Development and Evaluation of a Database of ... - ACS Publications

Development and Evaluation of a Database of...

0 downloads 93 Views 1MB Size

Critical Review pubs.acs.org/est

Development and Evaluation of a Database of Dietary Bioaccumulation Test Data for Organic Chemicals in Fish Jon A. Arnot*,†,‡ and Cristina L. Quinn†,§ †

ARC Arnot Research and Consulting Inc., 36 Sproat Avenue, Toronto, Ontario, Canada, M4M 1W4 Department of Physical and Environmental Sciences, University of Toronto Scarborough, 1265 Military Trail, Toronto, Ontario, Canada, M1C 1A4

‡

S Supporting Information *

ABSTRACT: Dietary bioaccumulation tests for ﬁsh have been conducted for about 40 years. Standardized test guidance has recently been developed. Test metrics of primary scientiﬁc and regulatory interest are the whole body depuration rate constant (kT), whole body growth corrected depuration rate constant (kTg), and corresponding chemical half-lives (t1/2 and t1/2g), dietary chemical absorption eﬃciency (AE), and biomagniﬁcation factor (BMF). A database of 3032 measurement end points for 477 discrete organic chemicals including 964 half-lives, 1199 AEs and 869 BMFs from 19 species (primarily trout and carp) was developed from the literature. Biological properties (e.g., organism weight, lipid content) and exposure conditions (e.g., temperature, feeding rate, dietary lipid content, exposure duration) are documented. Test chemicals range in molar mass from 120 to 1423 g·mol−1 with log octanol−water partition coeﬃcients (KOW) ranging from 0.8 to 14.3; 50% of the database entries are for polychlorinated biphenyls. The measured end points are derived from various protocols and sources of variability are described. The data are evaluated and categorized using proposed data quality (conﬁdence) criteria derived from the standardized test protocol providing initial guidance for data users. Half-lives range from 0.13 to 2600 days; however, approximately 54% have an identiﬁable source of uncertainty. The data suggest that chemicals absorbed from the gastrointestinal tract with a log KOW ≥ ∼5 and at least as high as ∼9 have biomagniﬁcation potential in ﬁsh. A mechanistic bioaccumulation model is compared to the measured data and used to illustrate the inﬂuence of growth and biotransformation rates on the BMF.

■

Dietary bioaccumulation testing in ﬁsh has been conducted for about 40 years led by the pioneering experiments of Hamelink et al.,1 Zitko,33−36 Bruggeman, Hutzinger, Opperhuizen, and colleagues;4,37−40 Niimi, Oliver, and colleagues;41−46 Gobas, Mackay, and colleagues;5,7,47−49 and Muir, Fisk, and colleagues,50−57 with contributions from several others. In 1987 Niimi41 reviewed biological half-lives of chemicals in ﬁsh. In 2008 Barber58 reviewed several models for describing bioaccumulation from dietary exposure with a focus on chemical assimilation eﬃciency. In 2012 the Organization for Economic Co-operation and Development (OECD) dietary bioaccumulation testing guidance for ﬁsh was adopted in the revision of Test Guideline 305.59 The new guidance59 provides context for a systematic review of dietary bioaccumulation testing methods, literature, and data. The objectives here are to collect dietary bioaccumulation testing data, critically examine the existing data and literature against the guiding principles in the new OECD 305 guidelines,59 and provide recommendations for the selection

INTRODUCTION Chemical bioaccumulation in ﬁsh and food webs is of scientiﬁc1−17 and regulatory18−21 interest. Various criteria and metrics are used for bioaccumulation assessment including the octanol−water partition coeﬃcient (KOW), bioconcentration factor (BCF), bioaccumulation factor (BAF), biomagniﬁcation factor (BMF), biota-sediment accumulation factor (BSAF), and the trophic magniﬁcation factor (TMF).22,23 Currently the majority of measured bioaccumulation data are ﬁsh BCFs determined in laboratory tests from chemical exposure in the surrounding water only.24 For hydrophobic chemicals, there are technical challenges obtaining reliable quality BCFs.24,25 In the environment, dietary exposures can be relatively greater than aqueous exposures for hydrophobic chemicals26,27 and there is a potential for biomagniﬁcation.28−31 For neutral organic chemicals, biomagniﬁcation occurs when there is an increase in chemical fugacity from the diet to the consumer and biomagniﬁcation is determined using fugacity ratios or lipid corrected BMFs (i.e., BMFL > 1 indicates biomagniﬁcation).22,23 Regulatory agencies can consider BMFs as part of a bioaccumulation assessment.32 Reliable testing data are desirable for bioaccumulation assessment, improved understanding of processes, and for the development and evaluation of models. © XXXX American Chemical Society

Received: December 22, 2014 Revised: March 23, 2015 Accepted: March 30, 2015

A

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

Environmental Science & Technology

Figure 1. Overview of the general OECD 305 dietary bioaccumulation test guidance methods and calculations59 including (A) the experimental stages of chemical exposure and depuration, initial data analysis to calculate (B) the depuration rate constant (kT) and (C) the growth rate constant (kG), followed by (D) the calculation of other metrics including the half-life (t1/2), the growth corrected depuration rate constant (kTg), the growth corrected half-life (t1/2g), the chemical concentration in ﬁsh at start of depuration (CF0), the chemical absorption eﬃciency (AE), the kinetic biomagniﬁcation factor (BMFK), and the growth corrected kinetic biomagniﬁcation factor (BMFKg) using measurements or estimates of the chemical concentration in diet (CD) and the dietary ingestion rate (I).

in which case kG is not included40 and k1 and kD have units μg· d−1. Dietary Bioaccumulation Test Methods and Metrics. Figure 1 summarizes the general OECD 305 dietary bioaccumulation test methods and metrics.59 Fish are ﬁrst fed a measured quantity of intentionally contaminated food in the exposure (dosing) phase (Figure 1A). There is no chemical exposure from the water; k1CWD = 0 in eq 1. This is followed by a depuration phase in which the ﬁsh are no longer exposed to the test chemical. The CF measurements in the depuration phase are used to calculate kT (Figure 1B). Following test guidance59 CFs are log or ln transformed and plotted as a function of time and linear regression of the data is performed, the slope being −kT (Figure 1B). The growth rate can be determined throughout the testing period or in the depuration phase (Figure 1C) and used to calculate the growth corrected elimination rate constant kTg (Figure 1D). The corresponding half-lives are t1/2 and t1/2g, respectively. Alternatively, kTg and t1/2g can be calculated from XF.40,41,44,59,60 The parameters AE, BMFK (kinetic), and BMFKg (kinetic and growth corrected) are often subsequently calculated from kT and kTg using CD, I and ﬁsh concentrations at the beginning of the depuration phase (CF0).59 BMFs should be growth and lipid corrected.59 The mean lipid fractions in the ﬁsh (LF) and diet (LD) are used to calculate BMFL as BMF/(LF/LD). Growth and lipid corrected kinetic BMFs are BMFKLg (also see Supporting Information, SI1). Data Collection and Evaluation. Database Development. The BMF is applicable for biomagniﬁcation assessment.22,23 The BCF, BAF, and TMF can be estimated from kT,41,59,61 and t1/2 has been proposed for bioaccumulation assessment.62 Data were considered relevant63 if at least one of kT, t1/2, kTg, t1/2g, AE or BMF was clearly measured and documented from a dietary exposure experiment. Over 100 published studies were collected from the peer-reviewed literature and the public domain (ﬁnishing ca. September 2013). Only tests with clearly deﬁned organic chemicals and species were considered. If reported, biological, and exposure conditions for which the OECD 305 provides guidance were

of data for models and bioaccumulation assessment. Theoretical concepts and terminology related to dietary bioaccumulation testing and data analysis are ﬁrst presented. The new database of test information comprising several study parameters is described. Based on the OECD 305 guidelines,59 data quality criteria are developed and applied to evaluate and categorize existing data. Finally, some recommendations for future testing and guidance revision are provided.

■

THEORY Chemical uptake and elimination in a ﬁsh can be expressed using a typical toxicokinetic (bioaccumulation) mass balance model as11,37 dC F/dt = k1C WD + kDC D − (k 2 + kE + kB + k G)C F = k1C WD + kDC D − k TC F

(1)

where dCF/dt is the net change in chemical concentration in the ﬁsh over time t (d), CF is the concentration in the ﬁsh (μg· kg−1), k1 is the chemical uptake rate constant from water (L(kg· d)−1), CWD is the dissolved chemical concentration in water (μg·L−1), kD is the chemical uptake rate constant from the diet (kg(kg·d)−1), and CD is the chemical concentration (μg·kg−1) in diet. kD is a product of the food ingestion rate (I; kg(kg·d)−1) and the chemical transfer eﬃciency from the diet (often referred to as the assimilation or absorption eﬃciency, α, AE or ED; unitless). The rate constants corresponding to potential decreasing chemical concentrations in the ﬁsh (d−1) are for elimination to water, fecal egestion, biotransformation and growth dilution as k2, kE, kB, and kG, respectively. The total chemical elimination rate constant is the sum of various individual loss processes as kT = k2 + kE + kB + kG. Following ﬁrst-order kinetics the total biological half-life of the chemical in the ﬁsh is t1/2 = ln2/kT. Growth dilution is a pseudoelimination process because the chemical is not actually eliminated from the organism and dCF/dt is a function of changes in biomass. The chemical mass balance in eq 1 can also be expressed in terms of chemical quantity (body burden approach) in the ﬁsh (XF; μg) B

