RESULTS

The recovery of hatchery-released juvenile salmon tagged with coded wire tags and comparison of the length and weight of juvenile salmon from the hatchery and estuaries was used to estimate the proportion of hatchery salmon collected from the estuary. The percent of the salmon tagged and released at the hatcheries, the estimated number of natural and hatchery-derived salmon, and the percent of tagged salmon recovered from all estuaries in 1989 are shown in Table 3. Additionally, the length and weight of fish sampled in 1989 and 1990 to assess contaminant exposure in their tissues are shown in Tables 4 and 5, respectively.

Hatchery fish were frequently sampled in some of the estuaries using our sampling design. This sampling was based on the proportion of recovered adipose-clipped juvenile chinook salmon and the similarity in the average length and weight of fish sampled in the hatcheries and the estuaries. For example, recovery of tagged fish in 1989 (Table 3) from the Duwamish Waterway (6.4%) was similar to the percentage of tagged juvenile salmon released in a large group from the Green River Hatchery (5.3% tagged of 4 million) just prior to estuary sampling. Additionally, in 1989 and 1990 the average length and weight of salmon sampled from the Green River Hatchery and the Duwamish Waterway were similar (Tables 4 and 5, respectively). In 1989, 7.5% of all the salmon released from the Kalama Creek Hatchery were tagged. In our particular sample, 21.7% of the Kalama Creek Hatchery fish were tagged, indicating that tagged fish were not uniformly distributed between the ponds at this hatchery at the time of sampling. Recovery of tagged salmon in the estuary was 11.8%, which indicates that a large proportion of hatchery salmon were being sampled in the Nisqually estuary. Those tags that were recovered were identified by the Washington Department of Fisheries to be from the Kalama Creek Hatchery. However, the average lengths and weights of Nisqually estuary juveniles were not statistically similar to the average lengths and weights of juvenile salmon sampled from Kalama Creek Hatchery in both 1989 and 1990 (Table 4 and 5). This dissimilarity may reflect the mixing and sampling of juvenile salmon from other sources in this estuary.

Chinook outmigrants from the Puyallup River system were captured primarily in the Milwaukee and City Waterways of Commencement Bay. No tagged fish were released from the Puyallup Hatchery, thus the sampling of hatchery-derived fish in the estuary could not be confirmed by recovery of tagged salmon. However, the average lengths and weights of juvenile salmon captured in 1989 and 1990 in Commencment Bay were similar to the size of salmon released from the hatcheries (Tables 4 and 5, respectively). It should be noted that a large proportion of tagged fish were recovered from Commencement Bay in 1990. The source of these fish sampled in the estuary is unknown at this time, but it is likely that salmon from the Nisqually River system were appearing in the Commencement Bay area.

Many of the salmon captured in the Snohomish estuary may not have been released from the Skykomish Hatchery. Tagged salmon were not released from the Skykomish Hatchery, thus the sampling of hatchery-derived fish in the estuary could not be confirmed. However, 20.5% of the fish released from the Tulalip Hatchery were tagged, and 6.0% of the fish captured in the Snohomish estuary were tagged, indicating that some fish of Tulalip Hatchery origin may have been captured during sampling in the Snohomish estuary. Tulalip Hatchery is located near Tulalip Bay which is approximately 5 miles north of the Snohomish River estuary. The average length and weight also indicates that fish of an origin other than the Skykomish Hatchery were also being sampled in the Snohomish estuary. Salmon captured in the Snohomish estuary were approximately 30 mm longer and weighed approximately 9 g more on average than fish released from the Skykomish Hatchery. The larger (>110 mm) juvenile salmon captured in the Snohomish estuary were not used for any of our investigations.

