Environmental Toxicology and Chemistry, Vol. 32, No. 12, pp. 2711–2718, 2013 # 2013 SETAC Printed in the USA
BIOACCUMULATION KINETICS OF POLYBROMINATED DIPHENYL ETHERS AND DECABROMODIPHENYL ETHANE FROM FIELD-COLLECTED SEDIMENT IN THE OLIGOCHAETE, LUMBRICULUS VARIEGATUS BAOZHONG ZHANG,yz HUIZHEN LI,yx YANLI WEI,yx and JING YOU*y yState Key Laboratory of Organic Geochemistry, Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, Guangzhou, China zSchool of Chemistry and Chemical Engineering, Henan University of Technology, Zhengzhou, China xUniversity of the Chinese Academy of Sciences, Beijing, China (Submitted 14 June 2013; Returned for Revision 18 July 2013; Accepted 26 August 2013) Abstract: The extensive use of polybrominated diphenyl ethers (PBDEs) and decabromodiphenyl ethane (DBDPE) has made them widespread contaminants in abiotic environments, but data regarding their bioavailability to benthic organisms are sparse. The bioaccumulation potential of PBDEs and DBDPE from field-collected sediment was evaluated in the oligochaete Lumbriculus variegatus using a 49-d exposure, including a 28-d uptake and a 21-d elimination phase. All PBDEs and DBDPE were bioavailable to the worms with biota–sediment accumulation factors (BSAFs) ranging from 0.0210 g organic carbon/g lipid to 4.09 g organic carbon/g lipid. However, the bioavailability of highly brominated compounds (BDE-209 and DBDPE) was poor compared with that of other PBDEs, and this was confirmed by their relatively low freely dissolved concentrations (Cfree) measured by solid-phase microextraction. The inverse correlation between BSAFs and hydrophobicity was explained by their uptake (ks) and elimination (ke) rate constants. While ke changed little for PBDEs, ks decreased significantly when chemical hydrophobicity increased. The difference in bioaccumulation kinetics of brominated flame retardants in fish and the worms was explained by their physiological difference and the presence of multiple elimination routes. The appropriateness of 28-d bioaccumulation testing for BSAF estimation was validated for PBDEs and DBDPE. In addition, Cfree was shown to be a good indicator of bioavailability. Environ Toxicol Chem 2013;32:2711–2718. # 2013 SETAC Keywords: Bioaccumulation kinetics sediment Lumbriculus variegatus
Polybrominated diphenyl ethers (PBDEs)
Decabromodiphenyl ethane (DBDPE)
Field
major uptake route of BDE-209 in fish [9]. Thus research on the bioavailability of BDE-209 to benthic invertebrates is required to better understand its transfer pathway from sediment to the aquatic food web, but available data regarding the bioavailability of BDE-209 have been controversial. Several studies claimed that uptake of BDE-209 by benthic annelids, such as Lumbriculus variegatus and Nereis virens, was minimal [10–12]. In contrast, La Guardia et al. [13] found that BDE-209 was the most dominant BFR in the bivalves and gastropods sampled from a river downstream of a wastewater treatment plant. Decabromodiphenyl ethane (DBDPE) was introduced as an alternative BFR of the now restricted deca-BDE [14–16]. A recent study found that DBDPE concentrations in sediment had increased at alarming rates, with a doubling time of less than 7 yr in Lake Michigan and Lake Ontario (USA–Canada) [16]. In addition, the detection of DBDPE in fish suggested its bioaccumulation potential in aquatic vertebrates [9,17]. To our knowledge, however, no data are available on the bioavailability of DBDPE to benthic invertebrates. The objective of the present study was to estimate the uptake and elimination rates of PBDEs and DBDPE in sediment collected from an electronic waste recycling site in South China in the oligochaete L. variegatus. The freely dissolved BFR concentrations in sediment porewater were also measured using solid-phase microextraction (SPME) fibers to better understand the potential extent of bioaccumulation. The impact of hydrophobicity and molecular weight of BFRs on their bioaccumulation was also explored. Finally, the appropriateness of conducting a 28-d bioaccumulation test for hydrophobic BFRs was assessed.
