Utilizing ²³⁴Th/²³⁸U disequilibrium to constrain particle dynamics in hydrothermal plumes in the Southwest Indian Ocean

1. Introduction

Hydrothermal events emit a large amount of hot fluid with metal elements into seawater (Ingebritsen and Evans, 2019). These metals precipitate by forming polymetallic sulfides (Ray et al., 2018) which will effectively absorb many other metals (e.g., Pb, Po, Th, Zn) (Liao et al., 2017; Pavia et al., 2019), phosphorus (Kadko et al., 1994), carbon and nitrogen (Toner et al., 2009) from ambient seawater. Metal-enriched particles are utilized by filter-feeders, resulting in an evident metal-accumulation, such as Hg, Cu, Pb, Cr, Ni and Zn, in mollusks which habited around the hydrothermal vents (Peng et al., 2011; Pancaldi et al., 2019). Hence, metal cycling in hydrothermal systems and similar submarine volcanic eruption (Lozano-Bilbao et al., 2018) have attracted extensive attentions according to their influences in associated ecosystem (Chakraborty et al., 2014). However, the influencing scope of the hydrothermal activities in deep ocean is difficult to define, due to the lack of technologies to constrain the resident timescale and the dispersal range of the metals (Yang et al., 2016).

Sediment trap is frequently used for particle collection directly from the upper ocean (Hung and Gong, 2010; Hung et al., 2012) or the deep water (Ran et al., 2015; Zhang et al., 2019). It can provide valuable information on sinking flux of the real sinking particles (Shih et al., 2019). In hydrothermal systems close to the vents, sediment traps have also been used to quantify the particle flux and to collect sulfides which could be used to characterize particle composition (German et al., 2002; Beaulieu et al., 2009). Hydrothermal plumes usually cover a large spatial scope and may vary spatiotemporally with the deep currents. Using sediment traps to track particle cycling in the hydrothermal plumes would cost high.

Radiogenic ²³⁴Th (T_1/2=24.1 d) is a tool for constraining the adsorption, sinking and resident timescales of elements in the upper ocean (Bi et al., 2013; Huang et al., 2013; Pavia et al., 2019). Since its particle-reactive nature in seawater (Lin et al., 2014) similar to some other metals (e.g., Pb, Po) (Boisson et al., 2001) and phosphorus (Kadko et al., 1994; Lin et al., 2012), ²³⁴Th is readily absorbed onto particulate matter and settles downwards (Buesseler et al., 1992; Yang et al., 2016). Thus, ²³⁴Th has been determined and used in the upper oceans especially in the euphotic zone to constrain biogenic carbon sinking to the ocean interior (i.e., the biological pump) (Ma et al., 2011). In deep oceans, the total ²³⁴Th is often found to be in equilibrium with its parent ²³⁸U (Benitez-Nelson et al., 2001a; Coppola et al., 2005; Anand et al., 2018) due to scarce particles and long residence time of ²³⁴Th. This observation seems to disable ²³⁴Th to trace the kinetics of element adsorption in deep oceans. However, previous studies reveal that ²³⁴Th in hydrothermal plumes is deficit, to a varying degree, with respect to ²³⁸U (Kadko et al., 1994; Owens et al., 2015; Pavia et al., 2019), ascribing to the abundant hydrothermal particles and enhanced metal adsorption. Therefore, ²³⁴Th-²³⁸U disequilibrium could be an efficient technique to constrain the resident timescales and dispersal range of particle-reactive metals emitted from the hydrothermal vents.

Recently, hydrothermal plumes at different depths from active vents were confirmed over the ultraslow-spreading Southwest Indian Ocean Ridge (SWIR) by the distributions of Fe, Mn, particulate content anomaly and zinc sulfide (Sun, 2011; Wang et al., 2012; Sun et al., 2014). These findings provide an opportunity for us to examine the influential scope of hydrothermal fluids and the residence and dispersal of metals along the plume. In this study, ²³⁴Th was determined in an identified hydrothermal plume over the SWIR in order to (1) compare the difference in radioactive ²³⁴Th contents between hydrothermal plume and general deep oceans; (2) quantify the resident timescale and the adsorption behavior of particle-associated metals; and (3) preliminarily evaluate the influence of hydrothermal plume on deep ocean carbon cycling and its environmental implications.

