Introduction

The ocean is a major global near-surface reservoir of halogens including chlorine and bromine (Chuck et al. 2005). Volatile halogenated organic compounds (VHOC) include chlorocarbons such as chloroform (CHCl3) and trichloroethylene (C2HCl3), which are significant marine sources of chlorine to the atmosphere (McCulloch et al. 1999; McCulloch 2003). VHOC also include bromocarbons such as bromodichloromethane (CHBrCl2), chlorodibromomethane (CHBr2Cl), and bromoform (CHBr3) that are important marine sources of biogenic reactive organic bromine (Chuck et al. 2005). Marine emissions of trace gases have important implications for atmospheric composition and chemistry (Hughes et al. 2013). VHOC are involved in atmospheric chemical processes; as such, they are considered an important group of compounds. Once emitted to the atmosphere, the halogen-containing VHOC are broken down by photolysis or interaction with OH− or Cl− radicals to release inorganic chlorine and bromine. These halogens take part in reactions that decrease ozone levels of the atmosphere (McGivern et al. 2000; Nielsen and Douglass 2001; Laturnus et al. 2002; Wang et al. 2007) and have implications for the climate (Carpenter and Liss 2000; Cincinelli et al. 2012). Therefore, field investigation data from various marine ecosystems are required to understand VHOC production processes and the role of marine VHOC in global climate change.

The presence of CHCl3 and C2HCl3 in the marine environment is usually related to human activities such as the use of pesticide, anti-freezing agents, dry cleaning fluids, and industrial reagents (Class and Ballschmiter 1986; Wang et al. 1995; Blake et al. 1996; McCulloch et al. 1999). In addition, these two chlorinated compounds are also produced naturally. For example, marine algae are important global sources for CHCl3 and C2HCl3 (Abrahamsson et al. 1995; Nightingale et al. 1995; Scarratt and Moore 1999). Brominated volatile halocarbons (e.g., CHBrCl2, CHBr2Cl, and CHBr3) are generally believed to have mainly a biogenic origin, and are broadly produced by macroalgae and microalgae in the marine environment (Nightingale et al. 1995; Moore et al. 1996; Ekdahl et al. 1998; Quack and Wallace 2003; Colomb et al. 2008; Paul and Pohnert 2011). Successive chlorine substitution as a natural source of CHBrCl2 and CHBr2Cl mainly refers to open ocean and deeper waters, i.e., aged waters (Class and Ballschmiter 1988; Moore and Tokarczyk 1993; Fogelqvist and Tanhua 1995; Sturges et al. 1997). However, if chlorination occurs in coastal waters, the two compounds will also be produced. Additionally, the use of seawater as a cooling medium in coastal nuclear power plants and other plants can lead to the production of a wide range of byproducts, such as CHBrCl2, CHBr2Cl, and CHBr3, which are collectively called trihalomethanes (Fogelqvist and Krysell 1991; Allonier et al. 1999; Quack and Wallace 2003; Zhou et al. 2005; Liu et al. 2011).

Knowledge of seawater and atmospheric concentrations of these trace gases enables calculations of the direction and magnitude of their fluxes across the sea–air interface. Previous studies have reported that the open ocean acts as a source of CHCl3, C2HCl3, CHBrCl2, CHBr2Cl, and CHBr3 to the atmosphere (Abrahamsson et al. 2004b; Moore 2001; Chuck et al. 2005; Quack et al. 2004), as well as a sink for CHBrCl2, CHBr2Cl, and CHBr3 (Chuck et al. 2005; Liu et al. 2013). Moreover, coastal regions have gained increased attention because the concentrations of the compounds in these regions are higher. For instance, coastal Antarctic waters and the western Atlantic coastal waters of the United States of America were found to be important sources of atmospheric CHBr3 (Carpenter et al. 2007; Liu et al. 2011).

In the last decade, the distributions, sources, sea-to-air fluxes of global oceans, and production mechanism of many VHOC have been widely studied in different marine environments (Christof et al. 2002; Huybrechts et al. 2005; Zhou et al. 2005; Carpenter et al. 2007; Liu et al. 2011). However, the research of VHOC along China coastal waters is still at initial stages (Lu et al. 2010; He et al. 2013a, b). The primary objectives of the four cruises in summer 2008 and spring 2009 were to investigate the occurrences, distributions, and temporal variations of five VHOC—CHCl3, C2HCl3, CHBrCl2, CHBr2Cl, and CHBr3 in the southern Yellow Sea (hereafter referred to as SYS) and estimate the fluxes of CHCl3, C2HCl3, and CHBr3 to the atmosphere.

