Environ. Sci. Technol. 2007, 41, 3083-3089
Lipid Reserve Dynamics and Magnification of Persistent Organic Pollutants in Spawning Sockeye Salmon (Oncorhynchus nerka) from the Fraser River, British Columbia BARRY C. KELLY,† SAMANTHA L. GRAY,† M I C H A E L G . I K O N O M O U , * ,† J. STEVE MACDONALD,‡ STELVIO M. BANDIERA,§ AND EUGENE G. HRYCAY§ Contaminant Sciences, Institute of Ocean Sciences, Fisheries and Oceans Canada (DFO), 9860 West Saanich Road, Sidney, British Columbia, Canada, V8L 4B2, CRMI Resource and Environmental Management, Simon Fraser University, Burnaby, British Columbia, Canada, V5A 1S6, and Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, British Columbia, Canada V6T 1Z3
Pacific sockeye salmon (Oncorhynchus nerka) can travel several hundred kilometers to reach native spawning grounds and fulfill semelparous reproduction. The dramatic changes in lipid reserves during upstream migration can greatly affect internal toxicokinetics of persistent organic pollutants (POPs) such as PCBs, PCDDs, and PCDFs. We measured lipid content changes and contaminant concentrations in tissues (liver, muscle, roe/gonads) and biomarker responses (ethoxyresorufin O-deethylase or EROD activity and CYP1A levels) in two Pacific sockeye salmon stocks sampled at several locations along their spawning migration in the Fraser River, British Columbia. Muscle lipid contents declined significantly with increasing upstream migration distance and corresponded to elevated lipid normalized concentrations of PCBs and PCDD/Fs in spawning sockeye. Post-migration magnification factors (MFs) in spawning sockeye ranged between 3 and 12 and were comparable to model-predicted MFs. ΣPCBs (150-500 ng‚g-1 lipid), ΣPCDD/Fs (1-1000 pg‚g-1 lipid) and 2,3,7,8-TCDD toxic equivalent or TEQ levels (0.1-15 pg‚g-1 lipid) in spawning sockeye were relatively low and did not affect hepatic EROD activity/CYP1A induction. Despite a 3-fold magnification, TEQ levels in eggs of spawning Fraser River sockeye did not exceed 0.3 pg‚g-1 wet wt, a threshold level associated with 30% egg mortality in salmonids. PCBs in Fraser River sockeye are comparable to previous levels in Pacific sockeye. In contrast to Pacific sockeye from more remote coastal locations, PCDDs and PCDFs in Fraser River sockeye were generally minor components (75%). The data suggest that (i) the Fraser River is not * Corresponding author phone:
[email protected]. † Fisheries and Oceans Canada. ‡ Simon Fraser University. § University of British Columbia. 10.1021/es061559n CCC: $37.00 Published on Web 03/31/2007
(250)
363-6804;
2007 American Chemical Society
e-mail:
a major contamination source of PCBs or PCDD/Fs and (ii) marine contaminant distribution, food-chain dynamics, and ocean-migration pathway are likely important factors controlling levels and patterns of POPs in returning Pacific sockeye. We estimate an annual chemical flux entering the Fraser River of up to 150 g for ∑PCBs and 40 mg for ∑PCDD/ Fs via returning sockeye. The results indicate that historical concentrations of PCBs and PCDD/Fs remain a potential threat to organism and ecosystem health on the west coast of Canada.
Introduction Anadromous Pacific sockeye salmon (Oncorhynchus nerka) are important components of both marine and freshwater ecosystems on Canada’s west coast. British Columbia’s Fraser River, a major migration route and spawning grounds for several Pacific sockeye stocks, is one of Canada’s largest rivers, flowing over 1,400 km through agricultural, industrial, and urbanized lands. After hatching, juvenile sockeye tend to remain in fresh water for 1-2 years before a 2-3 year openocean feeding migration. During spring/summer adults return to their natal streams to spawn and die (i.e., semelparous reproduction). This upstream migration can span several hundred kilometers and demands a phenomenal investment of the organism’s energy requirements (1, 2). Pacific sockeye salmon (like all fish) generally accumulate persistent organic pollutants (POPs) such as polychlorinated biphenyls (PCBs), polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), and organochlorine pesticides (OCPs) through dietary exposure and to a lesser extent from water via respiratory uptake (i.e., bioconcentration). As >95% of their biomass is acquired from Pacific ocean prey consumption (1), POP residues in returning adult sockeye are primarily of marine origin. However, recent evidence indicates some Washington State salmon populations may acquire significant in-shore contaminant burdens due to increased residence time in the relatively industrialized Puget Sound (3). Other studies of POPs in returning Pacific salmon have demonstrated the ecotoxicological significance of the upstream migration phenomenon, including (i) biotransport and (ii) magnification and elevated toxicity of POPs during spawning (2, 4, 5). As adult Pacific sockeye do not feed after entering freshwater, energy reserves (invested mostly into gonad maturation) are typically depleted during their spawning migration (1, 6). When organisms utilize lipid reserves during periods of starvation or migration, chemical residues become concentrated and/or mobilized to other tissues/viscera via blood perfusion. Thus, while total wet weight body burdens of POPs remain constant, a magnification of lipid normalized concentrations ultimately occurs (2, 4, 7). Fraser River sockeye can deplete lipid reserves by as much as 90% following upstream migration (6). A significant magnification of lipid normalized contaminant concentrations in these spawning sockeye is therefore anticipated. This additional amplification of PCBs, PCDD/Fs (due to lipid reserve depletion) may be toxicologically significant for spawning salmon as numerous studies have demonstrated toxic effects in fish at concentrations comparable to field exposure levels, including impacts on olfactory function and migration behavior (8) and reproductive success (9-14). There have been recent concerns regarding the health of several Canadian Pacific sockeye stocks. For example, Weaver Creek sockeye (a late-run timing group) returning to the VOL. 41, NO. 9, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3083
FIGURE 1. Map showing sampling locations along the Fraser River migration route. Early Stuart sites are shown as rectangles and sites for Weaver run sites are shown as ovals. Fraser River have exhibited abnormal migration behavior, elevated in-river mortality, and reduced spawning success (15). In general, exposure and effects of contaminants in Fraser River sockeye have not been fully evaluated. In this paper, we report measurements of (i) lipid contents and tissue residue concentrations of PCBs, PCDDs, PCDFs, and 2,3,7,8TCDD toxic equivalents (TEQs), and (ii) biochemical responses (hepatic ethoxyresorufin O-deethylase (EROD) activity and cytochrome P450 (CYP1A) levels) in two Fraser River sockeye stocks (Early Stuart and Weaver Creek) sampled at several locations during their upstream migration in 2001. We also apply a simple steady-state model to predict postmigration magnification factors (MFs), which we then compare to observed MFs in these fish. The study aims to assess the influence of lipid reserve depletion on contaminant magnification and related toxicological implications for spawning Fraser River sockeye salmon.
