In Vivo Passive Sampling of Nonpolar Contaminants in Brown Trout

Sep 10, 2013 - In Vivo Passive Sampling of Nonpolar Contaminants in Brown Trout. (Salmo trutta). Ian John Allan,* Kine Bæk, Thrond Oddvar Haugen, Kat...
1 downloads 0 Views 965KB Size
Download PDF
Article pubs.acs.org/est

In Vivo Passive Sampling of Nonpolar Contaminants in Brown Trout (Salmo trutta) Ian John Allan,* Kine Bæk, Thrond Oddvar Haugen, Kate Louise Hawley, Andreas Sven Høgfeldt, and Adam David Lillicrap Oslo Centre for Interdisciplinary Environmental and Social Research, Norwegian Institute for Water Research, Gaustadalléen 21, NO-0349, Oslo, Norway S Supporting Information *

ABSTRACT: Equilibrium passive sampling through in vivo implantation can help circumvent complex extractions of biological tissues, provide more accurate information on chemical contaminant burden based on the fugacity of a chemical in an organism rather than conventional normalization to lipid content, and improve the assessment of contaminant bioaccumulation potential. Here, we explored the feasibility of in vivo implantation for the passive sampling of neutral hydrophobic contaminants through the insertion of a silicone tag into brown trout (Salmo trutta). Implanted fish from the upper reaches of the River Alna (Oslo, Norway) were relocated to a polluted section of the river for a 28 day caged exposure. “Whole fish” lipidsilicone distribution coefficients (Dlip‑sil) were calculated for chlorinated compounds measured in whole fish and in silicone tags of 13 fish. Dlip‑sil ranged from 13.6 to 40.0 g g−1 for polychlorinated biphenyl congeners 28−156 (CB28 and CB156), respectively, and are in close agreement with literature in vitro lipid phase and tissue-based lipid-silicone partition coefficients. After dissection a further of eight fish, muscle and liver samples were analyzed separately. Muscle-based Dlip‑sil values similar to the whole fish data were observed. However, lipid-normalized concentrations in the liver tended to be lower than in muscle for most compounds (by up to 50%). Values of whole fish Dlip‑sil for brominated diphenyl ethers determined for three fish were in the range of 8.6−51 g g−1 and in agreement with chlorinated substances. Finally, fugacity ratios calculated from equilibrium concentrations in fishimplanted and water-exposed silicone provided information on the bioaccumulation for chlorinated compounds as well as for some polycyclic aromatic hydrocarbons. Equilibrium passive sampling through in vivo implantation can allow the comparison of a chemical’s activity or fugacity in biotic as well as abiotic environmental compartments and at different trophic levels up to humans.

■

INTRODUCTION The assessment of the bioaccumulation of nonpolar nonionized contaminants such as polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyls (PCBs) or polybrominated diphenyl ethers (PBDEs) into aquatic organisms generally relies on solvent-based exhaustive extractions of contaminants in either the whole organisms or specific tissues. Extraction and cleanup procedures for biological matrices can be complex and lengthy with extracts requiring extensive cleanup before analysis. Further normalization of concentrations to the lipid content of the matrix being analyzed is undertaken assuming that lipids are the main component responsible for a tissue or organism’s fugacity capacity. In addition, the total lipid content of the sample measured during sample preparation is only a crude descriptor of the actual lipid content and composition of an organism or matrix.1 It does not distinguish between lipid types and may not necessarily be inclusive of all types of lipids. Possible differences in the capacity of different types of lipids such as membrane or storage lipids, or as a result of the lipid composition found in different species at different trophic levels © 2013 American Chemical Society

means that normalizing contaminant concentrations to the lipid content of the matrix being extracted is of limited value.2 This approach does also not take into account other phases such as proteins that may exhibit significant sorption capacity particularly relevant for organisms of low lipid content.3 These factors tend to weaken comparisons of contaminant concentrations between different organisms and at different trophic levels.2 When a thermodynamic equilibrium is reached between two environmental phases or compartments, all phases have equal fugacity, f or activity.4,5 Recently, the application of in vitro passive sampling with silicone polymer immersed in lipids6 or inserted into lipid-rich fish tissues7,8 showed that it was possible to obtain an equilibrium between PCB concentrations in lipids Received: Revised: Accepted: Published: 11660

April 24, 2013 September 10, 2013 September 10, 2013 September 10, 2013 dx.doi.org/10.1021/es401810r | Environ. Sci. Technol. 2013, 47, 11660−11667

