Chemosphere 144 (2016) 2097e2105
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Polybrominated diphenyl ethers in sediments from the Southern Yellow Sea: Concentration, composition profile, source identification and mass inventory Guoguang Wang a, Jialin Peng a, Xiang Xu b, Dahai Zhang a, Xianguo Li a, * a b
Key Laboratory of Marine Chemistry Theory and Technology, Ministry of Education, Ocean University of China, Qingdao, 266100, China College of Chemistry and Pharmaceutical Sciences, Qingdao Agricultural University, Qingdao, 266109, China
h i g h l i g h t s BDE-209 is a dominant congener in the SYS, indicating high usage of Deca-BDE in China. PBDE concentrations increased from the coastal areas to the central mud area. Continental runoff and atmospheric deposition contributed 69.0% and 31.0% of PBDEs in the SYS, respectively. The SYS is a relatively weak sink of PBDEs in the world.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 July 2015 Received in revised form 20 October 2015 Accepted 22 October 2015 Available online xxx
The Southern Yellow Sea (SYS) is believed to be influenced by the contaminants from mainland China and the Korean peninsula. Here we report the first record about concentrations of polybrominated P diphenyl ethers (PBDEs) in the sediments of the SYS. The concentrations of 7PBDEs (BDE-28, 47, 99, 1 100, 153, 154, 183) and BDE-209 were 0.064e0.807 ng g (dry weight) and 0.067e1.961 ng g1 with a mean value of 0.245 ng g1 and 0.652 ng g1, respectively. These are distinctively low compared with the PBDE levels previously reported in other regions of the world. PBDE concentrations gradually increased from the coastal areas to the central mud area. BDE-209 was the dominant congener, accounting for 70.2 e91.6% of the total PBDEs. Congener profiles of PBDEs were similar to those in sediments from the Bohai Sea (BS), Laizhou Bay and modern Yellow River, which might be a tentative indication that they shared similar sources. Principal component analysis (PCA) revealed that PBDEs in the SYS were mainly from continental runoff (69.0%) and atmospheric deposition (31.0%). Depth profile of PBDEs in a sediment core collected from the edge of the central mud area showed that concentration of BDE-209 rapidly increased in recent years, which is in accordance with the replacement in demand and consumption of Penta- and Octa-BDEs by the Deca-BDE. Compared with BS, East China Sea, Erie and Ontario, the SYS was a relatively P weak sink of PBDEs (0.102e1.288 t yr1 for 7PBDEs and 0.107e3.129 t yr1 for BDE-209) in the world. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Polybrominated diphenyl ethers Distribution Composition profile Source identification Mass inventory The Southern Yellow Sea
1. Introduction Polybrominated diphenyl ethers (PBDEs), a class of brominated flame retardants (BFRs), have been widely used in the world since 1970s (Li et al., 2006a). They are added in a variety of consumer products such as electronic components, plastics, textiles, furnishing foam and fabrics to reduce the fire hazards by interfering
* Corresponding author. E-mail addresses: [email protected] (G. Wang), [email protected] (X. Li). http://dx.doi.org/10.1016/j.chemosphere.2015.10.088 0045-6535/© 2015 Elsevier Ltd. All rights reserved.
with the combustion of the polymeric materials (USEPA, 2008). Categorized by the average bromine contents, Penta- (predominantly BDE-47 and BDE-99), Octa- (predominantly BDE-183) and Deca-BDEs (predominantly BDE-209) are three main commercial PBDE products. In China, PBDEs are one group of the most produced and commonly used BFRs. The domestic production of BFRs in 2000 was up to 1000 tons and Deca-BDE accounted for the major proportion (Mai et al., 2005). In addition, BFRs imported from other countries also contributed to a portion (although the exact amount is unknown) of the market in China, because BFRs produced in three of the largest manufactures (i.e., Great Lakes Chemical,
2098
G. Wang et al. / Chemosphere 144 (2016) 2097e2105
Indianapolis, IN; Albemarle Chemical, Richmond, VA; and Dead Sea Chemical, Beer-Sheva, Israel) were sold to China. Horribly, for China, the annual demand of BFRs had been increasing at an estimated rate of 8% in 2000 (Mai et al., 2005). PBDEs can be easily released into surrounding environments during the production, usage and dismantling processes, because they are added in products without chemical bonding (Rahman et al., 2001). Consequently, PBDE residues have been widely € ller et al., 2011; Melymuk et al., 2012), dusts found in air (Mo (Whitehead et al., 2013), soils (Cetin, 2014; Shi et al., 2014), sediments (Lee et al., 2014; Zhu et al., 2014c), water (Hui-mei et al., 2013), marine organisms (Li et al., 2010; Zhu et al., 2014a) and human tissues (Mitchell et al., 2012; Chen et al., 2014). Due to their chemical stability, tendency for bioaccumulation and potential toxicity to the ecosystem, PBDEs have caused increasing attention around the world. In animal tests, some congeners of PBDEs were related with endocrine disruption, thyroid hormone disorders, reproductive/developmental toxicity and carcinogenicity (McDonald, 2002). Because of these adverse effects, the production and usage of Penta- and Octa-BDEs were banned in Europe and some factories in the USA voluntarily phased out PBDE manufacturing in 2004. PBDEs, like other persistent organic pollutants (POPs), can be transported to the coastal areas via a variety of processes including atmospheric deposition, land runoff and direct industrial and domestic wastewater discharges. Because of their lipophilic, organophilic and hydrophobic characteristics, PBDEs adsorb strongly to soils and sediments. A few studies have focused on surveying PBDEs in sediments from the coastal seas adjacent to mainland China (Mai et al., 2005; Chen et al., 2006; Pan et al., 2010; Li et al., 2012). However, to date, no studies of PBDEs contamination in sediments of the Southern Yellow Sea (SYS), lying between the Chinese mainland and the Korean Peninsula, have been reported. Although the Yellow River and Yangtze River don't input directly into the SYS, the two rivers distinctly inflow the sources of sediments into the SYS (Lim et al., 2007). According to the previous reports (CSOA, 2007; MEPPRC, 2007), about 7.60% of the industrial and domestic toxic waste materials generated by land-based activities ended in the SYS. Consequently, the objectives of this study are to determinate the PBDE concentrations, as well as the spatial and temporal distribution, to identify PBDE sources and to estimate the mass inventory of PBDEs in the SYS. 2. Materials and methods
Standard solutions of 2,4,5,6-tetrachloro-m-xylene (TCmX) and PCB-209 were used as recovery surrogates. BDE-77 was added as the internal standard for quantification of targeted PBDEs. All solvents including methanol, n-hexane and isooctane were of HPLC grade. The neutral silica gel was heated to 180 C and held for 3 h, cooled down to room temperature, then deactivated with 5% (w/w) distilled water before being stored in sealed containers. Sodium sulfate was baked at 450 C for 4 h and preserved in sealed containers after cooling. 2.3. Extraction and analysis The sample extraction was performed based on an ultrasonically assisted alkaline hydrolysis extraction method established previously in our laboratory (Wang et al., 2015). Briefly, 10 g of freezedried and ground sediment samples were spiked with TCmX and PCB-209 and processed in an ultrasonic bath in 2.0 M NaOHmethanol solution for 30 min. Activated copper was added to remove sulfur during the extraction. Then 50 mL distilled water and 50 mL n-hexane, in turn, were added. The extracts of organic phase were separated from the mixture and concentrated to 2e3 mL using a rotary vacuum evaporator. Concentrated extracts were cleaned and fractionated by a chromatographic column packed with neutral silica gel (7 g, 5% water deactivated) and anhydrous sodium sulfate (3 g) from bottom to top. The PBDE fractions were eluted with 70 mL n-hexane, concentrated to about 0.5 mL and dried out under a gentle stream of nitrogen. Finally, all the extracts were solvent-exchanged to isooctane to a constant volume of 100 mL with BDE-77 as the internal quantification standard. Sample analysis was performed with a Shimadzu 2010plus gas chromatograph coupled with electron capture detector (GC-ECD). A DB-5MS (30 m 0.25 mm i.d. 0.25 mm film thickness) capillary column was used for the determination of PBDE congeners except for BDE-209. The oven temperature was initiated at 100 C for 1 min, ramped at 10 C/min to 190 C, 5 C/min to 255 C for 0.5 min, then up to 300 C at 2 C/min and held for 10 min. Automatic injection of 1 mL sample was conducted in the splitless mode. Helium was used as the carrier gas at a flow rate of 3 mL min1. Injector and detector temperatures were fixed at 290 C and 315 C, respectively. For BDE-209, a DB-5 (15 m 0.25 mm i.d. 0.10 mm film thickness) capillary column was used. The oven temperature was initiated at 100 C for 1 min, then ramped at 8 C/min to 310 C and held for 24 min. Injector and detector temperatures were fixed at 300 C and 325 C, respectively.