Materials and Methods Sample Collections. Figure 1 shows sampling locations along the Fraser River migration route. Female and male fish were caught by seine, gill-net, or dip-net between June and October, 2001 (Table 1). Dorsal muscle, liver, and gonads/ roe were excised from the fish. Early Stuart (ES) sockeye were caught at five locations: Port Renfrew (PR), Whonnock (Whon), Yale (Y), Hell’s Gate (HG), and Gluskie Creek (GC), over ∼1200 km total distance. Weaver Creek sockeye were caught at two sites, Harrison River (HR) and ∼10 km upstream on their spawning grounds at Weaver Creek (WC). Age and stock origin were determined by the Pacific Salmon Commission using scale analysis, which has proven a highly accurate technique for these BC sockeye populations (16). Fish were 4 or 5 years of age. Contaminant and lipid content analyses were determined on individual and composite tissue samples. Composites were of fish of the same age. CYP1A 3084
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 9, 2007
levels and EROD activity were determined on liver samples from individual male and female fish from Yale, Gluskie Creek, Harrison River, and Weaver Creek. Contaminant and Lipid Content Analysis. Details regarding our methodology for extraction, cleanup, and quantification of target analytes (i.e., PCBs, PCDDs/Fs) by gas chromatography/high-resolution mass spectrometry (GC-HRMS) are reported elsewhere (4, 17, 18) and summarized in the Supporting Information (SI). Tissue lipid contents were determined gravimetrically using an extracted 5 g subsample (wet weight). Lipid contents (% lipid) were expressed as a percentage of the original wet tissue weight. Biochemical Response Measurements. Hepatic microsomes were prepared by differential centrifugation from individual fish. Microsomal pellets were suspended in 0.25 M sucrose and aliquots were stored at -75 °C prior to analysis. Microsomes were analyzed for total CYP content, EROD activity, and CYP1A protein levels. Total cytochrome P450 was determined from the carbon monoxide difference spectrum, and protein concentration was measured by the method of Lowry et al. (19). EROD activity was measured using a spectrofluorometric assay as described by Burke et al. (20). Immunoblotting and densitometric quantitation of CYP1A was done using polyacrylamide gel electrophoresis (PAGE) as described by Laemmli (21). Samples were applied to the gels at 20-40 µg protein per lane. Microsomal proteins resolved by SDS-PAGE were transferred electrophoretically onto nitrocellulose and probed with antibody as described previously (22). Membranes were incubated with rabbit antitrout CYP1A peptide IgG at a concentration of 10 µg‚mL-1 (23). Bound primary antibody was located using alkaline phosphatase-conjugated goat anti-rabbit IgG. Immunoreactive proteins were detected by reaction with a substrate solution containing 0.01% nitro blue tetrazolium, 0.05% 5-bromo-4-chloro-3-indolyl phosphate, and 5 mM MgCl2 in 0.1 M Tris-HCl buffer, pH 9.5. Assay conditions were optimized to ensure that color development did not proceed beyond the linear range of the phosphatase reaction. Staining intensities of the immunoreactive bands were quantified with a pdi 420 oe scanning densitometer integrated with Quantity One v. 3 software (pdi Inc., Huntington Station, NY). The amount of immunoreactive protein was determined from the integral of the optical density of the stained band. Staining intensities of bands on each blot were normalized with a purified rat hepatic CYP1A1. A purified rat CYP1A1 sample was included on every gel as an internal standard to correct for blot-to-blot and day-to-day variability. The CYP1A protein bands were expressed as relative optical density per mg protein, i.e., ROD (mg protein)-1 rather than in terms of microsomal content (e.g., pmol‚mg-1) because purified salmon CYP1A