INTRODUCTION
Polybrominated diphenyl ethers (PBDEs) have been used extensively as brominated flame retardant (BFR) additives in a variety of consumer products. The main commercial formulations are penta-, octa-, and deca-BDEs [1]. The PBDEs are persistent and bioaccumulate in wildlife and humans. Recent studies have suggested that PBDEs may cause neurotoxicity and thyroid disruption [2,3], and thus the ubiquitous presence of PBDEs in the environment has raised concerns about their risk [4]. Consequently, the production of penta- and octa-BDE was phased out in the early 2000 s. The use of deca-BDE is still under debate. Because of the possibility of forming more toxic, less brominated PBDEs through debromination, deca-BDE was restricted for use in Europe and the United States. However, the use of deca-BDE is legal in some regions including China. Meanwhile, current-use consumer products will continue to be a large PBDE reservoir even after their production has ceased. Although BDE-209, the major constituent of the deca-BDE formulation, has been detected at elevated concentrations in abiotic environments [4,5], it was once considered not to be bioavailable to aquatic organisms due to its high hydrophobicity (octanol–water partition coefficient [log KOW] ¼ 9.87) and molecular weight (959) [6,7]. A growing number of studies, however, have reported relatively high levels of BDE-209 in fish [5,8,9], and dietary exposure by preying on invertebrates was a All Supplemental Data may be found in the online version of this article. * Address correspondence to
[email protected]. Published online 3 September 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/etc.2384 2711
2712
Environ Toxicol Chem 32, 2013 MATERIALS AND METHODS
Chemical and reagents
A standard mixture containing 8 PBDE congeners (BDE-28, -47, -99, -100, -153, -154, -183, and -209) was purchased from AccuStandard, and a standard solution of DBDPE was obtained from Cambridge Isotope Laboratories. The structures and physicochemical properties of target BFRs are given in Supplemental Data, Figure S1 and Table S1. Two isotopelabeled polychlorinated biphenyls (PCBs), 13C-CB-141 and 13CCB-208, were purchased from Cambridge Isotope Laboratories and used as the surrogate and the internal standard, respectively, for analyzing BFRs by gas chromatography/mass spectrometry (GC/MS). Hexane (high-performance liquid chromatography grade) was purchased from Burdick and Jackson. Dichloromethane and acetone (analytical grade) were obtained from Tianjin Chemical Reagents and redistilled before use. Silica gel and alumina sorbents were cleaned with methanol and dichloromethane, dried, and baked at 180 8C and 250 8C for 6 h, respectively. Anhydrous Na2SO4 was used to remove residual water from the extracts and was baked at 450 8C for 4 h before use. Reconstituted water was prepared by dissolving various salts in deionized water and overnight aeration prior to use [18]. Sediment collection and characterization
Field-contaminated sediment was collected from a stream adjacent to an electronic waste recycling site in Qingyuan, South China (QY). This area has a history of more than 3 decades of electronic waste disposal and currently houses approximately 1200 recycling facilities [19]. Meanwhile, another sediment was collected from a drinking water reservoir in Conghua, China (CH). The CH sediment did not exhibit toxicity to benthic organisms [20], and it was used as the control and the uncontaminated sediment in the elimination phase. The top 5 cm of sediment was collected using a spade shovel and sieved through a 2-mm sieve to remove rocks or other large debris. The sediments were immediately transported to the laboratory, homogenized, and passed through a 500-mm sieve to remove macrofauna. The subsamples for bioaccumulation testing and chemical analysis were stored at 4 8C and 20 8C, respectively. Total organic carbon (TOC) and black carbon (BC) of the sediments were analyzed using an ElementarVavio EL III elemental analyzer. After removing inorganic carbonates with 10% HCl, the sediments were heated at 60 8C for approximately 24 h or at 450 8C for 4 h and used for the TOC and BC quantification, respectively. The QY and CH sediments had TOCs of 1.80 0.13% and 2.75 0.07% and BC contents of 0.103 0.007% and 0.093 0.003%, respectively. Bioaccumulation kinetic tests
Bioaccumulation kinetics of BFRs were evaluated by exposing L. variegatus to the sediment. The worms were cultured in accordance with a US Environmental Protection Agency (USEPA) standard protocol [18] at the Guangzhou Institute of Geochemistry, Chinese Academy of Sciences. No target BFRs were detected in the worms from the culture. The experiments were conducted in 400-mL beakers that contained approximately 120 g wet sediment following the standard method [18], and 300 mL reconstituted water was used as the overlying water. The beaker was placed in an automated water delivery system overnight to allow for the sediment to settle before 30 worms were added. The experiments were performed under a 16:8-h light:dark photoperiod at 23 l 8C,
