Black Carbon Inclusive Multichemical Modeling of PBDE and PCB

Sep 9, 2010 - Engineering and Applied Chemistry, University of Toronto,. Toronto, ON, Canada, M5S 3E5, Centre for Environment,. University of Toronto ...
0 downloads 0 Views 376KB Size
Download PDF
Environ. Sci. Technol. 2010, 44, 7548–7554

Black Carbon Inclusive Multichemical Modeling of PBDE and PCB Biomagnification and -Transformation in Estuarine Food Webs CAROLINA DI PAOLO,† NILIMA GANDHI,‡ S A T Y E N D R A P . B H A V S A R , §,| M A R T I N E V A N D E N H E U V E L - G R E V E , ⊥,# A N D A L B E R T A . K O E L M A N S * ,†,# Aquatic Ecology and Water Quality Management Group, Department of Environmental Sciences, Wageningen University, P.O. Box 47, 6700AA, Wageningen, The Netherlands, Division of Environmental Engineering, Department of Chemical Engineering and Applied Chemistry, University of Toronto, Toronto, ON, Canada, M5S 3E5, Centre for Environment, University of Toronto, Toronto, ON, Canada, M5S 3E8, Environmental Monitoring and Reporting Branch, Ontario Ministry of the Environment, 125 Resources Road, Etobicoke, ON, Canada, M9P 3 V6, Deltares, Department of Marine and Coastal Systems, P.O. Box 177, 2600 MH Delft, The Netherlands, and IMARESsInstitute for Marine Resources & Ecosystem Studies, Wageningen UR, P.O. Box 68, 1970 AB IJmuiden, The Netherlands

Received April 19, 2010. Revised manuscript received August 16, 2010. Accepted August 31, 2010.

Bioavailability and bioaccumulation of polybrominated diphenylethers (PBDEs) are affected by adsorption on black carbon (BC) and metabolism in biota, respectively. Recent studies have addressed these two processes separately, illustrating their importance in assessing contaminant dynamics. In order to properly examine biomagnification of polychlorinated biphenyls (PCBs) and PBDEs in an estuarine food-web, here we set up a black carbon inclusive multichemical model. A dual domain sorption model, which accounted for sorption to organic matter (OM) and black carbon (BC), was used to estimate aqueous phase concentrations from the measured chemical concentrations in suspended solids. We adapted a previously published multichemical model that tracks the movement of a parent compound and its metabolites in each organism and within its food web. First, the model was calibrated for seven PCB congeners assuming negligible metabolism. Subsequently, PBDE biomagnification was modeled, including biotransformation and bioformation of PBDE congeners, keeping the other model parameters the same. The integrated model was capable of predicting trophic magnification factors (TMF) within error limits. PBDE metabolic half-lives ranged 21-415 days and agreed to literature * Corresponding author: [email protected], phone: +31 317 483201. † Wageningen University. ‡ Department of Chemical Engineering and Applied Chemistry, University of Toronto. § Centre for Environment, University of Toronto. | Ontario Ministry of the Environment. ⊥ Deltares. # IMARESsInstitute for Marine Resources & Ecosystem Studies. 7548

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 19, 2010

data. The results showed importance of including BC as an adsorbing phase, and biotransformation and bioformation of PBDEs for a proper assessment of their dynamics in aquatic systems.

Introduction Polybrominated diphenylethers (PBDEs) have been widely used in industrial products such as textiles, plastics, building materials, and electronics, and were the first group of brominated flame retardants detected in fish in the early 1980s (1). A few years later, they were identified as global contaminants when they were detected in top-predators from remote areas (2). Recently, tetra-, penta-, hexa-, and heptaBDE were listed as persistent pollutants under the Stockholm convention (3). Because of risks to the environment and to human health, the use of penta- and octa-BDE has already been banned in Europe (4). Deca-BDE (BDE 209), however, was not restricted in electrical and electronic products (5) because of poor characterization of its persistence, bioaccumulation and toxicity in aquatic and terrestrial ecosystems (6). Similarly to polychlorinated biphenyls (PCBs), PBDE congeners are highly persistent and bioaccumulative, with octanol-water partition coefficients (logKOW) between 5.5 and 9 (7, 8). In addition, major uncertainties exist with respect to environmental and metabolic transformation of PBDEs by organisms. Higher-brominated PBDE congeners such as BDE 209 are subjected to both biotic and abiotic debromination that result in the formation of lower-brominated congeners, including more toxic banned compounds (9-12). Moreover, biotransformation and bioformation affect the bioaccumulation and biomagnification in fish as demonstrated by in vivo and in vitro studies (13, 14). Earlier PBDE food web models did not address biotransformation, reasoning that the metabolic transformation and its quantification were not clearly supported by experimental data (15, 16). However, debromination and bioformation of PBDEs have been observed in common carp (Cyprinus carpio) (17, 18), rainbow trout (Oncorhynchus mykiss) (19), and lake trout (Salvelinus namaycush) (8). In order to investigate rates of PBDE biotransformation in aquatic food-webs, Gandhi et al. (20) presented a multichemical food-web model that is capable of accounting both biotransformation and bioformation. The model was used to assess dynamics of PBDEs (BDE 47, BDE 99, BDE 100, BDE 153) in a simple food-web of Arctic char (Salvelinus alpinus). In a subsequent study, the model was enhanced for dynamic application and used to further refine biotransformation rates of 13 PBDEs in juvenile lake trout (Salvelinus namaycush) by successfully reproducing congener-specific experimentally measured fish accumulation for low and high doses during the uptake and depuration phases (21). Modeling solid-water partitioning remains an important issue in prospective risk assessment because of relatively expensive and complicated direct measurements of freely dissolved aqueous concentrations and because of the importance of uptake through ingestion of solids for many hydrophobic chemicals. In this context, recent inclusion of the strong binding of chemicals to black carbon (BC) significantly improved estimates of aqueous phase concentrations from measured concentrations in suspended solids (22). The added value of BC-inclusive food chain accumulation modeling has been shown for PCBs and metabolisable compounds such as polycyclic aromatic hydrocarbons (PAHs) (23, 24). 10.1021/es101247e

