Geophysical Research Letters

Export flux of POC in the main stream of the Kuroshio

Authors


Abstract

[1] Particulate organic carbon (POC) export fluxes were measured using both a drifting sediment trap and a 234Th approach (ratio of POC/234Th multiple 234Th flux) in the main steam of the Kuroshio in August, 2006. The POC fluxes measured by sediment traps and the 234Th approach at 65, 90, 120 and 140 m ranged from 20 to 54 mg m−2 d−1, and 26 to 67 mg m−2 d−1, respectively. The results show that using POC/234Th ratios in intermediate particles (10–53 μm) and sinking particles is more comparable to the POC flux measured by sediment traps. Therefore, the POC export fluxes within the euphotic zone estimated by sediment trap and 234Th approaches are complementary methods for measuring POC flux in the Kuroshio.

1. Introduction

[2] The export flux of particulate organic carbon (POC) plays a crucial role in the transfer of carbon between the atmosphere and the world oceans. Therefore, an accurate estimate of POC export flux within the euphotic zone is important for our understanding of the biogeochemical cycling of carbon in the ocean. Sediment traps are often used to measure POC export flux directly, despite possible biases by hydrodynamic and biological (swimmer) effects [Gardner, 1980; Lee et al., 1988; Karl and Knauer, 1989; Jurg, 1996]. A knowledge of the extent of 234Th deficiency relative to 238U in the euphotic zone, combined with an assessment of the POC/234Th ratio times the calculated 234Th flux in sinking particles, has been the basis for estimating POC fluxes in various marine environments [e.g., Buesseler et al., 1992; Cochran et al., 1995; Murray et al., 1996; Benitez-Nelson et al., 2000, 2001]. Recent studies have shown that the ratios of POC/234Th can vary with water depth, particle size, particle composition and different hydrographic regimes [e.g., Cochran et al., 1995; Baskaran et al., 1996, 2003; Buesseler, 1998; Santschi et al., 1999, 2003; Guo et al., 2002; Moran et al., 2003; Hung et al., 2004; Trimble and Baskaran, 2005], while 234Th and sediment trap approaches are complementary means for estimating the export flux of POC in the upper ocean because sediment trap efficiency can be evaluated by the 234Th flux in the water column. In recent years, a combination of 234Th and sediment traps has been widely used to estimate the POC flux in the open ocean [Rodier and Le Borgne, 1997; Murray et al., 1996, 2005; Hernes et al., 2001; Hung et al., 2004].

[3] The Kuroshio, the world's second-largest ocean current, originates from the North Equatorial Current and flows northward at a speed of 1–3 knots off the east coast of Taiwan toward Japan along the continental slope bordering the East China Sea [Nitani, 1972; Gong et al., 1999]. Previous research has characterized the Kuroshio Current as a nutrient-depleted water with correspondingly low chlorophyll a (Chl a) and low primary productivity (PP) (104 g C m−2 yr−1) all year round [Gong et al., 1999]. Although the Kuroshio Current flows through a considerable region almost covering the whole North Pacific Ocean, to our knowledge no information on POC export flux in the main stream of the Kuroshio have been presented. In this paper, we first report the POC export flux as determined by drifting sediment traps and the 234Th approach in the main stream of Kuroshio.

