NOAA Technical Memorandum NMFS-NWFSC-11

CHEMICAL CONTAMINANTS IN GRAY WHALES (Eschrichtius robustus) STRANDED IN ALASKA, WASHINGTON, AND CALIFORNIA, U.S.A. Usha Varanasi1, John E. Stein1, Karen L. Tilbury1, James P. Meador1, Catherine A. Sloan1, Donald W. Brown1, Sin-Lam Chan 1, and John Calambokidis2 1National Marine Fisheries Service Northwest Fisheries Science Center Environmental Conservation Division 2725 Montlake Blvd. E. Seattle WA 98112 2Cascadia Research Collective 218 1/2 West Fourth Avenue Olympia, WA 98501 August 1993 U.S. DEPARTMENT OF COMMERCE Ronald H. Brown, Secretary National Oceanic and Atmospheric Administration John A. Knauss, Administrator National Marine Fisheries Service William W. Fox, Jr., Assistant Administrator for Fisheries

INTRODUCTION

METHODS

Sample Collection

Analytical Methodologies

Chlorinated Hydrocarbons and Polycyclic Aromatic Contaminants

Toxic and Essential Elements

Percent Lipid

Quality Assurance Measures for Toxic and Essential Elements, Chlorinated Hydrocarbons, and Polycyclic Aromatic Contaminants

Statistical Methodology

RESULTS AND DISCUSSION

Percent Lipid

Chlorinated Hydrocarbons

Polycyclic Aromatic Contaminants

Toxic and Essential Elements

Regional Patterns in Contaminant Profiles

Sex-related Differences in Contaminant Profiles

Pattern of Accumulation of PCB Congeners

CONCLUSIONS

ACKNOWLEDGMENTS

CITATIONS

APPENDIXES

A - Concentrations of Individual Chlorinated Hydrocarbons in Gray Whalec Samples, Method Blanks, and Quality Control Materials

B - Concentrations of Aromatic Contaminants in Gray Whale Samples, Method Blanks, and Quality Control Materials

C - Concentrations of Elements in Gray Whale Samples and Method Blank

Gray whale (Eschrichtius robustus) populations in the eastern North Pacific have increased at an annual rate of close to 3% since cessation of commercial exploitation and now number over 20,000 (Buckland et al. 1993, Reilly 1992), which is close to their historical population size. These marine mammals make an annual round-trip migration between their breeding grounds in Mexican waters (along Baja California) and their feeding grounds in more northern waters which range from northern California to Alaska (Rice and Wolman 1971). The southbound migration to the breeding grounds occurs in December and January along the West Coast and the northbound migration from February through May (Pike 1962). Gray whales generally fast during the breeding season in Mexico and during their migrations (Rice and Wolman 1971). Their body mass, overall fat content, girth, and blubber thickness are significantly lower during the northbound migration than during the southbound migration (Rice and Wolman 1971). Though the majority of gray whales feed in the Bering and Chukchi Seas in Alaska (Rice and Wolman 1971), some animals spend extended periods in the spring and summer feeding in coastal waters of California, Oregon, Washington, and British Columbia (Nerini 1984; Rice and Wolman 1971; Sumich 1984; Mallonee 1991; Patten and Samaras 1977; Darling 1984; Calambokidis et al. 1991, 1992). Up to 17 gray whales have been documented entering Puget Sound, Washington, in a year and some have spent up to 4 months in the area (Calambokidis et al. 1991, 1992, 1993). Some of these whales return in multiple years with two whales seen in Puget Sound in three consecutive years (Calambokidis et al. 1993). Further, between 1986 and 1991, 5 of 23 gray whales individually identified while alive in Puget Sound were subsequently found dead (Calambokidis et al. 1991, 1992). This high proportion of gray whale deaths may be due either to the whales entering Puget Sound in poor health or to the exposure to contaminants affecting the whales' health. There has not been adequate information to determine which factor explains the apparent high rate of gray whales deaths in Puget Sound.

Gray whales feed primarily on benthic prey, though feeding on pelagic prey has also been documented (Nerini 1984). The whales use suction to engulf sediments and prey from the bottom, then filter out water and sediment through their baleen plates and then ingest the remaining prey (Nerini 1984). This feeding method often results in the ingestion of sand and other bottom materials (Rice and Wolman 1971). The dominant prey of gray whales in feeding grounds in Alaska are ampeliscid amphipods (Ampelisca macrocephala) though a variety of other benthic prey items are also consumed (Rice and Wolman 1971, Nerini 1984). Recent studies of gray whale feeding in northern Puget Sound have revealed predation on ghost shrimp (Callianassa californiensis) (Weitkamp et al. 1992). Thus, the potential exists for exposure to sediment associated contaminants if gray whales feed in urban embayments.

