Deep-Sea Research Part II 167 (2019) 34–45
Contents lists available at ScienceDirect
Deep-Sea Research Part II journal homepage: www.elsevier.com/locate/dsr2
Methane distributions and sea-to-air ﬂuxes in the Pearl River Estuary and the northern South China sea
Wangwang Yea,b, Guiling Zhanga,b,∗, Wenjing Zhenga, Honghai Zhanga, Ying Wuc a
Key Laboratory of Marine Chemistry Theory and Technology, Ministry of Education, Ocean University of China, Qingdao, 266100, PR China Laboratory for Marine Ecology and Environmental Science, Qingdao National Laboratory for Marine Science and Technology, Qingdao, 266071, China c State Key Laboratory of Estuarine and Coastal Research, East China Normal University, 200062 Shanghai, PR China b
A R T I C LE I N FO
A B S T R A C T
Keywords: Methane Air-sea exchange South China sea Estuary Shelf
We measured dissolved methane (CH4) at various depths in the Pearl River Estuary and the shelf, slope, and basin of the northern South China Sea (SCS) during cruises from May to July 2014, October 2014, June 2015, and March 2017. High CH4 concentrations (9.4–430.3 nM) were observed in the Pearl River Estuary. There was a negative correlation between the CH4 concentration and the salinity in the surface waters of the Pearl River Estuary, indicating rapid CH4 loss via air-sea exchange and aerobic CH4 oxidation. The CH4 concentrations in the surface waters ranged from 1.7 to 5.4 nM in the shelf and slope region and from 2.2 to 3.0 nM in the basin of the northern SCS. The injection of a Pearl River plume rich in CH4 and driven by eddies had a profound impact on the CH4 distribution in the shelf and slope region in June 2015. The vertical proﬁles of CH4 showed spatial variations that can be attributed to complex interactions among physical, chemical, and biological processes. The subsurface CH4 maxima occurred ubiquitously in the euphotic zones, probably due to in situ CH4 production in anoxic microniches and aerobic CH4 formation from potential substrate precursors such as dimethylsulfoniopropionate (DMSP). The gross CH4 production rate in the mixed layer was estimated at 0.11 nM d-1 in the summer and -0.19 nM d-1 in the autumn. The surface waters of the Pearl River Estuary were oversaturated in CH4 with respect to atmospheric equilibrium having a saturation level of 5066 ± 5908% (mean ± SD) in early autumn and 6166 ± 3611% in early summer. The CH4 ﬂuxes (estimated using the W14 relationship) were estimated at 314.3 ± 464.9 μmol m-2 d-1 in early autumn and 184.2 ± 187.5 μmol m-2 d-1 in early summer. In the shelf and slope areas, the CH4 saturation was 178 ± 37% in early autumn, 181 ± 60% in early summer, and 129 ± 14% in winter. The sea-to-air CH4 ﬂuxes were 8.0 ± 4.3 μmol m-2 d-1 in autumn, 4.1 ± 5.2 μmol m-2 d-1 in summer, and 1.1 ± 0.8 μmol m-2 d-1 in winter. The CH4 saturation ratio was less variable in the deep basin, ranging from 108% to 184% with an average of 140 ± 16%. The CH4 ﬂux in the deep basin area was 1.9 ± 1.2 μmol m-2 d-1. Overall, we estimate the annual CH4 emission from the SCS as 9.9 × 109 mol yr-1. Hence the SCS is a net source of atmospheric CH4.
1. Introduction Methane (CH4) is a potent greenhouse gas and accounts for about 20% of greenhouse gas-mediated global warming (IPCC, 2013). Although the atmospheric CH4 budget is constrained reasonably well, the non-anthropogenic CH4 emissions are still poorly identiﬁed and have large uncertainties (Dlugokencky et al., 2011). The oceans are considered a non-anthropogenic source of atmospheric CH4. They exhibit great temporal and spatial variability regarding CH4 emissions. For example, the estuaries (2.6 Tg yr-1) are more intense sources of atmospheric CH4 than the open ocean (0.4-1.8 Tg yr-
1 ), although the surface area of estuaries only accounts for a small part of global ocean area (Rhee et al., 2009; Borges and Abril, 2011). However, the continental shelves (∼13.0 Tg yr-1) are the most intense source compared to the estuaries and open ocean (Bange et al., 1994; Liss and Johnson, 2014). Estuarine water contains hundreds of nanomoles of CH4 and is an important source of atmospheric CH4 due to the fast ventilation of the shallow water column, particularly during the mixing with saline water (Borges et al., 2016; Schmale et al., 2005, 2018; Zhang et al., 2008a). The inner regions of continental shelves receive large amounts of terrestrial organic matter and nutrients from rivers, creating conditions favorable for methanogenesis in the water
∗ Corresponding author. Key Laboratory of Marine Chemistry Theory and Technology, Ministry of Education, Ocean University of China, Qingdao, 266100, PR China. E-mail address: [email protected]
https://doi.org/10.1016/j.dsr2.2019.06.016 Received 25 July 2018; Received in revised form 15 May 2019; Accepted 23 June 2019 Available online 25 June 2019 0967-0645/ © 2019 Elsevier Ltd. All rights reserved.
Deep-Sea Research Part II 167 (2019) 34–45
W. Ye, et al.
column as well as in the sediments (Zhang et al., 2008a; PfeiﬀerHerbert et al., 2015; Borges et al., 2018b). Thus, the shelf regions are of great importance for oceanic CH4 emissions due to having high CH4 production rates and a fast ventilation of the well-mixed water column (Bange et al., 1994; Borges et al., 2018b; Schmale et al., 2018). However, these estimates have large uncertainties due to the limited spatial and temporal coverage of the measurements, as well as the diﬀerent parameterizations of gas transfer velocity (Borges and Abril, 2011). In the open ocean, far from land, the dissolved CH4 at the surface is less variable. Nevertheless, the observed supersaturations of CH4 in the surface water cannot be attributed to the high-CH4 coastal waters (Reeburgh, 2007). This phenomenon of CH4 accumulation in the open ocean is commonly termed the “oceanic CH4 paradox”, which is still controversially discussed (de Angelis and Lee, 1994; Karl and Tilbrook, 1994; Karl et al., 2008; Damm et al., 2010; Grossart et al., 2011; Repeta et al., 2016). Recent studies found that CH4 can be formed by the decomposition of methyl-rich organic phosphorus or sulfur compounds (i.e. methylphosphonate (MPn) and dimethylsulfoniopropionate (DMSP)) in aerobic surface waters under speciﬁc conditions, such as nutrient limitation (Damm et al., 2008, 2010; Karl et al., 2008; Repeta et al., 2016). The South China Sea (SCS) covers an area of 3.5 × 106 km2, has an average depth of 1200 m, and is the largest marginal sea in the world. It is a semi-closed region that has connections to the northwestern Paciﬁc Ocean through the Luzon Strait and to the southeastern Indian Ocean through the Strait of Malacca. The Pearl River (Zhujiang), which has the second largest discharge in China, enters the SCS and provides high loads of nutrients, sediments, and CH4 (Chen et al., 2008). The Pearl River Estuary (deﬁned as the four eastern-most tributaries that discharge their waters into the Lingdingyang) (Su, 2004) has an area of over 2000 km2 and is quite shallow with a water depth between 2 and 10 m. Several previous studies reported the distributions and sea-to-air ﬂuxes of CH4 in the Pearl River Estuary and the northern SCS (Chen et al., 2008; Zhou et al., 2009; Tseng et al., 2017). However, data on the CH4 dynamics from the Pearl River Estuary to the deep basin of the SCS are still rather limited and little is known about seasonal variations and the vertical distribution of CH4 and its controlling factors. The objective of this study is to examine the spatial distribution of CH4 from the Pearl River Estuary to the northern SCS, identify possible sources and factors that aﬀect the distribution of dissolved CH4, and estimate the overall CH4 emissions from the SCS into the atmosphere. The results of this study will improve our understanding of the contribution of this region to global oceanic CH4 emissions.
An open research cruise (NORC2014-05) was conducted onboard the RV “Dong Fang Hong 2” from 20 May to 17 July 2014 and was supported by the ship-time sharing project of National Natural Science Foundation of China (NSFC). Seawater samples from various depths were collected using 12-L Niskin bottles mounted on a Sea-Bird 911 plus CTD rosette at 5 stations (D8, G3, G4, H5, H7) in the deep basin with water depths in the range of 3800–5000 m and station P4 in the West Philippine Sea, which is east of the Luzon Strait. Surface waters were collected at the deep basin west of the Luzon Strait to improve the spatial resolution of the air-sea ﬂuxes of CH4. Discrete surface water samples for chlorophyll-a (chl-a) analyses were collected and measured ﬂuorimetrically following the method of Holm-Hansen et al. (1965) during October 2014 and June 2015. Total suspended matter (TSM) and particulate organic carbon (POC) were measured as described by Wu et al. (2013) during October 2014 and June 2015. The water samples for determining the dimethylsulﬁde (DMS), DMSP, and dimethylsulfoxide (DMSO) concentrations were collected during June 2014. DMS was analyzed using the cryogenic purge-and-trap technique (Yang et al., 2008). DMSP was analyzed by alkaline cleavage to DMS at a 1:1 ratio (Yang et al., 2008). Dimethylsulfoxide (DMSO) was converted into DMS by adding cobaltdosed sodium borohydride (NaBH4) as described by Yang et al. (2014) and was analyzed immediately using the same technique. The seawater samples for the CH4 analysis were transferred from the Niskin bottles to 100-mL glass bottles by an overﬂow of about 1.5–2 times the bottle volume without bubbles. Then, 1 mL of supersaturated HgCl2 was added to inhibit microbial activity and the containers were sealed with butyl rubber stoppers and aluminum caps to prevent gas exchange. All bottles were stored in the dark and analyzed within 60 days at the laboratory. The concentrations of dissolved CH4 were determined by a gas-stripping method using a Shimadzu GC-14B gas chromatograph that was equipped with a ﬂame ionization detector (FID) (Zhang et al., 2004). In brief, after purging with high purity N2, the samples were passed through a drying tube with calcium chloride to remove the water vapor. Then, CH4 was separated on a 3 m × 3 mm i.d. stainless steel column packed with 80/100 mesh Porapak Q and measured with an FID (Zhang et al., 2004). The FID responses were calibrated every 3 h by using known volumes of CH4 standards (2.05, 4.22, and 50.4 ppmV) from the Research Institute of China National Standard Materials. The CH4 retention time was about 3 min and the precision of this method was about 3% (Zhang et al., 2004).