Environmental Science & Technology

Article

environment or that of their prey(s). For these organisms, organism-water fugacity ratios above unity can occur and are representative of biomagnification processes. Ecological factors such as home range and feeding habits, possible biotransformation processes, and nonequilibrium conditions can affect the levels of contaminants in these organisms. The aim of this pilot study was to evaluate the feasibility of in vivo implantation passive sampling to measure the internal fugacity of nonpolar nonionized contaminants in brown trout (Salmo trutta) over 28 days. The hypothesis was that contaminants accumulating in the fish would subsequently partition into polydimethylsiloxane tags inserted in the body cavity of brown trout. Data from the silicone tags for chlorinated and brominated compounds were evaluated against fish tissue concentrations (on a lipid weight basis) and literature values of lipid-silicone partition coefficients (Klip‑sil). Fugacity ratios for chlorinated substances and for PAHs calculated from silicone exposed to the fish and in water were used to estimate the degree of “equilibrium” established between a compounds concentration in fish and in water.

and that in lipid-immersed silicone (or polydimethylsiloxane) with equal fugacity in the two media (eq 1): flip = fsil

(1)

with f lip and fsil, contaminant fugacities (expressed in Pa) in the lipid phase and in the silicone, respectively. Equation 1 can be rewritten in terms of contaminant concentrations, Clip and Csil (mol m−3) and fugacity capacity of the two media, Zlip and Zsil (mol m−3 Pa−1), yielding lipid-silicone partition coefficients, Klip‑sil: C lip Csil

=

Z lip Zsil

= Klip − sil

(2)

Values of Klip‑sil for olive, Tobis and seal oils representing lipids found at different trophic levels were in the range of 10− 50 g g−1 for PCBs and other chlorinated pesticides.6 The application of in vitro equilibrium sampling with polydimethylsiloxane in intact fish tissues and Dugong blubber resulted in the measurement of lipid-silicone distribution coefficients, Klip‑sil in a similar range.8−10 Passive sampling methods have been used for two decades for the sampling of nonpolar nonionized and possibly bioaccumulative substances in water and sediments.11−13 In vivo passive sampling using solid phase microextraction fibres inserted into the muscle of living fish has been undertaken for the measurement of some pharmaceutical compounds and pesticides.14,15 This technique has more recently been applied to the measurement of contaminant bioaccumulation and elimination kinetics in fish with a reduced number of organisms needed.16 When sampling (such as in these studies) is undertaken in the kinetic mode of uptake,16 tight control over exposure conditions (e.g., exact physical location of the inserted fiber) and an understanding of factors influencing sampling rates are needed. In vivo implantation passive sampling has recently allowed the measurement of the fugacity of chlorinated and brominated persistent organic pollutants in humans.17 At equilibrium, the fugacity of the chemical in the polymeric material is equal to that in the organism, itself composed of a range of tissues/phases of differing volumes, Vi (m3) and fugacity capacities Zi: fsil =

M ∑ VZ i i

■

MATERIAL AND METHODS Solvent and Standards. Ultrapure water was from an Elgastat Maxima HPLC Deionization option 3 system. HPLCgrade dichloromethane and pentane were from Rathburn. Chromasolv-quality isopropanol was from Riedel-de-Haen and HPLC grade cyclohexane was from J.T. Baker. Standards for PAHs and their deuterated homologues from Chiron were of analytical-grade with purities of >99% for PAHs and >99.5% for deuterated PAHs. Analytical-grade standards and surrogate standards for organochlorinated compounds (OCs) and PBDEs were from LGC/Promochem. The full list of chemicals under study is given in Supporting Information (SI) Table S1. Silicone Tag and Strip Preparation. Medical grade Silastic A tubing (4.9 mm external diameter, 2.6 internal diameter) purchased from Cole-Parmer/Tekmo was used to prepare 2.5 cm long tags. The silicone tubing was Soxhlet extracted with ethyl acetate for 24 h prior to use. The silicone tags were then further cleaned in methanol before being kept in the freezer at −20 °C until use. A total of 25 tags were prepared and their mass was on average 0.382 g (1.5% RSD). Tags were not spiked with performance reference compounds. AlteSil silicone strips (2.5 cm wide, 60 cm long and 0.5 mm thick) from Altec Products Ltd. were also Soxhlet extracted for 24 h prior to use. Samplers were further soaked in methanol before spiking with performance reference compounds (deuterated PAHs). Spiking of PRC was performed according to procedures previously published using a methanol:water solution.19,20 PRCs with concentrations between 1 and 6 μg sampler−1 (relative standard deviation 3.5, according to Rs = βsil Ksw−0.08, with K sw the sampler-water partition coefficient for AlteSil silicone.22,24 More details are given in SI. The calculation of fugacity ratios through the comparison of contaminant concentrations in silicone tags from in vivo exposure (through implantation) with water-exposed silicone samplers may be a useful approach to assess contaminant bioaccumulation in an organism. Equilibrium between the concentration of contaminants in water-exposed silicone samplers and that in water is generally not reached for