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

Environmental Science & Technology

Since kTg can be calculated from kT and kG (Figure 1D, eq S3), errors in kT or kG can propagate into kTg. When kT ≫ kG, errors in kTg correspond to errors in kT; however, when kT and kG are similar in value, errors in kG can inﬂuence kTg (see further discussion below). For chemicals with relatively slow elimination rate constants and high biomagniﬁcation potential (e.g., PCBs), errors in kT (or kTg) can lead to errors in BMFK (or BMFKg). If kT (or kTg) is overestimated (i.e., elimination is actually slower than estimated) then BMFK (or BMFKg) will be underestimated. Uncertainty in the AE and BMFK calculations is also a function of the uncertainty of the input parameters speciﬁc to the test. Uncertainty analysis could address uncertainty in bioaccumulation metrics if estimates of error for the parameters used in the calculations are available.68 The format for presenting the data review is as follows. First, the rationale for including the data assessment categories and assessment criteria is made by referring to the OECD guidance59 or by providing a brief explanatory statement. This is followed by some description of the parameters reported in the database and how the data are distributed into the data quality scores described below (e.g., summary statistics) and some discussion on the test category topic. Klimisch et al.63 suggested four classiﬁcations for assessing data reliability and these general principles are considered here. According to Klimisch et al.,63 data are classiﬁed as “reliable without restriction” if the study complies with nationally or internationally accepted guidelines or is generally conducted following these methods and key test information is described in suﬃcient detail. Data generated from studies that meet the basic scientiﬁc principles of the guidelines and include adequate documentation but lack key details or do not suﬃciently comply with general guideline requirements are classiﬁed as “reliable with restrictions”. Data are classiﬁed as “not reliable” when key methodological considerations are clearly omitted or signiﬁcantly diﬀer from the guidelines or when key documentation or data evaluation reveals unacceptable performance of the test or data analysis. The database does not include any entries that only reported end points,63 and good laboratory practice (GLP)69 was not considered here because the vast majority of the data are not GLP certiﬁed.63 The overarching objectives of this data quality assessment are to develop some criteria, evaluate the existing literature, and identify data quality issues and signiﬁcant deviations from current testing guidance. The data quality assignments are subject to error and misinterpretation due to incomplete and uncertain information and, by necessity, require professional judgment. While we maintain the general principles of the “Klimisch approach” we use the classiﬁcations (scores) High (H), Medium (M), and Low (L) conﬁdence to provide relative screening indicators of perceived data quality and consistency with OECD guideline principles, that is, H > M > L. Some data screened L may actually be less uncertain than some data screened H or M, or vice versa. Data points screened as L may still be useful for certain purposes with due consideration of the issues identiﬁed in this review. Table 1 summarizes general requirements for test data to be categorized H and the database (Supporting Information) includes a table detailing the scoring methods. Chemical Information. The guiding principle for identifying data with high quality chemical information is that chemical speciﬁc analyses provide test results to a deﬁnable chemical structure. A commonly used representation for a chemical structure is a SMILES notation and is the criterion used here.

documented in the database. For example, the database includes ﬁsh mass and lipid content, feeding rate, growth rate, water temperature, diet (or dose) concentration, lipid and protein content of diet, and tissue analyzed. The database also includes chemical identiﬁcation information (e.g., chemical name, chemical abstract service registration number (CAS#), and Simpliﬁed Molecular-Input Line-Entry System (SMILES)64 notation), and physical-chemical properties (e.g., molar mass, KOW obtained from65 with measured chemical properties selected preferentially over predictions, unless otherwise noted66,67). The database, including other test details and primary source references, is presented in the SI in a Microsoft Excel spreadsheet. The database contains 3032 measured test end points for 477 chemicals (based on CAS#). There are 964 half-lives, 1199 AEs and 869 BMFs (632 are BMFLg). There are 1512 study entries and some studies have more than one measurement end point from the same test for the same chemical. For example, there are 642 study entries with values for t1/2, AE and BMF. The data are measured from 19 ﬁsh species; the majority of entries are from trout species (56%) and common carp (19%). The chemicals range in molar mass from 120 to 1423 g·mol−1 (median = 326) and log KOW from 0.8 to 14.3 (median = 6.6). Fifty percent of the data entries are for polychlorinated biphenyls (PCBs), approximately 15% are other legacy organochlorine pollutants (e.g., hexachlorobenzene, dioxins, furans, mirex), 13% are hydrocarbons, and about 10% are brominated ﬂame retardants, predominantly polybrominated diphenyl ethers (PBDEs). Approximately 10% of the chemicals in the database are ionogenic. Data Assessment Methods. The database includes critical review of the 3032 test end points following the data quality assessment criteria outlined in categories of test information below. Previous methods for assessing ecotoxicology63 and bioaccumulation24,25 data and the OECD guidelines for dietary bioaccumulation testing59 form the basis for developing the data quality assessment methods. The guidelines59 were developed from protocols implemented and reﬁned over the past few decades. Because few studies to-date followed the guidelines, only some of the fundamental principles were considered. The methods evolved through the literature review. Categories of test information selected for review were based on OECD guidelines,59 the degree to which testing parameters were reported in the literature, some sensitivity analysis for guideline calculations, and professional judgment. This review cannot cover all aspects of the test and data analysis. A simple, generic sensitivity analysis (see Supporting Information SI-2 for details) was used to evaluate the signiﬁcance of potential errors in typical AE and BMFK calculations (Supporting Information eqs S4 and S6) outlined in OECD guidance59 (Figure 1D). The analysis was for three hypothetical chemicals representing relatively slow, moderate and fast chemical elimination (kT); t1/2s are 100, 10, and 1 day, respectively. For all three chemicals, the AE and BMFK calculations are very sensitive to changes in CD and CF0. The AE calculation is sensitive to I but BMFK is not because I is in both the numerator and denominator. For relatively persistent chemicals (typically slow elimination), AE is relatively insensitive to changes in kT while BMFK is quite sensitive to changes in kT. For chemicals with relatively fast elimination, AE is sensitive to changes in kT and BMFK is insensitive to changes in kT. For chemicals with moderate elimination rates, AE and BMFK are moderately sensitive to changes in kT. C

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

SMILES: Simpliﬁed Molecular-Input Line-Entry System; kT: depuration rate constant; kTg: growth corrected depuration rate constant; CD: chemical concentration in diet; I: dietary ingestion rate; bw: body weight; AE: chemical absorption eﬃciency; GIT: gastrointestinal tract; BMFSS: steady-state biomagniﬁcation factor; BMFKg: growth corrected kinetic biomagniﬁcation factor; BMFK: kinetic biomagniﬁcation factor.

measured and reported test temperature that is tolerable for the test species species documented and body size and lipid content measured and reported

Chemical purity and radiochemical purity were not considered and we assumed that all of the published analytical methods are appropriate. If only the total radioactive residues are quantiﬁed and the chemical is subject to biotransformation it is possible that metrics include quantiﬁcation of the parent chemical and the metabolite thus confounding the value for the parent chemical.24,25,70 Parent chemical speciﬁc analyses are required for H classiﬁcation. Radiolabeled chemicals were used for 3% of the study entries. Most tests using radiolabeled chemicals that were not subject to parent chemical speciﬁc analyses are considered L. For some chemicals (e.g., PCBs), biotransformation rates are generally very slow; therefore, tests using total radioactivity analysis for these chemicals are considered M if evidence suggested negligible biotransformation occurred. Chemical Mixtures. Test guidance59 states that more than one chemical can be tested in a single test provided certain criteria are fulﬁlled. The ring test for the OECD dietary guidelines71 included simultaneous exposures to ﬁve chemicals with log KOWs ranging from 4.5 to 6.1 (Supporting Information Table S2). Some tests have been conducted for a single chemical but most tests have been conducted for mixtures of chemicals and/or congeners, for example, PCBs, 51,54,72 PBDEs,73 and mixtures of congeners and isomers (e.g., chlorinated paraﬃns),52,53,55 and mixtures of stereoisomers (e.g., hexabromocyclodecane, HBCD).74 Some evidence suggests that multiple chemicals can be tested at the same time (coexposure) without substantial diﬀerences in the results16,17,71 but this is not always the case. Tests using chemical mixtures were scored H as long as there is no reported evidence of cometabolism, enzyme saturation, enzyme inhibition, or eﬀects of bioformation or isomerization or toxicity on the results. Co-dosing PBDE congeners can result in potential misinterpretation of bioaccumulation parameters because of debromination.73,75−78 When ﬁsh are coexposed to certain mixtures of BDE congeners, the bioaccumulation parameters for a less brominated BDE congener can include quantiﬁcation of the less brominated congener and debrominated metabolites from certain, more brominated, congeners. The issues of isomerization and chirality are not addressed in current guidance59 but, when applicable, require some consideration when testing because of isomer speciﬁc biotransformation rates and reported diﬀerences in data for racemic mixtures.57,79 When there is evidence of bioformation or bioisomerization the data are initially scored L. If methods are applied to address bioformation or bioisomerization, the data are scored M. An exception is that possible diﬀerences in PCB data as a result of chirality (8 congeners) are recognized,57 but not considered here. Biological Information. Test guidelines recommend starting the experiment with ﬁsh of similar age, body size and lipid content.59 Body size8,41 and lipid content24,37,80 can inﬂuence test metrics; therefore, it is important to measure and report these parameters. Study entries that include measurements for test organism mass and lipid contents are categorized H (1479/ 1512 for mass, 98%; 1332/1512 for lipid content, 88%). Reported initial ﬁsh mass range from 0.1 to 1000 g (median = 10 g) and ﬁnal reported ﬁsh mass range from 3.8 g to 3500 g (median = 370 g). Mean “whole body” lipid contents range from 1.0% to 9.9% (median = 5.3%) and analyzed tissue lipid contents range from 0.43% (muscle) to 15.4% (not speciﬁed). Study entries that do not report mass or body size or lipid content are L (2% for mass; 12% for lipid content). In rare

a

growth correction documented for log KOW > 5; no identiﬁed source of error in the methods or calculations fed clean (uncontaminated) food while depurating in clean water in ﬂow through systems with no detectable background contamination (exposure and depuration) no reported evidence of bioformation or isomerization, or toxicity that inﬂuence the results

approach to steady-state conﬁrmed for BMFSS and high quality kTg (or kT) and AE for BMFKg (or BMFK)

data analysis

whole body (or whole body minus GIT or whole body minus GIT and liver) analyzed and r2 ≥ 0.7 for kT or kTg

experimental conditions

fed commercial feed (lipid content reported) for ≥7 day (exposure), I at 0.5−3% bw·d−1, CD measured for BMF and AE

chemical and biological information

chemical speciﬁc analysis for a deﬁnable structure (SMILES)