2. Materials and methods
The saturation (R, %) and sea-to-air ﬂux (F, μmol m-2 d-1) of the CH4 were estimated using the following equations:
2.2. Saturation and ﬂux calculations
2.1. Sampling strategies and analysis Three multidisciplinary cruises were conducted in the continental shelf and slope region (water depths of 100–2000 m, Fig. 1a–c) of the northern SCS by the RV “Nan Feng” during 10–29 of October in 2014, 10–30 of June in 2015, and 17–31 of March in 2017 in the framework of the comprehensive study program “Living-resource and Ecosystem Dynamics on the slope of the South China Sea” (LEDS). The sampling strategy is aimed at resolving the spatial distribution of various biogeochemical variables in the slope region of the SCS, identifying the diurnal vertical transportation of mesopelagic ﬁsh, and understanding the response of the biogeochemical variables to the active transport of carbon by the mesopelagic ﬁsh. Seawater samples were collected at diﬀerent depths using 5-L Niskin bottles mounted on a Sea-Bird 911 plus conductivity-temperature-depth (CTD) rosette. Surface waters were collected at a depth of ∼5 m and bottom waters were typically collected at ∼10 m above the seaﬂoor. Discrete surface waters (∼5 m) were collected from the shipboard underway sampling system to observe the variation of CH4 in the Pearl River Estuary and to improve the spatial resolution in the shelf and slope regions (Fig. 1a and b).
R = Cobs / Ceq × 100%
F = k w × (Cobs − Ceq)
where Cobs is the observed concentration of dissolved CH4 and Ceq is the air-equilibrated seawater CH4 concentration, which was calculated for the in situ temperature and salinity using the solubility data of Wiesenburg and Guinasso (1979). The atmospheric CH4 mixing ratios (1.84 ppm for 2014 and 1.85 ppm for 2015) from the National Oceanic and Atmospheric Administration/Earth System Research Laboratory (NOAA/ESRL) Global Monitoring Division in situ program were used to calculate Ceq. Previous researchers have derived many empirical relationships for estimating kw, the gas transfer coeﬃcient (cm h-1), which is a function of wind speed and the Schmidt number (Sc). Wanninkhof (2014) (W14) recently updated the most frequently used method (Wanninkhof, 1992), reﬂecting advances over the last two decades in quantifying the gas transfer coeﬃcient. There are no direct observations of the gas transfer coeﬃcient in the SCS and Pearl River Estuary; therefore, the relationships of Wanninkhof (2014) were used to compute kw in this study. For the daily average wind speeds, the National 35
Deep-Sea Research Part II 167 (2019) 34–45
W. Ye, et al.
Fig. 1. Study area and sampling locations. Blue triangles: conductivity/temperature/dissolved oxygen (CTD) stations; red crosses: surface water stations; green star: diurnal stations. (a) October 2014, (b) June 2015, (c) March 2017, (d) June 2014, and (e) Satellite-derived Sea Surface Height (SSH) in June 2015 (http://marine. copernicus.eu/). (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the Web version of this article.)
3. Results and discussion
Centers for Environmental Prediction (NCEP) Reanalysis data from the NOAA ESRL (http://www.esrl.noaa.gov/psd/data/gridded/data.ncep. reanalysis.surface.html) were used for estimations of the gas transfer coeﬃcients.
3.1. CH4 in the Pearl River Estuary and the SCS and comparison with previous studies Table 1 summarizes the CH4 measurements in the surface waters of the Pearl River Estuary, the shelf and slope, and basin regions of the SCS in this study and previously published papers. The CH4 concentrations in the Pearl River Estuary were high and showed substantial spatial and temporal variations that ranged from 9.4 to 430.3 nM in October 2014 and from 52.4 to 305.0 nM in June 2015. The mean CH4 concentration was 20% higher in June 2015 than in October 2014 and was more than twice that reported for September 2006 (61.4 ± 56.1 nM) (Zhou et al., 2009). These large CH4 diﬀerences between diﬀerent surveys in the summer can be attributed to diﬀerences in the freshwater discharge and tides, which inﬂuence the hydrodynamics and the bottom hypoxia of the estuary. For example, water discharge from Xijiang, the largest tributary of the Pearl River, was 1.1 times higher during June 2015 (40 × 109 m3) than in September 2003 (19 × 109 m3) and 1.7 times higher than in September 2006 (15 × 109 m3) respectively (River Sediment Bulletin of China). However, the highest CH4 value of about 3000 nM was observed during September 2003 (Chen et al., 2008). During the survey in the Pearl River Estuary in September 2003, severe bottom water hypoxia occurred and low oxygen (5 μmol L-1) was observed even in the surface water (Chen et al., 2008). Hence high CH4 was produced in the bottom waters and the sediments may have been transported to the surface when tides broke the stratiﬁcation (Cai et al., 2013). The CH4 concentrations were much lower in the continental shelf and slope region (1.8-7.0 nM) than in the Pearl River Estuary and showed less spatial variations (Table 1). The temperatures in October 2014 ranged from 26.8 ° to 27.9 °C and the salinity from 33.5 to 34.1, with increases from the east to the west (Fig. 2a and b). In contrast, the temperature was higher in early summer (June 2015) and ranged from 29.4 ° to 31.1 °C. The mean CH4 levels in October and June were
2.3. Deﬁnition of mixed layer depth The dissolved CH4 may be lost from the seasonal thermocline via vertical mixing and gas exchange. Thus, the mixed layer depth (MLD) is important to calculate the mass balance of CH4 in the mixed layer. The threshold method was used to determine the MLD. The potential density diﬀerences of ∼0.03 kg m3 (σ-σ0) and the temperature diﬀerences of 0.5 °C (T-T0) were used as the criteria to identify the base of the MLD, where σ (T) and σ0 (T0) are the densities (temperature) at the base of the MLD and at the depth of 10 m respectively. This deﬁnition has been used by numerous previous researchers (de Boyer Montégut et al., 2004; Weller et al., 2013). The more rigorous deﬁnition of 0.125 kg m3 (Levitus, 1982) is a visual inspection of a representative sample of proﬁles after a review of the mixed layer processes (de Boyer Montégut et al., 2004). The MLD based on our criterion was not particularly sensitive to a criterion ranging from 0.03 to 0.15 kg m3 with diﬀerences less than 3–7 m. 2.4. Statistical analysis Principal component analysis (PCA) is a multivariate statistical analysis technique, in which a group of correlated variables is transformed into a new group of variables that are uncorrelated or orthogonal to each other (Jackson, 1991). The datasets of the four cruises were analyzed using PCA and the correlation matrix. Prior to the statistical analysis, the variables were standardized to eliminate the inﬂuence of the diﬀerences in data magnitude and measurement scales (Webster, 2009). A VARIMAX rotation was performed to aid with the interpretation of the results. Factors with eigenvalues of 0.8 or greater are selected and considered signiﬁcant. 36
Deep-Sea Research Part II 167 (2019) 34–45 Ma and Cui (2013) This study This study This study Rehder and Suess (2001) Tseng et al. (2017) Ma and Cui (2013) This study 0.14-19.5 8.6 ± 1.4 6.0 ± 2.8 5.1 ± 1.6 4.5 6.3 ± 0.9 0.14-19.5 5.7 ± 1.1 6.3a 8.0 ± 4.3b 4.1 ± 5.2b 1.1 ± 0.8b 0.4a 4.3 ± 3.9b 9.7a 1.9 ± 1.2b 60-299 178 ± 37 181 ± 60 129 ± 14 110 NG 60-299 140 ± 16 1.8-6.5 3.4 ± 0.7 3.3 ± 1.1 2.6 ± 0.3 1.9-2.1 3.4 ± 1.0 1.8-2.5 2.6 ± 0.4 NG 33.9 ± 0.2 33.2 ± 1.0 34.2 ± 0.4 NG 32.9 ± 1.03 33.5 ± 0.2 33.6 ± 0.2 NG 27.5 ± 0.3 30.5 ± 0.6 25.2 ± 0.6 29.0 NG 29.5 ± 0.4 28.9 ± 0.6 46 15 16 16 14 30 22 63 Deep Basin
Continental Shelf and Slope
The dissolved CH4 in the Pearl River Estuary was negatively correlated with the salinity and distance from the river mouth in October 2014 and June 2015 (Fig. 3); this result is similar to the results reported by Chen et al. (2008) and Zhou et al. (2009). The CH4 had the most rapid loss rate at the beginning of the mixing, e.g. from ∼400 nM at the freshwater endmember to ∼100 nM at the zone with a salinity of 10 in October; this was attributed to the strong diﬀusion to the atmosphere (as discussed in Section 3.5) and the aerobic CH4 oxidation. As the salinity continued to increase, the CH4 concentrations tended to vary less. The nonlinear plot of CH4 versus salinity is one of the major patterns of CH4 concentration along salinity gradients (Borges and Abril, 2011). Indeed, this mixing behavior is commonly found in many estuaries, such as the Thames, the Loire, and the Changjiang (Middelburg et al., 2002; Zhang et al., 2008b). However, the CH4 outgassing in the Mekong River (southern SCS) showed a linear pattern in spring but nonlinear pattern in autumn, indicating seasonality (Borges et al., 2018a). Our results were 7 times higher in autumn but 2 times lower in summer than those in the Mekong River in the same seasons (Borges et al., 2018a). The CH4 diﬀerences between the Pearl River and the Mekong River may probably due to the quantity of freshwater discharge (Borges et al., 2018a). Although the CH4 level decreased rapidly with increasing the distance from the river mouth, the riverine input still had a large impact on the CH4 distribution in the shelf and slope regions of the northern SCS. The CH4 levels decreased to 19.4 nM at the surface water with a salinity of 29.6 in October and to 52.4 nM at the surface water with a salinity of 30.3 in June (Fig. 3). Those CH4 values are still ∼5–15 times higher than the mean CH4 value observed in the shelf and slope region, which means the shelf water (salinity < 32.0) contains higher levels of CH4 than other water masses (such as the SCS water). Hence the distribution of CH4 in the slope region of the northern SCS is inﬂuenced by cross-shelf transport driven by eddies (He et al., 2016). Actually, mesoscale eddies are a common hydrographical feature in the northern SCS, which is surrounded by cyclonic gyres driven by the East Asian monsoon (Gan and Qu, 2008; Hu et al., 2011). Those mesoscale eddies are highly active in the shelf (∼33 yr-1) (Xiu et al., 2010) and are
NG: Not Given. a Kw was estimated by the equation from Wanninkhof (1992). b Kw was estimated by the W14 equation.
