Organophosphorus flame retardants in a typical freshwater food web: Bioaccumulation factors, tissue distribution, and trophic transfer

Journal Pre-proof Organophosphorus flame retardants in a typical freshwater food web: Bioaccumulation factors, tissue distribution, and trophic transfer Yin-E. Liu, Xiao-Jun Luo, Pablo Zapata Corella, Yan-Hong Zeng, Bi-Xian Mai PII:

S0269-7491(19)31342-9

DOI:

https://doi.org/10.1016/j.envpol.2019.113286

Reference:

ENPO 113286

To appear in:

Environmental Pollution

Received Date: 14 March 2019 Revised Date:

18 September 2019

Accepted Date: 19 September 2019

Please cite this article as: Liu, Y.-E., Luo, X.-J., Corella, P.Z., Zeng, Y.-H., Mai, B.-X., Organophosphorus flame retardants in a typical freshwater food web: Bioaccumulation factors, tissue distribution, and trophic transfer, Environmental Pollution (2019), doi: https://doi.org/10.1016/ j.envpol.2019.113286. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Ltd.

Graphical Abstract

1 2

Organophosphorus flame retardants in a typical freshwater food web: bioaccumulation factors, tissue distribution, and trophic transfer

3 4

Yin-E Liu a,b, Xiao-Jun Luo a,*, Pablo Zapata Corella a, Yan-Hong Zeng a,

5

Bi-Xian Mai a

6 7

a

State Key Laboratory of Organic Geochemistry and Guangdong Key Laboratory of

8

Environmental Resources Utilization and Protection, Guangzhou Institute of

9

Geochemistry, Chinese Academy of Sciences, Guangzhou 510640, China

10

b

University of Chinese Academy of Sciences, Beijing 100049, China

11

*

Address correspondence to: [email protected]

1

12

Abstract

13

Water, sediment, and wild aquatic species were collected from an electronic

14

waste (e-waste) polluted pond in South China. This study aimed to investigate the

15

bioaccumulation, tissue distribution, and trophic transfer of organophosphorus flame

16

retardants (PFRs) in these aquatic organisms. The concentrations of PFRs detected in

17

the analyzed organisms were between 1.7 and 47 ng/g wet weight (ww). Oriental river

18

prawn and snakehead exhibited the highest and lowest levels, respectively. Tri-n-butyl

19

phosphate (TnBP), tris(2-chloroethyl) phosphate (TCEP), tris(2-chloroisopropyl)

20

phosphate (TCPP) and triphenyl phosphate (TPhP) were dominant contaminants,

21

accounting for approximately 86% of the total sum. The mean values of

22

bioaccumulation factors (BCFs) and logarithmic biota-sediment accumulation factors

23

(log BSAFs) for individual PFRs varied from 6.6 to 1109 and from -2.0 to 0.41,

24

respectively. Both log BCFs and log BSAFs of PFRs were significantly and positively

25

correlated with their octanol-water partitioning coefficient (log KOW). The

26

concentrations of PFRs in tissues of large mud carp and snakehead were significantly

27

and positively correlated with the lipid content (each p < 0.05) and the liver, kidney,

28

and gill exhibited high PFR levels. When the concentration was expressed on a lipid

29

basis, liver exhibited the lowest level, indicating the probable effects of metabolism.

30

Significantly positive correlation was also found between lipid content and total PFR

31

concentration in muscle of all aquatic organisms, given the strong correlation between

32

lipid content and the concentration of TnBP. Trophic magnification factors (TMF) of

33

TnBP and TPhP were lower than 1 (0.57 and 0.62), indicating that these PFRs

34

undergo trophic dilution in this aquatic food web.

35 36

Keywords: Organophosphorus flame retardant; Aquatic organisms; Bioaccumulation;

37

Tissue distribution; Trophic dilution

38 39 40

Capsule: The present study revealed the bioaccumulation and biomagnification po-

41

tentials of PFRs in aquatic organisms and provided basic data for the internal expo-

42

sure of PFRs in organisms.

43 44

2

45

1. Introduction

46

In recent years, organophosphorus flame retardants (PFRs) have become more

47

and more widely used as suitable alternative flame retardants in a variety of

48

commercial products, with the ban on the use of penta- and octa- polybrominated

49

diphenyl ethers (PBDEs) (Van der Veen and de Boer, 2012). Total global consumption

50

of PFRs increased from 500,000 t in 2011 to 680,000 t in 2015 (Van der Veen and de

51

Boer, 2012; Wei et al., 2015). In 2007, the demand for PFRs was approximately

52

70,000 t in China, and was growing at an average annual rate of 15% (Wei et al.,

53

2015). Since they were used as basic end-products by direct mixing into materials

54

rather than chemical bonding, PFRs can easily escape from the material and enter the

55

environment (Sundkvist et al., 2010). As a result, PFRs have been ubiquitously

56

detected in different environmental matrixes including water, air, dust, sediment, soil

57

and biota (Stapleton et al., 2009; Fries and Mihajlović, 2011; Van den Eede et al.,

58

2011; Tan et al., 2016; Zhang et al., 2018; Liu et al., 2019).

59

Considering their widespread use, the prevalence of PFRs in the environment

60

and their potential toxicity to organisms (i.e. neurotoxicity, carcinogenicity and

61

reproductive toxicity), nowadays, PFRs as emerging contaminants are attracting more

62

and more attention from environmental researchers, especially regarding their internal

63

exposure to living organisms (WHO, 1990, 2000; Van der Veen and De Boer, 2012;

64

Hou et al., 2016). However, data about the occurrence of PFRs in biota is limited,

65

especially on the bioaccumulation, tissue distribution and trophic transfer of PFRs,

66

which are key criteria in the assessment of internal exposure and potential risk of

67

PFRs to organisms. Hou et al., (2017) measured 8 PFRs in whole-body samples and

68

various tissues of three freshwater fish species from Beijing, China, to investigate the

69

bioaccumulation and tissue distribution of PFRs. Trophic magnification factors (TMF)

70

have been investigated in three aquatic food webs from Western Scheldt Estuary,

71

Netherlands (Brandsma et al., 2015) and Taihu Lake, China (Zhao et al., 2018; Wang

72

et al., 2019). In addition, several studies have investigated the bioaccumulation of

73

PFRs using laboratory model organisms (Sasaki et al., 1981, 1982; Wang et al., 2017;

74

Tang et al., 2019). The bioaccumulation potential of PFRs seems to be different from

75

PBDEs (Sundkvist et al., 2010; Kim et al., 2011; Brandsma et al., 2015; Malarvannan

76

et al., 2015), and trophic transfer of PFRs in different food webs was inconsistent

77

(Kim et al., 2011; Brandsma et al., 2015; Zhao et al., 2018; Wang et al., 2019).

3

78

Therefore, it is necessary to conduct additional studies on the bioaccumulation and

79

trophodynamics of PFRs for a better evaluation.

80

Qingyuan is one of the largest electronic waste (e-waste) recycling areas in South

81

China, and it has been proven that various e-waste-associated organic pollutants (i.e.

