Chemosphere 118 (2015) 293–300

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Trophic magnification of chlorinated flame retardants and their dechlorinated analogs in a fresh water food web De-Gao Wang a,⇑, Ming-Xing Guo a, Wei Pei a, Jonathan D. Byer b, Zhuang Wang c a

Department of Environmental Science and Engineering, Dalian Maritime University, Dalian, PR China Life Science and Chemical Analysis, Leco Corporation, 3000 Lakeview Ave., St. Joseph, MI, USA c Jiangsu Key Laboratory of Atmospheric Environment Monitoring and Pollution Control (AEMPC), School of Environmental Science and Engineering, Nanjing University of Information Science and Technology, Nanjing, PR China b

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

this food web as well as mirex and dechlorine plus.  Dechlorinated product has higher food web magnification potential than dechlorine plus.  Environmental risk assessment is needed because of biomagnification.

Concentration

 Chlordene plus was biomagnifying in

Chlordene Plus

Trophic level

a r t i c l e

i n f o

Article history: Received 27 March 2014 Received in revised form 14 September 2014 Accepted 18 September 2014 Available online 19 October 2014 Handling Editor: Myrto Petreas Keywords: Dechlorane Dechlorinated analogs Bioaccumulation Food web magnification

a b s t r a c t Chlorinated flame retardants, particularly dechlorane plus (DP), were widely used in commercial applications and are ubiquitous in the environment. A total of seven species of aquatic organisms were collected concurrently from the region of a chemical production facility in Huai’an, China. DP and structurally related compounds including mirex, dechloranes 602, 603, 604, chlordene plus (CP), DP monoadduct (DPMA), and two dechlorinated breakdown products of DP, decachloropentacyclooctadecadiene (antiCl10-DP) and undecachloropentacyclooctadecadiene (anti-Cl11-DP), were detected in these aquatic organisms. Nitrogen stable isotope ratios were also measured to determine the trophic levels of the organisms. Trophic magnification factors (TMFs) for these chemicals were calculated with values ranging from 1.0 to P 3.1. TMFs for CP, mirex, anti-DP, and DP were statistically greater than 1, showing evidence of biomagnification in the food web. Concentration ratios of anti-Cl11-DP to anti-DP showed a significant relationship with trophic level, implying that anti-Cl11-DP had a higher food-web magnification potential than its precursor. The biota-sediment accumulation factors and TMFs for DP demonstrated stereoselectivity, with syn-DP having a greater bioaccumulation potential than anti-DP in the aquatic environment. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction

⇑ Corresponding author. Tel./fax: +86 411 8472 3185. E-mail address: [email protected] (D.-G. Wang). http://dx.doi.org/10.1016/j.chemosphere.2014.09.057 0045-6535/Ó 2014 Elsevier Ltd. All rights reserved.

Chlorinated flame retardants (CFRs) such as dechlorane plus (DP) have been used in commercial polymer products like electrical wires and cables, plastic roofing materials, and connectors used

