PREFACE

Several U.S. West Coast populations of Pacific salmon are dwindling and, in some cases, declining so seriously that they have been listed under the Endangered Species Act. A number of factors, such as dramatic habitat loss and overfishing, are believed to be the major contributors to the problem. However, other environmental factors may also play a role. One of these is the chemical contamination of urban estuaries through which some juveniles pass on their migration to sea. To address this pollution factor, scientists within the Environmental Conservation (EC) Division of the Northwest Fisheries Science Center initiated a multidisciplinary project to assess the levels of contaminant exposure and signs of adverse biological effects in migrating juvenile chinook salmon (Oncorhynchus tshawytscha) from polluted urban estuaries of Puget Sound, Washington.

Scientists in the EC Division scientists previously documented that certain bottom-dwelling fish species, such as English sole (Pleuronectes vetulus), feeding from and living on contaminant-laden sediments in polluted areas of Puget Sound bioaccumulate or biotransform toxic contaminants and show a number of adverse biological effects (e.g., reproductive impairment and a variety of liver lesions, including neoplasms). Moreover, healthy English sole exposed in the laboratory to toxic chemicals extracted from urban sediments develop, over a period of 18 months, many of the same lesions found in the livers of fish sampled from polluted areas. These field and laboratory studies have helped establish cause-and-effect relationships between chemical contaminant exposure and certain adverse biological effects in benthic marine fish.

Because juvenile salmon generally reside in urban estuaries for a brief time (several weeks to months), they seemed at little risk compared to chronically exposed resident species. Furthermore, being primarily water-column inhabitants, it did not seem likely that salmon would be as directly or as significantly exposed to contaminant-laden bottom sediments as benthic fish.

Over the past few years, however, a deeper understanding has been gained of the subtle biochemical and biological responses occurring in marine organisms exposed to contaminants. These early responses (bioindicators) are believed to serve as harbingers of later, more serious effects. Such understanding has led EC Division scientists to develop more sophisticated analytical, biochemical, and immunological techniques that should lead to a better assessment of whether the brief residency of juvenile salmon in polluted estuaries results in long-term consequences for their health and survival.

Funding for this study was jointly provided by the Environmental Protection Agency (EPA) and the National Oceanic and Atmospheric Administration (NOAA) in 1989 and 1990. NOAA carried on the the study reported herein in 1991.

INTRODUCTION

Estuaries serve as important habitats for salmon during the juvenile stage of their life cycle (Healy 1982). Outmigrating juvenile salmon use estuaries as an area of refuge from predators, as an environment with a rich food supply to support rapid growth, and as a transition area to adjust to a marine habitat (Dorcey et al. 1978, Simenstad et al. 1982). However, estuaries located near urban centers also serve as depositories for both point and nonpoint sources of chemical contaminants from municipal and industrial activities. The chemicals from these sources are known to accumulate in the bottom sediments (Dexter et al. 1985). Several estuaries within Puget Sound, Washington, are located near urban centers and are used by migrating salmonids as areas of residence. These include the Duwamish Waterway entering Elliott Bay in Seattle, the Puyallup River entering Commencement Bay in Tacoma, and the Snohomish River entering Port Gardner Bay in Everett (Fig. 1). During their outmigration period, juvenile salmon undergo numerous physiological adaptations and adjustments. Thus, additional stresses, such as exposure to toxic chemicals while in the estuarine environment, could prove harmful to these juvenile salmon.

Of the five species of Pacific salmon, chinook salmon (Oncorhynchus tshawytscha) are most dependent upon estuaries during the early stages of their life cycle. Smolting chinook salmon reside in estuaries for a longer period than other species of salmonids (Thom 1987, Healy 1982). The residence time of juvenile chinook salmon in estuaries is variable and depends upon many factors, such as their age when they enter an estuary. Juvenile chinook have been reported to reside in some estuaries for up to 6 months (Reimers 1973, Levy and Northcote 1982, Simenstad et al. 1982). One to two months, however, appears to be an average duration of residence of juvenile chinook salmon in Puget Sound estuaries (Simenstad et al. 1982).