90 and 110% for PAHs and chlorinated and brominated compounds. Passive Sampling Data Handling. The lipid content of the fish used in this study were relatively low, therefore minimal lipid absorption may be expected. Silicone tags were not weighed following exposure, so it is not possible to know gravimetrically whether significant lipid absorption took place. Nonetheless contaminant concentrations in silicone tags were not corrected for potential influence of absorbed lipids on lipidsilicone partition coefficients.6 Measurements in the fish (reported on a lipid weight basis) and in the silicone tags enabled the estimation of contaminant distribution coefficients, Dlip‑sil (g g−1) between fish/fish tissues (on a lipid basis) and silicone tags. For silicone strips exposed to the river water, a boundary layer-controlled uptake model was used for estimating sampling 11663

dx.doi.org/10.1021/es401810r | Environ. Sci. Technol. 2013, 47, 11660−11667

Environmental Science & Technology

Article

randomly selected fish (SI Tables S4 and S5). Median values of Dlip‑sil for PBDEs are given in Table 2. The variability of in vivo Dlip‑sil for BDE28 (median = 4.6 g g−1; range 3.4−18 g g−1) for the three fish is higher than that found for PCBs with a similar hydrophobicity (Table 2). Median Dlip‑sil were 28 (18−38), 39 (31−40) and 54 g g−1 (31−68) for BDE47, BDE100, and BDE99, respectively. Logarithms of lipid-silicone distribution coefficients (logKlip‑sil) for PBDEs are in agreement with those for OCs when plotted as a function of logKow (SI Figure S3). It is interesting to note that in vivo Dlip‑sil values are relatively constant for the range of hydrophobicity from logKow of 5.0 to over 6.0. A mild increase in lipid-silicone distribution coefficients can be observed from logKow of 6 to 7.2. Our data was not corrected for possible amounts of lipid absorbed by the silicone and comparison were made with uncorrected in vitro Klip‑sil from Jahnke et al. (2008).6 The reason for this is that PDMS-absorbed lipids may not necessarily have an influence on Klip‑sil.26 Some uncertainty can also result from potential difference in the capacity of polydimethylsiloxane or silicones from different sources to absorb neutral hydrophobic compounds.24 Fish−Silicone Tag Equilibrium. An important question is whether the silicone tags have reached some form of equilibrium with surrounding tissues. Assuming a case where most of the resistance to mass transfer into the silicone tag is in the polymer, approximate times needed for silicone tags to equilibrate would be in the range of 1−5 days for PCBs according to a simple version of mass transfer model (see SI Figure S4). This means that, for our in vivo measurements, times to equilibrium may be longer than that. The measurement of in vivo Dlip‑sil through implantation is in excellent agreement with published data and tends to support the equilibrium nature of the sampling undertaken here. The agreement between in vivo Dlip‑sil and literature values of in vitro Klip‑sil from lipid-immersed silicone assays (e.g., for hexachlorobenzene) together with the lower variability of whole fish concentrations when normalized to the lipid content (not shown) tends to show that lipids were the main component responsible for fugacity capacity in our brown trout. Correlation coefficients for lipid-normalized whole fish− silicone tag concentrations are consistently higher than those based on a wet weight basis (SI Table S6). This supports the notion that concentrations in fish (lipid-normalized) and in silicone tags are likely to be related. While for lipid-rich tissues, “in tissue” exposures are expected to reach equilibrium within a few hours, equilibration times over seven days have been observed for less lipid-rich tissues.9 True equilibrium is difficult to define here since the fish were thinning (Table 1) during exposure meaning that the system was dynamic. While the brown trout showed a relatively low overall lipid content (2.2% wet weight basis) after exposure, the 14% body weight loss over the 28 days caging (Table 1) means that the lipid content and composition at the start of the experiment may have been different due to lipid mobilization during starving.27 Starvation of seawater-reared brown trout for a period of eight weeks has been shown to result in a 10% body weight loss.28 In that study, losses of both lipids and protein were observed.28 Lipid losses may help remobilize contaminants within the fish and this can affect internal equilibrium and the time needed for tags to equilibrate with the fish. The proportion of lipids in visceral tissues was shown to be significantly higher than in muscle tissues or in the liver.28 This means silicone tags implanted in the ventral part of the fish