enzyme is not available for use as a calibration standard for quantification. Data Analysis and Statistics. Concentration data are reported as ng‚g-1 lipid for PCBs and pg‚g-1 lipid for PCDDs, PCDFs, and 2,3,7,8-TCDD toxic equivalents (TEQs). TEQs were determined for individual PCDD/Fs and PCBs using World Health Organization toxic equivalency factors (WHOTEFs) developed for assessing 2,3,7,8-TCDD toxicity in fish (24). One-way analyses of variance (ANOVA) and Tukey’s HSD comparison tests were performed to evaluate differences between chemical concentrations or biomarker responses observed in pre- and post-migration sockeye. Magnification Factors (MFs). Observed migration-related magnification factors (MFs) for a given tissue i were determined as the ratio of lipid normalized concentrations observed at post- and pre-migration locations (i.e., MFi ) Ci,post/Ci,pre, lipid weight). Also, using the model presented by Debruyn et al. (4) we predicted MFs using measured preand post-migration tissue weights (Wi) and corresponding pre- and post-migration lipid fractions (Li). We used the
TABLE 1. Mean Concentrations of ∑PCB (ng‚g-1 Lipid), and ∑PCDD, ∑PCDF, and ∑TEQs (pg‚g-1 Lipid) in Tissues of Early Stuart and Weaver Sockeye Salmon during Their Upstream Migration in 2001 (* Indicates Significantly Higher Concentrations (p < 0.05) in Spawning Salmon Compared to Pre-spawning Locations; NA ) Not Analyzed) method
date
sex
tissue
% lipid
n
10.9 12.7
0.25 ( 0.02 0.14 ( 0.05 0.60
NA NA NA
NA NA NA
NA NA NA
NA NA NA
96.8 ( 18.1 54.3 81.4 ( 22.8 138.4 ( 63.4 111.6 ( 50.9 119.5 ( 13.7
7.57 7.95 3.50 6.90 ( 1.06 5.75 ( 0.59 36.4 ( 7.04
14.4 9.74 5.76 12.6 ( 2.36 12.0 ( 0.53 14.4 ( 0.91
0.26 ( 0.11 0.22 0.16 ( 0.04 0.43 ( 0.23 0.82 ( 0.004 0.13 ( 0.08
NA NA NA
NA NA NA
NA NA NA
NA NA NA
422.3 ( 102.1* 275.6 ( 124.7* 160.9 ( 17.6* 574.8 ( 317.9* 326.4 ( 134.6*
37.5 * 35.8 ( 34.4* 11.9 ( 0.32* 27.7 ( 21.4* 23.5 ( 3.1*
9.56 ( 13.5 10.2 ( 2.34* 30.3 ( 7.18* 36.4 ( 13.5*
1.15 ( 0.21* 1.02 ( 0.86* 0.42 ( 0.21* 2.36 ( 1.28* 5.02 ( 1.0*
Weaver Creek Run 1.38 ( 0.21 (n )3) 3 495.7 ( 473.9 2.92 ( 0.20 (n )2) 2 135.6 ( 47.7 10.27 ( 0.9 (n)2) 2 133 ( 27.1 1.08 ( 0.5 (n ) 2) 2 335.3 ( 148.5 2.76 ( 0.06 (n )2) 2 125.3 ( 72.6 0.22 ( 0.01 (n )2) 2 503.5 ( 148.3
123.5 ( 156.4 24.0 ( 10.8 26.7 ( 27.2 983.2 ( 1361 1149 ( 1600 433.1 ( 40.7
4.76 ( 6.73 3.06 ( 4.32 7.46 ( 1.88 50.7 78.3
5.48 ( 8.13 0.38 ( 0.28 0.89 ( 0.73 14.7 ( 19.6 0.80 ( 0.78 1.07 ( 1.15
31.7 25.07 ( 5.48 8.61 ( 2.39 31.6 110.5
20.8 ( 14.5 18.7 ( 2.7 5.26 ( 0.24 10.6 24.4
3.05 ( 2.85 2.98 ( 1.18 0.29 ( 0.02 1.17 6.03
Whonnockb
gill-net July 5
F F M
muscle 7.4 ( 0.8 (n ) 6) 6 roe 13.0 ( 0.2 (n ) 5) 5 muscle 13.0 ( 0.2 (n ) 6) 5
Yalea
dip-net July 10
F F F M M M
muscle 4.0 ( 1.8 (n ) 3) liver 6.2 ( 1.0 (n ) 3) roe 12.4 ( 0.8 (n ) 6) muscle 5.0 ( 1.6 (n ) 9) liver 10.0 ( 2.5 (n ) 3) gonad 0.9 ( 0.2 (n ) 8)
Hell’s Gatea
dip-net July 12
F F M
muscle 3.2 ( 1.5 (n ) 6) 5 roe 13.6 ( 1.8 (n ) 3) 3 muscle 3.1 ( 0.1 (n ) 6) 5
Gluskie Creeka
dip-net July 31
F F F M M
muscle liver roe muscle liver
F F F M M M
muscle liver roe muscle liver gonad
F F F M M
muscle 1.45 ( 0.23 (n )2) liver 5.16 ( 0.47 (n )3) roe 9.40 ( 0.01 (n )2) muscle 1.12 (n ) 1) gonad 0.86 (n ) 1)
Weaver Creeka
Sept 20
dip-net Oct 15
soma c roe gonad
ΣPCDFs ΣTEQ (pg‚g-1 lipid) (pg‚g-1 lipid)
4.02 3.37
seine
Harrison Riverb seine
ΣPCDDs (pg‚g-1 lipid)
Early Stuart Run 15.5 ( 0.2 (n ) 3) 3 85.4 ( 29.1 13.3 ( 1.2 (n ) 3) 4 72.7 ( 16.8 0.65 (n ) 1) 1 176.4
Port Renfrewa
June 29 F F M
ΣPCBs (ng‚g-1 lipid)
1.2 ( 0.2(n ) 5) 4.5 ( 0.7(n ) 2) 8.0 ( 0.9(n ) 3) 1.9 ( 0.6 (n ) 5) 5.5 ( 0.8(n ) 3)
5 1 5 6 2 2
4 2 4 3 2
2 3 2 1 1
365.3 ( 134.2 182.2 ( 25.8 94.3 ( 3.79 360.2 429.8
a Samples consisted of individual fish and multiple composites (2-3 pooled fish per analysis). per analysis). c Soma represents whole fish minus roe/gonads.