The presence of chemically contaminated sediments in the Duwamish Waterway coupled with the extended residence time of juvenile chinook salmon in estuaries prompted a study by McCain et al. (1990) to examine the degree to which juvenile fall chinook salmon are exposed to toxic chemicals such as aromatic hydrocarbons (AHs) and chlorinated hydrocarbons (CHs). Stomach contents and livers of juvenile salmon captured from the Duwamish Waterway were analyzed for levels of polychlorinated biphenyls (PCBs), and stomach contents and bile were analyzed for parent AHs and polar metabolites of AHs, respectively. These findings were compared to concentrations of chemicals in chinook salmon from the Nisqually estuary, a nonurban area known to be minimally contaminated, and the Kalama Creek Hatchery, which releases salmon smolts into the Nisqually River (Fig. 1). The findings showed that juvenile chinook salmon from the Duwamish Waterway were exposed to substantially higher concentrations of contaminants than juvenile chinook from the Nisqually River system. However, juvenile salmon specifically from the Green River Hatchery (the Green River is the upper reach of the Duwamish Waterway) were not examined in this study, thus making a direct comparison of increased contaminant levels in fish from the Duwamish somewhat difficult. Additionally, it was not clear if these findings could be extended to salmon outmigrating through other urban estuaries in Puget Sound.

In 1989, a cooperative study initiated by the National Marine Fisheries Service (NMFS) and the Environmental Protection Agency (EPA) was designed to confirm the findings by McCain et al. (1990) and to determine if juvenile chinook salmon, migrating through estuaries located near other Puget Sound urban centers, including the Puyallup River estuary and the Snohomish River estuary, were similarly exposed to chemical contaminants. Juvenile chinook salmon from the Nisqually River estuary and the hatcheries for each of the sampled rivers were considered reference fish. The study was also expanded to include measurement of early biochemical responses to contaminant exposure and measurement of biological effects that represent significant physiological processes. These biochemical responses, known as bioindicators, have been shown to reflect the degree of exposure of particular contaminants by marine organisms in a polluted environment as well as to indicate some of the early biological consequences of chemical exposure. The indicators of contaminant exposure (Stein et al. 1992) include levels of hepatic PCBs and biliary levels of fluorescent aromatic compounds (FACs), a semiquantitative measure of exposure to AHs. Biochemical responses to chemical exposure include hepatic cytochrome P-450 activities, measured by changes in aryl hydrocarbon hydroxylase (AHH) and ethoxyresorufin-O-deethylase (EROD) activities; and hepatic DNA-xenobiotic adducts, detected by ³²P-postlabeling. Aryl hydrocarbon hydroxylase and EROD activities represent inducible enzymes whose activities increase rapidly after exposure to AHs, whereas adduct levels represent the long-term cumulative effects of exposure to AHs and damage to the DNA. Biological parameters that were measured included effects on immune function and effects on growth and long-term survival of juvenile salmon. An effective immune system is a critical physiological process in fish that has been shown to become impaired due to contaminant exposure (McLeay and Gordon 1977, Arkoosh and Kaattari 1987, Bengtsson et al. 1988). Similarly, exposure to contaminants has been shown to significantly reduce survival and growth in a number of marine species (Swartz et al. 1985, Barron and Adelman 1984, Casillas et al. 1992).

The results to date show that outmigrating juvenile chinook salmon from both the Duwamish Waterway and from the Puyallup estuary in Commencement Bay exhibited consistent evidence of exposure to contaminants. Juvenile chinook salmon from the Snohomish estuary, another urban estuary, also appeared to be exposed to contaminants, but to a much lesser degree than salmon from the Duwamish Waterway or Puyallup estuary. In addition, when held in tanks with flow-through seawater for a period of several months, juvenile salmon from the Duwamish Waterway suffered significantly more mortalities, exhibited reduced growth, and showed evidence of impaired immune function when compared to salmon from either the Green River Hatchery (the primary source of salmon for the Duwamish Waterway) or to salmon from the nonurban Nisqually system. The biological consequences of chemical exposure to juvenile chinook salmon in the estuary may place additional stresses on these fish that may affect their long-term health and survival as they enter the marine environment.