substances with logKow > 5−6 over commonly used monthly exposures.20,25 For these substances, equilibrium concentrations in silicone that can be anticipated if samplers were left for a sufficiently long period of time are easily calculated from estimated Cw and Ksw values.24 Despite differences in logKsw values for the two silicones (from different suppliers for the tags and sheets) generally not more than 0.1 log unit for PAHs and PCBs,24 data were corrected for these differences.

■

RESULTS AND DISCUSSION Whole Fish−Silicone Distribution Coefficients for Chlorinated Compounds. To assess the relevance of internal chemical fugacity of these substances absorbed in the brown trout and measured with implanted silicone tags, we compared fish concentrations on a lipid weight basis with those in the silicone tags (data in SI Tables S2 and S3). “Whole fish” lipidsilicone distribution coefficients, Dlip‑sil (g g−1) were measured for 13 fish and data are given in the boxplot of Figure 1. Data are only given for those above limits of quantification. Dlip‑sil values for PCBs and chlorinated organics are in the range 10 to 50 g g−1. Median values for Dlip‑sil (n = 13) presented in Figure 1 are given in Table 2 (together with median absolute deviations). A very low variability in Dlip‑sil can be observed for substances such as HCB and CB28 up to CB118. Median absolute deviations (n = 13) were generally between 10 and 20% of median values for all substances listed in Table 1. The lowest median absolute deviations were for HCB and CB52. On average, the RSD was close to 20% for all substances. The increase in Dlip‑sil from CB28 to CB118 was observed to be, at a maximum, a factor of 2 for an increase in logKow of about one log unit. More variability was seen for CB153 to CB156 (Figure 1) and only minor increases in Dlip‑sil can be observed. CB180 was not found in the silicone tags while PeCB was close to limits of quantification. These values are in excellent agreement with in vitro Klip‑sil from Jahnke et al. (2008)6 who measured the partitioning of organochlorinated compounds between lipids and polydimethylsiloxane (PDMS) immersed in three lipids of differing composition and origin. Since relatively similar partitioning to the three types of lipids (olive, Tobis, and seal oils) was observed by these researchers, we used their generic (average) values for comparison. Values uncorrected for lipid uptake (Dlipid,pdms in Jahnke et al., 20086) were selected and are given in Table 2 (but termed Klip‑sil here). For example, the median in vivo whole fish Dlip‑sil for brown trout for HCB is in close agreement with the generic in vitro Klip‑sil of 13.5 g g−1. For the remaining chlorinated compounds of Table 2, the deviation between median Dlip‑sil for brown trout and generic Klip‑sil6 was on average below 20% (range of 6−34%). In vitro equilibrium passive sampling measurements of PCBs have also been undertaken by direct insertion of the PDMS phase into the fish tissue.8 Resulting in vitro Klip‑sil for Norwegian Atlantic salmon, Baltic Sea eel and Finnish lake eel calculated from Jahnke et al. (2011)8 are also provided in Table 2. Values of Dlip‑sil for brown trout are in closest agreement with Klip‑sil for the Norwegian Atlantic salmon and the Baltic Sea eel, with deviations on average of 17 and 21%, respectively. For the Finnish lake eel, Klip‑sil for CB28 to CB118 were consistently 50% higher than our results for brown trout. Differences were minimal however for CB153 and C138. To our knowledge, lipid-silicone distribution coefficients for PBDEs have not been reported in the literature. Here, the concentration of polybrominated diphenyl ether congeners was measured in tags and whole fish homogenates for three 11664

dx.doi.org/10.1021/es401810r | Environ. Sci. Technol. 2013, 47, 11660−11667

Environmental Science & Technology

Article

sufficient to observe equilibrium between contaminant concentrations in the silicone sheets and the freely dissolved concentration in water, particularly for analytes with logKow > 5 (see SI Table S11). The concentration at equilibrium could, however, be calculated from the estimated freely dissolved concentrations in water and the sampler-water partition coefficients (Ksw). Estimated freely dissolved concentrations for selected PAHs and OCs are given in Table S11 in the SI. Concentrations of OCs tend to be similar to those measured the previous year at this site in the Alna while PAH concentrations were substantially lower than in 2011.21 PAH concentrations in silicone tags are given in SI Table S12. Fugacity ratio equivalents were calculated as the ratios of equilibrium concentrations measured in the 13 fish-exposed silicone tags (data not corrected for possible lipid content) over those estimated for the silicone sheets exposed to the water. Results are shown in Figure 2 for PAHs and OCs. For these