294

D.-G. Wang et al. / Chemosphere 118 (2015) 293–300

in computers and televisions for more than 40 years (Sverko et al., 2011). DP and the dechlorinated breakdown products of DP, decachloropentacyclooctadecadiene (anti-Cl10-DP) and undecachloropentacyclooctadecadiene (anti-Cl11-DP), have been detected in human blood, bird eggs, dolphin, and sediments (Sverko et al., 2008; de la Torre et al., 2012; Yan et al., 2012; Ben et al., 2013). Structurally related CFRs, including mirex, dechlorane (Dec) 602, Dec 603, Dec 604, chlordene plus (CP), and DP monoadduct (DPMA), have been reported in biota and other environmental matrices (Sverko et al., 2010, 2011; Shen et al., 2012; Suebring et al., 2013). Occurrences of CFRs in biota samples have increased recently throughout the world (Sverko et al., 2011; Xian et al., 2011; Feo et al., 2012). The increasing trend may be explained by two main factors: production and use increased starting in 2007 because of a new manufacturing facility in China, where production is about 3000 tons per year (Wang et al., 2010); and (2) their physical–chemical properties and persistence in the environment tend to accumulation in the food web with increasing trophic level (TL) (Wu et al., 2010). Bioaccumulation is an inherent property expressing a chemical’s capacity to accumulate in organisms and is an important factor for risk assessment to the environment and human health. Bioaccumulation is referred to as a process in which the chemical concentration in an organism achieves a level that exceeds that in the respiratory medium, the diet, or both. Bioaccumulation of chemicals can be assessed by several parameters such as the bioconcentration factor (BCF), bioaccumulation factor (BAF), biomagnification factor (BMF), and Trophic magnification factors (TMFs). TMF is the ‘‘gold standard’’ for assessing biomagnification potential of Persistent Organic Pollutants (POPs) for those chemicals that have been in commerce long enough to detect them in environmental samples. TMFs represent the average prey to predator transfer of POPs through food webs, rather than the individual species biomagnification metrics that are highly variable from one predator–prey combination to another (Gobas et al., 2009). The bioaccumulation and biomagnification potential of DP in freshwater food webs have been well documented in several recent studies (Tomy et al., 2007, 2008; Wu et al., 2010; Zhang et al., 2011). However, there is little information about the bioaccumulation potential of the dechlorinated analogs of DP and other related compounds. Recently, we reported very high concentrations of DP in air, soil, sediment and biota samples collected from a new DP manufacturing facility in China (Wang et al., 2010). Additionally, some related compounds including Mirex and Dec 602 were measured in these samples. These results spurred our efforts to examine a series of additional related DP compounds, including Dec 603, Dec 604, CP, DPMA, and two dechlorinated analogs of DP, anti-Cl10-DP (aCl10DP) and anti-Cl11-DP (Cl11DP), in biota from a canal in South of China to assess and compare their bioaccumulation processes in a fresh water food web. The major objectives of the present study were to (1) determine concentrations of CFRs in different biota samples collected from the region of the manufacturing facility, and (2) to evaluate the bioaccumulation and biomagnification potential of these compounds in the freshwater aquatic environment. 2. Materials and methods 2.1. Sample collection A total of seven species of aquatic organisms were collected concurrently in Huai’an in Jiangsu province of China in May, 2010, from the Beijing-Hangzhou Grand Canal. Wild aquatic species included two invertebrates, river snail (Viviparus) and freshwater shrimp (Macrobrachium nipponense); five fish species, bleeker (Pseudolaubuca sinensis), loach (Misgurnus anguillicauda-

tus), crucian carp (Carassius auratus), common carp (Cyprinus carpio), and northern snakehead (Channa argus).To eliminate individual diversity, composite samples were used comprised of five individuals of the same species. Detailed information on the biota sample name, number, and lipid content are highlighted in Table 1. After collection, all samples were stored at 20 °C until chemical analysis. 2.2. Chemicals and reagents All solvents used were of pesticide grade purity (J.T. Baker, USA). Silica gel (pore size 60 Å, 70–230 mesh) was purchased from Merck (Merck, Germany). Dec 602 (95%), Dec 603 (98%), and Dec 604 (98%) were purchased from Toronto Research Chemical Inc. (Toronto, ON, Canada). CP, DP, DPMA, aCl10DP, and aCl11DP were obtained from Wellington Laboratories Inc. (Guelph, ON, Canada), and Mirex was obtained from Cambridge Isotope Laboratories Inc. (Andover, MA). The surrogate and internal standards used for all compounds were 2,20 ,4,40 ,6,60 -hexachlorobiphenyl (CB-155) and octachloronaphthalene (OCN) purchased from Accustandard Inc. (New Haven, CT). 2.3. Extraction and analyses Fish muscle tissues were homogenized individually. Soft tissues of two freshwater shrimps and five river snails for each group were homogenized for CFR concentration analysis. Approximately 5 g of each sample was mixed with ashed anhydrous sodium sulfate, spiked with the surrogate standard CB-155, and Soxhlet-extracted in a mixture of hexane/acetone (1:1, v/v) for 24 h. An aliquot of each extract was removed for gravimetric lipid determination, while the remaining extract was added to a separatory funnel and treated three times with 98% H2SO4 for lipid removal. The lipid-free eluate was purified using fully activated neutral silica (7 g) capped with anhydrous sodium sulfate (2 g). The column was prewashed with 60 mL of dichloromethane (DCM):hexane (1:1, v/v). Then 2 mL of extract was loaded onto the column, and the target compounds were eluted using 70-mL of DCM:hexane (1:1, v/v). The eluant was rotary-evaporated to approximately 4 mL, solvent-exchanged into isooctane and reduced to 3). The names, structures, retention times, mass ions, and ion ratios are summarized in Appendix (Table S1). 2.4. Stable isotope analysis For stable isotope analysis, muscle tissues of the aquatic organisms were freeze-dried and ground to homogeneous powders with a mortar and pestle. A 0.8–1.0 mg sample was weighed into a tin