There is increasing evidence that chemical pollution in coastal areas near urban centers may be responsible for a variety of deleterious biological effects in aquatic species, from liver tumors and reproductive dysfunction (infertility, spawning failure) in bottomfish that reside on contaminated sediments (Varanasi et al. 1992a) and altered immune function in juvenile salmon after only a brief residency in waters of polluted estuaries (Arkoosh et al. 1991) to reproductive dysfunction in marine mammals (DeLong et al. 1973, Duinker et al. 1979, Reijnders 1986). In Puget Sound, sediments in several urban and industrialized bays have elevated concentrations of anthropogenic chemicals (Krahn et al. 1986). Several field and laboratory studies have provided considerable evidence that anthropogenic compounds present in contaminated sediments are probable causative agents for hepatic tumors, related lesions, and reproductive dysfunction (Myers et al. 1987, Johnson et al. 1989, Casillas et al. 1991). Thus, the stranding of a gray whale near Port Angeles, Washington, on the Strait of Juan de Fuca in 1984 heightened public concern that chemical contaminants in sediments may have been responsible for its death. Chemical analyses, conducted in our laboratory, of tissues (liver, blubber, kidney, brain) and stomach contents of this whale revealed that the concentrations of chlorinated hydrocarbons (CHs) such as polychlorinated biphenyls (PCBs), 1,1,1-trichloro-2,2-bis (p-chlorophenyl) ethanes (DDTs), and a number of toxic elements (e.g., mercury and lead) were at levels well below toxicological concern and also well below the concentrations reported in most cetaceans and pinnipeds (Wagemann and Muir 1984). The only exception was that the concentrations of aluminum were relatively high in both the liver and brain of this gray whale. It was not possible, however, to compare the results with published values. There is little information on a broad spectrum of contaminants in gray whale tissues. Wolman and Wilson (1970) reported the presence of DDT and its metabolites in 6 of 23 gray whales taken off of San Francisco, California, during their northern and southern migrations. The concentrations of DDTs, 1,1-dichloro-2,2-bis (p-chlorophenyl) ethanes (DDDs), and 1,1-dichloro-2,2-bis (p-chlorophenyl) ethenes (DDEs) in blubber of these whales ranged from 22 to 360 ng/g wet weight, whereas it was reported that liver did not contain any of these chlorinated pesticides. Schaffer et al. (1984) reported concentrations of DDTs of 470 ng/g wet weight in blubber of a gray whale sampled in southern California in 1976. Total PCBs were not detected in the blubber (<230 ng/g wet weight).

In recent years, a number of gray whales have stranded in Puget Sound, as noted above, raising, once again, the concern that chemical pollution may have played a role in their deaths. However, demonstrating a causal link between pollution and strandings of marine mammals is particularly difficult because of inherent problems of availability of a sufficient number of tissue samples from both healthy and stranded animals and the inability to conduct controlled laboratory studies with live animals, particularly the large marine mammals. In addition to these sampling and experimental difficulties, the lack of detailed information on the biology and migration patterns of gray whales makes it very difficult to make a definitive assessment of the role of chemical pollution in their mortality. Nevertheless, the stranding of marine mammals and the increased public and scientific awareness of the potential impact of anthropogenic chemicals make it imperative that this issue be evaluated using the best possible strategy and state-of-the-art methodologies.

There has been considerable controversy around the role of pollutants in the deaths of gray whales in Puget Sound (Calambokidis 1992). Primary factors that have been cited to support a link to pollutants in these deaths has been the poor condition of the liver in some of these animals, the bottom-feeding behavior of this species, and the presence of contaminants in stomach contents and tissues (Fouty 1984). The basis for most of these conclusions has been challenged (Calambokidis 1992), but in the absence of better information, the question of the role of contaminants has largely been unresolved.

In the case of the recent gray whale strandings, tissue samples were collected from a total of 22 animals stranded at locations in Puget Sound, along the Strait of Juan de Fuca and Strait of Georgia, along the outer Washington Coast, on Kodiak Island, Alaska, and in San Francisco Bay, California, from 1988 through 1991. These sites represent a wide range of chemical contamination in bottom sediment, from the relatively pristine Alaskan waters to the urbanized areas of Puget Sound and San Francisco Bay (Varanasi et al. 1989a). We obtained stomach contents, liver, and blubber tissues from many of these animals with the assumption that the chemical profiles of CHs and essential and toxic elements in the stomach contents would reflect the most recent exposure, and the profiles in liver and blubber would reflect longer-term bioaccumulation of contaminants. The results from analyses of these samples should provide some insight into the relationship, if any, between chemical contamination at the site of stranding and concentrations and profiles of selected contaminants in various tissues. Moreover, to determine whether a particular group of contaminants preferentially accumulated in specific organs such as brain or kidney, these organs were also analyzed.