Table 1. General Criteria to Classify Data As High (H) Quality (See Text and Supporting Information for more Details for This Category and Others)a

Environmental Science & Technology

D

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

Environmental Science & Technology

1000 to 1 000 000 ng·g−1).59 Chemical concentrations should be measured in the food to conﬁrm the exposure doses and to reliably calculate BMF and AE (see above and Supporting Information SI-2). For the BMF and AE end points, studies had to clearly report that CD was measured to be categorized H. Food concentrations were measured for 91% of the study entries (H for BMF and AE end points). Food concentrations ranged from 0.04 to 5 000 000 ng·g−1 (median = 83 ng·g−1) on a dry weight (dw) basis. Measuring or reporting CD was not considered a data quality issue for half-life determination as long as the dose did not cause toxic eﬀects and was not expected to exceed its solubility in the diet (see Supporting Information SI-3). Doses were assumed below levels that could cause saturation of biotransformation systems because current data for enzyme kinetics in ﬁsh are quite limited.92 Finalized guidance59 suggests a feeding rate (I) of 1−2% body weight per day (bw·d−1) to minimize rates of change in somatic or lipid tissues. Guidance for correcting the mean feeding rate for growth during the test is also available.59 Results from slightly diﬀerent feeding rates conducted during the OECD ring trial were inconclusive of possible diﬀerences in I on the BMFL;71 however, some studies have shown the inﬂuence of diﬀerent feeding rates on the BMF and AE.48 Reported daily (continuous) feeding rates range from 0.4 to 4.2% body weight (median = 1.5% bw·d−1). Tests with exposure durations >1 day and with I between 0.5 and 3% bw·d−1 are assumed H (n = 1154) and rates > or < this range are M (n = 259). Studies that only administered a single dose of the test chemical (or ≤1 day exposure) had feeding rates ranging from 0.5 to 12.4% body weight (median = 5.5%). If feeding rates were not reported or were reported to ﬂuctuate, the study entries are L for BMF and AE calculations (n = 94). Since the calculations assume feeding rates are the same for all ﬁsh in the tank, it is necessary to closely monitor the ﬁsh during feeding (see above and Supporting Information SI-2).59 This assumption is not easily conﬁrmed or testable and is an expected source of variability in the test results. Experimental Design. Flow through conditions are recommended and semistatic methods can also be used with due consideration.59 A critical assumption for high quality kT or kTg, and AE and BMF data is that the ﬁsh are not exposed to test chemical in water during the uptake phase (resulting in bioconcentration) and that they are not exposed to any test chemical during the depuration phase. One method to minimize exposure from the water during the depuration phase is to use ﬂow through tanks and to ensure that all uneaten food is removed immediately after feeding. If the ﬁsh are in ﬂow through tanks during the depuration phase the study entry is H (∼85%). If the text described a ﬁltered system, or it was inferred to be a ﬂow through or semistatic design the entries are M (∼10%). If the depuration phase used a static test design or when the depuration conditions could not be determined the data entries are L (∼5%). Background Contamination. Detectable levels of test chemical in control ﬁsh, depuration food or depuration water are reported for 44% of the study entries and the issue of background contamination is a source of uncertainty. There are practical limitations to ensuring clean diet and water during the depuration phase, largely determined by the chemical-speciﬁc detection limit. Background contamination in water, ﬁsh, and diet is particularly challenging for ubiquitous contaminants (e.g., PCBs). For persistent hydrophobic chemicals, even a small quantity of chemical in the water can contribute to the

cases (4/1512) lipid contents are measured and used in the calculations but not documented (scored M). Diﬀerences in ﬁsh species, age and sex may result in diﬀerences in chemical bioaccumulation.81,82 Guidance lists a variety of species and sizes recommended for testing59 but does not include speciﬁcations for age and sex. The sex of the ﬁsh is rarely reported and it is implied ﬁsh are not reproducing during a standard test. Species type, life-stage, and sex are not addressed here. Guidance recommends that the methods used for measuring lipid contents be documented59 because diﬀerences are recognized;83 however, potential diﬀerences in results due to diﬀerences in lipid analysis methods were not considered. In summary, study entries with high quality biological information include documented species with some reported measurement of body size and lipid content. Although not addressed in the present review due to a general lack of reporting, excessive mortality (e.g., > 10%) and reduced growth rates in dosed groups are other data quality parameters that could be considered in future studies. Water Conditions. Guidance provides a list of suggested temperature ranges for the listed test species and the water temperature must remain relatively constant (±2 °C) and the dissolved oxygen concentration must be ≥60% saturation. Diﬀerences in water temperature can result in diﬀerent halflives for the same chemical in the same ﬁsh species41,84 due to biological factors such as changes in respiration85,86 and chemical factors such as changes in partitioning properties.87−89 Experimental temperatures ranged from 4 to 26 °C (median = 15 °C). Fish cannot survive or can be stressed outside of certain temperature ranges. While some tests were conducted at temperatures outside of the range of recommended temperatures for that species,59 no exposure temperatures were deemed unsuitable. Consistency in temperature throughout the test and dissolved oxygen concentration were not considered for data quality assessment; however, studies that did not report the exposure temperature are scored L (n = 40). Fifty-ﬁve percent of the study entries reported dissolved oxygen concentrations with values ranging from 7.2 to 10.9 mg·L−1 (median = 9.7 mg·L−1); all documented measurements were >7 mg·L−1 or >70% saturation. When reported, values for pH ranged from 7 to 8.5 (median = 7.1). Organic carbon content, ionic strength, and salinity may be important for some tests but were not considered here. Diet Type and Lipid Content. Guidelines59 recommend using dosed commercial pelleted ﬁsh food for consistency in the test and documentation of lipid content (and protein) in the food. Study entries using commercial ﬁsh food (e.g., pellets) are considered H (1232/1512) and nonstandardized diets (e.g., gelatin capsules, live feed) are M (278/1512). Studies that did not document the dosing matrix are L (n = 2). When reported, the lipid content of the diet ranged from 0.2 to 24% with a median value of 14%. The literature is somewhat conﬂicted as to whether or not dietary lipid content inﬂuences half-lives, BMFs and AE.29,30,48,49,90,91 Besides inﬂuencing growth rates, dietary lipid content may also aﬀect half-lives as a result of enhanced or reduced fecal elimination and is required to calculate BMFL. If lipid content in the diet was reported, the study entry is scored H (70%), if not reported the entry is M (30%). Food Concentrations and Feeding Rate (Dose). The administered dose is a product of CD (assumed homogeneous throughout the supplied food) and feeding rate. Guidance for the chemical concentration in the diet is somewhat vague (i.e., E

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

Environmental Science & Technology

Figure 2. (A) An illustration of model predicted half-lives as a function of the octanol−water partition coeﬃcient (KOW) and diﬀerent methods for growth correction. (B) The distribution of measured half-lives in the database as a function of data quality scores (H: High; M: Medium; L: Low).

ﬁsh body burden due to high bioconcentration. If there is no detectable background contamination the data are H. If studies report evidence of exposure to the chemical during the depuration phase either in the diet or in the water the entry is initially classiﬁed L. For example, a dietary testing study involving 13 pesticides (ranging in log KOW 1.5−5)93 reported measured concentrations of the test chemical in the water during the uptake phase and during the depuration phase. Studies that made eﬀorts to correct for background contamination are considered M because the results are still relatively uncertain. Cannibalism during the depuration phase has been reported40 and such data are scored L. When tests did not measure or report measurements for background levels the data are tentatively assumed H. Test Duration and Sampling Design. Guidance59 recommends at least 7−14 days of dietary exposure. Exposure durations reported in the literature range from 1 to 286 days (median = 21 days). Twenty-four percent of the study entries are 1 day exposures. Depuration durations reported in the database range from 0 to 300 days (median = 25 days). When possible, the ratio of the depuration time to the half-life was calculated (n = 943) and these ratios range from 0.035 to 250 (median = 1.92). Lower ratios may contribute error in half-life estimation and approximately 25% of the data have ratios < 1 (i.e., the depuration period was 4. Depuration sampling should include several time points (i.e., n ≥ 3). Guidelines recommend 5−10 ﬁsh sampled per time point;59 however, the number of ﬁsh sampled is not considered here. Tissue Analysis. Guidelines59 recommend using the whole body for chemical analysis but speciﬁc tissues can be analyzed with some considerations. Tests that analyze whole body (or carcass) are considered H. Tests that analyze “whole body minus the gastrointestinal tract (GIT)”, “whole body minus the GIT and liver”, or some variation (i.e., large majority of the body analyzed) are tentatively considered H. The implications of this classiﬁcation assumption could be further evaluated. These “whole body” tests account for 94% of the study entries. If tissue speciﬁc data (muscle, liver, kidney, etc.) are used for analysis they should be interpreted carefully because of diﬀerences that may occur in internal distribution. If tissue lipid contents were reported for neutral organic chemicals so that the tissue data could be lipid corrected the data are scored M. For certain ionizable organics, such as perﬂuorinated chemicals, traditional lipid correction may not be appropriate94−97 and bioaccumulation data from tissue speciﬁc analyses

without lipid correction for these chemicals are currently considered M. Growth Rate Constants and Growth Correction. Fish growth rates are aﬀected by biological factors (e.g., species, lifestage) and test conditions (e.g., temperature, feeding rates), and vary among individuals in test and environmental populations.28,98 Guidance recommends growth correction or the body burden approach59 to mitigate this variability so that chemicals can be evaluated in a consistent manner. Forty three percent of the total study entries (including AE and BMF studies that did not measure depuration) do not include a measured kG and 4% of the total entries report growth as “not signiﬁcant”. Reported kG ranges from −0.0014 to 0.125 d−1 (median = 0.0126 d−1). The median growth rate (1.3% bw·d−1) and the median feeding rate (1.5% bw·d−1) highlight the relative eﬃciency of biomass production for ﬁsh in laboratory settings. A kinetic bioaccumulation model is parametrized with median values from the database to illustrate the inﬂuence of growth dilution on half-lives determined under typical laboratory conditions (see Supporting Information SI-4 for further details). Figure 2A shows diﬀerent model simulations for half-lives (kB = 0 for all). The black line shows the t1/2s for chemicals when the model is parametrized using median database values, i.e., kG = 0.013 d−1 (no growth correction). The green line shows the expected relationship for t1/2gs using growth correction or the body burden approach, i.e., kG = 0 (growth correction). Diﬀerences between t1/2 and t1/2g become signiﬁcant when log KOW > ∼4.5−5. At log KOW ∼6 the t1/2 is about 40 d and t1/2g is about 130 days (factor 3.3 diﬀerence). At log KOW ∼8 the t1/2 is about 50 days and t1/2g is about 1200 days (>factor 20 diﬀerence). When the t1/2s are not growth corrected (black line) the values are similar at log KOW ∼6 and ∼8 because kG is controlling kT (see Supporting Information Figure S1 and related discussion). Chemicals with log KOW > 5 require growth correction to be considered H. Chemicals with log KOW > 5 with no apparent growth correction are considered L. Guidance59 suggests using a linear least-squares correlation calculated from ln(ﬁsh weight) against time to calculate kG and a single kG is often applied for growth correction. Because individual ﬁsh grow at diﬀerent rates in a test population there will be uncertainty in applying a single growth rate to the test population. Variability and uncertainty in k G can be considerable even in the same test, for example, coeﬃcient of determination (r2) of 0.5 and 0.6 for growth rate model ﬁts.57 There are diﬀerent growth correction methods and models F