2. Materials and Methods

[4] Seawater and size-fractionated suspended and sinking (via sediment traps) particulate matter from the Kuroshio Current (bottom water depth 1947 m, 121.7°E, 22.86°N) were collected aboard the R/V Ocean Researcher II during August 18–21, 2006 (Figure 1). Temperature was recorded using a SeaBird model SBE9/11 plus conductivity-temperature-depth (CTD) recorder and salinity was determined with an Autosal salinometer. Nitrate, phosphate, Chl a concentrations and light-saturated PP were determined according to Gong et al. [1996, 2003]. Suspended particles, defined as those particles retained by filtration with the following filter cut-off: 10 μm, 50 μm, 150 μm (to remove swimmers), were collected using sequential filtration and the size fractionated particles (up to several hundred liters) were then re-filtered onto silver filters for measurements of 234Th and POC [Santschi et al., 2006]. In addition, particulate (>1 μm) and dissolved (<1 μm) 234Th were also determined with large water volumes (up to 1,100 liters) as described by Guo et al. [2002] and Hung et al. [2004]. Briefly, large volumes of seawater were collected by a submersible pump equipped with a series of polypropylene filters with pore sizes of 1 μm followed by two MnO2-impregnated 0.5 μm pore sized filters [Buesseler et al., 1992; Baskaran et al., 1993, 1996; Cochran et al., 1995]. The three-filter set-up allowed for the collection of two different size fractions of 234Th (>1 and <1.0 μm). The activity of 234Th was measured either by a low background beta detector (RISØ GM-25-5, Denmark, Benitez-Nelson et al. [2001]) or a Canberra ultra high purity Germanium well detector [Santschi et al., 1999; Guo et al., 2002]. Sinking particles were collected with a drifting sediment trap array attached to a surface buoy, and were deployed for 24 h at depths of 65, 90, 120, 140 m [Santschi et al., 2003; Hung et al., 2003, 2004]. Prior to analyses, swimmers on all filters were carefully removed using forceps. Sinking particles were filtered through silver filters for the determination of POC and 234Th concentrations. POC concentrations from both suspended and sinking particles were quantified on a Perkin-Elmer CHNS/O elemental analyzer after concentrated HCl-fumed [Guo and Santschi, 1997].

Figure 1.

Map showing bathymetry, current vectors, and sampling locations within the Kuroshio Current. Red circles indicate the location of sediment trap deployment and recovery. Base map derived from historical summer means taken from the NCOR (National Center for Ocean Research) Ocean Data Bank.

3. Results and Discussion

3.1. Hydrographic Settings and Disequilibrium of 234Th and 238U

[5] Based on year-round investigations of the Kuroshio Current by Gong et al. [1999], it is known that Kuroshio surface water has a stable temperature range (28–29°C) (Figure 2) and salinity (33.8–34.8). Previous work has reported that nitrate concentrations are often below detection limits (0.3 μM) within the mixed layer of the Kuroshio throughout the year [Gong et al., 1999]. In this study, both nitrate and phosphate concentrations were very low in the surface layer suggesting that only little nutrient was transported to the euphotic zone (Figure 2). Typically the maximum Chl a concentration in Kuroshio surface waters was <0.2 mg m−3, while the maximum Chl a measured here was 0.52 mg m−3, more than a factor of 2 higher than previously documented average values (Figure 2) [Gong et al., 1999]. Correspondingly elevated PP was also observed, indicating that Kuroshio surface waters are more productive than in the past [Gong et al., 1999]. Currently, we are unable to explain the possible reasons due to limited data. The depth of the euphotic zone (1% of surface light intensity) in the Kuroshio is approximately 110 ± 10 m [Gong et al., 1999]. Here we used 120 m as the basis to estimate POC export flux in the water column.

Figure 2.

Vertical profiles of temperature (T), salinity (S), nitrate (NO3), phosphate (PO4), chlorophyll a (Chl a), and primary productivity (PP) in the Kuroshio.

[6] Activity concentrations of particulate (1–10 μm, 10–50 μm and >50 μm), dissolved and total (particulate + dissolved) 234Th as well as 238U (238U (dpm L−1) = 0.0709 × salinity, Chen et al. [1986]) are shown in Table 1 and Figure 3. Total 234Th was deficient relative to its parent, 238U, in the uppermost water column (0–120 m). In general, small particles (1–10 μm) had the highest percentage (36–84%) of particulate 234Th, followed by intermediate (12–62%, 10–50 μm) particles and then the large particles (>50 μm). This disequilibrium between 234Th and 238U demonstrates the depth within the euphotic zone (0–120 m) where 234Th was effectively scavenged by marine particles.