B. Zhang et al.
and approximately 150 mL of overlying water was changed twice daily. The organisms were not fed throughout the testing. Water quality parameters including pH, temperature, conductivity, and dissolved oxygen were monitored daily, and ammonia concentrations were measured weekly. The CH sediment was included as the control for bioaccumulation testing. The 49-d kinetic exposure consisted of a 28-d uptake and a 21-d elimination phase. In the uptake phase, 30 worms per replicate were exposed to test sediment, and 3 replicates were sampled at specific time intervals. Sampling times for the uptake phase were 1 d, 3 d, 5 d, 7 d, 14 d, and 28 d. At the conclusion of the uptake phase, the organisms from the remaining replicates were sieved from the test sediment and transferred to uncontaminated CH sediment to initiate the elimination phase of the testing. The last uptake sampling time (28 d) served as time 0 for the elimination phase. Three replicates were sampled per time over the course of the elimination phase, with sampling at 3 d, 7 d, and 21 d. When a replicate was terminated, the worms were sieved from the sediments, transferred to clean reconstituted water for 6 h to purge the guts, weighed, and frozen at –20 8C until analysis. One worm per replicate was used to quantify the lipid content of L. variegatus using a spectrophotometric method [21]. Before analysis, the worm was extracted with a mixture of methanol and chloroform and digested with concentrated H2SO4. Then color was developed with vanillin-phosphoric acid reagents. The remaining worms were used for analyzing body residues of BFRs. Sediment and tissue sample preparation
Concentrations of target BFRs in sediment were analyzed in triplicate before and after the 28-d exposure. After addition of the surrogate 13C-CB-141 and several pieces of activated copper sheets, 5 g of freeze-dried sediment was extracted with a mixture of acetone and hexane (1:1, v:v) for 48 h in a Soxhlet apparatus. The extract was concentrated to approximately 1 mL using a Zymark TurboVap, solvent-exchanged to hexane, and then concentrated to approximately 2 mL. The extract was purified with a self-packed column with a 6-mm internal diameter. The column was packed from the bottom to the top with 2 cm of 3% deactivated neutral alumina, 0.5 cm of 3% deactivated neutral silica gel, 1 cm of 25% sodium hydroxide silica gel, 0.5 cm of 3% deactivated neutral silica gel, 5 cm of 44% sulfuric acid silica gel, and 1 cm of anhydrous Na2SO4. Target BFRs were eluted from the column with 10 mL of hexane and 30 mL of 50% dichloromethane in hexane. The eluent was concentrated to 100 mL, and the internal standard 13C-CB-208 was added before GC/MS analysis. Concentrations of BFRs in L. variegatus were also quantified using GC/MS. After the surrogate was added, the weighed worms were sonicated 3 times with 10 mL of acetone. The ultrasonic processor was operated using 5 cycles of 10-s pulses at 600 W. The extracts were combined, filtered through anhydrous Na2SO4, solvent-exchanged to hexane, evaporated to approximately 2 mL, and then cleaned using the self-packed columns as described above. The clean extracts were evaporated to near dryness and rediluted to 50 mL of hexane containing the internal standard. Solid-phase microextraction
The freely dissolved BFRs in sediment porewater (Cfree) were measured using a previously developed SPME method [22]. Disposable SPME fibers coated with 10 mm of polydimethylsiloxane (PDMS) were purchased from Fiberguide and had a phase
Bioavailability of brominated flame retardants in sediment
Environ Toxicol Chem 32, 2013
volume of 0.069 mL/cm. Before use, the fibers were protected with stainless steel envelopes with 110-mm holes and sequentially washed with methanol and distilled water 3 times each. The fibers (30 cm) were placed into a glass vial containing 10 g wet sediment, and the vials were shaken at 120 rpm at 23 8C. The fibers were serially sampled in triplicate, and the sampling times were similar to those in the bioaccumulation testing. After retrieval from the sediment, the fibers were washed with distilled water, dried, and extracted by sonication with 3 mL of 50% dichloromethane in hexane for 5 min. The extraction was repeated twice, and the extracts were combined, cleaned with concentrated H2SO4, evaporated nearly to dryness, rediluted to 50 mL of hexane containing the internal standard, and analyzed on GC/MS. Instrumental analysis
