Polybrominated Diphenyl Ethers (PBDEs) and
Polychlorinated Biphenyls (PCBs) in
Lake Michigan Salmonids.
Environ. Sci. Technol., v.35, n.6 14feb01
JON B. MANCHESTER-NEESVIG,*, †
KARLIS VALTERS,‡ AND
WILLIAM C. SONZOGNI †
Water Chemistry Program, University of Wisconsin-Madison,
660 North Park Street, Madison, Wisconsin 53706, and
Department of Environmental Chemistry,
Wallenberg Laboratory, Stockholm University,
S-106 91 Stockholm, Sweden
* Corresponding author phone: (608) 265-4182; fax (608) 2620454; email@example.com.
† University of Wisconsin-Madison.
‡ Stockholm University.
Polybrominated diphenyl ethers (PBDEs) have been used extensively over the past two decades as flame retardants in most types of polymers. Many measurements of PBDEs in various environmental matrices from Sweden, Holland, Japan, and elsewhere have been reported, but few measurements are available for North America. PBDEs in 21 coho and chinook salmon taken from Lake Michigan tributaries in 1996 were measured for this study. The salmon samples were extracted and initially analyzed for polychlorinated biphenyl (PCB) congeners. It was demonstrated for these samples that the same extract fraction contains PBDEs. Six PBDE congeners were observed in all 21 samples, and the rank order of concentration of these congeners was similar to that in commercial mixtures of PBDEs. The average concentration across all samples of the sum of PBDE congeners was 80.1 ng/g of wet weight or 2440 ng/g of lipid. This is much less than the average sum PCB concentration (1450 ng/g of wet weight; 43100 ng/g of lipid). However, the average concentration of the most abundant PBDE congener (IUPAC BDE-49: 52.1 ng/g wet, 1590 ng/g of lipid) was about one-third of the average concentration of the most abundant PCB congener (IUPAC CB-153: 149 ng/g wet, 4550 ng/g of lipid). On the basis of an extensive literature survey, the concentrations of PBDEs reported here are among the highest in the world for salmon in open waters. The concentrations of PBDEs and PCBs are both correlated with fish length and mass, but not with lipid content. The concentrations of PBDEs and PCBs are highly correlated in individual fish, implying that PBDEs are as prevalent as PCBs in Lake Michigan.
Polybrominated diphenyl ethers (PBDEs) are added to many common items to reduce flammability. In recent years, however, an unexpected consequence of the widespread use of PBDEs has been observed. PBDEs are contaminating many parts of the global environment (1). An increasing body of evidence suggests that these compounds bioaccumulate and induce undesirable biological outcomes. Researchers in Sweden, Holland, and Japan have measured PBDEs in several environmental compartments, and some work has been done in North America. However, assessment of the prevalence of PBDEs in the Great Lakes region is almost nonexistent.
Background on PBDEs.
PBDEs are one type of a class of products known as flame retardants. Brominated (and chlorinated) flame retardants perform their intended function by decomposing at high temperature and liberating halogen atoms. These atoms are very effective reducing agents and free radical inhibitors and so quench oxidation reactions (2). For maximum effectiveness, the flame retardant must decompose at a temperature -50 °C lower than the material it is protecting (3). Organic bromine compounds fulfill this requirement for many synthetic polymers. Thus, the use of brominated flame retardants (BFRs) has greatly increased with increasing use of plastics.
BFRs are used in all kinds of plastic resins except polyacetals (4). Examples of products that contain BFRs are electronic circuit boards and cases, furniture, building materials, textiles, carpets, and vehicles. BFRs are incorporated into plastics either through reaction or addition (5). Reactive combination produces plastics with covalently bonded flame retardants that are less likely to leave the product. In contrast, additive flame retardants are only mixed with the polymer resin and so continually migrate out of the final product (6). PBDEs are additive flame retardants; thus, the appearance of these compounds in the environment is not surprising. The major types of polymers that are protected with PBDEs are, in descending order of importance, high-impact polystyrene, acrylonitrile butadiene styrene (ABS), flexible polyurethane foam, textile coatings, wire and cable insulation, and electrical/ electronic connectors. PBDEs are added to these polymers at levels ranging between 5 and 30% (7).
