Long-term variation of mesopelagic biogenic flux in the central South China Sea: Impact of monsoonal seasonality and mesoscale eddy

Author’s Accepted Manuscript Long-term Variation of Mesopelagic Biogenic Flux in the Central South China Sea: Impact of Monsoonal Seasonality and Mesoscale Eddy Hongliang Li, Martin G. Wiesner, Jianfang Chen, Zheng Lin, Jingjing Zhang, Lihua Ran www.elsevier.com

PII: DOI: Reference:

S0967-0637(16)30255-2 http://dx.doi.org/10.1016/j.dsr.2017.05.012 DSRI2802

To appear in: Deep-Sea Research Part I Received date: 12 August 2016 Revised date: 16 May 2017 Accepted date: 23 May 2017 Cite this article as: Hongliang Li, Martin G. Wiesner, Jianfang Chen, Zheng Lin, Jingjing Zhang and Lihua Ran, Long-term Variation of Mesopelagic Biogenic Flux in the Central South China Sea: Impact of Monsoonal Seasonality and Mesoscale Eddy, Deep-Sea Research Part I, http://dx.doi.org/10.1016/j.dsr.2017.05.012 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Long-term Variation of Mesopelagic Biogenic Flux in the Central South China Sea: Impact of Monsoonal Seasonality and Mesoscale Eddy Hongliang Li 1, 2, Martin G. Wiesner 3, Jianfang Chen 2, 4*, Zheng Lin 4, Jingjing Zhang 2, Lihua Ran 2 1

State Key Laboratory of Marine Geology, Tongji University, Shanghai 200092, P.R. China; Laboratory of Marine Ecosystem and Biogeochemistry, State Oceanic Administration, Hangzhou 310012, P.R. China; 3 Institute of Geology, University of Hamburg, Hamburg D-20146, Germany; 4 State Key Laboratory of Satellite Ocean Environment Dynamics, Hangzhou 310012, P.R. China. 2

*

Corresponding author, ; Tel.: +86-571-81963211; Fax: +86-571-81963213. E-mail: [email protected]

Abstract: The East Asian Monsoon and mesoscale eddies are known to regulate primary production in South China Sea (SCS), the largest tropical marginal sea; however, their contribution to the deep biogenic flux are yet to be quantified. Based on 7-year time series sediment trap observations at the depth of 1200 m in the central SCS, we used the monthly average sinking biogenic fluxes to evaluate the impact of the monsoon and mesoscale cyclonic eddies on biogenic fluxes in combination with remote sensing physical parameters. The monthly average particulate organic carbon (POC) and opal fluxes, ranging from 3.0–5.2 and 14.8–34.9 mg m−2 d−1, respectively, were higher during the northeastern monsoon period. This corresponded to the deeper mixed layer depth and higher net primary production in this area, due to nutrient replenishment from the subsurface induced by monsoon transition and surface cooling. In contrast, lower POC and opal fluxes occurred 1

during well-stratified inter-monsoon periods. In addition, CaCO3 flux (23.6–37.0 mg m−2 d−1) exhibited less seasonality and was assumed to originate from foraminifera. In terms of the long-term record, the combined effect of cyclonic eddies and mixing in the upper ocean could effectively regulate the temporal variation in the biogenic flux. In particular, the opal and POC fluxes in cyclonic eddies were 116% and 41% higher on average, respectively, than those during the non-cyclonic eddy period. Since the cyclonic eddies mainly occurred during the northeastern monsoon period, their contribution to biogenic flux via diatom blooms might overlap the regular winter flux peak, which could make the biological carbon pump more efficient at CO2 sequestration during this period thus amplifying the impact of seasonal transition. Keywords: sinking biogenic flux; seasonal variation; mesoscale eddy; biological pump; South China Sea

1 Introduction The net uptake of atmospheric CO2 by marine organisms through photosynthesis in the sunlit layer and the subsequent transport of the sinking particles into the ocean interior are key components of the biological carbon pump in the ocean [Volk and Hoffert, 1985]. The flux of sinking particulate organic carbon (POC) into the deep ocean (>1000 m) removes CO2 from the atmospheric reservoir for a period of centuries to millennia [Sigman and Boyd, 2000]. Furthermore, the biogenic components (POC, CaCO3 and opal) are precursors of geological sedimentary proxies for paleoclimate studies. Understanding of

2

the processes that control the biogenic flux in the deep ocean is crucial to comprehend how the biological pump regulates atmospheric CO2 and consequently affects global climate system [Archer and Maier-Reimer, 1994]. The strength of the export production is thought to be usually governed by the magnitude of net primary production (NPP) [Eppley and Peterson, 1979] and nutrients supplied by multi-scale physical forcing represent one of the key factors controlling the magnitude of NPP and subsequent particle fluxes [Honjo et al., 1999]. In this study, we focus on the temporal variability of the particle fluxes in the central South China Sea (SCS) basin which is situated between the Western Pacific Warm Pool and the Tibetan Plateau with an area of about 3.5 × 106 km2 and a maximum depth of 5500 m. It is also the largest semi-enclosed marginal sea in Southeast Asia (Fig. 1a). Generally, the upper ocean circulation of the SCS basin is controlled by the seasonal reversal of the East Asia monsoon [Su, 2004]. The northeasterly monsoon (NEM) in winter (November to March) drives a cyclonic gyre over the entire SCS basin with an intensified southward jet along the western boundary. In contrast, the basin-scale circulation splits into a weak cyclonic gyre north of 12°N, while an anti-cyclonic gyre in the southern SCS basin forced by the southwesterly monsoon (SWM) prevails from June to August in summer. In response to the forcing of East Asia monsoon, the lowest chlorophyll a (Chl-a) concentration has been observed during spring inter-monsoon period (SIM, Fig. 1b) from April to May [Wong et al., 2007]. While a medium Chl-a concentration in the SCS basin was accompanied by high biomass in the jet off Vietnam during the SWM and autumn inter-monsoon (AIM, from September to October) periods (Fig. 1c and 1d). Conversely, more nutrients supplied 3

diapycnally from the thermocline fuel the highest productivity during the NEM period in winter (Fig. 1e) [Liu et al., 2002]. Besides the seasonal East Asian Monsoon transient, some randomly occurring non-seasonal ocean–atmosphere physical processes would also affect the biogeochemistry, such as mesoscale eddies [Zhou et al., 2013], short-term internal waves [Pan et al., 2012], dust deposition [Wang et al., 2012], tropical cyclones [Lin et al., 2003] and the inter-annual El Niño [Liu et al., 2013]. For example, mesoscale eddies are often observed in the SCS, with a frequency of about 17 per year during the last 20 years [Xiu et al., 2010]. Among these, cyclonic cold eddies generally contain nutrient-rich subsurface water, which plays a crucial role in the replenishment of nutrients and in stimulating phytoplankton bloom in the oligotrophic SCS basin [Tang et al., 1999; Ning et al., 2004; Chen et al., 2006; Chen et al., 2007]. Sparse in-situ data from time series observations result in limited study of the vertical particle flux below the euphotic layer in the SCS basin [Wiesner et al., 1996; Chung et al., 2004; Chen, 2005a; Hung and Gong, 2010; Liu et al., 2014]. The total particle fluxes, POC, CaCO3, opal and amino acid fluxes of the trap-collected particles in the northern, central and southern SCS basin have been shown to display clear temporal variability and are mainly related to monsoon transitions [Wiesner et al., 1996; Chen, 2005a; Lahajnar et al., 2007]. In addition, apparent temporal variation but no regular seasonality in δ15N ratios was found in the trap-collected particles in the same suite of particles [Gaye et al., 2009]. Similarly, fluxes of specific plankton, such as diatoms, coccolithophores, trichodesmium, foraminifera and radiolaria could be subjected to the forcing of the East Asian monsoon [Wang et al., 2000; Chen, 2005b, Chen et al., 2006, 2007; Lin and Hsieh, 4