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

Environmental Science & Technology

through test systems at tolerable temperatures for the test species and there are no identiﬁed signiﬁcant sources of error in the methods or calculations. Absorption Eﬃciency. Absorption eﬃciency data are summarized and evaluated separately depending on whether or not the exposure period met the OECD recommendation of at least 7 days.59 Shorter-term exposure duration tests deviate from standardized guidance, but AE can be calculated during the linear uptake phase between days 1 and 3 of exposure.59 A generally applicable screening criterion is that AEs > 1.0 (100%) are suspect indicating a source of error and are thus classiﬁed L. Plausible explanations for AE > 1.0 are the (bio)formation of the chemical in the organism (or perhaps the GIT) or additional unintentional exposure during the test or analytical errors. In some cases AE values >1.0 are reported for certain chemicals for which bioformation is not expected to occur102 and there is no obvious explanation. Measured doses (quantities for CD and I) are required for H classiﬁcation for most AE data (76% of the entries). When the methods for calculating AE were not reported the entries are assumed L (3% of entries). Based on the sensitivity analysis described above, AEs calculated from L kT (or kTg) are tentatively classiﬁed M. Supporting Information Figure S3A shows 436 AEs (exposure durations 100%. Many of the AE data scored M do not use pelleted food and are inconsistent with guidance;59 however, these data can still provide useful information for chemical uptake from dietary exposures. For example, there is a cluster of high AEs (0.8−1.0) categorized M. The ﬁsh were fed live food (e.g., carp fed mayﬂies) at 5.5% bw for 1 day72 suggesting that either live food or higher feeding rates can result in relatively higher shortterm exposure AEs compared to other exposure regimes. Supporting Information Figure S3B shows 754 AEs from exposure durations ≥7 days ranging from 0 to 4.65 with a median of 0.45 (H = 225, M = 280, L = 287). The median, average and standard deviation for the H and M data are 0.42, 0.40, and 0.22 respectively. Exposure durations range from 8 to 247 days. Dietary types are also variable, but most (93%) are derived using some form of pelleted commercial ﬁsh food (following standardized guidelines). About 2.2% of the longerterm exposure AEs are >100% and categorized L. In one study, depuration data from a diﬀerent experiment were used to derive AE and BMFs37 and are scored L. According to guidance,59 AE can be calculated using Supporting Information eq S4 and kT with wet weight (ww) concentrations and feeding rates (Figure 1) or at the linear uptake phase of exposure (e.g., tissues analyzed between days 1 and 3). Another method in the guidelines for calculating AE (and BMFK) and often reported in the literature is to use kTg, lipid corrected feeding rates (I), lipid corrected ﬁsh concentrations and iterative nonlinear regression.50−57 Examples of variability in AE for PCB 64 (ranging from 0.48 to 1.01) 5 0 , 5 1 , 7 2 and PBDE 99 (ranging from 0 to

reported in the literature and there is an apparent growth correction error in some studies50−57,99−101 (see Supporting Information SI-5 for further details). When simulating this error with the model using median values from the database (red line in Figure 2A) the half-life diﬀerences are 6, the halflife diﬀerences are >2-fold. For chemicals with log KOW > 6 documented with this apparent growth correction error, the kTg and t1/2g are tentatively screened L to indicate this potential source of uncertainty. For persistent hydrophobic chemicals, changes in lipid content during the test may contribute more uncertainty in the results than somatic growth; an issue not considered here because lipid growth is not commonly reported. Some studies for these types of chemicals used lipid corrected concentrations and feeding rates for the bioaccumulation metric calculations50−57 thus addressing the issue of changing lipid content. The use of lipid corrected concentrations may mitigate the potential growth correction error. Further data analysis may address these uncertainties. Depuration Rate Constants and Half-Lives. Figure 2B shows the distribution of the 964 t1/2 data (growth corrected and not growth corrected) as a function of the data quality evaluation (H = 231, M = 217, L = 516). Half-lives range from 0.13 to 2600 days (median = 41 days) corresponding to depuration rate constants ranging from 5.3 to 0.00027 d−1 (median = 0.017 d−1). Two negative t1/2s and 14 t1/2s reported as “> 1000 days” are scored L. Eighty six percent of the 964 t1/2s are reported on a growth corrected basis. The apparent growth correction error (see Supporting Information SI-5) occurs in approximately 45% of the t1/2g entries and corresponds with the high frequency count of chemicals with longer half-lives classiﬁed L (Figure 2B). Guidance59 includes methods for calculating the bioaccumulation metrics and it suggests the data be examined for outliers and that standard statistical procedures be conducted to examine and present the results; however, there are no recommendations for the reliability of kT, t1/2, kTg, or t1/2g calculations. If statistics from the test are available, the propagation of error (uncertainty analysis) in key bioaccumulation metrics could be calculated. Relative standard errors for half-lives range from 0.1% to 100% with an average of about 20%; however, standard errors have not been regularly reported (n = 382). The r2 associated with depuration or half-life data is reported more frequently (n = 754). In the depuration regression, r2 reﬂects the variance described by time. A high r2 for kT from multiple time points implies (but does not conﬁrm) that depuration may follow ﬁrst-order kinetics (often assumed) and is reasonably well described by time. A low r2 indicates (but does not conﬁrm) that time may not be the primary factor and that non-ﬁrst-order kinetics, or some source of variability or uncertainty may be confounding factors (e.g., background contamination, ﬂuctuating internal distribution kinetics). Based on review of general relationships between relative standard errors and r2 and data availability, r2 was tentatively selected to identify possible data quality issues as r2 ≥ 0.7 (H = 46%); 0.5 ≤ r2 < 0.7 (M = 12%); r2 < 0.5 (L = 15%). Studies that do not report r2 are assumed M (27%). More rigorous evaluation methods should be considered if more data are available for analysis. In general, high quality kT, t1/2, kTg and t1/2g data are derived from experiments in which ﬁsh were regularly (daily) fed clean, commercial ﬁsh food while depurating in clean water in ﬂow G

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

Environmental Science & Technology

Figure 3. Measured and modeled biomagniﬁcation factors (BMFs) as a function of the octanol−water partition coeﬃcient (KOW). (A) Measured BMFs categorized to be of High (H) and Medium (M) quality, (B) measured BMFs categorized to be Low quality (L), and (C) all measured BMFs for polychlorinated biphenyls (PCBs) and hexachlorobenzene (HCB). Modeled BMFs are lipid corrected based on representative (not study speciﬁc) conditions to illustrate (D) the inﬂuence of the biotransformation rate constant (kB,N; d−1) and growth rate constant (kG; d−1) on the BMFs. The legend for model simulations in (A) also applies to (B) and (C).

0.60)73,75,78,101,103 are described in Supporting Information SI-6 with additional discussion. Barber58 also reported substantial variability and uncertainty in measured AE estimates and models. The database includes chemicals with diﬀerent biotransformation rates that likely contribute to variability in AE. Due to the variability in methods such as dosing (pellets, live food, gavage, etc.), exposure duration (1−247 days) and calculations, it is challenging to determine speciﬁc diﬀerences and controlling factors for AE from the overall data set. Test designs based on guidelines59 that focus on clearly deﬁned manipulations in exposure (treatments) are best suited for improved understanding in AE to better ascertain the potential inﬂuence of diet type,72 dietary lipid content,49,72,90 feeding rate,48,49,71,72 chemical concentration,48,51,54 species,73,75,103 exposure duration, and chemical structure. BMF Data. Figures 3A (H = 186 and M = 131) and 3B (L = 552) show 869 BMFs (ww or lw) reported with or without growth correction. Thirty-nine percent of the BMFs in Figure 3A and B are for PCBs. Only PCB and HCB BMFs (most are BMFLg) are plotted in Figure 3C, regardless of their data quality score. Thirty BMFKLgs for HCB scored H range from 1.1 to 4.8 (median = 1.8) indicating a plausible range for test variability and a positive control chemical for BMF tests. The BMFs range from 2 × 10−6 to 89 with a median of 1.3. Twenty three values reported as “” are scored L (not shown). Approximately 63% of the BMFs are categorized L. Two hundred twenty BMFs are calculated as concentration ratios (i.e., CF/CD) and thus to be scored H require conﬁrmation that steady-state was approached; however, 60% of the CF/CD BMFs either do not conﬁrm steady-state or report that

steady-state was not approximated and are scored L. BMFK calculations that used M kT or kTg and M AE data are categorized M. BMFK calculations are scored L if the kT or kTg or AE data are considered L. Many of the BMFKs are scored L because of potential errors in kTg. For neutral organic chemicals, BMFLs can indicate an increase or decrease in chemical potential (or fugacity) between the diet and the consumer, i.e. BMF L > 1 or < 1 respectively.22,23,28 Approximately 80% are reported as BMFL or can be calculated using reported lipid contents. When only a ww BMF can be calculated from otherwise H quality test data, the BMF entry is considered M because lipid content in the ﬁsh and the food can inﬂuence the BMF and lipid correction is suggested.59 Commercial pelleted ﬁsh food is typically about 13−15% lipid and most laboratory ﬁsh are about 3−8% lipid, thus under these conditions BMFLs are often about 2−5 times higher than the ww BMFs. Some BMFs are scored L because the calculations did not follow standard methods or there are identiﬁable errors or uncertainties. The units and calculations for the BMF are not reported for the vast majority but they can often be inferred for BMFKs if the units for I are reported (Supporting Information eqs S6 and S7); BMF units are kg-lipid·kg-lipid−1 if I is lw-food· lw-ﬁsh−1·d−1.56 In some cases I is based on ﬁsh-ww and foodlw;73 therefore, these nonstandardized BMF units of kg-lw·kgww−1 are L. Maximum theoretical lipid and growth corrected BMFLgs in ﬁsh are about 8−1530,31,106 and higher measured values suggest a potential error. For example, BMFs > 20 and as high as 46 were reported for some PBDE congeners.73 These reported BMFs appear to be the result of bioformation (debromination) from coexposure to higher brominated H