Chen et al. (2008) Zhou et al. (2009) This study This study Rehder and Suess (2001) Chen et al. (2008) Zhou et al. (2009) Tseng et al. (2017) NG 3.5 5.7 ± 0.9 3.8 ± 1.0 4.0 NG 10.7 7.0 ± 1.3 NG 63.5 ± 32.2b 314.3 ± 464.9b 184.2 ± 187.5b 4.9a NG 15.6 ± 8.0b 12.0 ± 7.4b NG 329-7896 5066 ± 5908 6166 ± 3611 201 NG 134-297 NG NG NG 28.1 ± 1.1 28.8 ± 0.5 NG 28.0-29.4 NG NG
Sep. 2003 Sep. 2006 Oct. 2014 June 2015 Sep. 1994 Sep. 2003 Sep. 2006 Sep. 2003; Jul. 2005; 2006 Apr.–May 2010 Oct. 2014 June 2015 Mar. 2017 Sep. 1994 Jul. 2007 Apr.–May 2010 June 2014 Pearl River Estuary
2 27 14 7 8 22 55 42
3.0-4.5 0.2-28.0 18.3 ± 11.1 9.3 ± 12.5 NG 33.4-33.7 NG 33.2 ± 1.1
23.3-2984 61.4 ± 56.1 112 ± 113 136 ± 85 2.0-6.5 3.0-7.0 3.8 ± 1.7 5.2 ± 2.1
Flux (μmol m-2 d-1) Surface CH4 (%) Surface CH4 (nM) Surface Salinity Surface T (°C) Surveying date
No. of stations
3.4 ± 0.7 nM and 3.3 ± 1.1 nM, respectively. The highest CH4 level (4.5 nM) occurred at station L05 in the early autumn (October 2014) and the highest CH4 levels in early summer of 2015 were concentrated at stations L09-L12, corresponding to the lowest salinity levels (Fig. 2e and f). However, the highest CH4 concentration in June 2015 was 5.4 nM at station L04 (which was near station L5 in October 2014). The CH4 was distributed rather uniformly in the winter (March 2017, Fig. 2i) and the lowest mean value of 2.6 nM corresponded to the lowest temperature (Table 1). In general, the range of our measured CH4 values (1.7–5.4 nM) is within the ranges reported by previous studies conducted in the northern continental shelf and slope region of the SCS. In the deep basin area, the temperatures decreased from north to south in June 2014 but the salinity was higher in the north (Fig. 2j and k). The surface CH4 concentrations ranged from 2.0 to 3.5 nM with an average of 2.6 ± 0.4 nM and the distribution was rather uniform. A higher CH4 concentration (3.8 nM) occurred near Luzon Island but lower values (2.0 nM) were observed in the center of the deep basin (> 2000 m). This is partly because the deep basin areas are farther away from land and rivers and the advective transport of CH4 was limited. Hence, the CH4 concentrations exhibited no obvious variations in this region and were ∼25% lower than those in the shelf and slope regions. Our CH4 results are comparable with those of previous studies (1.8–3.4 nM, Table 1) in the same region but the CH4 level of 2.6 nM in June 2014 was slightly higher than that in September 1994 (2.0 nM) and April–May 2010 (2.2 nM) but lower than that in July (3.4 nM). This variation may be due to seasonal temperature diﬀerences or the variability of the analytical methods. 3.2. Eﬀect of riverine input on the distribution of CH4 in the slope region of the SCS
Table 1 Surface saturation and sea-to-air ﬂux of CH4 in the Pearl River Estuary and the shelf and deep basin of the SCS (mean ± SD).
Wind speed (m s-1)
W. Ye, et al.
Deep-Sea Research Part II 167 (2019) 34–45
W. Ye, et al.
Fig. 2. Spatial proﬁles of temperature (°C), salinity, and CH4 (nM) in the surface waters during October 2014 (a-c), June 2015 (d-f), March 2017 (g-i), and June 2014 (j-l).
observed during the survey in June 2015 (Fig. 1e) (Chen et al., 2016). It was conﬁrmed by satellite remote sensing and by in situ observations of the low salinity in the slope region that the eddy-entrained Pearl River plume was injected into the SCS along stations L9-L12 (He et al., 2016) (Fig. 2e). Due to the inﬂuence of this plume water, high CH4 concentrations of about 4.0 nM were observed along stations L9-L12 (Fig. 2f). In contrast, the CH4 concentrations were less than 3.0 nM along the other two sections dominated by the SCS water or the Kuroshio water (Li et al., 2018). Moreover, other than a direct cross-shelf transport of rich CH4, the eddy-entrained plume has an indirect impact on the CH4 distribution in the slope or deep basin. For example, phytoplankton blooms were observed in the slope and basin of the northern SCS during June 2015 and were mainly induced by the injection of a nutrient-rich plume (He et al., 2016). In fact, even without the transport of nutrients by the plume, the cyclonic eddy itself in the SCS, like other cold eddies in the Northern Hemisphere, is characterized by enhanced chl-a concentrations and primary production due to the upwelling subsurface nutrients and can also trigger phytoplankton blooms (Chen et al., 2007). Weller et al. (2013) conducted time-series observations in the Paciﬁc Ocean and suggested that CH4 can be accumulated during eddy-induced blooms due to the increase in the potential substrates. Hence, regardless of whether the blooms were triggered by plume injection or the eddy itself, the side eﬀect of such an event on the CH4 dynamics cannot be neglected and needs to be reconsidered in the eddyactive SCS. Further, such an injection is not a single or speciﬁc event.