Stomach Contents

Aromatic hydrocarbons--Chemical analyses of stomach contents for selected aromatic hydrocarbons (AHs), including primarily 2-5 benzenoid ring aromatic compounds, and chlorinated hydrocarbons (CHs), including primarily polychlorinated biphenyls (PCBs), were done to assess possible dietary exposure of juvenile salmon to anthropogenic chemicals. The results showed significantly higher concentrations of total AHs, reported as low and high molecular weight AHs (LAHs and HAHs and their alkylated counterparts), in stomach contents of juvenile salmon from the Duwamish Waterway and Puyallup River estuary compared to fish from the Nisqually River estuary or hatcheries in both sampling years (Table 6). In contrast, the concentrations of LAHs in stomach contents of juvenile salmon from the Snohomish estuary, also an urban estuary, were not significantly different from concentrations of LAHs in stomach contents of fish found in the Nisqually estuary or corresponding hatchery fish. The concentrations of HAHs in stomach contents of salmon from the Snohomish estuary were significantly higher than concentrations of HAHs in stomach contents of fish found in the Nisqually estuary or corresponding hatchery fish. However, the concentration of HAHs in stomach contents of salmon caught in the Snohomish estuary were considerably less than the concentrations of HAHs in stomachs of fish captured from the Duwamish Waterway or Puyallup River estuary. Results for individual AHs found in stomach contents of juvenile salmon sampled in 1989 and 1990 by GC/MS are reported in Appendix Tables A-1 and A-2, respectively.

Results of analysis for CHs other than PCBs such as the pesticides 4,4'-DDE, 4,4'-DDD and 4,4'-DDT generally showed low concentrations in stomach contents of salmon at all sites in 1989 and 1990 (Table 6). No significant difference in summed concentrations of other selected CHs (see Table 2 for a listing of the selected pesticides) was observed for fish sampled from the estuaries compared to fish from the hatcheries. Results for individual CHs found in stomach contents of juvenile salmon sampled in 1989 and 1990 are reported in Appendix Tables A-3 and A-4, respectively.

Because fish metabolize AHs extensively, the AHs don't accumulate in tissues but are found in the bile as metabolites (Varanasi et al. 1989b). Bile was analyzed for FACs by HPLC/fluorescence at wavelengths specific for benzo[a]pyrene to estimate exposure of juvenile chinook salmon to AHs. In 1989, concentrations of FACs in bile were significantly higher in fish from the Duwamish Waterway and Puyallup estuary compared to concentrations in fish from the Nisqually estuary and the hatcheries (Fig. 2). Moreover, levels of biliary FACs in salmon from the Snohomish estuary were significantly higher than FACs in fish from the Nisqually estuary, a nonurban area. In 1990, the concentrations of FACs in bile of salmon from both the Duwamish Waterway and Puyallup estuary were significantly greater than the concentrations of FACs in salmon from the Nisqually estuary and the hatcheries (Fig. 2).

Liver

Chlorinated hydrocarbons, such as PCBs, are commonly found in sediments of highly industrialized urban waterways and bioaccumulate in tissues of fish inhabiting these areas, because they are not effectively metabolized by fish to excretable polar metabolites (Stein et al. 1984). The concentrations of hepatic PCBs in liver of juvenile salmon from the Duwamish Waterway, the Puyallup estuary, the Snohomish estuary, and the Nisqually estuary were significantly higher than the concentrations in liver of juvenile salmon from the hatcheries for both sampling years (Fig. 3). In addition, the concentrations of hepatic PCBs in juvenile salmon from the Duwamish Waterway and the Puyallup estuary were significantly higher than concentrations found in livers of juvenile salmon from the Nisqually estuary for both years. Results for individual CHs found in liver of juvenile salmon sampled in 1989 and 1990 are reported in Appendix Tables A-5 and A-6, respectively.

Exposure of fish to organotins is of concern because of the high toxicity of these compounds to marine species (Lee 1985, Snoeij et al. 1987). Recently, methods have been developed in our laboratories that allow for accurate measurement of mono-, di-, and tributyltins in tissues along with appropriate quality assurance procedures (Krone et al. 1989) . Results of the analysis of selected liver tissue composites from juvenile salmon sampled in 1989 for butyltins are shown in Table 7. Mono-, di-, tri-, and tetrabutyltins were not detected in most samples of juvenile salmon analyzed. Butyltins were detected only in fish from the Duwamish Waterway; however, the concentrations were near the limit of detection. Additionally, the concentrations of butyltins detected were low compared to the concentrations in adult English sole (Pleuronectes vetulus), a benthic species from the Duwamish Waterway, which had concentrations of butyltins that were up to three times as great as those in the juvenile chinook salmon from the Duwamish Waterway (Krone et al. 1989 a,b). Because of the low concentrations of butyltins found in liver of salmon in 1989, additional analyses were not conducted in 1990.