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

Environmental Science & Technology

log KOW 9.0; BMFKLg = 1.4). Binox M is ionogenic (acid pKa = 12); however, it exists almost entirely in the neutral form at environmental and physiological pH. The laboratory data suggest that chemicals with a log KOW ≥ ∼5 and at least as high as ∼9 have biomagniﬁcation potential in ﬁsh. This relationship is supported by the total data set for persistent chemicals in Figure 3C (measured and estimated log KOWs 5.0 to 9.1 with BMFLg > 1). There are also many chemicals within and beyond the log KOW range of 5−9 that have BMFs < 1. A kinetic bioaccumulation model (Supporting Information SI-4) was used to calculate BMFs at steady-state. The model is parametrized with median database values (i.e., not test speciﬁc conditions) and kB and kG are initially set to zero. The model predicted BMFLgs (black line Figure 3) are generally in good agreement with the H and M measured data with respect to log KOW ≥ ∼5 and BMFLg > 1. The model reveals that chemicals with log KOW < ∼5 do not biomagnify in ﬁsh from dietary exposures alone (under laboratory conditions) because chemical elimination rates to water (i.e., k2) are suﬃciently fast to counter-act gastrointestinal biomagniﬁcation (Supporting Information SI-4). The maximum modeled BMFLg is ∼10 and is in good agreement with most measured maximum H or M BMFLgs. If kG and kB are zero and the chemical is absorbed from the GIT, the model suggests there is no decline in BMFLg at very high KOW (i.e., log KOW > 8). The model BMFLg predictions are supported by theoretical maximum BMFs for ﬁsh;30,31,106 however, the time to approach steady-state for such chemicals is long (years) and may not be approximated within the organism’s lifetime. There is uncertainty in the model and in the reliability of high values of KOW, and it may not be possible to reliably determine very slow growth rates. The model is used to show the inﬂuence of growth dilution on the BMFL (see also Supporting Information Figure S5). At median laboratory growth rates (kG = 0.013 d−1; kB = 0) chemicals with a range of log KOW from about 5.6−7.3 will have BMFs > 1 and the maximum BMFL is about 1.2. When kG is lowered 100 fold, that is, ∼0.00013 d−1, chemicals with a range of log KOW from about 5.1−9.8 will have BMFs > 1 and the maximum BMFL is about 8.4. Measured and modeled ﬁeld BMFL and TMF data from aquatic food webs, although variable, are generally > 1 for PCBs and other persistent chemicals with log KOW ranging from about 5 to at least 8.7.22,109−112 Without growth correction, some chemicals that biomagnify in the environment may not have BMFLs > 1 in the test. This supports guidance59 recommendations for the growth correction of laboratory BMFs. The model calculations highlight the sensitivity of the BMF to kG and the importance of accurate growth correction methods for persistent hydrophobic chemicals. Model simulations are also provided to illustrate how changes in kB can reduce BMFs. Figure 3D shows that a kB,N (biotransformation rate constant normalized to a 10 g ﬁsh at 15 °C81) of ∼0.05 d−1 is required for laboratory BMFL < 1 independent of KOW and kG. This kB,N corresponds to a biotransformation half-life of about 14 days. When the BMFLg is not at a maximum value, kB does not have to be as fast for BMF < 1 (e.g., if AE is reduced or k2 plays a role). The kB,N of 0.05 d−1 is speciﬁc to the other parameters used in the simulation (e.g., lipid contents, body size, feeding rates, default AE, dietary exposure only); therefore, if these values change, the kB,N associated with the BMF threshold of 1 will also change.

congeners and other errors (e.g., nonstandardized BMF units73) and are scored L. Potential errors in reported BMFs can be identiﬁed through corroboration of the calculations using test data, if available. For example, if AE, I, lipid contents and kT (or kTg) are provided they can be used to recalculate BMFKL (or BMFKLg) using Supporting Information eq S6 (or Supporting Information eq S7). If the recalculated BMF does not equal or approximate the reported BMF this indicates a potential source of error. Some inconsistency in a recalculated BMF may exist because of diﬀerences in how the data were treated, for example, if the BMF is derived from lipid corrected ﬁsh and diet concentrations measured throughout the test rather than using the average values from the test. When data were available BMFs were recalculated and compared to the reported BMFs to ﬂag discrepancies. We scored ≥ 50% diﬀerence between recalculated BMFs and reported BMFs as L and ≥ 20% and 1 have been measured for chemicals with log KOW 4.5−9.0. Only one of the 12 BMFsKLg for the log KOW 4.5 chemical (musk-xylene)16,71 is >1. The median BMFKLg for musk-xylene is 0.43 indicating some potential uncertainty in the musk-xylene BMF > 1. The second lowest KOW chemical with a measured BMFLg > 1 considered H or M is PCB 4 (log KOW 5.0; BMFLg = 4.6). The highest log KOW chemical with a BMFLg > 1 considered H or M is 4,4′methylenebis[2,6-bis(2-methyl-2-propanyl)phenol] (Binox M: I

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

■

■

RECOMMENDATIONS The past 40 years of dietary testing data reﬂect signiﬁcant research to improve bioaccumulation science. Data reporting and analyses have been quite variable and standardized test guidance59 should foster the development of more consistent data sets. Experimentalists and data users should review the testing guidance,59 related supporting studies,71 current theoretical models30,31,58,106 and consider the data quality issues raised herein. The database and proposed data quality assessment methods are expected to evolve. Guidance, critical analysis and GLP provide a means to address uncertainty but uncertainty cannot be eliminated, for example, OECD 305 ring test.71 Guidance includes criteria for “valid” experimental conditions but further guidance to address uncertainty in the test calculations would improve data interpretation and use. Uncertainty analysis could be considered as warranted, if the required data are documented. Authors and journals are encouraged to publish more details of the experiments as Supporting Information (archived online)113 to improve data interpretation and scientiﬁc understanding, to reduce animal testing and expense, and to reduce potential error in decision-making (e.g., peer-reviewed publications are often used for bioaccumulation assessment). The OECD 305 guidance59 provides a point of departure for the careful manipulation of speciﬁc (independent) test variables (e.g., feeding rates, food type, exposure duration) to gain further insight into underlying sources of variability in bioaccumulation mechanisms and data. Tests that seek improvements in mechanistic knowledge should use persistent chemicals (e.g., PCBs) to reduce the confounding issue of biotransformation in data interpretation. The OECD dietary ring test71 included a relatively narrow and low range of log KOW (i.e., from 4.5 to 6.1). Future evaluations of the guideline methods should include higher log KOW chemicals (i.e., > 8). For persistent chemicals, data analyses with body burdens (XF) should be compared to results obtained using growth correction methods. Reference chemicals could be useful for benchmarking test data17,114 and the benchmark chemical(s) should have similar KOW to the test chemical(s). Chemicals are suggested as a reference for AE calculations59 and a positive control should be considered for the BMF calculations (i.e., HCB BMFLgs < 1 indicate a source of error). Chemicals subject to bioformation from coexposed chemicals or isomers subject to bioisomerization should be tested individually to obtain more reliable information. Testing chemical classes that are poorly represented in the database could provide more insight into the biotransformation capacity of ﬁsh and expand the applicability domain of current predictive models.115−118 Additional species tested with the same chemical(s) would provide further insights into interspecies variability for biomagniﬁcation and biotransformation potential. Evaluated models can reduce unnecessary animal testing. In vitro119−122 and in silico115−118 kB estimates can be used to parametrize BMF models to screen and prioritize chemicals for BMF testing and to aid test design (e.g., hypothesize suitable depuration durations). Finally, further evaluations of the laboratory testing data and models with monitoring data123 could better ascertain the ecological relevance and eﬀectiveness of the dietary testing methods and the data they provide for identifying and regulating bioaccumulative chemicals.

Critical Review

ASSOCIATED CONTENT

S Supporting Information *

Model description, additional results and discussion, and the evaluated database. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: 1-416-462-0482; fax: 1-416-462-0482; e-mail: jon@ arnotresearch.com. Present Address §

Golder Associates Ltd., Vancouver, British Columbia, Canada

Notes

The authors declare no competing ﬁnancial interest.

■

ACKNOWLEDGMENTS We acknowledge funding support from Environment Canada and ILSI-Health and Environmental Sciences (HESI). We thank HESI project monitoring team members (Mark Bonnell, Tom Parkerton, Lawrence Burkhard, Dan Merckel, Michelle Embry, Kent Woodburn, and Ken Drouillard) for comments on earlier drafts and Aaron Fisk and Andrea Buckman for providing datasets for analysis.

