Marine Chemistry 158 (2014) 5968

Contents lists available at ScienceDirect

Marine Chemistry
journal homepage: www.elsevier.com/locate/marchem

Mercury and methylmercury ux estimation and sediment distribution in an industrialized urban bay
Eunhee Kim a,b, Seam Noh a, Yong-gu Lee a, Sampa Rani Kundu b, Byeong-Gweon Lee b, Kihong Park a, Seunghee Han a,
a b

School of Environmental Science and Engineering, Gwangju Institute of Science and Technology (GIST), Gwangju, 500-712, Republic of Korea Department of Oceanography, Chonnam National University, Gwangju, 500-757, Republic of Korea

a r t i c l e

i n f o

a b s t r a c t
Our objectives of this study were to estimate mercury (Hg) ux to Masan Bay, an industrialized urban bay in Korea, and to examine important factors inuencing the Hg distribution and transformation in Masan Bay sediment. As the rst attempt to estimate Hg ux for Masan Bay, we found out that stream water discharge was the prime source of Hg, contributing 76% of the total input. Estimating the methyl Hg (MeHg) ux showed that groundwater discharge is the major source, contributing 53% of the total input. The large MeHg input through groundwater discharge emphasizes the importance of the in-situ sediment production of MeHg. The fraction of MeHg over total Hg (%MeHg) in the bay sediment, ranging from 0.10% to 1.5%, showed a strong negative correlation with the sediment organic matter content (2.814% as a loss on ignition). However, when different urban and industrialized estuarine sediments were compared, sediment organic matter either promoted or constrained %MeHg in sediment, suggesting that the role of sediment organic matter for the net Hg methylation varies, relying on the biogeochemical conditions of estuarine sediments. Despite the large variations in Hg and MeHg concentrations among different urban and industrialized estuarine sediments, relatively small variations were determined for the %MeHg, i.e., 0.52% on average. 2013 Elsevier B.V. All rights reserved.

Article history: Received 15 March 2013 Received in revised form 17 October 2013 Accepted 5 November 2013 Available online 19 November 2013 Keywords: Mercury Methylmercury Methylation Organic matter Flux Estuary

1. Introduction Mercury (Hg) is a persistent global pollutant that undergoes various biogeochemical processes (e.g., oxidation, reduction, methylation, and demethylation) in the aquatic environment. A key biogeochemical process is Hg methylation, which forms methylmercury (MeHg) by a diverse range of microorganisms including sulfate-reducing bacteria, iron-reducing bacteria, methanogens, and Firmicutes (Gilmour et al., 1992; Benoit et al., 1999; Kerin et al., 2006; Han et al., 2010; Gilmour et al., in press). Recently, the gene orthologs of hgcAB were found to be present in conrmed methylators, suggesting that possessing hgcAB may predict Hg methylating capability of microorganisms (Gilmour et al., 2013; Parks et al., 2013). MeHg, a major form of organic Hg, is more toxic and biomagnies more efciently than inorganic Hg within aquatic food chains. Coastal and estuarine sediments are major MeHg repositories and may act as either a sink or a potential source of MeHg for the overlying water (Benoit et al., 1998; Mason et al., 1999; Sunderland et al., 2004; Kim et al., 2006; Hollweg et al., 2009). Previous studies showed that the Hg methylation potential in coastal/marine sediments was closely related to sediment organic matter (OM) and sulde concentrations, which inuences Hg partitioning

Corresponding author. Tel.: +82 62 715 2438; fax: +82 62 715 2434. E-mail address: shan@gist.ac.kr (S. Han). 0304-4203/$ see front matter 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.marchem.2013.11.004

and bioavailability to methylating bacteria (Sunderland et al., 2004; 2006; Han et al., 2007, 2011; Hammerschmidt et al., 2008; Hollweg et al., 2009; Hsu-Kim et al., 2013). Sediment OM content appeared to be more important in controlling Hg methylation potential at low-sulde conditions (b 10 M) than at high-sulde conditions (Sunderland et al., 2006; Hammerschmidt et al., 2008; Hollweg et al., 2009). In highsulde sediment (N 10 M), MeHg production was associated with the dissolved Hg speciation (Benoit et al., 2001, 2003; Drott et al., 2007; Hollweg et al., 2009; Han et al., 2011). Cell culture, chemical equilibrium models, and eld data have suggested that high levels of dissolved sulde decrease Hg bioavailability to Hg-methylating bacteria due to the formation of hydrophilic Hg-sulde species (Benoit et al., 2001; Drott et al., 2007). A recent review by Hsu-Kim et al. (2013) emphasizes the role of Hg-sulde nanoparticles in Hg methylation, suggesting that Hg methylation cannot be sufciently explained with conventional chemical equilibrium models. Masan Bay, a semi-enclosed bay located on the southeastern coast of Korea (Fig. 1), is surrounded by highly populated cities (Masan, Changwon, and Jinhae) and large industrial complexes, including petrochemical, electrical, plastic, and metal industries (Moon et al., 2009). Large inputs from these industrial and municipal efuents cause environmental problems due to high nutrient loadings as well as organic and metal contamination (Hyun et al., 2007; Moon et al., 2008; Hong et al., 2010; Jang et al., 2011). Additionally, sluggish water circulation, ascribed to weak currents (23 cm s 1), results in high accumulation

60

E. Kim et al. / Marine Chemistry 158 (2014) 5968

understand the environmental factors that govern Hg methylation potential in urban bays. 2. Materials and methods 2.1. Sample collection 2.1.1. Surface seawater and sediment collection Surface water samples were collected from eight stations (Fig. 1) on ve occasions in 2009 and 2010 (March, May, and August in 2009 and January and May in 2010). Water samples were collected using a Go-Flo bottle (General Oceanics, USA) from about 0.5 m below the water surface. The Go-Flo bottle was lled with trace-metal grade 12 M hydrochloric acid (5% v/v) overnight, rinsed with high purity (18.2 Mcm) Milli-Q water several times, and then kept in a plastic bag before use. Water samples from the Go-Flo bottles were transferred to Teon bottles for the THg and MeHg analyses and polypropylene (PP) bottles to determine the total suspended solids (TSS) and particulate organic matter (POM). Ancillary parameters (temperature, dissolved oxygen [DO], and salinity) were measured with a handheld multimeter (YSI Incorporated, USA) on-site. The pH was measured either on-site or immediately after returning to the laboratory using a pH meter (Thermo Scientic) after calibration. All water samples transferred to the Teon and PP bottles were kept in coolers with ice until they were transported to the laboratory. Samples were acidied with trace-metal grade 12 M hydrochloric acid (0.5% v/v) and stored at 4 C until the THg and MeHg analyses. The unltered water samples collected for THg in August (2009) and May (2010) were analyzed for MeHg as well. To determine the TSS and POM, the water samples were ltered through pre-weighed 0.4 m glass ber lters (Advantec), and the lters were dried at 60 C overnight. POM was calculated from the difference between the dry weight and loss on ignition (at 550 C for 4 h, Kim et al., 2004). Surface sediments (about 15 cm) were collected using a Ponar grab sampler, on ve occasions previously mentioned (Fig. 1). Samples were mixed homogeneously on board and transferred to acid-cleaned polypropylene containers. Sediments were stored frozen until the THg and MeHg analyses were conducted. In addition to the THg and MeHg analyses, the sediments collected in May 2010 were analyzed for AVS. The Hg benthic ux was calculated in sediment samples collected in acrylic cores with a 12-cm diameter from inner and outer bay sites (Fig. 1) in May 2011. The cores were placed inside a nitrogen-lled glove box before the overlying water was removed. After the overlying water was removed, the cores were sectioned to 2 cm down the core. Pore waters were extracted from sediments by centrifugation, and the supernatant was ltered through 0.45-m syringe lters in a nitrogen-lled glove box (Mason et al., 1998). Bottom seawater samples were collected, together with sediment cores from the same sites. The ltered pore waters and seawaters were acidied to a 0.4% (v/v) nal concentration with 12 M hydrochloric acid (trace-metal grade) and stored at 4 C in a dark environment until the analysis was conducted. All analyses of the water for THg and MeHg were performed within a month after the sample was collected. 2.1.2. Collection of inowing water In order to estimate the wet atmospheric deposition of Hg, four rainwater samples were collected in May and August 2011 from Gwangju, located in southwestern Korea (about 200 km from Masan Bay). The rainwater was collected using acid-cleaned glass funnels (15 cm diameter) connected to a 2 L acid-cleaned Teon bottle. The rainwater samples were immediately preserved after collection by adding trace-metal grade 12 M hydrochloric acid (0.4% v/v). Before each sampling, the Teflon bottles and glass funnels were soaked in a hot 4 M hydrochloric acid bath (trace-metal grade) overnight and rinsed with Milli-Q water. The Teon bottles were lled with trace-metal grade 12 M hydrochloric acid (0.4% v/v) and double-bagged. The glass funnels were kept in a plastic bag until use.