 2010 American Chemical Society

Published on Web 09/09/2010

The present study couples the previous multichemical model (20) with a BC-inclusive chemical partitioning submodel, and estimates bioaccumulation and biomagnification of PCBs and PBDEs along the food-webs of prey of a colony of common terns (Sterna hirundo) from the Harbour of Terneuzen in the Western Scheldt estuary (The Netherlands). The coupled model considers (a) chemical-specific biotransformation and bioformation rates for each organism, and (b) measured concentrations in particulate matter along with BC levels. The major goals of the study were to estimate characteristic time scales for biotransformation and bioformation and to test the model in its capability to predict trophic magnification factors (TMFs). Modeling was performed for two different locations in the estuary in order to investigate effects of local variations on biomagnification.

Materials and Methods Data Set. Samples were collected from two locations near the harbor of Terneuzen: location HT (inside the Harbour of Terneuzen) and location MP (Middelplaat, a sand flat in the main estuarine channel) (25). Samples represent the diet of common terns and associated food-webs (Table S1, Figure S1, Supporting Information (SI)), including shrimp (Crangon crangon), anchovy (Engraulis encrasicolus), herring (Clupea harengus), sandeel (Ammodytes sp.), whiting (Merlangius merlangus), bib (Trisopterus luscus), Goby sp., and seabass (Dicentrarchus labrax). Data set included measured PCB and PBDE concentrations on solids, fractions of organic carbon and BC, stable isotopes data and other physical-chemical data for the two locations (SI Tables S2 and S3). Concentrations in lugworm (Arenicola marina) and mysids (Mysis sp.) were calculated from previously measured biota-sedimentaccumulation-factors (BSAF) from the same study area (26-28). Concentrations in zooplankton were calculated from previously measured biomagnification-factors (BMF) for C. harengus-zooplankton (29). BC percentages in sediment and suspended particulate matter were 0.24% and 1.05% for location MP and 0.73% and 1.38% for location HT, respectively (25). Measured concentrations of seven PCBs (PCB 28, PCB 52, PCB 101, PC 118, PCB 138, PCB 153, PCB 180) and seven PBDEs (BDE 28, BDE 49, BDE 47, BDE 100, BDE 99, BDE 126, BDE 209) in solids (sediment, suspended particulate matter) were corrected for fractions of organic and black carbon (see eq 4, below). Concentrations in biota were normalized to lipid content (biota, ng/g/fLIP) (25). Trophic Levels. Trophic levels (TL) were calculated from stable isotope data (δ15N) in sampled organisms (δ15NORG) and in cockle (Cerastoderma edule, δ15NCOCKLE) (25) using eq 1: TL ) 2 + [(δ15ΝORG - δ15ΝCOCKLE)/3.4]

(1)

In the absence of measured isotope data, TL was obtained from the literature (SI Table S1). Trophic Magnification Factor. Measured TMF was calculated as TMF ) eb, where b is the slope of the linear relationship: ln C ) a + (b × TL)

(2)

where C is the normalized measured or modeled chemical concentration in an organism and TL is the measured (not estimated) trophic level for that organism (30-32). Metabolic Index (MI). MI is an indicator of metabolic transformation and was used to guide estimates of relative metabolic rates in the model. MI gives PBDE biomagnification relative to a tracer PCB, according to the following equation (20, 33):

n

MI ) predator(BDEX /PCB153)/

∑ R (BDE /PCB i

X

153)

i

(3) where BDEX and PCB153 are the concentrations (ng/g/fLIP) of each x PBDE congener and of PCB 153 respectively; n is the number of prey i, and Ri is the diet fraction of each prey i in the total diet of predator (SI Table S1). MI value greater than one indicates that biomagnification for the PBDE congener in question is more than that for PCB 153; whereas MI < 1 indicates biotransformation (20, 33). Modeling Aqueous Phase Concentrations. Aqueous phase chemical concentrations were estimated from corresponding measured concentrations in solids (sediment or suspended particulate matter) and black carbon content according to the following (22, 23, 34): nf,BC-1 CS /C0,W ) fOCKOC + fBCKF,BCC0,W