model to predict MFs for two scenarios of chemical redistribution kinetics (i.e., slow versus rapid redistribution). For relatively slow internal chemical redistribution, each tissue (i) is viewed as an isolated compartment and the lipid normalized MF for each tissue is calculated as
b
Multiple composite samples (2-3 pooled fish
(2)
and Gluskie Creek. This corresponds to a relative lipid loss in muscle and roe equal to 84% and 38%, respectively. ES female sockeye appear to have a greater degree of lipid loss compared to males, which may be due to a greater need for lipid during roe production. This is supported by the fact that lipid content in ES female roe only drops slightly during the upstream migration (Table 1). For Weaver sockeye, no significant reductions in lipid content were detected between Harrison River and the nearby spawning grounds at Weaver Creek. However, muscle tissue lipid contents in Weaver sockeye were low (∼1%) at these locations and were comparable to those observed in ES sockeye at their spawning grounds. The data suggest lipid losses in Weaver fish likely occurred prior to sampling in the lower Fraser River. In general, our observations of decreasing lipid reserves are in agreement with previous studies of returning Pacific sockeye salmon (2, 4, 6) and other migratory fish (25).
Lipid Reserve Depletion in Returning Sockeye Salmon. Table 1 summarizes lipid contents (% lipid) observed in the salmon tissues from Early Stuart and Weaver Creek sockeye during their 2001 upstream migration. Early Stuart sockeye showed significant reductions (p < 0.05) of lipid content in muscle, liver, and roe/gonads during migration to their spawning grounds at Gluskie Creek (∼1200 km). Lipid content in female sockeye decreased from 7.4% to 1.3% in muscle tissue and from 13.3% to 8.0% in roe between Whonnock
Levels and Congener Profiles of PCBs, PCDDs, and PCDFs. Table 1 summarizes ∑PCB, ∑PCDD, and ∑PCDF concentrations and corresponding TEQ levels in the various tissues of returning Fraser River sockeye salmon. Concentrations of individual congeners are available in Table S3 in the SI. ∑PCB concentrations ranged between 100 and 500 ng‚g-1 lipid in female muscle tissue and from 80 to 160 ng‚g-1 lipid in roe. ∑PCBs in male sockeye ranged between 140 and 600 ng‚g-1 lipid in muscle and 120 and 500 ng‚g-1 lipid in the gonads. Tetrachloro (Cl4)-heptachloro (Cl7) PCB homologues were dominant in sockeye tissues, typically comprising 90%
MFi ) Ci,post/Ci,pre ) Wi,pre‚Li,pre ÷ Wi,post‚Li,post
(1)
For relatively rapid chemical redistribution, a high degree of chemical exchange results in equivalent lipid normalized concentrations between tissues (e.g., CMUSCLE ) CLIVER ) CROE) and hence the lipid normalized MFs in the various tissues are equivalent. MFs for this second scenario are for whole body (i.e., MFB) and are determined from the sum of all tissue weights (Wi) and lipid contents (Li):
MFB ) CB,post/CB,pre )
∑W
i,pre‚Li,pre
÷
∑W
i,post‚Lipost
Results and Discussion
VOL. 41, NO. 9, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3085
of ∑PCBs. Predominant congeners included Cl3-CB28, Cl3CB31, Cl4-CB52, Cl5-CB95, Cl5-CB101, Cl5-CB110, Cl5-CB118, Cl6-CB149, Cl6-CB153, Cl6-CB138, Cl7-CB187/182, and Cl7CB180. Congener specific R153 ratios (i.e., CBX/CB153) observed in muscle, liver, and roe/gonads of male and female sockeye at pre-spawning sites (Yale and Harrison River) and spawning grounds (Gluskie Creek and Weaver Creek) are summarized in Table S3 and Figure S1 (Supporting Information). The R153 data show PCB congener profiles are generally comparable among different tissues and between locations. Specifically, dominant Cl4-Cl7 congeners (e.g., CB52, CB110, CB118, CB138, CB153, CB180, etc.) exhibit a similar pattern among tissues and between pre- and post-migration sites. However, in some cases there appears an enrichment of lower chlorinated congeners (i.e., Cl2-Cl5 CBs) such as CB8/5, CB18, CB28, CB31, CB52, CB95, and CB101. For example, lower chlorinated CBs exhibit relatively high R153 values in female roe compared to muscle and liver, especially at postmigration sites (Figure S1). A similar enrichment of Cl2-Cl5 CBs is apparent in male gonads compared to liver and muscle. This indicates that lower chlorinated congeners, which tend to exhibit faster internal kinetics compared to higher chlorinated congeners, may be transferred more efficiently from muscle/liver into roe/gonads. Further, this may be particularly true during periods of migration. It should be noted that pattern changes involving Cl2 and Cl3 CBs are to be viewed with some caution due to the potential inaccuracy associated with quantification of those congeners. Specifically, low recoveries of those more volatile CBs during standard sample extraction and cleanup procedures can often elicit less accurate quantification, even when using isotopelabeled internal surrogates (See SI for details). ∑PCDD and ∑PCDF concentrations were very low (and often not detectable), with levels approximately 1000 times lower than ∑PCBs, ranging between 7 and 100 pg‚g-1 lipid in female tissues and 10-1000 pg‚g-1 lipid in male tissues. Levels and frequency of detection of PCDDs and PCDFs in Fraser River sockeye (present study) were somewhat lower than those reported in Pacific sockeye returning to Great Central Lake (GCL) on Vancouver Island, British Columbia (4). For example, 2,3,4,7,8-PeCDF, which was routinely observed in Vancouver Island Sockeye, only was occasionally detected in Fraser River sockeye tissue samples. While this relatively potent PCDF congener (WHO-TEF for fish ) 0.5) does coelute with three other less toxic PCDFs (i.e., 1,2,4,8,9 + 1,3,4,8,9 + 1,2,3,6,9-PeCDFs) when using a standard DB-5 GC capillary column, further