Fig. 1. A map of the study area and sampling locations.

of contaminants especially in the upper bay (Hyun et al., 2007; Hong et al., 2009). A wastewater treatment plant (WWTP) was constructed to mitigate the environmental problems in Masan Bay and has been in operation since 1994 (Fig. 1; Moon et al., 2009). The WWTP treats 2.6 105 tons of wastewater; mostly domestic, daily, and efuents are discharged through an underwater outfall pipe (Moon et al., 2009, Fig. 1). A recent study showed a fairly high concentrations of total Hg (THg) and MeHg in some benthic sh and invertebrate species collected from Masan Bay (Kim et al., 2012). Submarine groundwater discharge (SGD) ranged from 4.8 106 to 5.7 106 m3 day 1 in Masan Bay, or approximately 540% to 840% of the surface water discharge (Lee et al., 2009). Owing to this high SGD rate, the inorganic phosphorus and silicate ux was approximately 23 fold higher through the SGD than through river water discharge (Lee et al., 2009). Previous studies have shown that Hg ux by SGD in several estuarine/coastal environments is significant, exceeding river discharge and atmospheric inputs (Bone et al., 2007; Laurier et al., 2007; Black et al., 2009). Lee et al. (2011) found that the THg and MeHg ux through SGD for Hwasun and Bangdu Bay located on Jeju Island, Korea, was substantially higher than that through atmospheric deposition and sediment diffusion. Given this, Hg inputs via SGD to Masan Bay, which may be larger than industrial/river discharge and atmospheric deposition, can be important. To date, however, no information is available for estimating the distribution and sources of THg and MeHg in Masan Bay. For these reasons, the objectives of this study were as follows: 1) to examine the spatial and temporal distribution of THg and MeHg in Masan Bay water and sediment, 2) to identify major sources of THg and MeHg in the bay water, and 3) to investigate correlations between biogeochemical factors in sediments, i.e., THg, acid volatile sulde (AVS) and OM, and the sediment %MeHg (the fraction of MeHg over THg) to

E. Kim et al. / Marine Chemistry 158 (2014) 5968

61

Surface freshwater inputs were estimated by collecting unltered water from two inowing streams located in urban (Changwon Stream) and rural (Usan Stream) areas in May and August 2011 (Fig. 1), using a peristaltic pump and Teon tubing connected to 500 mL pre-acidcleaned Teon bottles. Unltered WWTP efuent samples were collected in May and August 2011 (Fig. 1), as described above. The stream water and wastewater efuent samples were immediately preserved by adding trace-metal grade 12 M hydrochloric acid (0.4% v/v) and stored at 4 C in a dark environment until analysis. The Hg in SGD was estimated by collecting groundwater samples from shallow wells in urban and rural areas in May and August 2011 (Fig. 1). Groundwater samples were collected in active seepage runoff areas in the intertidal zone. After the suspended debris and particles in shallow wells were discarded, ltered groundwater samples were collected using a peristaltic pump and Teon tubing connected to a Millipore groundwater lter capsule (0.45-m pore size, hydrophilic polyethersulfone membrane) in 500-mL acid-cleaned Teon bottles. The groundwater samples were immediately preserved after collection by adding trace-metal grade 12 M hydrochloric acid(0.4% v/v) and stored at 4 C. All analyses of the water samples for THg and MeHg were performed within a month after the sample was collected. 2.2. Sample analysis 2.2.1. THg in water and sediment The THg in water samples was analyzed by adding bromine monochloride (0.5% v/v) to the acidied samples at least 12 h before analysis for oxidation, followed by hydroxylamine hydrochloride to destroy free halogens (Method 1631, EPA, 1998). The samples were then further reduced with stannous chloride to convert Hg(II) to Hg(0), which was separated from the solution by nitrogen purging collected onto a gold trap. The Hg in the gold trap was thermally desorbed with argon owing and quantied with cold-vapor atomic uorescence spectrometry (CVAFS, EPA, 1998; Bloom and Fitzgerald, 1988). Analysis of duplicate water samples yielded a mean relative percent difference (RPD) of 16 6.7% (n = 4). The detection limits, based on 3 standard deviations of bubbler blanks (n = 14), were 0.20 pM. Spike recoveries averaged 100 11% (n = 9). Surface sediment samples were digested in an oven at 60 C overnight in 5.0 mL of a mixture of concentrated sulfuric acid/nitric acid (3:7 v/v). Similar to the water analysis, the digestate was further oxidized by bromine monochloride, and excess oxidant was neutralized with hydroxylamine hydrochloride before the analysis (Method 1631, EPA, 1998; Bloom and Crecelius, 1983). The samples were quantied using CVAFS, as described above. Analysis of the duplicate samples yielded a mean RPD of 7.0 9.8% (n = 10). The detection limits were 0.25 pmol g 1, based on 3 standard deviations of the digestion blank measurements (n = 6). Recovery of certied reference material (IAEA433; 0.84 0.09 nmol g 1 and MESS-3; 0.46 0.05 nmol g 1) averaged 111 5.2% (n = 4) and 90 2.2% (n = 3), respectively. All sediment data presented here are expressed on a dry weight basis. 2.2.2. MeHg in water and sediment Water and sediment samples were distilled with 1.0 mL of tracemetal grade 9 M sulfuric acid and 0.5 mL of 20% (w/w) trace-metal grade potassium chloride (Horvat et al., 1993). After distillation, the

samples were kept at 4 C and analyzed within 2 days. For analysis, a 1.0% (w/v) sodium tetraethylborate (0.1 mL) solution was added to the distillate in bubblers to convert the nonvolatile MeHg to gaseous methylethylmercury (Bloom, 1989). The volatile adduct was then purged from the solution and recollected on a solid phase column (Tenax) at room temperature. The methylethylmercury was thermally desorbed from the column, separated on a gas chromatographic (GC) column, and converted to elemental Hg in a pyrolytic column at the GC outlet. Samples were quantied with CVAFS. Analysis of duplicate samples typically gave a mean RPD of 9.4 8.0% (n = 5) for sediment. The detection limit, based on 3 standard deviations of the distillation blank measurements, was 7.4 fM (n = 10) for water and 0.018 pmol g 1 (n = 9) for sediments. Recovery of certied reference material, IAEA-433 (0.85 0.35 pmol g 1), averaged 86 12% (n = 7). Spike recoveries yielded 101 15% (n = 12) and 98 16% (n = 10) for water and sediment, respectively. All sediment data are reported on a dry weight basis. 2.2.3. AVS and sediment OM content Protocols for AVS analysis have been described elsewhere (Kim et al., 2006). Briey, an aliquot sample was weighed and added to a tared three-neck ask. The asks were immediately purged with ultra high purity nitrogen gas to minimize exposure to oxygen. After 6.0 M deoxygenated hydrochloric acid was added, the samples were distilled for 1.52 h at room temperature with nitrogen gas ow. Sulde volatilized from distillation was trapped in sulde antioxidant buffer (SAOB) solution and measured using an ion specic sulde electrode (Baumann, 1974; Allen et al., 1991; EPA, 1996). The detection limit with sulde electrode was 0.1 M. A calibration curve with an r2 of at least 0.99 was achieved, and the spike recovery averaged 103% (n = 2). The percentage of OM content in the sediment samples was determined as loss on ignition (LOI) to 550 C for 4 h (Benoit et al., 1998; Kim et al., 2006). 2.2.4. Statistical analysis The statistical signicance of the spatial and temporal variations in the Hg concentrations was tested using one-way analysis of variance (ANOVA, SPSS, 2003). When the normality and equal variances did not pass for ANOVA, the data were log-transformed. For multiple comparisons of means, Scheff post hoc test was used (SPSS, 2003). The Pearson productmoment correlation coefcients (r) were used to examine a simple linear correlation between two variables (SPSS, 2003). All the statistical results were reported as signicant at a level of p b 0.05, and SPSS version 12 was used for all statistical analyses. 3. Results and discussion 3.1. Surface water characteristics Table 1 shows the surface water characteristics of Masan Bay. The overall average water depth for the study area was 16 9.0 m (7.240 m, data not shown). Average salinity in August, a typical rainy season in Korea, was lower than the rest of the sampling occasions (Table 1). A similar result was found in previous studies for Masan Bay (Jang et al., 2011) and Gwangyang Bay (Jang and Han, 2011) in Korea. The temperature uctuation was relatively large among seasons, with the highest temperatures in August (25 0.3 C) and the lowest