Materials and methods

Study area

The SYS, located in the northwestern Pacific Ocean, is an important marginal, semi-closed sea along China’s coast in the mid-latitude zones. The SYS has a total surface area of 30.9 × 104 km2 and an average depth of 46 m. The SYS opens east into the Pacific Ocean with the Shandong Peninsula to the north and two developed provinces to the west (Fig. 1). This sea is influenced by several rivers, including the Xinyi River, Huaihe River, and Sheyang River. The Xinyi River is a seasonal flood river in the Huaihe River Basin. The annual mean runoff of this river is 16.8 billion cubic meters (Peng et al. 2010). Treated wastewater from Xuzhou and Suqian cities is transferred to the Xinyi River via channelization project and then flows into the SYS to ensure the water quality of the South-to-North Water Diversion Project in Jiangsu Province (Ruan et al. 2008). The major pollution source of the river is industrial wastewater, followed by domestic sewage (Zhang et al. 2005). The volume of treated wastewater from cities discharged to the Xinyi River is 410,000 t d−1 and the pollutant load (pollutants bearing capacity) is 14,965 t a−1 (Ruan et al. 2008). The Huaihe River, one of the main rivers in eastern China, flows through Henan, Anhui, and Jiangsu provinces, and eventually enters the SYS in Yancheng, which is an important industrial city in the north of Jiangsu (Chen 2009). The Sheyang River, located in the city of Yancheng and surrounding industrial areas, is also a significant contributor to the contamination of the estuary (Peng et al. 2010). The SYS is heavily influenced by a circulation system consisting of Yellow Sea Cold Water Mass (YSCWM), Yellow Sea Warm Current (YSWC), and Yellow Sea Coastal Current (YSCC) on the western side (Feng et al. 1999; Yuan et al. 2008). The SYS is also an active zone with interactions among land, sea, and atmospheric processes. The SYS has a very complicated system because of its shallow depth, little deep-water exchange, relatively poor ability of self-purification, complex hydrological conditions, and strong human activities (Feng et al. 1999).

Fig. 1
figure 1

Study area and sampling stations. Plus sign denotes July 2008 cruise, white circle denotes August 2008 cruise, white square denotes March 2009 cruise, and white up-pointing triangle denotes May 2009 cruise. Regional circulation models (YSCC Yellow Sea Coastal Current, YSWC Yellow Sea Warm Current) are indicated with arrows (cited from Bian et al. 2013), and the YSCWM area in summer is shadowed (cited from Wei et al. 2010). The location of the Tianwan nuclear power plant is indicated by star

Full size image

Sampling stations

A total of four cruises were carried out aboard the R/V Dong Fang Hong 2 during summer 2008 and spring 2009 in the SYS. The study area and sampling stations are shown in Fig. 1. The concentrations of the five VHOC in the surface waters were measured at all stations during these four cruises. The first cruise was conducted at the end of the blooming season of Enteromorpha prolifera (Chlorophyta, Ulvophyceae) during 22–26 July 2008 within 120.5–122°E, 33.5–36°N. The blooming of green algae E. prolifera occurs in the SYS during summer season every year, but the affected area has been found to be largest in 2008. According to satellite data, E. prolifera originated from the coasts of Yancheng and Lianyunguang, Jiangsu Province in mid May 2008. In late May 2008, the green alga had grown rapidly, and green patches covered about 1,200 km2. Subsequently, the green alga moved into the middle of the Yellow Sea near to Qingdao, Shandong Province. The large green patches reached the coast of Qingdao in late June 2008 and formed a massive green tide covering 600 km2 along the coast of Qingdao (Liu et al. 2009). The green tide lasted for 2 weeks, and more than 10,000 persons were involved in the cleanup and removal of over 1 million tons of algae from the coast of Qingdao (China Ocean News 2008). The second cruise was performed during 10–14 August 2008 within 120.5–123°E, 34.5–36°N. The third and fourth cruises were carried out during 20–24 March 2009 and 9–16 May 2009 within 120.7–123°E, 34.5–36°N and 120.5–123°E, 33–36°N, respectively. Detailed information about the sampling stations and hydrographic and meteorological parameters in the four cruises is described in Table 1.

Table 1 Description of sampling stations, water depth, and wind speed data in the SYS during the four cruises
Full size table

Analyses of VHOC and chlorophyll a in seawater

Surface seawater samples (average depth of 2 m) were collected by a Seabird 911 CTD Rosette (General Oceanics Co.) sampler. The concentrations of VHOC in the samples were analyzed immediately after collection using a cryogenic purge and trap system coupled to a gas chromatograph with an electron capture detector (GC-ECD). Seawater sample was drawn into a 100 ml glass syringe without ambient air. The sample was directly injected into a glass bubbling chamber at 40 °C and stripped with ultrapure nitrogen at a flow rate of 100 ml min−1 for 12 min. The extracted gases passed through two glass tubes containing anhydrous Mg(ClO4)2 and ascarite to remove water vapor and carbon dioxide, respectively. Subsequently, the extracted gases were trapped in an inert stainless steel tube immersed in liquid nitrogen. After extraction, the stainless steel tube was heated and the desorbed gases were introduced into a GC/ECD (Agilent 6890N) for analysis. Quantification of the target compounds was carried out based on retention times and peak areas of calibration standards. GC/ECD chromatograms of a working standard mixture of VHOC showed that their peaks were separated in the chromatograms (Fig. S1). The calibration standards were diluted twice with methanol to prepare working standard solutions. Microliter volumes of the working standards solutions were injected into the N2-purged seawater and subjected to the same procedure as seawater samples. The analytical precisions of the target compounds were generally in the range of 7–15 % in routine sample analysis. Their detection limits were 0.1–0.5 pmol L−1. The overall accuracy of the method for the target compounds, as determined by repeated analyses of low concentration standards (n = 7), was between 5 and 18 %. He et al. (2013a, b) further described the details of the method of analyses of VHOC.