295

capsule for the simultaneous determination of nitrogen ratios. Nitrogen isotope ratios were measured using an elemental analyzer-isotope ratio mass spectrometer (Delta V Advantage; Thermo Fischer, USA). Two replicates of each sample were analyzed, and the relative standard deviation was less than 0.2%. Stable isotope ratios were expressed in d notation as parts per thousand (‰) deviations from the international standards AIR (nitrogen), according to the Eq. (1),

d15 N ¼ ½ð15 N=14 NÞSample =ðð15 N=14 NÞAir  1Þ  1000‰

ð1Þ

2.5. Trophic magnification factor calculations The relative TL of each sample was calculated from d15N by using an enrichment factor (DN) of 3.4‰. TL was determined relative to the primary consumer (snail), which was assigned a TL of 2 (Fisk et al., 2001; Post, 2002). For each individual sample of invertebrates and fish, TL was determined using the following equation (Eq. (2)):

TLconsumer ¼ 2 þ ½ðd15 Nconsumer  d15 Nprimary consumer Þ=DN

ð2Þ

15

The values of d N for each sample were measured to calculate the TL (Table 1). TMFs were estimated as the slope (b) of the lipid normalized contaminant concentration regressed onto the TL (Eqs. (3) and (4)).

Ln ½Concentration ¼ a þ bTL

ð3Þ

TMF ¼ eb

ð4Þ

TMF < 1 implies that contaminants are not taken up by the organism or that they are metabolized; whereas, TMF > 1 indicates that contaminants are biomagnifying up the food chain. 2.6. Quality assurance/quality control Calibration standards were prepared in the range of 1– 50 ng mL1 except for DP isomers. Due to very high DP concentrations in all samples, a range of 1–50 000 ng mL1 for DP isomers standards were prepared to keep all concentrations within the linear range. CB-155 surrogate standard recoveries ranged from 52% to 85% and the final concentrations were corrected using the surrogate standard recoveries in all samples. Laboratory procedural blanks (Ashed anhydrous sodium sulfate) were used to

100000 10000 1000 100 10

C De P c6 0 De 2 c6 0 D 3 an ec 6 ti0 Cl 4 an 10-D tiCl P 11 -D P sy nDP an tiD To P tal DP

1

M ire x DP M A

Concentrations (ng g-1 lipid weight)

0.76 ± 0.08 0.51 ± 0.06 0.81 ± 0.05 0.72 ± 0.15 0.51 ± 0.06 0.81 ± 0.08 0.68 ± 0.05

fanti DP

85 700 ± 3300 92 600 ± 1600 64 400 ± 3700 10 500 ± 2600 93 000 ± 5600 19 300 ± 500 3010 ± 330 20 400 ± 6600 45 400 ± 4500 12 000 ± 2800 3200 ± 2700 45 500 ± 4500 3700 ± 160 968 ± 244 2340 ± 410 1010 ± 50 860 ± 90 126 ± 12 2470 ± 430 238 ± 33 144 ± 31 1.89 ± 0.38 329 ± 38 314 ± 22 3.28 ± 0.68 5.69 ± 1.96 150 ± 24 46.4 ± 21 1380 ± 180 4910 ± 760 2250 ± 500 982 ± 233 1960 ± 450 1720 ± 330 19 800 ± 3900 14.2 ± 3.1 65.5 ± 14.8 29.2 ± 4.5 31.4 ± 7.8 43.6 ± 9.7 17.3 ± 5.8 13.3 ± 4.9 3.2 ± 0.1 4.1 ± 0.1 3.1 ± 0.1 2.0 ± 0.2 3.2 ± 0.4 2.1 ± 0.1 3.3 ± 0.2 7 2 6 25 7 14 5 Common carp Snakehead Crucian carp River snail Bleeker Shrimp Loach