separation of these PCDFs on additional columns (e.g., DB-DIOXIN, DB5-ms, and Cp Sil 88) generally show that 2,3,4,7,8-PeCDF is by far the dominant compound in environmental and biological samples (See SI for details). In contrast, PCB levels in Fraser River sockeye were comparable to those levels reported in Vancouver Island sockeye. The fact that contaminant concentrations in sockeye from the more industrialized Fraser River system are comparable to (and in some cases lower than) levels observed in sockeye from more remote coastal regions (i.e., Vancouver Island) suggests historical PCB and PCDD/F loadings in the Fraser River are not a major contaminant source for these migrating fish. 2,3,7,8-TCDD Toxic Equivalents (TEQs). Table 1 shows total 2,3,7,8-TCDD toxic equivalents (i.e., ∑TEQs, hereafter referred as TEQs represent the sum of individual PCB, PCDD, and PCDF TEQs) for Early Stuart and Weaver Creek sockeye salmon. TEQs ranged between 0.2 and 5.5 pg‚g-1 lipid in female muscle and 0.4 and 15 pg‚g-1 lipid in male muscle. For Early Stuart sockeye, TEQs in female roe were 0.14 pg‚g-1 lipid at Port Renfrew, 0.16 pg‚g-1 lipid at Yale, and 1.15 pg‚g-1 lipid at Gluskie Creek. Similarly for Weaver sockeye, TEQs in female roe ranged from 0.29 to 0.89 pg‚g-1 lipid at Weaver 3086
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 9, 2007
Creek and Harrison River, respectively. Dioxin-like PCB congeners typically exhibited 85-95% of the total TEQ in the sockeye tissue samples. PCB126 typically contributed 6080% of the total ∑PCB-TEQ, while the overall rank order for TEQ contribution of PCBs was PCB126 > PCB118 >PCB105 > PCB156 > PCB169. TEQs in Fraser River sockeye are lower than TEQs reported in Vancouver Island sockeye (4). The primary reason for higher TEQs in Vancouver Island sockeye was the presence of 2,3,4,7,8-PeCDF, which typically contributed 40-50% of TEQ in those fish. In the present study, 2,3,4,7,8-PeCDF was generally not a major component of TEQ and hence dioxin like PCBs typically was the majority (>75%) of the total TEQ. Thus, in contrast to Vancouver Island sockeye, it appears Fraser River sockeye do not significantly accumulate this relatively toxic PCDF congener at any stage of their life history. Magnification of PCBs, PCDDs, and PCDFs in Spawning Sockeye Salmon. ∑PCB, ∑PCDD, and ∑PCDF concentrations increased significantly (p < 0.05) in Early Stuart sockeye at their spawning grounds compared to pre-spawning locations. ∑PCBs in female muscle increased from 96.8 ( 18.1 ng‚g-1 lipid at Yale to 422.3 ( 102 ng‚g-1 lipid at Gluskie Creek. Similarly, ∑PCBs in male muscle increased from 138.4 ( 63.4 ng‚g-1 lipid at Yale to 574.8 ( 318 ng‚g-1 lipid at Gluskie Creek. Elevated TEQs were also observed in spawning ES sockeye, increasing from 0.26 ( 0.11 to 1.15 ( 0.21 pg‚g-1 lipid in female ES sockeye muscle. TEQs increased from 0.43 ( 0.23 to 2.36 ( 1.28 pg‚g-1 lipid in male ES muscle. No significant increases of PCBs, PCDD/Fs, or TEQs were detected in Weaver sockeye. Observed lipid normalized magnification factors (MFs) of PCBs, PCDD/Fs were between 2 and 12 in spawning ES sockeye (Table S4 in the SI). Following migration from Port Renfrew to Gluskie Creek MFs of PCB153 in spawning ES female sockeye muscle, liver, and roe were 5.8, 3.9, and 2.1, respectively. PCB153 MFs in ES male sockeye were 7.9 in muscle. Comparable post-migration MFs (between 1.9 and 9.7) were reported for various PCB and PCDD/F congeners in Pacific sockeye salmon from coastal British Columbia (4). Similarly, PCB data for returning Alaskan sockeye salmon (2) correspond to MFs of 3.7 for muscle and 2.1 for roe. The cause of this magnification is likely the result of lipid reserve depletion during upstream migration (see Table 1). Numerous studies have shown the depletion of organism lipids reduces the overall sorptive capacity for lipid soluble organic contaminants, causing a concentrating effect and hence elevated lipid normalized tissue concentrations (2, 4, 7, 26). This contaminant magnification mechanism (i.e., due to lipid reserve depletion) can occur only if chemical elimination kinetics are very slow compared to the rate of biomass loss in the organism. Chemical accumulations in fish are delineated by competing rates of chemical uptake and elimination, including uptake via food (kD) and water through the gills (k1), and elimination via feces (kE), metabolic transformation (kM), and water through the gills (k2). Debruyn et al. (4) presented a simple model for quantifying PCB and PCDD/Fs kinetics and magnification in upstream migrating salmon. During migration, feeding and fecal elimination equal zero (kD and kE ) 0). Also, chemical exchange between fish and water (gill uptake ) k1, gill elimination ) k2) and metabolic transformation (kM) are negligible for recalcitrant PCBs, PCDD/Fs. Tissue residue concentrations (Ci) are then largely determined by the degree of biomass loss and subsequent redistribution between tissues. Figure 2 illustrates the observed changes in tissue weight (Wi) and lipid contents (Li) and corresponding concentration and tissue residue burdens of PCB153 in muscle, liver, and roe for a migrating female sockeye salmon between Port Renfrew (ocean) to Yale (∼180 km) and arriving at Gluskie Creek (spawning grounds). Following migration