Fig. 3. Correlation between CH4 concentration and salinity/distance from the river mouth in the surface waters of the Pearl River Estuary.
potential drivers for transporting the Pearl River plume to the slope or even to the deep basin. Fortunately, a cyclonic eddy (on the east of study area) and an anticyclonic eddy (on the west of study area) were 38
Deep-Sea Research Part II 167 (2019) 34–45
W. Ye, et al.
Fig. 4. Vertical distribution of temperature (red circles), salinity (blue crosses), DO (purple squares), chl a (green crosses), and CH4 (black triangles) in the SCS during October 2014 (a and b), June 2015 (c and d), March 2017 (e and f), and June 2014 (g and h). (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the Web version of this article.)
parameters (loadings 0.89, -0.88, 0.87 for temperature, salinity, and DO, respectively). The loadings of the parameters associated with the sinking particles, e.g. TSM (0.69) and POC (0.57), along with CH4 (0.60) for the ﬁrst factor indicate that the physical processes were signiﬁcantly correlated with the CH4 concentration. The second factor is loaded strongly by chl-a (0.89) and CH4 (0.62) and to a lesser extent by salinity (0.29), suggesting that this factor represents the inﬂuence of the biological processes on the CH4 variability because the chl-a concentration is usually correlated with phytoplankton biomass. The vertical proﬁle patterns of temperature, salinity, DO, chl-a, and CH4 in June 2015 (Fig. 4c and d) were similar to those in October but showed greater variability at certain depths (e.g. at the depth of 40–60 m, the chl-a concentrations were 0.2–0.8 mg m-3 in October but 0.2–2.0 mg m-3 in June). Generally, the temperatures declined gradually from the surface to the bottom and the salinity increased signiﬁcantly within the upper 150 m and then decreased slightly towards the bottom. The DO distribution was almost the same as in October and had a high correlation with temperature (r=0.95, p < 0.01, Pearson, Supplementary Table S4). There were similar CH4 maxima at 80–120 m, corresponding to the chlorophyll maximum layer. However, abnormally high CH4 concentrations (Fig. 4d) were observed in the mesopelagic zone (400–1000 m) with a potential density of 26.33 kg m3 (Fig. 6b). Several previous studies have observed abnormally high CH4 concentrations occurring at moderate depths (500–800 m) in the SCS and attributed this to a CH4 release from sediments and/or gas hydrates (Chen and Tseng, 2006; Zhou et al., 2009; Tseng et al., 2017). Our observations are consistent with those results and support the ﬁnding that active CH4 release from the sediments exists in the continental slope of the northern SCS. Zhang et al. (2014) observed deepsea sediment transport processes driven by mesoscale eddies in the northern SCS. They recorded an increase in suspended sediments that were trapped and transported from the southwest of Taiwan by mesoscale eddies. Considering the mesoscale eddy-pair observed in the northern SCS during June 2015, the abnormally high CH4 concentrations observed at 400–1000 m in this study may be related to the high CH4 release caused by an increase in suspended sediments driven by the
For example, He et al. (2016) reported the injection of a Pearl River plume into the SCS in the summer of 2010. Pfeiﬀer-Herbert et al. (2015) estimated that 32% of the riverine CH4 input into the Columbia River Estuary (western U.S.) was transported into the ocean. Overall, the injection of a Pearl River plume rich in CH4 may have a profound inﬂuence on the carbon cycle and the CH4 dynamics in the slope region or even in the basin of the northern SCS. 3.3. Vertical distribution of dissolved CH4 in the shelf and slope region of the SCS Fig. 4 shows the vertical distribution of temperature, salinity, DO, chl-a, and CH4 in the shelf and slope region of the SCS for 11 stations in October and 16 stations in June and March respectively. In the early autumn (Fig. 4a and b), the water temperature and salinity were wellmixed in the upper 100 m. Subsequently, the temperatures declined gradually from the depth of 100 m to the bottom but the salinity increased with increasing depth; the maximum occurred at around 120 m and then the values decreased slightly to the bottom. The dissolved oxygen (DO) was well-mixed in the upper 100 m and decreased at the depth of 100–800 m, which agrees with the changes in the temperature with depth. Below the oxygen minimum zone (600–800 m) where the microbial consumption is lower than in the mesopelagic zone (Naqvi et al., 2010), the DO increased with decreasing temperature, indicating an increase in the solubility. The chlorophyll maximum layer was present at ∼80–120 m in autumn and indicates that the phytoplankton accumulates in this layer. The surface CH4 concentrations varied signiﬁcantly as discussed in Section 3.1. The vertical CH4 maxima (which correlate with high levels of chl-a) were present at ∼100 m in autumn. Below the peaks, the CH4 decreased gradually to less than 2.0 nM at the bottom. The CH4 peaks can be inﬂuenced by multiple factors. Thus, the PCA was used to determine the correlations of those factors and their impact on the CH4 distribution. Fig. 5a reﬂects the loading plots for October 2014. Two factors were extracted and they accounted for 75% of the variance in the data (Supplementary Table S1). From these plots, it can be deduced that the ﬁrst factor is related to the three hydrological 39
Deep-Sea Research Part II 167 (2019) 34–45
W. Ye, et al.
Fig. 5. Loading plots corresponding to the ﬁrst two factors following VARIMAX rotation of the principal components during (a) October 2014, (b) June 2015, (c) March 2017, and (d) June 2014. Black triangles CH4; Red circles, temperature; Blue crosses, salinity; Purple square, DO; Green pluses, chl a; Pink stars, TSM; Orange circle pluses, POC; Magenta ﬁlled circle, DMS; Dark brown diamond, DMSP; Sea green square plus, DMSO. (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the Web version of this article.)
varied little at each depth and were below 4.0 nM in the whole water column. The signiﬁcant correlation between temperature and CH4 (r=0.76, p < 0.05, Pearson, Supplementary Table S6) indicates that temperature is an essential factor inﬂuencing the CH4 distribution in the winter. The PCA results (Fig. 5c) show that the ﬁrst factor is loaded strongly by temperature (0.98), DO (0.94), and CH4 (0.86), suggesting the characteristics of the soluble gases.
eddies. Another CH4 source at these depths may be related to the large amounts of mesopelagic ﬁsh observed at 200–1000 m in this survey; many of these species ascends to the surface at dusk to feed and inhabit deeper waters (400–600 m) during the daytime, where they produce fecal pellets and provide favorably anaerobic micro-environments (guts, fecal pellets) for methanogenesis. Although the POC maxima were observed at water depths of 400–600 m (Zhou et al., unpublished data), we currently have no direct evidence to support this view. The PCA results (Fig. 5b) indicate that more than 45% of the data variability is still explained by physical variables such as temperature (0.65) and salinity (-0.85) which were signiﬁcantly loaded on the ﬁrst factor (Supplementary Table S3). Correspondingly, the TSM (0.81) and POC (0.93) are also strongly loaded on the ﬁrst factor, which does not appear to inﬂuence the CH4 concentrations because CH4 is only loaded 0.03 on the ﬁrst factor. This is quite diﬀerent from the situation in October 2014 and is likely due to the more complex conditions in June 2015 (e.g. eddies and plume injection as discussed above). Factor 2 only shows a strong loading by CH4 (0.92), along with temperature (0.67) and DO (0.72), indicating the characteristics of the soluble gases. In fact, there were no signiﬁcant correlations between the CH4 and any other single parameter in June 2015 (Supplementary Table S4); this may further support the view that the CH4 concentrations were inﬂuenced by multiple factors. The vertical distribution of temperature, salinity, DO, and CH4 in March 2017 (Fig. 4e and f) was similar to that in autumn. In general, the temperature decreased gradually from the surface to the bottom and the salinity increased with increasing water depth in the upper 120 m, then decreased slightly until a depth of 500 m. The CH4 concentrations
3.4. Depth proﬁles of CH4 in the deep basin of the SCS Stations H5, H7, and D8 (June 2014) are in the deep basin region of the SCS (Fig. 4g) where the depth is greater than 3800 m. The temperature and salinity in these stations follow similar patterns (Fig. 6), suggesting that the water column in this region has a stable hydrography. The temperatures declined sharply from the surface to ∼1500 m, then decreased slightly to the bottom. The salinity had maxima (∼34.5) at depths of 150–200 m, decreased gradually from 200 to 500 m, and was 34.6 at the bottom. However, the CH4 distributions in these water columns had slight diﬀerences. For example, the CH4 maxima in the subsurface layer varied among diﬀerent locations. The CH4 maxima (∼6.5 nM) in the thermocline were higher at the center of the basin (stations H5 and H7) than in the west of the Luzon Strait (D8) (∼4.5 nM). The vertical proﬁles of CH4 below 500 m also showed diﬀerent patterns. The CH4 at station H5 was stable at ∼0.5 nM in the deep water but ﬂuctuated at station H7 around ∼1.0 nM below 500 m and decreased to 0.7 nM at the bottom. At station D8, weak CH4 maxima occurred between 500 and ∼1000 m. Below 1000 m, the CH4 level decreased gradually with increasing depth and 40
Deep-Sea Research Part II 167 (2019) 34–45
W. Ye, et al.
Fig. 6. T-S diagram and CH4 concentrations in the SCS during (a) October 2014, (b) June 2015, (c) March 2017, and (d) June 2014. SW: Shelf Water; SCSW: South China Sea Water; KW: Kurishio Water (Li et al., 2018).