2. Materials and methods

2.1 Sampling

Sampling station CTD3 is located right over the ultraslow-spreading SWIR with a water depth of about 1 800 m (Fig. 1). Within 1 000–1 600 m, elevated suspended particulate matter (SPM) contents comparing with the overlying water column and the existence of zinc sulfide (ZnS) mineral during the same cruise (Sun, 2011; Sun et al., 2014) identified the hydrothermal effluent plume from three active vents on the SWIR (Tao et al., 2012).

Figure  1.  Sampling location over the Southwest Indian Ocean Ridge.

DownLoad: Full-Size Img PowerPoint

Seawater and SPM were collected onboard the R/V Dayang I during the DY115-21 cruise in February 2010. The contents of SPM were quantified through the difference in the weight between dried particles at 40°C and wet particles (Sun, 2011). In total, thirteen depths (i.e., 3 m, 50 m, 100 m, 280 m, 300 m, 500 m, 800 m, 1 000 m, 1 400 m, 1 450 m, 1 500 m, 1 550 m, and 1 600 m) were occupied for ²³⁴Th determination. Below 1 200 m, SPM contents indicated evident hydrothermal plume signals (Fig. 2, Sun et al., 2014), thus high-resolution sampling with 50 m intervals was conducted to collect typical plume information. In the euphotic zone (0–100 m), ²³⁴Th was also sampled to examine the difference between the surface ocean and the hydrothermal plume by comparing their characteristics.

Figure  2.  Vertical distributions of dissolved, particulate and total ²³⁴Th concentrations (a), ²³⁴Th_T/²³⁸U ratios (b) and suspended particulate matter (SPM) concentrations (c). SPM data are obtained from Sun (2011) at the same station.

DownLoad: Full-Size Img PowerPoint

2.2 ²³⁴Th determination

At each specific depth, about 8 L of seawater were filtered through a quartz fiber membrane (combusted at 450°C for 4 h) with 1 μm pore size (QMA, Whatman) to collect particles for particulate ²³⁴Th (i.e., ²³⁴Th_P) measurement. Particle-contained membranes were dried at 60°C immediately and counted using a low-level beta counter with the counting efficiency of 41% (Yang et al., 2016) until the net counting errors were less than ±6%. A second counting was conducted after 150 d for quantifying other beta emitters in order to remove their contribution to the first obtained ²³⁴Th counts.

Dissolved ²³⁴Th (i.e., ²³⁴Th_D) was concentrated from 4 L of filtrate using the small-volume method (Benitez-Nelson et al., 2001b). In brief, the pH value of filtrate was adjusted to 9.0 using ammonium hydroxide, then KMnO₄ and MnCl₂ solution were added to form MnO₂ particle. After 6 h, MnO₂ precipitate was separated from solution via filtration using QMA membrane. The recovery was 95.7%±1.0% (mean±SD, n=5) (Yang et al., 2015a). After dried at 60°C, dissolved ²³⁴Th activity was measured by the low-level beta counter in a manner similar to particulate ²³⁴Th but with ±1σ counting error. The concentrations of both dissolved and particulate ²³⁴Th were calculated based on the two measurements and corrected for recovery, blank counts, and sampling time. The total ²³⁸U activities were calculated using the newly updated ²³⁸U-salinity relationship (Owens et al., 2011). The uncertainties of dissolved and particulate ²³⁴Th measurement were propagated from the counting errors. The total ²³⁴Th and ²³⁴Th/²³⁸U ratio uncertainties were propagated from dissolved and particulate ²³⁴Th.