1.32 0.95 1.26 0.6 1.03 1.03 2.06

8.7 ± 0.7 11.7 ± 0.4 8.4 ± 0.5 4.6 ± 0.6 8.6 ± 1.3 4.9 ± 0.3 8.9 ± 0.9

21.1 ± 8.0 122 ± 29 29.0 ± 8.1 63.4 ± 10.1 29.2 ± 5.5 13.4 ± 4.9 79.3 ± 8.7

29.8 ± 6.5 128 ± 18 66.5 ± 10.4 38.0 ± 7.7 61.4 ± 10.7 38.3 ± 6.0 37.0 ± 3.8

13.3 ± 4.3 119 ± 23 79.8 ± 5.7 17.3 ± 3.0 35.6 ± 5.3 44.3 ± 5.7 29.6 ± 10.4

6.03 ± 0.96 232 ± 15 165 ± 15 12.7 ± 1.9 25.6 ± 8.0 100 ± 13 49.1 ± 14.5

66 700 ± 10 700 47 200 ± 6080 52 400 ± 5900 7300 ± 200 47 500 ± 7800 15 500 ± 1900 2040 ± 190

P anti-DP syn-DP aCl11DP aCl10DP Dec 604 Dec 603 CP Dec 602 Mirex DPMA TL d15N(‰) Lipid (%) n Sample

Table 1 Concentrations (mean ± standard error) of chlorinated flame retardants (ng g1 lipid weight), number of individual samples collected in this study (five snails and two shrimps were pooled in each analyzed group), lipid content (mean ± standard error), nitrogen stable isotopes (d15N), and trophic level (TL) for samples collected from the surrounding environment of the DP facility.

D.-G. Wang et al. / Chemosphere 118 (2015) 293–300

Fig. 1. Concentrations (ng g1 lw) of chlorinated flame retardants and dechlorinated analogs in the aquatic organisms collected from the DP manufacturing facility in China.

296

D.-G. Wang et al. / Chemosphere 118 (2015) 293–300

monitor background contamination from sample handing and processing. Procedural blanks were extracted and analyzed identically to the samples. A procedural blank was run for every 12 samples. Only trace levels of DP isomers were detected in the procedural blanks. The limits of quantification (LOQ) were calculated as 10 times signal-to-noise ratios (Table S1). 3. Results and discussion 3.1. DP and dechlorinated analogs Concentrations of DP and its dechlorinated analogs based on lipid weight (lw) in the aquatic organisms are given in Table 1 and Fig. 1. DP isomers were detected in all samples. Total DP P ( DP) concentrations in the collected samples varied from 3 010 to 93 000 ng g1 lw. The concentrations reflected point source contamination exposure of the aquatic organisms in this canal. These higher concentrations of DP compared to previous studies were attributed to direct emission from the manufacturing facility. Recently, Zhang et al. (2013) reported very high DP concentrations in blood and hair samples collected from the facility’s manufacturing workers (Zhang et al., 2013). DP can be dechlorinated into lower chlorinated DPs (Wang et al., 2013). Two of lower chlorinated DPs, aCl11DP and aCl10DP, were previously reported in sediment, fish, human hair, serum, breast milk samples (Zheng et al., 2010; Zhao et al., 2011; Ben et al., 2013; Sühring et al., 2013). In the present study, the average concentration of aCl10DP was 84.4 ng g1 lw, which was similar to those of DPMA, CP, Decs 602, 603, and 604 (paired t-test, all values of p > 0.05), for all organisms collected from this canal. The monodechlorinated product of DP, aCl11DP, had a mean concentration of 1030 ng g1 lw, which was 12 times higher than aCl10DP in the samples. The dechlorinated species can be formed by the photodegradation and/or thermal degradation of DP. The metabolism of DP in vivo and vitro has been examined (Tomy et al., 2008; ChabotGiguere et al., 2013) and no hypothesized dechlorinated species have been determined to date. Therefore, the occurrence of the dechlorinated analogs, aCl10DP and aCl11DP, in the aquatic organisms originated from bioaccumulation from the environment and/ or food rather than biotransformation from DP in vivo. These observations indicated that DP and its dechlorinated analogs could bioaccumulate in aquatic organisms, and have become pollutants in the region of the manufacturing facility. The concentration ratios of aCl11DP to anti-DP provide further insight into its bioaccumulation potential. The concentration ratios of aCl11DP to anti-DP in aquatic organisms ranged from 0.015 to 0.070 with a mean value of 0.032. These values are similar to previously reported ratios in human serum from e-waste recycling workers (mean of 0.017) (Ren et al., 2010). The ratios of aCl10DP to anti-DP ranged from 0.0001 to 0.02 with a mean value of 0.006. Limited quantitative data is available for aCl10DP in biota samples and no other studies could be compared. But the concentration ratios indicated that the concentrations of aCl10DP and aCl11DP were far lower than those of anti-DP, implying anti-DP were very stable in the environment. 3.2. Mirex Except for DP isomers, mirex concentrations were the highest measured CFR in the aquatic organisms, with values ranging from 982 to 19 800 ng g1 lw. Guerra et al. (2011) reported mirex concentrations in peregrine falcon eggs from Canada and Spain ranging between 94–1353 ng g1 lw and 2.4–78 ng g1 lw, respectively (Guerra et al., 2011). De la Torre et al. (2012) measured

