Halogenated Flame Retardants in Predator and Prey Fish From theLaurentian Great Lakes: Age-Dependent Accumulation and TrophicTransferGuanyong Su,†,‡,§ Robert J. Letcher,*,†,§ Daryl J. McGoldrick,∥ and Sean M. Backus∥

†Ecotoxicology and Wildlife Health Division, Environment and Climate Change Canada, National Wildlife Research Centre, CarletonUniversity, Ottawa, Ontario K1A 0H3, Canada‡Jiangsu Key Laboratory of Chemical Pollution Control and Resources Reuse, School of Environmental and Biological Engineering,Nanjing University of Science and Technology, Nanjing 210094, P. R. China§Department of Chemistry, Carleton University, Ottawa, Ontario K1S 5B6, Canada∥Water Science & Technology Directorate, Science and Technology Branch, Environment and Climate Change Canada,Canada Centre for Inland Waters, Burlington, Ontario L7S 1A1, Canada

*S Supporting Information

ABSTRACT: The identification, persistence, accumulation and trophic transfer of25 polybrominated diphenyl ether (PBDE) congeners, 23 non-PBDE halogenatedflame retardants (NPHFRs), 4 polybrominated-diphenoxybenzenes (PB-DiPhOBzs) and6 methoxylated (MeO−) PB-DiPhOBzs were investigated in predator and prey fishcollected in 2010 from sites in the North American Great Lakes of Ontario (n = 26)and Erie (n = 39). Regardless of locations or species, 20 PBDEs and 12 NPHFRswere quantifiable in at least one of the 65 analyzed samples, and polybrominated-1,4-diphenoxybenzenes (PB-DiPhOBzs) and MeO-PB-DiPhOBzs were not detectable in anyof analyzed samples. Among the FRs, the greatest concentrations were the ∑PBDE,ranging from 1.06 (Rainbow Smelt, Lake Erie) to 162 (Lake Trout, Lake Ontario) ng/gwet weight (ww), which was followed by mean HBCDD concentrations ranging ND to17.3 (Lake Trout, Lake Ontario) ng/g ww. The remaining FRs were generally notdetectable or at sub-ppb levels. In most of cases, FR concentrations in samples from LakeOntario were greater than those from Lake Erie. Strong and significant positive linear relationships occurred betweenlog-normalized FR concentrations (ww or lipid weight (lw)) and ages of the top predator Lake Trout (n = 16, from LakeOntario), and the estimated FR doubling ages (T2) were 2.9−6.4 years. For Walleye from Lake Erie, significantly positive linearrelationships were also observed for some FRs, but the linear relationships generally became negative after FR concentrationswere normalized with lipid weight. This study provides novel information on FR accumulation in aquatic organisms, and for thefirst time, significant positive linear relationships are reported between log-normalized FR concentrations (lw or ww) and ages ofLake Trout from the Great Lakes.

■ INTRODUCTION

Flame retardants (FRs) are a grouping of chemicals that areused to hinder the ignition and spread of fire, and thus areadded to manufactured materials such as plastics, electronicproduct, textiles, surface finishes, and coatings.1,2 Going backseveral decades, legacy FRs include commercial mixtures of penta/octa bromodiphenyl ethers (penta/octa-BDEs), hexabromobi-phenyls (PBBs) and hexabromocyclododecane (HBCDD).Between 2009 and 2014 these FRs were listed as persistentorganic pollutants (POPs; Annex A) under the StockholmConvention due to their proven bioaccumulation, long-rangetransport, and adverse biological effect activity.3 Environmentalpersistence and impacts of these legacy FRs are likely to con-tinue since they remain in the products that contain them.4−6

Increasing restrictions on legacy FRs has led to the increasingdemand for “novel” FRs as replacements, and are thus emergingFRs of environmental concern, for example, Dechlorane Plus

(DDC−COs), decabromodiphenyl ethane (DBDPE), 1,2-bis-(2,4,6-tribromophenoxy)ethane (BTBPE), 2-ethylhexyl-2,3,4,5-tetrabromobenzoate (EHTBB), tetrabromobisphenol A-bis(2,3-dibromopropylether) (TBBPA-DBPE).7,8

Environmental monitoring studies have demonstrated thatboth legacy and emerging FRs are pervasive in the environmentsamples (i.e., bird eggs, water, fish, atmospheric dust) fromthe Laurentian Great Lakes of North America, which are animportant part of North America’s ecosystem and the largestsurface freshwater system on Earth.9,10 Su et al. recentlyinvestigated 14 PBDEs and 23 non-PBDEs halogenated FRs(NPHFRs) in 115 herring gull egg samples collected in 2012 or

Received: May 5, 2017Revised: June 15, 2017Accepted: June 21, 2017Published: June 21, 2017

Article

pubs.acs.org/est

© 2017 American Chemical Society 8432 DOI: 10.1021/acs.est.7b02338Environ. Sci. Technol. 2017, 51, 8432−8441

2013 from 20 colonies spanning the Great Lakes basin, andfound that PBDEs, HBCDD, syn- and anti-DDC−CO, BB-153,BB-101, 2,4,6-tribromophenyl allyl (TBP-AE), pentabromo-moethylbenzene (PBEB), BTBPE, α-1,2-dibromo-4-(1,2-dibro-moethyl)-cyclohexane (TBECH), β-TBECH, pentabromo-p-xylene (pTBX), octabromo-1,3,3-trimethyl-1-phenyl indane(OBTMPI) and hexabromobenzene (HBB) were quantifiablein at least one of these analyzed samples.5 Most importantly,concentrations of some FRs (i.e., decabromodiphenyl ether(BDE-209), HBCDD and DDC−COs) were significantlygreater than measured in earlier egg samples, emphasizing theimportance continued monitoring of these FRs in the LaurentianGreat Lakes.5 The occurrence of FRs has been frequentlyreported in Laurentian Great Lakes fish samples. Zhu et al.showed concentrations of these FRs in Great Lakes Lake Troutincreased exponentially between 1980 and 2000.11 Similarly,PBDEs, DDC−COs, PBEB, EHTBB, is(2-ethylhexyl)-tetrabro-mophthalate (BEHTBP), and TBBPA-BDBPE were reported inatmospheric particle phase samples from the same areas in theGreat Lakes.12,13

Environmental assessments evaluate the risks of chemicalexposure and bioaccumulation in humans and biota.14 Trophictransfer in Great Lakes aquatic biota has been well-examinedfor some FRs from some specific locations. In a recentstudy, PBDE concentrations were determined in a mixedfood web of native and non-native aquatic species in LakeErie, and non-native prey species (e.g., rainbow smelt) werefound to significantly contribute to PBDE biomagnification.15

Trophic transfer of DDC−COs was also reported in a marinefood web from Liaodong Bay, China, where lipid equivalentconcentrations of anti-DDC−CO were positively correlatedwith trophic level and with a trophic magnification factor(TMF) of 5.6, suggesting the trophic magnification potentialof anti-DDC−CO.16 In Lake Winnipeg (MB, Canada), in theexamination of PBDEs, HBCDD, DBDPE, and BTBPE, strongpositive linear relationships were found for BDE-47, BDE-209,and DBDPE concentrations in relation to trophic level, alsosuggesting these FRs biomagnify in the Lake Winnipeg foodweb.17

The present study has the following objectives: (1) to investi-gate the contamination levels of 48 legacy and replacement FRs(25 PBDEs and 23 non-PBDE halogenated flame retardants(NPHFRs)) in Lake Ontario and Lake Erie, the LaurentianGreat Lakes; (2) to evaluate whether bioaccumulation of FRs in

fish (i.e., Lake Trout in Lake Ontario, Walleye in Lake Erie)is related with biological characteristics (i.e., sex, age); and(3) to examine the relationships between FR concentrationsand trophic levels in aquatic biota from the Laurentian GreatLakes.