Correspondingly, the inﬂuence of temperature appears to be spread over factors 1 and 2, suggesting the eﬀect of temperature on the solubility of dissolved gases (such as DMS) and the microbial activities (such as the biological production of DMS). For the second factor, CH4 exhibits a strong positive loading (0.92), indicating that the CH4 concentration in the deep basin has a signiﬁcant correlation with the biogeochemical cycling of DMS. Damm et al. (2010) found that CH4 can be formed by the decomposition of methyl-rich organic sulfur compounds (DMS/DMSP) in aerobic surface waters under speciﬁc conditions, such as nutrient limitation. Thus, we speculate that the DMS/DMSP derived from algae might function as precursors for CH4 production in the SCS; however, this requires further investigation. Station P4 is in the east of Luzon Strait, where a branch of the Kuroshio (marginal current of the northwest of Paciﬁc Ocean) ﬂows westward. The proﬁles of temperature and salinity at station P4 show the inﬂuence of the Kuroshio with high temperature (29.6°C) and salinity (34.9) in the surface layer and the lowest salinity (34.3) in the middle layer (∼500 m) (Fig. 4h). The salinity increased slightly below
then increased more rapidly toward the bottom (> 3500 m), suggesting the presence of additional CH4 sources in the sediments. Previous studies showed the presence of cold seeps and gas hydrates on the continental slope of the northern SCS (Wu et al., 2005), which provides abundant CH4 to the water column. However, there is no other source of CH4 in deep water (2500–3500 m) and the concentration fell to ∼1.0 nM due to slow oxidation. In June 2014, we measured the vertical proﬁles of DMS, DMSP, and DMSO at station H4 (116°E, 17°N), which is close to H5 (116°E, 16°N). The temperature and salinity have signiﬁcant correlations at various depths (r2=1.00 and 0.97, p < 0.01, Pearson); therefore, we postulate that the CH4 proﬁles were similar at station H4 and H5. Thus, PCA is used to determine the substantial inﬂuence of a sulfur compound (Fig. 5d). Factor 1 still accounted for over 50% of the variability and was strongly loaded by temperature and salinity (0.85 and -0.99 respectively, Supplementary Table S7). The second factor implies a link between the microbial metabolism of the methyl-sulfur compound and the biogeochemical cycle of CH4 as the loadings of DMS, DMSP, and DMSO were 0.92, 0.42, and 0.71 respectively. 41
Deep-Sea Research Part II 167 (2019) 34–45
W. Ye, et al.
Table 2 CH4 budget in the mixed layer. The net production rates of CH4 were calculated at diurnal stations after a 24-h observation period. The gross CH4 production represents the overall daily change in CH4 concentration in the mixed layer, including sea-to-air losses, vertical exchange with deep water (a positive value indicates downward transport and a negative value indicates upward transport), and CH4 oxidation. In all calculations, we considered the advective transport as negligible. The CH4 oxidation rates were calculated from a ﬁrst-order parameterization, in which the CH4 oxidation rate (OR)=0.0153*[CH4] (Ward and Kilpatrick, 1990). A positive value indicates CH4 production and a negative value indicates CH4 consumption. The deﬁnition of mixed layer depth (MLD) is given in section 2.3. Date
Mixed layer depth (m)
Sea-to-Air Flux (nM d-1)
Vertical Exchange (nM d-1)
CH4 oxidation rate (nM d-1)
Net CH4 production rate (nM d-1)
Gross CH4 production rate (nM d-1)
Oct. 26 2014 June 25 2015
during June 2015, the coexistence of chl a and the CH4 maxima in the water column at all stations indicate the presence of abundant CH4 precursors (such as DMS/DMSP) in the euphotic zone. Recent research suggested that CH4 can be produced from the decomposition of DMS/ DMSP/DMSO in the oxygenated upper oceans (Damm et al., 2010; Zindler et al., 2013). Most recently, Schmale et al. (2018) indicated that the CH4 accumulations in the upper oxygenated water column in the Baltic sea might be inﬂuenced by changes in the dominant copepod species and food web structure. Hence, the distribution and production of CH4 appear to be a complex process aﬀected by physical, chemical, and biological factors. A mass balance in the mixed layer can be used to describe the sources and sinks of the CH4. The CH4 formation rate in the mixed layer can be calculated by the sea-to-air gas exchange, the advective and vertical transport, and methanotrophic oxidation. We postulate that the gross CH4 production rate is a function of the observed increase in concentration during a certain period (net CH4 production/consumption) plus losses to the atmosphere, vertical exchange with deep water, lateral losses, and CH4 oxidation. We can thus estimate the gross CH4 production rates in the mixed layer for the diurnal stations in October 2014 and June 2015 (Table 2). Given that the water mass was relatively stable in this area (Figs. 4 and 6), we assume the lateral losses of CH4 were negligible. The vertical CH4 transport was estimated based on a non-advective Fickian turbulent diﬀusion model on the gradient of measured CH4 concentration (Schmale et al., 2018). In brief, the vertical CH4 transport (μM m2 day-1) was estimated using the following equation: V(CH4)=Kp*d[CH4]/dh, where Kp (m2 s-1) is the diapycnal diﬀusivity or eddy viscosity, which is 10-3 in the SCS as reported by Lien et al. (2005). [CH4] is the mean CH4 concentration for a 24-h observation period at the depth of h (m) below the mixed layer. The CH4 oxidation rate (OR) was computed using a ﬁrst-order equation (OR=0.0153*CCH4) as reported by Ward and Kilpatrick (1990), where CCH4 is the mean CH4 concentration in the mixed layer. According to our diurnal observations, the net CH4 production rate was -0.24 nM d-1 in autumn and 0.06 nM d-1 in summer (Table 2). The sea-to-air loss accounts for 0.07–0.19 nM d-1 (Table 2). Hence the gross production seems to be nearly balanced throughout the day and was estimated at -0.19 nM d-1 in autumn and 0.11 nM d-1 in summer. These estimates show great seasonality and are comparable to those in subtropical ocean waters (0.04-0.06 nM d-1) (Weller et al., 2013), in which a phytoplankton bloom in a subtropical mesoscale eddy enhanced the CH4 production. Other than anaerobic methanogenesis in the microniches, our observations indicated that the oxygen was present in the mixed layer (3.5–4.0 mL L-1) at the diurnal stations, which means that the aerobic CH4 production may be another contributor to the excess CH4 in the upper water column. The high CH4 production rate in the mixed layer (particularly in summer) may result from these processes. First, DMSP is mainly produced by some marine phytoplankton and then released to well-oxygenated seawater where it serves as an initial substrate for CH4 production. Simó et al. (2002) found that the DMSP synthesis in oceanic phytoplankton was a diurnal process and proportional to photosynthesis. Hence, photosynthesis may increase DMSP in algal cells, which are then released to the ambient seawater via cell lysis and may serve as a
500 m but the temperature decreased with increasing depth and was extremely low in the deep water. The CH4 concentration was only 3.0 nM in the surface water, increased gradually in the mixed layer, and reached a maximum (4.3 nM) at 80 m. the CH4 concentration decreased thereafter with increasing depth but reached a maximum at ∼500 m (the depth with the lowest salinity). The temperature-salinity (T-S) diagram shows that the physical conditions are quite diﬀerent at station P4 and the SCS (Fig. 6d); this is likely due to the inﬂuence of the Kuroshio. Ye et al. (2016) reported similar observations for the East China Sea, where the Kuroshio inﬂuenced the CH4 distribution. Our data show that the CH4 decreased with increasing depth until 2500 m but ﬂuctuated slightly between 1000 and 1500 m. We did not collect water samples in this region below 2500 m due to technical diﬃculties. 3.5. Subsurface CH4 maxima and estimation of mixed layer CH4 production Subsurface CH4 maxima at depths of 25–200 m were observed ubiquitously in the water column of the slope and basin of the SCS (Fig. 4). However, these CH4 maxima appeared at diﬀerent depths with various concentrations. In the shelf and slope, the subsurface CH4 maxima occurred at 70–90 m in October 2014 with an overall mean of 4.1 nM. In the early summer, the CH4 peaks were observed at the water depth of 40–50 m with an average of 4.5 nM (station L04-L06). However, the CH4 maxima at stations L11 and L12 were shallower (25 m) probably due to the cyclonic eddy (Fig. 1e), where colder and saltier water (e.g., 27.4 °C and 34.0 at station L11 compared with 29.7 °C and 33.8 at station L06) upwelled to the subsurface and increased the CH4 peaks. In March 2017, the CH4 peaks were weakened due to the wintertime mixing and the overall mean was 3.4 nM at depths of 100–200 m. Subsurface CH4 maxima in the deep basin occurred at depths of 80–100 m in June 2014 and the overall average concentration was 6.0 nM. Thus, the depth and magnitude of the CH4 maximum showed seasonality and can be determined by the hydrographic and in situ conditions. The temperature may be an important factor that inﬂuences the values of the CH4 maxima because CH4 correlates well with temperature during all cruises (r=0.53–0.76, p < 0.05, Pearson). The study in the Belgian coastal zone showed that the change of CH4 is a function of temperature (Borges et al., 2018b). The authors supposed that higher temperature increases methanogenesis and can also enhance the ﬂux of CH4 from gassy sediments (Borges et al., 2018b). Glissman et al. (2004) reported the in situ CH4 formation rate increased with temperature in the range of 0-30 °C. Our results show that the temperature corresponding to the subsurface CH4 maximum was 23.6 °C in early autumn and 26.4 °C in early summer. Hence, higher temperatures may have enhanced the CH4 production rates in June 2015, resulting in a higher concentration of the subsurface CH4 maximum in June than in October. In addition, the CH4 maxima may also be related to the contents of the sinking and suspended particles. In October 2014, CH4 is positively correlated with TSM (r=0.53, p < 0.05, Pearson, Supplementary Table S1) and POC (r=0.58, p < 0.05, Pearson, Supplementary Table S1). A plausible explanation is that sinking and suspended matters provide favorable conditions for methanogenesis and therefore release more CH4 into the ambient water column. According to our observations 42