2.1. Sample collection
2.4. Quality assurance/quality control
A detailed illustration of the sampling sites is presented in Fig. 1. A total of 24 surface sediments (0e3 cm) were collected by using a Van Veen grabber in the SYS during November to December in 2013. In addition, a sediment core of approximately 36 cm long was collected at F4 site (122 000 0000 E, 35 000 0000 N, the red circle in Fig. 1) from the edge of the central mud area, and cut in 2 cm interval to yield 18 core sediment samples. All of the sediments were wrapped in solvent-rinsed aluminum foil, transported to the laboratory on ice and stored at 20 C until further analysis.
For each batch of 10 samples, a procedural blank (in which a filter paper identical to that used to wrap the sediments was solvent-extracted and processed in the same way as the samples), a matrix-spiked sample (a sediment spiked with 8 PBDE congeners, together with the same sediment without PBDEs spiked, was used for calculation of matrix-spiked recovery) and a sample duplicate were processed. Results showed that only trace amounts of BDE-47 and BDE-99 were found in procedural blanks, and they were appropriately subtracted from those in the sample extracts. The surrogate recoveries in 24 samples ranged from 77.3% to 110.8% with an average of 89.4 ± 10.1% for TCmX and from 84.4% to 111.6% with an average of 93.7 ± 9.5% for PCB-209. The mean recoveries in spiked matrix samples were 83.2% ± 13.5% for all targeted PBDE congeners except that of BDE-209, which was 73.7 ± 15.2%. Reported concentrations were not surrogate recovery corrected. Relative standard deviations (RSDs) for duplicated samples were 11.7%e20.4% with an average of 15.2%.
2.2. Materials A standard mixture of PBDEs (BDE-CSM) selected for quantitative analysis were purchased from AccuStandards, Inc. (USA). The BDE-CSM contains 8 PBDE congeners (IUPAC Nos. 28, 47, 99, 100, 153, 154, 183 and 209) which are treated as the compounds of priority control by US Environmental Protection Agency (USEPA).
G. Wang et al. / Chemosphere 144 (2016) 2097e2105
2099
Fig. 1. Locations of sampling sites (red dots) in the SYS. Circulation systems (dark blue arrows) and mud areas (light purple) are modified after Liu et al. (2007) and Cheng et al. (2004). (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this article.)
2.5. Source identification Principal component analysis (PCA) - multiple linear regression (MLR) did not need a priori knowledge of the number of sources and their compositions. So that it was widely used to identify the sources of pollutants in the environment. In this paper, PCA was conducted using factor extraction with the total variances >90% based on correlation matrix. Detailed descriptions were presented by Çamdevýren et al. (2005). 3. Results and discussion 3.1. Concentrations Summary of the PBDE concentrations in surface sediments from the SYS is presented in Table 1. Most of the PBDE congeners were detected in all sediment samples, except for BDE-28, 153 and 183 with the detection frequencies of 95.8%, 58.3% and 87.5%, respectively, indicating that PBDEs were widespread contaminants in the P SYS. Concentrations of 7PBDEs (sum of BDE-28, 47, 99, 100, 153, 154, 183) ranged from 0.064 to 0.807 ng g1 dry weight with a mean value of 0.245 ng g1. BDE-209 concentrations ranged from 0.067 to 1.961 ng g1 dry weight with an average of 0.652 ng g1, P and were 1e3 times higher than those of 7PBDEs, indicating that BDE-209 was the predominant congener of PBDE pollutants in the SYS sediments, which was consistent with the usage status of DecaBDE in China. PBDE concentrations in the SYS sediments were compared with
3.2. Spatial and temporal distributions
Table 1 Concentrations of 8 PBDE congeners in the SYS surface sediments. Congeners
BDE-28 BDE-47 BDE-99 BDE-100 BDE-153 BDE-154 BDE-183 P c 7PBDEs BDE-209 a
Concentrations (ng/g) a
Min.
Max.
Median
Mean
ndb 0.013 0.006 0.010 nd 0.010 nd 0.064 0.067
0.049 0.069 0.091 0.088 0.106 0.620 0.186 0.807 1.961
0.008 0.031 0.016 0.024 0.004 0.081 0.037 0.243 0.819
0.007 0.031 0.023 0.026 0.006 0.082 0.029 0.245 0.652
Mean-geometric mean value of 24 samples. nd-not detected, which was calculated as zero. P 7PBDEs-the sum concentrations of BDE-28, 47, 99, 100, 153, 154 and 183 in each site. b c
those reported for several areas from China and other countries (Table S1). Although the numbers of PBDE congeners analyzed in various studies were different, BDE-28, 47, 99, 100, 153, 154, 183 and 209 in sediment samples have been widely found. The PBDE concentrations (without BDE-209) reported in our study were comparable to those in sediments from Bohai Sea (BS), China (Pan et al., 2010), Yangtze River Delta (YRD), China (Chen et al., 2006), Korea coast (Moon et al., 2007) and Tokyo Bay, Japan (Minh et al., 2007), while slightly lower than South China Sea, China (Mai et al., 2005), East China Sea (ECS), China (Li et al., 2012), Xiamen areas, China (Li et al., 2010), Taiwan (Jiang et al., 2011), Goseong, Korea (Lee et al., 2014), Spain (Eljarrat et al., 2005), and Great Lakes, USA (Song et al., 2005), and much lower than Hong Kong mangrove (Zhu et al., 2014c), Pearl River Delta (PRD), China (Chen et al., 2013) and Lake Shihwa, Korea (Lee et al., 2014). For BDE-209, the levels in our study were comparable to those in South China Sea, China (Mai et al., 2005) and Taiwan (Jiang et al., 2011), but much lower than those in rapidly industrialized and urbanized areas such as ECS, China (Li et al., 2012), BS, China (Pan et al., 2010) and PRD, China (Chen et al., 2013), Korea coast (Moon et al., 2007), Goseong, Korea (Lee et al., 2014), Lake Shihwa, Korea (Moon et al., 2012), Great Lakes, USA (Song et al., 2005), Spain (Eljarrat et al., 2005) and Tokyo Bay, Japan (Minh et al., 2007). These results indicated a relatively low to moderate level of PBDE contamination in the SYS at present. However, considering the adverse effect of PBDEs on ecological system and human health, a systematic monitoring and risk assessment programs are desirable in the SYS.