were likely exposed to tissues with a higher lipid content (viscera and belly flap) than if exposed to the muscle or liver. This means that time to equilibrium may be shorter than those expected for in vitro in tissue exposures to tissues of low lipid content. As opposed to silicone phases exposed to nonliving matrices, blood circulation and organism movements during in vivo exposures are likely to play an important role in the time needed for an inserted reference phase such as silicone to come to equilibrium with its surrounding tissues. Results from intraperitoneal injection of PCB-spiked oil into Round Goby suggested a rapid intertissue distribution of PCBs in these small fish.29 Muscle- and Liver-Silicone Distribution Coefficients. Liver- and muscle-based Dlip‑sil (SI Table S7, S8 and S9) were obtained for the 8 largest fish (median values for OCs given in Table 2). The range of Dlip‑sil measured for fish muscle overlaps well with values from “whole fish” measurements for most substances under study. Ranges of values observed for muscle tissues are wider than “whole fish” data. Liver-silicone distribution coefficients are for most substances, lower than those for muscle tissues (Figure S5 in the SI). Most pronounced differences are for CB101, CB118, CB153 and CB138. Concentrations in fish were the lowest and closest to limits of detection for CB156 and this may be why data for this congener show high variability. For HCB, CB28 and CB52, Dlip‑sil values are in close agreement. For example, median Dlip‑sil values for HCB are 13.9, 13.1, and 12.2 for “whole fish”, muscle and liver, respectively (Table 2). As shown on SI Figure S1, fish used for muscle and liver measurement were larger than those for “whole fish” analysis. The good agreement between in vivo Dlip‑sil obtained for small and larger fish (muscle) indicate that the measurement is independent of the fish size. It is perhaps not surprising to observe the closest overlap of Dlip‑sil between the whole fish data and the muscle data. Despite a higher lipid content of the liver samples (Table 1), the proportion of muscle tissues (and lipids) in the fish compared with that of the liver is much higher. Once inserted into the fish, the silicone tag was positioned in the peritoneal cavity, near the viscera and close to both muscle tissues and the liver. Despite aiming to implant all tags from the same incision point, these were not necessarily located in exactly the same position in all fish and this may have affected the variability observed between paired fish-tag data. It may be that the lipid composition of the muscle tissues and of the liver differ, particularly after fasting. For example, saturated fatty acid mobilization upon starving appear to take place preferentially in the viscera for rainbow trout.27 Different lipids may also exhibit different capacities to absorb contaminants resulting in differences in in vivo Dlip‑sil.2 In our study the lipid content of whole fish, liver and muscle was measured during the matrix extraction for contaminant analysis. It is a relatively crude surrogate of the actual lipid content and precludes further detailed discussion of the reasons for differences in lipidnormalized contaminant concentrations in liver and muscle tissues. Fish−Water Fugacity Ratios. Contaminant concentrations in the silicone tags (assumed to be at equilibrium), a measure of the fugacity of chemicals in the fish, can be compared with those found in the silicone sheets at equilibrium with freely dissolved concentrations in water. Details of the calculation of freely dissolved concentration and estimation of equilibrium concentrations in silicone (and their associated uncertainties) are given in SI. The exposure time was not

Figure 2. Fugacity ratios calculated as the ratios of equilibrium silicone concentrations measured in fish-exposed tags (n = 13 “whole fish”) and in water-exposed sheets. Data points represent individual tag concentrations divided by the mean of triplicate water-exposed silicone sheet measurements.