In 1989, juvenile chinook salmon (Oncorhynchus tshawytscha) were sampled from hatcheries and the respective estuaries of four river systems in Puget Sound. These included the Green-Duwamish, Puyallup, Nisqually, and Snohomish River systems (Fig. 1). In 1990, the hatcheries and estuaries of the Green-Duwamish River, Puyallup River, and the Nisqually River were sampled for juvenile chinook salmon; fish from the Snohomish system were not sampled. In 1991, fish were sampled only from the hatchery and estuary of the Green-Duwamish River system.

Juvenile chinook salmon were collected from the hatcheries just prior to their release into the river and collected from their respective estuaries a minimum of 2 weeks after the initial hatchery release. This schedule allowed time for transit of juveniles to the estuaries. Weitkamp and Campbell (1980) estimated that the transit time from hatchery to estuary for chinook salmon was 15-19 days in the Green River-Duwamish system. Sampling dates at the hatcheries and at all estuarine sites and the number of fish collected during each sampling period are shown in Table 1. Juvenile chinook salmon were collected at the hatcheries from late May to early June and at the estuaries from June through early July of each year. Fish at the estuarine sites were captured using a 30-m beach seine. It was necessary to find sites where the beach seine could be successfully deployed. Typically, these sites were areas with little debris and with moderate currents. Further, if sites were not productive, alternative sites were sampled. In the various estuaries, the number of juvenile salmon captured ranged from 4 to approximately 550 fish per sampling day.

Juvenile chinook salmon collected at the hatcheries and in the estuaries were placed in 94-L ice chests filled with fresh or salt water, respectively. Water was continuously aerated using air stones and small battery-powered portable aerators. Salmon were transported to the NMFS's Montlake facility in Seattle for immediate tissue and fluid sampling, as described below, or to the NMFS's Mukilteo Marine Laboratory, a salt water facility. Mortalities of salmon during transportion were minimal. Fish held at our Mukilteo Marine Laboratory were used for depuration, survival, growth, or immune function studies as described below.

Juvenile salmon collected at the hatcheries for the depuration, immune function, growth and survival studies were acclimated to full-strength seawater at the Mukilteo Marine Laboratory over a 5-day period. Juvenile salmon were placed in 750-L circular fiberglass tanks, initially containing seawater adjusted to 5 parts per thousand (ppt), and held for 24 hours. Salinity was adjusted daily in increasing 5 ppt increments until the fifth day, when juveniles were adjusted to full-strength salt water (28-30 ppt). Flow was maintained at 12 L/minute. Juvenile chinook salmon captured in the estuaries were held initially in 15 ppt seawater the first day. Salinity was adjusted daily in increasing 5 ppt increments until the third day, when juveniles were adjusted to full strength seawater (28-30 ppt). Salmon were fed daily at 3% of their body weight with Oregon Moist Pellet (OMP; Moore-Clark, La Conner, WA) for a 10-day period and then fed a prophylactic diet for control of disease at 3% of their body weight with OMP supplemented with oxytetracycline (4 g 100g-¹ of diet) for another 10-day period. This prophylactic feeding regime was repeated every 20 days.

Analysis for organic chemicals in livers, whole bodies (minus gastrointestinal tract), and stomach contents were done according to procedures described by MacLeod et al. (1985) and Krahn et al. (1988). The AHs determined by gas chromotography with mass spectrometric detection (GC/MS) are listed in Table 2. The CHs (Table 2) were analyzed using gas chromotography (GC) with electron capture detection (ECD) and GC/MS for chemical confirmation on selected samples. The steps of the chemical analysis are briefly described below.

Tissue samples were extracted according to the procedures of MacLeod et al. (1985). Briefly, 3 g of tissue or stomach contents were added to a centrifuge tube containing sodium sulfate, methylene chloride, and the surrogate standards for analyzing AHs and CHs. The mixture was then macerated with a Tekmar® Tissumizer and the resulting extract filtered through a column of silica and alumina. The organic extract was then concentrated to 1 mL.