3.2.1. Spatial distribution P Concentrations of 7PBDEs and BDE-209 both increased from the coastal areas to the central mud area (see Fig. 2). Similar observations were also reported for organo-chlorine pesticides (OCPs; Hu et al., 2011a, b) and polychlorinated biphenyls (PCBs; Duan et al., 2013) in this area. Organic pollutants with high lipotropy and hydrophobicity (such as PBDEs, PCBs, PAHs and so on) are usually considered to be preferentially bounded to particles with higher TOC (total organic carbon) contents and smaller particle sizes (Cornelissen et al., 2005). According to Hu et al. (2011b) and Duan et al. (2013), TOC contents increase, and the grain size decreases from the coastal areas to the central mud area in the SYS. ConseP quently, the distributions of 7PBDEs and BDE-209 may be associated with the distributions of TOC and gain size.
2100
G. Wang et al. / Chemosphere 144 (2016) 2097e2105
Fig. 2. Spatial distributions of
P
7PBDEs
(a) and BDE-209 (b) in surface sediments of the SYS using Kriging by Surfer 11.0.
P Over the entire studied area, the highest levels of 7PBDEs and BDE-209 were all observed at the central mud area associated with smaller particle size, higher TOC and TON (total organic nitrogen) contents (Duan et al., 2013). This TOC dependent post-depositional sorption environment in the SYS was usually confined by the homogenous environmental conditions (Hu et al., 2011a, b). Hu. (1984) and Shi et al. (2002) reported that the YS cold water mass, YS warm current and the coast currents dominated the sedimentary environments of the central mud area in the SYS. It is well known that there was no local coastal river directly inputting to the central mud deposits of the SYS, and the SYS was far away from the direct sources in the BS. Consequently, the effects of shelf mud deposition (i.e., the hydrodynamic transport and depositional mechanism of the fine-grained sediments) and atmospheric deposition (Hu et al., 2011a; Zhao et al., 2013) were the main factors influencing the distribution and fate of PBDEs in the SYS. During these processes, PBDEs associated with fine-grained particles could be transported to the central mud area together with the particuP late matters. However, the lowest levels of 7PBDEs and BDE-209 were found at 32 N and 34 N, respectively. In the area near 32 N, a thick layer of sand sediments with lower TOC contents was deposited due to the resuspension and transportation of sediments from the old Yellow River and Yangtze River (Zhang et al., 2007). According to Marinari et al. (2006) and Meyers (1997), the TOC/TON ratio is used to reflect the sources of OMs in marine sediments, and the ratio varies from 4 to 10 in the OMs released by marine phytoplankton, while the ratio exceeds 10 in terrestrial OMs. In this study, the ratio ranged from 5.4 to 7.7 in the area near 34 N (Duan et al., 2013), indicating the OMs mainly deriving from marine sources. Additionally, the sediments mainly originated from the old Yellow River Delta in this area (Milliman et al., 1986). PBDEs have been used as BFR since 1970s and mainly derive from anthropogenic activities (Li et al., 2006a). Consequently, these led to the lower levels of BDE-209 in the area near 34 N. Relatively higher levels of PBDEs were also observed in the sediments near the Yangtze River Estuary (YRE). This can be attributed to the direct waste discharges and surface runoff into the Yangtze River (Hui et al., 2009), which runs through lots of highly populated and urbanized cities, such as Shanghai, Nanjing and so on. And those wastes deposited at the YRE are major contributors to pollutants (Feng et al., 2004) such as PBDEs. 3.2.2. Temporal distribution A sediment core collected from the edge of the central mud area was analyzed to examine PBDE temporal distribution. The historical P records of 7PBDEs and BDE-209 in Fig. 3 showed that PBDEs were found in all sediment layers, indicating that PBDEs had existed in
the sediments of the SYS and the areas continuously received inputs of PBDEs throughout years. P Concentrations of the 7PBDEs (Fig. 3a) were not significantly different at the depth of 18 cme36 cm. It is not surprising that the demand of PBDEs was low at that corresponding time period due to PBDEs used as flame retardants since 1970s. In addition, vertical mixing such as turbulence during typhoon season (Zhu et al., 2014c; Drage et al., 2015) may cause homogenization of pollutants in the sediments. It was also possible that some PBDEs may have been degraded or debrominated by microorganisms and physio-chemical processes because of relatively long existing time at the bottom of the sediment core. From the depth of 18 cm-12 cm, concentrations increased sharply with an apparent peak value of 0.57 ng g1 dry weight observed at the depth of 12 cm. The widespread usages of technical Penta- and Octa-BDEs as the commercial BFR products might be responsible. The concentrations then decreased thereafter from the depth of 8 cm to the surface, which might be attributed to the banning of production and usage of Penta- and Octa-BDEs in recent years. The concentrations of BDE-209 (Fig. 3b) increased gradually from the bottom to the surface, suggesting that the input of BDE209 increased year by year. These increases coincided with the high consumption of Deca-BDE for the BFRs market in China. A distinctive increase was observed at the upper layers, indicating that large quantities of BDE-209 were possibly input in the studied area in recent years and/or an increase in the sedimentation rate P occurred in the top layers. Compared with 7PBDEs, the more rapid and obvious increase of the concentrations of BDE-209 at the upper layers indicated that the historical records of these two groups were different, and the demand of Penta- and Octa-BDEs was largely replaced by the Deca-BDE in recent years. Similar rapid increases of BDE-209 in sediments were also found in sediment cores collected from other regions in PRD (Chen et al., 2007) and Hong Kong mangrove sediments (Zhu et al., 2014c). As the production and usage of PBDEs continue in the studied area, the emission of PBDEs to the environment is unavoidable. So that the SYS must be closely monitored to have a better understanding of these problems. 3.3. Congener patterns As shown in Fig. S1, BDE-209 was the predominant congener in the surface sediments of the SYS, accounting for 70.2e91.6% of the total PBDEs in approximately 62.5% of all samples. The predominance of BDE-209 was widely reported in marine sediments from BS (Pan et al., 2010), coastal ECS (Li et al., 2012), PRD (Chen et al., 2013) and other regions of the world (Moon et al., 2012; Zhu
G. Wang et al. / Chemosphere 144 (2016) 2097e2105
Fig. 3. Historical profiles of
P
7PBDEs
(a) and BDE-209 (b) in a sediment core from the SYS.
et al., 2014c). This is expected due to the fact that the consumption and production of Deca-BDE were much higher than Penta- and Octa-BDEs around the world. In addition, BDE-209 has a higher octanolewater partition coefficient (log Kow z 10) than other congeners (Rahman et al., 2001) and was deemed to possess low bioavailability and adsorbed strongly to marine sediments, especially the coastal sediments. P Percentages of tri-to hepta-BDE congeners in 7PBDEs were presented in Fig. 4. Among the 7 PBDE congeners detected, BDE-154 and BDE-183 were the dominated congeners (except for sites E1 and E3), followed by BDE-47, BDE-100 and BDE-99, in agreement with other studies (Pan et al., 2010, 2011). The relatively high abundances of BDE-154 and BDE-183 are likely attributed to the usage of Octa-BDE, with BDE-183 often taken as indicative of the presence of Octa-BDE commercial formulation (La Guardia et al., 2006). While the abundances of BDE-47, BDE-99 and BDE-100, the main congeners of Penta-BDE, indicated the source contribution of Penta-BDE commercial formulation. The ratio of sum of BDE-47 þ BDE-99 þ BDE-100 to the sum of BDE-153 þ BDE-154 has been used to assess the relative contributions of Penta-BDE and Octa-BDE to environmental contamination (Hites, 2004). In this study, the ratios ranged from 0.22 to 3.71 with a mean value of 0.91, indicating a slightly higher source contribution of Octa-BDE. The mean value in this study was comparable to those in the sediments from Laizhou Bay (0.97) (Jin et al., 2008), which may be an indication that sediments from Laizhou Bay are parts of the PBDE sources of the SYS via YS coastal currents. It is slightly lower than that in the sediments from the modern Yellow
Fig. 4. Percentages of tri-to hepta-BDE congeners to ments of the SYS.