2007; Wan et al., 2010]. However, long-term variations in biogenic components were found to be significant in the long term time series observations in sediment trap in the deep SCS, such as episodic high sinking flux during inter-monsoon period or extremely low and homogeneous flux throughout the year during the 1997–1998 El Niño event [Chen, 2005a]. Hence, the temporal variability in the sinking biogenic particle flux is probably the result of a synergistic response to multi-scale physical processes in the SCS. The objective of this study was to explore the regular seasonal pattern of sinking biogenic particle flux based on 7-year sediment trap observations, which provide insight into the response of the biological carbon pump to non-seasonal physical processes and rapid global change in the SCS. In this study, we identified the factors regulating temporal variation in sinking biogenic flux of the SCS. In particular, we quantified the relative contribution of cyclonic mesoscale eddies to the export of organic matter.

Figure 1. (a) The South China Sea (SCS) bathymetry and locations of the SCS-C station and the South-East Asian Time-series Study (SEATS) Station (black star). The dotted square indicates the 1° × 1° study area centered at the SCS-C station in the central SCS basin. The solid and the dashed curves with arrows indicate the cyclonic gyre in the winter and summer jets in the SCS basin, respectively. Satellite-derived climatology of Chl-a during (b) spring inter-monsoon (April–May), (c) southwestern monsoon (June–August), (d) autumn inter-monsoon (September–October), and (e) northeastern monsoon (November–March of the next year) periods in the SCS. The filled red inverted triangle represents the location of the SCS-C station in the central SCS.

5

2 Methods 2.1 Sediment trap samples The sediment trap moorings were deployed at the SCS-C station (115.0°E, 14.5°N) in the central SCS where the water depth is 4200 m, from March 1992 to April 1999 (Fig. 1a). A Mark VI (McLane, USA) time series sediment trap with a collection area of 0.5 m2 was deployed at 1200 m with a sample collection interval of 27–30 days. According to the model provided by Siegel and Deuser (1997), the potential trajectories of sinking particles have been estimated in the central SCS (See the auxiliary material). From June 1995 to June 1996 the trap depth was at 1830 m, and the biogenic components were normalized to 1200 m according to the power law by Martin et al. [1987]. The sample collection and processing procedures have been described by Wiesner et al. (1996) and Lahajnar et al. [2007]. The collecting cups were filled with trap-depth filtered seawater before trap deployment. Analytical grade NaCl (35 g l−1) and HgCl2 (3.3 g l−1) were added to the 250-ml polyethylene collecting bottles to minimize diffusive processes and inhibit bacterial degradation. After trap recovery, the wet samples were passed through a 1-mm mesh nylon sieve to remove zooplankton. Subsequently, the <1 mm fractions were split into four equal aliquots for analysis by a high-precision rotary splitter (McLane WSD-10), and filtered onto pre-weighed polycarbonate filters (0.45 µm pore size). Finally, the samples were dried at 45°C for 72 h and weighed for measurements of total particle flux. 2.2 Elemental analysis 6

The elemental analytical procedures followed the methods described by Wiesner et al. [1996] and Lahajnar et al. [2007]. The determinations of particulate total carbon (TC) and particulate nitrogen (PN) were performed using a Carlo Erba Science 1500 CNS Analyzer with a standard deviation of <0.15% and <0.005%, respectively. Particulate inorganic carbon (PIC) concentration was measured by a Wösthoff Carmhograph 6 with a standard deviation <1%. The samples were heated in 2.0 M phosphoric acid and the evolving CO2 was vented into a 0.05 M NaOH solution. The change in conductivity was measured and calibrated against a standard of pure CaCO3. The CaCO3 was determined from PIC by multiplying the molar mass ratio 8.3 of CaCO3 to C. The POC was calculated as the difference between TC and PIC, while particulate organic matter (POM) was converted from POC by multiplying the empirical factor of 1.8 [Müller et al., 1986]. Biogenic silica analysis was performed using a modified method of Mortlock and Froelich [1989]. The dry sample was first treated with 10% H2O2 solution and 1M HCl solution to eliminate organic matter and CaCO3. Then it was leached with 2 M Na2CO3 solution in water bath at 85°C for 5 h. The dissolved silicate concentration of the leachate was analyzed with the silicomolybdenum blue method using a spectrophotometer. Opal content was converted from biogenic silica by multiplying the molar mass ratio of 2.4 [Mortlock and Froelich., 1989]. Finally, the lithogenic material flux was calculated as the difference of total particle flux and biogenic fluxes (CaCO3, POM and opal). 2.3 Remote-sensing and modeling data To explore the regulating mechanisms linked to the particle flux seasonal variation, relevant physical and biogeochemical parameters were obtained from satellite and model 7

analysis. Monthly wind speed (WS) data with a spatial resolution of 0.25° × 0.25° were assembled from the Blended Sea Winds data provided by the National Climate Data Center

of

the

National

Oceanic

and

Atmospheric

Administration

(NOAA)

(http://www.ncdc.noaa.gov/thredds/dodsC/) for the period from January 1992 to December 1999. Monthly net heat flux (HF) data with spatial resolution of 1.0° × 1.0° for the period of January 1992 to December 1999 were obtained from the Comprehensive Ocean

and

Atmosphere

Data

Set

(http://apdrc.soest.hawaii.edu/dods/public_data/Reanalysis _Data/GODAS/pentad/). The monthly mixed layer depth (MLD) from January 1992 to December 1999 was calculated from the monthly temperature data set [Ishii and Kimoto, 2009] with a 1° × 1° grid, based on the definition that the MLD is the depth where the temperature is 0.5°C lower than the sea surface temperature [Monterey and Levitus, 1997]. For WS, HF and MLD, the climatological monthly mean value was calculated for each month of the year, and their anomalies were computed as the difference between the monthly value and the climatological mean for the particular month in the year. The daily sea surface height anomaly (SSHA) data with space resolution of 0.25° × 0.25° for the period of 1993 to 1999 were from the US Navy Modular Ocean Data Assimilation System (MODAS) product, which was derived from the TOPEX/POSEIDON(T/P), European Remote Sensing (ERS) and Geosat Follow On (GFO). Monthly SSHA data were computed from the daily dataset. The daily SSHA data from May to September 1992 were obtained from the