The annual world production in the mid-1990s of flame retardants was 600000 metric tonnes (8). Brominated organic compounds represent one-fourth of the total (150000 tonnes). About one-third of the BFRs are PBDEs (50000 tonnes), another third are tetrabromobisphenol A, with the remainder made up of many bromine products, such as polybrominated biphenyls. Most of the worldwide demand for these products is distributed between North America (40%), the Far East (30%), and Europe (25%) (2). Eight manufacturers produce the entire supply of PBDEs; these companies are located in France, Great Britain, Israel, Japan, Holland, and the United States.
Chemistry, Production, and Xenobiology of PBDEs.
PBDEs are structurally similar to polychlorinated biphenyls (PCBs), although PBDEs are generally more polar than PCBs because of the presence of the oxygen atom and the resulting asymmetry about the horizontal axis. The numbering of carbon atoms in PBDE molecules is identical to that in PCB molecules, and congeners are named with identical numbering systems.
Only a small number of the 209 possible PBDE congeners are found in the commercial mixtures. Three main types of products are commonly produced, varying in degree of bromination: pentabromodiphenyl ether, octabromodiphenyl ether, and decabromodiphenyl ether. Estimates for 1999 indicate that global production consisted of 82% decaBDE, 6% octaBDE, and 12% pentaBDE (9). The names of the types of commercial produces are somewhat misleading; the "pentabromo" product is predominately an almost equal mixture of tetraBDE and pentaBDE congeners, and the "octabromo" product consists of heptaBDE and octaBDE congeners. Congeners with fewer than four bromines are not found in significant quantity in the technical mixtures. Commercial mixtures of PBDEs are lipophilic, with log Kow values between 6.5 and 10. Vapor pressures for these mixtures are low (10-6-10-9 Torr). PBDEs are resistant to physical, chemical, and biological degradation (10).
Recently, the PBDE congeners present in Bromkal 70-5DE, a common and widespread commercial PBDE flame retardant, have been identified and quantified (11). Knowledge of the PBDE congeners present in this commercial mixture provides a list of pure compounds likely to be found in any sample from a PBDE-contaminated area. Conversely, the presence of these compounds in an environmental sample indicates probable contamination by PBDE flame retardants.
Because of their chemical similarity to dioxins and PCBs, there is a large and growing body of research describing the toxic aspects of PBDEs in the environment. A detailed review of this literature is beyond the scope of this discussion and may be found in the literature (1,10,12). Unlike dioxins and PCBs, PBDEs are not strong inducers of the aryl hydrocarbon hydroxylase (Ali) receptor. However, PBDEs are potent compounds with respect to thyroid function, and decreases of free thyroxin hormone (T4) have been observed in animals (including humans) after exposure to PBDEs.
PBDE congeners have been quantified in samples of human milk collected in Sweden during different periods from 1972 to 1997 (13). Concentrations of PBDEs in the milk increased > 50-fold during the past 25 years. PBDEs also have been measured in human adipose tissue; for example, Haglund et al. (14) reported as much as 8800 ng/kg of lipid in a 74-year-old Swedish male, and a study by the Wisconsin State Laboratory of Hygiene (15) found detectable quantities of PBDEs in blood serum samples from Lake Michigan fish consumers.
Recently, our laboratory has quantified several PBDE congeners in salmon taken from southern Lake Michigan. Concentrations of PCBs were also measured in the same salmon tissue. In this paper, we present these results and a comparison of levels of PBDEs and PCBs in these fish.
Twenty-one salmon were collected from two Lake Michigan tributaries during the fall of 1996. Sixteen coho (Oncorhynchus kisutch) salmon were collected on October 22 from the Kewaunee River, Wisconsin, and five Chinook (O. tshawytscha) salmon were collected on October 16 in Strawberry Creek, a Door County tributary to Sturgeon Bay of Lake Michigan. Nine Cohn salmon were female, two were male, and five were not sexed. Three Chinook salmon were female and two were male.