82

polychlorinated biphenyls (PCBs), PBDEs or short chain chlorinated paraffins

83

(SCCPs)) are present here at high concentrations (Chen et al., 2011; Luo et al., 2015;

84

Huang et al., 2018). In the present study, water, sediment and wild aquatic species

85

were collected from an enclosed e-waste polluted pond in Qingyuan, South China.

86

The concentrations of 10 PFRs were determined in all analyzed samples, to investi-

87

gate the species-specific bioaccumulation, tissue distribution profile, and trophic

88

transfer of PFRs in aquatic species.

89 90

2. Materials and methods

91

2.1. Sample collection

92

Fish and invertebrates were collected in December 2014 from an enclosed

93

freshwater body pond, located in Longtang Town, Qingyuan County, Guangdong

94

province, South China. A total of 138 individual aquatic organisms, including 50 ori-

95

ental river prawn individuals (Macrobrachium nipponense, grouped in 5 pooled sam-

96

ples), 16 crucian carp individuals (Carassius auratus, grouped in 5 pooled samples),

97

65 mud carp individuals (Cirrhinus molitorella, divided into two groups based on dif-

98

ferent size criteria: 5 individuals were the large size group (body length: 49 ± 3.0 cm,

99

weight: 1800 ± 230 g) and 5 pooled samples were the small size group (body length:

100

8.2 ± 1.1 cm, weight: 5.3 ± 1.5 g)), 2 catfish individuals (Clarias batrachus), and 5

101

snakehead individuals (Ophiocephalus argus) were collected.

102

The studied pond has an area of 5,000 square meters and a depth of 2 meters, and

103

the bottom of this pond is filled with abandoned e-wastes. Further information about

104

the sampling point, can be found in our previous study (Wu et al., 2008). Three water

105

samples and two surface sediment samples were simultaneously collected from the

106

pond. In the present study, large mud carps were used for investigating tissue distribu-

107

tion of PFRs and the small size group was used for investigating trophic transfer of

108

PFRs in the aquatic food web. The muscle of all organisms, and skin, gill, liver, kid-

109

ney and bladder of large mud carp and snakehead were freeze-dried, homogenized

110

and stored separately at -20 °C until analysis. Detailed information of the analyzed

111

samples is provided in the Supporting Information (SI, Table S1). 4

112 113

2.2. Sample treatment and instrumental analysis

114

10 PFR chemicals (including tri-iso-propyl phosphate (TiPP), tri-n-propyl phos-

115

phate (TnPP), TnBP, TCEP, TCPP, tris(2-chloro-1-(chloromethyl)ethyl) phosphate

116

(TDCPP), TPhP, 2-ethylhexyl diphenyl phosphate (EHDPP), tris(2-ethylhexyl) phos-

117

phate (TEHP) and tricresyl phosphate (TCrP)) were selected as targets in the present

118

study. The sample treatment for PFRs followed our previous studies (Tan et al., 2016;

119

Liu et al., 2018). More details are provided in the SI.

120

The instrumental analysis of PFRs was conducted on a gas chromatograph cou-

121

pled to a triple quadrupole mass spectrometer (GC-MS/MS) equipped with an electron

122

ionization (EI) source and a DB-5 capillary column (30 m×0.25 mm×0.25 µm) ac-

123

cording to previous studies (Poma et al., 2018; Liu et al., 2019) with minor modifica-

124

tions. The detailed information of MS/MS quantitation parameters for each chemical

125

is provided in the SI, Table S2.

126

Stable-carbon (δ13C) and stable-nitrogen (δ15N) isotopes for biota samples (ca.

127

0.5 mg freeze-dried) were measured with a Flash EA 112 series elemental analyzer.

128

Total organic carbon (TOC) for sediment was measured with a Vario EL III elemental

129

analyzer.

130 131

2.3. Quality assurance and quality control

132

To minimize contamination of samples with PFRs, all the glassware was baked

133

(at 450 °C) for 5 h and then rinsed with acetone, dichloromethane and n-hexane, the

134

evaporation and SPE equipments were placed in a pre-cleaned fume hood (Liu et al.,

135

2018). During the sample analysis, one procedural blank was run in every batch of

136

samples (n = 12). Only small amounts of TCEP, TCPP, and TPhP were detected in 6

137

procedural blanks and the final concentrations were blank-corrected. Details on levels

138

of blank contamination are in the SI, Table S3. Multi-level calibration curves (2-2000

139

ng/mL) were run with satisfied linearity (R2 > 0.99) for each chemical. A standard

140

solution (100 ng/mL for each PFR) was injected 3 times every day to monitor the sen-

141

sitivity of the instrument. The recoveries of surrogate standards, expressed as mean

142

values ± standard deviation (mean ± SD) in all analyzed samples were: 72 ± 7.8% for

143

TnPP-D21, 76 ± 16% for TnBP-D27, 70 ± 19% for TCPP-D18, and 90 ± 5.6% for

144

TPhP-D15. Method quantitation limits (MQLs) were calculated as the mean value

145

plus three times the standard deviations detected in the procedural blanks. For chemi5

146

cals that were not detectable in the blanks, the MQLs were set to be concentrations

147

that would produce signal-to-noise ratios of 10. The MQLs of PFRs for organisms,

148

water, and sediment ranged from 0.013 to 2.0 ng/g ww, from 0.23 to 20 ng/L, and

149

from 0.056 to 9.1 ng/g dry weight (dw), respectively. Details on linearity and MQLs

150

are in the SI, Table S4.

151 152

2.4. Data analysis

153

2.4.1. Characterization of food web using stable isotope analysis Stable carbon and nitrogen isotope abundances were expressed as δ13C (‰) and

154 155

δ15N (‰).

156

δ13C (‰) = ((13C/12Csample)/(13C/12Cstandard) 1) × 1000

(1)

157

δ15N (‰) = ((15N/14N sample) /(15N/14Nstandard) 1) × 1000

(2)

The

158

13

12

C/ C standard and

15

14

N/ N standard values were based on the reference

15

159

nitrogen for δ N and Vienna Pee Dee Belemnite for δ13C. The precision of the tech-

160

nique was ± 0.08% (SD) for δ13C and ± 0.07% (SD) for δ15N.

161

TMFs for PFRs in the food web were calculated using the following equations

162

(Eq. (3) and Eq. (4)), and using the slope of the curve formed by the representation of

163

δ15N values versus the logarithm of the PFR concentration (Ln C).

164

Ln C = a + b × δ15N

(3)

165

TMF = e

b

(4)

166

Where C represents the concentrations of PFRs on a wet weight basis.

167 168

2.4.2. Bioaccumulation factor calculation

169

The bioaccumulation factor (BCF) and biota-sediment accumulation factor

170

(BSAF) were used to assess the degree of bioaccumulation of the target compounds in

171

aquatic organisms. Both of them were calculated using Eq. (5) and (6) below:

172

BCF = Cbiota/Cwater

(5)

173

BSAF = Cbiota/Csediment

(6)

174

Where Cbiota, Cwater in Eq. (5) represent the PFR concentrations in biota on a wet

175

weight basis and the average dissolved PFR concentrations in water, respectively. Cbi-

176

ota,

177

and in sediment normalized by total organic carbon, respectively. BCF or BSAF val-

178

ues can only be calculated for chemicals that can be detected in both organisms and

179

water or in both organisms and sediments.