2.3 Calculations of kinetic parameters of ²³⁴Th and POC fluxes

In the oceans, adsorption and sinking of particle-reactive radionuclides (e.g., ²³⁴Th, ²¹⁰Pb and ²¹⁰Po) were following the first-order reaction kinetics (Bacon et al., 1976; Buesseler et al., 1992). Thus, the adsorption and sinking of ²³⁴Th in both euphotic zone and hydrothermal plume were expressed as follows:

where A_ThD, A_ThP, and A_U are the activities of ²³⁴Th_D, ²³⁴Th_P, and ²³⁸U (Bq/m³); λis the decay constant of ²³⁴Th (0.029 d^–1), k represents the adsorption rate constant (d^–1). Hence, λA_U reflects the generation rate of ²³⁴Th_D via ²³⁸U decay (Bq/(m³·d)). λA_ThD and λA_ThP represent the decay terms of dissolved and particulate ²³⁴Th. kA_ThD is the adsorption rate of ²³⁴Th_D. φ is the sinking rate constant of ²³⁴Th_P ( d^–1). φA_ThP denotes the sinking rate of ²³⁴Th_P. The residence times of dissolved and particulate ²³⁴Th are calculated from the reciprocals of adsorption and sinking rate constants (d). At the steady-state, the adsorption, sinking rate constants and residence times of ²³⁴Th can be calculated, as well as the adsorption and sinking rates.

The active particle dynamics in the hydrothermal plume over the SWIR suggest that particulate matter and its combined metals would quickly settle to bottom water and seafloor, and consequently shape different environment regions from general deep oceans (Yang et al., 2016). This study attempt to reveal the characteristics of particulate components sinking using limited data based upon the widely used flux calculation (Buesseler et al., 1992):

where F_absorb and F_sinking are the adsorption and sinking fluxes of ²³⁴Th (Bq/(m²·d)), respectively; $ {\displaystyle\int }_{1\;400}^{1\;600}{A}_{\mathrm{U}}{\rm{d}}z $, $ {\displaystyle\int }_{1\;400}^{1\;600}{A}_{\mathrm{T}\mathrm{h}\mathrm{D}} {\rm{d}}z$ and $ {\displaystyle\int }_{1\;400}^{1\;600}{A}_{\mathrm{T}\mathrm{h}\mathrm{T}}{\rm{d}}z $ represent the total amounts of ²³⁸U, ²³⁴Th_D, and the total ²³⁴Th (i.e., ²³⁴Th_T) (Bq/m²) in the hydrothermal plume (1 400–1 600 m).

The sinking fluxes of SPM, POC and PON were calculated through the proposed equation (Buesseler et al., 1992; Yang et al., 2016):

where F_i,sinking represents the fluxes of SPM, POC and PON, and i denotes the contents of SPM, POC and PON.

3. Results

In surface euphotic water (0−100 m), the total ²³⁴Th concentrations ranged from 9.5 Bq/m³ to 30.8 Bq/m³ with an average of (22.8±3.3) Bq/m³ (Fig. 2). The averaged ²³⁴Th_T/²³⁸U ratios was 0.55±0.08, corresponding to a typical deficit of ²³⁴Th with respect to ²³⁸U in the upper water of Indian Ocean (Yang et al., 2016; Anand et al., 2018). The lowest ²³⁴Th_T value was found at depth 50 m, matching with the highest SPM content in the euphotic zone. In the twilight zone (i.e., 100–1 000 m), ²³⁴Th_T activity was overall equal to ²³⁸U activity (Fig. 2) with the ²³⁴Th_T/²³⁸U ratios ranged from 0.92 to 1.06 (avg. 0.98±0.04), showing an equilibrium feature as widely observed in the mesopelagic in the Indian and Atlantic sectors of the Southern Ocean (Coppola et al., 2005; Roca-Martí et al., 2017), Pacific Ocean (Charette et al., 1999; Buesseler et al., 2009) and Atlantic Ocean (Owens et al., 2015). It is notable that ²³⁴Th_T activity showed lower value than ²³⁸U activity between 1 400 m and 1 550 m below the twilight zone, ranging from 28.7 Bq/m³ to 35.0 Bq/m³ and averaging 33.2 Bq/m³. Therefore, the ²³⁴Th_T/²³⁸U ratios varied from 0.73 to 0.88 with an average of 0.83±0.04, contrasting with the usual equilibria between ²³⁴Th and ²³⁸U observed in bathypelagic waters, e.g., the North Pacific Subtropical Gyre (Benitez-Nelson et al., 2001a), the Northeast Atlantic (Schmidt, 2006), and the northern Indian Ocean (Anand et al., 2018).