(4)

where CS (µg/kg) and C0,W (µg/L) are chemical concentrations in solids and water, respectively; fOC (kg OC/kg dry-weight) and fBC (kg BC/kg dry-weight) are fractions of organic carbon (OC) and black carbon (BC) in solids, respectively; KOC (L/kg OC) is the partitioning coefficient to organic carbon calculated as logKOC ) 0.41logKOW (for logKOW values, SI Table S4); KF,BC [(µg/kg BC)/(µg/L)n] is the Freundlich constant for sorption to BC, estimated from logKF,BC ) 0.928logKOW+0.080 (for PCBs) and logKF,BC ) 1.088logKOW - 1.365 (for PBDEs). These regressions were calculated from BC, total sediment, and passive sampler based aqueous phase PCB and PBDE data for harbor sediments from The Netherlands, as detailed in SI Figure S2 (35). nf,BC is the Freundlich coefficient for sorption to BC, estimated at 0.7 (22-24, 35). Because of estuarine mixing, spatial variability of fBC in suspended solids was assumed to be limited. Nevertheless, we evaluated the uncertainty in adsorption to BC in the default scenario (“Default-BC”) by testing additional “HighBC” and “Low-BC” scenarios. In these scenarios fBC was varied (50%, nf,BC was varied between 0.6 and 0.8, and logKF,BC was varied within its 90% confidence interval. Model Description. The multichemical food-web model is described in detail by Gandhi et al. (20). Briefly, it is a fugacity-based model that considers fugacity f (Pa) as linearly related to concentration C (mol/m3) through fugacity capacity Z (mol/Pa m3), by C ) fZ. Values of Z were calculated from partition coefficients (e.g., KOW) for different compartments such as water, solids, octanol or lipid, nonlipid organic matter, organisms, diet, and gut. Transport processes considered for organisms were chemical exchange through water respiration; net chemical uptake from diet and bioformation gain; chemical loss due to egestion, growth dilution, and metabolic transformation including debromination. A Dvalue (mol/Pa h) is calculated for each transport process by multiplying Z and G-values (m3/h). Each taxonomic group was assigned a chemical mass balance equation: VZdfA /dt ) fBDBA - fADAOut

(5)

where V (m3) is the volume of the environmental medium or taxonomic group; A and B are the two media of chemical exchange; BA indicates the input from B to A; and AOut is the sum of all D values responsible for chemical loss from compartment A. For this application, the model is run in steady-state, and calculates concentrations within each organism and transport rates within and across organisms. At steady-state, eq 5 for each taxonomic group is simplified as shown: fI ) WI(xWfW + xSfS) +

∑A f

(6)

JI J

VOL. 44, NO. 19, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7549

FIGURE 1. Biotransformation path of the major PBDE congeners considered in the model application, constructed as an improvement of a previously published pathway (21) by including four new congeners (BDE 156, BDE 126, BDE 49, BDE 17) and updating debromination fractions. Assuming that all the debromination products of higher PBDEs are lower congeners, the numbers adjacent to the arrows indicate the fractions of debromination of a higher congener that will form lower ones, from a total of 1.0. where I is the organism; J is its prey; WI and AJI are fugacity factors for respiration of water and uptake from diet, respectively; W and S are for water and sediment, respectively; x is a respiration fraction; and ∑ is for sum across all taxonomic groups. WI and AJI are calculated as DWI/ DToti and DAJI/DToti respectively, where DToti is the sum of the D values for elimination through respiration, egestion, growth dilution. and metabolic transformation. The organism-specific total chemical half-life is calculated as 0.693DToti/VZ, which includes the metabolic half-life (HL) of that chemical. p numbers of eq 6 were formulated, one for each of p taxonomic groups. To apply the model for q chemicals, p × q equations were formulated, one for each q chemical in each p taxonomic group. The set of equations were solved for fugacities using a general matrix form Af ) E, in which A is the diet matrix p × p; f is the vector of fugacities for each taxonomic group and chemical; and E is a respiration vector (36). Model Calibration and Evaluation. Model calibration was targeted to minimize differences between modeled and measured concentrations of PCBs and PBDEs across all taxonomic groups, which were evaluated by residual error and tendency for under or overestimation, as 7550