lipophilic compounds that partition preferentially to the fish lipids, ratios close to one would tend to indicate that the concentration in the fish is close to steady-state, or apparent equilibrium with the freely dissolved concentration in water. Ratios falling below 1 indicate that the fish is “underconcentrated” with respect to the water phase. This can occur as a result of low accumulation rates (far from equilibrium), biotransformation or dilution (e.g., through growth). Ratios above one would represent cases of biomagnification or rapid weight loss. Spatiotemporal inconsistencies in wild fish and passive sampler exposures could also result in ratios deviating from one.30 Ratios for most PCB congeners, HCB, PeCB and some PAHs such as anthracene, dibenzothiophene, phenanthrene, and fluorene were generally close to one. The excellent agreement of in vivo Dlip‑sil with literature values of Klip‑sil and fish−water fugacity ratios close to one tend to indicate that the concentration of these contaminants in the fish is in apparent equilibrium with that in the water. Fish−water fugacity ratios for acenaphthene and naphthalene suggest that the fish are “over-concentrated” for these two substances. It is likely that fish and water-exposed silicone sheets are representative of different ambient freely 11665

dx.doi.org/10.1021/es401810r | Environ. Sci. Technol. 2013, 47, 11660−11667

Environmental Science & Technology

Article

weights, fish tag information, raw data for OCs and PBDEs concentrations in silicone tags and biota samples, modeling of time to equilibrium for tags under polymer controlled-uptake, comparison of OC concentrations in fish liver and muscle samples, calculation of freely PAH and OC dissolved concentrations and their uncertainty, PAH concentrations in tags and logKow values for OCs and PBDEs and their sources. This information is available free of charge via the Internet at http://pubs.acs.org/.

dissolved concentrations in water for these compounds. Since the concentration of moderately hydrophobic substances such as naphthalene in silicone sheet reaches equilibrium with that in water very fast, data from silicone sheets is representative only of the last few days of exposure. Fugacity ratios for fluoranthene and pyrene are over an order of magnitude below one. This is not surprising since fluoranthene and pyrene are efficiently metabolized and have low half-lives in fish.31 Ratios for pyrene are close to an order of magnitude lower than those observed for fluoranthene, and this is in agreement with the fish biotransformation half-life of 0.556 and 2.57 days for pyrene and fluoranthene given by the BCFBAF v3.01 (EPI Suite v4.10).31 Of all PAHs detected and quantified in the silicone tags, fluoranthene and pyrene are the only ones whose fugacity ratios appear to be affected by metabolism in the fish. The less hydrophobic PAHs are generally close to the 1:1 relationship. Recent modeling32 showed that steady-state bioconcentration factors (BCFs) for substances with logKow > 5.0 are expected to be significantly influenced by hepatic biotransformation. In contrast these authors explained that hepatic biotransformation rates lower than rates of chemical elimination through fish gills for chemicals with logKow < 3 have negligible impact on BCF values. This means that hepatic transformation may have a limited impact on BCF values for these compounds.32 The measured fish−water ratio of equilibrium silicone concentration for pyrene of 0.003 is in agreement with pyrene fugacity ratios summarized in Burkhard et al. (2011)33 generally well below one for fish. Implications for Future Studies. In vivo equilibrium passive sampling of neutral hydrophobic compounds can be used to measure the internal fugacity or activity in fish and other organisms.17 This measure of internal fugacity can be compared with the fugacity of the chemical in other (a)biotic compartments measured with the same polymer (with the same fugacity capacity). Such fugacity ratios can provide relevant information on contaminant bioaccumulation potential within an organism, accounting for all components contributing to the organism’s contaminant absorption capacity (eq 3). Such a method has the potential to be applied to organisms with lower lipid content, although organism size and practical experimental time frame and time to equilibrium will remain crucial factors that require further investigations. When investigating organisms at different trophic levels across food webs, in vivo sampling may need to be combined with in vitro tissue-based measurements since not all organisms are amenable to implantation. The PRC procedure may be developed in the future to assess equilibrium or steady-state conditions between the organism and the polymeric phase. This way, tissue-silicone distribution of “native”/naturally accumulated contaminants and PRCs may be compared. The interpretation of results from “whole organism” ecotoxicological testing, BCF/BAF measurements, or caged organism exposures may benefit from a measurement of contaminant activity/fugacity inside the organism. This in vivo study confirms the validity of literature K or Dlip‑sil values and supports their use for the calculation of abiotic measures of lipid-normalized contaminant concentrations in biota based on passive sampling in water or sediments for comparison with actual biota concentrations.

■

■

AUTHOR INFORMATION

Corresponding Author

*(I.J.A.) Phone: +47 22 18 5100; fax: +47 22 18 5200; e-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Alfhild Kringstad and Etienne Vermeirssen are thanked for support with the analytical work for this study and for discussions of the fish caging, respectively. We thank the Norwegian Research Council for funding this work through NIVA’s basic research funding for innovation. We also thank four anonymous reviewers for comments that helped improve this manuscript.