In the present study, we included measurements of CHs, selected toxic and essential elements (e.g., mercury, lead, zinc, copper) and polycyclic aromatic contaminants (PACs), which consisted primarily of polycyclic aromatic hydrocarbons and the dibenzothiophenes (Table 1). Because CHs such as PCBs, DDTs and chlordanes are among the most widespread and persistent chemical contaminants in the near coastal environment (Varanasi et al. 1992b) and because of their lipophilicity and resistance to metabolism, these pollutants tend to bioaccumulate in aquatic organisms, particularly in lipid-rich tissues of marine mammals. Several toxic and essential elements were measured because of their toxicological significance and their possible accumulation in certain tissues of marine mammals. For example, mercury is nephrotoxic in mammals and it has been suggested that aluminum may alter brain function (Goyer 1986). Additionally, because gray whales feed on benthic organisms, a feeding strategy unique among baleen whales, stomach contents were analyzed for CHs, selected toxic and essential elements, and PACs to provide insight into sources and levels of these compounds available through diet. Further, because of the extensive metabolism by mammals and fish of contaminants such as the polycyclic aromatic hydrocarbons (Varanasi et al. 1992b and 1989b, Lee et al. 1972, Stegeman et al. 1981), the parent compounds generally are not detected in tissues, but may be present in the stomach contents of bottom feeding gray whales. Stomach contents consist of benthic invertebrates that do not efficiently biotransform PACs, as well as incidentally ingested sediment; the sediments from many urban areas contain elevated levels of parent PACs (Varanasi et al. 1989a).

Overall, the findings from this study showed that the concentrations of chemical contaminants in tissues of bottom feeding gray whales were substantially lower than the concentrations measured in certain pinnipeds and toothed cetaceans (Odontoceti) whose diets consist largely of fish. The findings also showed that there were no statistically significant region-specific differences in tissue concentrations or profiles of CHs and selected elements.

Samples of blubber, liver, kidney, brain, and stomach contents were collected by various scientists (see Acknowledgments) from 22 dead, beached gray whales from March 1988 to June 1991 (Fig.1, Table 2). It should be noted, however, that complete sets of tissue samples were not available for each whale. There was little information on the exact time of death; however, it was estimated that the time between death and necropsy ranged from days (1-2) to approximately 1 month; thus, the integrity of the gray whale tissue samples was poor in some cases due to the extended time between death and sampling. Five whales were stranded in the Puget Sound area, three on the Washington coast along the Strait of Juan de Fuca, one at Point Roberts along the Strait of Georgia, seven were stranded at sites along the Washington outer coast, four in San Francisco Bay, and two in Alaska. Fourteen of the whales were males and six were females; information on the sex of the Alaskan whales was not available. The ages of the whales were not determined. Twenty of the whales were measured and lengths ranged from 790 cm to 1,300 cm (Table 2).

More detailed biological information is available for 8 of the animals (CRC 334, CRC 397, CRC 398, CRC 401, CRC 395, CRC 402, CRC 332, CRC 337) than for the other 14. When considering this subset of eight whales, the sampling of all but one falls within the time period during or following the northbound migration. The single exception was whale CRC 395 sampled on 10 February 1991, which had been dead for several weeks, and had died during the period of the southbound migration. Four of these eight animals (CRC 334, CRC 397, CRC 398, CRC 401) had been individually identified while alive in Puget Sound using photographs of natural markings. They were first seen alive from 33 to 67 days prior to when they were found dead and were seen from 1 to 14 times while alive. All four were observed engaged in foraging behavior during one or more of these sightings. This technique of identifying and tracking gray whales has been used previously in biological studies of gray whales (Darling 1984, Calambokidis et al. 1991).

Chlorinated Hydrocarbons and Polycyclic Aromatic Contaminants

The analytical methodologies and quality assurance procedures for CHs and PACs were those used in the National Benthic Surveillance Project of NOAA's National Status and Trends Program (Krahn et al. 1988), except that the procedure for these analytes was modified to facilitate removal of interfering lipids, especially in blubber tissue. Briefly, tissue (1-3 g) and stomach contents (1-5 g) were macerated with sodium sulfate and methylene chloride. The methylene chloride extract was filtered through a column of silica gel and alumina, and the extract concentrated for further cleanup. The cleanup was done using size exclusion chromatography with high performance liquid chromatography (HPLC) (flow rate of 5 mL/min). A methylene chloride fraction containing the CHs and PACs was collected and then exchanged into hexane as the volume was reduced by evaporation to approximately 1 mL. The extracts were analyzed by capillary column gas chromatography (GC) with an electron capture detector for CHs. The PACs were determined by GC/mass spectrophotometry (MS) quantitation as outlined by Burrows et al. (1990). Chlorinated hydrocarbon peak identifications were confirmed on selected samples using GC/MS with selected ion monitoring.