The fish were frozen following collection. Prior to analysis a ~100 g steak was cut just in front of the dorsal fin to produce a "standard" steak that included skin, muscle, bone, and organ tissues. The standard steak was thawed and homogenized with dry ice flakes in a blender to produce a slurry. PBDE and PCB congener concentrations, water content, and lipid content were determined on ~8-10 g of the homogenized slurry. The methodological protocol detailing sample treatment and processing for PCB congener residues, water content, and lipid content determinations have been presented elsewhere (16). Briefly, organic compounds present in the fish tissue homogenate were extracted with dichloromethane using a column extraction technique. Lipids were removed using gel permeation chromatography, and silica gel column chromatography (described below) was used to separate PCBs and PBDEs from other organic compounds. Solvents were of residue analysis grade from EM Science. PCB congeners 14, 65, and 166 (Ultra Scientific) were added to all samples before extraction to monitor recoveries. Percent recoveries, with one standard error in parentheses, were 86% (7), 71% (2), and 109% (3), for PCB congeners 14, 65, and 166, respectively. Relative standard deviations of a triplicate analysis of one fish tissue were 6, 7, and 9% for lipid content, total PCB, and total PBDE, respectively.
The salmonids were collected and analyzed for PCBs as part of a larger project examining the trophic transfer of PCBs, and the extracts were stored at -30°C. These extracts were subsequently analyzed for PBDE congeners. Thus, the extraction scheme was not chosen specifically for PBDEs. Therefore, it was necessary to ensure that the separations done as part of the cleanup sequence allowed PBDEs to be collected in the PCB fraction.
Several methodologies have been published describing the extraction of PBDEs from biological matrices, and different cleanup schemes can result in a number of final extracts containing different sets of organic contaminant analytes (17). PCBs are often separated from PBDEs using a silica gel column, with PBDEs appearing in the second eluate. However, we have found that PBDEs may elute in the first fraction with PCBs and be successfully quantified.
A recovery experiment was done to confirm that PCBs and PBDEs might be collected in the same silica gel fraction. One milliliter of a PBDE spike solution was placed on a 5 g, 1 cm diameter silica gel column. The silica was prepared by drying at 130°C for 24 h and then precisely deactivated with 3% w/w water. The column was eluted with 50 mL of hexane (EM Science). All (98%) of the nine PBDE congeners were recovered in this first silica fraction.
The extracts containing PCBs and PBDEs were run on a 30 m DB-5 column (J&W Scientific; Hewlett-Packard 5890 GC) ramped from 100 °C at 1 °C/min to 250 °C to separate PBDE congeners from PCB congeners. This methodology successfully separated BDE-47 from CB-180. The remaining PBDE congeners eluted at later retention times at which there was no PCB congener interference.
We do not suggest that this methodology is preferred over other separation schemes. Instead, we have shown that PBDE congeners may be quantified in extracts originally prepared for PCB analysis. Moreover, the appearance of PBDEs in a particular cleanup fraction may depend on the sample matrix.
PBDE standards for GC analysis were obtained from Dr. Åke Bergman, Professor of Environmental Chemistry at Stockholm University (18). The standards were synthesized in Dr. Bergman's laboratory and are now commercially available (Cambridge Isotope Laboratories, Andover, MA). The standards were used to create a gas chromatograph three-point calibration for nine PBDE congeners (BDE-28, -47, -66, -85, -99, -100, -138, -153, and -154). The Wisconsin State Laboratory of Hygiene confirmed the presence of PBDEs in several extracts using a Finnigan Magnum ion-trap mass spectrometer (GC-ITMS).
PBDE congeners were detected in all salmon samples taken from Lake Michigan. Six of the nine congeners in the calibration were found in every sample, and three congeners (BDE-28, BDE-85, and BDE-138) were not found in any sample. The results are summarized in Table 1.
Congener BDE-47 was the most abundant congener, as has been the case in many other environmental samples (see below). The proportions of each of the six detected congeners were very uniform among samples. The concentration of each congener in a sample, normalized to the sum of all PBDE congeners in the sample, is displayed in Figure 1 for all 21 samples. The PBDE congeners in Bromkal 705DE, and their weight fraction, are given in Table 2, and some of these congeners (those in the salmon samples) are also shown in Figure 1. The rank order of the congener proportions in the salmon samples is similar to that found in the commercial PBDE flame retardant Bromkal 70-5DE, although differences are apparent. A thorough understanding of these measurements will require a more complete assessment of the possible sources of the PBDEs in the salmon and knowledge of the fate of these compounds once in the environment.