Table 1 Characteristics of Masan Bay surface water. Data are shown as the average standard deviation. Temp (C) Mar 2009 May 2009 Aug 2009 Jan 2009 May 2009 ND: not determined. 11 17 25 5.5 17 0.2 0.3 0.3 1.1 0.7 pH ND 8.4 0.04 8.4 0.15 ND 8.2 0.08 Salinity (ppt) 32 32 29 31 30 0.4 1.0 0.4 0.5 0.9 DO (mg L1) ND 7.4 8.1 9.2 8.5 1.4 1.8 2.0 0.7 TSS (mg L 1) 12 17 53 16 50 9.8 8.3 10 11 13 POM (mg L 1) 5.7 4.1 15 5.5 14 7.0 3.0 3.7 1.7 3.6

62

E. Kim et al. / Marine Chemistry 158 (2014) 5968

temperatures in January (5.5 1.1 C). The average DO in surface water remained relatively constant throughout all sampling events. As shown in Table 1, the TSS concentration was higher in August 2009 and May 2010 than for the other sampling occasions. A similar pattern was observed for POM, which was positively and signicantly (r = 0.92, p b 0.05, n = 27) correlated with the TSS. TSS was negatively and signicantly (r = 0.80, p b 0.05, n = 27; Fig. 2) correlated with salinity, which suggests that the high TSS in August 2009 and May 2010 is attributable to increased terrestrial inputs and/or sediment suspension during rainfall events. 3.2. Hg concentrations in surface water Unltered THg concentrations showed signicant temporal but not spatial variation (Fig. 2a). Here, the data for THg and MeHg for August

6 5 4

3 2 1 0 1 2 3 4 5 6 7 8

Station

2009 were obtained from Kim et al. (2012). The average THg concentration in August was 3.7 1.0 pM, which is signicantly (p b 0.05) higher than the concentrations in March (2.2 0.48 pM) and January (1.8 0.51 pM). The overall average THg in unltered water was 2.9 1.0 pM, which was in good agreement with other estuaries in the southern coast of Korea (Jang and Han, 2011). THg in unltered water was positively and signicantly (r = 0.63, p b 0.05, n = 27; Fig. 2b) correlated with the TSS, as was found in other estuarine coastal waters (Conaway et al., 2003; Balcom et al., 2008; Jang and Han, 2011; Ci et al., 2011). This, together with the negative correlation between the TSS and salinity (Fig. 2c), indicates that the THg level in the surface seawater increases along with freshwater discharge via river input of THg and/or sediment resuspension in the shallow estuarine zone. MeHg in unltered water was measured in August 2009 and May 2010 (Fig. 3). The data for August 2009 were obtained from Kim et al. (2012). The overall average MeHg (%MeHg as a fraction of MeHg over THg) in the surface water was 0.17 0.13 pM (4.6 3.7%). The MeHg levels for Masan Bay were within the same order of magnitude but somewhat lower than those in the San Francisco Bay, USA (0.45 0.05 pM; Conaway et al., 2003) and the NY/NJ Harbor, USA (0.52 0.09 pM; Balcom et al., 2008). The average MeHg concentrations were slightly higher in August than May, averaging 0.20 0.13 and 0.10 0.13 pM, respectively (Fig. 3a). The %MeHg in August (5.6 3.7% with a range of 2.0 to 14%) was approximately two times higher than that in May (2.6 3.2% with a range of 0.777.3%; Fig. 3b). It could be the result of either enhanced microbial activity at higher temperature or planktonic scavenging and consequent remobilization of Hg, as was reported by Luengen and Flegal (2009). The same report showed that maximum concentrations of particulate and dissolved MeHg occurred when the bloom was decayed, likely the consequence of enhanced sediment MeHg production through creation of anoxic conditions (Luengen and Flegal, 2009). Unlike THg, no

THg (pM)

6 5 4

0.5

THg (pM)

0.4

MeHg (pM)
0 20 40 60 80

3 2 1

0.3

0.2

0.1 0

TSS (mg/L)

0.0 1 2 3 4 5 6 7 8

80 70 60

Station

16 14 12

TSS (mg/L)

50

30 20 10 0 26 28 30 32 34
Mar 2009 May 2009 Aug 2009 Jan 2010 May 2010

%MeHg Salinity (ppt)

40

10 8 6 4 2 0 1 2 3 4 5 6 7 8

Fig. 2. (a) Unltered THg concentrations in surface water; (b) a correlation of THg with TSS; and (c) a correlation of TSS with salinity. Missing data in (a) indicate no sample collection.

Station
Fig. 3. (a) Unltered MeHg concentrations and (b) %MeHg in surface water.

E. Kim et al. / Marine Chemistry 158 (2014) 5968

63

noteworthy correlations were observed between MeHg and TSS, POM, or salinity, suggesting that freshwater inow did not affect the surface water MeHg distributions. 3.3. Hg ux 3.3.1. Atmospheric deposition The ux estimates are presented in Tables 2 and 3. The wet atmospheric deposition uxes of THg and MeHg were determined by a similar method used in previous studies (Lee et al., 2011; Rahman et al., 2013). We multiplied the 18-year average rainfall in Korea between 1992 and 2010 (1.56 0.66 m yr 1, Korea Meteorological Administration) by the bay area (80 km2) as well as the mean THg and MeHg concentrations in rainwater collected in Gwangju (Table 2). As seen in Table 3, the THg and MeHg inputs through wet atmospheric deposition were estimated to be 4.0 2.2 mol yr 1 (or 50 28 nmol m 2 yr 1) and 0.016 0.008 mol yr 1 (or 0.20 0.10 nmol m 2 yr 1), respectively. For dry deposition of THg, we used the literature value of 0.040 0.013 mol m 2 yr 1 determined for THg in Japan (Sakata and Marumoto, 2005) because no Hg dry deposition data are available for Korea. The THg input through dry atmospheric deposition was 3.2 1.0 mol yr 1 or 40 13 nmol m 2 yr 1 (Table 3). MeHg ux through dry deposition was assumed to be negligible. A recent study on the Hg ux in California coast showed that depositions of THg and MeHg via fog water accounted signicant portions of THg (742%) and MeHg (6199%) atmospheric deposition (Weiss-Penzias et al., 2012). However, the frequency of fog occurrence, mainly depends on the dew point temperature and sea surface temperature, is relatively low in Korean coastal zones, particularly in the eastern and southern coast (Cho et al., 2000). In the southern coastal zone, the mean frequency of sea fog occurrence was 4 days in July, the peak season, and less than 1 day from September to March, according to the 10-yr average data (Cho et al., 2000). Masan Bay Hg input via fog water deposition might be insignicant as compared to the wet and dry deposition. Nonetheless, THg and MeHg depositions should be measured directly in the Masan Bay in the future to reduce uncertainties in the estimates for atmospheric THg and MeHg ux. 3.3.2. Sediment diffusion The diffusive uxes of THg and MeHg were estimated using the mean THg and MeHg concentration gradients over the pore water and seawater, at the sedimentwater interface, using the equation J = Ds (dC/dz), where J = diffusive ux in nmol cm 2 s 1, = porosity (dimensionless), Ds = sediment diffusion coefcient in cm2 s 1, dC = difference in the concentration of THg and MeHg in the sediment pore water and near-bottom water in nmol g 1, and dz = depth difference between the sediment pore water and the near-bottom water in cm (Gill et al., 1999; Hammerschmidt et al., 2004; Hollweg et al., 2009; Duong and Han, 2011). The diffusion coefcient, Ds, was dened assuming sediment porosity b 70%, such that Ds = D0/2, where is sediment tortuosity (dimensionless), and D0 is the diffusion coefcient of Hg and MeHg in seawater of 5.0 10 6 cm2 s 1 and 1.2 10 5 cm2 s 1,