■ MATERIALS AND METHODS

Standards and Chemicals. The 48 target FRs (25 PBDEsand 23 other flame retardants (NPHFRs)), along with their fullchemical names and chemical structures are shown in Figure 1.All chemicals and standards were purchased from WellingtonLaboratories Inc. (Guelph, ON, Canada), with the exception ofoctabromo-1,3,3-trimethyl-1-phenylindane (OBTMI) whichwas provided by Dr. Åke Bergman (Stockholm University,Sweden). The standards of 3 polybrominated-1,4-diphenox-ybenzenes (PB-DiPhOBzs) and 5 methoxylated (MeO−)PB-DiPhOBzs were synthesized and kindly provided byAccuStandard Inc. (New Haven, CT), and the specific chemicalstructures are provided in Supporting Information (SI) Figure S1.

Sample Collection. Collections made by Environment andClimate Change Canada (ECCC) as part of ongoing contaminantmonitoring and surveillance activities in the Great Lakes, werethe source of all aquatic samples in the present study. Detailedinformation on sampling, sample preparation, and storagemethods have been described previously,18 and we have alsoreported in detail elsewhere all biological information on 65aquatic samples.19 In brief, a total for the 65 aquatic bioticsamples were collected in 2010 from Lake Ontario(n = 26) and Lake Erie (n = 39) (SI Figure S2). The aquaticsamples from Lake Ontario were comprised of six species,including Alewife (Alosa pseudoharengus; n = 2), DeepwaterSculpin (Myoxocephalus thompsonii; n = 2), Lake Trout(Salvelinus namaycush; n = 15), Rainbow Smelt (Osmerusmordax; n = 2), Round Goby (Neogobius melanostomus; n = 2)and Slimy Sculpin (Cottus cognatus; n = 2). The aquatic samplesfrom Lake Erie were comprised of nine species, includingEmerald Shiner (Notropis atherinoides; n = 3), Freshwater Drum(Aplodinotus grunniens; n = 3), Lake Trout (Salvelinusnamaycush; n = 7), Rainbow Smelt (Osmerus mordax;n = 3), Round Goby (Neogobius melanostomus; n = 3), TroutPerch (Percopsis omiscomaycus; n = 3); Walleye (Sander vitreus;n = 10), Whiter Perch (Morone americana; n = 3) and YellowPerch (Perca f lavescens; n = 3). After capture, the fish wereimmediately frozen on dry ice and transported to the laboratory

Figure 1. Chemical structures of the target flame retardant compounds in this study. The hydrogen atoms are omitted for clarity.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b02338Environ. Sci. Technol. 2017, 51, 8432−8441

8433

where they are partially thawed, weighed, measured, and sexedby visual assessment of their gonadal tissues. Ages of LakeTrout and Walleye were determined using coded wire tagswhen present. For fish not containing tags, ages were estimatedusing scale samples. All remaining portions of the fish, includinginternal organs, are then homogenized for FR analysis.Flame Retardant (FR) Analysis. Determination of these

target 48 brominated FRs in biotic samples was carried out inthe Letcher Laboratories at the NWRC (ECCC, Ottawa,Canada), and details on sample extraction and instrumentalanalysis have been described in detail in our previouspublications.5,6,20,21 In brief, approximately 2.0 g of the biotichomogenate (wet weight, ww) was accurately weighed andhomogenized with precleaned diatomaceous earth (DE). Themixture was spiked with BDE-30 and -156 and 13C12−BDE-209as internal standards, and subjected to accelerated solventextraction (ASE) using dichloromethane (DCM). ASE extractswere evaporated down to 2 mL following by residual moistureremoved by filtering through sodium sulfate. A 10% volumewas taken and lipid content was determined gravimetrically.The remaining extract volume was concentrated under gentlenitrogen, and subject to high performance-gel permeationchromatography (HP-GPC) (Waters, Milford, MA) that wasoperated using DCM at a flow rate of 5 mL/min. The collectedfraction was re-evaporated to approximately 0.5 mL, and furthercleaned-up on a silica LC-Si SPE cartridge (500 mg X 6 mL;6 g; J.T. Baker) with 8 mL of 5% DCM/hexane (v/v) solvent.22

The collected fraction was evaporated under gentle nitrogenflow down to 1 mL, solvent exchanged into 2,2,4-trimethyl-pentane (TMP) and re-evaporated under nitrogen to around250 μL. After quantitative transfer to a preweighed brown glassGC vial with insert and cap, the final sample fraction was thenready for analysis by gas chromatography (6890 GC) -singlequadrupole mass spectrometry (5973N MS) (GC-MS; (AgilentTechnologies, Mississauga, ON, Canada)) operated in theelectron capture negative ion (ECNI) mode. A 15 m DB-5 HTGC column (0.25 mm i.d., 0.1 mm, J&W Scientific, Agilent)was used. The sample injector was operated at 240 °C and inthe pulsed-splitless mode. The initial GC oven temperaturewas 100 °C and held for 2 min, then to 250 °C at a rate of25 °C/min, then to 260 °C at 1.5 °C/min, and finally to325 °C at 25 °C/min. This final temperature was held for7 min. Brominated FR quantification was achieved via selectedion monitoring (SIM) for m/z 79Br− and 81Br−. However, theSIMs for BDE-209 (m/z 487), 13C12−BDE-209 (m/z 495), andthe syn- and anti-DDC−CO isomers (m/z 652) were different.For quality assurance and quality control, for each batch of

biotic samples, one fish (from Ottawa market) tissue samplefortified with target chemicals (Figure 1 and SI Figure S1) wasalso analyzed to ensure good recoveries of the target analytes.During the analysis, some background contamination wasobserved for BDE-47, BDE-100, BDE-99, and BDE-209. Thus,for each batch of extractions, one blank sample was also includedto investigate the possible background contamination during theanalysis. FRs in the samples were background subtracted if FRcontamination was present and the background peak responsewas >5% of that in a given fish sample fraction (i.e., BDE-209).The method limits of detection (MLODs) were based on asignal-to-noise ratio (S/N) of 3 and ranged from 0.001 to0.1 ng/g ww depending on the specific FR. Recoveries of theinternal standard FRs were generally >80% and <120%. FRquantification was via an internal standard approach, and thusconcentrations were corrected for recovery.

FR Concentration As a Function of Fish Growth. Thedoubling age (T2) was defined as the rate that FR concentrationsdoubled annually as a function of growth year of the fish. T2 wasderived from the slope of the plots of natural log FR concentra-tions (wet or lipid weight) versus fish age (in years) as follows:

= + ×A SLn[FR] [age]

=T Ln2/S2

where S and A is the slope and intercept of fitted curves,respectively. [FR] and [age] were FR concentrations (expressedas ng/g ww or ng/g lw) and fish ages, respectively.When the slope (S) of a fitted curve was lower than 0, the

half-life (T1/2) over the growth period/age could be calcu-lated. T1/2 was defined as a growth-age time period when theFR concentrations decreased by 50% in fish bodies. T1/2 wascalculated as follows:

= −T Ln2/S1/2

Fish Trophic Levels (TLs) and Trophic MagnificationFactors (TMFs). The present aquatic samples had beenanalyzed in another study for organophosphate ester (OPE)FRs.19 As we already reported, the stable isotope determi-nations for the 15/14N and 13/12C ratios were by theEnvironmental Isotope Laboratory (EIL; University of Waterloo,ON, Canada).23 TLs were assigned relative to plankton using theequation described in previous publications.19 Insufficientmaterial remaining after chemical analysis meant that δ15Nanalysis for plankton samples was not possible. Thus, 2009-collected Lake Ontario plankton (n = 4; 8.65 ‰) was used inthe present study.19 Individual sample TLs were determinedaccording to TLsample = 2 + (δ15Nsample - δ

15Nplankton) ÷ ΔN.Here, ΔN is trophic enrichment factor that is estimated to be3.4‰ based on a previous publication.24

The TMFs were based on the entire food web of LakeOntario (n = 26) or Lake Erie (n = 39), and derived from theslope of the plots of natural log concentrations (wet weight orlipid weight) versus TL as follows:

= + ×Ln[FR] A S [TL ]sample

= eTMF S

where [FR] is FR concentration (ng/g ww) in the aquaticsamples; S is the slope of the plots of natural log concentrations(wet weight) versus TL.