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 19, 2010

described in detail in the SI. The BC-inclusive multichemical model was first calibrated for PCBs using diet preferences of organisms (SI Table S1) and fixed physicalchemical properties (SI Table S4), by optimizing the organisms’ growth and food ingestion rates using literature values as a guide (SI Table S5). Notably the metabolic transformation of PCBs was assumed negligible in this calibration. The PCB-calibrated model was then applied to PBDEs, now including biotransformation loss and bioformation gain, but using organism related parameters from the PCB calibration. A PBDE biotransformation and bioformation pathway (Figure 1) was modified from a previously published pathway (21) by including four new congeners (BDE 156, BDE 126, BDE 49, BDE 17) following the previous assumptions (20). The model was further calibrated by adjusting PBDE congener-specific HL values and debromination fractions (Figure 1) using the previously modeled HL values (20, 21) and the calculated MI values (SI Table S6) as a guide. Model performance was evaluated by running several test scenarios. Test scenarios were evaluated by comparing the residual error (TotalRE, %), calculated as shown:

totalRE )



m

∑ i)1

(

n

∑ [(log C

modx

- log CMEASx)/log CMEASx]2

x)1

n m

)

i

where logCMODx and logCMEASx are log transformed modeled and measured concentrations, respectively, of each chemical x in each organism i; n is the number of compounds; and m is the number of organisms (SI Table S7A). Next, we modeled and evaluated the following three PBDE biotransformation scenarios using the calibrated model: (a) with biotransformation-loss and bioformation-gain; (b) with biotransformation-loss but no bioformation-gain; and (c) neither biotransformation nor bioformation included. The first scenario (a) was further modeled by estimating aqueous chemical concentrations from their concentrations in sediment (CSED f CW) versus the suspended particulate matter (CPM f CW) that were compared with and without taking the chemical adsorption to BC into account (with BC or BC ) 0), resulting in four additional test-scenarios: (a1) CSED f CW/BC ) 0; (a2) CPM f CW/BC ) 0; (a3) CSED f CW/WithBC; and (a4) CPM f CW/WithBC. As discussed in the following section, the scenario a4 that used the CPM based BC-inclusive multichemical model configuration provided the best results (SI Table S7B) and hence was used in subsequent calibration and modeling of TMF.

Results and Discussion Model Performance. Modeled PBDE concentrations that considered biotransformation-loss and bioformation-gain showed the best fit to measured concentrations (SI Table S7A). As such, this scenario along with other model concepts and design was adopted in all subsequent model runs. This agrees to the earlier model applications (20, 21) and implies that accounting for biotransformation and bioformation as specified in Figure 1 is important in assessing dynamics of this group of chemicals. Inclusion of adsorption to BC also resulted in a better agreement to observed concentrations. The improvement in modeled estimates was up to factors of 2 and 4 for PCBs and PBDEs, respectively (WithBC test scenarios, SI Table S7A). Quality of fit also improved by estimating CW from CPM instead of from CSED, except for PBDEs in location HT. These results are consistent with earlier model studies where BC appeared to improve model performance for bioaccumulation of PCBs and PAHs (23, 34). Inclusion of BC does not really increase the number of fitting parameters in the model, because the BC term in eq 4 dominates over the OC term so that the BC parameters in fact replace the previous OC parameters. In the Default-BC scenario, finetuning of the chemical-specific gut absorption efficiencies and biotransformation rates (PBDEs) resulted in a further decrease of residual error by factors of 2-3. In both the lowand high-BC scenarios, the error values decreased by ∼1.5 times. In addition, in the default-BC scenario most of the compounds and organisms also presented a good fit, indicated by low residual error (0.1-6%), and agreement of modeled concentrations with measured values within 1 order of magnitude (SI Figure S3). Agreement was within a factor of 5 in 82% (MP/PCB), 89% (MP/PBDE), 93% (HT/PCB) and 65% (HT/PBDE) of the cases (SI Table S8A/8B). Levels of PBDEs in D. labrax from location HT (RE ) 20%) were considered outliers. Modeling Trophic Magnification Factors (TMF). Measured and modeled TMF ranged 0.85-3.03 (TMFMEAS) and 0.86-2.46 (TMFMOD) for PCBs; and 0.15-2.31 (TMFMEAS) and 0.94-1.67 (TMFMOD) for PBDEs, respectively (Figure 2). Confidence intervals (95%CI, SI Table S9) showed that only the measured TMF for PCB 28 differed from the other PCBs.