Cleanup of the concentrated 1-mL organic extracts was done using a high-performance liquid chromatograph (HPLC). Samples were injected on a Spectra-Physics Model 8800 HPLC equipped with two preparatory size Phenomenex (containing Phenogel,100-Å size-exclusion packing) columns in series and the analytes monitored by ultraviolet (UV) detection. The mobile phase consisted of helium-gassed, methylene chloride solvent, run at a flow rate of 7 mL/minute for 20 minute at ambient temperature. The fraction containing the AHs and CHs were collected according to Krahn et. al. (1988). The methylene chloride fraction volume, reduced by evaporation to about 0.1 mL, was then exchanged into hexane. Standards were then added before analysis by GC.

The organic solvent extracts of stomach contents were analyzed for AHs by GC/MS according to MacLeod et al. (1985). The organic extracts of stomach contents and liver were analyzed for CHs (PCBs and pesticides) by capillary column GC with electron capture detection (ECD). Representative samples were also analyzed by GC/MS to confirm the identifications of the CHs.

The concentrations of AHs in stomach contents are reported as sums of low molecular weight AHs (LAHs, 2-3 rings) and high molecular weight AHs (HAHs, > 3 rings) and their alkylated counterparts (Table 2). The concentrations of PCBs in stomach contents and liver are reported as the sum of PCB homolog classes, from trichlorobiphenyls to decachlorobiphenyls (Table 2). Concentrations of other CHs in stomach contents are also reported as the sum of selected pesticides. Concentrations of individual AHs, classes of PCBs, and other CHs determined in stomach contents are reported in the Appendix. The quality assurance procedures using method blanks and percent recoveries of surrogate standards of a National Institute of Science and Technology (NIST) control material are also reported in the Appendix.

Analysis of liver tissue for butyltins was conducted according to the method described by Krone et al. (1989 a,b). Briefly, the organotin chlorides are extracted from liver tissue (by homogenization with a Tissumizer) using methylene chloride with 0.1% tropolone as a complexing agent. The extracted organotins are converted to their n-hexyl derivatives through the Grignard reaction (Morrison and Boyd 1973) and then cleaned up by two chromatographic steps: 1) eluting the sample extracts through a glass column containing 4.5 g each of alumina and silica, with 20 mL pentane, and 2) loading the eluate onto an amino Sep-Pak® and eluting with 3 mL of pentane. The pentane eluate is then concentrated to 1 mL and transferred to a GC vial for GC/MS analysis. Quality assurance procedures using blanks and recovery of tripropyltin showed no detection of butyltins in blanks, and percent recoveries (98 ± 19) for tripropyltin were above acceptable levels.

Bile was analyzed by the HPLC/UV method of Krahn et al. (1986) to estimate the exposure of juvenile salmon to fluorescent aromatic compounds (FACs) such as AHs with 2-5 benzenoid rings. Bile is injected directly into a HPLC equipped with a reverse-phase analytical column. The polar analytes (primarily metabolites of AHs) in bile were separated using a water/methanol gradient (100% water containing 5 µL acetic acid/L, to 100% methanol) and monitored by two fluorescence detectors in series. The excitation/emission wavelengths of one detector was set to 290/335 nm (where metabolites of naphthalene (NPH) fluoresce) and the other set to 380/430 nm (where the metabolites of benzo[a]pyrene (BaP), pyrene and fluoranthene fluoresce). The levels of biliary FACs are reported as equivalents of known concentrations of BaP or NPH standards on the basis of biliary protein, because recent sudies (Collier and Varanasi 1991) have shown that such a normalization can, to a large extent, account for changes in the levels of FACs due to differences in the feeding status of some fish. The concentrations of biliary protein were measured by the method of Lowry et al. (1951) using bovine serum albumin as the standard.