FIGURE 2. Illustration showing body composition changes and tissue distribution/burdens of PCB 153 in muscle, liver, and roe of migrating female sockeye salmon at Port Renfrew, Yale, and Gluskie Creek. Numbers in shaded areas represent percent lipid (Li). Numbers above tissue compartments represent grams of wet weight tissue (Wi). Numbers inside tissue compartments represent lipid normalized chemical concentration (Ci , ng‚g-1 lipid) and tissue burdens (ng) in brackets. aNote: Data for muscle and liver at Port Renfrew represent measured concentrations for whole body (minus gonads). these female sockeye exhibit a 30% decrease in muscle tissue weight and a 330% increase in roe weight. Post-migration lipid content changes in muscle are substantial (8-1% decrease), while lipid content changes in roe are relatively small (13-8% decrease). By applying the magnification model to these female sockeye for a recalcitrant PCB congener (PCB153), assuming rapid internal chemical redistribution (Scenario 1), the predicted magnification factor (MF) is 3.5 for muscle, liver, and roe, which is comparable to observed PCB153 MFs in those tissues (2.1-5.8). Assuming no chemical exchange between tissues (Scenario 2), the predicted PCB 153 MF was 10.9, 2.0, and 0.5 in muscle, liver, and roe, respectively. The apparent over-estimation of MFs in muscle and the under-estimation of MFs in liver and roe (for this latter modeling scenario) suggest that redistribution of POPs from muscle to other tissues and roe likely occurs in these migrating salmon. This is supported by the fact these salmon experienced a 370 ng loss of PCB153 from muscle tissue and a corresponding positive chemical flux into roe (240 ng) and liver (10 ng) between Port Renfrew and Gluskie Creek. Transfer of POPs from adult to eggs has been demonstrated in salmon (2, 11), other fish species (9-11), and birds (7). Russell et al. (27) argued that lipid normalized POP concentrations in muscle and roe of oviparous organisms should be similar (i.e., CMUSCLE /CROE ) 1.0). However, the lipid normalized muscle/egg ratios in the present study were typically between 2 and 5 (Table S3). The muscle/roe ratio for PCB153 in female sockeye at Gluskie Creek was 3.2. The data suggest that while mobilization and redistribution of POPs can occur during periods of lipid reserve depletion, the relatively slow kinetics may never result in a chemical equilibrium between tissues and eggs. Regardless, there remains a substantial degree of maternal transfer of POPs to eggs during the spawning migration. For example, elevated TEQs in roe of females at Gluskie Creek (0.42 pg‚g-1 lipid) corresponds to a tissue burden of 10.7 pg, which is 43% of their total TEQ body burden (25 pg) during spawning.
FIGURE 3. CYP1A protein content (ROD‚(mg protein)-1 and EROD activity (pmole‚min-1‚mg-1) for sockeye salmon from the Weaver and Early Stuart runs. Asterisk (*) indicates significantly greater ROD values (p < 0.05) in spawning sockeye compared to prespawning fish. Toxicological Implications. Figure 3 shows hepatic CYP1A levels and EROD activity for spawning and pre-spawning sockeye from the Early Stuart and Weaver Creek sockeye runs. No significant changes in EROD activity were observed in fish between pre- and post-spawning locations. Significant increases (p < 0.05) in CYP1A levels were observed in spawning sockeye compared to pre-spawning fish. For example, the CYP1A levels in female salmon from the Early Stuart run increased from 0.8 ROD (mg protein)-1 at a prespawning site (Yale) to 1.8 ROD (mg protein)-1 at Gluskie VOL. 41, NO. 9, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3087
FIGURE 4. Observed TEQs (pg‚g-1 lipid) in female sockeye eggs sampled from Early Stuart and Weaver Creek sockeye salmon. Bars are arithmetic mean TEQs and error bars represent standard deviations (SD). Horizontal lines represent TEQs concentrations in salmonid eggs associated with 30% egg mortality (14) and endocrine disruption effects in seals (18). TEQs in roe of Vancouver Island sockeye are from ref 4. Data from the Great Lakes between 1986 and 1997 include Chinook salmon (Oncorhynchus tshawytscha) from Lake Michigan (30), Lake trout (Salvelinus namaycush) from Lake Michigan (29), and Lake trout from Lakes Erie, Superior, Huron, and Ontario (28). British Columbia killer whale data are from ref 18. Creek. However, the fact that there was no significant enhancement of EROD activity indicates there was no effects of aryl hydrocarbon receptor (AhR) agonists during the spawning migration. The increased CYP1A levels in these spawning sockeye salmon may be the result of depleted somatic protein (∼50% decline), (6), hence concentrated CYP1A levels. Figure 4 shows that a 3-fold magnification of lipid normalized TEQs occurs in Early Stuart female sockeye eggs from 0.14 pg‚g-1 lipid (0.018 wet wt) in the open ocean to 0.42 pg‚g-1 lipid (0.032 wet wt) during spawning. Despite this magnification and increased contaminant burden in roe, egg TEQs remained below the 30% mortality threshold level of 0.3 pg‚g-1 wet wt previously reported in salmonid eggs (14). Previous studies of Vancouver Island sockeye observed similar magnification of egg TEQs (∼3 fold), which ultimately were elevated to toxicologically significant concentrations (e.g., 0.3 pg‚g-1 wet wt in roe at Robertson Creek) (4), primarily due to a high TEQ contribution from 2,3,4,7,8-PeCDF. In the present study, the detection of 2,3,4,7,8-PeCDF in roe in Weaver Creek sockeye at Harrison River was the reason for the relatively high TEQs in those fish compared to other Fraser River sockeye samples. The data indicate that the presence and preferential accumulation of relatively more toxic PCDD/F congeners (such as 2,3,4,7,8-PeCDF exhibiting high TEFs) may be key for achieving toxicological thresholds associated with egg mortality and reproductive success of BC sockeye populations. Figure 4 shows TEQs in Pacific salmon roe from the west coast of Canada are substantially lower than observations in salmonids from the Great Lakes (100-2000 pg‚g-1 lipid), (28-30), which may have caused significant population declines during the 1980s (10). Also, TEQs in Fraser River sockeye were 50-100 times lower than TEQs we reported in killer whales (Orcinus orca) from coastal British Columbia (which surpassed endocrine disruption 3088