Target BFRs were analyzed on a Shimadzu QP-2010-plus series negative chemical ionization GC/MS with methane used for the reaction gas. Separation of the analytes was achieved using a DB-5HT column (15 m 0.25 mm, 0.1-mm film thickness), and helium was used as the carrier gas at a flow rate of 1.5 mL/min. Injection of 5 mL of extract was performed in splitless mode at 290 8C. The temperature of the ion source and transfer line was set at 200 8C and 280 8C, respectively. The oven was initially set at 110 8C, held for 1 min, heated to 200 8C at 20 8C/min, held for 5 min, heated to 310 8C at 10 8C/min, and held for 10 min. Qualification of the analytes was based on simultaneous detection of the target and qualifier ions within 1% of retention time windows in the selective ion monitoring mode. Ion fragments of a mass-to-charge ration (m/z) of 487 and m/z 489 were used as the target and qualifier ions for BDE-209; m/z 79 and m/z 81 for DBDPE and less brominated PBDE congeners; m/z 372 and m/z 374 for 13C-CB-141; and m/z 474 and m/z 476 for 13C-CB-208, respectively. Quantification was based on internal calibration, with 13CCB-208 being the internal standard. The reporting limit was defined as the product of the lowest concentration of calibration standards and the volume of solution used for instrumental analysis dividing by the mass of the sample extracted
2713
(Supplemental Data, Table S2). To check the performance of the analytical procedures, quality control samples including a blank and a matrix-spiked sample were analyzed every 12 samples. No target BFRs were detected in the blanks. In addition, 13C-CB-141 was added to the samples before extraction as the surrogate, and its recovery was 86 11% for all samples. Data analysis
Biota–sediment accumulation factors (BSAFs) were used as estimated values to describe the bioaccumulation potential of sediment-associated contaminants [23]. Two types of BSAFs were calculated: BSAF28d from the standard 28-d bioaccumulation testing and BSAFss (at steady state) from the kinetic estimation. The calculation of BSAF28d assumed that the steady state was reached for BFRs between sediment organic carbon (OC) and organism lipid within 28 d (Equation 1) BSAF28d ¼
C b28;lipid C s;OC
ð1Þ
where Cb28,lipid and Cs,OC are lipid-normalized BFR concentrations in the organism at 28 d, and OC-normalized BFR concentrations in sediment, respectively. Because sediment concentrations were not significantly different before and after 28-d exposure, the mean values at the beginning and the end of the exposure were used (Table 1). The BSAFss was obtained using a kinetic approach, as shown in Equation 2, and ks (g OC/[g lipid d]) and ke (d1) are uptake and elimination rate constants, respectively BSAFss ¼
ks ke
ð2Þ
To estimate ks and ke, kinetic data in the uptake and elimination phases were simultaneously fitted to a first-order one-compartment model (Equation 3) with the assumptions of passive partitioning between the sediment and the organism, no
Table 1. Polybrominated diphenyl ether (PBDE) and decabromodiphenyl ethane (DBDPE) in sediment collected from an electronic waste recycling site in South Chinaa Cb (ng/g lipid) Contaminant BDE-28 BDE-47 BDE-99 BDE-100 BDE-153 BDE-154 BDE-183 BDE-209 DBDPE a
Log KOWb
Cs (ng/g dry wt)
Steady state
28 d
Cfiber (ng/mL PDMS)
6.24 6.80 7.38 7.09 7.86 7.61 8.61 9.87 11.6
2.63 0.21 25.4 2.4 48.3 4.3 9.17 1.30 20.1 0.7 16.4 1.6 17.3 1.8 1866 160 314 26
316 39 5733 756 2413 302 824 142 679 68 1082 160 306 48 10 640 2773 364 0.54
314 24 5685 405 2649 139 813 50 693 50 1101 157 315 49 8980 470 334 23
19.1 1.5 41.2 2.5 43.3 2.1 10.4 0.9 9.44 0.81 8.33 9.01 372 24 63.3 3.6
Cfree (pg/L)c 54.5 4.2 49.6 3.0 21.6 1.0 8.05 0.72 2.25 0.19 2.90 0.682 4.11 0.27 0.0491 0.0028
Cs,bioavailable (ng/g dry wt)d 1.32 0.10 1.83 0.11 1.22 0.06 0.367 0.033 0.181 0.016 0.195 0.096 1.47 0.10 0.0634 0.0036
PBDE and DBDPE sediment residues (Cs) are given in ng/g dry weight; their concentrations in Lumbriculus variegatus (Cb) are given in ng/g lipid at the steady state and at 28 d; and their concentrations in polydimethylsiloxane (PDMS) fiber (Cfiber) are given in ng/mL PDMS at equilibrium. The estimated freely dissolved chemical concentrations in sediment porewater (Cfree; pg/L) and the readily bioavailable fractions of target compounds in sediment [Cs,bioavailable]; ng/g dry wt) are also presented. Data are reported as mean standard deviation of 3 replicates. Total organic carbon content in sediment was 1.80% 0.13% and lipid content in the organism was 1.20% 0.13%. b Octanol–water partition coefficients (KOW) of PBDEs were from Bao et al. [6], and the KOW value of DBDPE was from Kierkegaard et al. [14]. c The Cfree was calculated by dividing Cfiber by the fiber–water partition coefficient (Kfw), which was calculated from log KOW using the equation of Stenzel et al. [26]: log Kfw ¼ 0.665 log KOW þ 1.394. d The Cs,bioavailable was calculated by multiplying Cfree by the organic carbon–water partition coefficient (Koc), which was calculated from log KOW using the equation of La Guardia et al. [13]: log Koc ¼ 0.332 log KOW þ 4.12.