Overall, the deficit extents of the total ²³⁴Th mirror the vertical distribution of SPM (data from Sun, 2011) in the water column over the SWIR (Fig. 2c). In the upper 200 m, low ²³⁴Th_T/²³⁸U ratios (²³⁴Th deficit) correspond to high SPM contents, indicating the predominant role of particles in absorbing and removing ²³⁴Th. Within 280–800 m depth, SPM showed low concentrations and matched with a weak removal of ²³⁴Th and around 1.0 of ²³⁴Th_T/²³⁸U ratios (Fig. 2c). At 1 000 m, high SPM contents were also observed with a slightly low ²³⁴Th_T/²³⁸U ratio. Below 1 200 m, the low ²³⁴Th_T/²³⁸U ratios and high SPM were simultaneously observed.

4. Discussion

4.1 Unusually low ²³⁴Th in the hydrothermal plume vs. normal deep water

In normal bathypelagic zone (1 000–4 000 m), available data suggest that ²³⁴Th is often comparable to ²³⁸U (Benitez-Nelson et al., 2001a; Coppola et al., 2005; Anand et al., 2018) due to scarce SPM and weak ²³⁴Th removal, which allow ²³⁴Th to reach equilibrium with ²³⁸U as observed in the mesopelagic at the study site (Fig. 2b). However, ²³⁴Th showed significant deficit with respect to ²³⁸U within the hydrothermal effluent plume (below 1 200 m) over the SWIR (Fig. 2b), indicating an effective removal of ²³⁴Th. In seawater, ²³⁴Th removal includes sinking with particles after adsorption (Yang et al., 2015a) and in situ decay. In oceanic settings with abundant particulate matter, for example in the productive euphotic zone and turbid benthic nepheloid layer (BNL), adsorption and sinking usually dominate ²³⁴Th removal and result in ²³⁴Th deficit (Rutgers van der Loeff et al., 2002; Bacon and Rutgers van der Loeff, 1989; Turnewitsch et al., 2008). In environments with less contents of particles, decay often dominates ²³⁴Th removal and leads to the equilibria between ²³⁴Th and ²³⁸U as observed in general bathypelagic water (Anand et al., 2018). Thus, the abnormal ²³⁴Th deficit is attributed to the enhanced ²³⁴Th removal via particle settling in the hydrothermal plume. In fact, a recent study observed intensive removal of thorium isotopes (i.e., ²³⁴Th, ²³⁰Th) in a hydrothermal plume (Hayes et al., 2015). As shown in Fig. 3, the ²³⁴Th_T/²³⁸U ratio decreases with elevated SPM concentration, supporting the effective removal of ²³⁴Th through on particulate matter in the studied hydrothermal plume.

Figure  3.  Relationship between the ²³⁴Th_T/²³⁸U ratio and SPM content.

DownLoad: Full-Size Img PowerPoint

Unlike the euphotic zone, less abundance of phytoplankton was found in the hydrothermal plume. Hence, particles in plume was less influenced by biogenic origin. Results showed that the elevated particles are mainly as sulfide minerals in the hydrothermal plume from the current study (Sun, 2011; Sun et al., 2014). Since absorption natures of sulfides were different from biogenic particles (i.e., particulate organic matter, biogenic silica, and bio-carbonate), they might show distinguishable affinity with ²³⁴Th. Indeed, the distribution coefficients (K_d) of ²³⁴Th in the photic zone (from 1.70×10⁵ L/kg to 4.52×10⁵ L/kg) were higher than 0.55×10⁵–1.75×10⁵ L/kg in the hydrothermal plume (Fig. 4), implying a weak affinity of sulfide minerals for ²³⁴Th comparable with biogenic particulate. Previous studies indicated that a strong affinity of organic matter with ²³⁴Th, followed by Mn- and Fe-oxides and carbonate, biogenic silica (Guo et al., 2002; Yang et al., 2009). While Hayes et al. (2015) reported that the K_d values of ²³⁰Th on Fe/Mn (hydr)oxides were 1–2 orders of magnitude greater than those for organic matter. Enriched Fe and Mn were reported in the hydrothermal plume over the SWIR (Wang et al., 2012), Fe- and Mn-contained sulfides seem not to be effective in removing ²³⁴Th as biogenic particles, probably due to their sulfide forms (Sun et al., 2014). This view is corroborated by the comparable SPM concentrations between the photic zone and hydrothermal plume (Fig. 2c) but lower K_d values of ²³⁴Th in the plume (Fig. 4).