Csediment in Eq. (6) represent the PFR concentrations in biota on a lipid weight basis

6

180 181

2.4.3. Tissue distribution difference analysis

182

The ratios of PFR concentrations in other tissues to the liver (OLR) were used to

183

gain an insight into the distribution of PFRs between liver and other tissues in organ-

184

isms (Sun et al., 2017). These ratios were calculated using Eq. (7):

185

OLR = Cother /(Cother + Cliver)

186

Where Cother, Cliver represent the lipid-normalized concentrations of PFRs in other

187

tissues (i.e. muscle, bladder, skin, kidney, gill) and liver, respectively. Value of OLR

188

that differ significantly from 0.5 indicate a significant difference in concentrations

189

between liver and other tissues (Sun et al., 2017).

(7)

190 191

2.4.4. Statistical treatment of the data

192

Statistical analyses were performed with IMB SPSS Statics 19.0 and Origin 8.0

193

software. Pearson correlations were used to assess the relationships between PFR

194

concentration and lipid content, log KOW and log BCF, log KOW and log BSAF, and Ln

195

C and δ15N values of the organisms. One-way ANOVA and cluster analysis were used

196

for comparisons of differences in PFR concentrations and compositions among dif-

197

ferent aquatic species. Detection limit divided by two was used to replace the unde-

198

tected values when conducting correlation analyses, since the censoring percentages

199

are below 15% (USEPA, 1998) in the present study. All differences with p < 0.05

200

were considered significant.

201 202

3. Results

203

Between 10 PFRs analyzed in the present study, TnBP, TCPP, TPhP, and TEHP

204

were detected in all samples, TCEP, EHDPP, and TCrP were found in > 90% of the

205

samples, and TDCPP was found in 50% of the samples. TiPP and TnPP were not de-

206

tected in any sample. The total PFR concentration in water was 255 ± 20 ng/L. PFR

207

concentrations in the two sediment samples were 83 and 187 ng/g dw, with TOC val-

208

ues of 1.5% and 3.4%, respectively. The levels of total PFRs in aquatic organisms

209

were between 1.7 and 47 ng/g ww, and the highest and lowest average concentrations

210

were detected in oriental river prawn (34 ng/g ww) and snakehead (4.3 ng/g ww),

211

respectively (Table 1). Generally, TnBP, TCPP, TCEP, and TPhP were the dominant

212

PFRs, collectively accounting for up to 86% of the total amount, and the composition

213

profile of PFRs exhibited inter- and intra-species differences (Figure 1). 7

214

The mean BCFs of TnBP, TCEP, TCPP, TDCPP, TPhP, EHDPP, TEHP and TCrP

215

in the present study had the following ranges: 20-228, 3.6-11, 20-171, 6.6-245, 18-898,

216

48-289, 70-1109 and 58-906, respectively (Table S5). Log BSAFs of PFRs ranged

217

from -2.0 to 0.41. Significantly positive correlations were found between log BCF and

218

log KOW, and between log BSAF and log KOW (Figure 2).

219

The tissue distribution of PFRs in snakeheads differed from that in mud carps.

220

The order of total PFR mean levels among tissues in snakeheads was liver (26 ng/g

221

ww) ≈ kidney (26 ng/g ww) > gill (17 ng/g ww) > skin (6.6 ng/g ww) > bladder (5.8

222

ng/g ww) > muscle (4.3 ng/g ww). The order in mud carps was gill (31 ng/g ww) ≈

223

liver (29 ng/g ww) > kidney (18 ng/g ww) and skin (17 ng/g ww) > muscle (10 ng/g

224

ww) > bladder (5.2 ng/g ww) (Table S6). The tissue distribution of PFRs was, to some

225

extent, similar to the lipid distribution in those tissues from these two fish species. For

226

example, the lipid content in tissues of mud carp followed this order: liver (10.5 ±

227

1.05%) ≈ gill (10.3 ± 0.58%) > kidney (5.6 ± 1.33%) > skin (2.79 ± 0.14%) > muscle

228

(1.69 ± 0.24%) > bladder (0.39 ± 0.04%). Significantly positive correlations between

229

lipid contents and concentrations of PFRs in tissues of large mud carp and snakehead

230

were found (each p < 0.05) (Figure 3). All calculated OLR values were larger than 0.5,

231

and snakehead exhibited higher OLR values than mud carp (Figure 4).

232

Significantly negative correlations were found between the δ15N values and loga-

233

rithmic transformed wet weight concentrations (Ln Cww) for total PFRs, TnBP and

234

TPhP (Figure 5), with calculated TMF values of 0.72, 0.57 and 0.62, respectively.

235

However, logarithmic transformed lipid-normalized concentrations (Ln Clw) were not

236

significantly correlated with δ15N values, Ln Clw of TnBP and TPhP showed a weak

237

negative correlation though (Figure 5).

238 239

4. Discussion

240

4.1 Levels and distribution patterns of PFRs

241

The levels of PFRs in aquatic organisms in this study were close to those in three

242

freshwater fish species (including mud carp, tilapia and plecostomus) from the Pearl

243

River Delta, South China (2.3-30 ng/g ww) (Liu et al., 2019), and higher than those in

244

fish collected in lakes of Canada (mean concentration of 1.6 ng/g ww) (McGoldrick et

245

al., 2014) and in crucian carp from the Nakdong River, South Korea (4.2-7.8 ng/g ww)

246

(Choo et al., 2018). However, the level of total PFRs ranged from 9.9 to 81 ng/g ww

247

(mean concentration of 38 ng/g ww) in freshwater fish from Western Scheldt Estuary, 8

248

Netherlands (Brandsma et al., 2015), which was higher than those in aquatic organ-

249

isms from the e-waste polluted pond in the present study. These reported results re-

250

flected to some extent the differences in PFR contamination levels among aquatic

251

organisms in different countries and regions. Of course, it could be affected by the

252

differences in fish species and size.

253

The pattern of PFRs in the present study was similar to some results previously

254

observed in fish (Kim et al., 2011; Hou et al., 2017). As reported by Zheng et al.,

255

(2015), TPhP and TCPP were also the most abundant PFR chemicals in indoor dust in

256

this e-waste area, accounting for more than 70% of the total level. In addition, Poma

257

et al., (2019) and Zheng et al., (2016) also found that TPhP and TCPP were two dom-

258

inant PFR chemicals in insects and home-produced eggs collected in the same area.

259

As we all know, TCPP is widely used in insulating materials, sealing foams and elec-

260

tronic equipment, and TPhP is an important plasticizer used in TVs, notebook com-

261

puters, and the manufacture of power sockets (Wei et al., 2015). There are a large

262

number of dismantled and recycled wires, household appliances and electronic mate-

263

rials at the studied site. These are likely the main sources of these PFR contaminants

264

(i.e. TCPP, TPhP) (Poma et al., 2019).