Colorado

Center

for

Astrodynamics

Research

(http://eddy.colorado.edu/ccar/data_viewer/index). The daily surface current data were 8

taken

from

the

Ocean

Surface

Current

Analysis

Real-time

(OSCAR)

(http://www.oscar.noaa.gov/) during January 1993 to December 1999, which provides real time reanalysis of currents averaged over the top 15 m at a spatial resolution of 1° × 1°. Monthly nitrate and Chl-a data of surface water for the period 1998 to 2012 were retrieved

from

the

NASA

Ocean

Biogeochemical

Model

(NOBM:

http://gmao.gsfc.nasa.gov/research/oceanbiology/), which was a comprehensive and interactive ocean biogeochemical model coupled with a circulation and radiative model for the global oceans [Gregg and Casey, 2007]. Monthly net primary production (NPP) of euphotic layer for the period 1998 to 2012 was from NASA Ocean Biology Distributed Active Archive Center (http://www.science.oregonstate.edu/ocean.productivity/index.php),

which was estimated by the vertically generalized productivity model (VGPM) [Behrenfeld and Falkowski, 1997]. The climatological monthly mean value of Chl-a and NPP was calculated for each month of the year. Both field observations [Ning et al., 2004; Tseng et al., 2005] and model study [Liu et al., 2002] showed a similar seasonal pattern of Chl-a and primary production presented in this study. Comparison with field data (0.1– 0.4 µM and 0.03 µM for DIN during the NEM period and rest of the year, reported by Wong et al. [2007]) at station SEATS, indicates that the modeled nitrate could faithfully demonstrate the seasonal variability in surface nutrients in the central SCS. Hence, it is appropriate to link the climatology of Chl-a, NPP and nitrate with the average monthly sinking biogenic flux.

9

3. Results 3.1 The long term flux record The time series of biogenic flux (CaCO3, opal and POC) during the period from April 1992 to April 1999 at SCS-C is shown in Fig. 2. The relative standard deviations of the time series fluxes, which are an index of the magnitude of temporal variation, were estimated to be approximately 37.0%, 36.0%, 57.2% and 36.6% for total biogenic matter, CaCO3, opal and POC fluxes, respectively. Overall, there were several important characteristics in the time series biogenic flux. The first and most notable is that there was a clear seasonal variation in flux but with substantial interannual variability in its timing and magnitude. For example, intermittent peaks of total biogenic flux generally occurred in SWM or NEM periods, while low values occurred in the rest of the year. More specifically, the total biogenic flux varied from 9.0 to 170.0 mg m−2 d−1, with an average of 77.7 ± 27.9 mg m−2 d−1 over 7 years. As the main components of biogenic flux, CaCO3 and opal fluxes were generally positively correlated to biogenic flux (r = 0.82, P < 0.01 and r = 0.87, P < 0.01, respectively). The CaCO3 flux ranged from 2.5 to 67.5 mg m−2 d−1, with an average of 32.2 ± 11.6 mg m−2 d−1, which accounted for 52.3% percent of the biogenic flux on average. Compared with the CaCO3 flux, the opal flux represented a smaller proportion of biogenic flux, and occasionally decoupled with CaCO3 (r = 0.43, P < 0.01). The opal average flux was 22.1 ± 12.6 mg m−2 d−1 over 7 years, with a range of 2.1 to 79.2 mg m−2 d−1. The trend of POC flux was almost in phase with the variability of CaCO3 and opal fluxes, varying from 0.8 to 9.7 mg m−2 d−1. According to the carrying coefficients of CaCO3 (0.327) and opal fluxes (0.649) based on a multiple linear 10

regression analysis (using stepwise method), siliceous plankton may be more important in the export of POC flux than calcareous plankton in the central SCS (See the auxiliary material). The second characteristic of the observed biogenic flux time series is the four distinct large total biogenic flux pulses measured as ≥150% of the averaged flux. These were interspersed among regular seasonal variations with comparable biogenic flux values of 117.6 mg m−2 d−1 in July 1992, 129.5 mg m−2 d−1 in September 1994, 104.9 mg m−2 d−1 in December 1994, 141.1 mg m−2 d−1 in January 1996. What calls for special attention is that the large flux pulse in September 1994 occurred during the AIM period, while others occurred during monsoon seasons. This interesting phenomenon indicates that not only the seasonally reversing monsoon but also intra-seasonal and interannual oceanographic processes, such as mesoscale eddies, may play important roles on the sinking biogenic flux.

Figure 2. Time series of monthly mean (a) total biogenic flux, (b) CaCO 3 flux, (c) opal flux, (d) particulate organic carbon (POC) flux and (e) lithogenic material flux during 1992–1999 at the SCS-C station. Yellow and gray shades indicate monsoon and inter-monsoon periods, respectively. On the x axis, "J" and "D" denote the months of June and December, respectively.

3.2 Climatological seasonal variability of biogenic flux The seasonal variability of biogenic flux was assessed by computing the average monthly fluxes from the weighted mean value of the different biogenic fluxes during the sampling period (Fig. 3.) The monthly means were divided into the SIM, SWM, AIM and NEM

11

periods. The average monthly total biogenic flux, opal and POC showed bimodal patterns, generally displaying peaks during the monsoon periods and low values during the inter-monsoon periods. The average monthly total biogenic flux, opal and POC fluxes exhibited the maxima (100.7, 34.9 and 5.2 mg m−2 d−1, respectively) in January during the NEM period, and the secondary peak in July during the SWM period. Compared with the maxima during the NEM period, the lowest values of total biogenic flux, opal and POC fluxes were 1.7, 2.4 and 1.7 times lower during the SIM period, respectively. In contrast, there was no distinctive seasonality observed in the averaged CaCO3 flux, exhibiting the maximum value (37.0 mg m−2 d−1) in July during the SWM period rather than in January during the NEM period like the opal and POC fluxes. Moreover, the lowest CaCO3 value was observed during the SIM period was 1.6 times lower than the maxima. The discrepancy in timing of peaks at the seasonal scale between the averaged CaCO3 and opal fluxes (ANOVA analysis, see the auxiliary material) suggested different temporal responses of calcareous and siliceous planktons to monsoon transition. According to the relative standard deviation of specific averaged monthly fluxes (15.8%, 15.1%, 26.0% and 15.5% for total biogenic matter, CaCO3, opal and POC fluxes, respectively), opal exhibited a greater seasonal variability than CaCO3 and POC fluxes. Additionally, the standard deviation of the averaged biogenic flux, CaCO3, POC and opal also showed similar temporal variation to the magnitude of particle flux at 10.5-29.6, 3.1-24.3, 0.7-2.6, 5.8-17.1 mg m−2 d−1, respectively. (Fig. 3). Similar to the particle flux trend, the standard deviation of the flux also exhibited two peaks during monsoon periods (with an average of 19.6, 12.2, 1.4 and 11.2 mg m−2 d−1 for averaged total biogenic particles, opal, POC, and 12

CaCO3 flux, respectively), reflecting the large inter-annual variability in the timing and duration of the monsoon periods of higher fluxes. In contrast, the standard deviation of the flux was at a low level during inter-monsoon periods (with an average of 12.6, 5.8, 0.9 and 9.5 mg m−2 d−1 for averaged total biogenic particles, opal, POC, and CaCO3 flux, respectively), indicating the uniformly low downward fluxes observed annually during inter-monsoon periods.