In the present study, the catalytic assays for AHH and EROD, were used to measure hepatic cytochrome P4501A (CYP1A) activity, the major inducible cytochrome P450 in fish (Goksøyr et al. 1991). Because hepatic AHH and EROD were highly correlated in juvenile salmon sampled in 1989 (r = 0.933, P = 0.0001) and 1990 (r = 0.809, P = 0.0001), the results for only hepatic AHH activity are shown (Fig. 4). Significantly higher hepatic AHH activities were observed for fish from the Duwamish Waterway and the Puyallup River estuary when compared to AHH activity in salmon sampled from the Nisqually estuary or from the hatcheries during both years of the study. Hepatic AHH activity in salmon from another urban area, the Snohomish estuary, was similar to the mean hepatic AHH activity for salmon from the Nisqually estuary or the hatcheries.

Hepatic Xenobiotic-DNA Adducts

The levels of hepatic xenobiotic-DNA adducts are a measure of the exposure, metabolism, and binding of hydrophobic contaminants, such as AHs, to a critical cellular macromolecule, DNA. The covalent binding of a chemical carcinogen to DNA is believed to be a critical step in the multistep process of chemical carcinogenesis (Swenberg et al. 1985, Poirier et al. 1991). The levels of hepatic xenobiotic-DNA adducts in juvenile chinook salmon sampled in 1989 and 1990 are shown in Figure 5. In 1989, hepatic DNA adduct levels were significantly higher in salmon from the Duwamish Waterway than adduct levels in fish from the Nisqually estuary and the hatcheries. However, the concentrations of DNA adducts in fish from the Puyallup estuary and the Snohomish estuary were not significantly different from concentrations of adducts in salmon from the hatcheries or the Nisqually estuary (Fig. 5). In 1990, the levels of hepatic xenobiotic-DNA adducts were significantly higher in salmon from the Duwamish Waterway and the Puyallup estuary than adduct levels in fish from the hatcheries and the Nisqually estuary.

In a preliminary laboratory experiment, juvenile chinook salmon from the Duwamish Waterway and the Green River Hatchery were held at our seawater research facility to assess the rate and extent of removal of PCBs in salmon from a PCB-contaminated site. The results showed no significant decline in the concentration of PCBs in the whole body (i.e., body burden) of juvenile chinook salmon from the Duwamish Waterway over 3 months of holding in filtered flowing seawater in the laboratory (Fig. 6). The PCB body burden of Duwamish fish at the termination of a 90-day holding period was still significantly higher than body burden of PCBs for juvenile chinook salmon from the Green River Hatchery. Results for individual classes of PCBs found in whole bodies of juvenile salmon sampled in 1989 by GC/MS are reported in Appendix Table A-7.

The concentration of total immunoglobulin in the plasma of juvenile chinook salmon collected from the hatcheries and the respective estuaries was determined for fish sampled in 1989. The concentration ranged from a high of approximately 36,000 immunoglobulin units/µL of plasma to a low of approximately 10,000 immunoglobulin units/µL of plasma (Fig. 7). However, the amount of total immunoglobulin in the plasma of salmon from each hatchery did not differ significantly from the amount of immunoglobulin in the plasma of salmon collected from their corresponding estuary.