Deep-Sea Research Part II 167 (2019) 34–45
W. Ye, et al.
enrichment in the photic zone and the correlations with DMSP/DMS/ DMSO, in combination with the biological views from previous studies implied that the excess of CH4 in the mixed layer in the northern SCS may be attributable to the in situ CH4 production in the anoxic microniches by methanogenic archaea and in the aerobic water column by heterotrophic bacteria. Unfortunately, we have no data on the dominant pathways for CH4 production in the water column of the SCS. Thus, further studies are still needed to provide an in-depth explanation of the CH4 paradox in the SCS. 3.6. Sea-to-air CH4 ﬂuxes The surface water of the Pearl River Estuary is supersaturated with CH4 and the CH4 saturation ranged from 5066 ± 5908% in October 2014 to 6166 ± 3611% in June 2015. Hence the estuary is a signiﬁcant source of atmospheric CH4. The sea-to-air CH4 ﬂuxes in this region show great variation, ranging from 12.3 to 1427.6 μmol m-2 d-1 (average: 314.3 ± 464.9 μmol m-2 d-1) in autumn and 52.4 to 539.2 μmol m-2 d-1 (average: 184.2 ± 187.5 μmol m-2 d-1) in summer based on the Wanninkhof (2014) relationship. Our results were much higher than those of Zhou et al. (2009) for the same area (Table 1) and were 1.5-2.5 times higher than those reported for the Mekong delta, indicating the CH4 was more dynamic in the Pearl River (northern SCS) than Mekong River (southern SCS) (Borges et al., 2018b). Besides, the CH4 ﬂuxes in the Pearl River Estuary are within the range of values in European estuaries (17.0-1352.0 μmol m-2 d-1, Upstill-Goddard and Barnes, 2016) and comparable to those reported for western Africa (251187 μmol m-2 d-1) (Koné et al., 2010; Borges et al., 2015). Our results were also 5-8 times higher than that reported in the northeastern Indian (Rao and Sarma, 2017). Geological diﬀerences, seasonal variations, and wind speeds also increase the uncertainties of these estimations. For example, Zhou et al. (2009) found a wide range of CH4 saturation in the Pearl River Estuary in September (329–7900%, Table 1) but the average wind speed (3.5 m s-1, Table 1) during their survey was ∼2 times lower than in the present study. Taken together, we estimate the CH4 emission rate in the Pearl River Estuary as 187.3 μmol m-2 d-1 by combining previous results with the observations from this study (Table 1). Considering the estimated area of the Pearl River Estuary (2000 km2) and the mean CH4 ﬂux, we estimate the CH4 emissions from this region as 1.4 × 108 mol yr-1 (Fig. 7); this is comparable to the emissions of the Changjiang Estuary (8.7 × 108 mol yr-1) (Zhang et al., 2008a). In the continental shelf and slope of the northern SCS, the CH4 saturation in October 2014 ranged from 131% to 261% with a mean of 178 ± 37%. In the early summer, the CH4 saturation ranged from 92% to 298% with a mean of 181 ± 60%. In the winter, the CH4 saturation ranged from 103% to 152% with a mean of 129 ± 14%. It should be noted that most regions in the northern SCS emit CH4 into the atmosphere in autumn and summer but only one station (L14 in June 2015, saturation: 92%) had a net uptake of CH4 during our cruises. Therefore, the continental shelf and slope of the northern SCS is a net source of atmospheric CH4 and the CH4 saturation in this area was comparable in autumn and summer but was higher than in winter (Table 1). The average daily wind speed for this region was 8.6 ± 1.4 m s-1 in October, 6.0 ± 2.8 m s-1 in June, and 5.1 ± 1.6 m s-1 in March (Table 1). We estimate the range of the sea-to-air CH4 ﬂuxes from the shelf and slope region (using the W14 equation) as 2.3 to 17.7 μmol m-2 d-1 (average: 8.0 ± 4.3 μmol m-2 d-1) in autumn, as -0.9 to 20.4 μmol m-2 d-1 (average: 4.1 ± 5.2 μmol m-2 d-1) in summer, and as 0.1 to 1.7 μmol m-2 d-1 (average: 1.1 ± 0.8 μmol m-2 d-1) in winter. Thus, in the continental shelf and slope of the northern SCS, the CH4 ﬂuxes in autumn were higher than in summer and spring (6.3 μmol m-2 d-1) (Ma and Cui, 2013) and were lowest in the winter. Zhou et al. (2009) reported a high CH4 emission rate (15.6 ± 8.0 μmol m-2 d-1) from the northern SCS in summer and attributed it to the high average wind speed (10.7 m s-1). Our results were also lower than those reported by Tseng et al. (2017) in the same region (12.0 ± 7.4 μmol m-2 d-1). Overall, we estimate the
Fig. 7. Mean sea-to-air CH4 ﬂux (W14) and annual CH4 emission in diﬀerent sampling regions.
substrate for CH4 production in the mixed layer. Second, it is widely accepted that the oxidation of DMS is the principal mechanism of DMSO synthesis (Lee et al., 1999). Zindler et al. (2013) reported that DMSO and its degradation products might serve as potential substrates for CH4 production in the oxygen-containing surface layer of the western Paciﬁc Ocean. Indeed, there were positive correlations of CH4 with DMS (r=0.84, p < 0.05, Pearson), DMSP (r=0.59, p < 0.05, Pearson), and DMSO (r=0.88, p < 0.05, Pearson) in the euphotic zone (Supplementary Table S8). Hence, we speculate that DMSP and DMSO derived from algae might serve as precursors for CH4 production in the SCS. Lastly, sunlight may also enhance the decomposition of DMS and increase the levels of methanogenic substrates, such as DMSP and DMSO, and thereby increase the CH4 production rate. Zhang and Xie (2015) reported a photochemical CH4 production rate of (1.98.1) × 108 mol yr-1 in the global ocean. This suggests that CH4 photoproduction via chromophoric dissolved organic matter (CDOM) may also contribute to the CH4 supersaturation in the ocean surface (Zhang and Xie, 2015). Methanogenesis is the biological production of CH4 mediated by methanogenic archaea. The methanogens can only use acetate, hydrogen and carbon dioxide or other methylated compounds as energy substrates (Thauer, 1998, 2008; Lang et al., 2015). The long-standing dogma was that all methanogens taxonomically belonged to the archaeal phylum of Euryarchaeota. However, recent studies challenged the traditional opinion and showed that methanogenesis happens within the Bathyarchaeota other than Euryarchaeota (Vanwonterghem et al., 2016). Whatever phyla the methanogens belong to, the methanogenesis is strictly due to processes present in marine sediments (Xiao et al., 2018). Nevertheless, Bianchi et al. (1992) ﬁrst reported the existence of living and active cells of methanogenic microbes in either fresh zooplankton fecal pellets or large settling particles containing numerous large size fecal pellets. This indicates that methanogenesis was present in micro-niches and can contribute to the CH4 paradox. Other than as a product via catabolism by methanogenic archaea, CH4 can also be produced in the oxygenated water as by-product when heterotrophic bacteria compete energy sources (such as DMSP and MPn) under nutrient-stressed conditions (Karl et al., 2008; Damm et al., 2010). For example, obvious CH4 production were observed under the incubation experiment when the seawaters from the oligotrophic Arctic Ocean spiked with DMSP (Damm et al., 2010). Meanwhile, the Archaea were negligible in the DMSP supplemented approaches while Bacteria became ∼100% of the community during the incubation suggest that the bacterial catabolism of DMSP was responsible for the CH4 accumulation (Damm et al., 2010). Thus, our observations of CH4 43
Deep-Sea Research Part II 167 (2019) 34–45
W. Ye, et al.
mean CH4 emission rate in the shelf and slope region as 7.4 μmol m-2 d-1 based on previous results and our calculated emission rates in this study (Table 1). Thus, based on the estimated area of 7.4 × 104 km2, the annual CH4 emissions is 2.0 × 108 mol yr-1 (Fig. 7). The CH4 saturation in the deep basin area of the SCS (June 2014) was much lower and less variable than that in the continental shelf and slope, which ranged from 108% to 184% (average: 140 ± 16%). We estimated the CH4 ﬂux as 1.9 ± 1.2 μmol m-2 d-1. The mean CH4 ﬂux from the basin was also high in autumn (1.9 ± 1.2 μmol m-2 d-1) and low in late summer (0.4 μmol m-2 d-1) (Rehder and Suess, 2001) but all ﬂuxes were lower than in the spring (9.7 μmol m-2 d-1) (Ma and Cui, 2013). Hence, we estimate the mean sea-to-air CH4 ﬂux from the deep basin of the SCS as 4.1 μmol m-2 d-1 based on previous studies and our data. Considering the estimated area of 2.9 × 105 km2, we, therefore, estimate the annual CH4 emission from this region as 4.3 × 108 mol yr-1 (Fig. 7). Taken together, based on the whole area of the SCS (∼3.5 × 106 km2) and the mean area-weighted CH4 emission rate (7.8 μmol m-2 d-1), we estimate the overall annual CH4 emission from the SCS as 9.9 × 109 mol yr-1. Thus, we found that the SCS accounts for about 1.0% of the total oceanic CH4 emissions (15.8 Tg yr-1, W14), which is comparable to its areal proportion (0.97% of the global ocean surface area). However, these estimations have great uncertainties due to spatial and seasonal variations, the lack of data in the diﬀerent seasons, the large variations in wind speeds, and the lack of data on CH4 release from seepages and hydrates. For example, Di et al. (2014) reported that natural gas seeps on the seabed of the northern SCS may be an important source of atmospheric CH4. Chen and Tseng (2006) also indicated that high levels of CH4 in the continental slopes of the SCS may have been released from sediments and/or originated from CH4 gas hydrates. Moreover, it was shown previously that the wind speed data from the NCEP reanalysis model are underestimated (Smith et al., 2001). Long-time observations of CH4 ﬂuxes in a Belgian coastal zone showed that the NCEP data had a considerable eﬀect on the ﬂux computations. The ﬂuxes were about 2.7 times lower when computed from the NCEP data than from in situ data (Borges et al., 2018b). Besides, the W92 relationship leads to a 1.2-fold higher ﬂux estimate than the W14 relationship. Further studies are needed to better quantify the sea-to-air CH4 ﬂuxes from the SCS.