P
7PBDEs
2101
in the surface sedi-
River Delta (1.41) (Chen et al., 2012), but much lower than PRD, South China (10.73) (Zou et al., 2007), USA (11.35) (Yun et al., 2008), €m et al., 2005) and the technical Penta-BDE Sweden (8.74) (Sellstro mixtures DE-71 and Bromkal 70-DE (10.01e11.93) (La Guardia., 2006), suggesting that the levels of the usage of Penta-BDE was low to moderate in our studied area. Although Deca-BDE was the predominant source of PBDE contamination in this area, the result also showed the minor contributions from Penta-BDE and Octa-BDEs. As shown in Fig. 4, the proportions of BDE-154 was distinctively higher at most sites. This could be attributable to the fact that Octaand Penta-BDEs, both of which contains BDE-154, was widely used in this area. And BDE-183 was prone to converting to BDE-154 under anaerobic bacteria condition (Davis and Stapleton, 2009). In addition, unlike the coastal sediments and especially the river sediments, the sediments in the SYS come from different sources, including atmospheric deposition as well as hydrodynamic transport, which provided a lot of fine particles. BDE-154 preferentially affiliated with fine-grained particles due to the higher Kow and lower rate of desorption from the fine particles (Rayne et al., 2003). It is interesting to note that distinctly higher proportions of BDE28 in sediments were observed than any commercial mixtures (Figs. 4 and 5). The degradation of higher PBDE congeners is a possible reason. Some studies had shown that the formation of BDE-28 was attributed to possible photodegradation of higher brominated congeners during atmospheric transportation (Bezares-Cruz et al., 2004), reductive debromination catalyzed by iron and iron sulfides (Keum and Li, 2005) and debromiantion by microorganism (Zhu et al., 2014b). PBDE congener patterns in most of the sites in the SYS presented only slight variations (Fig. 4), indicating these congeners in this study originated from the common sources or went through similar processes such as transformation (degradation) of PBDEs. The significant correlations among PBDE congeners (except for BDE-28) (R2 ¼ 0.476e0.864, p < 0.01) (Table S2) also suggested that the PBDEs in this area were likely to have similar sources. It is probably due to the fact that there is no large local coastal river that directly input to the SYS. Consequently, the impact of point source is negligible, and the mixture of hydrodynamic transportation and atmospheric deposition may be the major sources of PBDEs in the sediments of the SYS. Compared with PBDEs in coastal sediments, there is more time allowed for PBDEs in the SYS to achieve equilibrium partitioning with the surrounding water phase. And PBDEs were prone to have the similar patterns during these processes. PBDE congener patterns of the sediments in the SYS, BS (Pan et al., 2010), ECS (Li et al., 2012), PRD (Mai et al., 2005), Laizhou Bay (Pan et al., 2011), the modern Yellow River (MYR) (Chen et al., 2012), YRD (Chen et al., 2006) and commercial mixtures (La
2102
G. Wang et al. / Chemosphere 144 (2016) 2097e2105
Fig. 5. Comparison of PBDE congener patterns in the sediments of the SYS with commercial mixtures (DE-71, 70-5DE and 79-8DE) (La Guardia et al., 2006) and other regions: BS (Pan et al., 2010), East China Sea (Li et al., 2012), Pearl River Delta (Mai et al., 2005), Laizhou Bay (Pan et al., 2011), the modern Yellow River (Chen et al., 2012) and YRD (Chen et al., 2006).
Guardia et al., 2006) were shown in Fig. 5. It is seen that the congener patterns in this study were similar to BS, Laizhou Bay and MYR, especially to the BS (except for a slightly higher content of BDE-154), suggesting that particles from BS may be the main sources of PBDEs in the SYS. It is known that most of the mud deposits in the central SYS are considered to be derived primarily from the MYR and the old Yellow River subaqueous delta based on the circulation pattern in the SYS (Yang and Youn, 2007). Laizhou Bay is one of the three major bays of BS, and the biggest manufacturing base for BFRs in China are scattering around the bay. Pollutants containing PBDEs were emitted and discharged and/or deposited to the Laizhou Bay. According to Pan et al. (2010, 2011) and Chen et al. (2012), the MYR and Laizhou Bay are the important sources for PBDEs in the BS. Consequently, the MYR and Laizhou Bay are also the main sources of PBDEs in the SYS. However, all of them were significantly different from those of commercial mixtures (i.e., DE-71, 70-5DE and 79-8DE), indicating that a mixing of different sources of commercial products or transformation may have occurred for PBDEs during the transportation or after deposition in the SYS. 3.4. Source identification Principal component analysis (PCA) (varimax rotated) was performed on the 8 individual PBDEs (the concentrations) and 24 sediment samples to investigate the sources of PBDE contamination. Firstly, PCA extracted PCs to explain the maximum variances among the 8 PBDEs in 24 sediment samples. Then, coupled with the distributions of the scores of each PC in 24 sediment samples, we looked for the maximum source of each PC in the SYS and minimized the interferences among the PCs. The score plots of apparent groups were presented in Fig. S2. Four PCs were extracted, accounting for 91.1% of the total variances, with 27.4%, 26.4%, 22.4% and 14.9% for PC1, PC2, PC3 and PC4, respectively. PC1 was dominated by BDE-47, BDE-183, BDE-209. As well known, they are indicative congeners of commercial Penta-, Octaand Deca-BDEs, respectively. Consequently, PC1 can be considered as the indicator of commercial Penta-, Octa- and Deca-BDEs. As for BDE-183 and BDE-209, they are high bromo-compounds with
higher values of Kow (log Kow z 7.14e9.97) and shorter range of transportation (Wania and Dugani, 2003). BDE-47 has a higher water solubility (ca. 1.02 104 mol/m3) and a smaller Kow (log Kow z 6.11) (Wania and Dugani, 2003), and thus can easily reach the equilibrium partitioning between water column and sediments during the long time of transport. These congeners from surface runoff can be easily adsorbed to coastal sediments and were transported by the migration of sediments. A distinctive feature was shown in Fig. 6a: PC1 mainly influenced the northern region, and the sample scores were significantly higher in the Shandong peninsular coastal areas. Therefore, PC1 represented the surface runoff and hydrodynamic transport of commercial products, accounting for 44.4% (calculated from the MLR) of the total PBDEs in the SYS. PC2 was dominated by BDE-99 and BDE-153. These two congeners were the major components of commercial Penta-BDE (Fig. 5). They had relatively moderate octanoleair partition coefficients (log Koa z 11.31e11.82) (Wania and Dugani, 2003). The affected areas were located at central Yellow Sea mud (CYSM) and near the southwestern Cheju Island mud (SWCIM) (Fig. 6b). Although the CYSM is composed of fine-grained muds, such a low accumulation rate in this area (
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Polybrominated diphenyl ethers in sediments from the Southern Yellow Sea: Concentration, composition profile, source identification and mass inventory Guoguang Wang a, Jialin Peng a, Xiang Xu b, Dahai Zhang a, Xianguo Li a, * a b
Key Laboratory of Marine Chemistry Theory and Technology, Ministry of Education, Ocean University of China, Qingdao, 266100, China College of Chemistry and Pharmaceutical Sciences, Qingdao Agricultural University, Qingdao, 266109, China
h i g h l i g h t s BDE-209 is a dominant congener in the SYS, indicating high usage of Deca-BDE in China. PBDE concentrations increased from the coastal areas to the central mud area. Continental runoff and atmospheric deposition contributed 69.0% and 31.0% of PBDEs in the SYS, respectively. The SYS is a relatively weak sink of PBDEs in the world.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 July 2015 Received in revised form 20 October 2015 Accepted 22 October 2015 Available online xxx