Quality assurance of BaP and NPH calibration standards run at the start of the analysis set had a relative standard deviation of 3.4% in 1989 and 2.2% in 1990. The number of bile composites for each sampling year was small enough that all samples were analyzed in a single set. A "bile pool" reference material was analyzed with each set in 1989 and 1990. Concentrations of FACs in the bile pool reference material during these analyses were within 10% of the mean concentration results of the previously determined FAC concentrations in the bile pool reference material. Analysis of blanks and replicate analyses showed that quality assurance tests were passed.

Hepatic microsomes for AHH and EROD activity were prepared by a slight modification of the procedure of Collier et al. (1986). Composite liver samples were homogenized with a Potter-Elvehjem homogenizer, using 4 mL of 0.25 M sucrose per gram wet weight. The homogenate was then centrifuged at 10,000 g for 20 minutes, and the resulting supernatant centrifuged at 100,000 g for 60 minutes. The resulting supernatant was discarded and the surface of the microsomal pellet rinsed with 1 mL of 0.25 M sucrose, and the pellet resuspended, using a Dounce homogenizer, in 1 mL of 0.25 M sucrose in 20% glycerol per gram of original liver weight. Protein was measured by the method of Lowry et al. (1951) using bovine serum albumin as a standard.

Aryl hydrocarbon hydroxylase activity was assayed by a method described by Collier et al. (1986). The reaction mixture consisted of the following: 450 µL 0.05 M Tris buffer, pH 7.5, 5 mM MgCl2, 1 mM NADPH, and 37.5 µL microsomal suspension (0.2-0.8 mg protein). The mixture was incubated at 25°C for 15 minutes, and the reaction started by adding 40 nmole ¹⁴C benzo[a]pyrene, 4.6 µ Ci/mole, in 12.5 µL acetone. Assays were run at 25°C for 15 minutes, and the reactions stopped by the addition of 1.0 mL of 0.15 M KOH in 85:15 volume:volume (v:v) DMSO:H2O. The reaction mixture was extracted with 15 mL hexane in triplicate to remove the unreacted substrate. A 150-µL aliquot of the DMSO: aqueous phase was then neutralized by adding 0.5 mL of 0.1 N HCL. After the addition of 5 mL liquid scintillation cocktail, the ¹⁴ C associated with the reaction products was determined by liquid scintillation spectroscopy. Quality assurance procedures included duplicate zero-time and boiled enzyme blanks for each set of assays. Each sample was run in duplicate and those samples showing more than 10% difference between duplicates were repeated.

Hepatic EROD activity was measured according to the method of Prough et al. (1978). Briefly, the assay (2 mL total volume) contained 0.1 M Tris buffer, pH 8.0; 0.5 µM ethoxyresorufin; 60 µM NADPH, and 0.1 to 0.5 mg microsomal protein. The reaction was initiated by the addition of the NADPH, and the formation of resorufin was continuously monitored by fluorescence at 530 nm (excitation) and 585 nm (emission) for 1 to 3 minutes. A blank (minus NADPH) was run with each set of samples.

Hepatic DNA for DNA-xenobiotic adduct determination was isolated by a chloroform/phenol extraction procedure of Gupta (1984), which included treatment with Proteinase K and both RNase A and T1 and precipitation of DNA with ethanol. The levels of hydrophobic xenobiotic-DNA adducts were determined using ³²P-postlabeling assay after adduct enrichment using the n-butanol procedure as described in Varanasi et al. (1989a). Briefly, 10 µg of DNA, determined spectrophotometrically (1 absorbance unit = 20 µg DNA/µL), was hydrolyzed to 3'-mononucleotides using micrococcal endonuclease and spleen phosphodiesterase (Gupta and Randerath 1988). The adducts were extracted with n-butanol and labeled in the 5' position by T4 polynucleotide kinase-catalyzed transfer of ³²phosphate from ³²P-labelled adenosine triphophate (ATP) (5 to 6 x 103 Ci/mmol). The remaining normal nucleotides and adducts were separated by anion-exchange thin-layer chromotography (TLC) on polyethyleneimine (PEI) cellulose sheets, and the ³²P-labeled adducts were detected by screen-enhanced autoradiography. Areas of the TLCs corresponding to DNA-xenobiotic adducts were excised and levels of ³²P measured by liquid scintillation spectroscopy. Salmon sperm DNA from Sigma Chemical Co. (St. Louis, MO) was carried through the ³²P-postlabeling assay and used to correct for background. Adduct levels are reported as nanomoles of adducts per mole of DNA bases. Total DNA bases in a sample were determined according to Gupta et al. (1982). In 1990, the n-butanol enhancement procedure was replaced by the nuclease P1 enhancement method according to Gupta and Randerath (1988). Studies have shown that the n-butanol and nuclease T1 methods give comparable results for several carcinogenic aromatic hydrocarbons (Gupta and Early 1988).