respectively (Gobeil and Cossa, 1993; Mason et al., 1993). For the current study, sediment porosity () was calculated using the following relationship: = (Mw/w)/[(Ms/s) + (Mw/w)], where Mw is the water weight, w is the water density (1.025 g cm 3), Ms is the dry sediment weight, and s is the sediment density (2.6 g cm 3) (Ram et al., 2009). The tortuosity of the sediment, 2 = 1.7, was derived by applying the relationship between porosity and tortuosity, = 1 ln(2) (Boudreau, 1996). The differences in the concentration of THg in the sediment pore water and near-bottom water (depth difference = 1 cm) were 6.74 pM for the inner bay (ST1 and ST2) and 12 pM for the outer bay (ST3ST5). Based on this concentration gradient, the nal benthic ux of THg was estimated to be 0.29 0.06 mol yr 1 using 20 km2 of the inner bay area and 60 km2 of the outer bay area. An area normalized ux was estimated to be 4.1 1.4 nmol m 2 yr 1. The differences in the MeHg concentration in the sediment pore water and the nearbottom water (depth difference = 1 cm) were 0.20 pM for the inner bay and 0.25 pM for the outer bay. The nal benthic ux of MeHg was 0.035 0.017 mol yr 1, with an area normalized ux of 0.39 0.14 nmol m 2 yr 1 (Table 2). THg and MeHg mobilization from sediment might be underestimated because only diffusive ux was considered in our calculation. Hammerschmidt and Fitzgerald (2008) showed that MeHg mobilization was mostly diffusional when the water column DO was less than 80% saturation in coastal sediments. In the same study, MeHg mobilization was enhanced when the DO was near saturation, which might be linked to increased activity of macrofauna that irrigate sediments (Hammerschmidt and Fitzgerald, 2008). Low DO conditions in bottom water are found every summer in Masan Bay (Lim et al., 2007). Given that the bottom DO is more than 80% saturation from October to April and that the experimental ratio of the THg (or MeHg) efux to diffusional ux is approximately 2 at saturated DO conditions as was shown by Hammerschmidt and Fitzgerald (2008), the overall THg (or MeHg) efux is calculated to be 160% of the current value (i.e., 0.47 mol yr 1 for THg and 0.056 mol yr 1 for MeHg). The underestimation seems to be sufciently small not to alter the order of the source contribution in Table 3. 3.3.3. Groundwater ux The magnitudes of the THg and MeHg input through SGD were calculated by multiplying the mean groundwater ux by the mean concentrations of THg and MeHg in coastal groundwater (Table 2). We used the SGD ux of (1.9 0.23) 109 m3 yr 1, which included the recirculating seawater and the terrestrially derived fresh groundwater (Lee et al., 2009). Given the above, THg and MeHg inputs through SGD were estimated to be 5.2 4.8 mol yr 1 (or 65 60 nmol m 2 yr 1) and 0.21 0.14 mol yr 1 (or 2.6 1.8 nmol m 2 yr 1) (Table 3), respectively. Our SGD ux of THg for Masan Bay was orders of magnitude higher than that for Hwasun and Bangdu Bays (0.23 0.14 and 0.23 0.20 mol yr 1, respectively) on Jeju Island (Lee et al., 2011). In contrast, the SGD THg ux for Masan Bay was considerably lower than that for Hampyeong Bay (18 12 mol yr 1), located in the southwestern part of Korea (Rahman et al., 2013). The difference in the SGD THg uxes for different estuarine systems was mainly driven by the SGD ux, not by the THg concentrations in SGD. 3.3.4. Stream water and wastewater efuent discharge The magnitudes of the THg and MeHg input through rural stream discharge were calculated by multiplying the rural stream ux (3.8 107 m3 yr 1 with a large seasonal variation, Cho and Chae, 1998) by the mean concentrations of THg and MeHg in unltered rural stream water (Table 2). Similarly, the magnitudes of the THg and MeHg input through urban stream discharge were calculated by multiplying the urban stream ux (1.1 108 m3 yr 1, Cho and Chae, 1998) by the mean concentrations of THg and MeHg in unltered urban stream water (Table 2). The large variations of THg concentration in river water seem to rely on the river ow rate (David et al., 2009; Kim

Table 2 THg and MeHg concentrations in rainwater and inowing waters to Masan Bay. Sample type Rain water Stream water THg (pM) MeHg (pM) 0.13 0.07 0.58 0.33 0.37 0.17 1.5 1.3 0.11 0.07 0.06 0.07 0.24 0.38

32 17 Rural stream (Usan Stream) 24 26 Urban stream (Changwon Stream) 360 492 Wastewater efuent 133 4.6 Groundwater 2.7 2.5 Bottom water Inner bay 0.82 Outer bay 1.33 Sediment pore water Inner bay 8.83 Outer bay 6.18 Values indicate mean standard deviation.

64 Table 3 THg and MeHg mass ux calculations in Masan Bay, Korea. Category THg Flux (mol yr1) Wet deposition Dry deposition Sediment diffusion SGD Stream WWTP Total input 4.0 3.2 0.29 5.2 59 6.1 78 2.2 1.0 0.060 4.8 40 0.2 40 Contribution (%) 5.2 4.1 0.37 6.7 76 7.8 100

E. Kim et al. / Marine Chemistry 158 (2014) 5968

MeHg Unit area ux (nmol m2 yr 1) 50 40 4.1 65 590 76 825 28 13 1.4 60 500 3 504 Flux (mol yr1) 0.016 0.035 0.21 0.064 0.067 0.39 0.008 0.017 0.14 0.007 0.059 0.15 Contribution (%) 4.1 9 53 16 17 100 Unit area ux (nmol m2 yr 1) 0.20 0.39 2.6 0.80 0.84 4.8 0.10 0.14 1.8 0.09 0.74 2.0

Values indicate mean standard deviation.

et al., 2011). In San Francisco Bay, THg concentrations in downstream SacramentoSan Joaquin River was approximately ve times higher during the high ow season, likely due to enhancement of contaminated sediment transport from the Cost Range Hg mines (David et al., 2009). Indeed, a temporal trend of THg in Usan (5.8 pM in May 2011 and 42 pM in August 2011) and Changwon (12 pM in May 2011 and 710 pM in August 2011) streams followed the trend of ow rate (Table 1). THg and MeHg discharge from streams could largely increase in the monsoon season (e.g., July and August) every year, therefore, we estimated the Hg river ux separately for dry and wet seasons. Here, May Hg data were used for dry ux estimation and August Hg data for wet ux estimation. Dry and wet river discharge rates were collected individually from Cho and Chae (1998). The overall Hg input through stream discharges calculated for the rural stream was 1.3 mol yr 1 (or 16 nmol m 2 yr 1) for THg and 0.027 mol yr 1 (or 0.33 nmol m 2 yr 1) for MeHg, and that for the urban stream was 57.8 mol yr 1 (or 723 nmol m 2 yr 1) for THg and 0.037 mol yr 1 (or 0.46 nmol m 2 yr 1) for MeHg. The magnitude of the THg and MeHg input through treated WWTP discharge was calculated by multiplying the mean wastewater ux (4.5 107 m3 yr 1, KEI Report, 1996) by the mean concentrations of

THg and MeHg in wastewater (Table 2). As a result, the THg and MeHg inputs through WWTP were 6.1 0.21 mol yr 1 (or 76 2.6 nmol m 2 yr 1 ) and 0.067 0.059 mol yr 1 (or 0.84 0.74 nmol m 2 yr 1), respectively (Table 3). 3.3.5. Relative contributions of Hg sources in Masan Bay Although more detailed information is needed for a rened mass balance calculation, the current estimate shows the relative importance of each source of THg and MeHg in Masan Bay. As seen in Table 3, the overall input estimation shows that 78 40 mol of THg enters Masan Bay annually. For THg, stream discharge was the dominant source, contributing 76% of the total input, followed by WWTP and SGD, contributing 7.8% and 6.7%, respectively. The mass ux estimate results are in agreement with the surface water distributions of THg: THg in surface water largely increased in August, particularly from ST1 to ST4 (Fig. 2a), during the rainy season when the TSS increased. The signicance of stream or river input as a source of coastal THg has been reported from the Hg mass ux results of large uvial ow estuaries (Tokyo Bay, Sakata et al., 2006; NY/NJ Harbor Estuary, Balcom et al., 2008; Long Island Sound, Balcom et al., 2004; Bay of Fundy, Sunderland et al., 2010).