Statistical Analysis. GraphPad Prism 5 software was usedfor statistical analysis and data visualization, and was conductedfor the FRs only with quantifiable concentrations. Linearregression analysis was carried out for the relationships betweenlog-normalized FR concentrations and fish ages, and also forthe curve fitting between log-normalized FR concentrations andTL. Unpaired t tests were performed to assess FR concen-tration differences between male and female fish. For allstatistical analysis, the significance level was set at 0.05. UsingR software, principal component analysis (PCA) was carriedout to assess for differences in PBDE congener patterns amongthe fish species. All PCA data sets were expressed as a per-centage of the ∑PBDE concentration. The R code wasobtained from the vignette source ‘ch-outlier.rnw’.

■ RESULTS AND DISCUSSIONOverall Concentrations of Polybrominated Diphenyl

Ethers (PBDEs). Twenty-one BDE congeners, including

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b02338Environ. Sci. Technol. 2017, 51, 8432−8441

8434

BDE-47, -100, -119, -154, -153, -99, -28, -49, -85/155, -209, -66,-183, -203, -207, -17, -15, -77, -205, -138, and -206, werequantifiable in at least one of the n = 65 fish samples collectedfrom Lake Ontario or Lake Erie, whereas BDE-3, -7, −71,and -181 were not detectable in any of analyzed samples(SI Table S1 and Table S3). Among these 65 analyzed bioticsamples, ∑PBDE concentrations varied dramatically with arange from 1.06 ng/g ww (33.2 ng/g lw; Rainbow Smelt fromLake Erie) to 162 ng/g ww (1070 ng/g lw; Lake Trout fromLake Ontario), and the BDE-47, -99, -100, -153, and -154congeners were the dominant congeners constituting >80%of the ∑PBDE concentrations of all of the 65 analyzed bioticsamples. The pattern of BDE congeners was generally com-parable with previous studies in the same area.5,25 Similar toprevious reports,17 in the present study different speciesexhibited large differences in PBDE concentrations. Forexample, the mean ∑PBDE concentrations in top predatorLake Trout from Lake Ontario was as great as 94.8 ng/g ww(522 ng/g lw), as compared to Deepwater Sculpin fromthe same area were as low as 4.58 ng/g ww (84.3 ng/g lw).Lake Trout from Lake Ontario contained significantly greater∑PBDE concentrations than those from Lake Erie. Greaterconcentrations in Lake Ontario fish as compared to those inLake Erie were consistent with a previous study that reported∑PBDE in edible portions of Great Lakes fish and the spatialtrend was Lake Ontario ≫ Erie ≈ Huron ≈ Superior.25 Thiswas also consistent with ∑PBDE levels in whole fish frombasin-wide sampling where the order was Lake Ontario >Superior > Michigan > Huron > Erie.As shown from the PCA, for the different fish species

there were contrasting BDE congener patterns (Figure 2A andSI Figure S3). Greater than 95% of the combined overall datavariability accounted for by PC 1 and PC 2 (93.6% and 3.1%for Lake Erie, 89.6% and 8.0% for Lake Ontario). As indicatedby PCA biplot of Lake Erie or Ontario fish, there are cleardifferences with the proportion of BDE-99 and BDE-47 to∑PBDE concentrations, whereas there were no clear differ-ences for the proportion of the other BDE congeners (i.e.,BDE-28, BDE-49, BDE-66, BDE-100, BDE-119, BDE-85/155,BDE-154, BDE-153, and BDE-209) among fish species. Especiallyfor fish from Lake Erie, the proportion of BDE-47 appeared tobe differentiated along PC1 with greater proportions of fishoccupying higher trophic levels, whereas BDE-99 appears todifferentiate along PC1 with higher proportions in fish occupying

the lower trophic levels (Figure 2A). Correlative relationshipswere further investigated between trophic levels and propor-tions of BDE-47 or BDE-99 to ∑PBDEs for n = 39 individualfish from Lake Erie (Figure 2B and Figure 2C). A significantand positive correlative relationship (Spearman r = 0.3617;p = 0.0236) was observed for BDE-47, whereas a significant andnegative correlative relationship (Spearman r = −0.4120;p = 0.0102) was observed for BDE-99. Relatively greateraccumulation of BDE-47 as compared to BDE-99 could be theresult of various factors including trophic magnification factors,uptake rate and possibly metabolism. Previous studies havedemonstrated that the lower brominated BDE congenersgenerally have greater trophic magnification factors, and thatsmaller fish sizes generally related to higher uptake rates.15,26,27

Overall Concentrations of Non-PBDE halogenatedFlame Retardants (NPHFRs). Among the 23 NPHFRsanalyzed, BB-153 (100%; detection frequencies), BB-101(83.1%), syn-DDC−CO (87.7%), anti-DDC−CO (89.2%),HBCDD (83.1%), pTBX (72.3%), PBEB (56.9%), TBP-AE(38.5%), PBT (16.9%), DBDPE (10.8%), α/β-DBE-DBCH(7.7%) and TBCT (6.2%) were quantifiable in at least one ofthe n = 65 biotic samples collected from Lake Ontario and Erie in2010, whereas DPTE, PBPAE, HBB, PBBA, EHTBB, HCDBCO,PBPA-DBPE, BTBPE, BEHTBP, and OBTMI were notdetectable in any of analyzed samples (Table 1 and SI Table S2).Used as an additive FR for over 30 years, hexabromocyclodoe-

cane (HBCDD) is found many products including upholsterytextiles, electrical equipment and polystyrene foam.28 HBCDD hasbeen frequently detected in Laurentian Great Lakes samples, thatis, bird eggs, fish, air or water samples.5,6,29 HBCDD wasquantifiable in 100% of present biotic samples collected from LakeOntario with concentrations ranging from 0.34 to 32.5 ng/g ww(16.1−191 ng/g lw), which are generally comparable to thebasin-wide detectable range of HBCDD (4−122 ng/g lipid) infish from the Laurentian Great Lakes.29 Like PBDEs inLake Ontario, the greatest mean HBCDD wet weight concen-trations were in Lake Trout (17.3 ng/g ww; 97.3 ng/g lw),which was followed by Slimy Sculpin (5.97 ng/g ww;135 ng/g lw), Alewife (2.52 ng/g ww; 21.7 ng/g lw), RainbowSmelt (2.44 ng/g ww; 62.9 ng/g lw), Deepwater Sculpin(1.59 ng/g ww; 28.5 ng/g lw) and Round Goby (0.46 ng/g ww;98.6 ng/g lw). Biotic samples from Lake Erie containedclearly lower HBCDD concentrations than those from LakeOntario. Within Lake Erie, the greatest mean HBCDD con-

Figure 2. Biplot from the principal component analysis (principal component (PC) 1:93.6%, PC2:3.1%) of polybrominated diphenyl ether (PBDE)congener composition patterns in different fish species (A), correlation analysis between trophic level and proportion of BDE-47 (B) or BDE-99(C) to ∑PBDEs in n = 39 individual fish samples, from Lake Erie, the Laurentian Great Lakes. In Figure 2A, ES, WP, LT, RS, TP, FD, RG, YP, andW represented Emerald Shiner, White Perch, Lake Trout, Rainbow Smelt, Trout Perch, Freshwater Drum, Round Goby, Yellow Perch, and Walleye,respectively. The number behind the fish species is the calculated trophic level (Figure 2A).