Figure 3. Averaged monthly fluxes of (a) biogenic flux, (b) CaCO 3 flux, (c) opal flux, (d) POC flux, (e) foraminifera flux, (f) particulate organic carbon (POC) to particulate inorganic carbon (PIC) mole ratio and, (g) opal/CaCO3 mole ratio during 1992–1999 at the SCS-C station. Error bars represent the standard deviations. Yellow and gray shades indicate monsoon and inter-monsoon periods, respectively.

3.3 Climatology of NPP, Chl-a and nitrate In this section, we assess the climatological seasonal variation of NPP, Chl-a and nitrate to explore the relationship between particle fluxes with upper ocean physical forcing and biogeochemical processes. As shown in Fig. 4, the climatological NPP presented a strong peak of 316.3 mg C m−2 d−1 in January during the NEM period and a noticeably weak peak in August during summer peak. In contrast, the minimum of NPP, 1.5 times lower than maxima, occurred in April during the SIM period. Similarly, the climatological surface Chl-a concentration ranged from 0.08 to 0.14 mg m−3, with maxima during the NEM period and the lowest value during the SIM period. Moreover, the climatological 13

surface nitrate concentration fluctuated from 0.02 to 0.15 µM, with a prominent winter peak during the NEM period and a secondary but barely noticeable peak from August to October. As a result, the climatological pattern of primary production in the euphotic zone agreed well with the most productive season in winter and the secondary productive season in summer, as observed in previous studies [Liu et al., 2002; Chen, 2005b].

14

Figure 4. Climatology of (a) net primary production (NPP), (b) Chl-a, (c) nitrate concentration, (d) mixed layer depth (MLD), (e) wind speed (WS) and (f) heat flux (HF) at the SCS-C station with a 1° × 1° grid. Among them, the Chl-a and nitrate concentrations are specific for surface layer.

4. Discussion 4.1 Factors controlling seasonal variation of biogenic flux It is widely recognized that nutrient availability in the euphotic zone fueling primary production is one of the key factors controlling temporal variations in deep ocean biogenic flux [Lampitt et al., 2010]. To eliminate the impact from non-seasonal factors, the monthly average POC, opal and CaCO3 fluxes (Fig. 3) were related to the climatological environmental data (WS, HF, MLD, nitrate, Chl-a and NPP, Fig. 4). The potential contribution of physical forcing, such as water mixing to nutrients supply, primary production and sinking biogenic fluxes, were considered on a seasonal basis in the central SCS. As expected, the climatological MLD was significantly correlated with WS (r = 0.6, P < 0.05) and HF (r = −0.6, P < 0.05). Their correlations suggest that vertical water mixing was regulated not only by wind stirring but also by surface cooling. Consequently, the surface nitrate concentration and NPP were significantly positively correlated with MLD (r = 0.91, P < 0.01; r = 0.89, P < 0.01), implying that nutrient availability in the upper layer was mainly controlled by MLD in the central SCS. This was consistent with Tseng et al. [2005], who indicated that the combined effect of surface cooling and wind-induced

15

mixing could effectively introduce nutrients from subsurface into the upper layer during the NEM period. Furthermore, significant positive correlations between NPP and POC (r = 0.59, P < 0.05), opal (r = 0.74, P < 0.01) and CaCO3 fluxes (r = 0.72, P < 0.01 indicated that primary production in the euphotic layer was responsible for the mesopelagic POC and opal fluxes. However, the MLD (25.6–39.7 m, Fig. 4e) during the SWM period is unable to exceed the top of the nutricline depth, which was shown to lack systematic seasonal variation, and remained around 50–70 m throughout the year at SEATS station in the northern SCS [Tseng et al., 2005]. Hence, limited nutrients could be introduced into the mixed layer by mixing from the subsurface to fuel the second maximum of the NPP (Fig. 4b) and mesopelagic biogenic fluxes (e.g., opal and POC, Fig. 3c and 3d) during the SWM period. The summer jet (Fig. 1a) induced by dual-pole off Vietnam with high biomass (Fig. 1c) could be a plausible candidate process transporting both nutrients and biogenic particles to the central SCS basin by horizontal advection [Chen and Wang, 2014]. More studies are needed to reveal the regulation of physical processes on sinking biogenic fluxes during the SWM in the central SCS. Table 1. Correlation coefficient (r) values from Pearson correlation analysis between monthly climatological mixed layer depth (MLD), surface nitrate, net primary production (NPP), total biogenic flux (Tot. Bio.), particulate organic carbon (POC), CaCO3 and opal fluxes at the SCS-C site in the central SCS.

MLD Nitrate NPP

MLD

Nitrate

NPP

1

0.91**

0.88

1

0.89

Tot. Bio.

**

0.88

**

0.71

1

0.80

Tot. Bio.

1

POC

** ** **

POC 0.82 0.61 0.59 0.90 1

CaCO3

** * * **

CaCO3 0.64 0.51 0.72 0.84 0.60 1

16

*

Opal 0.88 0.74

** ** *

0.74 0.93 0.91

** ** ** ** **

0.58

*

Opal

1

* P < 0.05 ** P < 0.01

The discrepancy between CaCO3 and opal and POC in their time of maximum monthly fluxes, in combination with the absence of any clear seasonality in CaCO3 flux, suggested that the calcareous phytoplankton growth due to nutrients input from subsurface was not the only factor controlling CaCO3 flux. CaCO3 originates from the shells of coccolithophores, foraminifera and pteropod in oceanic environments. However, the coccolithophores are more abundant during the NEM period than during the SWM period in the SCS [Chen, 2005b]. Hence, calcareous zooplankton might contribute to CaCO3 flux, except coccolithophores, during the early SWM period. The monthly average foraminifera flux of the same suite of trap samples (Fig. 3e) from May 1993 to June 1996 revealed that higher biomass of foraminifera appeared in warm season (March to August), with a maximum in July. The surface dwelling species of G. ruber (~30 m) and G. sacculifer (~50 m) dominated the foraminifera, accounting for more than 60% of total foraminifera fluxes [Chen et al., 2006]. Although foraminifera are grazers of phytoplankton, their POC/PIC mole ratio has been found to be low because of their heavy calcareous shells. [Loubere et al., 2007]. As expected, the climatological opal/CaCO3 and POC/PIC mole ratios were low in July and August (Fig. 3f and Fig. 3g), suggesting that the carbonate counter pump was enhanced due to foraminifera blooms during the SWM period, which could partly offset the positive effect of CO2 absorption by the biological organic carbon pump.