0.8

b
MeHg (pmol/g)
1 2 3 4 5 6 7 8

2.5 2.0 1.5 1.0 0.5 0.0 1 2 3 4 5 6 7 8

0.6

THg (nmol/g)

0.4

0.2

0.0

Station

Station

1.6 1.4 1.2

16 14 12

30 25

% MeHg

1.0 0.8 0.6 0.4 0.2 0.0 1 2 3 4 5 6 7 8

10 15 8 6 4 2 0 1 2 3 4 5 6 7 8 10 5 0

Station

Station

Fig. 4. (a) Sediment THg; (b) MeHg; (c) %MeHg; and (d) OM content and AVS (open circle) in Masan Bay. Missing data indicate no sample collection.

AVS ( mol/g)

OM (%LOI)

20

E. Kim et al. / Marine Chemistry 158 (2014) 5968

65

THg (nmol/g)

The MeHg input estimate shows that a total of 0.39 0.15 mol of MeHg enters the Masan Bay annually (Table 3). In contrast to THg, the MeHg input through SGD was dominant, contributing 53% of the total input, followed by WWTP and stream, contributing 17% and 16%, respectively. This result agrees with that of other estuarine systems, suggesting that most MeHg is produced in situ from anoxic sediment and is then transported to overlying water (Balcom et al., 2004; Mason et al., 1999). Indeed, our results showed that the MeHg distribution in surface seawater was not associated with TSS or THg, implying that freshwater inow was not a major source of MeHg. More MeHg was transported from sediment to the overlying water through groundwater discharge than through diffusion, although our mass ux calculation should be further rened. Our results showed the importance of SGD MeHg ux, although this might be a local phenomenon associated with large groundwater discharge in the Korean peninsula, as others have found (Kim et al., 2005; Hwang et al., 2005; Waska and Kim, 2010). 3.4. Distribution of Hg in sediment As shown in Fig. 4a, sediment THg concentrations were signicantly (p b 0.05) higher at the upper sites (ST1ST4) than at the lower bay (ST5ST8), with ST2 the highest (0.65 0.07 nmol g 1). High sediment THg in the upper bay could be attributed to the large THg input through stream TSS, which agreed with the mass ux estimate that stream water input was the major contributor of THg to Masan Bay. Distribution of sediment MeHg, however, did not show a similar pattern to THg (Fig. 4b), implying that the TSS supply from stream water was not a major source of sediment MeHg. The sediment %MeHg (the fraction of MeHg over THg) increased toward the lower bay (ST5ST8; Fig. 4c). Similar to THg, the spatial differences in MeHg and the %MeHg appeared to be larger than the temporal variations. However, this should be interpreted carefully because our study did not measure the vertical prole of MeHg or the %MeHg, which was typically controlled by the depth of oxic/anoxic boundary layer and the degree of organic matter degradation. As seen in Fig. 4d, the sediment OM content was slightly more enriched in the upper bay (ST1ST4: 11.0 1.2%, on average, with a range of 9.0% to 13.6%) than in the lower bay (ST5, 78: 8.4 1.6%, on average, with a range of 5.9% to 10.4%). Sediment OM content was lowest at ST6 with an average of 4.2 1.4% (2.7%5.3%). Overall, sediment OM content in Masan Bay was comparable to other estuarine systems such as the Chesapeake Bay (USA, Hollweg et al., 2009) and the NJ/NY Harbor (USA, Hammerschmidt et al., 2008) but somewhat higher than the Seine (France) and Medway (UK) estuaries (Ouddane et al., 2008). As mentioned, AVS was measured for May 2010 (Fig. 4d). AVS showed spatial differences, with the highest AVS at ST1 and the lowest at ST6 where the sediment OM content was lowest. AVS concentration ranged from 1.1 to 28 mol g 1, averaging 12 8.8 mol g 1. Our data fell within a range similar to the range of 1.5 to 22 mol g 1 for the Chesapeake
800
Mar 2009 May 2009 Aug 2009 Jan 2010 May 2010

Bay (Hollweg et al., 2009) and 0.01 to 44 mol g 1 for the NY/NJ Harbor (Hammerschmidt et al., 2008), although these comparisons should be interpreted carefully because our surface (015 cm) sediment samples were homogenized before AVS analysis. Sediment THg concentrations were positively and signicantly (r = 0.77, p b 0.05, n = 34; Fig. 5) correlated with sediment OM content, which explained 59% of the variability in the THg concentrations.

a
10
Persian Gulf San Francisco Bay NY/NJ Harbor Msan Bay

0 0 2 4 6 8 10 12 14 16

OM (%LOI)

b
50
Persian Gulf San Francisco Bay NY/NJ Harbor Masan Bay

40

MeHg (pmol/g)

30

20

10

0 0 2 4 6 8 10 12 14 16

OM (%LOI)

c
2.0

Persian Gulf San Francisco Bay NY/NJ Harbor Masan Bay

1.5

% MeHg

600

1.0

THg (pmol/g)

400

0.5

200

0.0 0 2 4 6 8 10 12 14 16

0 0 2 4 6 8 10 12 14 16

OM (%LOI) OM (%LOI)
Fig. 6. Relationships of (a) THg; (b) MeHg; and (c) %MeHg with OM among different urban and industrialized coastal sediments. Individual data are from the references shown in Table 4. Solid linear regression lines represent San Francisco Bay and NJ/NY Harbor sediments and a dashed linear regression line represents Masan Bay and Persian Gulf.

Fig. 5. Correlation between sediment THg and OM in Masan Bay.

66

E. Kim et al. / Marine Chemistry 158 (2014) 5968

Table 4 Comparisons of THg, MeHg, %MeHg, and sediment organic matter content as LOI in surface sediment from industrial and urban estuarine coastal systems. Location Masan Bay, Koreaa Persian Gulf, Iranb San Francisco Bay, USAc NJ/NY Harbor, USAd Seine Bay, Francee Medway Bay, UKe
a b c d e

n 34 24 25 25 10 10

THg (nmol g 1) 0.34 0.18 1.10 4.86 1.55 2.85 0.20 (0.020.73) 0.05 (0.120.28) 0.57 (0.102.80) 2.45 (1.119.79) 0.6 0.95

MeHg (pmol g1) 1.12 1.25 3.62 23.9 4.15 16.0 0.42 (0.191.88) 0.47 (0.502.00) 4.03 (0.1017.0) 12.3 (3.049.1) 1.20 4.75

MeHg/THg (%) 0.50 0.70 0.35 0.56 0.27 0.56 0.35 (0.081.46) 0.21 (0.331.11) 0.40 (0.041.70) 0.36 (0.171.68) 0.20 0.50

OM (%LOI) 9.47 2.45 (2.7013.6) 1.98 0.58 (0.693.10) 2.02 0.78 (0.343.79) 8.01 2.67 (2.7011.5) 1.725.17 3.105.17

This study (top 15 cm). Agah et al. (2009), sample (top 510 cm) collection in January 2004. Conaway et al. (2003), sample (top 5 cm) collection in July 2000. Hammerschmidt et al. (2008), sample (top 36 cm) collection in August 2002 and February/May 2003. Ouddane et al. (2008), sample (top 20 cm) collected in July 2003 (Seine) and September 2003 (Medway).

Sediment THg in Masan Bay appears to be strongly associated with sediment OM content controlling THg distribution and perhaps partitioning. A similar relationship has been found for various estuarine systems (Benoit et al., 2006; Conaway et al., 2003; Heyes et al., 2006; Han et al., 2007; Hammerschmidt et al., 2008). However, the sediment MeHg concentration was not signicantly correlated with either the sediment THg or OM content. The sediment %MeHg was negatively and signicantly (r = 0.68, p b 0.05, n = 34; Fig. 6c) correlated with the sediment OM content in Masan Bay, which explained 46% of the variability in the %MeHg. Given that the %MeHg is a good indicator of the net MeHg production rate found by other studies (Kim et al., 2006; Sunderland et al., 2006; Hollweg et al., 2009), our results suggest that sediment MeHg production in Masan Bay is negatively related to sediment OM content. This result is similar to what Hollweg et al. (2009) found from the Chesapeake Bay. The same study demonstrated that the vertical zonation of Hg methylation varied between sampling sites, condensed into the upper few cm in the organic rich upper bay and broader and deeper in the lower bay, shelf and slope (Hollweg et al., 2009).