■

REFERENCES

(1) Randall, R. C.; Lee, H.; Ozretich, R. J. Evaluation of selected lipid methods for normalizing pollutant bioaccumulation. Environ. Toxicol. Chem. 1991, 10 (11), 1431−1436. (2) van der Heijden, S. A.; Jonker, M. T. O. Intra- and interspecies variation in bioconcentration potential of polychlorinated biphenyls: Are all lipids equal? Environ. Sci. Technol. 2011, 45 (24), 10408− 10414. (3) Endo, S.; Bauerfeind, J.; Goss, K. U. Partitioning of neutral organic compounds to structural proteins. Environ. Sci. Technol. 2012, 46 (22), 12697−12703. (4) Mackay, D. Correlation of bioconcentration factors. Environ. Sci. Technol. 1982, 16 (5), 274−278. (5) Mackay, D. Finding fugacity feasible. Environ. Sci. Technol. 1979, 13 (10), 1218−1223. (6) Jahnke, A.; McLachlan, M. S.; Mayer, P. Equilibrium sampling: Partitioning of organochlorine compounds from lipids into polydimethylsiloxane. Chemosphere 2008, 73 (10), 1575−1581. (7) Ossiander, L.; Reichenberg, F.; McLachlan, M. S.; Mayer, P. Immersed solid phase microextraction to measure chemical activity of lipophilic organic contaminants in fatty tissue samples. Chemosphere 2008, 71 (8), 1502−1510. (8) Jahnke, A.; Mayer, P.; Adolfsson-Erici, M.; McLachlan, M. S. Equilibrium sampling of environmental pollutants in fish: Comparison with lipid-normalized concentrations and homogenization effects on chemical activity. Environ. Toxicol. Chem. 2011, 30 (7), 1515−1521. (9) Jahnke, A.; Mayer, P.; Broman, D.; McLachlan, M. S. Possibilities and limitations of equilibrium sampling using polydimethylsiloxane in fish tissue. Chemosphere 2009, 77 (6), 764−770. (10) Jin, L.; Gaus, C.; van Mourik, L.; Escher, B. I. Applicability of passive sampling to bioanalytical screening of bioaccumulative chemicals in marine wildlife. Environ. Sci. Technol. 2013, 47 (14), 7982−7988. (11) Vrana, B.; Mills, G. A.; Allan, I. J.; Dominiak, E.; Svensson, K.; Knutsson, J.; Morrison, G.; Greenwood, R. Passive sampling techniques for monitoring pollutants in water. TrAC, Trends Anal. Chem. 2005, 24 (10), 845−868. (12) Mayer, P.; Tolls, J.; Hermens, L.; Mackay, D. Equilibrium sampling devices. Environ. Sci. Technol. 2003, 37 (9), 184A−191A.

ASSOCIATED CONTENT

S Supporting Information *

List of chemicals of interest in this study, further details and limits of detection for the various analyses, fish length and 11666