■

REFERENCES

(1) Hamelink, J. L.; Waybrant, R. C.; Ball, R. C. A proposal: Exchange equilibria control the degree chlorinated hydrocarbons are biologically magnified in lentic environments. Trans. Am. Fish. Soc. 1971, 100 (2), 207−214. (2) Chiou, C. T.; Freed, V. H.; Schmedding, D. W.; Kohnert, R. L. Partition coefficients and bioaccumulation of selected organic chemicals. Environ. Sci. Technol. 1977, 11 (5), 242−257. (3) Mackay, D. Correlation of bioconcentration factors. Environ. Sci. Technol. 1982, 16 (5), 274−278. (4) Bruggeman, W. A.; Opperhuizen, A.; Wijbenga, A.; Hutzinger, O. Bioaccumulation of super-lipophilic chemicals in fish. Toxicol. Environ. Chem. 1984, 7 (3), 173−189. (5) Gobas, F. A. P. C.; Muir, D. C. G.; Mackay, D. Dynamics of dietary bioaccumulation and faecal elimination of hydrophobic organic chemicals in fish. Chemosphere 1988, 17 (5), 943−962. (6) Thomann, R. V. Bioaccumulation model of organic chemical distribution in aquatic food chains. Environ. Sci. Technol. 1989, 23 (6), 699−707. (7) Gobas, F. A. P. C.; Zhang, X.; Wells, R. Gastrointestinal magnification: The mechanism of biomagnification and food chain accumulation of organic chemicals. Environ. Sci. Technol. 1993, 27 (13), 2855−2863. (8) Hendriks, A. J.; van der Linde, A.; Cornelissen, G.; Sijm, D. T. H. M. The power of size. 1. Rate constants and equilibrium ratios for accumulation of organic substances related to octanol-water partition ratio and species weight. Environ. Toxicol. Chem. 2001, 20 (7), 1399− 1420. (9) Burkhard, L. P.; Cook, P. M.; Mount, D. R. The relationship of bioaccumulative chemicals in water and sediment to residues in fish: A visualization approach. Environ. Toxicol. Chem. 2003, 22 (11), 2822− 2830. (10) Czub, G.; McLachlan, M. S. A food chain model to predict the levels of lipophilic organic contaminants in humans. Environ. Toxicol. Chem. 2004, 23 (10), 2356−2366. (11) Arnot, J. A.; Gobas, F. A. P. C. A food web bioaccumulation model for organic chemicals in aquatic ecosystems. Environ. Toxicol. Chem. 2004, 23 (10), 2343−2355. (12) Borgå, K.; Fisk, A. T.; Hargrave, B.; Hoekstra, P. F.; Swackhamer, D.; Muir, D. C. G. Bioaccumulation factors for PCBs revisited. Environ. Sci. Technol. 2005, 39 (12), 4523−4532. J

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

Environmental Science & Technology (13) Jonker, M. T. O.; van der Heijden, S. A. Bioconcentration factor hydrophobicity cutoff: An artificial phenomenon reconstructed. Environ. Sci. Technol. 2007, 41 (21), 7363−7369. (14) Nichols, J. W.; Bonnell, M.; Dimitrov, S. D.; Escher, B. I.; Han, X.; Kramer, N. I. Bioaccumulation assessment using predictive approaches. Integr. Environ. Assess. Manage. 2009, 5 (4), 577−597. (15) 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 use of a probabilistic modeling approach. Integr. Environ. Assess. Manage. 2012, 8 (1), 42−63. (16) Inoue, Y.; Hashizume, N.; Yoshida, T.; Murakami, H.; Suzuki, Y.; Koga, Y.; Takeshige, R.; Kikushima, E.; Yakata, N.; Otsuka, M. Comparison of bioconcentration and biomagnification factors for poorly water-soluble chemicals using common carp (Cyprinus carpio l.). Arch. Environ. Contam. Toxicol. 2012, 63 (2), 241−248. (17) Adolfsson-Erici, M.; Åkerman, G.; McLachlan, M. S. Measuring bioconcentration factors in fish using exposure to multiple chemicals and internal benchmarking to correct for growth dilution. Environ. Toxicol. Chem. 2012, 31 (8), 1853−1860. (18) U.S. EPA. Toxic Substances Control Act (1976); U.S. Environmental Protection Agency: Washington, DC, 1976. (19) Government of Canada. Canadian Environmental Protection Act, 1999. In Canada Gazette Part III, November 4, 1999. (20) UNEP. Final Act of the Conference of Plenipotentiaries on the Stockholm Convention on Persistent Organic Pollutants; United Nations Environment Program: Geneva, Switzerland, 2001; p 44. (21) European Parliament. Regulation (EC) 1907/2006 of the European Parliament and of the Council Of 18 December 2006 Concerning The Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH), Establishing a European Chemicals Agency, Amending Directive 1999/45/Ec and Repealing Council Regulation (Eec) 793/93 and Commission Regulation (EC) 1488/94 as well as Council Directive 76/769/EEC and Commission Directives 91/155/eec, 93/67/eec, 93/105/ec and 2000/21/ec. Oﬀ. J. Eur. Union, 30.12.2006. L 396/1-849, 2006. (22) 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. 2012, 8 (1), 17−31. (23) Gobas, F. A. P. C.; de Wolf, W.; Burkhard, L. P.; Verbruggen, E.; Plotzke, K. Revisiting bioaccumulation criteria for POPs and PBT assessments. Integr. Environ. Assess. Manage. 2009, 5 (4), 624−637. (24) Arnot, J. A.; Gobas, F. A. P. C. A review of bioconcentration factor (BCF) and bioaccumulation factor (BAF) assessments for organic chemicals in aquatic organisms. Environ. Rev. 2006, 14 (4), 257−297. (25) Parkerton, T. F.; Arnot, J. A.; Weisbrod, A. V.; Russom, C.; Hoke, R. A.; Woodburn, K.; Traas, T.; Bonnell, M.; Burkhard, L. P.; Lampi, M. A. Guidance for evaluating in-vivo fish bioaccumulation data. Integr. Environ. Assess. Manage. 2008, 4 (2), 139−155. (26) Thomann, R. V.; Connolly, J. P.; Parkerton, T. F. An equilibrium model of organic chemical accumulation in aquatic food webs with sediment interaction. Environ. Toxicol. Chem. 1992, 11 (5), 615−629. (27) Qiao, P.; Gobas, F. A. P. C.; Farrell, A. P. Relative contributions of aqueous and dietary uptake of hydrophobic chemicals to the body burden in juvenile rainbow trout. Arch. Environ. Contam. Toxicol. 2000, 39 (3), 369−377. (28) Connolly, J. P.; Pedersen, C. J. A thermodynamic-based evaluation of organic chemical accumulation in aquatic organisms. Environ. Sci. Technol. 1988, 22 (1), 99−103. (29) Gobas, F. A. P. C.; Wilcockson, J. B.; Russell, R. W.; Haffner, G. D. Mechanism of biomagnification in fish under laboratory and field conditions. Environ. Sci. Technol. 1999, 33 (1), 133−141.

(30) Debruyn, A. M. H.; Gobas, F. A. P. C. A bioenergetic biomagnification model for the animal kingdom. Environ. Sci. Technol. 2006, 40 (5), 1581−1587. (31) Drouillard, K. G.; Paterson, G.; Liu, J.; Haffner, G. D. Calibration of the gastrointestinal magnification model to predict maximum biomagnification potentials of polychlorinated biphenyls in a bird and fish. Environ. Sci. Technol. 2012, 46 (18), 10279−10286. (32) ECHA. Guidance on Information Requirements and Chemical Safety Assessment. Chapter r.7c: Endpoint Speciﬁc Guidance. Version 1.1; European Chemicals Agency: Helsinki, November 2012; p 239. (33) Zitko, V. Uptake and Excretion of Chlorinated and Brominated Hydrocarbons by Fish, Canada Fisheries and Marine Service Technical Report, 1977; pp 1−14. (34) Zitko, V. The accumulation of polybrominated biphenyls by fish. Bull. Environ. Contam. Toxicol. 1977, 17 (3), 285−292. (35) Zitko, V.; Hutzinger, O. Uptake of chloro- and bromobiphenyls, hexachloro- and hexabromobenzene by fish. Bull. Environ. Contam. Toxicol. 1976, 16 (6), 665−673. (36) Zitko, V. The uptake and excretion of mirex and dechloranes by juvenile atlantic salmon. Chemosphere 1980, 9 (2), 75−78. (37) Bruggeman, W. A.; Matron, L. B. J. M.; Kooiman, D.; Hutzinger, O. Accumulation and elimination kinetics of di-, tri-, and tetra chlorobiphenyls by goldfish after dietary and aqueous exposure. Chemosphere 1981, 10 (8), 811−832. (38) Opperhuizen, A.; Damen, H. W. J.; Asyee, G. M.; Van Der Steen, J. M. D.; Hutzinger, O. Uptake and elimination by fish of polydimethylsiloxanes (silicones) after dietary and aqueous exposure. Toxicol. Environ. Chem. 1987, 13 (3−4), 265−285. (39) Opperhuizen, A.; Schrap, S. M. Uptake efficiencies of two polychlorobiphenyls in fish after dietary exposure to five different concentrations. Chemosphere 1988, 17 (2), 253−262. (40) Sijm, D. T. H. M.; Seinen, W.; Opperhuizen, A. Life-cycle biomagnification study in fish. Environ. Sci. Technol. 1992, 26 (11), 2162−2174. (41) Niimi, A. J. Biological half-lives of chemicals in fishes. Rev. Environ. Contam. Toxicol. 1987, 99, 1−46. (42) Niimi, A. J.; Dookhran, G. P. Dietary absorption efficiencies and elimination rates of polycyclic aromatic hydrocarbons (PAHs) in rainbow trout (Salmo gairdneri). Environ. Toxicol. Chem. 1989, 8 (8), 719−722. (43) Niimi, A. J.; Oliver, B. G. Biological half-lives of polychlorinated biphenyl (PCB) congeners in whole fish and muscles of rainbow trout (Salmo gairdneri). Can. J. Fish. Aquat. Sci. 1983, 40 (9), 1388−1394. (44) Niimi, A. J.; Oliver, B. G. Biological half-lives of chlorinated dibenzo-p-dioxins and dibenzofurans in rainbow trout (Salmo gairdneri). Environ. Toxicol. Chem. 1986, 5 (1), 49−53. (45) Niimi, A. J.; Oliver, B. G. Influence of molecular weight and molecular volume on the dietary absorption efficiency of chemicals by fish. Can. J. Fish. Aquat. Sci. 1988, 45 (2), 222−227. (46) Niimi, A. J.; Palazzo, V. Temperature effect on the elimination of pentachlorophenol, hexachlorobenzene and mirex by rainbow trout (Salmo gairdneri). Water Res. 1985, 19 (2), 205−207. (47) Clark, K. E.; Gobas, F. A. P. C.; Mackay, D. Model of organic chemical uptake and clearance from fish from food and water. Environ. Sci. Technol. 1990, 24 (8), 1203−1213. (48) Clark, K. E.; Mackay, D. Dietary uptake and biomagnification of four chlorinated hydrocarbons by guppies. Environ. Toxicol. Chem. 1991, 10 (9), 1205−1217. (49) Gobas, F. A. P. C.; McCorquodale, J. R.; Haffner, G. D. Intestinal absorption and biomagnification of organochlorines. Environ. Toxicol. Chem. 1993, 12 (3), 567−576. (50) Buckman, A.; Wong, C.; Chow, E.; Brown, S.; Solomon, K.; Fisk, A. Biotransformation of polychlorinated biphenyls (PCBs) and bioformation of hydroxylated PCBs in fish. Aquat. Toxicol. 2006, 78 (2), 176−185. (51) Buckman, A. H.; Brown, S. B.; Hoekstra, P. F.; Solomon, K. R.; Fisk, A. T. Toxicokinetics of three polychlorinated biphenyl technical mixtures in rainbow trout (Oncorhynchus mykiss). Environ. Toxicol. Chem. 2004, 23 (7), 1725−1736. K