Figure 3.

(a) Vertical profiles of particulate and dissolved 234Th and 238U (dpm L−1) in the Kuroshio. (b) Ratios of POC/234Th in different size fractions (1.0–10, 10–50, >50 μm) in the Kuroshio.

Table 1. Particulate 234Th (dpm L−1) in the Different Suspended Particle Fractions
Depth, mPart. 234Th, >1.0 μmPart. 234Th, 1.0–10 μmPart. 234Th, 10–50 μmPart. 234Th, >50 μm
50.07361%38%1%
250.10736%62%2%
500.07776%24%3%
750.13662%32%4%
1000.12276%23%1%
1250.12084%12%2%

3.2. Ratios of POC/234Th, Th Flux and Trapping Efficiency

[7] The ratios of POC/234Th in different size fractions (1.0–10, 10–53 and >53 μm) with depth are shown in Figure 3. Average ratios of POC/234Th in small (1.0–10 μm), intermediate (10–50 μm) and large particles (>50 μm) showed large variations (42.5 ± 33.9, 3.7 ± 2.1, 36.4 ± 20.7 μmol dpm−1, respectively) within the euphotic zone. However, average ratios of POC/234Th in intermediate (10–50 μm) particles below 50 m showed variations (3.9 ± 1.8 μmol dpm−1) (Table 2). Ratios of POC/234Th in sinking particles collected from 65, 90, 120 and 140 m were 4.3 ± 0.3, 2.8 ± 0.1, 1.3 ± 0.1 and 1.2 ± 0.1 μmol dpm−1 (Table 2), respectively.

Table 2. POC/234Th in Different Suspended Particle Fractions and in Sinking Particlesa
Depth, m234Th flux, Traps234Th flux, ModelTrap eff., %Ratio aRatio bRatio c
  • a

    a, sinking particles; b, 10–50 μm; c, 1.0–10 μm. The unit of 234Th flux is dpm m−2 d−1. Trap efficiency, measured 234Th flux/calculated 234Th flux. The unit of ratio is μmol dpm−1. Trap eff., trap efficiency; n.a., data not available.

651045 ± 251312 ± 10280 ± 64.3 ± 0.33.4 ± 1.138.3 ± 2.2
901420 ± 301546 ± 11492 ± 72.8 ± 0.11.9 ± 0.416.2 ± 1.1
1201754 ± 331704 ± 143103 ± 91.3 ± 0.11.9 ± 0.210.2 ± 0.6
1401454 ± 291871 ± 10778 ± 51.2 ± 0.1n.a.n.a.

[8] The POC export flux is derived from 234Th fluxes by:

equation image

where (POC/234Th)p is the ratio of POC/234Th (μmol dpm−1) measured in different sized particles (p) collected at a given depth, [flux of 234Th] is the calculated flux of 234Th derived from the deficiency of 234Th relative to its parent 238U in the water column. The 234Th fluxes at 65, 90, 120 and 140 m were estimated to be 1312, 1546, 1704 and 1871 dpm m−2 d−1 (Table 1) by a simple one-dimensional box model [Coale and Bruland, 1985; Buesseler et al., 1992; Guo et al., 2002; Baskaran et al., 2003; Trimble and Baskaran, 2005] assuming steady state conditions and negligible impacts by advective or diffusive processes. Calculated 234Th fluxes were used to evaluate the trapping efficiency of the drifting sediment traps by comparing them with measured 234Th fluxes. Trapping efficiencies of the drifting sediment traps at 65, 90, 120 and 140 m were 80 ± 4%, 92 ± 7%, 103 ± 9% and 78 ± 5% (Table 2), with an average value of 88 ± 10%, suggesting that the drifting sediment traps caught nearly all sinking particles in the Kuroshio and that both 234Th fluxes from traps and the model are in good agreement.