The average concentration of total PBDE (sum of observed congeners) in the salmon was 80.1 ng/g of wet weight (2440 ng/g of lipid). The salmon also contained an average of 1450 ng/g of wet weight total PCB (43100 ng/g of lipid), or ~18 times as much PCB as PBDEs. However, the PCB total is the sum of as many as 90 congeners, whereas the PBDE total is the sum of 6 congeners. A more useful comparison might be between single congeners. The average concentration of the most abundant PCB congener in the salmon (congener CB153) was 149 ng/g of wet weight (4550 ng/g of lipid). This is only ~3 times higher than the average concentration of BDE-47 in these samples (52.1 ng/g wet, 1590 ng/g of lipid). A recent report (19) presents concentrations of PBDE congeners measured in several individual Steelhead trout taken from Lake Michigan. The authors report an average total PBDE (sum of BDE-28, -47, -99, -100, -153, -154) in the trout muscle of 41 ng/g of fresh weight (3000 ng/g of lipid), or about half the average total PBDE found in the fish described in this paper. Also, the relative amounts of the congeners are almost identical to that reported here (percent of total for BDE-47, 56%; BDE-99, 19%; BDE-100, 12%; BDE154, 6.6%; BDE-153, 3.6%).
TABLE 1. Concentrations of PBDE Congeners, Total PBDEs, and PCB Congener 153 in Lake Michigan Salmon (Nanograms per Gram of Wet Weight except As Noted
% total total PBDEs sample lipid BDE-47 BDE-66 BOE-100 BDE-99 BDE-154 BDE-153 PBDEs (ng/g of lipid) CB-153 OLS-1 3.03 67.3 2.1 13.0 10.8 6.0 3.6 103 3390 183 OLS-2 3.64 57.6 2.1 10.8 8.6 4.7 2.8 86.6 2380 184 OLS-3 5.79 51.2 1.9 9.9 8.5 4.3 2.6 78.5 1360 136 OLS-4 3.16 26.0 1.4 5.2 7.2 2.8 1.8 44.6 1410 82.3 OLS-5 3.43 51.7 2.2 9.7 6.9 4.5 2.7 77.8 2270 172 OLS-6 4.04 70.4 1.8 13.6 10.7 6.2 3.5 106 2630 187 OLS-7 5.96 37.1 1.3 6.5 5.9 3.3 2.0 56.1 942 101 OLS-8 2.27 31.0 1.4 5.6 7.7 3.2 2.0 50.8 2240 85.0 OLS-9 2.11 55.4 1.4 10.1 8.4 4.6 2.7 82.6 3920 150 OLS-10 6.12 29.3 1.2 5.4 6.7 2.8 1.8 47.3 773 85.8 OLS-11 3.04 50.4 1.7 8.9 7.3 4.3 2.5 75.1 2470 143 OLS-12 2.15 51.4 1.6 9.8 7.8 4.7 2.7 78.1 3630 133 OLS-13 2.52 40.4 1.5 7.3 10.4 3.6 2.3 65.5 2600 108 OLS-14 2.71 45.4 1.6 8.0 8.7 3.9 2.3 69.8 2580 127 OLS-15 3.24 51.3 1.5 9.4 7.3 4.4 2.5 76.3 2360 127 OLS-16 4.03 45.6 1.4 8.1 9.0 3.8 2.3 70.2 1740 109 OLS-17 4.35 59.9 1.5 10.9 10.6 5.0 2.9 90.8 2090 169 OLS-18 1.83 95.1 2.5 18.8 18.9 8.5 4.8 149 8120 244 OLS-19 5.29 33.9 1.4 6.5 7.9 3.5 2.1 55.3 1050 125 OLS-20 7.19 78.7 1.6 14.6 13.4 6.4 3.5 118 1640 183 OLS-21 5.79 65.5 2.4 12.2 13.5 4.5 2.5 101 1740 212 av 3.89 52.1 1.7 9.7 9.3 4.5 2.7 80.1 2440 149 SD 1.56 17.1 0.4 3.4 3.0 1.4 0.7 25.3 1550 43.9 max 7.19 95.1 2.5 18.8 18.9 8.5 4.8 148 8120 244 min 1.83 26.0 1.2 5.2 5.9 2.8 1.8 44.6 773 82.3
FIGURE 1. Relative amount of PBDE congeners in 21 salmon samples and in Bromkal 70-5DE (Bromkal data from ref 111. Congener concentrations are expressed as a fraction of total PBDE.