Measured TMFs for PBDEs were not significantly different from each other. Taking errors in the regression slopes into account, it can be concluded that biomagnification was apparent for all PCBs (TMF > 1) except PCB 28 in location HT; for some PBDEs at location MP, and for BDE 47 at location HT. All values of measured TMFs for PBDEs were lower than those for PCBs, which could be due to the limited trophic length of the system and the complex feeding relationships, including omnivory. Additionally, metabolism of PBDEs also contributed to their lower TMFs compared to those for PCBs. Considering the scatter in the data, modeled TMFs from the Default-BC scenario agreed well to measured values for PBDEs at location MP (Figure 2, SI Table S9) and for PCBs in location HT. TMFs for PCBs at location MP were slightly underestimated (TMFMEAS ) 1.31-3.03; TMFMOD ) 0.86-2.45), which can be explained by the overestimation of PCB concentrations in organisms at TL 1.3-2.3; and underestimation in TL > 3.3 (ΣPCBsMEAS:ΣPCBsMOD ) 1.31-4.04, except from E.encrausilocus). PBDEs in location HT on the contrary were overestimated in Default-BC due to underestimation in TL ) 1.2-2.3 and overestimation in TL > 3.3; and agreed well with the high-BC scenario. Obviously, modeled TMFs improved by a factor of 2 when regressed using only the values of CMOD that presented good fits to measured values (e.g., CMEAS:CMOD ) 0.4-2.5). Measured and modeled TMFs showed a curvilinear trend when plotted against logKOW (Figure 2), with their maximum values at logKOW ) 6.8 (PCB 153, BDE 47). Similar TMFlogKOW relationships have previously been reported for PCBs (37, 38), and curvilinear relationships between log-half-life or log-assimilation-efficiency versus logKOW in fish (39). The model appropriately accounted for this effect in modeled gut absorption efficiencies that followed parabolic relationships with molecular size and logKOW (21) (Table 1). Bioaccumulation and Biotransformation Profiles. Chemical profiles (ng/g/fLIP) were correctly modeled for both PCBs and PBDEs (SI Table S10, PCB 153 > PCB 138 > PCB 101/PCB 180 > PCB 118/PCB 52 > PCB 28 and BDE 209 > BDE 47 > BDE 49/BDE 99/BDE 100 > BDE 28 > BDE 126). Measured BDE 209 had the highest concentrations among PBDEs, which is in contrast to previous studies that identified BDE 47 to be prevalent in invertebrates and fish from the Western Scheldt (16, 26, 27). We believe this is because the chemical analyses in the current work were performed for whole organism homogenates that included the gut content (25). The high concentrations of BDE 209 may be related to solids and algae inside the digestive systems, suggesting that particle ingestion is an important route of exposure for this compound. In contrast, the trend in measured PBDE profiles (SI Table S10) noticeably changed in the high trophic-level fish Goby (location MP, BDE 47 > BDE 209 > BDE 100 > BDE 49 > BDE 99) and D. labrax (location HT, BDE 47 > BDE 209 > BDE 100 > BDE 49 > BDE 126 > BDE 28 > BDE 99), suggesting that a significant biotransformation of BDE 209 in the gut can form an additional exposure route for lower-brominated PBDE congeners such as BDE 47. Biomagnification of BDE 47, and potential biotransformation of BDE 99 and BDE 209 were suggested by several indicators. First, MI values for Goby and D. labrax are >1 for BDE 47 and ,1 for BDE 99 and BDE 209 (SI Table S9). Second, BDE 47 levels were underestimated and BDE 209/BDE 99 levels were overestimated (CMEAS/CMOD BDE 47 Goby and D. labrax .1, CMEAS/CMOD BDE 209 ,1). Third, the profiles of prevalent compounds in both fish agreed to this pattern and therefore these fish species are suggested to metabolize BDE 99 and BDE 209, and to form BDE 47. Model-calibrated biotransformation half-lives (HLs; Table 1) in fish (TL 3-4) were 1.5-2 times lower (BDE 28, BDE 100, BDE 99), 1.5 times higher (BDE 47), or the same (BDE 209) as reported for juvenile rainbow trout following a dietary VOL. 44, NO. 19, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7551

FIGURE 2. Plots of measured (-b-) and modeled (default-BC: ---O---; high-BC: --+--; low-BC: -x-) values of trophic magnification factors (TMF) of PCBs and PBDEs (BDE 209 not included) from the location MP and location HT, plotted versus respective values of logKOW. Error bars indicate values of 95% CI for TMFMEAS.

TABLE 1. Model Calibrated Metabolic Half-Lives (HL) in Hours (h) and Gut Absorption Efficiencies (GAEs) for Each PBDE in Trophic Levels 3 to 4 (TL 3-4) from the Current Model Application model calibrated metabolic half-lives (HL) current model application compounda

TL 3-4 (h)

BDE 28 BDE 49 BDE 47 BDE 100 BDE 99 BDE 126 BDE 209

500 10000 1500 3000 3000 3000 1000

model calibrated gut absorption efficiencies (GAEs)

previous model applications Arctic char (h)

50000 10000 7000

b

current model application c

juvenile lake trout (h)

TL 3-4 (%)

1000

28 49 47 35 35 35 30

1000 5000 5000 1000

previous model applications Arctic charb (%)

juvenile lake troutc (%) 45

90 70 70

45 45 45

10

35

a

For TL 1-4, HL values were considered as negligible, with the value of 100 years considered as negligible biotransformation. b From ref 20. c From ref 21.