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b02338Environ. Sci. Technol. 2017, 51, 8432−8441

8435

Table

1.Arithmetic

Means

andRanges(ng/gwet

Weight)

OfIndividu

alNon

-PBDEHalogenated

Flam

eRetardants(N

PHFR

s)and∑PBDEconcentrations

inFish

from

Lakesof

Ontario

andErie,

theLaurentian

Great

Lakes

lipid/%

Aa

LbWc

Nd

δ15N

TBPA

EpT

BX

PBEB

BB-101

BB-153

HBCDD

syn-DDC-CO

anti-DDC-CO

∑PB

DEs

lakes

common

name

38.5%g

72.3%

56.9%

83.1%

100.0%

83.1%

87.7%

89.2%

100%

Ontario

Alewife

12.1 (10.1−

14.1)

NAe

18.3

58.3

211.9 (11.7−

12.2)

0.046

(0.040−

0.046)

0.0054

(0.0054−

0.0054)

ND

0.029(N

D−

0.058)

0.10 (0.097−

0.11)

2.52 (2.28−

2.75)

0.010

(0.010−0.010)

0.012

(0.011−0.013)

11.7 (9.28−

11.7)

Deepw

ater

Sculpin

5.43 (4.92−

5.95)

NA

8.5

7.9

215.2 (15.1−

15.3)

0.0053

(0.0048−

0.0059)

0.013

(0.008−

0.019)

ND

ND

0.051

(0.051−

0.052)

1.59 (0.99−

2.19)

0.017

(0.016−0.018)

0.021

(0.020−0.022)

4.58 (4.13−

5.03)

Lake

Trout

18.1 (7.23−

26.4)

6.6(3−10)64.8

4310

9F/6M

16.8 (15.7−

18.5)

0.091(N

D−

0.173)

0.010(N

D−

0.028)

ND

0.31

(ND−0.74)

0.80 (0.17−

1.40)17.3 (2.7−32.5)

0.034(N

D-

0.079)

0.036(N

D-

0.086)

94.8 (19.1−

162)

Rainbow

Smelt

4.29 (2.50−

6.07)

NA

18.1

7.9

215.5 (15.0−

15.9)

0.025

(0.016−

0.034)

ND

ND

0.019(N

D−

0.039)

0.084

(0.036−

0.133)

2.44 (1.93−

2.94)

0.009

(0.008−0.010)

0.010

(0.007−0.012)

15.5 (7.50−

23.5)

Round

Goby

0.55 (0.27−

0.83)

NA

10.2

16.8

214.3 (14.3−

14.3)

0.0054

(0.0018−

0.0090)

0.030

(0.002−

0.059)

ND

0.020(N

D−

0.040)

0.041

(0.039−

0.043)

0.46 (0.34−

0.57)

0.016

(0.013−0.019)

0.034

(0.028−0.040)

5.53 (5.01−

6.06)

SlimyScul-

pin

3.85 (2.46−

5.25)

NA

910.5

217.1 (17.0−

17.2)

0.017

(0.016−

0.019)

0.049

(0.038−

0.060)

ND

ND

0.31 (0.29−

0.34)5.97 (1.95−

10.0)

0.11 (0.11−

0.11)

0.11

(0.10−

0.11)16.1 (12.6−

19.6)

Erie

Emerald

Shiner

2.19 (1.78−

2.75)

NA

2.19

2.19

32.19 (1.78−

2.75)

ND

ND

0.023

(0.019−0.027)

0.064

(0.055−0.070)

0.057

(0.046−

0.072)

0.45 (0.34−

0.53)

0.005

(0.003−0.007)

0.005

(0.004−0.006)

3.60 (3.35−

3.72)

Freshw

ater

Drum

2.18 (0.91−

3.53)

NA

22.8

127

39.67 (9.21−

10.3)

ND

0.026

(0.011−

0.049)

0.007

(0.006−0.008)

0.037

(0.033−0.040)

0.13 (0.13−

0.15)ND

0.010

(0.006−0.014)

0.010

(0.008−0.012)

3.14 (3.04−

3.21)

Lake

Trout

14.5 (12.3−

17.6)

4.38 (3−10)

54.0

2290

4F/3M

6.13 (4.05−

8.91)

ND

0.009

(0.004−

0.022)

0.029

(0.014−0.091)

0.14

(0.09−

0.33)0.18 (0.13−

0.42)2.76 (1.20−

6.26)

0.004(N

D-

0.011)

0.005(N

D-

0.011)

11.8 (7.21−

33.7)

Rainbow

Smelt

3.57 (3.10−

4.43)

NA

11.4

9.4

310.3 (10.3−

10.3)

ND

0.003

(0.002−

0.005)

0.003

(0.002−0.004)

0.019

(0.017−0.022)

0.015

(0.012−

0.018)

0.23 (0.20−

0.27)

ND

0.001(N

D-

0.003)

1.15 (1.06−

1.32)

Round

Goby

0.77 (0.60−

0.99)

NA

7.73

8.05

311.0 (10.6−

11.2)

ND

0.005

(0.004−

0.007)

0.005(N

D-0.008)0.031

(0.025−0.040)

0.13 (0.08−

0.18)ND

0.006

(0.004−0.007)

0.006

(0.005−0.006)

3.71 (2.60−

5.17)

Trout

Perch

0.94 (0.75−

1.16)

NA

9.07

6.9

310.9 (10.6−

11.5)

ND

ND

0.006

(0.006−0.007)

0.028

(0.023−0.032)

0.18 (0.16−

0.22)0.23

(ND-

0.35)

0.008

(0.006−0.009)

0.10 (0.009−0.012)

4.16 (3.90−

4.64)

Walleye

9.58 (3.23−

14.0)

3.4(2−6)

47.3

1420

5F5M

9.11 (2.49−

12.6)

ND

0.013

(0.005−

0.027)

0.055

(0.040−0.074)

0.20

(0.11−

0.29)0.19 (0.13−

0.25)1.19

(ND-

2.57)

0.006(N

D-

0.011)

0.007(N

D-

0.015)

11.0 (6.50−

17.3)

White

Perch

4.58 (4.44−

4.76)

NA

15.0

47.2

39.43 (8.39−

10.0)

ND

0.023

(0.015−

0.029)

0.010

(0.009−0.010)

0.062

(0.050−0.081)

0.08 (0.06−

0.10)0.24

(ND-

0.42)

0.010

(0.009−0.012)

0.011

(0.011−0.011)

3.27 (2.52−

4.58)

Yellow

Perch

0.96 (0.88−

1.09)

NA

16.5

45.4

310.7 (9.52−

11.8)

ND

ND

0.011(N

D-0.020)0.088

(0.073−0.104)

0.21 (0.15−

0.32)0.19

(ND-

0.58)

0.008

(0.005−0.011)

0.010

(0.008−0.012)

5.45 (4.37−

6.24)

a“A”means

age(unit:year).b“L”means

length

(unit:cm

).c “W”means

weight(unit:g).d“N

”means

number,samplesize.“9F/6M

”means

that

thesamplesize

is15

with

9femaleand6male

fishes.e“N

A”means

notavailable.f “ND”means

notdetectable.gdetectionfrequency.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b02338Environ. Sci. Technol. 2017, 51, 8432−8441

8436

centrations (wet weight) were for Lake Trout (2.76 ng/g ww).HBCDD was not detectable in any of Freshwater Drum andRound Goby samples.Commercial production of polybrominated biphenyls (PBBs),

an additive FR, began in the 1970s, and its manufacture wasdiscontinued in the United States in 1976 due to the agriculturecontamination episode in Michigan in 1973−1974.30 Despitethe short period of production and usage, in the present studythe detection frequencies of BB-101 and -153 were as high as83.1% and 100%, respectively, and indicating their persistence inLake Ontario and Erie fish. The high detection frequencies ofBB-101 and BB-153 are consistent with those reported inGreat Lakes herring gull eggs.5 The greatest mean BB-101 and−153 concentrations (wet weight) were 0.31 and 0.80 ng/g ww,respectively, both of which were determined in Lake Troutsamples of Lake Ontario. Like BB-101 and BB-153 in herringgull eggs,5 there was a significant linear correlation relationship(p < 0.0001; r2 = 0.8786) between BB-101 and BB-153concentrations in fishes from Lake Ontario and Lake Erie.DDC−COs isomers are used as polychlorinated FRs and

produced by Oxychem. Initial reports of these DDC−COisomers in the Great Lakes environment was in 2006, wheresyn- and anti-DDC−CO were found in air samples.31 In thepresent study, syn- and anti-DDC−CO were quantifiable in87.7% and 89.2% of the 65 analyzed fish samples from the LakeOntario and Lake Erie, and concentrations of syn- oranti-DDC−CO were consistently at subppb level. These highdetection frequencies but low contamination levels of DDC−COs were in a good agreement with previous Great Lakes fishmonitoring studies.29 In the present fish the concentrations ofanti-DDC−CO were in general greater than for syn-DDC−CO.This is similar to DDC−CO isomer profiles reported in fish29

and herring gull eggs32 from sites spanning the Great Lakes.Also, our results were similar to that of ring-billed gull liver andplasma from birds sampled in the St. Lawrence River, Canada.33