The Southern Yellow Sea (SYS) is believed to be influenced by the contaminants from mainland China and the Korean peninsula. Here we report the first record about concentrations of polybrominated P diphenyl ethers (PBDEs) in the sediments of the SYS. The concentrations of 7PBDEs (BDE-28, 47, 99, 1 100, 153, 154, 183) and BDE-209 were 0.064e0.807 ng g (dry weight) and 0.067e1.961 ng g1 with a mean value of 0.245 ng g1 and 0.652 ng g1, respectively. These are distinctively low compared with the PBDE levels previously reported in other regions of the world. PBDE concentrations gradually increased from the coastal areas to the central mud area. BDE-209 was the dominant congener, accounting for 70.2 e91.6% of the total PBDEs. Congener profiles of PBDEs were similar to those in sediments from the Bohai Sea (BS), Laizhou Bay and modern Yellow River, which might be a tentative indication that they shared similar sources. Principal component analysis (PCA) revealed that PBDEs in the SYS were mainly from continental runoff (69.0%) and atmospheric deposition (31.0%). Depth profile of PBDEs in a sediment core collected from the edge of the central mud area showed that concentration of BDE-209 rapidly increased in recent years, which is in accordance with the replacement in demand and consumption of Penta- and Octa-BDEs by the Deca-BDE. Compared with BS, East China Sea, Erie and Ontario, the SYS was a relatively P weak sink of PBDEs (0.102e1.288 t yr1 for 7PBDEs and 0.107e3.129 t yr1 for BDE-209) in the world. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Polybrominated diphenyl ethers Distribution Composition profile Source identification Mass inventory The Southern Yellow Sea
1. Introduction Polybrominated diphenyl ethers (PBDEs), a class of brominated flame retardants (BFRs), have been widely used in the world since 1970s (Li et al., 2006a). They are added in a variety of consumer products such as electronic components, plastics, textiles, furnishing foam and fabrics to reduce the fire hazards by interfering
* Corresponding author. E-mail addresses: [email protected] (G. Wang), [email protected] (X. Li). http://dx.doi.org/10.1016/j.chemosphere.2015.10.088 0045-6535/© 2015 Elsevier Ltd. All rights reserved.
with the combustion of the polymeric materials (USEPA, 2008). Categorized by the average bromine contents, Penta- (predominantly BDE-47 and BDE-99), Octa- (predominantly BDE-183) and Deca-BDEs (predominantly BDE-209) are three main commercial PBDE products. In China, PBDEs are one group of the most produced and commonly used BFRs. The domestic production of BFRs in 2000 was up to 1000 tons and Deca-BDE accounted for the major proportion (Mai et al., 2005). In addition, BFRs imported from other countries also contributed to a portion (although the exact amount is unknown) of the market in China, because BFRs produced in three of the largest manufactures (i.e., Great Lakes Chemical,
2098
G. Wang et al. / Chemosphere 144 (2016) 2097e2105
Indianapolis, IN; Albemarle Chemical, Richmond, VA; and Dead Sea Chemical, Beer-Sheva, Israel) were sold to China. Horribly, for China, the annual demand of BFRs had been increasing at an estimated rate of 8% in 2000 (Mai et al., 2005). PBDEs can be easily released into surrounding environments during the production, usage and dismantling processes, because they are added in products without chemical bonding (Rahman et al., 2001). Consequently, PBDE residues have been widely € ller et al., 2011; Melymuk et al., 2012), dusts found in air (Mo (Whitehead et al., 2013), soils (Cetin, 2014; Shi et al., 2014), sediments (Lee et al., 2014; Zhu et al., 2014c), water (Hui-mei et al., 2013), marine organisms (Li et al., 2010; Zhu et al., 2014a) and human tissues (Mitchell et al., 2012; Chen et al., 2014). Due to their chemical stability, tendency for bioaccumulation and potential toxicity to the ecosystem, PBDEs have caused increasing attention around the world. In animal tests, some congeners of PBDEs were related with endocrine disruption, thyroid hormone disorders, reproductive/developmental toxicity and carcinogenicity (McDonald, 2002). Because of these adverse effects, the production and usage of Penta- and Octa-BDEs were banned in Europe and some factories in the USA voluntarily phased out PBDE manufacturing in 2004. PBDEs, like other persistent organic pollutants (POPs), can be transported to the coastal areas via a variety of processes including atmospheric deposition, land runoff and direct industrial and domestic wastewater discharges. Because of their lipophilic, organophilic and hydrophobic characteristics, PBDEs adsorb strongly to soils and sediments. A few studies have focused on surveying PBDEs in sediments from the coastal seas adjacent to mainland China (Mai et al., 2005; Chen et al., 2006; Pan et al., 2010; Li et al., 2012). However, to date, no studies of PBDEs contamination in sediments of the Southern Yellow Sea (SYS), lying between the Chinese mainland and the Korean Peninsula, have been reported. Although the Yellow River and Yangtze River don't input directly into the SYS, the two rivers distinctly inflow the sources of sediments into the SYS (Lim et al., 2007). According to the previous reports (CSOA, 2007; MEPPRC, 2007), about 7.60% of the industrial and domestic toxic waste materials generated by land-based activities ended in the SYS. Consequently, the objectives of this study are to determinate the PBDE concentrations, as well as the spatial and temporal distribution, to identify PBDE sources and to estimate the mass inventory of PBDEs in the SYS. 2. Materials and methods
Standard solutions of 2,4,5,6-tetrachloro-m-xylene (TCmX) and PCB-209 were used as recovery surrogates. BDE-77 was added as the internal standard for quantification of targeted PBDEs. All solvents including methanol, n-hexane and isooctane were of HPLC grade. The neutral silica gel was heated to 180 C and held for 3 h, cooled down to room temperature, then deactivated with 5% (w/w) distilled water before being stored in sealed containers. Sodium sulfate was baked at 450 C for 4 h and preserved in sealed containers after cooling. 2.3. Extraction and analysis The sample extraction was performed based on an ultrasonically assisted alkaline hydrolysis extraction method established previously in our laboratory (Wang et al., 2015). Briefly, 10 g of freezedried and ground sediment samples were spiked with TCmX and PCB-209 and processed in an ultrasonic bath in 2.0 M NaOHmethanol solution for 30 min. Activated copper was added to remove sulfur during the extraction. Then 50 mL distilled water and 50 mL n-hexane, in turn, were added. The extracts of organic phase were separated from the mixture and concentrated to 2e3 mL using a rotary vacuum evaporator. Concentrated extracts were cleaned and fractionated by a chromatographic column packed with neutral silica gel (7 g, 5% water deactivated) and anhydrous sodium sulfate (3 g) from bottom to top. The PBDE fractions were eluted with 70 mL n-hexane, concentrated to about 0.5 mL and dried out under a gentle stream of nitrogen. Finally, all the extracts were solvent-exchanged to isooctane to a constant volume of 100 mL with BDE-77 as the internal quantification standard. Sample analysis was performed with a Shimadzu 2010plus gas chromatograph coupled with electron capture detector (GC-ECD). A DB-5MS (30 m 0.25 mm i.d. 0.25 mm film thickness) capillary column was used for the determination of PBDE congeners except for BDE-209. The oven temperature was initiated at 100 C for 1 min, ramped at 10 C/min to 190 C, 5 C/min to 255 C for 0.5 min, then up to 300 C at 2 C/min and held for 10 min. Automatic injection of 1 mL sample was conducted in the splitless mode. Helium was used as the carrier gas at a flow rate of 3 mL min1. Injector and detector temperatures were fixed at 290 C and 315 C, respectively. For BDE-209, a DB-5 (15 m 0.25 mm i.d. 0.10 mm film thickness) capillary column was used. The oven temperature was initiated at 100 C for 1 min, then ramped at 8 C/min to 310 C and held for 24 min. Injector and detector temperatures were fixed at 300 C and 325 C, respectively.