4.2 Control factors on the long-term record of biogenic flux 17

4.2.1 Decoupling of large biogenic flux pulses with water mixing Our results suggest that an increase in the monthly average sinking biogenic flux occurred when the MLD deepened, especially during the winter NEM period. However, prominent interannual variability in biogenic fluxes (POC, opal and CaCO3) was demonstrated by the large relative standard deviations (Fig. 2). Moreover, time series data of MLD showed significant interannual variability (relative standard deviation of 41.1%), which was correlated with WS (r = 0.47, P < 0.01) and HF (r = −0.46, P < 0.01) during the study period (Fig. 5). The weak correlations between MLD and opal (r = 0.37, P < 0.01) or POC fluxes (r = 0.23, P < 0.05) implied that the fluctuation of the MLD mediated the production of siliceous plankton and organic matter export over the long-term. However, CaCO3 flux was independent of variability in MLD (P > 0.05). To more directly assess the relationship between long-term biogenic flux fluctuation and water mixing in the upper ocean, we calculated the anomaly of biogenic fluxes (CaCO3, opal and POC) according to the difference between the monthly biogenic fluxes and the climatological mean for the particular month in the year. As shown in Fig. 6, the pulses of CaCO3, opal and POC fluxes at random occurred during the study period and lacked the same pattern. According to correlation analysis, the anomaly of POC flux was primarily influenced by the opal flux (r = 0.77, P < 0.01), followed by CaCO3 flux (r = 0.61, P < 0.01). However, two exceptions occurred in June and August 1997, where the POC peaks were 1.6 and 1.5 times higher than the monthly average values. Unlike the other POC peaks that corresponded to higher opal flux, these peaks corresponded with an elevation of CaCO3 flux but without any clear peak in opal flux. This may be related to the 1997–1998 El Niño 18

events. Counter-intuitively, the biogenic flux anomaly (CaCO3, opal and POC) was uncoupled with WSA, HFA and MLDA (Fig. 7), indicating that these extra high sinking biogenic fluxes could not be driven by the vertical water mixing. Instead, those may be related to other physical processes and could trigger high productivity and subsequent high sinking biogenic particle fluxes, especially for opal and POC flux.

Figure 5. Time series of monthly (a) wind speed (WS), (b) net heat flux (HF) and (c) mixed layer depth (MLD) during 1992-1999 at the SCS-C station. Yellow and gray shades indicate monsoon and inter-monsoon periods, respectively.

Figure 6. Time series of monthly flux anomaly of (a) total biogenic flux, (b) CaCO 3 flux, (c) opal flux and (d) POC flux anomaly during 1992–1999 at the SCS-C station. The flux anomalies were calculated according to the difference between the monthly biogenic fluxes and the climatological mean for the particular month in the year.

Figure 7. Time series of monthly wind speed (WSA), heat flux anomaly (HFA), the mixed layer depth (MLD) and sea surface height anomaly (SSHA) during 1992–1999 at SCS-C station. The anomaly for each variable was computed as the difference between the monthly value and the climatological mean for the particular month in the year.

4.2.2 Role of mesoscale eddies on positive biogenic flux pulses The positive effect on the biological production of cyclonic eddies (cold core) is resulted 19

by pumping nutrients from subsurface [Sweeney et al., 2003]. And the "new" nutrients induced by cyclonic eddies generally trigger diatom biomass increase [Wang et al., 2016], which may enhance the downward export of biogenic particles in the ocean sometimes. Hence, we attempted to relate the potential role of cyclonic eddies on the large particle flux pulses in the central SCS. Mesoscale eddies in the SCS are discernible from satellite-derived SSHA data [Chelton and Schlax, 2003]. Four distinct cyclonic eddies (SSHA < −15 cm) were observed from 1992 to 1999 in the study area based on the time series SSHA data at station SCS-C (Fig. 7) and spatial maps of SSHA in the SCS. The characteristics of these eddies (e.g. age, magnitude and radius) based on daily SSHA data were shown in Table 2. Interestingly, these four cyclonic eddies were accompanied by the positive anomaly of opal and POC fluxes (i.e. large flux pulses peaked in July 1992, November 1992, December 1994 and January 1996), which was inversely related to the negative SSHA. The first cyclonic eddy generated at the beginning of June 1992 peaked on 20th July (SSHA < −20 cm, Fig. 8a) and weakened at the end of July. Correspondingly, the opal flux elevated 1.4 and 2.4 times in June and July compare with climatological seasonal value. In November 1992, the second cyclonic eddy (SSHA < −25 cm, Fig. 8b) occurred at the station SCS-C. This cold eddy gradually weakened with northward movement over time, and the trap was located in the periphery of the eddy (SSHA > −10 cm) in December. As a consequence, the opal flux increased by 1.8 times during November. Similarly, the third and fourth cyclonic eddies occurred during winter from November to January of next year in 1994 and 1995, with comparable intensity (SSHA < −20 cm, Fig. 8d, e). Again, opal 20

fluxes increased 2.4 and 3.6 times their climatological monthly fluxes in December 1994 and January 1996. Simultaneously, POC flux increased 2.4, 1.3, 1.2 and 2.0 times due to the above four cyclonic eddies in chronological sequence, respectively. In contrast, there was a weak anticyclonic eddy (SSHA = 10 cm) in the trap area from August to October 1994, when 2.3 and 2.2 times higher opal and POC fluxes were documented in September than their climatological monthly fluxes. The anticyclonic eddy was not responsible for this flux pulse, which resulted in downwelling and oligotrophy. However, the anticyclonic eddy was surrounded by three cyclonic eddies. On the southwestern side, an extremely strong cyclonic eddy (SSHA < −30 cm, Fig. 8c) was located 327 km away from the trap, while other relatively weak two eddies (SSHA = −10 cm) were on the northwestern and northeastern sides (Fig. 8). The divergent current associated with the cyclonic eddy, combined with the convergent effect caused by the anticyclonic eddy above the trap was thought to be responsible for the enhanced biogenic flux by advection nutrients or particles from the adjoining cold eddy to the study area. As expected, the positive opal and POC flux anomalies occurred during the mature and decaying stage of these cyclonic eddies. This phenomenon is consistent with the conceptual model of the biological effect of a cyclonic eddy proposed by Sweeney et al. (2003).