Figure  4.  Contrasting distribution coefficients of ²³⁴Th between general photic waters (i.e., euphotic and twilight water) and hydrothermal plume.

DownLoad: Full-Size Img PowerPoint

Like ²³⁴Th, particle-reactive elements such as Pb, Po, and P are easy to be absorbed by particulate matter in seawater (Feely et al., 1990; Yang et al., 2013, 2015b) and then sink to seafloor (Boisson et al., 2001; Ma et al., 2017). The enhanced ²³⁴Th removal (Fig. 2) indicated that hydrothermal plume is a special oceanic setting characterized by active particle dynamics. Many metals from the hydrothermal effluent would settle down to seafloor below the plume, which might increase the metal accumulation in benthic mollusks (Kádár et al., 2007). Indeed, the available study indicated increased Pb flux in the hydrothermal vent area (Boisson et al., 2001). Thus, given particle-reactive metals, their accumulation in organisms in the hydrothermal plume was probably different from the general aphotic deep ocean, intensive investigation was needed before polymetallic sulfides mining.

4.2 Radioactive ²³⁴Th adsorption and sinking in the hydrothermal plume

Thermodynamically, distribution coefficients illustrated the difference in the adsorption of ²³⁴Th on particulate matter between the euphotic zone and hydrothermal plume (Fig. 4). Here, this study also attempt to examine the possible differences in the kinetics during ²³⁴Th adsorption and sinking in the two different types of oceanic environments.

In the hydrothermal plume, the adsorption constants varied from 0.007 d^–1 to 0.012 d^–1 with an average of (0.009±0.001) d^–1 (Fig. 5), comparable to (0.010±0.002) d^–1 (ranging from 0.008 d^–1 to 0.012 d^–1) in another hydrothermal plume within 2 000–3 100 m over the SWIR (Yang et al., 2016). However, these constants were much lower than (0.060±0.008) d^-1 (from 0.019 d^–1 to 0.142 d^–1) in the euphotic zone (Table 1), indicating that sulfides in the hydrothermal plume need longer time to absorb ²³⁴Th than biogenic particles in the euphotic zone. In fact, the residence times of dissolved ²³⁴Th varied from 82 d to 150 d with an average of (115±19) d (Table 1). By contrast, they showed a range of 7–52 d, averaging (37±4) d in the euphotic zone. Obviously, approximately 3 fold higher residence time supported the slow transfer of ²³⁴Th from dissolved form to particulate phase in the hydrothermal plume. Although only a few studies have been reported the residence times of ²³⁴Th in hydrothermal plumes, the results were consistent with the reported values. Yang et al. (2016) reported (108±19) d in a hydrothermal plume versus (60±22) d in the euphotic zone over the SWIR. In addition, Kadko et al. (1994) showed dissolved ²³⁴Th residence time of 18–102 d over the North Cleft segment of the Juan de Fuca Ridge. Owens et al. (2015) presented the mean residence time of 100 d in the Trans-Atlantic Geotraverse (TAG) hydrothermal plume. Pavia et al. (2019) reported the adsorption constants of 0.003 d^–1 to 0.008 d^–1, corresponding to 122–302 d in a hydrothermal plume over the Southeast Pacific Ocean Ridge. These available data suggested that particle-reactive metals seem to reside over tens of days in the hydrothermal plumes. Such a timescale allows particle-reactive metals, either emitted from hydrothermal vents or captured from ambient seawater during plume dispersal, to spread over a large spatial scale and the bathypelagic environments could be influenced extensively.