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

Environmental Science & Technology (52) Fisk, A. T.; Cymbalisty, C. D.; Bergman, A.; Muir, D. C. G. Dietary accumulation of C-12- and C-16-chlorinated alkanes by juvenile rainbow trout (Oncorhynchus mykiss). Environ. Toxicol. Chem. 1996, 15 (10), 1775−1782. (53) Fisk, A. T.; Cymbalisty, C. D.; Tomy, G. T.; Muir, D. C. G. Dietary accumulation and depuration of individual C10-, C11- and C14polychlorinated alkanes by juvenile rainbow trout (Oncorhynchus mykiss). Aquat. Toxicol. 1998, 43 (2−3), 209−221. (54) 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 (5), 951− 961. (55) Fisk, A. T.; Tomy, G. T.; Cymbalisty, C. D.; Muir, D. C. G. Dietary accumulation and quantitative structure-activity relationships for depuration and biotransformation of short (C-10), medium (C-14), and long (C-18) carbon-chain polychlorinated alkanes by juvenile rainbow trout (Oncorhynchus mykiss). Environ. Toxicol. Chem. 2000, 19 (6), 1508−1516. (56) Muir, D. C. G.; Yarechewski, A. L. Dietary accumulation of four chlorinated dioxin congeners by rainbow trout and fathead minnows. Environ. Toxicol. Chem. 1988, 7 (3), 227−236. (57) Wong, C. S.; Lau, F.; Clark, M.; Mabury, S. A.; Muir, D. C. G. Rainbow trout (Oncorhynchus mykiss) can eliminate chiral organochlorine compounds enantioselectively. Environ. Sci. Technol. 2002, 36 (6), 1257−1262. (58) Barber, M. C. Dietary uptake models used for modeling the bioaccumulation of organic contaminants in fish. Environ. Toxicol. Chem. 2008, 27 (4), 755−777. (59) OECD. OECD Guidelines for Testing Chemicals. Test No. 305: Bioaccumulation in Fish: Aqueous and Dietary Exposure; Organization for Economic Co-operation and Development: Paris, 2012. (60) Brooke, D. N.; Crookes, M. J.; Merckel, D. A. S. Methods for predicting the rate constant for uptake of organic chemicals from water by fish. Environ. Toxicol. Chem. 2012, 31 (11), 2465−2471. (61) Mackay, D.; Arnot, J. A.; Gobas, F. A. P. C.; Powell, D. E. Mathematical relationships between metrics of chemical bioaccumulation in fish. Environ. Toxicol. Chem. 2013, 32 (7), 1459−1466. (62) Goss, K.-U.; Brown, T. N.; Endo, S. Elimination half-life as a metric for the bioaccumulation potential of chemicals in aquatic and terrestrial food chains. Environ. Toxicol. Chem. 2013, 32 (7), 1663− 1671. (63) Klimisch, H. J.; Andreae, M.; Tillmann, U. A systematic approach for evaluating the quality of experimental toxicological and ecotoxicological data. Regul. Toxicol. Pharmacol. 1997, 15, 1−5. (64) Weininger, D. SMILES, a chemical language and informationsystem.1. Introduction to methodology and encoding rules. J. Chem. Inf. Comput. Sci. 1988, 28 (1), 31−36. (65) U.S. EPA. Estimation Programs Interface (EPI) Suite for Microsoft®, Windows, Ver. 4.1., Released October, 2011; U. S. Environmental Protection Agency: Washington, D.C., 2011. (66) Hawker, D. W.; Connell, D. W. Octanol-water partition coefficients of polychlorinated biphenyl congeners. Environ. Sci. Technol. 1988, 22 (4), 382−387. (67) Wania, F.; Dugani, C. B. Assessing the long-range transport potential of polybrominated diphenyl ethers: A comparison of four multimedia models. Environ. Toxicol. Chem. 2003, 22 (6), 1252−1261. (68) Morgan, M. G.; Henrion, M. Uncertainty: A Guide to Dealing with Uncertainty in Quantitative Risk and Policy Analysis; Cambridge University Press: Cambridge, Mass, 1990. (69) OECD. Series on Principles of Good Laboratory Practice and Compliance Monitoring, ENV/MC/CHEM(98)17; Organisation for Economic Co-operation and Development: Paris, France, 1998. (70) Franke, C. How meaningful is the bioconcentration factor for risk assessment? Chemosphere 1996, 32 (10), 1897−1905. (71) OECD. Series on Testing and Assessment: Ecotoxicity Testing, No. 175. Part I - Validation Report of a Ring Test for the OECD 305 Dietary Exposure Bioaccumulation Fish Test and Part II - Additional Report including Comparative Analysis of Trout and Carp Results, Environment

Directorate; Organization for Economic Co-operation and Development: Paris, July 26, 2012. (72) Liu, J.; Haffner, G. D.; Drouillard, K. G. The influence of diet on the assimilation efficiency of 47 polychlorinated biphenyl congeners in Japanese koi (Cyprinus carpio). Environ. Toxicol. Chem. 2010, 29 (2), 401−409. (73) 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 (5), 1496−1504. (74) Law, K.; Palace, V. P.; Halldorson, T.; Danell, R.; Wautier, K.; Evans, B.; Alaee, M.; Marvin, C.; Tomy, G. T. Dietary accumulation of hexabromocyclododecane diastereoisomers in juvenile rainbow trout (Oncorhynchus mykiss) I: Bioaccumulation parameters and evidence of bioisomerization. Environ. Toxicol. Chem. 2006, 25 (7), 1757−1761. (75) 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 (1), 112−119. (76) 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 (15), 4653−4658. (77) Stapleton, H. M.; Letcher, R. J.; Li, J.; Baker, J. E. Dietary accumulation and metabolism of polybrominated diphenyl ethers by juvenile carp (Cyprinus carpio). Environ. Toxicol. Chem. 2004, 23 (8), 1939−1946. (78) 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 (10), 3724−3731. (79) Du, M. M.; Lin, L. F.; Yan, C. Z.; Wang, C.; Zhang, X. Enantiomer-specific bioaccumulation and depuration of hexabromocyclododecanes in zebrafish (Danio rerio). J. Hazard. Mater. 2013, 248, 167−171. (80) Geyer, H.; Scheunert, I.; Korte, F. Relationship between the lipid content of fish and their bioconcentration potential of 1,2,4trichlorobenzene. Chemosphere 1985, 14 (5), 545−555. (81) 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. (82) Schultz, I. R.; Hayton, W. L. Interspecies scaling of the bioaccumulation of lipophilic xenobiotics in fish: An example using trifluralin. Environ. Toxicol. Chem. 1999, 18 (7), 1440−1449. (83) Smedes, F. Determination of total lipid using non-chlorinated solvents. Analyst 1999, 124, 1711−1718. (84) Buckman, A. H.; Brown, S. B.; Small, J.; Muir, D. C. G.; Parrott, J.; Solomon, K. R.; Fisk, A. T. Role of temperature and enzyme induction in the biotransformation of polychlorinated biphenyls and bioformation of hydroxylated polychlorinated biphenyls by rainbow trout (Oncorhynchus mykiss). Environ. Sci. Technol. 2007, 41 (11), 3856−3863. (85) Thurston, R. V.; Gehrke, P. C. Respiratory Oxygen Requirements of Fishes: Description of OXYREF, A Data File based on Test Results Reported in the Published Literature. In Proceedings of an International Symposium; Russo, R. C.; Thurston, R. V., Eds.; Sacramento, California, 1990; pp 95−108. (86) Huey, R. B.; Kingsolver, J. G. Variation in universal temperature dependence of biological rates. Proc. Natl. Acad. Sci. U. S. A. 2011, 108 (26), 10377−10378. (87) Li, N. Q.; Wania, F.; Lei, Y. D.; Daly, G. L. A comprehensive and critical compilation, evaluation, and selection of physical-chemical property data for selected polychlorinated biphenyls. J. Phys. Chem. Ref. Data 2003, 32 (4), 1545−1590. (88) Shiu, W. Y.; Ma, K. C. Temperature dependence of physicalchemical properties of selected chemicals of environmental interest. II. Chlorobenzenes, polychlorinated biphenyls, polychlorinated dibenzoL