TABLE 2. PBDE Congeners in a Common Brominated Flame Retardant
PBDE structure concn (°/u, w/w) 47 2,2',4,4'-tetraPBDE 37 99 2,2',4,4',5-pentaPBDE 35 100 2,2',4,4',6-pentaPBDE 6.8 153 2,2',4,4',5,5'-hexaPBDE 3.9 154 2,2',4,4',5,6'-hexaPBDE 2.5 85 2,2',3,4,4'-pentaPBDE 1.6 138 2,2',3,4,4',5'-hexaPBDE 0.41 66 2,3',4,4'-tetraPBDE 0.22 28 2,4,4'-triPBDE 0.11 17 2,2',4-triPBDE 0.022 183 2,2',3,4,4',5',6-heptaPBDE detected, not quantified
Measurements of PBDEs in the Environment. Many reported measurements of PBDEs in environmental matrices are listed in Table 3. Although this is not a complete list, the tabulated values provide an overview of PBDE levels in the global environment. These values may be compared with those reported here for Lake Michigan salmon.
The values in Table 3 are a representative listing of PBDE environmental measurements as reported in the literature. The table reveals that the Baltic region is the most thoroughly researched area with respect to PBDEs in the global environment. In contrast, very little research on PBDEs in the Laurentian Great Lakes of North America has been published. However, it has recently been suggested that a lake trout fish homogenate from Lake Ontario (originally developed as a certified reference material for dioxin research) may be appropriate for use as a PBDE certified reference material because it contains easily quantified levels of PBDEs (20). Also, one of the earliest reports of PBDEs in an environmental sample described the identification of PBDE residues in bird eggs from the southern United States (Texas and Louisiana) (21). Despite this early start, there now appears to be a lack of information regarding PBDEs in the Great Lakes region in particular and the United States in general.
Available measurements of PBDEs indicate that environmental levels are not decreasing. For example, the level of total PBDEs in Baltic salmon collected in Sweden in 1995 was equal to that measured in sea trout taken from the southern coast of Sweden in 1979 (17, 22), and the amount of PBDE in pike muscle from the Viskan River in Sweden actually increased between 1981 and 1995 (6, 22).
Part of the increase in PBDEs in Viskan River pike may have been due to differences in location along the river. It is apparent from Table 3 that PBDEs concentrations are correlated with human populations and, so, are most likely of anthropogenic origin. Concentrations of PBDEs in similar matrices are higher in rivers than in open seas. There is also a decreasing gradient of concentrations moving from south to north in the Scandinavian samples, that is, from populous Europe toward the Arctic. For example, fish muscle concentrations decrease from the Baltic Sea to the Bothnian Sea and are lower still in Lake Storvindeln, Lapland (23, 24). The values for PBDEs in cod liver in the southern, central, and northern North Sea present the best example of this trend. Concentrations in this single matrix decrease by almost 2 orders of magnitude along this transect (25).
There is evidence in Table 3 that PBDEs bioconcentrate. For example, fish concentrations are much higher than sediment concentrations in samples collected at the same time and location in the Viskan River (6). Also, Osprey concentrations in Sweden are higher than the fish they feed on, providing some evidence for biomagnification (23).