3.5. Correlation between %OM and %MeHg in coastal sediments Table 4 shows comparisons of THg, MeHg, %MeHg, and sediment OM content from various industrialized and urban coastal systems in temperate and arid zones. For consistency, we converted sediment organic carbon to sediment OM content (as % LOI), using a conversion factor of 1.72 (assuming that 58% of OM is organic carbon; Frangipane et al., 2009). In Masan Bay and the NJ/NY Harbor, the sediment OM content was higher overall, on average, and more variable than other estuarine systems (Table 4, Fig. 6). Sediment THg showed a signicant (r = 0.82, p b 0.05, n = 51) positive correlation with sediment OM content for the San Francisco Bay and the NJ/NY Harbor, and discretely for the Persian Gulf and Masan Bay (r = 0.72, p b 0.05, n = 60; Fig. 6a). Sediment OM explained 67% and 52% of the variability in THg levels for the highTHg sites (San Francisco Bay and NJ/NY Harbor) and the low-THg sites (Persian Gulf and Masan Bay), respectively. Although the MeHg concentration in the San Francisco Bay and NJ/NY Harbor sediments was positively and signicantly (r = 0.91, p b 0.05, n = 50) correlated with the sediment OM content, the MeHg in the Persian Gulf and Masan Bay did not show any noteworthy relationship (Fig. 6b). The %MeHg, as a proxy of MeHg production, decreased signicantly (r = 0.53, p b 0.05, n = 58; Fig. 6c) as the sediment OM increased in the low-THg sites, the Masan Bay and the Persian Gulf. This might be attributable to the fact that: 1) inorganic Hg bioavailability is a limiting factor for MeHg production, and 2) bioavailable Hg concentration decreases as sediment OM increases, as was shown by a few estuarine studies (Hammerschmidt and Fitzgerald, 2004; Hammerschmidt et al., 2008). The negative correlation between %MeHg (or Hg methylation rate) and OM was also found from the Long Island Sound and Chesapeake Bay sediments (Hammerschmidt and Fitzgerald, 2004; Hammerschmidt et al., 2008; Hollweg et al., 2009). In Long Island Sound, partitioning coefcients

for THg and MeHg were related positively to the OM content of deposits, suggesting that OM reduces bioavailability of Hg and attenuate MeHg production in sediment (Hammerschmidt et al., 2008). In Chesapeake Bay, increased vertical expanses of MeHg production in the low OM sites caused higher MeHg inventories in offshore sediments (Hollweg et al., 2009). In contrast, %MeHg showed a weak positive (r = 0.39, p b 0.05, n = 50; Fig. 6c) correlation with sediment OM content for the high-THg sites, the San Francisco Bay and the NJ/NY Harbor. A constructive role for OM in the sediment %MeHg is explained by the fact that elevated sediment organic matter content may stimulate formation of anoxic zone and anaerobic bacterial activities there (Lambertsson and Nilsson, 2006; Schartup et al., 2013). In similar to this, when MeHg production was determined from the eleven estuarine sites in the east US coast, sediment organic matter showed positive linear relation with Hg methylation rate, implying that OM supports microbial communities and perhaps enhances Hg methylation (Schartup et al., 2013). Up to dates, only limited study has compared different urban/industrialized estuarine system in order to understand the role of sediment OM in Hg methylation potential from the empirical relationship (Schartup et al., 2013). The previous literatures on each urban estuarine system mostly suggested negative effects of OM on the Hg methylation potential (Hammerschmidt and Fitzgerald, 2004; Hammerschmidt et al., 2008; Hollweg et al., 2009). Our results however suggest that the correlation between sediment OM and MeHg production is not straightforward. The limiting factor for the sediment MeHg production appear to vary, relying on the biogeochemical conditions of each coastal system, and sediment OM may either increase %MeHg via enhancing microbial activities or decrease %MeHg via limiting Hg bioavailability. It agrees to the recent suggestion by Schartup et al. (2013): a reduction

0.04

0.03

Persian Gulf San Francisco Bay NY/NJ Harbor Masan Bay Seine estuary Medway estuary

MeHg (nmol/g)

0.02

0.01

y=0.0052x-0.0011 R2=0.95
0.00 0 2 4 6 8

THg (nmol/g)
Fig. 7. Relationship between average concentrations of MeHg and THg in surface sediments from various urban and industrialized coastal environments. Error bars present standard deviations of the reported measurements. Data are shown in Table 4.

E. Kim et al. / Marine Chemistry 158 (2014) 5968

67

in sediment OM content will not necessarily lead to increase net Hg methylation, as was suggested by the sediment eutrophication hypothesis. Finally, we found signicant (r2 = 0.953, p b 0.05) positive relation between the THg and MeHg concentrations, on average (Fig. 7). The slope of the relationship (the average ratio of MeHg to THg) was 0.0052 or 0.52%, similar to the ratio of MeHg to THg found in the literature (Hammerschmidt and Fitzgerald, 2006; Schartup et al., 2013). It would be possible to use THg loads as a proxy to estimate MeHg concentration in urban/industrialized estuarine sediments. Although our study contributes to better understanding of MeHg distributions in urban estuarine sites, further studies are necessary for in-depth investigation of the relationship between MeHg production and the sediment biogeochemical factors. Since sediments are a major location for Hg methylation in estuarine systems, better estimates of MeHg sediment ux are also critical to ultimately reduce human exposure to MeHg. Additionally, the effects of ongoing sediment management programs (e.g., sediment dredging and/or nutrient control programs) in urban bays should be carefully evaluated to avoid increased MeHg sediment ux. 4. Summary The overall Hg input estimate showed that a total of 78 40 mol of THg and 0.39 0.15 mol of MeHg enters the Masan Bay annually. The stream discharge and SGD appear to be a dominant source of THg and MeHg, respectively. Our ux estimate generally agrees with the THg and MeHg distribution in water and sediment. Sediment OM content appeared to have a signicant impact on THg and %MeHg distributions for Masan Bay. When various coastal sediments in global urban/industrial areas were compared, sediment OM content showed an ambiguous role in either increasing or decreasing MeHg production: sediment OM may either increase Hg methylation potential via enhancing microbial activities or decrease Hg methylation potential via limiting Hg bioavailability. Further research is necessary to better understand the critical biogeochemical factors for the MeHg production in coastal systems. Acknowledgment The authors thank Nam-hyun Kim, Myung-gil Choi, and Soo-Chan Ahn from CNU, Indar Ar from GIST, and Tae-Hoon Kim and Intae Kim from SNU for their support in collecting samples. We also thank Hyunkyung Lee for the AVS analysis. This work was supported by the Korea Research Foundation Grants funded by the Korean Government (KRF2008-532-C00040 and KRF-2012R1A2A206046793). References
Agah, H., Elskens, M., Fatemi, S., Ow, F., Baeyens, W., Leermakers, M., 2009. Mercury speciation in the Persian Gulf sediments. Environ. Monit. Assess. 157, 363373. Allen, D., Hayhurst, A.N., 1991. The kinetics of the reaction between calcium oxide and hydrogen sulphide at the temperatures of uidized bed combustors. Symposium (International) on Combustion, 23, pp. 935941. Balcom, P.H., Fitzgerald, W.F., Vandal, G.M., Lamborg, C.H., Rolfhus, K.R., Langer, C.S., et al., 2004. Mercury sources and cycling in the Connecticut River and Long Island Sound. Mar. Chem. 90, 5374. Balcom, P.H., Hammerschmidt, C.R., Fitzgerald, W.F., Lamborg, C.H., O'Connor, J.S., 2008. Seasonal distributions and cycling of mercury and methylmercury in the waters of New York/New Jersey Harbor Estuary. Mar. Chem. 109, 117. Baumann, E.W., 1974. Determination of parts per billion sulde in water with the suldeselective electrode. [4] Anal. Chem. 46, 13451347. Benoit, J.M., Gilmour, C.C., Mason, R.P., Riedel, G.S., Riedel, G.F., 1998. Behavior of mercury in the Patuxent river estuary. Biogeochemistry 40, 249265. Benoit, J.M., Gilmour, C.C., Mason, R.P., Heyes, A., 1999. Sulde controls on mercury speciation and bioavailability to methylating bacteria in sediment pore waters. Environ. Sci. Technol. 33, 951957. Benoit, J.M., Mason, R.P., Gilmour, C.C., Aiken, G.R., 2001. Constants for mercury binding by dissolved organic matter isolates from the Florida Everglades. Geochim. Cosmochim. Acta 65, 44454451. Benoit, J., Gilmour, C.C., Heyes, A., Mason, R., Miller, C., 2003. Geochemical and biological controls over methylmercury production and degradation in aquatic systems. In: Chai, Y., Braids, O.C. (Eds.), Biochemistry of Environmental Important Trace Elements. American Chemical Society, Washington, DC, pp. 262297.