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 9, 2007
effect levels observed in seals, ∼255 pg‚g-1 lipid; 18). Although contaminant concentrations in salmon from this study did not exceed adverse effect levels, historical loadings of PCBs and PCDD/Fs remain a potential threat to the health of this coastal marine food web. Our findings indicate that lipid reserve depletion during upstream migration can cause a significant magnification of PCB and PCDD/F concentrations in tissues of spawning Fraser River sockeye salmon. The internal kinetics of POPs during spawning migration appears to be relatively rapid and can result in chemical exchange between tissues and eggs. For spawning Fraser River sockeye, the resulting 2,3,7,8TCDD toxic equivalent concentrations in developing embryos are below threshold levels for the onset of egg survival effects. The Fraser River does not appear to be a major source of these PCB and PCDD/F residues in these returning sockeye. However, based on observed concentrations (Table 1) and annual return estimates for Early Stuart and Weaver sockeye (100,000-500,000 fish/year), we roughly estimate annual upstream biotransport fluxes of 0.6-3 g‚yr-1 for ∑PCB and 0.1-0.4 mg‚yr-1 for ∑PCDD/Fs into the Fraser River. These flux estimates could be 50 times higher, considering the numerous other stocks returning to the Fraser each year (i.e., 20-30 million fish/year). Chemical fate and distribution studies are needed to better assess the impacts of this POP biotransport mechanism for salmon eggs/fry and other resident organisms of the Fraser River ecosystem. It is also important to note that other globally distributed dioxin-like compounds and/or endocrine disrupting chemicals (EDCs) such as polychlorinated napthalenes (PCNs), polybrominated diphenyl ethers (PBDEs), hexachlorocyclohexanes (HCHs), endosulfans, dialkyl phthalate esters (DPEs), and perfluoroalkyl compounds (PFCs) are undoubtedly present in tissues of these salmon. Similar magnification and maternal transfer of these compounds during spawning may
therefore enhance toxicological impacts on returning adults and early life stage development. Unlike legacy pollutants such as PCBs and PCDD/Fs, many of these other currentuse chemicals of concern can be extensively discharged into urban/agricultural receiving waters such as the Fraser River. Our future research on Fraser River sockeye will therefore focus on accumulation patterns and the cumulative and/or synergistic effects of PCBs, PCDD/Fs, and other EDCs of emerging concern on the reproductive health and population dynamics of these vital Pacific salmon populations.
Acknowledgments We gratefully acknowledge Dave Barnes, Dave Patterson, and Cory Dubetz for sample collection, the DFO Regional Contaminants Laboratory staff for sample analyses and technical assistance, and Pacific Salmon Commission, DFO, and the National Sciences and Engineering Research Council of Canada (NSERC) for financial support. We also thank the five anonymous reviewers for their insightful evaluations of this paper.
Supporting Information Available Supporting text, Tables S1, S2, S3, and S4, and Figure S1. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Groot, C., Margolis, L. Eds. Pacific Salmon Life Histories; UBC Press: Vancouver, BC, 1991. (2) Ewald, G.; Larsson, P.; Linge, H.; Okla, L.; Szarzi, N. Biotransport of organic pollutants to an inland Alaska lake by migrating sockeye salmon (Oncorhynchus nerka). Arctic 1998, 51, 40-47. (3) Missildine, B.; Peters, R.; Chin-Leo, G.; Houck, D. Polychlorinated biphenyl concentrations in adult chinook salmon (Oncorhynchus tshawytscha) returning to coastal and Puget Sound hatcheries of Washington State. Environ. Sci. Technol. 2005, 39, 6944-6951. (4) Debruyn, A. M.; Ikonomou, M. G.; Gobas, F. A. Magnification and toxicity of PCBs, PCDDs, and PCDFs in upriver-migrating Pacific salmon. Environ. Sci. Technol. 2004, 38, 6217-6224. (5) Krummel, E. M.; Macdonald, R. W.; Kimpe, L. E.; Gregory-Eaves, I.; Demers, M. J.; Smol, J. P.; Finney, B. Blais, J. M. Aquatic ecology: delivery of pollutants by spawning salmon. Nature 2003, 425, 255-256. (6) Gilhousen, P. Energy sources and expenditures in Fraser River sockeye salmon during their spawning migration. Int. Pac. Salmon Fish Commission Bull. 1980, 22, 1-51. (7) Findlay, G. M.; DeFreitas, A. S. M. DDT movement from adipocyte to muscle cell during lipid utilization. Nature 1971, 229, 63-65. (8) Moore, A.; Waring, C. P. Sublethal effects of the pesticide Diazinon on olfactory function in mature Atlantic salmon. J. Fish Biol. 1996, 48, 758-775. (9) Guiney, P. D.; Melancon, M. J., Jr.; Lech, J. J.; Peterson, R. E. Effects of egg and sperm maturation and spawning on the distribution and elimination of a polychlorinated biphenyl in rainbow trout (Salmo gairdneri). Toxicol. Appl. Pharmacol. 1979, 47, 261-272. (10) Niimi, A. J. Biological and toxicological effects of environmental contaminants in fish and their eggs. Can. J. Fish Aquat. Sci. 1983, 40, 306-312. (11) Miller, M. A. Maternal transfer of organochlorine compounds in salmonines to their eggs. Can. J. Fish. Aquat. Sci. 1993, 50, 1405-1413. (12) Walker, M. K.; Peterson, R. E. Potencies of polychlorinated dibenzo-p-dioxin, dibenzofuran, and biphenyl congeners, relative to 2,3,7,8-tetrachlorodibenzo-p-dioxin, for producing early life stage mortality in rainbow trout (Oncorhynchus mykiss). Aquat. Toxicol. 1991, 21, 219-238. (13) Zitko, V.; Saunders, R. L. Effect of PCBs and other organochlorine compounds on the hatchability of Atlantic salmon (Salmo salar) eggs. Bull. Environ. Contam. Toxicol. 1979, 21, 125-130.