TABLE 3. Environmental Measurements of PBDEs
collection total PBDEs year matrix location (ng/g wet)a ref Scandinavia 1979 sea trout Klosterfjorden, Sweden 15 1980 eel muscle Viskan River, Sweden 215 22 1981 pike muscle Viskan River, Sweden 124 1981 pike liver Viskan River, Sweden 9680 1986 whitefish muscle Lake Storvindeln, Lapland 0.2 1986 herring muscle Bothnian Sea 7 1987 herring muscle Baltic Sea 23 23 1987 Arctic char muscle Lake Vättern, Sweden 27 1982-1986 osprey muscle Sweden, various locations 86 1987 surface sediment Bornholm Deep, Baltic Sea 0.56 28 1988 sewage sludge Gothenburg, Sweden 256 1991 salmon muscle Umeå River, Sweden 287 14 1994 human adipose Sweden 14c 1995 pike muscle Viskan River, Sweden 358 6 surface sediment Viskan River, Sweden 46 1995 salmon muscle Baltic Sea, taken from 14 1995 salmon egg Dalälven River, 9 17 1995 salmon blood Sweden 6 Japan 1981-1985 mussel Osaka Harbor 17.4 1981-1985 surface sediments Neyo River, Osaka 57.26 29 1981-1985 sardine Wakayama 0.8 United States 1980 black skimmer eggs Isle Dernier, Louisiana detected d 27 1987 dolphin blubber Atlantic coast, Virginia 200 30 1991 carp muscle Buffalo River, New York 18.7 31 Eastern North Atlantic Ocean 1977-1987 cod liver North Sea (north) 26 1977-1987 cod liver North Sea (central) 54 25 1977-1987 cod liver North Sea (south) 170 1995 sperm whale blubber North Atlantic Ocean 99 1995 sperm whale liver North Atlantic Ocean 4 1995 minke whale blubber North Atlantic Ocean 122 1995 dolphin blubber North and Wadden Seas 7700 26 1995 dolphin liver North and Wadden Seas 31 1995 harbor seal blubber North and Wadden Seas 1083 1995 harbor seal liver North and Wadden Seas 20 a Some cited references report concentrations normalized to lipid content. These values have been converted here to wet weight, using data from the reference, to facilitate comparisons between studies. Also, in several cases, multiple values for the same matrix within a study were averaged. b These values have units of ng/g dry. c These values are normalized to lipid content. d Not quantified.
Although regional contamination of the environment by PBDEs has been apparent for some time, new evidence indicates that these compounds are now globally distributed, reaching even the deep oceans, as evidenced by measurements of PBDEs in the blubber and liver of deep-feeding sperm whale (26). The presence of PBDEs in the deep ocean suggests that PBDEs, like PCBs, have become ubiquitous environmental contaminants. This possibility was suggested 12 years earlier (27).
PBDEs have been quantified in sediments of rivers and seas. Sediments often provide an environment conducive to chemical and biological degradation, so the presence of PBDEs in sediments underscores their resistance to degradation. Also, sediment records can sometimes be used to reconstruct historical inputs of persistent compounds. The value listed in Table 3 for Baltic Sea surface sediments is one of a set of values generated from measurements of laminated sediments in this region (28). Depth profiles generated from these measurements indicate that PCBs and DDT, after accumulating rapidly several decades ago, are now at relatively constant values in these sediments. In contrast, the profile for PBDEs suggests that these compounds have only begun to accumulate in the past two decades, and there is no indication that they have reached maximum values. Also, the researchers state that surface values for individual PBDE congeners are of the same magnitude as PCB congeners and that the PBDE concentrations are increasing more rapidly than PCB concentrations ever increased.
The Lake Michigan salmon average PBDE value presented here may be compared with similar values in Table 3. Lake Michigan is a large body of water (57100 km2, maximum depth of 300 m) and, although smaller and containing freshwater, might be compared with the Baltic Sea. A recent measurement of total PBDE in salmon muscle in the Baltic (14 ng/g wet) (17) is about one-sixth the average value measured for Lake Michigan. Although this is a useful comparison, it is obviously limited by many differences between the two salmon data sets. However, given the evidence indicating that PBDEs bioconcentrate and that the salmon in each system are at the top of their respective food webs, one might conclude that Lake Michigan is more contaminated with PBDEs than the Baltic Sea.
Comparison of PBDE and PCB Concentrations in Lake Michigan Salmon.
The concentrations of numerous PCB congeners were also measured in the salmon samples discussed here. Thus, it is possible to compare the concentrations of PBDEs and PCBs in the same fish taken from Lake Michigan.
Figures 2 and 3 show the measured concentrations of total PBDEs and PCBs as a function of fish length and fish mass. A line representing the results of a linear regression is shown for each data set in both figures. Regression parameters are listed in Table 4.
Several trends are visible in Figures 2 and 3. Both PBDEs and PCBs are positively correlated with fish length and fish mass. The coefficients of determination, r2, indicate that the correlations are moderate. Also, the correlations are very similar for each contaminant with respect to each independent variable; length is better correlated with each contaminant than is mass, and PCB is more strongly correlated with length and mass than is PBDE.