Brominated benzenes, including DPTE, EHTBB, PBP-DBPE,PBBA, HBB, PBPAE, PBT, PBEB, TBCT, pTBX, BEHTBP,OBTMI, TBPAE, have been used as FRs in various applications,that is, thermoset polyester resin, polybutyleneterephthalate,paper, textiles, electronics.7,34 Previous studies have reportedthe occurrence of HBB, pTBX and PBEB in Great Lakesatmospheric samples.10,35,36 Our present study detectedTBPAE (38.5%), pTBX (72.3%), PBEB (56.9%), PBT(16.9%), TBCT (6.2%) in at least one of 65 analyzed fishsamples, however, we could not detect any of the otherbrominated benzenes in any of the analyzed samples. PBEB wasnot detectable in any samples from Lake Ontario, but had highdetection frequency (37 out of 39) for Lake Erie fish sampleswith concentrations ranging from ND to 0.091 ng/g ww.The greatest mean PBEB concentration was 0.055 ng/g ww inLake Erie Walleye samples. Compared to PBEB, TBPAE wasquantifiable in 25 out 26 fish samples from Lake Ontario, butwas not detectable in any of 39 Lake Erie fish samples.The greatest mean TBPAE concentrations were 0.091 ng/g wwin Lake Trout samples from Lake Ontario. The low concen-trations of brominated benzenes in the present fish were inagreement with a recent monitoring study of fish or bird eggsamples collected in the Great Lakes basin.5,6,29

Decabromodiphenyl ethane (DBDPE) is an additive FR andstarting in the early 1990s was used as a FR replacement for thedeca-BDE formulation.7 DBDPE was detected in only 7 of65 fish samples in the present study. This result is consistent witha previous study where DBDPE was found in 3 of the 15 analyzed

fish samples.29 1,2-dibromo-4-(1,2-dibromoethyl)cyclohexane(TBECH) is an additive FR and contains equal amounts oftwo major diastereomers, α- and β-TBECH.7,37 α-/β-TBECHwere detected in 5 of 65 present fish samples. The greatest wetconcentration was 0.86 ng/g ww in one Lake Trout samplefrom Lake Erie. In a recent herring gull egg monitoring study,TBECH isomers were generally not detectable or at concen-trations not exceeding 0.48 ng/g ww.5

None of the target PB-DiPhOBz and MeO-PB-DiPhOBzcongeners were detectable in any of the 65 fish samples. Chenet al. first reported MeO-PB-DiPhOBzs in the eggs of herringgulls from colony sites in the Laurentian Great Lakes, andhypothesized that the MeO-PB-DiPhOBzs are degradation/metabolism products and sourced from the FR tetradecabro-mo-1,4-diphenoxybenzene (TeDB-DiPhOBz).38,39 More recentstudies have reported that TeDB-DiPhOBz has a shortphotolytic half-life in the order of minutes in organic solventsand when exposed to sunlight, and that enzymatic hydrox-ylation can occur to the photolytic and lower brominatedPB-DiPhOBzs products of TeDB-DiPhOBz as demonstrated inin vitro biotransformation assays using harvested wild herringgull and adult male Wister-Han rat liver.40−43 In the presentstudy, the lack of detection of PB-DiPhOBzs and MeO-PB-DiPhOBzs in Great Lake forage and predatory fish suggeststhat aquatic derived dietary items are also not a source ofPB-DiPhOBz congeners to herring gulls.38,39

Age- and Sex-Dependent Accumulation of FlameRetardants. Relationships between individual characteristics offish (i.e., sex, age, length, and weight) and FR concentrationswere examined for the Lake Trout from Lake Ontario (n = 15;nine female and six male), and the Walleye from Lake Erie(n = 10; five female and five male). The remaining species inour food web study were not included due to their small samplesizes.For most of detected FRs, significant differences were

observed on their concentrations between male and femaleWalleye from Lake Erie. It should be emphasized that sig-nificant differences were also observed for physiologicalcharacteristics (i.e., ages, length, and weight) between maleand female Walleye fishes. Among the FR concentrations forLake Trout from Lake Ontario, a significant difference (t test, p< 0.05) between male and female fishes was only observed forBB-101 (female > male). These results suggested that sex didnot influence FR concentrations in both Lake Trout andWalleye. This finding is consistent with a study on the influenceof fish sex on mercury/total-PCB concentrations in whichsignificant differences between sexes were detected in less than25% of the tests conducted.44

FR concentrations (wet weight) increased as a function ofincreasing age of the Lake Trout or Walleye. Significantly positivelinear correlative relationships were observed for 16 FRs(BDE-28, BDE-49, BDE-47, BDE-66, BDE-100, BDE-119,BDE-99, BDE-95/155, BDE-154, BDE-153, BDE-183,∑PBDEs, BB-101, BB-153, and HBCDD) in Lake Troutfrom Lake Ontario between log-normalized concentration(wet weight) versus fish age, and T2 values ranged from 2.9(BDE-47) to 3.9 (BDE-49) years (Table 2 and Figure 3). ForFR concentrations (wet weight) in Walleye from Lake Erie,significant positive linear correlative relationships were foundfor BDE-66, BDE-119, BDE-99, BDE-85/155, BDE-154,BDE-153, pTBX, BB-153, and HBCDD, and T2 ranged from2.0 (HBCDD) to 5.5 (BB-153) years (Table 2 and Figure 4).For ∑PBDE wet weight concentrations in Walleye, there was a

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b02338Environ. Sci. Technol. 2017, 51, 8432−8441

8437

positive linear correlative relationship but it was not statisticallysignificant (Pearson r = 0.5682; p = 0.0866) (Figure 4).To better understand the relationships between FR concen-

trations and fish ages, the FR concentrations were further nor-malized with lipid weight. For Lake Trout from Lake Ontario,there were significant positive linear correlative relationshipsfor all eight FRs between log-normalized concentration (lipidweight) versus fish age. The T2 values ranged from 5.2 (BDE-47) to6.4 (BDE-49) years, which were 2-times greater than those derivedfrom wet weight for each of FRs (Table 2 and SI Figure S4).

However, for lipid-corrected concentrations of FRs in Walleyefrom Lake Erie, these linear age relationships became nega-tive for all FRs, and where these negative relationships werestatistically significant for, e.g., BDE-28, BDE-49, BDE-47,BDE-100, ∑PBDEs, PBEB, BB-101, and BB-153, and with T1/2

values ranging from 2.3 (PBEB) to 4.6 (∑PBDEs) (Table 2and SI Figure S5).Relationships between chemical contaminant concentrations

and fish ages have been well established for mercury,45 andknowledge on the relationships between FR concentrations and

Table 2. Calculated Doubling Ages (T2) or Half-Life Years (T1/2) from the Slope of the Plots of Natural Log FR Concentrations(Wet or Lipid Weight) vs Age

lake trout (Lake Ontario; n = 15) walleye (Lake Erie; n = 10)

wet weight basis lipid weight basis wet weight basis lipid weight basis

chemicals Sc p T2 S p T2 S p T2 S p T1/2

PBDEs BDE-28 0.189 0.0039 3.7 NS NS −0.2349 0.0021 3.0BDE-49 0.18 0.0061 3.9 NS NS −0.2245 0.0051 3.1BDE-47 0.2362 0.0001 2.9 0.1275 0.0051 5.4 NS −0.1646 0.0178 4.2BDE-66 0.1867 0.0022 3.7 NS 0.2234 0.0329 3.1 NSBDE-100 0.2431 <0.0001 2.9 0.1344 0.0031 5.2 NS −0.1621 0.016 4.3BDE-119 0.1835 0.0002 3.8 NS 0.2124 0.0156 3.3 NSBDE-99 0.2249 0.001 3.1 0.1162 0.024 6.0 0.3159 0.0194 2.2 NSBDE-85/155 0.1858 0.0078 3.7 NS 0.2062 0.0059 3.4 NSBDE-154 0.2165 0.0001 3.2 0.1078 0.0137 6.4 0.2144 0.0053 3.2 NSBDE-153 0.222 0.0005 3.1 0.1132 0.0194 6.1 0.1503 0.0363 4.6 NSBDE-183 0.1914 0.0015 3.6 NS ND NS∑PBDEs 0.2297 0.0002 3.0 0.1232 0.0046 5.6 NS −0.1514 0.0255 4.6

NPHFRs pTBX NSa NS 0.2343 0.0129 3 NSPBEB NDb ND NS −0.3056 0.0002 2.3BB-101 0.2289 0.0001 3.0 NS NS −0.1869 0.0086 3.7BB-153 0.2197 0.0003 3.2 0.1109 0.0294 6.3 0.1271 0.0186 5.5 −0.158 0.0225 4.4HBCDD 0.2228 0.0003 3.1 0.1141 0.0294 6.1 0.348 0.0033 2.0 NS

a“NS” means no significant correlation was observed. b“ND” means this FR was not detectable in fishes. c“S” means slope of the fitted curves.