To establish when the peak of a specific antibody response in the plasma of juvenile salmon occurred, the kinetics of the primary in vivo response to a standard hapten, TNP, were monitored over time (10 weeks) with a quantitative ELISA (Arkoosh and Kaattari 1990) using juvenile salmon captured from the Green River and Kalama Creek Hatchery and the Duwamish Waterway and Nisqually estuary. A significant increase (peak response) in the anti-TNP titer (1860 ± 900 units of anti-TNP activity/µL of plasma) in the plasma of the Green River Hatchery salmon occurred at 4 weeks post primary injection. In contrast, the significant increase (peak response) in the anti-TNP titer (850 ± 390 units of anti-TNP activity/mL of plasma) in the plasma of the Duwamish Waterway salmon occurred between 6 and 8 weeks post primary injection (Fig. 8). Because of a limited number of salmon available to sample, salmon from the Kalama Creek Hatchery and the Nisqually estuary were sampled for plasma levels of anti-TNP antibodies only at 4 and 9 weeks after exposure to the protein-conjugated hapten (antigen). Although the levels in the anti-TNP titer were comparable to concentrations observed for salmon from the Green River-Duwamish Waterway system, there were no statistical differences in the anti-TNP titer response in the Kalama Creek Hatchery and Nisqually estuary juvenile salmon at 4 and 9 weeks post primary injection (Fig. 9).

In Vitro Immune Response: Field-Exposed Salmon

To determine if juvenile chinook salmon leukocytes could generate a primary and secondary in vitro B cell response to TNP-KLH and TNP-LPS, a modified Mishell-Dutton (1967) culture system and the Cunningham modification of the Jerne hemolytic plaque assay (Jerne et al. 1963) was used (Kaattari et al. 1986). Various doses near the optimal antigen concentrations were used in the event that one of the antigen doses became suboptimal during the experimental period. However, there were no significant differences in the number of PFCs per culture produced with respect to the doses of antigen used in the assays. Leukocytes from the anterior kidney of juvenile salmon collected from the Green River Hatchery, Kalama Creek Hatchery and the Nisqually estuary were able to generate a significantly higher secondary response with TNP-KLH (Fig. 10) than that produced during the primary response. The number of primary PFCs per culture generated with TNP-KLH in salmon from the Green River Hatchery, Kalama Creek Hatchery and Nisqually estuary ranged from 1 to 5 PFCs per culture. Upon a second exposure to TNP-KLH, the PFC response increased significantly (9 to 12 times more PFCs per culture) for the juvenile chinook salmon from all three sites. However, this heightened secondary PFC response relative to the number of primary PFCs per culture to TNP-KLH did not occur with leukocytes from the anterior kidney of primed Duwamish Waterway juvenile salmon (Fig. 10).

The number of PFCs per culture generated during the secondary response in anterior kidney leukocytes with TNP-LPS was significantly higher than the primary response in juvenile salmon from all four locations (Fig. 11). However, there was a statistical difference between the secondary PFC response per culture with the anterior kidney leukocytes from the Green River Hatchery and Duwamish Waterway chinook salmon. Salmon from the hatchery produced a significantly higher secondary PFC response per culture than juvenile salmon captured from the Duwamish Waterway. However, the secondary in vitro PFC response elicited in salmon from the Kalama Creek Hatchery and Nisqually estuary with TNP-LPS was not significantly different (Fig. 11) from each other.

Similarly, leukocytes from the spleen of juvenile salmon collected from the Green-Duwamish system and the Nisqually system were able to produce a significantly higher secondary response to TNP-LPS compared to the primary response (Fig. 12). Unlike the anterior kidney's PFC response, suppression of immunological memory was not found in the splenic PFC response of juvenile chinook salmon from an urban estuary when compared to juvenile salmon from the respective hatchery or salmon from the Nisqually River system. Because an insufficient number of cells could be harvested, the splenic primary and secondary PFC response to TNP-KLH was not performed.

In Vitro Immune Response: Laboratory-Exposed Salmon

To substantiate that chemical contaminants were in part responsible for the alterations of immune function observed in Duwamish Waterway juvenile salmon in 1990, Green River Hatchery juvenile chinook were exposed directly to organic solvent extracts of DWSE by intraperitoneal injection in 1991. These sediment extracts contain AHs and PCBs, contaminants known to alter immune function in vertebrates. Lymphocytes from the anterior kidney of chinook salmon, exposed to either the DWSE or the carrier control, were able to produce an enhanced in vitro secondary PFC response (Fig. 13). An approximate 115 to 129% increase in the number of PFCs per culture was observed in the secondary response when compared to the primary response in these juvenile salmon. However, lymphocytes from the spleen of chinook salmon exposed to DWSE were unable to produce an enhanced in vitro secondary PFC response, while the splenic lymphocytes from the control group of salmon were able to produce a secondary in vitro response (Fig. 13). The level of stimulation of the secondary response in the control group increased approximately 250% when compared to the primary response of splenic lymphocytes of juvenile chinook salmon.