Figure  5.  Comparisons of adsorption constant (a), removal constant (b), adsorption rate (c), removal rate (d) of ²³⁴Th, residence times of dissolved (e) and particulate (f) ²³⁴Th between the euphotic zone and hydrothermal plume.

DownLoad: Full-Size Img PowerPoint

Table  1.  Adsorption (k) and removal constants (φ) of ²³⁴Th, adsorption (kA_ThD) and removal rates (φA_ThP) in the euphotic zone and hydrothermal plume, residence times of dissolved and particulate ²³⁴Th (τ_ThD, τ_ThP)

Zone	Depth/m	k/d–1	φ/d–1	kAThD/(Bq·m–3·d–1)	φAThP/(Bq·m–3·d–1)	τThD/d	τThP/d
Euphotic zone	0	0.020±0.003	0.048±0.011	0.482±0.062	0.302±0.063	51±8	21±5
	50	0.142±0.024	0.367±0.041	0.989±0.033	0.917±0.034	7±1	3±0
	100	0.019±0.003	0.109±0.021	0.474±0.063	0.375±0.064	52±8	9±2
	mean	0.060±0.008	0.175±0.016	0.649±0.031	0.531±0.032	37±4	115±19
Hydrothermal plume	1 400	0.007±0.002	0.020±0.021	0.216±0.075	0.140±0.076	150±54	19±10
	1 450	0.008±0.003	0.071±0.033	0.240±0.077	0.171±0.077	132±43	14±6
	1500	0.012±0.003	0.291±0.083	0.342±0.072	0.311±0.073	82±19	3±1
	1 550	0.010±0.003	0.035±0.016	0.303±0.075	0.166±0.076	97±26	29±13
	mean	0.009±0.001	0.113±0.024	0.275±0.037	0.197±0.038	11±2	16±5

 | Show Table

DownLoad: CSV

In the hydrothermal plume, the sinking rate constants of ²³⁴Th_P spanned an order of magnitude, varying from 0.020 d^–1 to 0.291 d^–1 (Fig. 5) and averaging (0.113±0.024) d^–1 (Table 1). These constants correspond to 3–29 d residence times (avg. (16±5) d). In another hydrothermal plume over the SWIR, the sinking rate constants varied from 0.031 d^–1 to 0.121 d^–1 and the residence times from 8 d to 32 d (Yang et al., 2016), comparable to the results of this study. In the euphotic zone, ²³⁴Th_P showed comparable sinking rate constants (from 0.048 d^–1 to 0.367 d^–1, avg. (0.175±0.016) d^–1) and residence times (3–21 d, avg. (11±2) d) (Table 1). It is clear that particles in the hydrothermal plume seemed to settle as fast as biogenic particulate did in the euphotic zone though their adsorption constants had 6-fold difference (Table 1). In addition, the sinking constants of ²³⁴Th were an order of magnitude higher than its adsorption constants (Table 1), indicating an effective sinking. Thus, hydrothermal particles probably play important role in transporting particle-reactive metals from deep seawater to sediment over the SWIR. Unlike the observation of this study, ²³⁴Th_P appeared to reside for longer time (>150 d) in the hydrothermal plumes over the Juan de Fuca Ridge and the TAG field (Kadko et al., 1994; Owens et al., 2015). This difference may result from a close distance from the sampling site to the active hydrothermal vents (Tao et al., 2012).

The adsorption rate of ²³⁴Th_D varied from 0.215 Bq/(m³·d) to 0.342 Bq/(m³·d) with an average of (0.275±0.022) Bq/(m³·d) (Table 1), comparable to (0.292±0.042) Bq/(m³·d) in another plume over the SWIR (Yang et al., 2016) and 0.300 Bq/(m³·d) at the TAG (Owens et al., 2015). However, these rates were lower than 0.295−0.783 Bq/(m³·d) estimated using dataset over the Juan de Fuca Ridge (Kadko et al., 1994). The sinking rate of ²³⁴Th_P ranged from 0.140 Bq/(m³·d) to 0.310 Bq/(m³·d) and averaged (0.197±0.022) Bq/(m³·d) (Fig. 5d), which was the same as observations in adjacent hydrothermal plume within 2 900–3 100 m over the SWIR (from 0.130 Bq/(m³·d) to 0.287 Bq/(m³·d), avg. (0.197±0.042) Bq/(m³·d) (Yang et al., 2016). Thus, it seems that the adsorption and sinking rates of ²³⁴Th have similar characteristics in hydrothermal plumes at different sites over the SWIR (Table 1; Yang et al., 2016) though they were different from those in the Pacific Ocean (Kadko et al., 1994). This difference indicated the different particle kinetics in terms of various hydrothermal plumes, probably due to the different composition of the hydrothermal particle from different vents.