3.3. Comparison of POC Export Flux as Determined by Sediment Traps and 234Th

[9] It is known that the sediment trap only provides a snap shot (several hours to a few days) for measuring POC flux and the 234Th approach provides information taking place over the mean-life of 234Th within the water column. However, a comparison of the POC fluxes obtained from sediment trap to that by the 234Th method would provide information on the stability of the ecosystem and export fluxes of POC. The POC export flux determined directly by sediment traps at 65, 90, 120 and 140 m, ranged from 54 to 20 mg C m−2 d−1 (Table 3). The 234Th-derived POC export flux varied from 67 to 26 mg C m−2 d−1, using the ratio of POC/234Th in sinking particles. If the POC export flux is calculated by using the ratio of POC/234Th in intermediate (10–50 μm) and small (1.0–10 μm) particles multiplied by the estimated 234Th flux at different integrated depths, one would obtain POC export from 35 to 54 mg C m−2 d−1, and 237–633 mg C m−2 d−1, respectively (Table 3). Based on these calculations, POC export flux measured directly by drifting sediment traps is in a good agreement with 234Th-derived POC export flux using the ratio of POC/234Th in sinking particles and intermediate particles. Others have reported similar successes using these approaches in various marine environments [Murray et al., 1996; Hernes et al., 2001; Hung et al., 2004].

Table 3. POC Export Flux Measured by Sediment Traps Directly and Estimated by 234Th Approach Using Different Ratios of [POC/234Th × 234Th flux]a
Depth, mExport, TrapsExport AExport BExport CAverage PPe Ratio Export/PP
  • a

    The unit of POC export flux is mg C m−2 d−1. The unit of PP is mg C m−2 d−1 and e ratio is equal to export flux/PP. n.a., data not available.

6553.5 ± 2.867.2 ± 7.054 ± 18633 ± 36391.30.14
9047.2 ± 0.951.3 ± 4.235.1 ± 7.8317 ± 22423.40.11
12027.9 ± 0.927.1 ± 3.139.1 ± 5.3237 ± 14430.00.06
14020.1 ± 1.725.9 ± 2.6n.a.n.a.430.00.05

[10] Average PP, measured in four seasons in the main course of the Kuroshio near the northeast coast of Taiwan ranged from 227 to 335 mg C m−2 d−1, with little seasonal variation [Gong et al., 1999]. Average PP in this survey ranged from 391 to 430 mg C m−2 d−1, from 65 m to 120 m (Table 3). The average PP at 120 m in this survey is approximately 1.5 times higher than that reported previously (430 vs. 285 mg C m−2 d−1). The derived e-ratio (export flux/PP) decreased with increasing depth from 0.14 to 0.05 (Table 3), with a value of 0.05 near the bottom (120 m) of euphotic zone which is similar to the reported data (0.04 ∼ 0.08) in other open ocean regions [Murray et al., 1996; Rodier and Le Borgne, 1997]. On the other hand, the result also demonstrates that dissolved organic carbon export flux into the deep ocean is expected to contemporaneously increase. To our knowledge the POC export flux has been never presented in the main stream of the Kuroshio. This is the first study to use floating sediment traps and 234Th approaches to measure POC flux in the main stream of the Kuroshio. More research is needed to understand the temporal and spatial variations of POC export flux in the Kuroshio.

Acknowledgments

[11] We appreciate the assistance of the crew of the R/V Ocean Research II, C. Xu, K. Roberts, W.T. Kuo, W.C. Chuang, M.Y. Lin, and P.G. Hsiu as well as the comments of K. Schwehr and K.M. Yeager. We also thank one anonymous reviewer who gave constructive comments which strengthened the paper. This research was supported by the National Research Council of Taiwan (NSC95-2611-M-019-001, NSC96-2611-M019-011, NSC95-2611-M-019-020-MY3, NSC95-2611-M-019-021-MY3) and the Center of Marine Bioscience and Biotechnology at the National Taiwan Ocean University.

Ancillary