Table 2. The life cycle, radius, minimum of SSHA and approximate stage and duration of large flux anomaly of the cyclonic eddies. Time

Life cycle

Approximate

Minimum of

21

Stage

Approximate duration

Radius

SSHA

of large flux anomaly

(months)

(km)

(cm)

(months)

1992.07

/

200

-20

/

/

1992.11

3

110

-28

Mature-Decaying

1-2

1994.09

2.5

220

-30

Mature-Decaying

1-2

1994.12

2.5

Cross basin*

-24

Mature

1.5-2.5

1995.12

3.5

Cross basin

-21.5

Mature-Decaying

1-3

Figure 8. Spatial maps of sea surface height anomaly (SSHA, m) overlaid with geostrophic currents during the high flux periods. Filled red triangles represent the location of the SCS-C station.

It is noteworthy that three of the five cyclonic eddies occurred during the NEM period with high primary production in the euphotic layer [Liu et al., 2002] and a subsequent high sinking biogenic flux and mesopelagic depth occurred in this study. The 18 cyclonic eddies (SSHA < −14 cm) were observed during the NEM period, and 6 in the SWM period from 1993 to 2000, while the rest 4 in inter-monsoon periods [Wang et al., 2003]. Consequently, about 64% of the cyclonic eddies appeared in the winter monsoon period in the entire SCS basin based on statistical analyses. Wang et al. [2003] and Lin et al. [2015] argued that the cyclonic eddies were preferentially generated in the monsoon season, particularly during the NEM period, because of the stronger wind stress curl and Kuroshio intrusion in winter. Hence, as pointed in this study, cyclonic eddies may further elevate 22

primary production and subsequent sinking biogenic fluxes by replenishing nutrients from the deep water, ultimately intensifying the biological carbon pump during monsoon seasons, especially during the productive NEM period in winter.

The potential contribution of cyclonic eddies to the sinking biogenic fluxes (POC and opal) to the mid-depth was estimated at the SCS-C station. As shown in Table 3, 44% to 59% of opal flux collected in the mesopelagic depth (~1200 m) resulted from diatom biomass induced by cyclonic eddy events in the study area. Unlike opal flux, the contribution of cyclonic eddies to the POC flux varied in a wide range from 17% to 54%. This might be attributed to the attenuation of organic matter caused by enhanced microbial activity [Jiao et al., 2014]. Also, some siliceous plankton (e.g. radiolarian/phaeodarian silica) which probably have higher Si:C ratio than diatom may also contribute to the large biogenic silica export. Therefore, the combined effect of mesoscale cyclonic eddy activities and vertical water mixing might make the SCS a more efficient biological pump in the monsoon seasons sometimes. A schematic model was presented in Fig. 9 to illustrate biogenic fluxes (opal and POC) export in the central SCS basin during cyclonic eddy period and non-cyclonic eddy conditions in winter time. In cyclonic eddy conditions, opal and POC fluxes were 57.6 mg m−2 d−1 and 6.2 mg m−2 d−1, respectively. In contrast, in non-cyclonic eddy conditions, opal and POC fluxes were 26.8 mg m−2 d−1 and 4.5 mg m−2 d−1, respectively. This indicates that silica and carbon export could be enhanced in the presence of the cyclonic eddy during the NEM period in the central SCS. Our findings were verified by the result from biogeochemical modeling in the SCS, which showed that the POC flux at the base of the mixed layer when cyclonic eddies were present was 41% 23

higher than those in the basin, whereas it was 31% lower in the presence of anti-cyclonic eddies [Xiu and Chai, 2011]. However, only 6.3% POC flux was contributed by cyclonic eddies at annual scale on average. This means that the impact of cyclonic eddy is occasional and regional, while monsoonal seasonality is the first order controlling factor on sinking biogenic flux in the SCS.

Figure 9. Schematic diagram depicting opal and POC fluxes (mg m−2 d−1) to the deeper ocean (~1200 m) in the central SCS during (a) non-cyclonic eddy condition and (b) cyclonic eddy period.

Table 3. Opal and POC flux during cyclonic eddy periods (July 1992, November 1992, September 1994, December 1994 and January 1996), and the contribution of eddy-associated flux in mg m−2 d−1 and %. The eddy-associated flux was calculated from the specific flux minus the climatological average monthly flux, and the percentage represented the portion of eddy-associated flux in the specific total flux. Specific Total Flux −2

Time

Eddy-Associated Flux

−1

−2

(mg m d )

−1

(mg m d )

Eddy-Associated Flux in percentage (%)

opal

POC

opal

POC

opal

POC

Jul-92

54.1

9.7

31.6

5.3

58.4

54.3

Nov-92

40.3

5.3

17.9

1.2

44.4

23.5

Sep-94

51.0

9.2

29.9

4.8

58.6

52.0

Dec-94

53.3

5.1

26.5

0.9

49.8

16.9

Jan-96

79.2

8.2

44.3

3.1

56.0

37.8

5. Conclusion The sinking biogenic particle flux collected by a sediment trap deployed in the central SCS provided insight into the response of the biological carbon pump to multi-scale physical processes. The climatological monthly bulk biogenic fluxes (opal and POC fluxes)

24

were well correlated to physical forcing factors (East Asian Monsoon and surface cooling). Bimodal patterns of opal and POC fluxes were observed, with the peak occurring during the NEM period and secondary peak during the SWM period. Significant correlations between POC fluxes and HF, WS, MLD, nitrate, Chl-a, NPP, opal and suggested that the diapycnal nutrient input from the subsurface, induced by wind stirring and surface cooling, was responsible for an increase in NPP in the euphotic layer and subsequent deep-ocean opal and POC fluxes. The climatological monthly CaCO3 flux had less seasonality and was not correlated with nitrate (P > 0.05), which may have been linked to an extra source of CaCO3 from foraminifera, besides coccolithophores. Few nutrients were replenished into the surface layer to sustain NPP and opal and POC fluxes since the MLD was shallower than the nitracline during the SWM period. More attention should be paid to physical processes that could introduce nutrients or biogenic material to the central SCS basin, e.g., summer jet off Vietnam during the SWM period. At long time scale, mesoscale cyclonic eddies represented one of the other factors regulating sinking biogenic flux in addition to the vertical mixing in the upper ocean. The opposite relationship between sinking particle pulses and SSHA confirmed the role of cyclonic eddies in mediating the biological carbon pump. During cyclonic eddy events， opal and POC fluxes driven by siliceous planktons could account for 53% and 37% on average, respectively. Since cyclonic eddies mainly occurred during the NEM period, we anticipate that sinking biogenic flux and carbon sequestration were further intensified in winter because of the combined effect of water mixing in the mixed layer and occasional cyclonic eddy events. 25

Acknowledgements This study was financially supported by the State Key R&D project of China (2016YFA0601100) and the National Natural Science Foundation of China (91128212 and 91528304), the Special Fund for Basic scientific research of the Second Institute of Oceanography, State Oceanic Administration (JT1501, JG1514 and JT1401), the National Programme on Global Change and Air-Sea Interaction (GASI-03-01-03-03) and the German Ministry of Education and Research (03G0095, 03G0114, 03G0513, 03G0532, 03F0604, 03F0727, 03F0673). This paper is a contribution to the Sino-German cooperation project “Impact of the ENSO-Monsoon System on the Biogeochemical Fluxes in the Northern South China Sea”. We would like to thank Jiliang Xuan, Ruibin Ding, Denglingtong Lu and Ronghua Chen help with the ancillary data collection and the crew of R/V Xianyanghong-5, R/V Xianyanghong-14 and R/V Sonne for their assistance in the deployment and recovery of the sediment trap moorings. We also wish to express our thanks to Kuanbo Zhou, Jie Xu and Ruiz-Pino Diana for their comments and suggestions.