4.3 The enhancement of ²³⁴Th and other particulate components sinking

The adsorption flux of ²³⁴Th was (52.4±5.4) Bq/(m²·d) in the hydrothermal plume, which was comparable to available ²³⁴Th adsorption flux in the plume over the SWIR ((46.0±5.6) Bq/(m²·d); Yang et al., 2016), at the TAG (45.4 Bq/(m²·d); Owens et al., 2015), but lower than 44.3 Bq/(m²·d) to 129.7 Bq/(m²·d) at the Juan de Fuca Ridge (Kadko et al., 1994). The sinking flux of ²³⁴Th was (36.2±5.4) Bq/(m²·d) at the study site, which was much higher than 16.0 Bq/(m²·d) reported for another SWIR plume (Yang et al., 2016). Combining with all the kinetic parameters (i.e., rate constant, rate, flux and residence time), it appeared that similar characteristics of the hydrothermal plume were seen over the ultraslow-spreading SWIR to TAG and large difference from the Juan de Fuca Ridge.

In addition to particle-reactive metals, the cycling of some particle components (e.g., SPM, nitrogen and carbon) could change with the hydrothermal plumes. Previous studies reported that particles have the capability to disperse hundreds of kilometers with the hydrothermal plumes (Siegel et al., 2008; Wang et al., 2012; Estapa et al., 2015). Based on a preliminary estimation using Eq. (5), the sinking flux of SPM out of the plume reached (7.5±1.3) g/(m²·d). This magnitude is comparable to the cross-shelf process induced high particle flux in the mesopelagic South China Sea (Ma et al., 2017), indicating the crucial role of hydrothermal plume in affecting the particle cycling in the ocean interior. Hydrothermal particles usually contain abundant Fe and Mn compounds (Sun et al., 2014), which were expected to induce DOC aggregation in the hydrothermal plumes (Toner et al., 2009) and POC could be effectively delivered to sediments (German et al., 2015). The only available high POC sinking flux from a hydrothermal plume also supports these views (Yang et al., 2016). In order to evaluate the role of the studied hydrothermal plume in affecting carbon and nitrogen cycling, this study conservatively adopt the minimum POC and PON values reported in an adjacent plume (Yang et al., 2016), i.e., 0.72 μmol/L (in terms of carbon) and 0.17 μmol/L (in terms of nitrogen). The calculated POC and PON fluxes were (7.4±1.3) mmol/(m²·d) (in terms of carbon) and (1.8±0.3) mmol/(m²·d) (in terms of nitrogen), respectively. These values seem to be comparable with reported values of (9.3±0.6) mmol/(m²·d) (in terms of carbon) and (2.2±0.6) mmol/(m²·d) (in terms of nitrogen) from another hydrothermal plume over the SWIR (Yang et al., 2016). The POC flux was much higher than those of <2 mmol/(m²·d) (in terms of carbon) observed at 2 000 m (Rixen et al., 2019) and comparable to the maximal POC flux (9.0 mmol/(m²·d) (in terms of carbon)) out of the euphotic zone in the northern Indian Ocean (Anand et al., 2018), indicating other POC contribution in addition to the surface ocean to the hydrothermal plumes. Also, the POC flux in our study was even higher than that of <6 mmol/(m²·d) (in terms of carbon) observed at 100–2 000 m in the South China Sea where was identified by cross-shelf lateral particle transport (Shih et al., 2019), indicating the significant influence of hydrothermal plume on POC sinking in deep oceans. The POC flux was calculated using the POC/²³⁴Th_P ratio on SPM collected by bottle sampler. Earlier research speculated that large particles (>53 μm) represent the sinking POC (Coppola et al., 2002). Recently, some studies found small size particles accounted for a major fraction of the sinking particles in the open oceans (Hung and Gong, 2010; Hung et al., 2012). Since the POC/²³⁴Th_P ratio sometimes varied with particle size (Hung et al., 2012), it is difficult to validate the specific size of representing the sinking POC currently. Bottle-collected particles might represent the upper limit of POC flux (Cai et al., 2008). Thus, this study’s estimates may constrain the maximal POC flux out of the studied plume. POC flux is usually less than 2 mmol/(m²·d) (in terms of carbon) in the general deep ocean (Smith and Rabouille, 2002; Ran et al., 2015). Obviously, POC and PON fluxes to the seafloor was increased by the hydrothermal plumes. Together with the radioactive ²³⁴Th flux, high fluxes of particulate-related components (e.g., metals, some nutrients, and organic matter) might be a driving force of fueling the flourishing hydrothermal organisms and metal accumulation on mollusks.