Because of the high mortality observed in 1990, observations on survival of juvenile salmon in 1991 were made only with fish from the Green River Hatchery and the Duwamish Waterway. A study was conducted to confirm the effects of an urban estuary on survival of juveniles observed in 1990 and to evaluate the effect of changes in our protocols of fish husbandry. Modifications to the protocol included placing fish in a reduced light environment, minimizing handling, using a different anaesthetic, and feeding at a reduced frequency. The proportion of surviving fish was significantly improved in 1991. With these changed protocols, survival of fish from the Green River Hatchery was 77% (n = 100) over a period of 84 days. With respect to survival of salmon from an urban estuary and the respective hatchery, there was no significant difference in the percent survival of juvenile salmon from the Duwamish Waterway and the Green River Hatchery after 40 days. However, under these improved holding conditions we were able to prolong the experiment and noted that the survival of Duwamish Waterway juvenile chinook salmon was significantly lower (59%, n = 100) than the survival of fish from the Green River Hatchery (77%, n = 100) over a period of 84 days (Table 8).

Growth of juvenile fall chinook from the different hatcheries and estuaries sampled in 1990 and 1991 are summarized in Table 9. Growth was assessed as the net increase in fork length and weight of individual juvenile chinook salmon. During 1990, a significantly smaller increase in length, but not weight, was observed for juvenile salmon captured in the Duwamish Waterway compared to fish taken from the Green River Hatchery during the 40-day holding period. However, a significantly smaller increase in length, as well as weight, was observed for juvenile salmon captured in the nonurban Nisqually estuary compared to salmon taken from the Kalama Creek Hatchery (the corresponding reference hatchery) during the 40-day holding period. No difference in the net increase in length or weight was observed for fish sampled from the Puyallup estuary compared to juveniles sampled from the Kalama Creek Hatchery. During 1990, growth of salmon from all the hatcheries and from the Duwamish Waterway and Puyallup estuary ranged from a 3.9 to 7.9% increase in length and a 31 to 38% in weight. However, during 1990, salmon from the Nisqually estuary increased only 1.5% in length and 14% in weight over the 40-day period. Juvenile salmon from the Nisqually system, in particular the estuary, were found to harbor a high infestation of the kidney fluke, Nanophaetes sp. (M. Myers, NMFS, pers. commun., July 1990) which may have reduced the growth of these salmon. Further studies will be needed to determine the effect of the fluke infestation on growth of juvenile chinook salmon.

In 1991, growth of juvenile salmon was significantly improved relative to growth of salmon studied in 1990. This was principally attributed to the improved holding conditions during the course of this study. Daily growth rates for length increased on average from 0.23 mm/day in 1990 to 0.46 mm/day in 1991 for Green River Hatchery fish. Similarly for fish from the Duwamish Waterway daily growth rates in length increased on average from 0.18 mm per day in 1990 to 0.37 mm per day in 1991. The increase in survival and the increases in rate of growth indicated that changes in the holding conditions instituted in 1991 improved our ability to evaluate the effect of contaminant exposure on growth and survival of these fish.

In 1991, only growth in the Green-Duwamish River system was studied, partly as a result of the difficulties in holding fish for extended periods of time as previously described. Juvenile salmon from the Green River Hatchery increased an average of 44% in length and 253% in weight over 84 days, whereas salmon from the Duwamish Waterway grew 34% in length and 242% in weight. Green River Hatchery fish grew significantly more in length than juveniles captured from the Duwamish Waterway. Green River Hatchery salmon (87.0 ± 5.3 mm, n = 100) at the beginning of the growth study were approximately 4 mm smaller in length compared to salmon captured from the Duwamish Waterway (90.9 ± 5.2 mm, n = 100); however, final lengths for Green River and Duwamish salmon were 125.6 ± 10.5 mm (n = 77) and 121.9 ± 14.0 mm (n = 59), respectively (Table 9).

Table of Contents