Fish length and mass are directly related to fish age. The correlations visible in Figures 2 and 3 indicate that Lake Michigan salmonids acquire PBDE and PCB over time scales of years, comparable to the lifespan of the fish, and not through acute exposure. This implies that PBDEs have been a part of the Lake Michigan environment for many years.
Fish concentrations of PBDEs and PCBs are plotted as a function of fish lipid content in Figure 4. There is no correlation between the concentration of contaminant and the amount of lipid in the fish. Although correlations of this type have been observed when average lipid content and average contaminant level between species of fish were compared, recent work has shown that this correlation does not hold within a given fish species (32). Our measurements support this observation. This may reflect relatively rapid changes in lipid content within individual fish as compared to the slow uptake and depuration of persistent organic contaminants within the same fish. Close examination of the data in Figures 2-4 reveals the covariant nature of PBDE and PCB concentrations in the fish examined here. This covariance is examined directly in Figure 5. It is clear from the regression in Figure 5 that the amount of each contaminant is highly correlated (r2 = 0.81) within each fish. A likely explanation for this correlation is that the contaminants are being acquired from the same source (Lake Michigan) and that net uptake is occurring at comparable rates. This would suggest that PBDEs have been a part of the Lake Michigan environment for many years and that, like PCBs, PBDEs are distributed throughout the lake.
FIGURE 2. Length of Lake Michigan salmon versus concentrations of total PBDEs and PCBs in the fish.
FIGURE 3. Mass of Lake Michigan salmon versus concentrations of total PBDEs and PCBs in the fish.
TABLE 4. Regression Parameters for PBDE and PCB Concentration Measurements
variable slope r 2 length vs PBDE 1.45 0.39 PCB 2.93 0.52 mass vs PBDE 65 0.32 PCB 133 0.45 lipid vs PBDE -7.8 0.002 PCB -1.3 0.00002 PBDE vs PCB 15.8 0.82
FIGURE 4. Lipid content of Lake Michigan salmon versus concentrations of total PBDEs and PCBs in the fish.
Figure 5 fifil~. ~lötöfihecö~ce~rä5~t~h~tr'~`'s'm~90rr~trl~ salmon versus the concentration of total PCBs in the same fish. The solid line is a linear regression on the data.
We thank Conrad Lamon for collecting and transporting salmon from the field to laboratory. We are grateful to the American Chemical Society for providing funds to support the work of KY. while at the Water Chemistry Program. This work was funded, in part, by the University of Wisconsin Sea Grant Institute under grants from the National Sea Grant College Program, the National Oceanic and Atmospheric Administration, the U.S. Department of Commerce, and the State of Wisconsin. Federal Grant NA90AA-D-SG469, Project R/WM-41.
(1) de Boer, J.; de Boer, K.; Boon, J. P. In The Handbook of Environmental Chemistry; Paasivirta, J., Ed.; Springer-Verlag: New York, 1999.
(2) Arias, P. Report to the Organization for Economic Cooperation and Development (OECD), Brussels, Belgium, 1992.
(3) Price, D. Proceedings from the Workshop on Brominated Aromatic Flame Retardants, Solna, Sweden, Oct 24-26; Swedish National Chemicals Inspectorate: 1989; pp 13-23.
(4) Teuerstein, A. Proceedings from the Workshop on Brominated Aromatic Flame Retardants, Solna, Sweden, Oct 24-26; Swedish National Chemicals Inspectorate: 1989; pp 49-52.
(5) Bergman, r1. Proceedings from the Workshop on Brominated Aromatic Flame Retardants, Solna, Sweden, Oct 24-26; Swedish National Chemicals Inspectorate: 1989; pp 13-23.
(6) Sellsträm, U.; Kierkegaard, A.; De Wit, C.; Jansson, B. Environ. Toxicol. Chem. 1998, 17, 1065-1072.
(7) World Health Organization. Environmental Health Criteria 162: Brominated Diphenyl Ethers; WHO: Geneva, Switzerland, 1994.
(8) Organization for Economic Cooperation and Development. Risk Reduction Monograph 3: Selected Brominated Flame Retardants, Environmental Monograph Series 97; Paris, France, 1994.