exposure (21); and were also lower than for arctic char in a simplified food-web (BDE 47: 33× lower; BDE 100: 3× lower; BDE 99: 2× lower) (20). Lower values of HLs imply higher biotransformation-loss in the presently studied organisms, while BDE 47 in contrast was found to suffer less metabolic transformation or higher bioformation-gain than in juvenile rainbow trout. The large difference for BDE 47 compared to arctic char is probably related to lower metabolic rates of fish in cold arctic environments (40). Differences between species and local conditions may yield differences in metabolic capacities. To our knowledge, no consistent set of PBDE HL estimates exists for the present studied organisms. Therefore, the current model-calibrated HL values and gut 7552

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 19, 2010

absorption efficiencies (Table 1) may guide future modeling and investigation of PBDE biotransformation. Differences between Locations MP and HT. Default modeling results generally showed good agreements with measured values for chemical concentrations in biota (SI Figure S3) and TMF (Figure 2), except for PBDEs at location MP, which gave better results for the high-BC scenario. Further, TMF values for both PCBs and PBDEs were lower at location HT than at location MP, suggesting a reduced bioavailability of chemicals due to sorption to solids. This agrees with the fact that location HT is within the Harbour of Terneuzen, existing as an almost closed system within the estuary, with longer residence times and higher weathering

of solids compared to location MP. On the other hand, location MP is an area in the main estuarine channel that is subjected to continuous upstream input of solids and resuspension due to tidal effect, thus creating highly dynamic conditions in which chemicals may not reach equilibrium between phases. Therefore, sorption to OM and BC might have been overestimated by eq 4, resulting in underestimation of bioavailable chemical concentrations and consequently the model results. Implications. This study reiterates that modeling of PBDE dynamics in aquatic organisms must consider the biotransformation and bioformation of compounds. Our study, however, recognizes the need for additional experimental studies on PBDE metabolic transformation in fish, the understanding of which is important for the improvement of model input parameters and for the derivation of their biotransformation pathways. Given that TMF has been advocated as the “gold standard” metric for understanding and indicating the biomagnification potential of chemicals (32, 41), our study showed the importance of considering the sorption of chemicals to BC in suspended solids when estimating bioavailable concentrations in estuaries. After all, considering sorption to BC improved modeled concentrations in biota and TMF values for both PCBs and PBDEs. Finally, this work advocates the usefulness of models to directly simulate TMFs, and to mechanistically analyze the impacts of chemical metabolism on their TMF values.

Supporting Information Available Additional 3 figures and 10 tables. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited ¨ .; Blomkvist, G. Polybrominated aromatic pollut(1) Andersson, O ants found in fish in Sweden. Chemosphere 1981, 10, 1051– 1060. (2) Jansson, B.; Asplund, L.; Olsson, M. Brominated flame retardants - Ubiquitous environmental pollutants. Chemosphere 1987, 16, 2343–2349. (3) Report of the Conference of the Parties of the Stockholm Convention on Persistent Organic Pollutants on the Work of Its Fourth Meeting; UNEP/POPS/COP.4/38: Geneva, 4-8 May 2009. (4) EU-Parliament; EU-Council, Directive 2003/11/EC of the European Parliament and of the Council of 6 February 2003, Official Journal of the European Union: 2003; pp L42/45-L42/46. (5) EU-Commission, Commission Decision of 13 October 2005 Amending for the Purposes of Adapting to the Technical Progress the Annex to Directive 2002/95/EC; Official Journal of the European Union: Luxembourg, 2005; L271/48-L271/50. (6) EU-Commission, Communication from the Commission on the Results of the Risk Evaluation of Chlorodifluoromethane, Bis(Pentabromophenyl)Ether and Methenamine and on the Risk Reduction Strategy for the Substance Methenamine (2008/C 131/ 04); Official Journal of the European Union: Luxembourg, 2008; pp C131/7-C131/12. (7) Mikula, P.; Svobodova´, Z. Brominated flame retardants in the environment: Their sources and effects (a review). Acta Vet. Brno 2006, 75, 587–599. (8) Tomy, G. T.; Palace, V. P.; Halldorson, T.; Braekevelt, E.; Danell, R.; Wautier, K.; Evans, B.; Brinkworth, L.; Fisk, A. T. Bioaccumulation, biotransformation, and biochemical effects of brominated diphenyl ethers in juvenile lake trout (Salvelinus namaycush). Environ. Sci. Technol. 2004, 38, 1496–1504. (9) Eriksson, J.; Green, N.; Marsh, G.; Bergman, A. Photochemical decomposition of 15 polybrominated diphenyl ether congeners in methanol/water. Environ. Sci. Technol. 2004, 38, 3119–3125. (10) He, J.; Robrock, K. R.; Alvarez-Cohen, L. Microbial reductive debromination of polybrominated diphenyl ethers (PBDEs). Environ. Sci. Technol. 2006, 40, 4429–4434. (11) Tokarz, J. A.; Ahn, M.-Y.; Leng, J.; Filley, T. R.; Nies, L. Reductive debromination of polybrominated diphenyl ethers in anaerobic sediment and a biomimetic system. Environ. Sci. Technol. 2008, 42, 1157–1164. (12) Shih, Y.-h.; Wang, C.-K. Photolytic degradation of polybromodiphenyl ethers under UV-lamp and solar irradiations. J. Hazard. Mater. 2009, 165, 34–38.