mirex concentrations from 7.63 to 275 ng g1 lw in dolphins off the coast of Brazil (de la Torre et al., 2012). Comparatively, mirex concentrations in biota samples in the present study were much higher than in North America, likely because mirex use as a flame retardant and pesticide was banned in the 1980s in North America. However, mirex is a commonly used pesticide for termite prevention and control in China. Although, the production and use of mirex has been decreasing in China. Production was approximately 667 tons in 2005, and China has basically eliminated the production and use of mirex since 2009 (China, 2007). 3.3. Dechlorane plus monoadduct DPMA may be formed through partial reactions during the synthesis of DP through the diadduct Diels–Alder process. DPMA was detected in the aquatic organisms with concentrations from 13.3 to 65.5 ng g1 lw. Previous studies have attempted to analyze DPMA in eels from German (Sühring et al., 2013), white stork eggs from Spain (Muñoz-Arnanz et al., 2011), and marine biota samples from Chile (Barón et al., 2013). However, they found that DPMA was below their detection limits in these studies. To date, only one study has detected DPMA in biota; peregrine falcon eggs from Canada and Spain had high concentrations of DPMA with a geometric mean value of 21.1 ng g1 lw in Spain and 30.2 ng g1 lw in Canada (Guerra et al., 2011), which were close to the concentrations measured in aquatic organisms in this study. However, the ratios of DPMA:DP in the present study ranged between 0.02% and 0.44%, 3–4 orders of magnitude lower than in the peregrine falcon eggs (1.17–12.3%). This is likely because of differences in the source of DP contamination. In China, high DP contamination was from a point-source caused by the manufacturing facility; whereas, this was not the case in Spain or Canada. 3.4. Chlordene plus CP was detected in all fish samples, and the concentrations ranged from 29.8 to 128 ng g1 lw. These concentrations were in the same order as Decs 602, 603 and 604. To date, only one study has reported CP concentrations in biota samples. De la Torre et al. (2012) reported low CP concentrations (mean = 0.13 ng g1 lw) in half of the dolphin samples off the coast of Brazil (de la Torre et al., 2012). CP concentrations in the present study were much higher than those data reported in the dolphins. CP is not manufactured intentionally or used in China; however, CP occurs as a byproduct in technical chlordane at 0.40% (Shen et al., 2011b). Between 1997 and 2001, total chlordane use was about 476 tons in Jiangsu Province (China, 2007), where our sampling site was located. According to this CP ratio of 0.40% in technical chlordane, about two tons CP has been used in this province. The relatively high CP concentrations in this study could be linked to the high use of chlordane for termite prevention and control in buildings in this region. 3.5. Dechloranes 602, 603, and 604 Mean concentrations of Decs 602, 603 and 604 in the aquatic organisms collected in this study were 51.0, 48.4, and 121 ng g1 lw, respectively. These data were very similar to concentrations of DPMA, CP, and aCl10DP (Fig. 1). Shen et al. (2010) reported concentrations of Decs 602, 603 and 604 in fish samples collected from the Great Lakes in the range of 0.47–34 ng g1 lw, 0.014– 0.55 ng g1 lw, and 0.063–13 ng g1 lw, respectively (Shen et al., 2010); much lower than those in the present study. To our knowledge, China has not manufactured these three compounds to date, and no records are currently available about their use. Dec 603 and Dec 604 have been identified as impurities in organic chlorinated