Figure 3. Chemical concentrations in Lake Trout of the Lake Ontario (ng/g wet weight; Log-transformed) versus fish ages (years) for selected flameretardants, see other parameters for the fitted curves in Table 2.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b02338Environ. Sci. Technol. 2017, 51, 8432−8441

8438

fish ages remains somewhat limited. Vives et al. reportedsignificant positive correlations between age and log-trans-formed liver concentrations for BDE-47, BDE-99, andBDE-100 in fish from high mountain lakes in Europe and inGreenland, and in a recent study measuring PBDEs in edibleportions of Great Lakes fish, researchers found that ∑PBDEconcentrations were related to fish length and lipid contentwhen all measurements including all 18 fish species werepooled.25,46 In the present Great Lakes fish study, clear andsignificant correlative relationships existed between fish ageversus log-normalized concentrations (wet weight or lipidweight) for Lake Trout and Walleye, and with measurabledoubling ages for the FRs under study (SI Figure S6 and S7).Trophic Magnification Factors (TMFs) of Flame

Retardants. Significant positive relationships (p < 0.05;slope > 0) were observed for most of the FRs from the plotof natural log wet weight concentrations versus TL (based onδ15N-values) for both Lake Ontario and Lake Erie fishes(SI Table S4). However, these significant positive relationshipsdisappeared when the FR concentrations were normalized withlipid weight (SI Table S5). The exceptions were BDE-28,BDE-47, BDE-119, BB-153, and HBCDD in Lake Ontariofishes, with TMF values of 1.60, 2.11, 2.33, 2.25, and 2.23,respectively, indicating their biomagnification potential in theLake Ontario aquatic food web. For fishes from Lake Erie, afterthe FR concentrations were normalized with lipid weight,the relationships for most of FRs were statistically significant(p < 0.05) but negative (slope < 0) (SI Table S5). Thedifference in the slopes of fitted curves (before and after lipidweight normalization) for Lake Erie fishes could be caused bythe difference of lipid content among different TL species.For example, we observed a positive and statistically significant(p < 0.0001) correlative relationship between lipid content andtrophic levels of fishes (n = 39) from Lake Erie (SI Figure S8).Based on previous studies, PBDEs appeared to show differentbiomagnification potentials in different food webs. For example,Kelly et al. determined PBDE concentrations in a CanadianArctic marine food web, and found that BDE-47 was the onlyBDE congener with a TMF statistically greater than 1.47

However, Hu et al. investigated PBDE trophodynamics in afreshwater food chain, and found that nine BDE congeners hadTMFs greater than 1.48 In a recent study that investigatedtrophic transfer of PBDEs in a mixed Lake Erie food web(i.e., native and non-native species), most of BDE congeners

showed TMFs greater than 1 with exceptions of BDE-99 andBDE-209.15 Different fish species that occupy the same relativeTL may have very different concentrations of contaminantsdepending the level of exposure and metabolic activity.49 In thepresent study, it is emphasized that the TMF values shouldbe interpreted with caution due to the lack of higher trophic(i.e., fish-eating bird) or lower trophic (i.e., invertebrate)species in the food web comparisons.

Implication for Future Studies. Over the past decadesenvironmental scientists have used fish as key indicators fordetermining the environmental health/pollution status ofaquatic ecosystems worldwide.11,50−54 Predatory and long-lived species of fish are especially useful for contaminant bio-monitoring. From surrounding waters as well as from their preyorganism diets, predatory fish are exposed to organic pollutantswith favorable bioaccumulation and biomagnification poten-tials.29,52,55 However, relationships between organic contami-nant concentrations and fish ages remain poorly understood.Our present study provides clear evidence that concentrationsof some FRs can increase as a function of the age of fish, that is,Lake Trout and Walleye, and we estimated the doubling agesfor wet or lipid weight-based concentrations of several FRs torange from 2.9 to 6.4 years in Lake Trout from the slopes of theplots of natural log concentrations versus fish age. The presentresults also demonstrated that concentrations of FRs in fishfrom the same water body can vary widely depending on the ageof the fish. In Lake Ontario for example,∑PBDE concentrationswere as low as 19.1 ng/g ww (264 ng/g lw) in a 3-year old LakeTrout, whereas in a 9-year old Lake Trout concentrations wereas great as 162 ng/g ww (1070 ng/g lw). The findings of thisstudy strongly suggest that fish age should be regarded as a keyfactor in future FR monitoring and surveillance in fish.Another critical fact related to human fish consumption in

the Great Lakes. Although numerous local guides to eatingfish have proposed that smaller fishes tend to contain lesscontaminants than larger fish of the same species,56 the under-lying knowledge is limited to several specific contaminants, thatis, mercury.45 Our results provide evidence that larger and olderLake Trout and Walleye contain greater concentrations ofFRs. Further research is warranted in studying the species-specific differences of FR bioaccumulation in Great Lakesfish, and determining the differences between whole-bodyhomogenates and edible parts of fishes in consideration of humanconsumption.

Figure 4. Chemical concentrations in Walleye of the Lake Erie (ng/g wet weight; Log transformed) versus fish ages (years) for selected flameretardants, see other parameters for the fitted curves in Table 2.

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b02338Environ. Sci. Technol. 2017, 51, 8432−8441

8439

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.est.7b02338.

Further details are given on concentrations (ng/g wwand ng/g lw) of individual PBDE congeners, concen-trations of (ng/g lw) of individual non-PBDE halo-genated FRs, trophic magnification factor (TMF, basedon both ng/g ww and ng/g lw), chemical structure oftarget PB-DiPhOBzs and MeO-PB-DiPhOBzs, samplinglocations, PCA analysis for PBDEs in fishes from LakeOntario, Log transformed FR concentrations (ng/g lw)versus fish (Lake Trout or Walleye) ages, biological char-acteristics (lipid content, length, weight) versus fish(Lake Trout or Walleye) ages (PDF)

■ AUTHOR INFORMATIONCorresponding Author*Phone: 1-613-998-6696; fax: 1-613-998-0458; e-mail: robert.letcher@canada.ca.ORCIDRobert J. Letcher: 0000-0002-8232-8565NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThe Chemicals Management Plan (ECCC), the NaturalScience and Engineering Research Council (NSERC) ofCanada (to R.J.L.) provided financial support. The collectionof the aquatic biota samples and subsequent processing wereaccomplished by Michael Keir, Mandi Clark, and Mary Malecki(ECCC)..