(13) Stapleton, H. M.; Brazil, B.; Holbrook, R. D.; Mitchelmore, C. L.; Benedict, R.; Konstantinov, A.; Potter, D. In vivo and in vitro debromination of decabromodiphenyl ether (BDE 209) by juvenile rainbow trout and common carp. Environ. Sci. Technol. 2006, 40, 4653–4658. (14) Benedict, R. T.; Stapleton, H. M.; Letcher, R. J.; Mitchelmore, C. L. Debromination of polybrominated diphenyl ether-99 (BDE99) in carp (Cyprinus carpio) microflora and microsomes. Chemosphere 2007, 69, 987–993. (15) Van Beusekom, O. C.; Eljarrat, E.; Barcelo´, D.; Koelmans, A. A. Dynamic modeling of food-chain accumulation of brominated flame retardants in fish from the Ebro river basin, Spain. Environ. Toxicol. Chem. 2006, 25, 2553–2560. (16) Veltman, K.; Hendriks, J.; Huijbregts, M.; Leonards, P.; Van Den Heuvel-Greve, M.; Vethaak, D. Accumulation of organochlorines and brominated flame retardants in estuarine and marine food chains: Field measurements and model calculations. Mar. Pollut. Bull. 2005, 50, 1085–1102. (17) Stapleton, H. M.; Alaee, M.; Letcher, R. J.; Baker, J. E. Debromination of the flame retardant decabromodiphenyl ether by juvenile carp (Cyprinus carpio) following dietary exposure. Environ. Sci. Technol. 2004, 38, 112–119. (18) Stapleton, H. M.; Letcher, R. J.; Baker, J. E. Debromination of polybrominated diphenyl ether congeners BDE 99 and BDE 183 in the intestinal tract of the common carp (Cyprinus carpio). Environ. Sci. Technol. 2004, 38, 1054–1061. (19) Kierkegaard, A.; Balk, L.; Tjarnlund, U.; de Wit, C. A.; Jansson, B. Dietary uptake and biological effects of decabromodiphenyl ether in rainbow trout (Oncorhynchus mykiss). Environ. Sci. Technol. 1999, 33, 1612–1617. (20) Gandhi, N.; Bhavsar, S. P.; Gewurtz, S. B.; Diamond, M. L.; Evenset, A.; Christensen, G. N.; Gregor, D. Development of a multichemical food web model: Application to PBDEs in Lake Ellasjøen, Bear Island, Norway. Environ. Sci. Technol. 2006, 40, 4714–4721. (21) Bhavsar, S. P.; Gandhi, N.; Gewurtz, S. B.; Tomy, G. T. Fate of PBDEs in juvenile lake trout estimated using a dynamic multichemical fish model. Environ. Sci. Technol. 2008, 42, 3724– 3731. (22) Koelmans, A. A.; Jonker, M. T. O.; Cornelissen, G.; Bucheli, T. D.; ¨ . Black carbon: The reverse Van Noort, P. C. M.; Gustafsson, O of its dark side. Chemosphere 2006, 63, 365–377. (23) Hauck, M.; Huijbregts, M. A. J.; Koelmans, A. A.; Moermond, C. T. A.; Van Den Heuvel-Greve, M. J.; Veltman, K.; Hendriks, A. J.; Vethaak, A. D. Including sorption to black carbon in modeling bioaccumulation of polycyclic aromatic hydrocarbons: Uncertainty analysis and comparison to field data. Environ. Sci. Technol. 2007, 41, 2738–2744. (24) Moermond, C. T. A.; Traas, T. P.; Roessink, I.; Veltman, K.; Hendriks, A. J.; Koelmans, A. A. Modeling decreased food chain accumulation of PAHs due to strong sorption to carbonaceous materials and metabolic transformation. Environ. Sci. Technol. 2007, 41, 6185–6191. (25) Van den Heuvel-Greve, M. Z., A. Identification and Trophic Transfer of Contaminants in Estuarine Food Webs; State of the Art Report 2008-2010, Deltares report Z4635/1200235.002; Deltares: Delft, 2010. (26) Van den Heuvel-Greve, M.; Leonards, P.; Vethaak, D. Dioxineonderzoek Westerschelde: Meting Van Gehalten Aan Dioxinen, Dioxineachtige Stoffen En Andere Mogelijke Probleemstoffen in Visserijproducten, Sediment En Voedselketens Van De Westerschelde; RIKZ: Middelburg, The Netherlands: 2006; 80 pp. (27) Van den Heuvel-Greve, M.; Withagen, L.; Leonards, P. E. G. Vervolgonderzoek Chemische Stoffen in Visserijproducten Uit De Westerschelde; Dioxines (w.o. Furanen), PCB’s, Gebromeerde Vlamvertragers En Geperfluoreerde Verbindingen, 2006; Rijkswaterstaat. Rijksinstituut voor Kust en Zee: Middelburg, the Netherlands: 2007; 64 pp. (28) Verslycke, T. A.; Vethaak, A. D.; Arijs, K.; Janssen, C. R. Flame retardants, surfactants and organotins in sediment and mysid shrimp of the Scheldt estuary (The Netherlands). Environ. Pollut. 2005, 136, 19–31. (29) Burreau, S.; Zebu ¨hr, Y.; Broman, D.; Ishaq, R. Biomagnification of PBDEs and PCBs in food webs from the Baltic Sea and the northern Atlantic Ocean. Sci. Total Environ. 2006, 366, 659–672. (30) Fisk, A. T.; Hobson, K. A.; Norstrom, R. J. Influence of chemical and biological factors on trophic transfer of persistent organic pollutants in the northwater polynya marine food web. Environ. Sci. Technol. 2001, 35, 732–738. (31) Borgå, K.; Fisk, A. T.; Hoekstra, P. F.; Muir, D. C. Biological and chemical factors of importance in the bioaccumulation and trophic transfer of persistent organochlorine contaminants in VOL. 44, NO. 19, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7553