2.1. Sample collection
2.4. Quality assurance/quality control
A detailed illustration of the sampling sites is presented in Fig. 1. A total of 24 surface sediments (0e3 cm) were collected by using a Van Veen grabber in the SYS during November to December in 2013. In addition, a sediment core of approximately 36 cm long was collected at F4 site (122 000 0000 E, 35 000 0000 N, the red circle in Fig. 1) from the edge of the central mud area, and cut in 2 cm interval to yield 18 core sediment samples. All of the sediments were wrapped in solvent-rinsed aluminum foil, transported to the laboratory on ice and stored at 20 C until further analysis.
For each batch of 10 samples, a procedural blank (in which a filter paper identical to that used to wrap the sediments was solvent-extracted and processed in the same way as the samples), a matrix-spiked sample (a sediment spiked with 8 PBDE congeners, together with the same sediment without PBDEs spiked, was used for calculation of matrix-spiked recovery) and a sample duplicate were processed. Results showed that only trace amounts of BDE-47 and BDE-99 were found in procedural blanks, and they were appropriately subtracted from those in the sample extracts. The surrogate recoveries in 24 samples ranged from 77.3% to 110.8% with an average of 89.4 ± 10.1% for TCmX and from 84.4% to 111.6% with an average of 93.7 ± 9.5% for PCB-209. The mean recoveries in spiked matrix samples were 83.2% ± 13.5% for all targeted PBDE congeners except that of BDE-209, which was 73.7 ± 15.2%. Reported concentrations were not surrogate recovery corrected. Relative standard deviations (RSDs) for duplicated samples were 11.7%e20.4% with an average of 15.2%.
2.2. Materials A standard mixture of PBDEs (BDE-CSM) selected for quantitative analysis were purchased from AccuStandards, Inc. (USA). The BDE-CSM contains 8 PBDE congeners (IUPAC Nos. 28, 47, 99, 100, 153, 154, 183 and 209) which are treated as the compounds of priority control by US Environmental Protection Agency (USEPA).
G. Wang et al. / Chemosphere 144 (2016) 2097e2105
2099
Fig. 1. Locations of sampling sites (red dots) in the SYS. Circulation systems (dark blue arrows) and mud areas (light purple) are modified after Liu et al. (2007) and Cheng et al. (2004). (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this article.)
2.5. Source identification Principal component analysis (PCA) - multiple linear regression (MLR) did not need a priori knowledge of the number of sources and their compositions. So that it was widely used to identify the sources of pollutants in the environment. In this paper, PCA was conducted using factor extraction with the total variances >90% based on correlation matrix. Detailed descriptions were presented by Çamdevýren et al. (2005). 3. Results and discussion 3.1. Concentrations Summary of the PBDE concentrations in surface sediments from the SYS is presented in Table 1. Most of the PBDE congeners were detected in all sediment samples, except for BDE-28, 153 and 183 with the detection frequencies of 95.8%, 58.3% and 87.5%, respectively, indicating that PBDEs were widespread contaminants in the P SYS. Concentrations of 7PBDEs (sum of BDE-28, 47, 99, 100, 153, 154, 183) ranged from 0.064 to 0.807 ng g1 dry weight with a mean value of 0.245 ng g1. BDE-209 concentrations ranged from 0.067 to 1.961 ng g1 dry weight with an average of 0.652 ng g1, P and were 1e3 times higher than those of 7PBDEs, indicating that BDE-209 was the predominant congener of PBDE pollutants in the SYS sediments, which was consistent with the usage status of DecaBDE in China. PBDE concentrations in the SYS sediments were compared with
3.2. Spatial and temporal distributions
Table 1 Concentrations of 8 PBDE congeners in the SYS surface sediments. Congeners
BDE-28 BDE-47 BDE-99 BDE-100 BDE-153 BDE-154 BDE-183 P c 7PBDEs BDE-209 a
Concentrations (ng/g) a
Min.
Max.
Median
Mean
ndb 0.013 0.006 0.010 nd 0.010 nd 0.064 0.067
0.049 0.069 0.091 0.088 0.106 0.620 0.186 0.807 1.961
0.008 0.031 0.016 0.024 0.004 0.081 0.037 0.243 0.819
0.007 0.031 0.023 0.026 0.006 0.082 0.029 0.245 0.652
Mean-geometric mean value of 24 samples. nd-not detected, which was calculated as zero. P 7PBDEs-the sum concentrations of BDE-28, 47, 99, 100, 153, 154 and 183 in each site. b c
those reported for several areas from China and other countries (Table S1). Although the numbers of PBDE congeners analyzed in various studies were different, BDE-28, 47, 99, 100, 153, 154, 183 and 209 in sediment samples have been widely found. The PBDE concentrations (without BDE-209) reported in our study were comparable to those in sediments from Bohai Sea (BS), China (Pan et al., 2010), Yangtze River Delta (YRD), China (Chen et al., 2006), Korea coast (Moon et al., 2007) and Tokyo Bay, Japan (Minh et al., 2007), while slightly lower than South China Sea, China (Mai et al., 2005), East China Sea (ECS), China (Li et al., 2012), Xiamen areas, China (Li et al., 2010), Taiwan (Jiang et al., 2011), Goseong, Korea (Lee et al., 2014), Spain (Eljarrat et al., 2005), and Great Lakes, USA (Song et al., 2005), and much lower than Hong Kong mangrove (Zhu et al., 2014c), Pearl River Delta (PRD), China (Chen et al., 2013) and Lake Shihwa, Korea (Lee et al., 2014). For BDE-209, the levels in our study were comparable to those in South China Sea, China (Mai et al., 2005) and Taiwan (Jiang et al., 2011), but much lower than those in rapidly industrialized and urbanized areas such as ECS, China (Li et al., 2012), BS, China (Pan et al., 2010) and PRD, China (Chen et al., 2013), Korea coast (Moon et al., 2007), Goseong, Korea (Lee et al., 2014), Lake Shihwa, Korea (Moon et al., 2012), Great Lakes, USA (Song et al., 2005), Spain (Eljarrat et al., 2005) and Tokyo Bay, Japan (Minh et al., 2007). These results indicated a relatively low to moderate level of PBDE contamination in the SYS at present. However, considering the adverse effect of PBDEs on ecological system and human health, a systematic monitoring and risk assessment programs are desirable in the SYS.