Surface seawater temperature, salinity, and local water depth were obtained from the Seabird 911 CTD Rosette while ship-based in situ wind speeds were measured by a ship-borne weather instrument (RM Young, USA) at about 10 m above sea surface. A total of 200 ml seawater samples were filtered through glass fiber filters (Whatman GF/F, 47 mm diameter) for chlorophyll a (Chl-a) analysis. Chl-a on the filter was extracted with 10 ml 90 % acetone for 24 h in the dark at 4 °C. The fluorescence of the extract was measured with a fluorescence spectrophotometer (F-4500, Hitachi Co., Japan) after centrifugation according to Parsons et al. (1984).

Saturation anomaly and sea-to-air flux calculations

Saturation anomaly (SA; %) is defined as percent difference from the equilibrium between gas concentration in surface seawater and the atmosphere. A positive SA indicates a net flux from the ocean to the atmosphere (Hughes et al. 2009) calculated by the equation

$$ SA \, \left( \% \right) \, = \, \left( {C_{w} {-}C_{a} /H} \right)/ \, C_{a} /H \times 100 $$

where C w is the measured surface seawater concentration, C a is the atmosphere concentration, and H is the dimensionless temperature-dependent Henry’s law constant. The five VHOC had no in situ measured atmospheric concentrations, so we adopted the mean atmospheric concentrations of 28.34 pptv for CHCl3, 27.32 pptv for C2HCl3, and 5.8 pptv for CHBr3 measured by He et al. (2013a) in the SYS. However, no data on the atmospheric concentrations of CHBrCl2 and CHBr2Cl were available in the study area. Henry’s law constants for CHCl3, C2HCl3, and CHBr3 were calculated from the relationships described by Moore et al. (1995) and Moore (2000) using the measured surface seawater temperature at each station. The sea-to-air fluxes of CHCl3, C2HCl3, and CHBr3 were calculated by the following equation (Liss and Slater 1974):

$$ F = k_{w} \left( {C_{w} {-}C_{a} /H} \right) $$

where k w is the gas transfer coefficient (m day−1) calculated as a function of wind speed and Schmidt number. k w is calculated for CHCl3, C2HCl3, and CHBr3, as described by Liss and Merlivat (1986), respectively, using the following equations:

$$ k_{w} = \, 0.17u\left( {Sc/600} \right)^{ - 2/3} \quad 0 < u \le 3.6 $$
$$ k_{w} = \, (2.85 \times u{-}9.65) \, \left( {Sc/600} \right)^{ - 1/2} \quad 3.6 < u \le 13 $$
$$ k_{w} = \, (5.9 \times u{-} \, 49.3) \, \left( {Sc/600} \right)^{ - 1/2} \quad 13 < u $$

where u is the wind speed (m s−1) and Sc is the Schmidt number. The temperature-dependent Schmidt numbers for CHCl3, C2HCl3, and CHBr3 were calculated using the equations given by Khalil et al. (1999).

Knowledge of the oceanic and atmospheric concentrations of the trace gases enables calculations of their saturation anomalies in the water phase. Therefore, the direction and magnitude of their fluxes across the air–sea interface may also be calculated. The surface seawater concentrations, in situ water temperatures and wind speeds, and sea-to-air fluxes of the three compounds were estimated based on referenced atmospheric concentrations. This type of flux calculation is subject to several uncertainties, e.g., in atmospheric mixing ratios, which are probably inconsistent with the actual values because these gases are quite sensitive to seasonal change. Therefore, the results should be regarded as approximate estimates of the magnitude of fluxes.

Results and discussion

Surface distributions of VHOC in summer 2008 and spring 2009

According to the temperature–salinity (T/S) plot (Fig. 2), the two water masses in the study area during the investigation could be identified as (1) cold and low-salinity water in the coastal region and (2) warm and high-salinity water in open sea. Two perennial oceanic currents were noted in the study area: (1) the southward YSCC, characterized by low temperature and salinity, (2) the northwestward YSWC, a branch of the Kuroshio into the SYS, with protruding feature of high temperature and salinity (Figs. 3a, 4a, 5a, 6a, 3b, 4b, 5b, 6b). The observed hydrographic feature of the water masses was consistent with previous observations (Feng et al. 1999; Yuan et al. 2008). In addition, YSCWM (central position 123°E, 35°N, Fig. 1) in the bottom water forms during spring, matures in summer, and decays in autumn. The YSCWM is affected by thermocline, tidal mixing, the YSWC, and unique topography in the study area (Weng et al. 1989; Ren and Zhan 2005).