2714
Environ Toxicol Chem 32, 2013
B. Zhang et al.
growth dilution, and sediment concentration remaining constant throughout the testing [24,25] C b;lipid ¼
k s C s;OC 1 eke t ke
ð3Þ
where Cb,lipid is lipid-normalized BFR concentration in organism at time t (d) and Cs,OC is OC-normalized BFR concentration in sediment (Table 1). Apparent steady state was defined as 95% of the equilibrium, and the time to reach apparent steady state (t95) was derived from ke. t 95 ¼
lnð1 0:95Þ 2:996 ¼ ke ke
ð4Þ
To ensure that equilibrium of a chemical among sediment OC, porewater, and fiber was reached, serial samplings were conducted for SPME measurements. Concentrations of BFRs in the fibers at equilibrium (Cfiber) were modeled using Equation 5 C t ¼ C fiber 1 ekt
ð5Þ
where Ct and Cfiber are BFR concentrations in the fibers (ng/mL PDMS) at time t (d) and at equilibrium, respectively, while k is the desorption rate constant (d1). The Cfree values were calculated by dividing Cfiber by the fiber–water partition coefficients (Kfw; Equation 6). The Kfw values were calculated from log KOW values [6,14] using Equation 7, as proposed by Stenzel et al. [26]. Using Equations 8 and 9, the readily bioavailable fractions of BFRs in sediment (Cs, bioavailable) and bioconcentration factors (BCFs) were attained, respectively. The OC–water partition coefficients (KOC) were derived from log KOW, as suggested by La Guardia et al. [13] (log KOC ¼ 0.322 log KOW þ 4.12) C free ¼
C fiber K fw
ð6Þ
log K fw ¼ 0:665 log K OW þ 1:394
ð7Þ
C s;bioavailable ¼ C free K OC
ð8Þ
BCF ¼
C b;lipid C free
ð9Þ
The uptake and elimination estimates by data fitting were performed using Micromath Scientist. Statistical differences among groups were compared by analysis of variance, and Tukey’s honestly significant difference test was introduced for further comparison. The level of significance was defined at a ¼ 0.05. RESULTS AND DISCUSSION
Levels of PBDEs and DBDPE in sediment
All target PBDEs were detected in QY sediment, and BDE209 was the most abundant, with a concentration of 1866 160 ng/g dry weight (Table 1). High concentrations of PBDEs in QY sediment were anticipated. Previous studies have shown that the sampling area was a world hotspot, heavily polluted by BFRs, and PBDE concentrations in the present study were consistent with previously reported data in the same area [5]. Unlike the well-studied PBDEs, information on the
occurrence and environmental behavior of DBDPE is sparse. He et al. [9] reported that DBDPE concentrations in sediments collected in South China ranged from not detectable to 1700 ng/g dry weight, with a mean of 200 ng/g dry weight. Wei et al. [15] analyzed DBDPE and BDE-209 in sediments from Arkansas (USA) and found that concentrations of DBDPE were approximately 1 order of magnitude less than those of BDE209. The concentration of DBDPE was 314 26 ng/g dry weight in QY sediment, and it was also around 6 times less than the BDE-209 concentration (Table 1). The recent upsurge in DBDPE concentrations in sediment has shown that more studies are needed on this alternative BFR of BDE-209 [16]. Target BFRs were also analyzed in CH sediment, which was used as the control and the sediment in the elimination phase. There was no detection of less brominated PBDEs and DBDPE, but trace amounts of BDE-209 were detectable in this sediment, although its concentration was less than the reporting limit. Bioaccumulation kinetics of PBDEs and DBDPE
The quality of the overlying water was monitored throughout the bioaccumulation kinetic testing, including dissolved oxygen (5.4 0.4 mg/L), pH (7.6 0.2), temperature (23.6 0.8 8C), conductivity (316 10 ms/cm), and ammonia (