3.2.1. Spatial distribution P Concentrations of 7PBDEs and BDE-209 both increased from the coastal areas to the central mud area (see Fig. 2). Similar observations were also reported for organo-chlorine pesticides (OCPs; Hu et al., 2011a, b) and polychlorinated biphenyls (PCBs; Duan et al., 2013) in this area. Organic pollutants with high lipotropy and hydrophobicity (such as PBDEs, PCBs, PAHs and so on) are usually considered to be preferentially bounded to particles with higher TOC (total organic carbon) contents and smaller particle sizes (Cornelissen et al., 2005). According to Hu et al. (2011b) and Duan et al. (2013), TOC contents increase, and the grain size decreases from the coastal areas to the central mud area in the SYS. ConseP quently, the distributions of 7PBDEs and BDE-209 may be associated with the distributions of TOC and gain size.
2100
G. Wang et al. / Chemosphere 144 (2016) 2097e2105
Fig. 2. Spatial distributions of
P
7PBDEs
(a) and BDE-209 (b) in surface sediments of the SYS using Kriging by Surfer 11.0.
P Over the entire studied area, the highest levels of 7PBDEs and BDE-209 were all observed at the central mud area associated with smaller particle size, higher TOC and TON (total organic nitrogen) contents (Duan et al., 2013). This TOC dependent post-depositional sorption environment in the SYS was usually confined by the homogenous environmental conditions (Hu et al., 2011a, b). Hu. (1984) and Shi et al. (2002) reported that the YS cold water mass, YS warm current and the coast currents dominated the sedimentary environments of the central mud area in the SYS. It is well known that there was no local coastal river directly inputting to the central mud deposits of the SYS, and the SYS was far away from the direct sources in the BS. Consequently, the effects of shelf mud deposition (i.e., the hydrodynamic transport and depositional mechanism of the fine-grained sediments) and atmospheric deposition (Hu et al., 2011a; Zhao et al., 2013) were the main factors influencing the distribution and fate of PBDEs in the SYS. During these processes, PBDEs associated with fine-grained particles could be transported to the central mud area together with the particuP late matters. However, the lowest levels of 7PBDEs and BDE-209 were found at 32 N and 34 N, respectively. In the area near 32 N, a thick layer of sand sediments with lower TOC contents was deposited due to the resuspension and transportation of sediments from the old Yellow River and Yangtze River (Zhang et al., 2007). According to Marinari et al. (2006) and Meyers (1997), the TOC/TON ratio is used to reflect the sources of OMs in marine sediments, and the ratio varies from 4 to 10 in the OMs released by marine phytoplankton, while the ratio exceeds 10 in terrestrial OMs. In this study, the ratio ranged from 5.4 to 7.7 in the area near 34 N (Duan et al., 2013), indicating the OMs mainly deriving from marine sources. Additionally, the sediments mainly originated from the old Yellow River Delta in this area (Milliman et al., 1986). PBDEs have been used as BFR since 1970s and mainly derive from anthropogenic activities (Li et al., 2006a). Consequently, these led to the lower levels of BDE-209 in the area near 34 N. Relatively higher levels of PBDEs were also observed in the sediments near the Yangtze River Estuary (YRE). This can be attributed to the direct waste discharges and surface runoff into the Yangtze River (Hui et al., 2009), which runs through lots of highly populated and urbanized cities, such as Shanghai, Nanjing and so on. And those wastes deposited at the YRE are major contributors to pollutants (Feng et al., 2004) such as PBDEs. 3.2.2. Temporal distribution A sediment core collected from the edge of the central mud area was analyzed to examine PBDE temporal distribution. The historical P records of 7PBDEs and BDE-209 in Fig. 3 showed that PBDEs were found in all sediment layers, indicating that PBDEs had existed in
the sediments of the SYS and the areas continuously received inputs of PBDEs throughout years. P Concentrations of the 7PBDEs (Fig. 3a) were not significantly different at the depth of 18 cme36 cm. It is not surprising that the demand of PBDEs was low at that corresponding time period due to PBDEs used as flame retardants since 1970s. In addition, vertical mixing such as turbulence during typhoon season (Zhu et al., 2014c; Drage et al., 2015) may cause homogenization of pollutants in the sediments. It was also possible that some PBDEs may have been degraded or debrominated by microorganisms and physio-chemical processes because of relatively long existing time at the bottom of the sediment core. From the depth of 18 cm-12 cm, concentrations increased sharply with an apparent peak value of 0.57 ng g1 dry weight observed at the depth of 12 cm. The widespread usages of technical Penta- and Octa-BDEs as the commercial BFR products might be responsible. The concentrations then decreased thereafter from the depth of 8 cm to the surface, which might be attributed to the banning of production and usage of Penta- and Octa-BDEs in recent years. The concentrations of BDE-209 (Fig. 3b) increased gradually from the bottom to the surface, suggesting that the input of BDE209 increased year by year. These increases coincided with the high consumption of Deca-BDE for the BFRs market in China. A distinctive increase was observed at the upper layers, indicating that large quantities of BDE-209 were possibly input in the studied area in recent years and/or an increase in the sedimentation rate P occurred in the top layers. Compared with 7PBDEs, the more rapid and obvious increase of the concentrations of BDE-209 at the upper layers indicated that the historical records of these two groups were different, and the demand of Penta- and Octa-BDEs was largely replaced by the Deca-BDE in recent years. Similar rapid increases of BDE-209 in sediments were also found in sediment cores collected from other regions in PRD (Chen et al., 2007) and Hong Kong mangrove sediments (Zhu et al., 2014c). As the production and usage of PBDEs continue in the studied area, the emission of PBDEs to the environment is unavoidable. So that the SYS must be closely monitored to have a better understanding of these problems. 3.3. Congener patterns As shown in Fig. S1, BDE-209 was the predominant congener in the surface sediments of the SYS, accounting for 70.2e91.6% of the total PBDEs in approximately 62.5% of all samples. The predominance of BDE-209 was widely reported in marine sediments from BS (Pan et al., 2010), coastal ECS (Li et al., 2012), PRD (Chen et al., 2013) and other regions of the world (Moon et al., 2012; Zhu
G. Wang et al. / Chemosphere 144 (2016) 2097e2105
Fig. 3. Historical profiles of
P
7PBDEs
(a) and BDE-209 (b) in a sediment core from the SYS.
et al., 2014c). This is expected due to the fact that the consumption and production of Deca-BDE were much higher than Penta- and Octa-BDEs around the world. In addition, BDE-209 has a higher octanolewater partition coefficient (log Kow z 10) than other congeners (Rahman et al., 2001) and was deemed to possess low bioavailability and adsorbed strongly to marine sediments, especially the coastal sediments. P Percentages of tri-to hepta-BDE congeners in 7PBDEs were presented in Fig. 4. Among the 7 PBDE congeners detected, BDE-154 and BDE-183 were the dominated congeners (except for sites E1 and E3), followed by BDE-47, BDE-100 and BDE-99, in agreement with other studies (Pan et al., 2010, 2011). The relatively high abundances of BDE-154 and BDE-183 are likely attributed to the usage of Octa-BDE, with BDE-183 often taken as indicative of the presence of Octa-BDE commercial formulation (La Guardia et al., 2006). While the abundances of BDE-47, BDE-99 and BDE-100, the main congeners of Penta-BDE, indicated the source contribution of Penta-BDE commercial formulation. The ratio of sum of BDE-47 þ BDE-99 þ BDE-100 to the sum of BDE-153 þ BDE-154 has been used to assess the relative contributions of Penta-BDE and Octa-BDE to environmental contamination (Hites, 2004). In this study, the ratios ranged from 0.22 to 3.71 with a mean value of 0.91, indicating a slightly higher source contribution of Octa-BDE. The mean value in this study was comparable to those in the sediments from Laizhou Bay (0.97) (Jin et al., 2008), which may be an indication that sediments from Laizhou Bay are parts of the PBDE sources of the SYS via YS coastal currents. It is slightly lower than that in the sediments from the modern Yellow
Fig. 4. Percentages of tri-to hepta-BDE congeners to ments of the SYS.