dx.doi.org/10.1021/es401810r | Environ. Sci. Technol. 2013, 47, 11660−11667

Environmental Science & Technology

Article

(13) Huckins, J. N.; Petty, J. D.; Booij, K. Monitors of Organic Chemicals in the Environment: Semipermeable Membrane Devices; Springer: New York, 2006. (14) Ouyang, G. F.; Oakes, K. D.; Bragg, L.; Wang, S.; Liu, H.; Cui, S. F.; Servos, M. R.; Dixon, D. G.; Pawliszyn, J. Sampling-rate calibration for rapid and nonlethal monitoring of organic contaminants in fish muscle by solid-phase microextraction. Environ. Sci. Technol. 2011, 45 (18), 7792−7798. (15) Zhou, S. N.; Oakes, K. D.; Servos, M. R.; Pawliszyn, J. Application of solid-phase microextraction for in vivo laboratory and field sampling of pharmaceuticals in fish. Environ. Sci. Technol. 2008, 42 (16), 6073−6079. (16) Adolfsson-Erici, M.; Akerman, G.; McLachlan, M. S. In-vivo passive sampling to measure elimination kinetics in bioaccumulation tests. Chemosphere 2012, 88 (1), 62−68. (17) Allan, I. J.; Bæk, K.; Kringstad, A.; Roald, H. E.; Thomas, K. V. Should silicone prostheses be considered for specimen banking? A pilot study into their use for human biomonitoring. Environ. Int. 2013, 59, 462−468. (18) Maenpaa, K.; Leppanen, M. T.; Reichenberg, F.; Figueiredo, K.; Mayer, P. Equilibrium sampling of persistent and bioaccumulative compounds in soil and sediment: Comparison of two approaches to determine equilibrium partitioning concentrations in lipids. Environ. Sci. Technol. 2011, 45 (3), 1041−1047. (19) Booij, K.; Smedes, F.; van Weerlee, E. M. Spiking of performance reference compounds in low density polyethylene and silicone passive water samplers. Chemosphere 2002, 46 (8), 1157− 1161. (20) Allan, I. J.; Harman, C.; Kringstad, A.; Bratsberg, E. Effect of sampler material on the uptake of PAHs into passive sampling devices. Chemosphere 2010, 79 (4), 470−475. (21) Allan, I. J.; Ranneklev, S. B. Occurrence of PAHs and PCBs in the Alna River, Oslo (Norway). J. Environ. Monit. 2011, 13 (9), 2420− 2426. (22) Rusina, T. P.; Smedes, F.; Koblizkova, M.; Klanova, J. Calibration of silicone rubber passive samplers: Experimental and modeled relations between sampling rate and compound properties. Environ. Sci. Technol. 2010, 44 (1), 362−367. (23) Booij, K.; Smedes, F. An Improved Method for Estimating in situ sampling rates of nonpolar passive samplers. Environ. Sci. Technol. 2010, 44 (17), 6789−6794. (24) Smedes, F.; Geertsma, R. W.; van der Zande, T.; Booij, K. Polymer-water partition coefficients of hydrophobic compounds for passive sampling: Application of cosolvent models for validation. Environ. Sci. Technol. 2009, 43 (18), 7047−7054. (25) Allan, I. J.; Booij, K.; Paschke, A.; Vrana, B.; Mills, G. A.; Greenwood, R. Field performance of seven passive sampling devices for monitoring of hydrophobic substances. Environ. Sci. Technol. 2009, 43 (14), 5383−5390. (26) Jahnke, A.; Mayer, P. Do complex matrices modify the sorptive properties of polydimethylsiloxane (PDMS) for non-polar organic chemicals? J. Chromatog., A 2010, 1217 (29), 4765−4770. (27) Jezierska, B.; Harzel, J. R.; Gerking, S. D. Lipid mobilization during starvation in the rainbow trout, Salmo gairdneri Richardson, with attention to fatty acids. Journal of Fish Biology 1982, 21, 681−692. (28) Regost, C.; Arzel, J.; Cardinal, M.; Laroche, M.; Kaushik, S. J. Fat deposition and flesh quality in seawater reared, triploid brown trout (Salmo trutta) as affected by dietary fat levels and starvation. Aquaculture 2001, 193 (3−4), 325−345. (29) O’Neil, J. A.; Johnson, T. B.; Drouillard, K. G. Validation of rapid assimilation of PCBs following IP dosing in the round goby (Neogobius melanostomus). Bull. Environ. Contam. Toxicol. 2013, 91 (2), 135−140. (30) Selck, H.; Drouillard, K.; Eisenreich, K.; Koelmans, A. A.; Palmqvist, A.; Ruus, A.; Salvito, D.; Schultz, I.; Stewart, R.; Weisbrod, A.; van den Brink, N. W.; van den Heuvel-Greve, M. Explaining differences between bioaccumulation measurements in laboratory and field data through the use of a probabilistic modeling approach. Integr. Environ. Assess. Manage. 2012, 8 (1), 42−63.

(31) Arnot, J. A.; Mackay, D.; Parkerton, T. F.; Bonnell, M. A database of fish biotransformation rates for organic chemicals. Environ. Toxicol. Chem. 2008, 27 (11), 2263−2270. (32) Nichols, J. W.; Huggett, D. B.; Arnot, J. A.; Fitzsimmons, P. N.; Cowan-Ellsberry, C. E. Towards improved models for predicting bioconcentration of well-metabolized compounds by rainbow trout using measured rates of in vitro intrinsic clearance. Environ. Toxicol. Chem. 2013, 32 (7), 1611−1622. (33) Burkhard, L. P.; Arnot, J. A.; Embry, M. R.; Farley, K J; Hoke, R. A.; Kitano, M.; Leslie, H. A.; Lotufo, G. R.; Parkerton, T. F.; Sappington, K. G.; Tomy, G. T.; Woodburn, K. B. Comparing laboratory and field measured bioaccumulation endpoints. Integr. Environ. Assess. Manage. 2011, 8 (1), 17−31.

11667

dx.doi.org/10.1021/es401810r | Environ. Sci. Technol. 2013, 47, 11660−11667