The T-independent antigen, trinitrophenylated lipopolysaccharide (TNP-LPS), was prepared as described by Jacobs and Morrison (1975). Briefly, Escherichia coli lipopolysaccharide (LPS), serotype 0111.B4 (Difco, Detroit, MI) was dissolved in 0.28 M cacodylate buffer and adjusted to a pH of 11.5. Picrylsulfonic acid (Sigma Chemical Co., St. Louis, MO) was added dropwise to a test tube containing the LPS solution. Upon coupling, the TNP-LPS solution was exhaustively dialyzed against 0.077 M phosphate buffered saline (PBS), pH 7.4, with a final dialysis against the media (Rosewell Park Memorial Institute (RPMI)-1640; Gibco, Grand Island, NY). The solution was then pasteurized for 30 minutes at 70°C and stored at 4°C in a sterile stoppered serum bottle.

The T-dependent antigen, TNP-keyhole limpet hemocyanin (TNP-KLH), was prepared as described by Rittenberg and Amkraut (1966). Briefly, KLH (Sigma Chemical Co., St. Louis, MO) was added to 0.28 M cacodylate buffer and mixed for 1 hour at room temperature. This mixture was added to a foil-wrapped tube and picrylsulfonic acid was added. The mixture was then dialyzed against PBS and a final change against RPMI-1640. The conjugate was filtered, sterilized, and stored in a sterile stoppered serum bottle. The conjugation ratio of TNP to KLH was determined to be 16 mmoles of TNP to 100 g KLH.

Trinitrophenylated-bovine serum albumin (TNP-BSA), used in the enzyme-linked immunosorbent assay (ELISA) for the quantification of anti-TNP antibodies, was prepared according to the method described by Garvey et al. (1977). Trinitrophenylation was achieved by first mixing picrylsulfonic acid in borate acid. The picrylsulfonic acid solution was added dropwise into a BSA solution under constant mixing. The solution was dialyzed extensively against 0.17 M borate buffered saline. At the end of dialysis, the mixture was sterilized by filtration and stored in a sterile stoppered serum bottle at 4°C.

Measurement of Total Immunoglobulin

Total plasma immunoglobulin concentrations were determined for juvenile chinook salmon collected during May to July, 1989. Blood samples from juvenile salmon from the Green River Hatchery, Kalama Creek Hatchery, Puyallup Hatchery, Skykomish Hatchery, Duwamish Waterway, Nisqually estuary, Puyallup estuary and Snohomish estuary were sampled at the Montlake facility. Upon arrival of the fish, blood samples were immediately collected from the caudal blood vessels after removal of the caudal peduncle (Hesser 1960). Blood samples were collected in heparinized micro-hematocrit tubes (Van Waters and Rogers, Seattle, WA) and one end of the tube was plugged in seal-ease (Cray Adams, Parsippany, NM). The samples were then spun down in a hematocrit centrifuge (International Equipment Company, Needham Heights, MA.) for 5 minutes. The hematocrit tubes were broken above the cell pellet, the plasma was removed and stored in microfuges tube at -20°C.