P
7PBDEs
2101
in the surface sedi-
River Delta (1.41) (Chen et al., 2012), but much lower than PRD, South China (10.73) (Zou et al., 2007), USA (11.35) (Yun et al., 2008), €m et al., 2005) and the technical Penta-BDE Sweden (8.74) (Sellstro mixtures DE-71 and Bromkal 70-DE (10.01e11.93) (La Guardia., 2006), suggesting that the levels of the usage of Penta-BDE was low to moderate in our studied area. Although Deca-BDE was the predominant source of PBDE contamination in this area, the result also showed the minor contributions from Penta-BDE and Octa-BDEs. As shown in Fig. 4, the proportions of BDE-154 was distinctively higher at most sites. This could be attributable to the fact that Octaand Penta-BDEs, both of which contains BDE-154, was widely used in this area. And BDE-183 was prone to converting to BDE-154 under anaerobic bacteria condition (Davis and Stapleton, 2009). In addition, unlike the coastal sediments and especially the river sediments, the sediments in the SYS come from different sources, including atmospheric deposition as well as hydrodynamic transport, which provided a lot of fine particles. BDE-154 preferentially affiliated with fine-grained particles due to the higher Kow and lower rate of desorption from the fine particles (Rayne et al., 2003). It is interesting to note that distinctly higher proportions of BDE28 in sediments were observed than any commercial mixtures (Figs. 4 and 5). The degradation of higher PBDE congeners is a possible reason. Some studies had shown that the formation of BDE-28 was attributed to possible photodegradation of higher brominated congeners during atmospheric transportation (Bezares-Cruz et al., 2004), reductive debromination catalyzed by iron and iron sulfides (Keum and Li, 2005) and debromiantion by microorganism (Zhu et al., 2014b). PBDE congener patterns in most of the sites in the SYS presented only slight variations (Fig. 4), indicating these congeners in this study originated from the common sources or went through similar processes such as transformation (degradation) of PBDEs. The significant correlations among PBDE congeners (except for BDE-28) (R2 ¼ 0.476e0.864, p < 0.01) (Table S2) also suggested that the PBDEs in this area were likely to have similar sources. It is probably due to the fact that there is no large local coastal river that directly input to the SYS. Consequently, the impact of point source is negligible, and the mixture of hydrodynamic transportation and atmospheric deposition may be the major sources of PBDEs in the sediments of the SYS. Compared with PBDEs in coastal sediments, there is more time allowed for PBDEs in the SYS to achieve equilibrium partitioning with the surrounding water phase. And PBDEs were prone to have the similar patterns during these processes. PBDE congener patterns of the sediments in the SYS, BS (Pan et al., 2010), ECS (Li et al., 2012), PRD (Mai et al., 2005), Laizhou Bay (Pan et al., 2011), the modern Yellow River (MYR) (Chen et al., 2012), YRD (Chen et al., 2006) and commercial mixtures (La
2102
G. Wang et al. / Chemosphere 144 (2016) 2097e2105
Fig. 5. Comparison of PBDE congener patterns in the sediments of the SYS with commercial mixtures (DE-71, 70-5DE and 79-8DE) (La Guardia et al., 2006) and other regions: BS (Pan et al., 2010), East China Sea (Li et al., 2012), Pearl River Delta (Mai et al., 2005), Laizhou Bay (Pan et al., 2011), the modern Yellow River (Chen et al., 2012) and YRD (Chen et al., 2006).
Guardia et al., 2006) were shown in Fig. 5. It is seen that the congener patterns in this study were similar to BS, Laizhou Bay and MYR, especially to the BS (except for a slightly higher content of BDE-154), suggesting that particles from BS may be the main sources of PBDEs in the SYS. It is known that most of the mud deposits in the central SYS are considered to be derived primarily from the MYR and the old Yellow River subaqueous delta based on the circulation pattern in the SYS (Yang and Youn, 2007). Laizhou Bay is one of the three major bays of BS, and the biggest manufacturing base for BFRs in China are scattering around the bay. Pollutants containing PBDEs were emitted and discharged and/or deposited to the Laizhou Bay. According to Pan et al. (2010, 2011) and Chen et al. (2012), the MYR and Laizhou Bay are the important sources for PBDEs in the BS. Consequently, the MYR and Laizhou Bay are also the main sources of PBDEs in the SYS. However, all of them were significantly different from those of commercial mixtures (i.e., DE-71, 70-5DE and 79-8DE), indicating that a mixing of different sources of commercial products or transformation may have occurred for PBDEs during the transportation or after deposition in the SYS. 3.4. Source identification Principal component analysis (PCA) (varimax rotated) was performed on the 8 individual PBDEs (the concentrations) and 24 sediment samples to investigate the sources of PBDE contamination. Firstly, PCA extracted PCs to explain the maximum variances among the 8 PBDEs in 24 sediment samples. Then, coupled with the distributions of the scores of each PC in 24 sediment samples, we looked for the maximum source of each PC in the SYS and minimized the interferences among the PCs. The score plots of apparent groups were presented in Fig. S2. Four PCs were extracted, accounting for 91.1% of the total variances, with 27.4%, 26.4%, 22.4% and 14.9% for PC1, PC2, PC3 and PC4, respectively. PC1 was dominated by BDE-47, BDE-183, BDE-209. As well known, they are indicative congeners of commercial Penta-, Octaand Deca-BDEs, respectively. Consequently, PC1 can be considered as the indicator of commercial Penta-, Octa- and Deca-BDEs. As for BDE-183 and BDE-209, they are high bromo-compounds with
higher values of Kow (log Kow z 7.14e9.97) and shorter range of transportation (Wania and Dugani, 2003). BDE-47 has a higher water solubility (ca. 1.02 104 mol/m3) and a smaller Kow (log Kow z 6.11) (Wania and Dugani, 2003), and thus can easily reach the equilibrium partitioning between water column and sediments during the long time of transport. These congeners from surface runoff can be easily adsorbed to coastal sediments and were transported by the migration of sediments. A distinctive feature was shown in Fig. 6a: PC1 mainly influenced the northern region, and the sample scores were significantly higher in the Shandong peninsular coastal areas. Therefore, PC1 represented the surface runoff and hydrodynamic transport of commercial products, accounting for 44.4% (calculated from the MLR) of the total PBDEs in the SYS. PC2 was dominated by BDE-99 and BDE-153. These two congeners were the major components of commercial Penta-BDE (Fig. 5). They had relatively moderate octanoleair partition coefficients (log Koa z 11.31e11.82) (Wania and Dugani, 2003). The affected areas were located at central Yellow Sea mud (CYSM) and near the southwestern Cheju Island mud (SWCIM) (Fig. 6b). Although the CYSM is composed of fine-grained muds, such a low accumulation rate in this area (