Benoit, P., Gratton, Y., Mucci, A., 2006. Modeling of dissolved oxygen levels in the bottom waters of the Lower St. Lawrence Estuary: coupling of benthic and pelagic processes. Mar. Chem. 102, 1332. Black, F.J., Paytan, A., Knee, K.L., de Sieyes, N.R., Ganguli, P.M., Gray, E., Flegal, A.R., 2009. Submarine groundwater dsicharge of total mercury and monomethylmercury to Central California coastal waters. Environmental Science & Technology 43, 56525659. Bloom, N.S., Crecelius, E.A., 1983. Determination of mercury in seawater at sub-nanogram per liter levels. Mar. Chem. 14, 4959. Bloom, N., Fitzgerald, W.F., 1988. Determination of volatile mercury species at the picogram level by low-temperature gas chromatography with cold-vapour atomic uorescence detection. Anal. Chim. Acta. 208, 151161. Bloom, N.S., 1989. Determination of picogram levels of methylmercury by aqueous phase ethylation, followed by cryogenic gas-chromatography with cold vapor atomic uorescence detection. Can. J. Fish. Aquat. Sci. 46 (7), 11311140. Bone, S.E., Charette, M.A., Lamborg, C.H., Gonneea, M.E., 2007. Has submarine groundwater discharge been overlooked as a source of mercury to coastal waters? Environmental Science & Technology 41, 30903095. Boudreau, B.P., 1996. The diffusive tortuosity of ne-grained unlithied sediments. Geochim. Cosmochim. Acta 60, 31393142. Ci, Z.J., Zhang, X.S., Wang, Z.W., Niu, Z.C., Diao, X.Y., Wang, S.W., 2011. Distribution and air sea exchange of mercury (Hg) in the Yellow Sea. Atmos. Chem. Phys. 11, 28812892. Cho, H.Y., Chae, J.W., 1998. Analysis on the characteristics of the pollutant load in Chinhae-Masan Bay. J. Korean Soc. Coast. Ocean Eng. 10 (1998), 132140. Cho, Y.K., Kim, M.-O., Kim, B.-C., 2000. Sea fog around the Korean Peninsula. J. Appl. Meteorol. 39, 24732479. Conaway, C.H., Squire, S., Mason, R.P., Flegal, A.R., 2003. Mercury speciation in the San Francisco Bay estuary. Mar. Chem. 80, 199225. David, N., McKee, L.J., Black, F.J., Flegal, R., Conaway, C.H., Schoellhamer, D.H., Ganju, N.K., 2009. Mercury concentrations and loads in a large river system tributary to San Francisco Bay, California, USA. Environ. Toxicol. Chem. 28, 20912100. Drott, A., Lambertsson, L., Bjrn, E., Skyllberg, U., 2007. Importance of dissolved neutral mercury suldes for methyl mercury production in contaminated sediments. Environmental Science & Technology 41, 22702276. Duong, H.V., Han, S., 2011. Assessment of benthic transfer and speciation of mercury in wetland sediments downstream from a sewage outfall. Ecol. Eng. 37, 989993. EPA, 1996. Method 9215: potentiometric determination of sulde in aqueous samples and distillates with ion-selective electrode. US EPA. EPA, 1998. Method 1631: mercury in water by oxidation, purge and trap, and cold vapor atomic uorescence spectrometry. US EPA. Frangipane, G., Pistolato, M., Molinaroli, E., Guerzoni, S., Tagliapietra, D., 2009. Comparison of loss on ignition and thermal analysis stepwise methods for determination of sedimentary organic matter. Aquat. Conserv. Mar. Freshwat. Ecosyst. 19, 2433. Gill, G.A., Bloom, N.S., Cappellino, S., Driscoll, C.T., Dobbs, C., Mcshea, L., Mason, R.P., Rudd, J.W.M., 1999. Environ. Sci. Technol. 33, 663669. Gilmour, C.C., Henry, E.A., Mitchell, R., 1992. Sulfate stimulation of mercury methylation in freshwater sediments. Environ. Sci. Technol. 26, 22812287. Gilmour, C.C., Podar, M., Bullock, A.L., Graham, A.M., Brown, S.D., Somenahally, A.C., 2013. Environ. Sci. Technol. (In Press). Gobeil, C., Cossa, D., 1993. Mercury in sediments and sediment pore water in the Laurentian trough. Can. J. Fish. Aquat. Sci. 50, 17941800. Hammerschmidt, C.R., Fitzgerald, W.F., 2004. Geochemical controls on the production and distribution of methylmercury in near-shore marine sediments. Environmental Science & Technology 38, 14871495. Hammerschmidt, C.R., Fitzgerald, W.F., 2008. Sedimentwater exchange of methylmercury determined from shipboaed benthic ux chambers. Mar. Chem. 109, 8697. Hammerschmidt, C.R., Fitzgerald, W.F., Lamborg, C.H., Balcom, P.H., Visscher, P.T., 2004. Biogeochemistry of methylmercury in sediments of Long Island Sound. Mar. Chem. 90, 3152. Hammerschmidt, C.R., Fitzgerald, W.F., 2006. Methylmercury in freshwater sh linked to atmospheric mercury deposition. Environmental Science & Technology 40, 77647770. Hammerschmidt, C.R., Fitzgerald, W.F., Balcom, P.H., Visscher, P.T., 2008. Organic matter and sulde inhibit methylmercury production in sediments of New York/New Jersey Harbor. Mar. Chem. 109, 165182. Han, S., Obraztsova, A., Pretto, P., Choe, K.Y., Gieskes, J., Deheyn, D.D., et al., 2007. Biogeochemical factors affecting mercury methylation in sediments of the Venice Lagoon, Italy. Environ. Toxicol. Chem. 26, 655663. Han, S., Narasingarao, P., Obraztsova, A., Gieskes, J., Hartmann, A.C., Tebo, B.M., Allen, E.E., Deheyn, D.D., 2010. Mercury speciation in marine sediments under sulfate-limited conditions. Environ. Sci. Technol. 44, 37523757. Han, S., Porrachia, M., Volpato, E., Gieskes, J., Deheyn, D., 2011. Mercury concentration and monomethylmercury production in sediment: effect of dredged sediment reuse on bioconcentration for ragworms. Water, Air, & Soil Pollution 219, 5968. Heyes, A., Mason, R.P., Kim, E.H., Sunderland, E., 2006. Mercury methylation in estuaries: insights from using measuring rates using stable mercury isotopes. Mar. Chem. 102, 134147. Hollweg, T.A., Gilmour, C.C., Mason, R.P., 2009. Methylmercury production in sediments of Chesapeake Bay and the mid-Atlantic continental margin. Mar. Chem. 114, 86101. Hong, S.H., Munschy, C., Kannan, N., Tixier, C., Tronzynski, J., Has-Moisan, K., Shim, W.J., 2009. PCDD/F, PBDE, and nonylphenol contamination in a semi-enclosed bay (Masan Bay, South Korea) and a Mediterranean lagoon (Thau, France). Chemosphere 77, 854862. Hong, S.H., Kannan, N., Jin, Y., Won, J.H., Han, G.M., Shim, W.J., 2010. Temporal trend, spatial distribution, and terrestrial sources of PBDEs and PCBs in Masan Bay, Korea. Mar. Pollut. Bull. 60, 18361841.