297

D.-G. Wang et al. / Chemosphere 118 (2015) 293–300

pesticides (OCPs) (Shen et al., 2011b). For instance, Dec 603 was detected in technical products of aldrin and dieldrin. However, only synthetic experiments were carried out in China and there is currently no industrial production of aldrin and dieldrin (China, 2007). Dec 604 was an impurity (2%) in a commercial product of mirex. Accordingly, about one ton of Dec 604 was produced as a byproduct in this province. The relatively high concentrations of Dec 603 and Dec 604 in this study could be linked to the high use of OCPs for termite prevention and control in buildings in this region. These observations indicated that Decs 602–604 could bioaccumulate in aquatic organisms. These potential sources of CFRs in China were different than in North America, which originated from their use as organic chlorinated pesticides and not direct use as flame retardants. 3.6. Trophic magnification factor The food web structure of the Beijing-Hangzhou Grand Canal was elucidated using nitrogen stable isotopes (Table 1). The aquatic food web consisted of species ranging over three trophic levels. The average of d15N for river snail was 4.6‰, and was assumed as the primary consumer (TL = 2.0). Trophic levels differed among organisms, with three distinct groups apparent. The river snail and shrimp occupied a relative low trophic position (means TL of 2.0 and 2.1). The crucian carp, common carp, bleeker, and loach were in the median trophic position with the range of 3.1–3.3. Snakehead was in the top trophic position with the value of 4.1. TMFs have been used to assess the food web magnification for entire food webs and are based on the relationship between TL and lipid-normalized contaminant concentration. All TMFs for these chlorinated flame retardants ranged between 1.0 and 3.1 (Table 2), showing they had biomagnification potential in this food web. However, significant relationships were observed for CP P (p < 0.01), Mirex (p < 0.03), anti-DP (p < 0.04), and DP (p < 0.01) only (Fig. 2), suggesting that these chemicals are biomagnifying up the food chain. To our knowledge, this is the first report on the TMF of CP in a food web. No correlation was found between trophic level and lipid content, suggesting that biomagnification was not attributable to lipid content effects but indeed occurred. Mirex has been reported to biomagnify in many fresh water and marine food webs (Kelly et al., 2007; Peng et al., 2014). The biomagnification of DP in food webs also has been reported in some studies. Wu et al. (2010) reported TMFs for syn-DP (11.3), antiP DP (6.5) and DP (10.2) in a reservoir food web in an electronic waste (e-waste) recycling site in South China (Wu et al., 2010); these values were approximately 3–5 times higher than what was observed in the present study. Tomy et al. (2007) reported a TMF of 2.5 for anti-DP in the Lake Winnipeg food web (Tomy et al., 2007). In the same food web, syn-DP was found to be diluted with increasing trophic level. They also observed no significant relationships between trophic level and concentration for either Table 2 Regression results, trophic magnification factors (TMF) and 95% confidence internal (CI) for the chlorinated flame retardants examined in this study. Compound

TMF (95% CI)