(9) Bromine Science Environmental Forum. Major Brominated Flame Retardants Volume Estimates; Brussels, Belgium, 2000.
(10) Nordic Council of Ministers. Polybrominated Diphenyl Ethers: Food Contamination and Potential Risks; Report from an NNT project group; TemaNord: Copenhagen, Denmark, 1998;1998: 503.
(11) Sjödin, A.; Jakobsson, E.; Kierkegaard, A.; Marsh, G.; Sellström, U. J. Chromatogr. 1998, 822, 83-89.
(12) Pijnenburg, A. M. C. M.; Everts, J. W.; de Boer, J.; Boon, J. P. Rev. Environ. Contam. Toxicol. 1995, 141, 1-26.
(13) Meironyté, D.; Bergman, Å.; Norén, K. Organohalogen Compounds; Swedish Environmental Protection Agency: Stockholm, Sweden, 1998; Vol. 35.
(14) Haglund, P. S.; Look, D. R.; Buser, H.; Hu, J. Environ. Sci. Technol. 1997, 31, 3281-3287.
(15) Mathew, J.; Adler, D. M.; Klisz, A.; West, L.; Sonzogni, W. C. State Laboratory of Hygiene, University of Wisconsin, Madison, 1996.
(16) Jackson, L. J.; Carpenter, S. R.; Manchester-Neesvig, J.; Stow, C. A. J. Great Lakes Res. 1998, 24, 808-821.
(17) Asplund, L.; Athanasiadou, M.; Sjödin, A.; Bergman, Ä.; Börjeson, H. Ambio 1999, 28, 67-76.
(18) Marsh, G.; Hu, J.; Jakobsson, E.; Rahm, S.; Bergman, Ä. Environ. Sci. Technol. 1999, 33, 3033-3037.
(19) Asplund, L.; Hornung, M.; Peterson, R. E.; Turesson, K.; Bergman, Ä. Organohalogen Compounds, Swedish Environmental Protection Agency: Stockholm, Sweden, 1999; Vol. 40.
(20) Sergeant, D. B.; Alaee, M.; Luross, J.; Ikonomou, M. G. Organohalogen Compounds; Swedish Environmental Protection Agency: Stockholm, Sweden, 1998; Vol. 35.
(21) Stafford, C. J. Chemosphere 1983, 12, 1487-1495.
(22) Andersson, Ö.; Blomkvist, G. Chemosphere 1981,10,1051-1060.
(23) Jansson, B.; Andersson, R.; Asplund, L.; Litzén, K.; Nylund, K.; Sellström, U.; Uvemo, U.; Wahlberg, C.; Wideqvist, U.; Odsjö, T.; Olsson, M. Environ. Toxicol. Chem. 1993, 12, 1163-1174.
(24) Sellström, U.; Janssen, B.; Kierkegaard, A.; De Wit, C. Chemo sphere 1993, 26, 1703-1718.
(25) de Boer, J. Chemosphere 1989, 18, 2131-2140.
(26) de Boer, J.; Wester, P. G.; Klamer, H. J. C.; Lewis, W. E.; Boon, J. P. Nature 1998, 394, 28-29.
(27) Janssen, B.; Asplund, L.; Olsson, M. Chemosphere 1987,16, 2343 2349.
(28) Nylund, K.; Asplund, L.; Jansson, B.; Jansson, P.; Litzén, K.; Sellstr6m, U. Chemosphere 1992, 24, 1721-1730.
(29) Watanabe, L; Kashimoto, T.; Tatsukawa, R. Chemosphere 1987, 16, 2389-2396.
(30) Kuehl, D. W.; Haebler, R.; Potter, C. Chemosphere 1991, 22, 1071-1084.
(31) Loganathan, B. G.; Kannan, K.; Watanabe, L; Kawano, M.; Irvine, K.; Kumar, S.; Sikka, H. C. Environ. Sci. Technol. 1995, 29,18321838.
(32) Stow, C. A.; Jackson, L. J.; Amrhein, J. F. Can. ). Fish. Aquat. Sci. 1997, 54, 1031-1038.
Received for review June 27, 2000. Revised manuscript received November 27, 2000. Accepted January 4, 2001.
If you have come to this page from an outside location click here to get back to mindfully.org