68

E. Kim et al. / Marine Chemistry 158 (2014) 5968 Luengen, A.C., Flegal, A.R., 2009. Role of phytoplankton in mercury cycling in the San Francisco Bay estuary. Limnol. Oceanogr. 54, 2340. Lim, D., Kim, Y.-O., Kang, M.R., Jang, P.-K., Shin, K., Jang, M., 2007. Variability of water quality and limiting factor for primary production in semi-enclosed Masan Bay, South Sea of Korea. Ocean and Polar Research 29, 349366 ((Korean)). Mason, R.P., Fitzgerald, W.F., Hurley, J., Hanson Jr., A.K., Donaghay, P.L., Sieburth, J.M., 1993. Mercury biogeochemical cycling in a stratied estuary. Limnol. Oceanogr. 38, 12271241. Mason, R., Bloom, N., Cappellino, S., Gill, G., Benoit, J., Dobbs, C., 1998. Investigation of porewater sampling methods for mercury and methylmercury. Environmental Science & Technology 32, 40314040. Mason, R.P., Lawson, N.M., Lawrence, A.L., Leaner, J.J., Lee, J.G., Sheu, G.R., 1999. Mercury in the Chesapeake Bay. Mar. Chem. 65, 7796. Moon, H.B., Yoon, S.P., Jung, R.H., Choi, M., 2008. Wastewater treatment plants (WWTPs) as a source of sediment contamination by toxic organic pollutants and fecal sterols in a semi-enclosed bay in Korea. Chemosphere 73, 880889. Moon, H.B., Choi, M., Choi, H.G., Ok, G., Kannan, K., 2009. Historical trends of PCDDs, PCDFs, dioxin-like PCBs and nonylphenols in dated sediment cores from a semienclosed bay in Korea: tracking the sources. Chemosphere 75, 565571. Ouddane, B., Mikac, N., Cundy, A.B., Quillet, L., Fisher, J.C., 2008. A comparative study of mercury distribution and methylation in mudats from two macrtidal estuaries: the Seine (France) and the Medway (United Kingdom). Appl. Geochem. 23, 618631. Parks, J.M., Johs, A., Podar, M., Bridou, R., Hurt Jr., R.A., Smith, S.D., et al., 2013. The genetic basis for bacterial mercury methylation. Science 339, 13321335. Rahman, M.D.M., Lee, Y., Kim, G., Lee, K., Han, S., 2013. Signicance of submarine groundwater discharge in the coastal mass balance of mercury in Hampyeong Bay, Yellow Sea. Chemospere 91 (3), 320327. Ram, A., Borole, D.V., Rokade, M.A., Zingde, M.D., 2009. Diagenesis and bioavailability of mercury in the contaminated sediments of Ulhas Estuary, India. Mar. Pollut. Bull. 58, 16851693. Sakata, M., Marumoto, K., 2005. Wet and dry deposition uxes of mercury in Japan. Atmos. Environ. 39, 31393146. Sakata, M., Marumoto, K., Narukawa, M., Asakura, K., 2006. Mass balance and sources of mercury in Tokyo Bay. J. Oceanogr. 62, 767775. Schartup, A.T., Mason, R.P., Balcom, P.H., Hollweg, T.A., Chen, C.Y., 2013. Methylmercury production in estuarine sediments: role of organic matter. Environ. Sci. Technol. 47, 695700. SPSS, 2003. SPSS Base 12.0 User's Guide. SPSS Inc. Sunderland, E.M., Gobas, F.A.P.C., Heyes, A., Branreun, B.A., Bayer, A.K., Cranston, R.E., 2004. Speciation and bioavailability of mercury in well-mixed estuarine sediments. Mar. Chem. 90, 91105. Sunderland, E.M., Gobas, F.A.P.C., Branreun, B.A., Heyes, A., 2006. Environmental controls on the speciation and distribution of mercury in coastal sediments. Mar. Chem. 102, 111123. Sunderland, E.M., Dalziel, J., Heyes, A., Branreun, B., Krabbenhoft, D., Gobas, F.A.P.C., 2010. Response of a macrotidal estuary to changes in anthorpogenic mercury loading between 1850 and 2000. Environ. Sci. Technol. 44, 16981704. Waska, H., Kim, G., 2010. Submarine groundwater discharge (SGD) as a main nutrient source for benthic and water-column primary production in a large intertidal environment of the Yellow Sea. J. Sea Res. 65, 103113. Weiss-Penzias, P.S., Ortiz, C., Acosta, R.P., Heim, W., Ryan, J.P., Fernandez, D., Collett, J.L., Flegal, A.R., 2012. Total and monomethyl mercury in fog water from the central California coast. Geophys. Res. Lett. http://dx.doi.org/10.1029/2011GL050324.

Horvat, M., Bloom, N.S., Liang, L., 1993. Comparison of distillation with other current isolation methods for the determination of methyl mercury-compounds in low-level environmental-samples: 1. Sediments. Anal. Chim. Acta. 281 (1), 135152. Hsu-Kim, H., Kucharzyk, K.H., Zhang, T., Deshusses, M.A., 2013. Mechanisms regulating mercury bioavailability for methylating microorganixmx in the aquatic environment: a critical review. Environmental Science & Technology 47, 24412456. Hwang, D.W., Lee, Y.W., Kim, G., 2005. Large submarine groundwater discharge and benthic eutrophication in Bangdu Bay on volcanic Jeju Island, Korea. Limnol. Oceanogr. 50, 13931403. Hyun, S., Lee, C.H., Lee, T., Choi, J.W., 2007. Anthropogenic contributions to heavy metal distributions in the surface sediments of Masan Bay, Korea. Mar. Pollut. Bull. 54, 10591068. Jang, J., Han, S., 2011. Importance of monsoon rainfall in mass uxes of ltered and unltered mercury in Gwangyang Bay, Korea. Sci. Total Environ. 409, 14981503. Jang, P.G., Shin, K., Chang, M., Kim, D., 2011. Spatial and temporal trends in water quality in response to sewage discharge in Masan and Hangam Bays, Korea. J. Coast. Res. 27, 144155. KEI report, 1996. A study on the total mass emissions managements for the reduction of contaminants in the coastal zone. Korea Environment Institute ((in Korean)). Kerin, E.J., Gilmour, C.C., Roden, E., Suzuki, M.T., Coates, J.D., Mason, R.P., 2006. Mercury methylation by dissimilatory iron-reducing bacteria. Appl. Environ. Microbiol. 72, 79197921. Kim, E.H., Mason, R.P., Porter, E.T., Soulen, H.L., 2004. The effect of resuspension on the fate of total mercury and methylmercury in a shallow estuarine ecosystem: a mesocosm study. Mar. Chem. 86, 121137. Kim, E.H., Mason, R.P., Porter, E.T., Soulen, H.L., 2006. The impact of resuspension on sediment mercury dynamics, and methylmercury production and fate: a mesocosm study. Mar. Chem. 102, 300315. Kim, E., Kim, H., Shin, K.-H., Kim, M.-S., Kundu, S.R., Lee, B.-G., et al., 2012. Biomagnication of mercury through the benthic food webs of a temperate estuary: Masan Bay, Korea. Environ. Toxicol. Chem. 31, 12541263. Kim, G., Ryu, J.-W., Yang, H.-S., Yun, S.-T., 2005. Submarine groundwater discharge (SGD) into the Yellow Sea revealed by 228Ra and 226Ra isotopes: implications for global silicate uxes. Earth Planet. Sci. Lett. 237, 156166. Kim, H., Noh, S., Kim, E., Poopal, R.K., Lee, H.Y., Han, S., 2011a. Bioconcentration of methylmercury in microzooplankton in a temperate river system. Environmental Toxicology & Chemistry 30, 28602867. Kim, M., Han, S., Gieskes, S., Deheyn, D., 2011b. Importance of organic matter lability for monomethylmercuyry production in sulfate-rich marine sediments. Sci. Total Environ. 409, 778784. Lambertsson, L., Nilsson, M., 2006. Organic material: the primary control on mercury methylation and ambient methylmercury concentrations in estuarine sediments. Environmental Science & Technology 40, 18221829. Laurier, F.J.G., Cossa, D., Beucher, C., Brvire, E., 2007. The impact of groundwater discharges on mercury partitioning, speciation and bioavailability to mussels in a coastal zone. Mar. Chem. 104, 143155. Lee, Y.W., Hwang, D.W., Kim, G., Lee, W.C., Oh, H.T., 2009. Nutrient inputs from submarine groundwater discharge (SGD) in Masan Bay, an embayment surrounded by heavily industrialized cities, Korea. Sci. Total Environ. 407, 31813188. Lee, Y.G., Rahman, M.D.M., Kim, G., Han, S., 2011. Mass balance of total mercury and monomethylmercury in coastal embayments of a volcanic island: signicance of submarine groundwater discharge. Environmental Science & Technology 45, 98919900.