Slope

r2

p

DPMA Mirex Dec 602 CP Dec 603 Dec 604 aCl10DP aCl11DP syn-DP anti-DP P DP

1.2 1.9 1.4 1.3 1.3 1.1 1.0 3.1 3.1 1.9 2.2

0.158 0.643 0.361 0.266 0.233 0.049 0.021 1.14 1.12 0.635 0.770

0.03 0.31 0.10 0.16 0.04 0.01 0.01 0.31 0.23 0.11 0.16

0.29 0.03 0.06 0.01 0.21 0.08 0.07 0.06 0.06 0.04 0.01

(0.87–1.6) (1.2–2.9) (0.99–2.1) (1.1–1.6) (0.87–1.8) (0.35–3.1) (0.53–1.97) (1.8–5.4) (1.6–5.8) (1.1–3.4) (1.2–3.9)

isomer in the Lake Ontario food web. These perceived differences may be explained by differences in species and their metabolism. TMFs of the other DP analogs measured in this study including the two dechlorinated DPs were not statistically different from 1, suggesting that these chemicals do not biomagnify in this food web. However, it is very difficult to be definitive that these compounds will not biomagnify in this food web based on TMFs only because their KOW are all larger than 105. In aquatic food webs, poorly metabolized hydrophobic chemicals with KOW between 105 and 108 generally have the highest biomagnifications potentials, while the biomagnifications potential of chemicals with KOW > 108 are dramatically decreased (Kelly et al., 2007). DP have been found to have the trophic biomagnifications potential in this food web with the KOW of 109.4, while the dechlorinated DPs were excepted to be higher biomagnifications potentials than DPs due to lower KOW with the less chlorine atoms. Recently, Peng et al. (2014) confirmed that the aCl11DP showed tropic magnification in a marine food web. However, they have not found the DPs biomagnify in this food web (Peng et al., 2014). In addition, the log-normalized concentration ratios of aCl11DP to anti-DP show a significant relationship with TL (Fig. 3, p < 0.001), implying that aCl11DP has a higher food web magnification potential than anti-DP. Many factors including the composition of the food web, the assimilation efficiency in the food web components, and even extrinsic conditions such as environmental levels may have led to different trophic biomagnification potentials among the food webs (Wu et al., 2009; Borgå et al., 2012a,b). 3.7. Isomeric rations of DP The isomer composition of DPs in the environmental samples varied from that of the technical products because of their biota isomer-selective uptake or elimination, biodegradation, and stereospecific photodegradation. The fraction of anti-DP (fanti) were calculated based on the concentrations of anti-DP divided by the sum of concentrations of DP isomers (Table 1). The values of fanti in the organisms ranged from 0.51 to 0.81 in this study, which were close to or lower than the values in commercial products of DPs (0.75–0.80) (Hoh et al., 2006; Qiu et al., 2007) or sediment samples (0.73) (Wang et al., 2010). The snakehead occupied the highest TL had the lowest fanti value of 0.51 and shrimp occupied the lowest TL had the highest fanti value of 0.81. A significant negative correlation was found between fanti and TL of the species (p < 0.01) (Fig. 3). The results can be attributed that the species have lower uptake efficiencies and higher depuration rate for the anti-DP compared to syn-DP (Tomy et al., 2008). This observation is consistent with that the fanti decreases with the TL in another freshwater food web in China (Wu et al., 2010) and strongly suggests that differential bioaccumulation of DP isomers. 3.8. Biota-sediment accumulation factors of DP The biota-sediment accumulation factor (BSAF) is a common empirical parameter describing accumulation of sediment-associated organic compounds into tissues of ecological receptors. The BSAF is calculated using Eq. (5),

BSAF ¼ C B =C S

ð5Þ 1

where CB is the concentration of DP in biota (ng g lipid wt) and CS is the concentration of DP in sediment (ng g1 organic carbon). Using the CB data from this study and the CS from our previous study reported on a dry weight basis and about 2% organic carbon content, BSAFs were estimated for syn- and anti-DP of aquatic species to sediments in the canal near the manufacturing facility. A theoretical BSAF value of 1.7 has been estimated based on partitioning of nonionic organic chemicals between tissue lipids and sediment

298

D.-G. Wang et al. / Chemosphere 118 (2015) 293–300

12.0

Ln Cocentrations

Ln Cocentrations

6.0 5.0 4.0 3.0 2.0 1.0

CP R = 0.16, p