Toxic and Essential Elements

The concentrations of 16 elements (Table 1) were determined using analytical methodologies and quality control procedures similar to those used in the National Benthic Surveillance Project of NOAA's National Status and Trends Program (Varanasi et al. 1989a). Briefly, thawed tissue (1.0-1.8 g) of liver, kidney, brain, and stomach contents was digested with 10 mL of concentrated ultra pure nitric acid for 2 hours at room temperature and subsequently heated in a microwave oven in a sealed Teflon bomb. The digestates were diluted with deionized water and the concentrations of elements were determined by atomic absorption spectrophotometry (mercury, arsenic, selenium, lead, iron, chromium, manganese, nickel, tin, silver) and inductively coupled argon plasma emission spectroscopy (copper, aluminum, zinc, cadmium, barium, strontium). In addition, the percent dry weight of the samples was determined by drying approximately 2 g of tissue in an oven (85°C) for 24 hours. After cooling the sample, the percent dry weight was calculated by dividing the weight of the dried tissue by the original wet sample weight and multiplying by 100.

Percent Lipid

To determine extractable lipids, an aliquot of the initial methylene chloride extract of tissue was filtered through filter paper containing approximately 5 g of diatomaceous earth as a filtering aid, and the solvent was removed from each sample using a rotary evaporator. After the solvent was removed, the mass of lipid was determined. The percent lipid was calculated by dividing the mass of lipid by the original sample wet weight and multiplying by 100. Using sea lion blubber (n = 5) and liver (n = 5) samples as test material, this procedure yielded results for total lipids comparable to those obtained using the method of Hanson and Olley (1963), a modification of the Bligh and Dyer (1959) method. The percent total lipids in the sea lion blubber determined by our method and the Hanson and Olley method were 84 ± 0.7% and 80 ± 2.2%, respectively, and for the sea lion liver were 2.7 ± 0.1% and 2.2 ± 0.2%, respectively.

Many studies of marine mammals report tissue concentrations of chemical contaminants on a wet weight basis, although the water content of tissues is highly variable and thus limits intertissue and interanimal comparisons. Herein, we discuss the data using the wet weight convention to compare to other literature values; however, percent lipid was determined for each sample and is reported in Appendix A. In addition, differences among studies in analytical methods and quality assurance measures make it difficult to rigorously compare contaminant concentrations in many cases.

Quality Assurance Measures for Toxic and Essential Elements,

Chlorinated Hydrocarbons, and Polycyclic Aromatic Contaminants

Quality control procedures included the use of standard reference materials (SRMs) and certified reference materials (CRMs), which allowed an evaluation of the accuracy of the analytical methods (Tables 3 and 4). In summary, the grand mean recoveries (± standard deviation) for selected analytes were 85 ± 35% for CHs and 110 ± 45% for the PACs, and duplicate analyses of four gray whale samples agreed within ± 12%. The recoveries were calculated from the mean recoveries for certain analytes in SRM 1974, which was analyzed in the present study (CHs, n = 6; PACs, n = 2), and from previous analyses of SRM 1974 (CHs, n = 15; PACs, n = 103 for parent compounds and n = 88 for alkylated PACs). The concentrations (wet weight) of CHs and PACs in the method blanks (n = 6) analyzed with the samples were low and near the limit of detection (CHs: 0.2-1 ng/g, PACs: 0.5-2 ng/g).

The mean recovery of toxic and essential elements from the CRMs was 103 ± 64% (n = 29), and the analyses of replicate CRMs (n = 113) agreed within ± 29%.

Statistical Methodology

The data on chemical concentrations were analyzed by one-way or multi-way analysis of variance (ANOVA), with the factors being site of stranding, sex of the animal, and the year of stranding when appropriate. The statistical analyses were somewhat limited by the availability of samples; for example, blubber was available from all 22 animals, but liver was collected from only 10 animals and the sex of 2 whales was unknown. Further, the data for organic contaminants and essential and toxic elements were log transformed (log (x + 1)) to reduce deviations from normality. The results of the statistical analyses were very similar whether concentrations were expressed on a wet weight, lipid normalized weight or dry weight basis, thus only the results for the analyses using concentrations based on wet weight are discussed in detail.