Fig. 2
figure 2

Temperature–salinity plot of the investigated stations in the SYS of the four cruises. Dashed lines denote density contours (unit kg m−3)

Full size image
Fig. 3
figure 3

Horizontal distributions of temperature (°C), salinity (‰), Chl-a (μg L−1), and the five VHOC (pmol L−1) in surface seawater of the SYS in July 2008

Full size image
Fig. 4
figure 4

Horizontal distributions of temperature (°C), salinity (‰), Chl-a (μg L−1), and the five VHOC (pmol L−1) in surface seawater of the SYS in August 2008

Full size image
Fig. 5
figure 5

Horizontal distributions of temperature (°C), salinity (‰), Chl-a (μg L−1), and the five VHOC (pmol L−1) in surface seawater of the SYS in March 2009

Full size image
Fig. 6
figure 6

Horizontal distributions of temperature (°C), salinity (‰), Chl-a (μg L−1), and the five VHOC (pmol L−1) in surface seawater of the SYS in May 2009

Full size image

The mean (range) concentrations of CHCl3, C2HCl3, CHBrCl2, CHBr2Cl, and CHBr3 in the surface water of the SYS were 364.1 (87.7–802.2), 11.0 (1.2–23.6), 207.2 (106.5–366.8), 28.9 (1.8–114.0), and 11.5 (6.0–18.4) pmol L−1 in July 2008; 64.1 (32.2–116.8), 12.4 (6.8–19.3), 135.3 (67.9–248.6), 25.0 (11.4–62.9), and 6.1 (3.7–9.5) pmol L−1 in August 2008; 22.9 (13.9–86.1), 11.6 (3.7–33.2), 17.9 (10.6–26.2), 4.2 (2.2–11.6), and 13.5 (8.6–29.6) pmol L−1 in March 2009; and 37.2 (2.5–85.2), 12.3 (4.4–42.0), 21.5 (7.2–29.9), 4.6 (0.8–11.0), and 9.4 (6.8–12.1) pmol L−1 in May 2009, respectively. The mean concentrations of CHCl3 in the four cruises were higher than that of He et al. (2013a) (16.90 pmol L−1 in the East China Sea), but the averages of CHCl3 except for the first cruise were much lower than that of Roy (2010) (325.85 pmol L−1 in the coastal eastern Arabian Sea). The mean concentrations of C2HCl3 in the four cruises were slightly lower than that of He et al. (2013a) (16.27 pmol L−1 in the East China Sea), although evidently higher than that of Singh et al. (1983) (2.28 pmol L−1 in the eastern Pacific Ocean). The mean concentrations of CHBrCl2 in the present study were significantly higher than those of Class et al. (1986) (0.61 pmol L−1 in the South Atlantic Ocean) and Bravo-Linares et al. (2007) (0.98 pmol L−1 in the Irish Sea). The averages for CHBr2Cl were higher in the summer and lower in the spring than that of Bravo-Linares et al. (2007) (13.91 pmol L−1 in the Irish Sea). However, the concentration ranges of CHBr2Cl in the four cruises were in agreement with previously published results (Class et al. 1986; Carpenter et al. 2000). The mean concentrations of CHBr3 in this study were a little lower than that of He et al. (2013a) (17.11 pmol L−1 in the East China Sea) and also lower than those measured in other coastal waters (Zhou et al. 2005; Hughes et al. 2009; Liu et al. 2011) but higher than that of Bravo-Linares et al. (2007) (2.92 pmol L−1 in the Irish Sea). The discrepancy in the concentrations may be caused by the change in the source strengths of the different regions and seasons.

The elevated concentrations of CHCl3 were observed in the southernmost region and the northeast area in July and August 2008, respectively (Figs. 3d, 4d). Haloforms like CHCl3 are formed through water chlorination and sewage treatment (Bellar et al. 1974; Rook 1974). These processes possibly account for the elevated levels of CHCl3 in the southern study area of the first cruise, which resulted from the effluents from urban areas such as Yancheng and discharges from the Huaihe River and Sheyang River. Moreover, in the second cruise, elevated CHCl3 concentrations appeared in the northeast area near the YSCWM (Fig. 4d), suggesting the influence of water mass movement. In spring 2009, the highest CHCl3 concentrations in the two cruises were found in the southern coastal waters (Figs. 5d, 6d). This high concentration might be attributed to anthropogenic input from terrestrial pollutants (Blake et al. 1996; McCulloch et al. 1999).