■ REFERENCES(1) Covaci, A.; Harrad, S.; Abdallah, M. A.; Ali, N.; Law, R. J.;Herzke, D.; de Wit, C. A. Novel brominated flame retardants: a reviewof their analysis, environmental fate and behaviour. Environ. Int. 2011,37 (2), 532−56.(2) Meng, X. Z.; Venkatesan, A. K.; Ni, Y. L.; Steele, J. C.; Wu, L. L.;Bignert, A.; Bergman, A.; Halden, R. U. Organic contaminants inChinese sewage sludge: A meta-analysis of the literature of the past 30years. Environ. Sci. Technol. 2016, 50 (11), 5454−66.(3) Stockholm Convention; The new POPs under the StockholmConvention; Website: http://chm.pops.int/Convention/ThePOPs/TheNewPOPs/tabid/2511/Default.aspx (accessed on March, 2017).(4) Arkoosh, M. R.; Van Gaest, A. L.; Strickland, S. A.; Hutchinson,G. P.; Krupkin, A. B.; Dietrich, J. P. Dietary exposure to individualpolybrominated diphenyl ether congeners BDE-47 and BDE-99 altersinnate immunity and disease susceptibility in juvenile chinook salmon.Environ. Sci. Technol. 2015, 49 (11), 6974−81.(5) Su, G.; Letcher, R. J.; Moore, J. N.; Williams, L. L.; Martin, P. A.;de Solla, S. R.; Bowerman, W. W. Spatial and temporal comparisons oflegacy and emerging flame retardants in herring gull eggs fromcolonies spanning the Laurentian Great Lakes of Canada and UnitedStates. Environ. Res. 2015, 142, 720−30.(6) Su, G.; Letcher, R. J.; Moore, J. N.; Williams, L. L.; Grasman, K.A. Contaminants of emerging concern in Caspian tern compared toherring gull eggs from Michigan colonies in the Great Lakes of NorthAmerica. Environ. Pollut. 2017, 222, 154−164.(7) Covaci, A.; Harrad, S.; Abdallah, M. A.; Ali, N.; Law, R. J.;Herzke, D.; de Wit, C. A. Novel brominated flame retardants: a reviewof their analysis, environmental fate and behaviour. Environ. Int. 2011,37 (2), 532−56.

(8) Iqbal, M.; Syed, J. H.; Katsoyiannis, A.; Malik, R. N.; Farooqi, A.;Butt, A.; Li, J.; Zhang, G.; Cincinelli, A.; Jones, K. C. Legacy andemerging flame retardants (FRs) in the freshwater ecosystem: Areview. Environ. Res. 2017, 152, 26−42.(9) Gauthier, L. T.; Hebert, C. E.; Weseloh, D. V.; Letcher, R. J.Dramatic changes in the temporal trends of polybrominated diphenylethers (PBDEs) in herring gull eggs from the Laurentian Great Lakes:1982−2006. Environ. Sci. Technol. 2008, 42 (5), 1524−30.(10) Venier, M.; Salamova, A.; Hites, R. A. Halogenated flameretardants in the Great Lakes environment. Acc. Chem. Res. 2015, 48(7), 1853−61.(11) Zhu, L. Y.; Hites, R. A. Temporal trends and spatial distributionsof brominated flame retardants in archived fishes from the GreatLakes. Environ. Sci. Technol. 2004, 38 (10), 2779−84.(12) Liu, L.; Venier, M.; Salamova, A.; Hites, R. A. A novel flameretardant in the Great Lakes atmosphere: 3,3′,5,5′-tetrabromobi-sphenol A bis(2,3-dibromopropyl) ether. Environ. Sci. Technol. Lett.2016, 3 (5), 194−199.(13) Salamova, A.; Ma, Y.; Venier, M.; Hites, R. A. High levels oforganophosphate flame retardants in the Great Lakes atmosphere.Environ. Sci. Technol. Lett. 2014, 1 (1), 8−14.(14) Arnot, J. A.; Gobas, F. A. P. C. A review of bioconcentrationfactor (BCF) and bioaccumulation factor (BAF) assessments fororganic chemicals in aquatic organisms. Environ. Rev. 2006, 14 (4),257−297.(15) Perez-Fuentetaja, A.; Mackintosh, S. A.; Zimmerman, L. R.;Clapsadl, M. D.; Alaee, M.; Aga, D. S. Trophic transfer of flameretardants (PBDEs) in the food web of Lake Erie. Can. J. Fish. Aquat.Sci. 2015, 72 (12), 1886−1896.(16) Peng, H.; Wan, Y.; Zhang, K.; Sun, J. X.; Hu, J. Y. Trophictransfer of dechloranes in the marine food web of Liaodong Bay,North China. Environ. Sci. Technol. 2014, 48 (10), 5458−5466.(17) Law, K.; Halldorson, T.; Danell, R.; Stern, G.; Gewurtz, S.;Alaee, M.; Marvin, C.; Whittle, M.; Tomy, G. Bioaccumulation andtrophic transfer of some brominated flame retardants in a LakeWinnipeg (Canada) food web. Environ. Toxicol. Chem. 2006, 25 (8),2177−86.(18) McGoldrick, D. J.; Clark, M. G.; Keir, M. J.; Backus, S. M.;Malecki, M. M. Canada’s national aquatic biological specimen bankand database. J. Great Lakes Res. 2010, 36 (2), 393−398.(19) Greaves, A. K.; Letcher, R. J.; Chen, D.; McGoldrick, D. J.;Gauthier, L. T.; Backus, S. M. Retrospective analysis of organo-phosphate flame retardants in herring gull eggs and relation to theaquatic food web in the Laurential Great Lakes of North America.Environ. Res. 2016, 150, 255−263.(20) Chen, D.; Martin, P.; Burgess, N. M.; Champoux, L.; Elliott, J.E.; Forsyth, D. J.; Idrissi, A.; Letcher, R. J. European starlings (Sturnusvulgaris) suggest that landfills are an important source ofbioaccumulative flame retardants to Canadian terrestrial ecosystems.Environ. Sci. Technol. 2013, 47 (21), 12238−12247.(21) Chen, D.; Letcher, R. J.; Martin, P. Flame retardants in eggs ofAmerican kestrels and European starlings from southern Lake Ontarioregion (North America). J. Environ. Monit. 2012, 14 (11), 2870−6.(22) Saito, K.; Sjodin, A.; Sandau, C. D.; Davis, M. D.; Nakazawa, H.;Matsuki, Y.; Patterson, D. G., Jr. Development of a accelerated solventextraction and gel permeation chromatography analytical method formeasuring persistent organohalogen compounds in adipose and organtissue analysis. Chemosphere 2004, 57 (5), 373−81.(23) Hebert, C. E.; Hobson, K. A.; Shutt, J. L. Changes in food webstructure affect rate of PCB decline in herring gull (Larus argentatus)eggs. Environ. Sci. Technol. 2000, 34 (9), 1609−1614.(24) McGoldrick, D. J.; Letcher, R. J.; Barresi, E.; Keir, M. J.; Small,J.; Clark, M. G.; Sverko, E.; Backus, S. M. Organophosphate flameretardants and organosiloxanes in predatory freshwater fish fromlocations across Canada. Environ. Pollut. 2014, 193, 254−61.(25) Gandhi, N.; Gewurtz, S. B.; Drouillard, K. G.; Kolic, T.;MacPherson, K.; Reiner, E. J.; Bhavsar, S. P. Polybrominated diphenylethers (PBDEs) in Great Lakes fish: Levels, patterns, trends and

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b02338Environ. Sci. Technol. 2017, 51, 8432−8441