(32)

(33)

(34)

(35)

(36)

7554

Arctic marine food webs. Environ. Sci. Technol. 2004, 23, 2367– 2385. Weisbrod, A.; Woodburn, K.; Koelmans, A. A.; Parkerton, T. F.; McElroy, A.; Borga, K. Evaluation of Bioaccumulation Using In Vivo Laboratory and Field Studies. Integr. Environ. Assess. Manage. 2009, 5, 598–623. Wolkers, H.; van Bavel, B.; Derocher, A. E.; Wiig, O.; Kovacs, K. M.; Lydersen, C.; Lindstrom, G. Congener-specific accumulation and food chain transfer of polybrominated diphenyl ethers in two arctic food chains. Environ. Sci. Technol. 2004, 38, 1667–1674. Moermond, C. T. A.; Zwolsman, J. J. G.; Koelmans, A. A. Black Carbon and ecological factors affect in situ biota to sediment accumulation factors for hydrophobic organic compounds in flood plain lakes. Environ. Sci. Technol. 2005, 39, 3101–3109. Koelmans, A. A.; Poot, A.; De Lange, H. J.; Velzeboer, I.; Harmsen, J.; van Noort, P. C. M. Estimation of in situ sediment to water fluxes of polycyclic aromatic hydrocarbons, polychlorobiphenyls and polybrominated diphenylethers. Environ. Sci. Technol. 2010, 44, 3014–3020. Campfens, J.; Mackay, D. Fugacity-based model of PCB bioaccumulation in complex aquatic food webs. Environ. Sci. Technol. 1997, 31, 577–583.

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 19, 2010

(37) Kelly, B. C.; Ikonomou, M. G.; Blair, J. D.; Surridge, B.; Hoover, D.; Grace, R.; Gobas, F. A. P. C. Perfluoroalkyl contaminants in an arctic marine food web: trophic magnification and wildlife exposure. Environ. Sci. Technol. 2009, 43, 4037–4043. (38) Wu, J.-P.; Luo, X.-J.; Zhang, Y.; Yu, M.; Chen, S.-J.; Mai, B.-X.; Yang, Z.-Y. Biomagnification of polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls in a highly contaminated freshwater food web from South China. Environ. Pollut. 2009, 157, 904–909. (39) Fisk, A. T.; Norstrom, R. J.; Cymbalisty, C. D.; Muir, D. C. G. Dietary accumulation and depuration of hydrophobic organochlorines: bioaccumulation parameters and their relationship with the octanol/water partition coefficient. Environ. Toxicol. Chem. 1998, 17, 951–961. (40) Steffensen, J. F.; Anthony, P. F.; John, F. S., Respiratory systems and metabolic rates. In Fish Physiology; Academic Press: New York, 2005; Vol. 22, pp 203-238. (41) Gobas, F. A. P. C.; de Wolf, W.; Burkhard, L. P.; Verbruggen, E. M. J.; Plotzke, K. Revisiting bioaccumulation criteria for POPs and PBT assessments. Integr. Environ. Assess. Manage. 2009, 5, 624–637.

ES101247E