of the vent fluid and hydrothermal plume in the KHF have been determined for the fi rst time, revealing the source and evolution of the fluid. The pH, B content, and δ11B are signifi cantly related from the vent fluid to the hydrothermal plume. The δ11B/B and pH/B ratios are stable over short distances (＜15 m) from the vent to the hydrothermal plume. The B content and δ11B value in the hydrothermal plume can be used to describe the diffusion process controlling the chemical composition of the hydrothermal plume in the seawater environment. Furthermore, the hydrothermal fluid and boron in the hydrothermal plume are derived mainly from seawater, with only a small amount from andesite, and the interaction time between the subseafloor fluid and andesite is short. In the fi rst geochemical study of Anachis sp. shells from the KHF. We found that element accumulation (K, Mn, Hg, and B) in crab shells is affected by gender via molting, and high metal concentration in snails may be ascribed to long metal accumulation time. LREEs in crab and snail shells originate from hydrothermal fluids.

(3) The formation mechanism of Fe-Si-Mnoxyhydroxides in the PACMANUS fi eld of the Manus basin in the western Pacifi c is revealed. Fe-Si-Mnoxyhydroxides show various fi lamentary microstructures. Thus, microorganisms play an important role in the formation of Fe-Si-Mnoxyhydroxides. During the mixing process of hydrothermal fluid and seawater, the mineralization of Fe oxidizing bacteria promoted the precipitation of Si and the existence of bacterial fi laments led to the enrichment of U in Fe-Si-Mn-oxyhydroxides, whereas a portion of the Fe-oxyhydroxides were encapsulated with the growth of Mn-oxyhydroxides. Smectite from the EPR near 13°N most likely formed by the reaction of hydrothermal Fe-oxyhydroxide with silica and seawater in metalliferous sediments. Furthermore, volcanism may be the main reason for the observed distribution and composition of hydrocarbons in the sample from the NOT.

(4) A new calculation method of the He/heat ratio is proposed. The contents of Re, Os, and REEs in the global seafloor sulfi de deposits are very low (～4 t of Re, ～8 kg of Os, and ～280 t of REEs). The boron flux is between 6.69×104and 1.32×105mol/a in the KHF. We estimated that the global He and heat fluxes are up to 500 kg/a and 1×1011W, respectively, and 0.3% of the ocean heat is provided by high-temperature hydrothermal activity on the seafloor, which provides a new theory and method for overcoming the major problem of hydrothermal geology on the seafloor (i.e., heat and material fluxes).

(5) We propose a multilayer magma chamber system to explain the complex plagioclase crystals in silicic rocks. Our published results indicated that one depleted source and two enriched sources contribute to the formation of MORBs from EPR between 1°S and 2°S. The fractionation of silicon and oxygen isotopes of basalts from the EPR near 13°N is influenced by the SiO2content in igneous rocks. The SWIR peridotites originated from a depleted mantle source magma and experienced partial melting.

Combined with the above work, an in-depth view of submarine hydrothermal geology is established, the basic concept of submarine hydrothermal system research is systematically defi ned, and the types of HPs (e.g., submarine hydrothermal sulfi de), are classifi ed in the published book. This provides a theoretical basis for the investigation of submarine hydrothermal geological process, its associated resources, and environmental effects.

7 DATA AVAILABILITY STATEMENT

All data generated and/or analyzed during this study are available from the corresponding author upon reasonable request.

8 ACKNOWLEDGMENT

We would like to thank the crews of the DY105-17, DY115-19, DY115-20, DY115-21, HOBAB2, HOBAB3, HOBAB4, and HOBAB5 cruises for helping in collecting samples. We thank Esther Posner, PhD, from Liwen Bianji, Edanz Editing China (www.liwenbianji.cn/ac), for editing the English text of a draft of this manuscript.