4.4 Implications for the hydrothermal plume

Based on current information, distinctive characteristics were observed in the hydrothermal plume from general deep oceans. First, the Fe- and Mn-rich particles, such as Fe- and Mn-sulfides (Sun et al., 2014) in the hydrothermal effluent plume (Pavia et al., 2019) can disperse with the plume over a large spatial range (Siegel et al., 2008; Wang et al., 2012; Estapa et al., 2015). During this dispersal, particle-reactive metals such as Th, Pb, Po, were absorbed on these particles and aggregation into POC by binding with DOC (Toner et al., 2009) over the timescale of days- to a hundred of days (Feely et al., 1990; Boisson et al., 2001; Owens et al., 2015; Yang et al., 2016). Then, metal-rich particles settled down to seafloor along the plume passage over the timescale of days to tens of days (this study and Yang et al. (2016)), possible causing the metal accumulation in benthic organisms. Inorganic particles acted simultaneously as the ballast and the settling of organic components induced. Finally, the enhanced sedimentation of particulate organic matter might deliver abundant metals to benthic ecosystems. The previous study suggested that POC source largely determines the activities of sediment-mixing benthos (Smith and Rabouille, 2002), the higher POC flux to the seafloor, the much deeper of the mixed layer in sediments (Smith et al., 2008). Thus, the enhanced organic matter flux may result in different sediment environments below the hydrothermal plumes comparing with general deep oceans.

As shown in Fig. 6, all these unusual characteristics probably develop a special environment region in deep oceans. Although this study have a glimpse of the hydrothermal effluent plume, it is still a veiled system. As reviewed by Santos et al. (2018), the development of new tools and technologies would help the understanding of deep-sea mining-related environmental assessment. This study indicated that radioactive nuclides, e.g., Th isotopes (Owens et al., 2015; Yang et al., 2016; Pavia et al., 2019), would be effective tool for evaluating particle and metal cycling in hydrothermal effluent plumes.

Figure  6.  The particle-reactive elements adsorption, sinking and its environmental implications in the hydrothermal plume.

DownLoad: Full-Size Img PowerPoint

5. Concluding remarks

Unusual deficits of radioactive ²³⁴Th with respect to its parent ²³⁸U allow us to examine the kinetics of particle-reactive metals in the hydrothermal plume over the SWIR. Unlike general deep oceans, the hydrothermal plume showed enhanced particle-reactive metal adsorption and sinking. Both adsorption and sinking constants of ²³⁴Th suggested that particle-reactive elements (e.g., Th, Pb, Po, P) could be effectively removed by hydrothermal particles from the plumes over the timescales of days to tens of days, and these elements were transported to bottom water and seafloor, which would increase the metal accumulation in benthic organisms. In addition, hydrothermal particles have the capability to aggregate DOC and deliver it efficiently to sediments to sustain the benthos. These results indicated that the hydrothermal plume, as a typical environment region, has its own characteristics different from general deep oceans, and thus intensive and extensive investigations were required before deep-sea mining activities.

Acknowlegements

We appreciate the reviewers for their constructive comments and the help from the crew of R/V Dayang I during sampling.