265

The composition profile of PFRs varied among the studied aquatic species (Fig-

266

ure 1). TCPP was the most abundant chemical in snakehead, accounting for 44% of

267

the total PFR level. TnBP was found to be the most abundant chemical in mud carp,

268

catfish, and crucian carp, while TCPP and TPhP (each one accounting for 35% of the

269

total sum) were the most abundant pollutants in oriental river prawn. Even in a given

270

species, the composition pattern of PFRs was different. For example, the small sized

271

mud carp had a higher proportion of TnBP (69%) than the large sized ones (47%),

272

while the large mud carp had higher proportions of halogenated PFRs (TCEP and

273

TCPP) and TPhP (29% and 19%, respectively) than the small ones (22% and 4.8%,

274

respectively). Similarly, the small catfish exhibited a higher proportion of TnBP (63%)

275

and lower proportion of halogenated PFRs (30%) than large catfish (with proportions

276

of 46% and 47%, respectively). These inter- and intra-species differences may be due

277

to the differences in the feeding habits and metabolic potential of PFRs among differ-

278

ent organisms.

279

However, when the composition profile of PFRs in each fish species was com-

280

pared using cluster analysis (Figure S1), the mud carp and crucian carp can be catego-

281

rized into one group with similar composition profiles, which could be attributed to 9

282

the similar feeding behavior of these two omnivorous fish species. Notably, the Eu-

283

clidean distance between carnivorous catfish and omnivorous fish (crucian and mud

284

carps) was shorter than the one between catfish and the also carnivorous snakehead.

285

Meanwhile, significant species-specific differences in the composition profiles of

286

PFRs in these two carnivorous fish species (catfish and snakehead) can be found in

287

the present study. The low number of catfish samples (n = 2) and individual differ-

288

ences between them (body weight of 1340 g for large catfish and 177 g for small)

289

could be the main reason for this abnormal result.

290 291

4.2. Bioaccumulation factors of PFRs in fish

292

The BCF values in the present study were close to those in killifish and crucian

293

carp, which were 1.1-6.15, 2.44, 47-113, 35-71.9 and 500 for TCEP, TCPP, TDCPP,

294

TnBP and TPhP, respectively (Sasaki et al., 1981; WHO, 1998; Choo et al., 2018).

295

Hou et al., (2017) reported that the average BCFs of TnBP, TCEP, TCPP, TDCPP,

296

TPhP, EHDPP and TEHP in freshwater fish from Beijing, China were 173, 34.7, 250,

297

27.8, 1008, 163 and 1983, respectively, which were comparable for these chemicals in

298

the present study. The log BSAF values in this study were comparable to those in fish

299

from previous studies (Giulivo et al., 2017; Hou et al., 2017), but much lower than

300

those from Choo et al., (2018) and Wang et al., (2019).

301

Among the different PFR chemicals analyzed in the present study, TPhP, TEHP,

302

and TCrP exhibited higher bioaccumulation potential, which exhibited relatively

303

higher BCF or BSAF values, but all calculated BCFs were below the REACH criteri-

304

on (> 2000) as a bio-accumulative chemical (European Union, 2008), or BSAF values

305

were lower than 1 (except TEHP in the snakehead and prawn, where mean values of

306

1.4 and 2.6 were found, respectively). Correlation analyses were conducted between

307

log BCF and log KOW, and between log BSAF and log KOW, to investigate the effect

308

of hydrophobicity properties of PFRs on their bioaccumulation potential. Significantly

309

positive correlations were found between log BCF and log KOW (r = 0.62, p < 0.01)

310

and between log BSAF and log KOW (r = 0.53, p < 0.01) (Figure 2), suggesting that

311

the accumulation of PFRs could be estimated by their hydrophobicity, but could also

312

be influenced by metabolism. Hou et al., (2017) also found that there was a weak but

313

significantly positive correlation between log BSAF and log KOW for 8 PFRs (log

314

KOW range of 1.44-9.49) in three freshwater fish species from Beijing, China. Wang et

315

al., (2019) reported that the log BSAF of PFRs in benthic invertebrates first increased 10

316

with log KOW in the range of 1.44-5.73 and then decreased.

317 318

4.3. Tissue distribution of PFRs in fish

319

Linear correlation analyses revealed that there were significant correlations be-

320

tween total PFR concentration and lipid content in the tissues of both snakeheads and

321

mud carps (Figure 3a), suggesting that chemical affinity of PFRs to lipids still plays a

322

significant role in the deposition of PFRs in tissues. With respect to individual chemi-

323

cals (except TDCPP with detection frequency < 50%), the chlorinated hydrocarbon

324

chains compounds (TCEP and TCPP), which have relatively low log KOW values

325

(1.63 and 2.89), had weak or no correlation with lipid content. However, TPhP,

326

EHDPP, TEHP, and TCrP, which have relatively high log KOW (> 4), showed signifi-

327

cant and strong correlations (Table 2). These results were consistent with previous

328

studies, which showed that chemicals with lower log KOW (< 3) have higher elimina-

329

tion speeds and shorter half-lives (t1/2) in organisms (Sasaki et al., 1981; Green et al.,

330

2007). In a laboratory exposure experiment using common carp as a model, Tang et

331

al., (2019) found that the percentage contribution of each PFR to total PFRs in all tis-

332

sues, except serum, was significantly and positively correlated with log KOW, although

333

PFRs are less hydrophobic than halogenated flame retardant such as PBDEs.

334

When the correlation analysis was conducted in individual tissues for aquatic

335

organisms, a significant correlation was also found between total PFR concentration

336

and lipid content (Figure 3b). This significant correlation was mainly derived by the

337

strong correlation between lipid content and concentration of TnBP. In the present

338

study, TnBP was the most abundant PFR chemical, and has relatively strong lipophilic

339

behaviour (log KOW = 4). Several studies have reported that total PFRs were not basi-

340

cally associated with lipids (Sundkvist et al., 2010; Kim et al., 2011; Chen et al., 2012;

341

Brandsma et al., 2015; Malarvannan et al., 2015; Hou et al., 2017). This could be due

342

to the difference in dominant compounds among different studies. For example,

343

Brandsma et al., (2015) found that tris(2-butoxyethyl) phosphate (TBOEP), TCPP and

344

TCEP were dominant PFRs in 34 samples from the Western Scheldt. Sundkvist et al.,

345

(2010) found that TCPP exhibit the highest levels among 11 PFR chemicals in differ-

346

ent marine and freshwater species, and Malarvannan et al., (2015) also reported that

347

TCPP was the most abundant component in the European eel from Flanders, Belgium,

348

accounting for 64% of the total level. Significant correlation between lipid content

349

and concentration was only found for TnBP (Figure 3b) and not for other compounds 11

350

in the present study. Furthermore, all aquatic samples were taken from a closed small

351

pond and shared a small motion range and the same pollution source in this study.

352

Whereas in the foregoing study, fish were collected from an open environment, such

353

as various rivers as well as a fish market, so these fish may have covered a large geo-

354

graphic area and can be exposed to different pollution sources. This could be one of

355

the factors to explain the differences found between the present study and previous

356

ones. However, other factors, such as metabolism and exposure pathway rather than

357

lipid content, could also play an important role in the deposition of PFRs in fish tis-

358

sues.