26

Reference Archer, D., and E. Maier-Reimer (1994), Effect of deep-sea sedimentary calcite preservation on atmospheric CO2 concentration, Nature, 36, 260-263, doi:10.1038/367260a0. Chen, J. F. (2005a), Biogeochemistry of settling particles in the South China Sea and its significance for paleo-environment studies, P.H.D. thesis, School of Ocean and Earth Science, Tongji University, Shanghai, P. R. China. Chen, Y. L. L. (2005b), Spatial and seasonal variations of nitrate-based new production and primary production

in

the

South

China

Sea,

Deep

Sea

Res.,

Part

I,

52(2),

319-340,

doi:10.1016/j.dsr.2004.11.001. Chen, R. H., Z. M. Jian, Y. L. Zheng, and J. F. Chen (2006), Seasonal variations of the planktonic foraminifera flux in the central South China Sea, Journal of Tongji University, 28(1), 73-77, doi:0253374X( 2000) 01- 0073- 05. Chen, L. Y. L., H. Y. Chen, and C. W. Chung (2007), Seasonal variability of coccolithophore abundance and assemblage in the northern South China Sea, Deep Sea Res., Part II, 54(14), 1617-1633, doi:10.1016/j.dsr2.2007.05.005. Chen, C., and G. Wang (2014), Interannual variability of the eastward current in the western South China Sea associated with the summer Asian monsoon, J. Geophys. Res. Oceans, 119(9), 5745-5754. Chelton, D. B., and M. G. Schlax (2003), The accuracies of smoothed sea surface height fields constructed from tandem satellite altimeter datasets, J. Atmos. Oceanic Technol., 20(9), 1276-1302. Chung, Y., H. C. Chang, and G. W. Hung (2004), Particulate flux and 210Pb determined on the sediment trap and core samples from the northern South China Sea, Cont. Shelf Res., 24(6), 673-691, doi:10.1016/j.csr.2004.01.003. Eppley, R. W., and B. J. Peterson (1979), Particulate organic matter flux and planktonic new production in the deep ocean, Nature, 282, 677-680, doi:10.1038/282677a0. Gaye, B., M. G. Wiesner, and N. Lahajnar (2009), Nitrogen sources in the South China Sea, as discerned from stable nitrogen isotopic ratios in rivers, sink particles, and sediments, Mar. Chem., 114(3), 72-85, doi:10.1016/j.marchem.2009.04.003. Honjo, S., J. Dymond, W. Prell, and V. Ittekkot (1999), Monsoon-controlled export fluxes to the interior of the Arabian Sea, Deep Sea Res., Part II, 46(8), 1859-1902, doi:10.1016/S0967-0645(99)00047-8. Hung, C. C., and G. C. Gong (2010), POC/234Th ratios in particles collected in sediment traps in the northern South China Sea, Estuarine, Coastal and Shelf Science, 88, 303-310, doi:10.1016/j.ecss.2010.04.008. Ishii, M., and M. Kimoto (2009), Reevaluation of historical ocean heat content variations with time-varying XBT and MBT depth bias corrections, J. Oceanogr., 65(3), 287-299. Jiao, N., Y. Zhang, K. Zhou, Q. Li, M. Dai, J. Liu, J. Guo, and B. Huang (2014), Revisiting the CO2 “source” problem in upwelling areas–a comparative study on eddy upwellings in the South China Sea, Biogeosciences,11(9), 2465-2475. Lahajnar, N., M. G. Wiesner, and B. Gaye (2007), Fluxes of amino acids and hexosamines to the deep South China Sea, Deep Sea Res., Part I, 54(12), 2120-2144, doi:10.1016/j.dsr.2007.08.009. 27

Lampitt, R. S., I. Salter, B. A. de Cuevas, S. Hartman, K. E. Larkin, and C. A. Pebody (2010), Long-term variability of downward particle flux in the deep Northeast Atlantic: causes and trends, Deep Sea Res., Part II, 57(15), 1346-1361, doi:10.1016/j.dsr2.2010.01.011. Lin, H. L., and H. Y. Hsieh (2007), Seasonal variations of modern planktonic foraminifera in the South China Sea, Deep Sea Res., Part II, 54(14), 1634-1644, doi:10.1016/j.dsr2.2007.05.007. Lin, I., W. T. Liu, C. C. Wu, G. T. Wong, C.Hu, Z. Chen, and K. K. Liu (2003), New evidence for enhanced ocean

primary

production

triggered

by

tropical

cyclone,

Geophys.

Res.

Lett.,

30(13),

doi:10.1029/2003GL017141. Lin, X., C. Dong, D. Chen, Y. Liu, J. Yang, B. Zou, and Y. Guan (2015), Three-dimensional properties of mesoscale eddies in the South China Sea based on eddy-resolving model output, Deep Sea Res., Part I, 99, 46-64. Liu, J., P. D. Clift, W. Yan, Z. Chen, H. Chen, R. Xiang, and D. Wang (2014), Modern transport and deposition of settling particles in the northern South China Sea: Sediment trap evidence adjacent to Xisha Trough, Deep Sea Res., Part I, 93, 145-155, doi:10.1016/j.dsr.2014.08.005. Liu, K. K., S. Y. Chao, P. T. Shaw, G. C. Gong, C. C. Chen, and T. Y. Tang (2002), Monsoon-forced chlorophyll distribution and primary production in the South China Sea: observations and a numerical study, Deep Sea Res., Part I, 49(8), 1387-1412, doi:10.1016/S0967-0637(02)00035-3. Liu, K. K., L. W. Wang, M. Dai, C. M. Tseng, Y. Yang, C. H. Sui, and S. M. Huang (2013), Inter-annual variation of chlorophyll in the northern South China Sea observed at the SEATS Station and its asymmetric

