Magnesium isotopic composition of submarine vent fluids from arc and back-arc hydrothermal systems in the western Pacific

Highlights

• Magnesium isotopic compositions of hydrothermal vent fluid were measured.

• Some fluids with extreme magnesium depletion showed low δ²⁶Mg.

• High-temperature hydrothermal sinks have little effect on the oceanic Mg isotopic composition.

• 7–26% of the riverine Mg input is estimated to be removed via high-temperature sink.

Abstract

Seafloor hydrothermal systems are important sinks for Mg as their concentration in high-temperature hydrothermal vent fluids decreases to near zero through the formation of secondary minerals. However, the problem of mass balance still persists regarding the global sink for Mg via sub-seafloor high-temperature hydrothermal reactions. In this study, we measured isotopic compositions of Mg in high-temperature hydrothermal vent fluids from 5 arc and back-arc systems in the western Pacific. This was done to better understand Mg behavior during hydrothermal circulation and to quantify the oceanic Mg cycle. The samples with significantly low Mg concentrations showed ²⁶Mg depletion due to Mg fractionation and Mg fixation during hydrothermal circulation. Addition of Mg to the permeable sub-seafloor during venting is the probable reason for the isotopic variation of Mg. The vent fluids (return flux) with lighter Mg isotopic compositions contain much smaller amounts of Mg than seawater. This confirms that the high-temperature hydrothermal sinks have a less significant effect on the oceanic Mg isotopic composition. Based on the simple steady state equations and low-temperature hydrothermal sinks of ²⁶Mg (−0.25‰ to 0.00‰ of δ²⁶Mg), 7–26% of the riverine Mg input is presumed to be removed through the high-temperature sink. This estimate is lower than that proposed by Mottl and Wheat (1994), which was 10–40%. Thus, other Mg sinks such as the low-temperature hydrothermal sink and/or the dolomite sink need to be considered.

Introduction

Magnesium plays an important role in addressing the long-term global carbon cycle which is supported by the covariation between seawater Mg/Ca and climate (Stanley and Hardie, 1998; Ries et al., 2006). Secular changes in the marine mineralogy are also regarded as a result of variation in seawater Mg chemistry throughout geologic time (Hardie, 1996; Higgins and Schrag, 2015; Wilkinson and Algeo, 1989). Conversely, the modern oceans have a constant Mg concentration (53 mM Holland et al., 1986) and isotopic composition (−0.83 ± 0.09‰, Ling et al., 2011) with a residence time of ~10 million years (Berner and Berner, 2012). The major source of Mg into the ocean is continental runoff; the largest Mg flux is delivered by rivers (5.5 Tmol/yr, estimate from Berner and Berner, 2012). However, there is little agreement on the oceanic Mg cycle because the magnitude and mechanisms regarding the removal of Mg in the ocean are poorly constrained. This mass imbalance was called the “Mg problem” (Drever, 1974) for a long time. Now, it has been recognized that seafloor hydrothermal systems are responsible for considerable amount Mg and sulfate removal. This is because the concentration of both species in high-temperature hydrothermal vent fluids decreases to near zero, as demonstrated by earlier experimental studies (Seyfried and Bischoff, 1977; Seyfried and Mottl, 1982). Several estimates support that the hydrothermal Mg sink, which forms Mg-containing secondary minerals, comprises more than half of the riverine Mg flux (Milliman, 1974; Elderfield and Schultz, 1996; Tipper et al., 2006b; Beinlich et al., 2014). However, the proportion of seawater Mg eliminated via seafloor hydrothermal reactions at high-temperatures is not clear.

In the 2000s, the development of Mg isotope measurement techniques broadened the understanding of Mg behavior. Relatively large mass differences between Mg isotopes (²⁴Mg, ²⁵Mg, and ²⁶Mg) facilitate isotopic fractionation by low-temperature processes. For example, the high-temperature geological processes – including planetary accretion processes, partial melting and magmatic differentiation – do not significantly change Mg isotopic compositions, leading to homogeneity (Teng et al., 2010). The fresh oceanic crust, such as the mid-oceanic ridge basalt (MORB) and oceanic island basalt (OIB), has δ²⁶Mg values similar to that of the mantle (−0.25 ± 0.07‰, Teng et al., 2010). On the contrary, Mg isotopic studies of the low-temperature geological processes such as secondary mineral formation (Galy et al., 2002; Tipper et al., 2006a) and weathering (de Villiers et al., 2005; Tipper et al., 2006a, Tipper et al., 2006b; Pogge von Strandmann et al., 2008) showed large Mg isotopic variations. δ²⁶Mg is a suitable proxy for determining mass-balance considerations of the oceanic input and output fluxes of Mg. In general, the net flux of δ²⁶Mg from terrestrial weathering has been negative (−1.09‰, Tipper et al., 2006b) relative to seawater; this isotopic offset is inherited from lower δ²⁶Mg outputs in the ocean, caused by either a Mg isotope fractionation of low-temperature Mg-bearing carbonates/clays or high-temperature hydrothermal alteration of primary silicate minerals. Previous studies showed the variable δ²⁶Mg values of the ocean floor environment. In deep sea sediments, the precipitation of Mg-bearing minerals leads to the decrease in Mg concentrations of pore-fluids with depth; but δ²⁶Mg values tend to change depending on different minerals (Higgins and Schrag, 2010): silicate minerals are enriched in the heavy Mg isotopes, while dolomite formation, regarded as one of the major Mg sinks in the oceans, prefers the light Mg isotopes.

The relative importance of the hydrothermal sink to the net oceanic Mg sink can be resolved as a function of the difference of the δ²⁶Mg of the hydrothermal and dolomite sinks. The Mg isotopic fractionation during hydrothermal ultramafic rock alteration suggests that sub-seafloor hydrothermal carbonation may contribute significantly to the Mg isotopic composition of seawater (Beinlich et al., 2014). A recent study demonstrated that formation of secondary minerals, saponite and calcite results in variable δ²⁶Mg values of the altered oceanic crust ranging from −2.76 to +0.21‰ (Huang et al., 2018). This indicates that significant Mg isotopic fractionation occurs during low-temperature alteration of the oceanic crust. Although high-temperature basalt alteration represents a sink in the global Mg mass-balance (10–40% loss of the riverine Mg input, thermal estimation by Mottl and Wheat, 1994), effects of host rock type, phase separation, and overlying sediment on Mg isotope fractionation and variations in the δ²⁶Mg values of the hydrothermal fluids are still under debate.

Much of the early research on the oceanic Mg sink was related to the missing Mg flux, and little attention was given to the behavior of Mg isotopes during high-temperature seafloor hydrothermal processes. This is because high-temperature hydrothermal systems completely remove Mg from the fluid, suggesting that Mg isotopic fractionation could be negligible (Higgins and Schrag, 2015). In order to investigate the Mg isotopic behavior during high-temperature hydrothermal circulation and quantify its contribution to the global Mg cycle, we present δ²⁶Mg data for eleven submarine vent sites from five arc and back-arc hydrothermal systems in the western Pacific. These are classified under three categories: sediment-starved sites (Vienna Woods and PACMANUS in the Manus Basin, Suiyo Seamount in the Izu-Bonin Arc, Alice Springs and Forecast Vent in the Mariana Trough), phase-separated sites (White Lady, Kaiyo, and LHOS in the North Fiji Basin) and sediment-hosted sites (JADE, Minami-Ensei, and CLAM in the Okinawa Trough). Our research can provide constraints on the seawater Mg chemistry in modern hydrothermal systems, mainly at back-arc tectonic spreading centers in different types of hydrothermal settings.