359

In the present study, relatively high concentrations of PFRs were found in liver

360

tissues of snakehead and mud carp. The livers also exhibited higher PFR concentra-

361

tions than the ones reported in crucian carp from the Nakdong River, South Korea

362

(Choo et al., 2018), in crucian carp and loach from Beijing, China (Hou et al., 2017)

363

and in Atlantic cod from Svalbard, Norway (Evenset et al., 2009). However, when the

364

concentration was expressed on the basis of lipid weight, liver exhibited lower levels.

365

As previously mentioned, metabolization could be the main reason for this observa-

366

tion because liver is the most important detoxification organ where PFRs can be read-

367

ily metabolized (Hou et al., 2017). All calculated OLR values were larger than 0.5 in

368

this study, indicating the effect of metabolism on the deposition of PFRs in the liver

369

(Figure 4). In both snakehead and mud carp, the OLRs for kidney were significantly

370

lower than those in other tissues (p < 0.05), suggesting that the kidney may also be

371

involved in the metabolism of PFRs. Snakehead exhibited higher OLR values than the

372

mud carp, which could be due to the higher metabolism potential of the former, con-

373

sidering that snakehead occupied a higher trophic level. Hu et al., (2016) and Ruus et

374

al., (2002) also have suggested that further metabolic transformation of phthalate es-

375

ters and organochlorines in aquatic organisms occupied higher trophic levels.

376

Significantly negative correlations between log KOW and OLR values for the

377

muscle tissue in both mud carp and snakehead and gill and kidney just in the snake-

378

head (each p < 0.05) were observed (Figure S2). Negative correlations between the

379

log KOW and OLRs for four organs (kidney, gill, muscle, and skin) in mud carp, and

380

between log KOW and OLRs for two organs (kidney and muscle) in snakehead were

381

found in our previous study which was focused on the tissue distribution of SCCPs in

382

the same investigated organisms (Sun et al., 2017). Additionally, the same trends were

383

also found in terrestrial organisms. For example, Zheng et al., (2014) and Li et al., 12

384

(2016) found significantly negative correlations between the ratios of muscle to liver

385

in neonate chicks and log KOW for halogenated organic chemicals, including PBDEs

386

and polybrominated biphenyls. These results indicated that liver preferentially accu-

387

mulates high lipophilic chemicals compared to other tissues.

388 389

4.4. PFRs in the food webs

390

Stable isotope analysis is integrated into diet measures for analyzing the structure

391

of food web and effectively helping to elucidate the trophic transfer of chemicals

392

which could be accumulated in organisms. δ15N is commonly used to determine the

393

trophic level of organisms, which usually increases together with δ15N values. As de-

394

scribed in our previous study (Sun et al., 2017), δ15N values of the studied aquatic

395

species usually increased in the following order: shrimp, omnivorous fish and carniv-

396

orous fish (Figure S3). Notably, small mud carp exhibited even higher δ15N values

397

than the carnivorous catfish, which could be influenced by the potentially different

398

food sources for these two fish species and the unrepresentative number of catfish

399

samples. Juvenile mud carp mainly feed on zooplankton, such as rotifers, copepods

400

and small cladocerans, while the adult fish mainly feed on phytoplankton. Size de-

401

pendence for the relative trophic position in the mud carp group was observed. The

402

small sized group of mud carp exhibited higher δ15N values than the large sized group;

403

both of them exhibited similar δ13C values though. Carnivorous fish (snakehead and

404

catfish) showed similar δ13C values than prawn, suggesting that prawn may be the

405

main food resource for both of them (Sun et al., 2017). Additionally, during the dis-

406

section process, we found that small mud carps often appeared in the stomach of the

407

snakeheads (Figure S4), suggesting that the small carp were also prey for snakeheads.

408

The large mud carps were excluded from the food web investigation in this study,

409

because they could not be prey of predators due to their large body size.

410

Due to the lack of baseline organisms (primary producers or primary consumers),

411

δ15N values of each organism were used to express relative trophic position in the

412

food web. Considering that the concentrations of total PFRs and TnBP were positively

413

correlated with the lipid content in muscle tissue, correlation analyses were conducted

414

between δ15N values and concentrations based on wet weight and between δ15N val-

415

ues and concentrations based on lipid weight. The Ln Cww for total PFRs, TnBP and

416

TPhP were significantly and negatively correlated with δ15N values. These TMFs

417

were tentatively calculated to be 0.72, 0.57 and 0.62, respectively, indicating that 13

418

these PFRs undergo trophic dilution rather than trophic magnification in this aquatic

419

food web. These negative correlations were mainly caused by the lower lipid content

420

and PFR levels of the predators.

421

As far as we know, there are only a few reports on the trophic transfer of PFRs in

422

aquatic food webs, with inconsistent results. Brandsma et al., (2015) reported the

423

trophic magnification for TBOEP, TCPP and TCEP in the benthic food web, with

424

TMFs of 3.5, 2.2 and 2.6, respectively (p < 0.05), while trophic dilution (p > 0.05)

425

was found for other PFR chemicals (i.e. TnBP, TPhP, EHDPP) in both pelagic and

426

total (benthic + pelagic) food webs. The structure of food webs could be the main

427

cause for this different finding. However, Zhao et al., (2018) found that TCPP, TDCPP

428

and tris(methylphenyl) phosphate (TMPP) underwent trophic dilution in the food web

429

from Taihu Lake, and Wang et al., (2019) only found that EHDPP was biomagnified

430

in the food web also from Taihu Lake, with a TMF of 3.6. Additionally, Kim et al.,

431

(2011) suggested that there was significant biomagnification for TPhP in the fish from

432

Manila Bay, although they did not calculate TMF values. Hence, the structure of food

433

webs, the size, feeding habits and habitat of organisms, and the potential of biotrans-

434

formation or metabolism for PFRs in different organisms, and even some environ-

435

mental parameters (such as the dissolved organic matter, suspended particles and the

436

temperature of water) (Sun et al., 2017; Wang et al., 2019) may all contribute to the

437

trophodynamics differences of the studied pollutants.

438 439

5. Conclusion

440

The present results have demonstrated that PFRs accumulate extensively in

441

aquatic organisms in an e-waste polluted pond, and the bioaccumulation of PFRs ex-

442

hibited species-specific profiles among the investigated aquatic species. Both log

443

BCFs and log BSAFs of PFRs displayed significantly positive correlations with log

444

KOW. The accumulation of PFRs could still be estimated by hydrophobicity, but could

445

also be influenced by metabolism and elimination. Significant and positive correla-

446

tions between PFR levels and lipid content of tissues were found in snakehead and

447

large mud carp, indicating that the affinity for lipids still plays a significant role in the

448

deposition of PFR in tissues. Total PFRs, TnBP and TPhP underwent trophic dilution

449

in the studied aquatic food web, with TMF values of 0.72, 0.57 and 0.62, respectively.

450

Due to the lack of investigation on PFR metabolites, it remains an insufficient under-

451

standing of the entire phenomenon of PFRs in the analyzed aquatic organisms, con14

452

sidering that some PFRs have high metabolization potential.

453 454

Acknowledgments

455

This study was financially supported by the National Science Foundation of

456

China (Nos. 41673100, 41877386, 41931290), the National Basic Research Program

457

of China (2015CB453102), Chinese Academy of Science (QYZDJ-SSW-DQC018),

458

Local Innovative and Research Teams Project of Guangdong Pearl River Talents Pro-

459

gram (2017BT01Z134) and Guangdong Foundation for Program of Science and

460

Technology Research (2017B030314057).