An ELISA was used to determine the relative concentration of plasma immunoglobulin as modified from Kaattari and Yui (1987). Briefly, wells of a 96-well ELISA plate (Costar, Cambridge, MA) were coated overnight with a monoclonal antibody against rainbow trout (Oncorhynchus mykiss) immunoglobulin. This monoclonal antibody (1-14) is known to cross react with salmonid antibodies (Kaattari and Yui 1987). The monoclonal antibodies were MAPS (monoclonal antibody purification system) purified (Bio-Rad, Richmond, CA) and the protein content of the solution was determined by the Lowry method (Lowry et al. 1951). The coating agent was removed and 50 µL of various dilution of the standard rainbow trout sera and the chinook plasma samples were added to the wells. The standard sera allowed for normalization of the data from day to day and was also used to assign immunoglobulin units per microliter of plasma for each sample tested. Also for quality assurance, another source of pooled rainbow trout sera was added to the wells. Although this method does not allow for an absolute quantification of total plasma immunoglobulin concentration, it does allow for relative quantification of immunoglobulin concentration of unknown samples. The plasma and sera were incubated and washed from the wells, and 100 µL of strepavidin-horseradish peroxidase was added for 30 minutes at room temperature. Substrate was then added as described by Arkoosh et al. (1991). A 10-minute kinetic-based ELISA was measured with a Titertek Multiskan® spectrophotometer (Flow Laboratories, McLean, VA). Results were reported as immunoglobulin units per µL of plasma.

Primary in vivo Anti-TNP Response

The primary in vivo anti-TNP response of juvenile chinook salmon was determined for fish collected from the Green-Duwamish and the Nisqually systems during the spring of 1990. Both estuary and hatchery chinook salmon were allowed to acclimate at the Mukilteo Marine Laboratory for a minimum 2-week period prior to injection with TNP-KLH as described in Arkoosh et al. 1991. To minimize stress during sampling and injecting, which can result in immunosuppression (Maule et al. 1989), salmon were anesthetized by adding 20 mg/L of tricaine methanesulphonate (MS-222; Sigma Chemical Co, St. Louis, MO) directly to an aerated tank with the seawater flow turned off. After the salmon were quiescent, they were injected intraperitoneally with either 100 µg of TNP-KLH (16 mmoles TNP/100 g of KLH) emulsified in Freund's complete adjuvant (FCA; Difco Laboratories, Detroit, MI) or with PBS emulsified in FCA. After the injection was completed, the seawater flow was reinstated, allowing the salmon to recover from the anesthetic.

Plasma samples were collected from 10 primed and 10 unprimed fish from the Green-Duwamish system at 2, 4, 6, 8, and 10 weeks post-injection for the detection of specific antibodies to TNP. Plasma samples were collected from primed and unprimed fish from the Nisqually system at 4 and 9 weeks post-injection for the detection of specific antibodies to TNP. Four weeks had been previously determined to be the peak of the in vivo anti-TNP response in juvenile chinook salmon (M. Arkoosh, NMFS. Pers. observ., June 1990). Plasma samples from five salmon were taken from the hatcheries and estuaries of each system before injection of the antigen to determine the presence of naturally occurring anti-TNP antibodies. These samples represent time zero.

An ELISA was used to determine anti-TNP activity per microliter in the chinook's plasma. This method was modified from Arkoosh and Kaattari (1990) and described in Arkoosh et al. 1991. Results were reported as units of anti-TNP activity per microliter of plasma. If the ELISA did not detect any anti-TNP antibody in the plasma sample, we assigned these samples one half the detection limit of this method for statistical analyses, which was 72 units of anti-TNP activity per microliter of plasma.