and 2016YFA0601302, by the National Science Foundation of China through Grant No. 41521064, and was funded by the Taishan Scholars Programme of Shandong Province and the Aoshan Talents Programme of the Qingdao National Laboratory for Marine Science and Technology (No. 2015ASTP-OS08). Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.dsr2.2019.06.016. References Bange, H.W., Bartell, U.H., Rapsomanikis, S., Andreae, M.O., 1994. Methane in the Baltic and North Seas and a reassessment of the marine emissions of methane. Glob. Biogeochem. Cycles 8, 465–480. Bianchi, M., Marty, D., Teyssié, J.-L., Fowler, S.W., 1992. Strictly aerobic and anaerobic bacteria associated with sinking particulate matter and zooplankton fecal pellets. Marine ecology progress series 55–60. Borges, A., Abril, G., 2011. 5.04-Carbon dioxide and methane dynamics in estuaries. Treatise on Estuarine and Coastal Science 5, 119–161 Biogeochemistry. Borges, A.V., Abril, G., Bouillon, S., 2018a. Carbon dynamics and CO2 and CH4 outgassing in the Mekong delta. Biogeosciences 15, 1093–1114. Borges, A.V., Champenois, W., Gypens, N., Delille, B., Harlay, J., 2016. Massive marine methane emissions from near-shore shallow coastal areas. Sci. Rep. 6, 27908. Borges, A.V., Darchambeau, F., Teodoru, C.R., Marwick, T.R., Tamooh, F., Geeraert, N., Omengo, F.O., Guérin, F., Lambert, T., Morana, C., 2015. Globally signiﬁcant greenhouse-gas emissions from African inland waters. Nat. Geosci. 8, 637. Borges, A.V., Speeckaert, G., Champenois, W., Scranton, M.I., Gypens, N., 2018b. Productivity and temperature as drivers of seasonal and spatial variations of dissolved methane in the Southern Bight of the North Sea. Ecosystems 21, 583–599. Cai, S., Zheng, S., Wei, X., 2013. Progress on the hydrodynamic characteristics and the hypoxia phenomenon in the Pearl River Estuary. Journal of Tropical Oceanography 32, 1–8. Chen, C.-T.A., Tseng, H.-C., 2006. Abnormally High CH4 Concentrations in Seawater at Mid-depths on the Continental Slopes of the Northern South China Sea. Chen, C.-T.A., Wang, S.-L., Lu, X.-X., Zhang, S.-R., Lui, H.-K., Tseng, H.-C., Wang, B.-J., Huang, H.-I., 2008. Hydrogeochemistry and greenhouse gases of the Pearl River, its estuary and beyond. Quat. Int. 186, 79–90. Chen, Y.-L.L., Chen, H.-Y., Lin, I.I., Lee, M.-A., Chang, J., 2007. Eﬀects of cold eddy on phytoplankton production and assemblages in Luzon strait bordering the South China Sea. J. Oceanogr. 63, 671–683. Chen, Z., Yang, C., xu, D., Xu, M., 2016. Observed hydrographical features and circulation with inﬂuences of cyclonic-anticyclonic eddy-pair in the northern slope of the South China Sea during June 2015. Journal of Marine Sciences 34, 10–19. Damm, E., Helmke, E., Thoms, S., Schauer, U., Nöthig, E., Bakker, K., Kiene, R.P., 2010. Methane production in aerobic oligotrophic surface water in the central Arctic Ocean. Biogeosciences 7, 1099–1108. Damm, E., Kiene, R.P., Schwarz, J., Falck, E., Dieckmann, G., 2008. Methane cycling in Arctic shelf water and its relationship with phytoplankton biomass and DMSP. Mar. Chem. 109, 45–59. de Angelis, M.A., Lee, C., 1994. Methane production during zooplankton grazing on marine phytoplankton. Limnol. Oceanogr. 39, 1298–1308. de Boyer Montégut, C., Madec, G., Fischer, A.S., Lazar, A., Iudicone, D., 2004. Mixed layer depth over the global ocean: an examination of proﬁle data and a proﬁle-based climatology. J. Geophys. Res.: Oceans 109. Di, P., Feng, D., Chen, D., 2014. In-situ and on-line measurement of gas ﬂux at a hydrocarbon seep from the northern South China Sea. Cont. Shelf Res. 81, 80–87. Dlugokencky, E.J., Nisbet, E.G., Fisher, R., Lowry, D., 2011. Global atmospheric methane: budget, changes and dangers. Phil. Trans. Math. Phys. Eng. Sci. 369, 2058–2072. Gan, J., Qu, T., 2008. Coastal jet separation and associated ﬂow variability in the southwest South China Sea. Deep Sea Res. Oceanogr. Res. Pap. 55, 1–19. Glissman, K., Chin, K.J., Casper, P., Conrad, R., 2004. Methanogenic pathway and archaeal community structure in the sediment of eutrophic lake dagow: eﬀect of temperature. Microb. Ecol. 48, 389–399. Grossart, H.-P., Frindte, K., Dziallas, C., Eckert, W., Tang, K.W., 2011. Microbial methane production in oxygenated water column of an oligotrophic lake. Proc. Natl. Acad. Sci. Unit. States Am. 108, 19657–19661. He, X., Xu, D., Bai, Y., Pan, D., Chen, C.-T.A., Chen, X., Gong, F., 2016. Eddy-entrained Pearl River plume into the oligotrophic basin of the south China sea. Cont. Shelf Res. 124, 117–124. Holm-Hansen, O., Lorenzen, C.J., Holmes, R.W., Strickland, J.D.H., 1965. Fluorometric determination of chlorophyll. ICES (Int. Counc. Explor. Sea) J. Mar. Sci. 30, 3–15. Hu, J., Gan, J., Sun, Z., Zhu, J., Dai, M., 2011. Observed three-dimensional structure of a cold eddy in the southwestern South China Sea. J. Geophys. Res.: Oceans 116. IPCC, 2013. Climate Change 2013: the Physical Science Basis. Cambridge University Press, New York 2013. Jackson, J.E., 1991. A User's Guide to Principal Components. Wiley. Karl, D.M., Beversdorf, L., Bjorkman, K.M., Church, M.J., Martinez, A., Delong, E.F., 2008. Aerobic production of methane in the sea. Nat. Geosci. 1, 473–478. Karl, D.M., Tilbrook, B.D., 1994. Production and transport of methane in oceanic particulate organic matter. Nature 368, 732–734.
4. Conclusions Dissolved CH4 in the northern SCS showed great spatial variations; the highest concentration occurred in the estuaries, followed by the shelf and slope region and the deep basin area. Estuarine mixing has a large eﬀect on the CH4 distribution in the Pearl River Estuary and the adjacent area. The vertical proﬁles of CH4 in the northern SCS have great spatial variations that are attributed to the complex interactions among physical, chemical, and biological processes. Subsurface CH4 maxima were present in the euphotic zones of most stations and likely resulted from in situ CH4 production in anoxic microniches and/or aerobic CH4 formation from substrate precursors. Our estimation of the CH4 production rate implies that the excess of CH4 in the mixed layer may be partly derived from the decomposition of certain CH4 substrates such as DMS, DMSP, and/or DMSO. The surface waters of the northern SCS were generally supersaturated with CH4 and the annual CH4 emission was estimated at 9.9 × 109 mol yr-1. Hence the SCS is a net source of atmospheric CH4. Acknowledgements The authors wish to thank the crews of the R/V “Dong Fang Hong 2” and “Nan Feng” and the colleagues from the Laboratory of Marine Biogeochemistry, Ocean University of China, for assistance in the collection of ﬁeld samples. This work was supported by the Ministry of Science and Technology of China through Grant Nos. 2014CB441502 44