responses

to

climate

oscillation,

Biogeosciences,

10(11),

7449-7462,

doi:10.5194/bg-10-7449-2013. Loubere, P., S. A. Siedlecki, and L. I. Bradtmiller (2007), Organic carbon and carbonate fluxes: Links to climate change, Deep Sea Res., Part II, 54(5), 437-446, doi:10.1016/j.dsr2.2007.02.001. Martin, J. H., G. A. Knauer, D. M. Karl, and W. W. Broenkow (1987), VERTEX: carbon cycling in the northeast Pacific, Deep Sea Res., Part I, 34(2), 267-285, doi:10.1016/0198-0149(87)90086-0. McGillicuddy, D. J., L. A. Anderson, N. R. Bates, T. Bibby, K. O. Buesseler, C. A. Carlson, C. S. Davis, C. Ewart, P. G. Falkowski, S. A. Goldthwait, D. A. Hansell, W. J. Jenkins, R. Johnson, V. K. Kosnyrev, J. R. Ledwell, Q. P. Li, and D. A. Siegel (2007), Eddy/wind interactions stimulate extraordinary mid-ocean plankton blooms, Science, 316(5827), 1021-1026. Monterey, G. I., and S. Levitus(1997), Climatological cycle of mixed layer depth in the world ocean, US Government Printing Office, NOAA NESDIS, 5. Mortlock, R. A., and P. N. Froelich (1989), A simple method for the rapid determination of biogenic opal in pelagic

marine

sediments,

Deep

Sea

Res.,

Part

I,

36(9),

1415-1426,

doi:10.1016/0198-0149(89)90092-7. Müller, P. J., E. Suess and C. A. Ungerer (1986), Amino acids and amino sugars of surface particulate and sediment trap material from waters of the Scotia Sea, Deep Sea Res. 33, 819–838. Ning, X. R., F. Chai, H. Xue, Y. Cai, Liu, C. and J. Shi (2004), Physical-biological oceanographic coupling influencing phytoplankton and primary production in the South China Sea, J. Geophys. Res., 109(C10), doi:10.1029/2004JC002365. 28

Pan, X., G. T. Wong, F. K. Shiah, and T. Y. Ho (2012), Enhancement of biological productivity by internal waves: observations in the summertime in the northern South China Sea, J. oceanogr., 68(3), 427-437, doi:10.1007/s10872-012-0107-y. Sigman, D. M., and E. A. Boyle (2000), Glacial/interglacial variations in atmospheric carbon dioxide, Nature, 407, 859-869, doi:10.1038/35038000. Su, J. L. (2004), Overview of the South China Sea circulation and its influence on the coastal physical oceanography

outside

the

Pearl

River

Estuary,

Cont.

Shelf

Res.,

24(16),

1745-1760,

doi:10.1016/j.csr.2004.06.005. Sweeney, E. N., D. J. McGillicuddy, and K. O. Buesseler (2003), Biogeochemical impacts due to mesoscale eddy activity in the Sargasso Sea as measured at the Bermuda Atlantic Time Series Study (BATS), Deep Sea Res., Part II, 50 (22–26), 3017–3039. Tang, D. L., I. H. Ni, D. R. Kester, and F. E. Muller-Karger (1999), Remote sensing observations of winter phytoplankton blooms southwest of the Luzon strait in the South China Sea, Mar. Ecol. Prog. Ser., 191, 43. Tseng, C. M., G. T. Wong, I. I. Lin, C. R. Wu, and K. K. Liu (2005), A unique seasonal pattern in phytoplankton biomass in low-latitude waters in the South China Sea, Geophys. Res. Lett., 32(8), doi:10.1029/2004GL022111. Tseng, C. M., G. C. Gong, L. W. Wang, K. K. Liu, and Y. Yang (2009), Anomalous biogeochemical conditions in the northern South China Sea during the El-Niño events between 1997 and 2003, Geophys. Res. Lett., 36, L14611, doi: 10.1029/2009GL038252. Volk, T., and M. I. Hoffert (1985), Ocean carbon pumps: Analysis of relative strengths and efficiencies in ocean-driven atmospheric CO2 changes, in The Carbon Cycle and AtmosphericCO2: Natural Variations Archean to Present, Geophys. Monogr. Ser., vol. 32, edited by E. T. Sundquist and W. S. Broecker, pp. 99-110, AGU, Washington, D. C. Wan, S., Z. M. Jian, X. R. Cheng, P. J. Qiao, and R. J. Wang (2010), Seasonal variations in planktonic foraminiferal flux and the chemical properties of their shells in the southern South China Sea, Sci. China Ser. D, 53(8), 1176-1187, doi:10.1007/s11430-010-4039-3. Wang, S. H., N. C. Hsu, S. C. Tsay, N. H. Lin, A. M. Sayer, S. J. Huang, and W. K. Lau (2012), Can Asian dust trigger phytoplankton blooms in the oligotrophic northern South China Sea? Geophys. Res. Lett., 39(5), doi:10.1029/2011GL050415. Wang R. J., J. Lin, L. F. Zheng, R. H. Chen, and J. F. Chen (2000), Siliceous microplankton fluxes and seasonal variations in the central South China Sea during 1993-1995: monsoon climate and El Niño responses, Chinese Sci. Bull., 45(23), 2168-2172, doi:10.1007/BF02886323. Wang, G. H., J. L. Su, and P. C. Chu (2003), Mesoscale eddies in the South China Sea observed with altimeter data, Geophys. Res. Lett., 30 (21), 2121, doi:10.1029/2003GL018532. Wang, L., B. Huang, K. P. Chiang, X. Liu, B. Chen, Y. Xie, Y. Xu, J. Hu, and M. Dai (2016), Physical-biological coupling in the western South China Sea: The response of phytoplankton community to a mesoscale cyclonic eddy, PloS one, 11(4), e0153735. Wiesner, M. G., L. F. Zheng, H. K. Wong, Y. Wang, and W. Chen, (1996), Fluxes of particulate matter in the 29

South China Sea, in Particle Flux in the Ocean, edited by V. Ittekkot, P. Schäfer, S. Honjo and P. J. Depetris, pp. 293-312, John Wiley & Sons, New York. Wong, G. T., C. M. Tseng, L. S. Wen, and S. W. Chung (2007), Nutrient dynamics and N-anomaly at the SEATS station, Deep Sea Res., Part II, 54(14), 1528-1545, doi:10.1016/j.dsr2.2007.05.011. Xiu, P., F. Chai, L. Shi, H. J. Xue, and Y. Chao (2010), A census of eddy activities in the South China Sea during 1993–2007. J. Geophys. Res. 115, C03012. http://dx.doi.org/10.1029/2009jc005657. Xiu, P., and F Chai (2011), Modeled biogeochemical responses to mesoscale eddies in the South China Sea. J. Geophys. Res. 116, C10006. http://dx.doi.org/10.1029/2010jc006800. Zhou, K., M. Dai, S. Kao, L. Wang, P. Xiu, F. Chai, and Y. Liu (2013), Apparent enhancement of 234Th-based particle export associated with anticyclonic eddies, Earth Planet. Sci. Lett., 381, 198-209, doi:10.1016/j.epsl.2013.07.039.

Highlights: 1. Long-term time series sinking biogenic flux to the mesopelagic zone of the Central South China Sea is reported.

2. Monsoon and surface cooling controlled the seasonality of sinking biogenic flux. 3. Cyclonic eddy associated sinking organic matter was quantified.

30