The high concentrations of C2HCl3 occurred along the coast in summer 2008 and displayed a decreasing trend from inshore to offshore sites (Figs. 3e, 4e). In spring 2009, the elevated concentrations of C2HCl3 appeared in open sea in March (Fig. 5e) and in the central coastal waters in May (Fig. 6e). Chemical decomposition for C2HCl3 can be ruled out because the compound is very stable against hydrolysis in the aquatic environment, with half-lives of the order of hundred years (Mabey and Mill 1978; Fogelqvist 1984). Therefore, in summer 2008, its high concentrations along the coastal waters mirrored the impact of anthropogenic activities on the coastal environment (Class and Ballschmiter 1986; Wang et al. 1995; Blake et al. 1996). Moreover, decreased C2HCl3 concentrations with the distance from the coast should be ascribed to the influence of the YSCC and river runoff in the study area (Feng et al. 1999; Yuan et al. 2008). In March 2009, elevated C2HCl3 concentrations were observed in the open sea with higher temperature and salinity relative to the coastal waters (Fig. 5a, b), suggesting the influence of YSWC. In addition, previous studies have found relatively high abundance of phytoplankton populations in this region during the same season despite low levels of Chl-a (Huang et al. 2006). This phytoplankton abundance indicated that high C2HCl3 concentrations in the open sea were associated with phytoplankton production, in accordance with the previous observation that marine microalgae could produce C2HCl3 (Abrahamsson et al. 1995; Scarratt and Moore 1999). The elevated concentrations of C2HCl3 in the central coastal waters in May 2009 might be caused by the comprehensive action of YSCC, terrestrial inputs, and water mass movement.

The concentrations of CHBrCl2 appeared to be obviously high in the southern coastal region and decreased gradually offshore in summer 2008 (Figs. 3f, 4f) as influenced by the river runoff and Yancheng and Lianyungang coastal terrestrial inputs. The concentration of CHBrCl2 was significantly correlated with CHCl3 (R = 0.66, p < 0.001) in July 2008 (Table 2). This result supported the theory that the same influence factors controlled the two compound distributions in the first cruise. In March 2009, relatively low sea surface temperature and salinity showed homogeneous mixing distribution patterns (Fig. 5a, b) because early spring maintained winter hydrologic characteristics (Tang et al. 1999). The high concentrations of CHBrCl2 were observed in the northeast study area (Fig. 5f). This observation might be related to the production of benthic algae or the emission of sediments that were carried to the surface seawater by vertical mixing of water masses (He et al. 2013a). In addition, high CHBrCl2 concentrations in the region corresponded well with high levels of Chl-a (Fig. 5c, f), which might be attributed to the influence of YSWC. YSWC could stimulate certain phytoplankton species to flourish and release more CHBrCl2 (Nightingale et al. 1995). CHBrCl2 revealed a decreasing trend from inshore to offshore sites in May 2009, particularly in the southern study area (Fig. 6f), contrary to the distribution pattern of sea surface salinity (Fig. 6b). This result suggested that CHBrCl2 distribution in this season was affected by the anthropogenic input from the low-salinity river water along the coast.

Table 2 Correlation coefficients of linear regression analyses between concentrations of the five VHOC and other oceanographic parameters in the SYS during the four cruises
Full size table

CHBr2Cl concentrations in the surface seawater during July 2008 showed distinct spatial variation with a declining trend from the coast to the central region (Fig. 3g). A linear relationship between CHBrCl2 and CHBr2Cl (R = 0.45, p < 0.05) in this cruise, as previously reported by Chuck et al. (2005), was also found in the present study (Table 2). This result indicated that the two compounds might have a common source. Elevated CHBr2Cl concentrations appeared in the southern study area in August 2008 with higher levels of Chl-a (Fig. 4c, g). Two possible explanations for these high values could be deduced. On one hand, this region was clearly affected by terrestrial input. On the other hand, high CHBr2Cl concentrations are caused by river discharges (e.g., Sheyang River, Huaihe River, and Xinyi River) with abundant nutrients that stimulate local phytoplankton growth and emission of CHBr2Cl (Chen 2009; Peng et al. 2010). In spring 2009, high values of CHBr2Cl were observed in the Lianyungang and Yancheng coastal waters, whereas low values were measured in the open sea (Figs. 5g, 6g). The elevated CHBr2Cl in the coastal waters might be attributed to the influence of anthropogenic inputs.