15

461

Reference

462

Brandsma, S.H., Leonards, P.E., Leslie, H.A., de Boer, J., 2015. Tracing

463

organophosphorus and brominated flame retardants and plasticizers in an

464

estuarine food web. Sci. Total Environ. 505, 22-31.

465

Chen, D., Letcher, R.J., Chu, S., 2012. Determination of non-halogenated, chlorinated

466

and brominated organophosphate flame retardants in herring gull eggs based on

467

liquid chromatography-tandem quadrupole mass spectrometry. J. Chromatogr. A

468

1220, 169-174.

469

Chen, M.Y., Luo, X.J., Zhang, X.L., He, M.J., Chen, S.J., Mai, B.X., 2011.

470

Chlorinated paraffins in sediments from the Pearl River Delta, South China:

471

spatial and temporal distributions and implication for processes. Environ. Sci.

472

Technol. 45, 9936-43.

473

Choo, G., Cho, H.S., Park, K., Lee, J.W., Kim, P., Oh, J.E., 2018. Tissue-specific

474

distribution and bioaccumulation potential of organophosphate flame retardants

475

in crucian carp. Environ. Pollut. 239, 161-168.

476

European

Union,

2008.

European

Union

Risk

Assessment

Report:

Tris

477

(1-chloro-2-propyl) Phosphate (TCPP), EINECS No: 237-158-7. Office for

478

Official Publications of the European Communities, Luxembourg, pp.

479

2019-2051.

480

Evenset, A., Leknes, H., Christensen, G.N., Warner, N., Remberger, M., Gabrielsen,

481

G.W., 2009. Screening of New Contaminants in Samples from the Norwegian

482

Arctic Rapport, Akvaplan-niva Report. Norwegian Pollution Control Agency,

483

Norway.

484 485

Fries, E., Mihajlović, I., 2011. Pollution of soils with organophosphorus flame retardants and plasticizers. J. Environ. Monit. 13, 2692-2694.

486

Giulivo, M., Capri, E., Kalogianni, E., Milacic, R., Majone, B., Ferrari, F., Eljarrat, E.,

487

Barcelo, D., 2017. Occurrence of halogenated and organophosphate flame

488

retardants in sediment and fish samples from three European river basins. Sci.

489

Total Environ. 586, 782-791.

490

Green, N., Schlabach, M., Bakke, T., Brevik, E.M., Dye, C., Herzke, D., Huber, S.,

491

2007. Screening of Selected Metals and New Organic Contaminants 2007.

492

Akvaplanniva report. Norwegian Pollution Control Agency, Norway.

16

493

Hou, R., Liu, C., Gao, X., Xu, Y., Zha, J., Wang, Z., 2017. Accumulation and

494

distribution of organophosphate flame retardants (PFRs) and their di-alkyl

495

phosphates (DAPs) metabolites in different freshwater fish from locations around

496

Beijing, China. Environ. Pollut. 229, 548-556.

497

Hou, R., Xu, Y., Wang, Z.J., 2016. Review of OPFRs in animals and humans:

498

Absorption, bioaccumulation, metabolism, and internal exposure research.

499

Chemosphere 153, 78-90.

500

Hu, X., Gu, Y., Huang, W., Yin, D., 2016. Phthalate monoesters as markers of

501

phthalate contamination in wild marine organisms. Environ. Pollut. 218,

502

410-418.

503

Huang, C.C., Zeng, Y.H., Luo, X.J., Tang, B., Liu, Y.E., Ren, Z.H., Mai, B.X., 2018.

504

Level changes and human dietary exposure assessment of halogenated flame

505

retardant levels in free-range chicken eggs: A case study of a former e-waste

506

recycling site, South China. Sci. Total Environ. 634, 509-515.

507

Kim, J.W., Isobe, T., Chang, K.H., Amano, A., Maneja, R.H., Zamora, P.B., Siringan,

508

F.P., Tanabe, S., 2011. Levels and distribution of organophosphorus flame

509

retardants and plasticizers in fishes from Manila Bay, the Philippines. Environ.

510

Pollut. 159, 3653-3659.

511

Li Z.R., Luo X.J., Huang L.Q., Mai, B.X., 2016. In ovo uptake, metabolism, and

512

tissue-specific distribution of chiral PCBs and PBDEs in developing chicken

513

embryos. Sci. Rep. 6, 36597.

514

Liu, Y.E, Huang, L.Q., Luo, X.J., Tan, X.X., Huang, C.C., Corella, P.Z., Mai, B.X.,

515

2018. Determination of organophosphorus flame retardants in fish by

516

freezing-lipid precipitation, solid-phase extraction and gas chromatography-mass

517

spectrometry. J. Chromatogr. A 1532, 68-73.

518

Liu, Y.E, Luo, X.J., Huang, L.Q., Zeng, Y.H., Mai, B.X., 2019. Organophosphorus

519

flame retardants in fish from Rivers in the Pearl River Delta, South China. Sci.

520

Total Environ. 663, 125-132.

521

Lynn, R.K., Wong, K., Garvie-Gould, C., Kennish, J.M., 1981. Disposition of the

522

flame retardant, tris(1,3-dichloro-2-propyl) phosphate, in the rat. Drug Metabol.

523

Dispos. 9, 434-441

524

Luo, X.J., Sun, Y.X., Wu, J.P., Chen, S.J., Mai, B.X., 2015. Short-chain chlorinated

525

paraffins in terrestrial bird species inhabiting an e-waste recycling site in South

526

China. Environ. Pollut. 198, 41-6. 17

527

Malarvannan, G., Belpaire, C., Geeraerts, C., Eulaers, I., Neels, H., Covaci, A., 2015.

528

Organophosphorus flame retardants in the European eel in Flanders, Belgium:

529

Occurrence, fate and human health risk. Environ. Res. 140, 604-10.

530

McGoldrick, D.J., Letcher, R.J., Barresi, E., Keir, M.J., Small, J., Clark, M.G., Sverko,

531

E., Backus, S.M., 2014. Organophosphate flame retardants and organosiloxanes

532

in predatory freshwater fish from locations across Canada. Environ. Pollut. 193,

533

254-261.

534

Minegishi, K., Kurebayashi, H., Nambaru, S., Morimoto, K., Takahashi, T., Yamaha,

535

T., 1988. Comparative studies on absorption, distribution, and excretion of flame

536

retardants halogenated alkyl phosphate in rats. Eisei. Kagaku. 34, 102-114.

537

Poma, G., Liu, Y., Cuykx, M., Tang, B., Luo, X. J., Covaci, A., 2019. Occurrence of

538

organophosphorus flame retardants and plasticizers in wild insects from a former

539

e-waste recycling site in the Guangdong province, South China. Sci. Total

540

Environ. 650, 709-712.

541

Ruus, A., Ugland, K.I., Skaare, J.U., 2002. Influence of trophic position on

542

organochlorine concentrations and compositional patterns in a marine food web.

543

Environ. Toxicol. Chem. 21, 2356-2364.