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

Environmental Science & Technology p-dioxins, and dibenzofurans. J. Phys. Chem. Ref. Data 2000, 29 (3), 387−462. (89) Opperhuizen, A.; Serne, P.; van der Steen, J. M. D. Thermodynamics of fish/water and octanol/water partitioning of some chlorinated benzenes. Environ. Sci. Technol. 1988, 22 (3), 286− 292. (90) Dabrowska, H.; Fisher, S. W.; Dabrowski, K.; Staubus, A. E. Dietary uptake efficiency of 2,2 ′,4,4 ′,5,5 ′-hexachlorobiphenyl in yellow perch and rainbow trout: Role of dietary and body lipids. Environ. Toxicol. Chem. 1999, 18 (5), 938−945. (91) Clark, K. E. The Bioaccumulation of Hydrophobic Organic Chemicals in Fish; University of Toronto: Toronto, ON, 1990. (92) Fitzsimmons, P. N.; Lien, G. J.; Nichols, J. W. A compilation of in vitro rate and affinity values for xenobiotic biotransformation in fish, measured under physiological conditions. Comp. Biochem. Physiol., C: Comp. Pharmacol. 2007, 145 (4), 485−506. (93) Lazartigues, A.; Thomas, M.; Banas, D.; Brun-Bellut, J.; CrenOlive, C.; Feidt, C. Accumulation and half-lives of 13 pesticides in muscle tissue of freshwater fishes through food exposure. Chemosphere 2013, 91 (4), 530−535. (94) Armitage, J. M.; Arnot, J. A.; Wania, F. Potential role of phospholipids in determining the internal tissue distribution of perfluoroalkyl acids in biota. Environ. Sci. Technol. 2012, 46 (22), 12285−12286. (95) Martin, J. W.; Mabury, S. A.; Solomon, K. R.; Muir, D. C. G. Bioconcentration and tissue distribution of perfluorinated acids in rainbow trout (Oncorhynchus mykiss). Environ. Toxicol. Chem. 2003, 22 (1), 196−204. (96) Ng, C. A.; Hungerbuhler, K. Bioaccumulation of perfluorinated alkyl acids: Observations and models. Environ. Sci. Technol. 2014, 48 (9), 4637−4648. (97) Ng, C. A.; Hungerbühler, K. Bioconcentration of perfluorinated alkyl acids: How important is specific binding? Environ. Sci. Technol. 2013, 47 (13), 7214−7223. (98) Burtnyk, M. D.; Paterson, G.; Drouillard, K. G.; Haffner, G. D. Steady and non-steady state kinetics describe polychlorinated biphenyl bioaccumulation in natural populations of bluegill (Lepomis macrochirus) and cisco (Coregonus artedi). Can. J. Fish. Aquat. Sci. 2009, 66 (12), 2189−2198. (99) Berntssen, M. H. G.; Glover, C. N.; Robb, D. H. F.; Jakobsen, J.V.; Petri, D. Accumulation and elimination kinetics of dietary endosulfan in atlantic salmon (Salmo salar). Aquat. Toxicol. 2008, 86 (1), 104−111. (100) Butt, C. M.; Muir, D. C. G.; Mabury, S. A. Biotransformation of the 8:2 fluorotelomer acrylate in rainbow trout. 1. In vivo dietary exposure. Environ. Toxicol. Chem. 2010, 29 (12), 2726−2735. (101) Munschy, C.; Héas-Moisan, K.; Tixier, C.; Olivier, N.; Gastineau, O.; Le Bayon, N.; Buchet, V. Dietary exposure of juvenile common sole (Solea solea l.) to polybrominated diphenyl ethers (PBDEs): Part 1. Bioaccumulation and elimination kinetics of individual congeners and their debrominated metabolites. Environ. Pollut. 2011, 159 (1), 229−237. (102) Arnoldsson, K.; Haldén, A. N.; Norrgren, L.; Haglund, P. Retention and maternal transfer of environmentally relevant polybrominated dibenzo-p-dioxins and dibenzofurans, polychlorinated dibenzo-p-dioxins and dibenzofurans, and polychlorinated biphenyls in zebrafish (Danio rerio) after dietary exposure. Environ. Toxicol. Chem. 2012, 31 (4), 804−812. (103) Burreau, S.; Axelman, J.; Broman, D.; Jakobsson, E. Dietary uptake in pike (Esox lucius) of some polychlorinated biphenyls, polychlorinated naphthalenes and polybrominated diphenyl ethers administered in natural diet. Environ. Toxicol. Chem. 1997, 16 (12), 2508−2513. (104) Niimi, A. J. Biological half-lives of chlorinated diphenyl ethers in rainbow trout (Salmo gairdneri). Aquat. Toxicol. 1986, 9 (2−3), 105−116. (105) Niimi, A. J.; Palazzo, V. Biological half-lives of eight polycyclic aromatic hydrocarbons (PAHs) in rainbow trout (Salmo gairdneri). Water Res. 1986, 20 (4), 503−507.

(106) Kelly, B. C.; Gobas, F. A. P. C.; McLachlan, M. S. Intestinal absorption and biomagnification of organic contaminants in fish, wildlife, and humans. Environ. Toxicol. Chem. 2004, 23 (10), 2324− 2336. (107) Tomy, G. T.; Thomas, C. R.; Zidane, T. M.; Murison, K. E.; Pleskach, K.; Hare, J.; Arsenault, G.; Marvin, C. H.; Sverko, E. Examination of isomer specific bioaccumulation parameters and potential in vivo hepatic metabolites of syn- and anti-dechlorane plus isomers in juvenile rainbow trout (Oncorhynchus mykiss). Environ. Sci. Technol. 2008, 42 (15), 5562−5567. (108) Du, M. M.; Lin, L. F.; Yan, C. Z.; Zhang, X. Diastereoisomerand enantiomer-specific accumulation, depuration, and bioisomerization of hexabromocyclododecanes in zebrafish (Danio rerio). Environ. Sci. Technol. 2012, 46 (20), 11040−11046. (109) Kelly, B. C.; Ikonomou, M. G.; Blair, J. D.; Morin, A. E.; Gobas, F. A. P. C. Food web-specific biomagnification of persistent organic pollutants. Science 2007, 317 (5835), 236−239. (110) Law, K.; Halldorson, T.; Danell, R.; Stern, G.; Gewurtz, S.; Alaee, M.; Marvin, C.; Whittle, M.; Tomy, G. Bioaccumulation and trophic transfer of some brominated flame retardants in a Lake Winnipeg (Canada) food web. Environ. Toxicol. Chem. 2006, 25 (8), 2177−2186. (111) Mackintosh, C. E.; Maldonado, J.; Hongwu, J.; Hoover, N.; Chong, A.; Ikonomou, M.; Gobas, F. A. P. C. Distribution of phthalate esters in a marine aquatic food web: Comparison to polychlorinated biphenyls. Environ. Sci. Technol. 2004, 38 (7), 2011−2020. (112) Walters, D. M.; Mills, M. A.; Cade, B. S.; Burkard, L. P. Trophic magnification of PCBs and its relationship to the octanol− water partition coefficient. Environ. Sci. Technol. 2011, 45 (9), 3917− 3924. (113) Harris, C. A.; Scott, A. P.; Johnson, A. C.; Panter, G. H.; Sheahan, D.; Roberts, M.; Sumpter, J. P. Principles of sound ecotoxicology. Environ. Sci. Technol. 2014, 48 (6), 3100−3111. (114) Xiao, R.; Adolfsson-Erici, M.; Åkerman, G.; McLachlan, M. S.; MacLeod, M. A benchmarking method to measure dietary absorption efficiency of chemicals by fish. Environ. Toxicol. Chem. 2013, 32 (12), 2695−2700. (115) Arnot, J. A.; Meylan, W.; Tunkel, J.; Howard, P. H.; Mackay, D.; Bonnell, M.; Boethling, R. S. A quantitative structure-activity relationship for predicting metabolic biotransformation rates for organic chemicals in fish. Environ. Toxicol. Chem. 2009, 28 (6), 1168−1177. (116) Brown, T. N.; Arnot, J. A.; Wania, F. Iterative fragment selection: A group contribution approach to predicting fish biotransformation half-lives. Environ. Sci. Technol. 2012, 46 (15), 8253−8260. (117) Kuo, D. T. F.; Di Toro, D. M. Biotransformation model of neutral and weakly polar organic compounds in fish incorporating internal partitioning. Environ. Toxicol. Chem. 2013, 32 (8), 1873− 1881. (118) Papa, E.; van der Wal, L.; Arnot, J. A.; Gramatica, P. Metabolic biotransformation half-lives in fish: QSAR modelling and consensus analysis. Sci. Total Environ. 2014, 470−471, 1040−1046. (119) Cowan-Ellsberry, C. E.; Dyer, S. D.; Erhardt, S.; Bernhard, M. J.; Roe, A. L.; Dowty, M. E.; Weisbrod, A. V. Approach for extrapolating in vitro metabolism data to refine bioconcentration factor estimates. Chemosphere 2008, 70 (10), 1804−1817. (120) Han, X.; Nabb, D. L.; Mingoia, R. T.; Yang, C.-H. Determination of xenobiotic intrinsic clearance in freshly isolated hepatocytes from rainbow trout (Oncorhynchus mykiss) and rat and its application in bioaccumulation assessment. Environ. Sci. Technol. 2007, 41 (9), 3269−3276. (121) 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. M

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Critical Review

Environmental Science & Technology (122) Nichols, J. W.; Schultz, I. R.; Fitzsimmons, P. N. In vitro−in vivo extrapolation of quantitative hepatic biotransformation data for fish. Aquat. Toxicol. 2006, 78 (1), 74−90. (123) Burkhard, L. P.; Cowan-Ellsberry, C.; Embry, M. R.; Hoke, R. A.; Kidd, K. A. Bioaccumulation data from laboratory and field studies: Are they comparable? Integr. Environ. Assess. Manage. 2012, 8 (1), 13− 16.

N

DOI: 10.1021/es506251q Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Development and Evaluation of a Database of ... - ACS Publications

Recommend Documents