(14) Giesy, J. P.; Jones, P. D.; Kannan, K.; Newsted, J. L.; Tillitt, D. E.; Williams, L. L. Effects of chronic dietary exposure to environmentally relevant concentrations to 2,3,7,8-tetrachlorodibenzo-p-dioxin on survival, growth, reproduction and biochemical responses of female rainbow trout (Oncorhynchus mykiss). Aquat. Toxicol. 2002, 59, 35-53. (15) Pacific Salmon Commission, Early migration of late-run Fraser River sockeye salmon; Meeting notes; Vancouver, BC, 2001. (16) Mike Lapointe, Pacific Salmon Commission, Vancouver, Canada; personal communication. (17) Ikonomou, M. G.; Fraser, T. L.; Crewe, N. F.; Fischer, M. B.; Rogers, I. H.; He, T.; Sather, P. J.; Lamb, R. L. A comprehensive multiresidue ultra trace analytical method based on HRGC/ HRMS, for the determination of PCDDs, PCDFs, PCBs, PBDEs, PCDEs, and organochlorine pesticides in six different environmental matrices. Can. Tech. Rep. Fish. Aquat. Sci. 2001, 2389, vii + 95 pp. (18) Ross, P. S.; Ellis, G. M.; Ikonomou, M. G.; Barrett-Lennard, L. G.; Addison, R. F. High PCB concentrations in free ranging killer whales (Orcinus orca): effects of age, sex, and dietary preference. Mar. Pollut. Bull. 2000, 40, 504-515. (19) Lowry, D.; Rosebrough, N.; Farr, A. Randall, R. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 1951, 193, 265-275. (20) Burke, M. T.; Thompson, S.; Elcombe, R. R.; Halpert, J.; Haaparanta, T. Mayer, R. T. Ethoxy-, pentoxy- and benzyloxyphenoxazones and homologues: a series of substrates to distinguish between different induced cytochromes P450. Biochem. Pharmacol. 1985, 34, 3337-3345. (21) Laemmli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970, 207, 680-685. (22) Towbin, H.; Staehelin, T.; Gordon, J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedures and some applications. PNAS 1979, 76, 4350-4354. (23) Lin, S.; Bullock, P. L.; Addison, R. F. Bandiera, S. M. Detection of cytochrome P4501A in several species using antibody against a synthetic peptide derived from rainbow trout cytochrome P4501A1. Environ. Toxicol. Chem. 1998, 17, 439-445. (24) Van den Berg, M.; Birnbaum, L.; Bosveld, A. T.; Brunstrom, B.; Cook, P.; Feeley, M.; Giesy, J. P.; Hanberg, A.; Hasegawa, R.; Kennedy, S. W.; Kubiak, T.; Larsen, J. C.; van Leeuwen, F. X.; Liem, A. K.; Nolt, C.; Peterson, R. E.; Poellinger, L.; Safe, S.; Schrenk, D.; Tillitt, D.; Tysklind, M.; Younes, M.; Waern, F. Zacharewski, T. Toxic equivalency factors (TEFs) for PCBs, PCDDs, PCDFs for humans and wildlife. Environ. Health Perspect. 1998, 106, 775-792. (25) Crawford, R. H.; Cusack, R. R.; Parlee, T. R. Lipid content and energy expenditure in the spawning migration of alewife (Alosa psuedoharengus). Can. J. Zool. 1986, 64, 1902-1907. (26) Addison, R. F. Organochlorine compounds and marine lipids. Prog. Lipid Res. 1982, 21, 47-71. (27) Russell, R. W.; Gobas, F. A. P. C. Haffner, G. D. Maternal transfer and in ovo exposure of organochlorines in oviparous organisms: A model and field verification. Environ. Sci. Technol. 1999, 33, 416-420. (28) Hickey, J. P.; Batterman, S. A. Chernyak, S. M. Trends of chlorinated organic contaminants in Great Lakes trout and walleye from 1978 to 1998. Arch. Environ. Contam. Toxicol. 2006, 50, 97-110. (29) USGS. Lake Trout Early Life Stage Mortality: Interactions of the Nutrient Thiamine and Dioxin-like PCBs and Their Mixtures Found in Green Bay; Final report of interagency agreement, Prepared for U.S. Fish and Wildlife Service; 14-48-0009-98-978; Washington, DC, 1999. (30) Williams, L. L.; Giesy, J. P. Relationships among concentrations of individual PCB congeners, 2,3,7,8-TCDD equivalents and rearing mortality of chinook salmon (Oncorhynchus tshawytscha) eggs from Lake Michigan. J. Great Lakes Res. 1992, 18, 108-124.
Received for review June 30, 2006. Revised manuscript received February 21, 2007. Accepted March 5, 2007. ES061559N
VOL. 41, NO. 9, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3089