The elevated concentrations were measured in the open waters for CHBr3 in the first cruise (Fig. 3h). However, as shown in Fig. 4h, the concentrations of CHBr3 presented a decreasing trend from inshore to offshore sites in the second cruise. High concentrations of CHBr3 appeared in the northeast study area in the third cruise and in the Qingdao and Yancheng coastal waters in the fourth cruise (Figs. 5h, 6h). These results suggested that various impact factors controlled the CHBr3 distributions in the different cruises. The elevated CHBr3 concentrations in the open sea with relatively high levels of Chl-a (Fig. 3c) during July 2008 might be related to the in situ biogenic sources. Locally elevated concentrations of CHBr3 in the open waters have generally been caused by the planktonic production associated with upwelling and, most likely, diatom abundance (Atlas et al. 1993; Baker et al. 1999). The upwelling current was most significant during this period and gave rise to high abundance of diatoms in this region (Zhang et al. 2002). Nevertheless, CHBr3 concentrations in the second cruise took on a decreasing pattern from inshore to offshore sites. This pattern could be attributed to the influence of Tianwan nuclear power plant effluent and Xinyi River input. As an example, larger Xinyi River flow was observed during the investigation period, with the highest value reaching 4,800 m3/s (Sun et al. 2010). In March 2009, elevated CHBr3 concentrations observed in the northeast study area corresponded well with the elevated Chl-a levels (Fig. 5c, h). These concentrations might be attributed to in situ biological production because of high phytoplankton biomass. However, poor correlation was noted between CHBr3 and Chl-a in the present study (Table 2) because CHBr3 production was specific to certain species of phytoplankton (Moore et al. 1996). CHBr3 concentrations in the surface seawater during May 2009 displayed a declining trend from the Qingdao and Yancheng coasts to the central region (Fig. 6h). This distribution pattern might be caused by anthropogenic inputs. For instance, a great number of disinfection facilities were located in coastal urban areas such as Qingdao and Yancheng cities. A wide range of byproducts, such as trihalomethanes from the disinfection of seawater, could be discharged into coastal waters (Zhou et al. 2005; Liu et al. 2011).

Relationships between VHOC and oceanographic parameters

A simple statistical approach of linear regression was used for the analysis of our field data in the four cruises to investigate the relationships between VHOC concentrations and oceanographic parameters. Table 2 lists the results of the linear regression analyses among the five VHOC, temperature, salinity, and Chl-a in surface seawater, which were performed using SPSS. 19.0 (SPSS Inc., IBM, USA). As shown in Table 2, no obvious correlation appeared between the VHOC and Chl-a concentrations in the surface seawater, which was in agreement with previous studies (Abrahamsson et al. 2004a; Chuck et al. 2005; He et al. 2013b). These results suggested that Chl-a content, which is a general indicator of phytoplankton biomass, is not necessarily associated with VHOC production from biological activities in the SYS. A weak correlation was noted between the Chl-a contents and the concentrations of CHBrCl2 and CHBr2Cl in July 2008 (Table 2). This correlation might imply the biological production of these gases during this period. However, results show that CHBr3 is produced by algae at much higher rates than other bromomethanes (Nightingale et al. 1995; Ekdahl et al. 1998; Quack and Wallace 2003). Relatively low concentrations of CHBr3 and lack of correlation with other trihalomethanes were observed in this cruise, which illustrated an anthropogenic rather than biological source of CHBrCl2 and CHBr2Cl in the SYS. In addition, enhanced nutrient levels observed in the study area during the wet season because of riverine inputs (e.g., Sheyang River, Huaihe River and Xinyi River) could stimulate algal bloom in the SYS (Chen 2009; Peng et al. 2010). Consequently, Chl-a is a better marker of river water inflows than an indicator of biological production of the two bromochloromethane. In the physical parameters, sea surface salinity was negatively correlated with C2HCl3 in July 2008 and CHBrCl2 in both August 2008 and May 2009 (Table 2), implying a direct impact of the river runoff on the distributions of C2HCl3 and CHBrCl2 in the study area. A negative correlation between sea surface temperature and CHBrCl2 concentrations was found in August 2008. This correlation further proved that the terrestrial inputs had greater influence on CHBrCl2 concentration and distribution in summer 2008. No relationship could be found between other VHOC and sea surface temperature or salinity because of the complicated interactions of various water masses in the study area.

In summary, the surface distribution patterns of the five VHOC were complicated and controlled by riverine runoffs (e.g., Sheyang River, Huaihe River, and Xinyi River), various current systems (e.g., the YSCC, YSWC, and YSCWM), anthropogenic input from terrestrial pollutants (e.g., water chlorination and sewage discharge), and biological processes such as phytoplankton production. Moreover, oceanographic environmental factors such as sea surface temperature and salinity also potentially have an important function in controlling these VHOC distributions.