In Vitro Immune Response: Laboratory-Exposed Salmon

Juvenile chinook salmon collected from the Green River Hatchery were used to examine the effects of a contaminated sediment extract from the Duwamish Waterway delivered by ip injection on the primary and secondary in vitro anti-TNP response. Salmon transferred to the Mukilteo Marine Laboratory were allowed to acclimate for a minimum 2-week period prior to the beginning of the immunological tests. Two sediment extract solutions were prepared according to Krahn et al. (1988) for the injections: solution 1 contained 614 µg/mL polycyclic aromatic hydrocarbons (PAHs) and 90.2 µg/mL PCBs; solution 2 contained 793 µg/mL PAHs and 97.7 µg/mL PCBs. Salmon were injected with 12 µL of either 200 mg equivalents of Duwamish Waterway sediment (DWSE) (extract solution 1), in emulphor 620/g of fish (1:1 volume/volume) or with 12 µL of the carrier (acetone: emulphor 620, 1:1 volume/volume) as a control. One week after the initial injection, the fish were injected for a second time with either the same concentration of sediment extract (test fish; solution 2) or acetone/emulphor 620 solution (control fish). Ten days after the second injection with sediment extractor carrier, bile from 10 fish of the control group and 20 fish of the sediment extract group were collected for analysis for FACs by HPLC/UV (Krahn et al., 1986). Five days after we collected bile samples for FAC analysis, salmon were injected with either 100 µg of the antigen TNP-KLH emulsified in Freund's complete adjuvent, or with PBS emulsified in Freund's complete adjuvent. Twelve weeks after the antigen or PBS injection, the Mishell Dutton and the Cunningham modification of the Jerne plaque assays were performed, as described above.

Growth and survival studies were begun approximately 10-14 days after juvenile chinook salmon were collected. This delay allowed time for fish to be acclimated to full-strength seawater and for the cessation of any mortalities due to the stress of capture and transport. During 1990, juvenile chinook salmon from the Kalama Creek Hatchery, Green River Hatchery, Nisqually estuary, the Puyallup estuary and the Duwamish Waterway were evaluated for differences in growth and survival. During 1991, only chinook salmon from the Green River Hatchery and the Duwamish Waterway were evaluated for differences in growth and survival.

For studies of growth rate we utilized PIT (passive integrated transponder) tags as described by Prentice et al. (1990). Groups of 50 or 100 chinook received PIT tags; these groups were also used to obtain estimates of survival rates. Use of individual PIT tags and an electronic readout of tag numbers allowed us to identify and calculate growth on individual fish, thus greatly reducing statistical variability. In 1990, juveniles were anesthetized with MS-222 solution in seawater for tag insertion. In 1991, we tested a relatively new anesthetic, Metomidate Wildlife® (metomidate hydrochloride, Wildlife Pharmaceuticals Inc. Fort Collins CO), in an attempt to further reduce the stress of handling. After juvenile chinook were anesthetized, the glass-encased PIT tags were loaded into a 12-gauge needle and inserted with a 20 cc syringe into the right ventral abdominal body cavity. Tag numbers were read with a standard reader (BioSonics Inc., Seattle, WA), initial fork lengths were measured to the nearest millimeter, and fish were then weighed to the nearest 0.1 gram. Length and weight measurements were made at monthly intervals during the 1990 study and at 6-week intervals during the 1991 study. Survival and growth measurements were intended to be conducted over a 4-month period; however, due to large increases in mortality, the maximum period was 40 days in 1990 and 84 days in 1991.

Analysis of variance (ANOVA) was used to determine the significance of differences among means for each of the parameters or endpoints. Statistical analyses of data for tissue contaminants, bioindicators, immunology, growth, and survival were performed using the StatView II and SuperANOVA statistical packages (Feldman and Gagnon 1986). Multiple means obtained for contaminants in tissues and bioindicators of chemical contaminant exposure were compared using Fisher's protected least significant difference test (Zar 1974). Tissue contaminant data for the hatcheries were combined in 1989 and 1990 due to low sample sizes for composites from some of the hatcheries and because there were generally no significant differences in concentrations of contaminants in fish tissues from the hatcheries. Net growth in length and weight of individual juvenile salmon were calculated by subtracting initial from final lengths and weights. Means of the parameters obtained from immunological and growth studies for fish from contaminated estuaries were compared to means of the parameters obtained for fish from hatcheries and the reference (Nisqually) estuary using Dunnett's multiple comparison test or Student's t-test (Zar 1974). The proportion of surviving juvenile chinook salmon was calculated as the final number of juveniles in the growth studies from each hatchery or estuary divided by the number of salmon at the initiation of the growth studies. Differences in percent survival for fish from the estuaries and hatcheries were evaluated using Chi-square analysis. Results of all statistical tests in this study were considered significant at P<&0.05.

Table of Contents