544

Sasaki, K., Takeda, M., Uchiyama, M., 1981. Toxicity, absorption and elimination of

545

phosphoric acid triesters by killifish and goldfish. Bull. Environ. Contam.

546

Toxicol. 27, 775-782.

547

Sasaki, K., Suzuki, T., Takeda, M., Uchiyama, M., 1982. Bioconcentration and

548

excretion of phosphoric acid triesters by killifish (Oryzeas latipes). Bull. Environ.

549

Contam. Toxicol. 28, 752-759.

550

Stapleton, H.M., Klosterhaus, S., Eagle, S., Fuh, J., Meeker, J.D., Blum, A., Webster,

551

T.F., 2009. Detection of organophosphate flame retardants in furniture foam and

552

US house dust. Environ. Sci. Technol. 43, 7490-7495.

553

Sun, R.X., Luo, X.J., Tang, B., Chen, L., Liu, Y., Mai, B.X., 2017. Bioaccumulation of

554

short chain chlorinated paraffins in a typical freshwater food web contaminated

555

by e-waste in south china: Bioaccumulation factors, tissue distribution, and

556

trophic transfer. Environ. Pollut. 222, 165-174.

557

Sundkvist, A.M., Olofsson, U., Haglund, P., 2010. Organophosphorus flame retardants

558

and plasticizers in marine and fresh water biota and in human milk. J. Environ.

559

Monit. 12, 943-951.

18

560

Tan, X.X., Luo, X.J., Zheng, X.B., Li, Z.R., Sun, R.X., Mai, B.X., 2016. Distribution

561

of organophosphorus flame retardants in sediments from the Pearl River Delta in

562

South China. Sci. Total Environ. 544, 77-84.

563

Tang, B., Poma, G., Bastiaensen, M., Yin, S.S., Luo, X.J., Mai, B.X., Covaci, A., 2019.

564

Bioconcentration and biotransformation of organophosphorus flame retardants

565

(PFRs) in common carp (Cyprinus carpio). Environ. Int. 126, 512-522.

566

USEPA, Office of Research and Development, 1998. Guidance for Data Quality

567

Assessment: Practical Methods for Data Analysis. EPA/600/ R-96/084,

568

Washington, DC.

569

Van Ael, E., Covaci, A., Das, K., Lepoint, G., Blust, R., Bervoets, L., 2013. Factors

570

influencing the bioaccumulation of persistent organic pollutants in food webs of

571

the scheldt estuary. Environ. Sci. Technol. 47, 11221-31.

572

Van den Eede, N., Dirtu, A.C., Neels, H., Covaci, A., 2011. Analytical developments

573

and preliminary assessment of human exposure to organophosphate flame

574

retardants from indoor dust. Environ. Int. 37, 454-461.

575

Van der Veen, I., de Boer, J., 2012. Phosphorus flame retardants: properties,

576

production, environmental occurrence, toxicity and analysis. Chemosphere 88,

577

1119-1153.

578

Wang, G., Shi, H., Du, Z., Chen, H., Peng, J., Gao, S., 2017a. Bioaccumulation

579

mechanism of organophosphate esters in adult zebrafish (Danio rerio). Environ.

580

Pollut. 229, 177-187.

581

Wang, X., Zhong, W., Xiao, B., Liu, Q., Yang, L., Covaci, A., Zhu, L., 2019.

582

Bioavailability and biomagnification of organophosphate esters in the food web

583

of Taihu Lake, China: Impacts of chemical properties and metabolism. Environ.

584

Int. 125, 25-32.

585

Wei, G.L., Li, D.Q., Zhuo, M.N., Liao, Y.S., Xie, Z.Y., Guo, T.L., Li, J.J., Zhang, S.Y.,

586

Liang, Z.Q., 2015. Organophosphorus flame retardants and plasticizers: sources,

587

occurrence, toxicity and human exposure. Environ. Pollut. 196, 29-46.

588

WHO, 1990. Environmental Health Criteria 209, Flame Retardants: Tris(chloropropyl)

589

Phosphate and Tris(2-chloroethyl) Phosphate World Health Organization.

590

Swizerland, Geneva.

591

WHO, 1998. Flame Retardants: Tris(Chloropropyl) Phosphate and Tris(2-

592

chloropropyl) Phosphate. Environmental Health Criteria 209. World Health

593

Organization, Geneva, Switzerland. 19

594

WHO, 2000. Flame Retardants: Tris(2-butoxyethyl) Phosphate, Tris(2-ethylhexyl)

595

Phosphate and Tetrakis(hydroxymethyl) Phosphonium Salts. Environmental

596

Health Criteria 218. World Health Organization, Geneva, Switzerland.

597

Wu, J.P., Guan, Y.T., Zhang, Y., Luo, X.J., Zhi, H., Chen, S.J., Mai, B.X., 2010a.

598

Trophodynamics of hexabromocyclododecanes and several other non-PBDE

599

brominated flame retardants in a freshwater food web. Environ. Sci. Technol. 44,

600

5490-5495.

601

Wu, J.P., Luo, X.J., Zhang, Y., Luo, Y., Chen, S.J., Mai, B.X., Yang, Z.Y., 2008.

602

Bioaccumulation

603

polychlorinated biphenyls (PCBs) in wild aquatic species from an electronic

604

waste (e-waste) recycling site in South China. Environ. Int. 34, 1109-1113.

605

of

polybrominated

diphenyl

ethers

(PBDEs)

and

Wu, J.P., Luo, X.J., Zhang, Y., Yu, M., Chen, S.J., Mai, B.X., Yang, Z.Y., 2009.

606

Biomagnification

of

polybrominated

diphenyl

ethers

(PBDEs)

and

607

polychlorinated biphenyls in a highly contaminated freshwater food web from

608

South China. Environ. Pollut. 157, 904-9.

609

Wu, J.P., Zhang, Y., Luo, X.J., Wang, J., Chen, S.J., Guan, Y.T., Mai, B.X., 2010b.

610

Isomerspecific bioaccumulation and trophic transfer of Dechlorane Plus in the

611

freshwater food web from a highly contaminated site, South China. Environ. Sci.

612

Technol. 44, 606-611.

613

Zhang Y., Zheng X.B., Wei L.F., Sun R.X., Guo H.Y., Liu X.Y., 2018. The distribution

614

and accumulation of phosphate flame retardants (PFRs) in water environment.

615

Sci. Total Environ. 630: 164-170.

616

Zheng, X.B., Luo, X.J., Zeng, Y.H., Wu, J.P., Chen, S.J., Mai, B.X., 2014.

617

Halogenated flame retardants during egg formation and chicken embryo

618

development: maternal transfer,

619

distribution. Environ. Toxicol. Chem. 33, 1712-9.

possible

biotransformation,

and

tissue

620

Zheng, X.B., Xu, F.C., Chen, K.S., Zeng, Y.H., Luo, X.J., Chen, S.J., Mai, B.X.,

621

Covaci, A., 2015. Flame retardants and organochlorines in indoor dust from

622

several e-waste recycling sites in South China: composition variations and

623

implications for human exposure. Environ. Int. 78, 1-7.