Seasonal variations in the concentrations of VHOC in the study area

The average concentrations of the five VHOC and Chl-a in the SYS during the four cruises are shown in Fig. 7. The average concentration of CHCl3 was (364.1 ± 176.0) pmol L−1 in July 2008, but reduced to (64.1 ± 23.8) pmol L−1 in less than a month. In March and May 2009, the concentration of CHCl3 changed slightly, but was 50 % lower than that in August 2008. No obvious variation was observed in the concentrations of C2HCl3, unlike CHCl3, in the four cruises. The overall variations of concentrations of CHBrCl2 and CHBr2Cl were similar to CHCl3. On the average, CHBrCl2 and CHBr2Cl concentrations in summer were 8.7 and 6.1 times higher than those in spring, respectively (Fig. 7). The highest average concentration of CHBr3 was found in March 2009, which is nearly two times higher than the lowest mean concentration observed in August 2008 (Fig. 7). Consequently, the VHOC (except C2HCl3) displayed significant seasonal variations. In the present study, an obvious seasonal variation was also observed for Chl-a. The mean concentration of Chl-a in summer was two times higher than that in spring (Fig. 7). River runoff, which brought massive terrestrial contaminants into the SYS, was obviously higher in summer than in spring (Chen 2009). In addition, enhanced input of nutrients by terrestrial runoff during the summer could stimulate phytoplankton growth (Wang et al. 2001), which might produce more VHOC. The VHOC seasonal variations may partially be explained by differences in river runoff, various anthropogenic contaminant inputs, and biological processes in different seasons.

Fig. 7
figure 7

Average concentrations of five VHOC and Chl-a in four cruises in the SYS. Error bars represent ±mean standard deviations of the average concentrations for each cruise

Full size image

Our investigation results in March and May 2009 were compared with published data of VHOC concentrations in spring 2011 from the SYS and the East China Sea to better understand the temporal variations of VHOC concentrations in the SYS (He et al. 2013b). Result showed that the average concentrations of CHCl3, C2HCl3, CHBr2Cl, and CHBr3 in spring 2009 were clearly lower than those in spring 2011. This discrepancy could be attributed to regional and source differences. As an example, four measured VHOC concentrations in the SYS and the East China Sea during spring 2011 were greatly influenced by the Yangtze River effluent (He et al. 2013b). In addition, the concentration difference of these VHOC in the four cruises was possibly caused by regional differences because the four cruises did not cover exactly the same area (Fig. 1). However, data from coastal and shelf regions covered by the four cruises shared the trend toward higher VHOC concentrations in summer. For example, the average concentrations of CHCl3, CHBrCl2, and CHBr2Cl at stations A1 and B1 were 98.7, 145.8, and 26.7 pmol L−1 in summer, whereas their spring values were 11.9, 17.5, and 4.9 pmol L−1, respectively. These results clearly indicated that the significant differences in these VHOC concentrations between summer and spring in the present study were caused mainly by temporal change rather than regional difference. In summer rainy season, a large amount of terrestrial pollutants and sewage flowed into the SYS by flood (Li et al. 2009; Sun et al. 2010). The surface seawater concentrations of VHOC except C2HCl3 revealed obvious seasonal variations, indicating that the indirect effects of significantly high river runoff and terrestrial inputs dominated the VHOC concentrations. Moreover, a large part of the coastal water and YSWC also brought abundant nutrients, stimulating local phytoplankton growth and emission of these VHOC. The nutrients even altered phytoplankton species composition and community structure (Li et al. 2009).

Saturation anomalies and sea-to-air fluxes of three VHOC during the four cruises

According to the above-described equation, the saturation anomalies of VHOC were calculated in the study area. The saturation anomalies of CHCl3, C2HCl3, and CHBr3 were 5,798.2 (1,383.4–13,017.6), 378.8 (−45.8 to 1,033.7), and 30.8 (−28.8 to 126.7) % in July 2008; 1,007.3 (478.4–1,861.7), 498.5 (219.4–796.6), and −32.5 (−61.3 to 24.0) % in August 2008; 50.1 (−5.9 to 444.3), 108.2 (−27.9 to 555.5), and −51.3 (−68.0 to 0.6) % in March 2009; and 259.3 (−75.8 to 764.4), 220.3 (19.0–964.6), and −47.1 (−66.2 to 28.4) % in May 2009, respectively. Results from the four cruises showed that CHCl3 and C2HCl3 were generally supersaturated in the SYS, with the exception of only a few locations (e.g., stations A19, C5, and D2) where CHCl3 and C2HCl3 were undersaturated in the surface seawater. By contrast, CHBr3 was undersaturated at most locations of the SYS except for the first cruise. In the present study, the saturation anomalies of CHCl3 and C2HCl3 were higher in summer 2008 than in spring 2009. The results were consistent with the seasonal variation in the sea surface temperature, which were 24.9 (19.2–26.8) and 26.8 (23.1–30.8) °C in summer 2008, 8.2 (5.8–10.4) and 15.1 (13.3–16.2) °C in spring 2009 during the four cruises, respectively (Table 3). The highest saturation anomalies of CHCl3 and C2HCl3 appeared in July and August 2008, which corresponded to the greatest level of CHCl3 (Fig. 6) and the highest temperature in the same cruise, respectively. Moreover, the maximum negative saturation anomaly of CHBr3 was observed in March 2009, in agreement with the highest concentration of CHBr3 during the same period (Fig. 6). Therefore, the large temporal discrepancies in the saturation anomalies of CHCl3, C2HCl3, and CHBr3 were mainly dependent on both sea surface temperature and VHOC concentration gradient between water and air. The observed positive saturation anomalies indicated that the SYS was a source of CHCl3 and C2HCl3 to the atmosphere during the investigation period. However, CHBr3 was generally undersaturated in surface seawater at most locations. The negative saturation anomalies of CHBr3 in the current study were consistent with the result of Chuck et al. (2005), who reported that CHBr3 was usually undersaturated in a broad area of the temperate Atlantic Ocean. Moreover, negative saturation anomalies and fluxes of CHBr3 were observed in the Florida Straits and along the east coast of Florida (Liu et al. 2011).