Deep-Sea Research Part II 167 (2019) 34–45
W. Ye, et al.
the south China sea and the Western Philippine Sea. Cont. Shelf Res. 135, 23–34. Upstill-Goddard, R.C., Barnes, J., 2016. Methane emissions from UK estuaries: re-evaluating the estuarine source of tropospheric methane from Europe. Mar. Chem. 180, 14–23. Vanwonterghem, I., Evans, P.N., Parks, D.H., Jensen, P.D., Woodcroft, B.J., Hugenholtz, P., Tyson, G.W., 2016. Methylotrophic methanogenesis discovered in the archaeal phylum Verstraetearchaeota. Nature Microbiology 1, 16170. Wanninkhof, R., 1992. Relationship between wind speed and gas exchange over the ocean. J. Geophys. Res.: Oceans 97, 7373–7382. Wanninkhof, R., 2014. Relationship between wind speed and gas exchange over the ocean revisited. Limnol Oceanogr. Methods 12, 351–362. Ward, B.B., Kilpatrick, K.A., 1990. Relationship between substrate concentration and oxidation of ammonium and methane in a stratiﬁed water column. Cont. Shelf Res. 10, 1193–1208. Webster, R., 2009. Statistics to support soil research and their presentation. Eur. J. Soil Sci. 52, 331–340. Weller, D.I., Law, C.S., Marriner, A., Nodder, S.D., Chang, F.H., Stephens, J.A., Wilhelm, S.W., Boyd, P.W., Sutton, P.J.H., 2013. Temporal variation of dissolved methane in a subtropical mesoscale eddy during a phytoplankton bloom in the southwest Paciﬁc Ocean. Prog. Oceanogr. 116, 193–206. Wiesenburg, D., Guinasso, N., 1979. Equilibrium solubilities of methane, carbon monoxide, and hydrogen in water and sea water. J. Chem. Eng. Data 24, 356–360. Wu, S., Zhang, G., Huang, Y., Liang, J., Wong, H.K., 2005. Gas hydrate occurrence on the continental slope of the northern South China Sea. Mar. Pet. Geol. 22, 403–412. Wu, Y., Bao, H.-Y., Unger, D., Herbeck, L.S., Zhu, Z.-Y., Zhang, J., Jennerjahn, T.C., 2013. Biogeochemical behavior of organic carbon in a small tropical river and estuary, Hainan, China. Cont. Shelf Res. 57, 32–43. Xiao, K.-Q., Beulig, F., Røy, H., Jørgensen, B.B., Risgaard-Petersen, N., 2018. Methylotrophic methanogenesis fuels cryptic methane cycling in marine surface sediment. Limnol. Oceanogr. 63, 1519–1527. Xiu, P., Chai, F., Shi, L., Xue, H., Chao, Y., 2010. A census of eddy activities in the South China Sea during 1993–2007. J. Geophys. Res.: Oceans 115. Yang, G.-P., Jing, W.-W., Kang, Z.-Q., Zhang, H.-H., Song, G.-S., 2008. Spatial variations of dimethylsulﬁde and dimethylsulfoniopropionate in the surface microlayer and in the subsurface waters of the South China Sea during springtime. Mar. Environ. Res. 65, 85–97. Yang, G.-P., Wang, X., Zhang, H.-H., Liu, C.-Y., 2014. Temporal and spatial variations of dimethylsulfoxide in the bohai sea and the yellow sea. J. Sea Res. 90, 33–43. Ye, W., Zhang, G., Zhu, Z., Huang, D., Han, Y., Wang, L., Sun, M., 2016. Methane distribution and sea-to-air ﬂux in the East China Sea during the summer of 2013: impact of hypoxia. Deep Sea Res. Part II Top. Stud. Oceanogr. 124, 74–83. Zhang, G., Zhang, J., Liu, S., Ren, J., Xu, J., Zhang, F., 2008a. Methane in the Changjiang (Yangtze River) Estuary and its adjacent marine area: riverine input, sediment release and atmospheric ﬂuxes. Biogeochemistry 91, 71–84. Zhang, G., Zhang, J., Ren, J., Li, J., Liu, S., 2008b. Distributions and sea-to-air ﬂuxes of methane and nitrous oxide in the North East China Sea in summer. Mar. Chem. 110, 42–55. Zhang, G.L., Zhang, J., Kang, Y.B., Liu, S.M., 2004. Distributions and ﬂuxes of methane in the east China sea and the yellow sea in spring. J. Geophys. Res.: Oceans 109. Zhang, Y., Liu, Z., Zhao, Y., Wang, W., Li, J., Xu, J., 2014. Mesoscale eddies transport deep-sea sediments. Sci. Rep. 4, 5937. Zhang, Y., Xie, H., 2015. Photomineralization and photomethaniﬁcation of dissolved organic matter in Saguenay River surface water. Biogeosciences 12, 6823–6836. Zhou, H., Yin, X., Yang, Q., Wang, H., Wu, Z., Bao, S., 2009. Distribution, source and ﬂux of methane in the western Pearl River estuary and northern South China sea. Mar. Chem. 117, 21–31. Zindler, C., Bracher, A., Marandino, C.A., Taylor, B., Torrecilla, E., Kock, A., Bange, H.W., 2013. Sulphur compounds, methane, and phytoplankton: interactions along a north–south transit in the western Paciﬁc Ocean. Biogeosciences 10, 3297–3311.
Koné, Y.J.M., Abril, G., Delille, B., Borges, A.V., 2010. Seasonal variability of methane in the rivers and lagoons of ivory coast (West Africa). Biogeochemistry 100, 21–37. Lang, K., Schuldes, J., Klingl, A., Poehlein, A., Daniel, R., Brune, A., 2015. New mode of energy metabolism in the seventh order of methanogens as revealed by comparative genome analysis of “candidatus methanoplasma termitum”. Appl. Environ. Microbiol. 81, 1338. Lee, P.A., de Mora, S.J., Levasseur, M., 1999. A review of dimethylsulfoxide in aquatic environments. Atmos.-Ocean 37, 439–456. Levitus, S., 1982. Climatological Atlas of the World Ocean, vol. 13. NOAA/ERL Professional Paper, Washington DC, pp. 173. Li, D., Zhou, M., Zhang, Z., Zhong, Y., Zhu, Y., Yang, C., Xu, M., Xu, D., Hu, Z., 2018. Intrusions of Kuroshio and shelf Waters on Northern slope of south China sea in summer 2015. J. Ocean Univ. China 17, 477–486. Lien, R.C., Tang, T.Y., Chang, M.H., D'Asaro, E.A., 2005. Energy of nonlinear internal waves in the South China Sea. Geophys. Res. Lett. 32. Liss, P.S., Johnson, M.T., 2014. Ocean-atmosphere Interactions of Gases and Particles. Springer. Ma, L., Cui, Y., 2013. Dissolved methane concentration and sea-to-air transfer ﬂux of dissolved methane in the upper seawater of the central and northern South China Sea. Journal ofTropical Oceanography 32, 91–101. Middelburg, J.J., Nieuwenhuize, J., Iversen, N., Høgh, N., de Wilde, H., Helder, W., Seifert, R., Christof, O., 2002. Methane distribution in European tidal estuaries. Biogeochemistry 59, 95–119. Naqvi, S.W.A., Bange, H.W., Farías, L., Monteiro, P.M.S., Scranton, M.I., Zhang, J., 2010. Marine hypoxia/anoxia as a source of CH4 and N2O. Biogeosciences 7, 2159–2190. Pfeiﬀer-Herbert, A.S., Kincaid, C.R., Bergondo, D.L., Pockalny, R.A., 2015. Dynamics of wind-driven estuarine-shelf exchange in the Narragansett Bay estuary. Cont. Shelf Res. 105, 42–59. Rao, G., Sarma, V., 2017. Inﬂuence of river discharge on the distribution and ﬂux of methane in the coastal Bay of Bengal. Mar. Chem. 197, 1–10. Reeburgh, W.S., 2007. Oceanic methane Biogeochemistry. Chem. Rev. 107, 486–513. Rehder, G., Suess, E., 2001. Methane and pCO2 in the Kuroshio and the South China Sea during maximum summer surface temperatures. Mar. Chem. 75, 89–108. Repeta, D.J., Ferron, S., Sosa, O.A., Johnson, C.G., Repeta, L.D., Acker, M., DeLong, E.F., Karl, D.M., 2016. Marine methane paradox explained by bacterial degradation of dissolved organic matter. Nat. Geosci. 9, 884–887. Rhee, T.S., Kettle, A.J., Andreae, M.O., 2009. Methane and nitrous oxide emissions from the ocean: a reassessment using basin-wide observations in the Atlantic. J. Geophys. Res.: Atmosphere 114. Schmale, O., Greinert, J., Rehder, G., 2005. Methane emission from high-intensity marine gas seeps in the Black Sea into the atmosphere. Geophys. Res. Lett. 32. Schmale, O., Wäge, J., Mohrholz, V., Wasmund, N., Ulf, G., Gregor, R., Matthias, L., Natalie, L.W., 2018. The contribution of zooplankton to methane supersaturation in the oxygenated upper waters of the central Baltic Sea. Limnol. Oceanogr. 63, 412–430. Simó, R., Archer, S.D., Pedrós-Alió, C., Gilpin, L., Stelfox-Widdicombe, C.E., 2002. Coupled dynamics of dimethylsulfoniopropionate and dimethylsulﬁde cycling and the microbial food web in surface waters of the North Atlantic. Limnol. Oceanogr. 47, 53–61. Smith, S.R., Legler, D.M., Verzone, K.V., 2001. Quantifying uncertainties in NCEP reanalyses using high-quality research vessel observations. J. Clim. 14, 4062–4072. Su, j., 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, 1745–1760. Thauer, R.K., 1998. Biochemistry of methanogenesis: a tribute to marjory stephenson: 1998 marjory stephenson prize lecture. Microbiology (Read.) 144, 2377–2406. Thauer, R.K., Kaster, A.-K., Seedorf, H., Buckel, W., Hedderich, R., 2008. Methanogenic archaea: ecologically relevant diﬀerences in energy conservation. Nat. Rev. Microbiol. 6, 579. Tseng, H.-C., Chen, C.-T.A., Borges, A.V., DelValls, T.A., Chang, Y.-C., 2017. Methane in