8440

implications for human exposure. Sci. Total Environ. 2017, 576, 907−916.(26) Song, W.; Li, A.; Ford, J. C.; Sturchio, N. C.; Rockne, K. J.;Buckley, D. R.; Mills, W. J. Polybrominated diphenyl ethers in thesediments of the Great Lakes. 2. Lakes Michigan and Huron. Environ.Sci. Technol. 2005, 39 (10), 3474−9.(27) Stapleton, H. M.; Letcher, R. J.; Li, J.; Baker, J. E. Dietaryaccumulation and metabolism of polybrominated diphenyl ethers byjuvenile carp (Cyprinus carpio). Environ. Toxicol. Chem. 2004, 23 (8),1939−46.(28) de Wit, C. A. An overview of brominated flame retardants in theenvironment. Chemosphere 2002, 46 (5), 583−624.(29) Guo, J.; Venier, M.; Salamova, A.; Hites, R. A. Bioaccumulationof dechloranes, organophosphate esters, and other flame retardants inGreat Lakes fish. Sci. Total Environ. 2017, 583, 1−9.(30) Reich, M. R. Environmental politics and science: the case ofPBB contamination in Michigan. Am. J. Public Health 1983, 73 (3),302−13.(31) Hoh, E.; Zhu, L.; Hites, R. A. Dechlorane plus, a chlorinatedflame retardant, in the Great Lakes. Environ. Sci. Technol. 2006, 40 (4),1184−9.(32) Gauthier, L. T.; Letcher, R. J. Isomers of Dechlorane Plus flameretardant in the eggs of herring gulls (Larus argentatus) from theLaurentian Great Lakes of North America: temporal changes andspatial distribution. Chemosphere 2009, 75 (1), 115−20.(33) Gentes, M. L.; Letcher, R. J.; Caron-Beaudoin, E.; Verreault, J.Novel flame retardants in urban-feeding ring-billed gulls from the St.Lawrence River, Canada. Environ. Sci. Technol. 2012, 46 (17), 9735−44.(34) Hoh, E.; Zhu, L.; Hites, R. A. Novel flame retardants, 1,2-bis(2,4,6-tribromophenoxy)ethane and 2,3,4,5,6-pentabromoethylben-zene, in United States’ environmental samples. Environ. Sci. Technol.2005, 39 (8), 2472−7.(35) Ma, Y.; Venier, M.; Hites, R. A. 2-Ethylhexyl tetrabromoben-zoate and bis(2-ethylhexyl) tetrabromophthalate flame retardants inthe Great Lakes atmosphere. Environ. Sci. Technol. 2012, 46 (1), 204−8.(36) Venier, M.; Ma, Y.; Hites, R. A. Bromobenzene flame retardantsin the Great Lakes atmosphere. Environ. Sci. Technol. 2012, 46 (16),8653−60.(37) Gauthier, L. T.; Potter, D.; Hebert, C. E.; Letcher, R. J.Temporal trends and spatial distribution of non-polybrominateddiphenyl ether flame retardants in the eggs of colonial populations ofGreat Lakes herring gulls. Environ. Sci. Technol. 2009, 43 (2), 312−7.(38) Chen, D.; Letcher, R. J.; Gauthier, L. T.; Chu, S.; McCrindle, R.Newly discovered methoxylated polybrominated diphenoxybenzeneshave been contaminants in the Great Lakes herring gull eggs for thirtyyears. Environ. Sci. Technol. 2012, 46 (17), 9456−63.(39) Chen, D.; Letcher, R. J.; Gauthier, L. T.; Chu, S.; McCrindle, R.Newly discovered methoxylated polybrominated diphenoxybenzeneshave been contaminants in the Great Lakes herring gull eggs for thirtyyears. Environ. Sci. Technol. 2012, 46 (17), 9456−63.(40) Chen, D.; Letcher, R. J.; Gauthier, L. T.; Chu, S.Tetradecabromodiphenoxybenzene flame retardant undergoes photo-lytic debromination. Environ. Sci. Technol. 2013, 47 (3), 1373−80.(41) Su, G.; Letcher, R. J.; Crump, D.; Farmahin, R.; Giesy, J. P.;Kennedy, S. W. Photolytic degradation products of two highlybrominated flame retardants cause cytotoxicity and mRNA expressionalterations in chicken embryonic hepatocytes. Environ. Sci. Technol.2014, 48 (20), 12039−46.(42) Su, G.; Letcher, R. J.; Crump, D.; Farmahin, R.; Giesy, J. P.;Kennedy, S. W. Sunlight irradiation of highly brominated polyphenylethers generates polybenzofuran products that alter dioxin-responsivemRNA expression in chicken hepatocytes. Environ. Sci. Technol. 2016,50 (5), 2318−2327.(43) Su, G.; Greaves, A. K.; Teclechiel, D.; Letcher, R. J. In vitrometabolism of photolytic breakdown products of tetradecabromo-1,4-diphenoxybenzene flame retardant in herring gull and rat livermicrosomal assays. Environ. Sci. Technol. 2016, 50 (15), 8335−43.

(44) Gewurtz, S. B.; Bhavsar, S. P.; Fletcher, R. Influence of fish sizeand sex on mercury/PCB concentration: importance for fishconsumption advisories. Environ. Int. 2011, 37 (2), 425−34.(45) Agusa, T.; Kunito, T.; Iwata, H.; Monirith, I.; Tana, T. S.;Subramanian, A.; Tanabe, S. Mercury contamination in human hairand fish from Cambodia: levels, specific accumulation and riskassessment. Environ. Pollut. 2005, 134 (1), 79−86.(46) Vives, I.; Grimalt, J. O.; Lacorte, S.; Guillamon, M.; Barcelo, D.Polybromodiphenyl ether flame retardants in fish from lakes inEuropean high mountains and Greenland. Environ. Sci. Technol. 2004,38 (8), 2338−44.(47) Kelly, B. C.; Ikonomou, M. G.; Blair, J. D.; Gobas, F. A.Bioaccumulation behaviour of polybrominated diphenyl ethers(PBDEs) in a Canadian Arctic marine food web. Sci. Total Environ.2008, 401 (1−3), 60−72.(48) Hu, G. C.; Dai, J. Y.; Xu, Z. C.; Luo, X. J.; Cao, H.; Wang, J. S.;Mai, B. X.; Xu, M. Q. Bioaccumulation behavior of polybrominateddiphenyl ethers (PBDEs) in the freshwater food chain of Baiyangdianlake, north China. Environ. Int. 2010, 36 (4), 309−15.(49) Borga, K.; Kidd, K. A.; Muir, D. C.; Berglund, O.; Conder, J. M.;Gobas, F. A.; Kucklick, J.; Malm, O.; Powell, D. E. Trophicmagnification factors: considerations of ecology, ecosystems, andstudy design. Integr Environ. Integr. Environ. Assess. Manage. 2012, 8(1), 64−84.(50) Batt, A. L.; Wathen, J. B.; Lazorchak, J. M.; Olsen, A. R.;Kincaid, T. M. Statistical survey of persistent organic pollutants: Riskestimations to humans and wildlife through consumption of fish fromU.S. rivers. Environ. Sci. Technol. 2017, 51 (5), 3021−3031.(51) De Silva, A. O.; Spencer, C.; Ho, K. C.; Al Tarhuni, M.; Go, C.;Houde, M.; de Solla, S. R.; Lavoie, R. A.; King, L. E.; Muir, D. C.; Fair,P. A.; Wells, R. S.; Bossart, G. D. Perfluoroalkylphosphinic acids innorthern pike (Esox lucius), double-crested cormorants (Phalacrocoraxauritus), and bottlenose dolphins (Tursiops truncatus) in relation toother perfluoroalkyl acids. Environ. Sci. Technol. 2016, 50 (20), 10903−10913.(52) Fakouri Baygi, S.; Crimmins, B. S.; Hopke, P. K.; Holsen, T. M.Comprehensive emerging chemical discovery: Novel polyfluorinatedcompounds in Lake Michigan trout. Environ. Sci. Technol. 2016, 50(17), 9460−8.(53) Lu, Z.; De Silva, A. O.; Peart, T. E.; Cook, C. J.; Tetreault, G. R.;Servos, M. R.; Muir, D. C. G. Distribution, partitioning andbioaccumulation of substituted diphenylamine antioxidants andbenzotriazole UV stabilizers in an urban creek in Canada. Environ.Sci. Technol. 2016, 50 (17), 9089−97.(54) Gebbink, W. A.; Bignert, A.; Berger, U. Perfluoroalkyl acids(PFAAs) and selected precursors in the Baltic Sea environment: Doprecursors play a role in food web accumulation of PFAAs? Environ.Sci. Technol. 2016, 50 (12), 6354−62.(55) Chu, S.; Letcher, R. J.; McGoldrick, D. J.; Backus, S. M. A newfluorinated surfactant contaminant in biota: Perfluorobutane sulfona-mide in several fish species. Environ. Sci. Technol. 2016, 50 (2), 669−75.(56) Eating Ontario Fish (2017−18); website: https://www.ontario.ca/page/eating-ontario-fish-2017-18#section-12 (accessed on May,2017).

Environmental Science & Technology Article

DOI: 10.1021/acs.est.7b02338Environ. Sci. Technol. 2017, 51, 8432−8441

8441