Table 3 Sea-to-air fluxes of CHCl3, C2HCl3, and CHBr3 in the current study compared with other study areas
Full size table

The sea-to-air fluxes of CHCl3, C2HCl3, and CHBr3 across the air–sea interface in the SYS were calculated according to the previously described equations. As shown in Table 3, the average sea-to-air fluxes of the three compounds were within the previously reported range, except CHCl3 in the first cruise with a considerably high flux associated with the highest mean concentration (364.1 pmol L−1). The wind speeds were 5.5 (3.0–8.0), 8.5 (3.8–11.6), 7.7 (3.0–11.6), and 7.4 (1.4–17.6) m s−1 during the four cruises from July 2008 to May 2009, respectively. The fluxes of CHCl3 in the four cruises showed an obvious seasonal variation, with the summer values largely exceeding the spring measurements. The mean flux of CHCl3 in summer was higher by a factor of 8.6 than in spring, whereas the mean concentration in summer was 7.1 times that of spring observations. This result indicated that the higher concentrations of dissolved CHCl3 in summer were responsible for the larger fluxes. The sea-to-air fluxes of C2HCl3 appeared to be the highest in August 2008, although the concentrations of C2HCl3 were not the highest, which was attributed to the maximum wind speed and sea surface temperature in this cruise. Furthermore, the sea-to-air fluxes of CHBr3 appeared to be most negative in March 2009, which corresponded to the minimum sea surface temperature and maximum concentration of CHBr3 during the same period. Consequently, the large temporal variabilities in the sea-to-air fluxes of VHOC were controlled not only by their surface concentrations, but also by wind speed and sea surface temperature.

Based on our data (Table 3), the average flux differences for CHCl3, C2HCl3, and CHBr3 between the two cruises of 2008 were –266.7, 19.0, and –10.3 nmol m2 day−1, respectively. The data above show that C2HCl3 fluxes increased, whereas CHCl3 and CHBr3 fluxes decreased in less than 1 month after the bloom of the green alga. This difference indicated that during the post-bloom period, more C2HCl3 was released into the atmosphere. The mean wind speed and seawater temperature in the second cruise were 3.0 m s−1 and 1.9 °C higher than those in the first cruise, respectively (Table 3). Although the surface concentrations of C2HCl3 decreased in the second cruise, increasing wind speed and temperature led to the higher fluxes.

The major uncertainty in the assessment of sea-to-air fluxes is in the calculation of gas transfer velocities, which depends on measured wind speed and sea surface temperature. In this work, the instantaneous wind speeds were used to calculate the gas transfer coefficient for the sea-to-air fluxes at each station. Therefore, the preliminary calculation fluxes of the three compounds should be regarded as instantaneous and approximate assessment. In addition, the introduction of these atmospheric mixing ratios also led to larger uncertainties in the sea-to-air flux estimates. Further study, including simultaneous measurement of concentrations of these VHOC in both aqueous and gaseous phases, is necessary to accurately estimate the sea-to air fluxes of these VHOC in the study area.

Conclusions

This study provided preliminary understanding of the spatio-temporal variations of five VHOC and their sea-to-air fluxes in the SYS. The distributions of these VHOC in the study area were complicated and influenced by river runoff, anthropogenic inputs, various current systems, and biological processes. Environmental factors (e.g., Chl-a, sea surface temperature, and salinity) potentially had a significant function in controlling VHOC distributions. Salinity appeared to be negatively correlated with C2HCl3 in July 2008 and CHBrCl2 in August 2008 and May 2009, whereas temperature was found to be negatively correlated with CHBrCl2 concentrations in August 2008. Fully understanding the influence of these oceanographic parameters on VHOC distributions is necessary in further studies. All VHOC except C2HCl3 revealed apparent seasonal variations in the study area, which was caused mainly by temporal change rather than regional difference. The calculation of saturation anomalies showed that CHCl3 and C2HCl3 were supersaturated, whereas CHBr3 was generally undersaturated in surface seawater at most locations during the study period. The calculation of sea-to-air fluxes indicated that the SYS was a significant source of CHCl3 and C2HCl3 to the atmosphere.