624

Zheng, X., Xu, F., Luo, X., Mai, B., Covaci, A., 2016. Phosphate flame retardants and

625

novel brominated flame retardants in home-produced eggs from an e-waste

626

recycling region in China. Chemosphere 150, 545-550.

20

627

Figure captions

628

Figure 1. Composition profiles of PFRs in aquatic organisms from the pond polluted

629

by e-waste in Qingyuan, South China. The error bars represent standard deviations

630 631

Figure 2. Correlations between (a) log BCF and log KOW, (b) log BSAF and log KOW

632

of PFRs in the analyzed aquatic organisms

633 634

Figure 3. Correlations between (a) the total PFR concentration and lipid content in

635

various tissues of the two fish species, (b) the concentrations of PFRs and lipid con-

636

tent in muscle tissue of all aquatic organisms

637 638

Figure 4. Comparisons of the ratios of other tissue to liver in two fish species. The

639

error bars represent standard deviations

640 641

Figure 5. Correlations between the PFR concentration (Ln transformed) and the

642

trophic levels of the aquatic organisms

21

75 60 45

30

30

15

15

0 75

0 75

Large mud carp

45

45

30

30

15

15

0 75

0 75

Prawn

TPhP

TDCPP

TnBP

TCrP

TEHP

0

EHDPP

15

0

TPhP

30

15

TDCPP

30

TCPP

45

TCEP

Crucian carp

60

45

TCPP

60

Small mud carp

60

TCEP

60

Snakehead

TCrP

45

TnBP

Relative abundance to total PFRs (%)

60

TEHP

Catfish

EHDPP

75

643

Figure 1. Composition profiles of PFRs in aquatic organisms from the pond polluted

644

by e-waste in Qingyuan, South China. The error bars represent standard deviations.

22

3.5

(a)

(b)

0.0

2.8

r = 0.62, p < 0.01

r = 0.53, p < 0.01

2.1

Log BSAF

Log BCF

-0.5

1.4

-1.5

Crucian carp Catfish Large mud carp Small mud carp Snakehead Prawn

0.7

-1.0

-2.0

0.0

-2.5 0

2

4

6

8

10

0

Log Kow

2

4

6

8

10

Log Kow

645

Figure 2. Correlations between (a) log BCF and log KOW, (b) log BSAF and log KOW

646

of PFRs in the analyzed aquatic organisms.

23

60

50

Mud carp Snakehead

50

(b)

r = 0.77, p < 0.01

40

30

20

TnBP PFRs

49

PFR concentration (ng/g ww)

Total PFR concentration (ng/g ww)

(a)

r = 0.57, p < 0.01

10

r = 0.44, p < 0.05

25 20 15 10

r = 0.79, p < 0.01

5 0

0 0

3

6

9

12

15

18

0

Lipid content (%)

1

2

3

4

5

Lipid content in muscle (%)

647

Figure 3. Correlations between (a) the total PFR concentration and lipid content in

648

various tissues of the two fish species, (b) the concentrations of PFRs and lipid con-

649

tent in muscle tissue of all aquatic organisms.

24

Ratio of other tissue to liver (Cother/(Cother+ Cliver))

1.0

Mud carp Snakehead 0.8

0.6

0.4

0.2

0.0

Muscle

Bladder

Skin

Kindey

Gill

650

Figure 4. Comparisons of the ratios of other tissue to liver in two fish species. The

651

error bars represent standard deviations.

25

4.5

TnBP TPhP PFRs

3.0

7.5

6.0

Ln C (ng/g lw)

Ln C (ng/g ww)

r = 0.63, p < 0.01 1.5

r = 0.61, p < 0.01 0.0

4.5

3.0 -1.5

r = 0.50, p < 0.05 1.5 -3.0 9

10

11

12

13

14

9

10

11

12

13

14

15

δ N (‰)

15

δ N (‰)

652

Figure 5. Correlations between the PFR concentration (Ln transformed) and the

653

trophic levels of the aquatic organisms.

26

654

Table 1. PFR concentrations (mean ± SD) in aquatic organisms (ng/g ww), water (ng/L), and sediments (ng/g dw). Sample /muscle N (a)

TCEP

TCPP

TDCPP

TPhP 0.32 ± 0.25

EHDPP

TEHP

TCrP

∑PFRs

Snakehead

5

1.0 ± 0.97

0.84 ± 0.22

1.8 ± 0.79

0.14 ± 0.23

Catfish

2

2.7 ± 0.23

0.51 ± 0.48

1.6 ± 0.77

ND

Large mud carp

5

4.7 ± 1.1

0.36 ± 0.081

2.5 ± 0.30

ND

2.1 ± 1.7

0.27 ± 0.096 0.023 ± 0.017 0.11 ± 0.050

10 ± 2.2

12 ± 1.3

0.39 ± 0.22

3.4 ± 0.74

0.11 ± 0.026

0.79 ± 0.40

0.20 ± 0.049 0.061 ± 0.010 0.12 ± 0.031

17 ± 1.7

0.48 ± 0.44 0.078 ± 0.041 0.24 ± 0.071

11 ± 1.7

Small mud carp 5(60)

655

TnBP

0.10 ± 0.18 0.043 ± 0.062 0.067 ± 0.069

4.3 ± 2.6

0.23 ± 0.010 0.081 ± 0.11 0.023 ± 0.011 0.036 ± 0.034

5.1 ± 1.6

Crucian carp

5(16)

6.1 ± 0.45

0.41 ± 0.15

2.4 ± 0.64

0.22 ± 0.45

1.2 ± 0.59

Prawn

5(50)

3.5 ± 3.8

1.1 ± 0.84

13 ± 10

3.2 ± 0.98

12 ± 4.8

0.24 ± 0.10

0.36 ± 0.047

0.56 ± 0.62

34 ± 13

Water

3

51 ± 10

99 ± 15

76 ± 8.7

13 ± 2.5

13 ± 1.5

1.7 ± 0.083

0.32 ± 0.17

0.62 ± 0.17

255 ± 20

Sediment

2

50 ± 67

18 ± 10

43 ± 24

4.5 ± 2.5

13 ± 7.1

4.0 ± 2.3

0.14 ± 0.074

1.1 ± 0.63

135 ± 74

N (a) Numbers of pooling samples (individual samples); ND, undetected values.

27

656

Table 2. Correlations between PFR concentrations and lipid contents in all tissues of

657

large mud carp and snakehead. Correlation TnBP

TCEP

0.82

0.048

0.44

-

0.52

0.79

0.53

0.47

0.000

0.81

0.015

-

0.004

0.000

0.003

0.011

0.35

0.32

0.46

-

0.54

0.085

0.34

0.32

0.080

0.11

0.017

-

0.004

0.68

0.088

0.12

Large mud carp

r p

Snakehead

r p

658

TCPP TDCPP TPhP EHDPP TEHP

-, not available.

28

TCrP

Highlights • PFR bioaccumulation exhibited species-specific profiles • TnBP, TCEP, TCPP, and TPhP were generally the dominant PFRs • Log BCFs and log BSAFs were significantly correlated with log KOW • PFR level was positively correlated with lipid content for a given species • Trophic dilution for TnBP and TPhP were found in the aquatic food web