University of Alaska Oil Spill-Related Research

Presented at 1999 oil spill symposium

Legacy of Oil Spill

Abstracts and Posters

This site contains the abstracts and posters of research conducted by scientists at the University of Alaska and presented at the March 1999 symposium "Legacy of an Oil Spill: Ten Years After Exxon Valdez." University of Alaska researchers are denoted by bold type.

Session Titles

Population and Recovery Status

PAPER

A Reexamination of Events Influencing the 1993 Pacific Herring (Clupea pallasi) Population Crash in Prince William Sound, Alaska

The 1989 Exxon Valdez oil spill in Prince William Sound, Alaska, was an anthropogenic perturbation occurring between two major natural pulse perturbations, the 1964 earthquake and the 1997-1998 El Nino. The Pacific herring (Clupea pallasi) population crashed in 1993 and is only beginning to recover in 1998. Identification of damage caused by the oil spill and commercial fisheries (bait, sac-roe, and roe on kelp) were confounded by the lack of life history information. Presently the primary source of mortality is attributed to an increase in a viral infection triggered by exposure to oil. However, it is unclear how this disease spread throughout the herring population or if the disease could independently have caused the crash. New information on herring life history enables a reexamination of events influencing the 1993 population crash and suggests mechanisms that may have spread the disease through the population. It also suggests alternative sources of mortality that may have acted independently or in combination with the disease to cause the herring collapse and slow recovery.

PAPER

Trend and Habitat Use of Harbor Seals in Prince William Sound, Alaska, after the Exxon Valdez Oil Spill

We used aerial counts to monitor the trend in numbers of harbor seals (Phoca vitulina richardsi) in Prince William Sound (PWS) following the 1989 Exxon Valdez oil spill. Repetitive counts were made at 25 haulout sites during the annual molt period each year from 1990 through 1997. A generalized linear model indicated that time of day, date, and time relative to low tide significantly affected seal counts. When Poisson regression was used to adjust counts to a standardized set of survey conditions, results showed a highly significant decline of 4.6% per year. The number of harbor seals on the trend count route in eastern and central PWS has been declining since at least 1984 with an overall population reduction of 63% through 1997.

During 1992 to 1997, we attached satellite tags to 63 harbor seals in PWS to investigate their movements and diving behavior. In southcentral PWS, adult females displayed strong site fidelity, seldom traveled, and made relatively short and shallow dives. Subadults traveled greater distances within and outside PWS, made deeper and longer dives, and utilized a greater variety of depths when diving. During this study there appeared to be a change in the feeding locations of seals during winter-spring. Before 1995, only 2 of 30 tagged seals traveled to the Copper River delta. During 1995 to 1997, 8 of 19 spent time there. This is consistent with fatty acid analysis indicating that seal diets changed in 1995. It is clear from these tagging studies that some harbor seals in PWS move considerable distances to feed during winter months. The distance from southcentral PWS, where most seals were tagged, to the Gulf of Alaska (either near Middleton Island or the Copper River delta) is more than 100 km. This is greater movement than has been reported for harbor seals in most other studies.

Oceanography, Carrying Capacity, and Long-Term Monitoring

PAPER

The Case for Bottom-Up Control of Food Web Dynamics in Prince William Sound

Does the amount produced and the nature of the phytoplankton community determine the production of upper trophic levels, or is there always enough phytoplankton of an edible species so that it does not matter? Our results, from the past four years of field data collected in the SEA project, indicate that the phytoplankton bloom in Prince William Sound varies in duration, distribution, and composition. The spring phytoplankton bloom, dominated by diatoms, begins in April, fueled by nitrate from winter mixing, and extends into early May, when grazing and nutrient depletion take control. The lake-river system, though interannually variable, is reflected in an isotopic gradient of the plankton, and spatial data indicate that central Prince William Sound is comprised of two regions. An inner sound is consistently separated from the outer sound by a physical (upwelling) boundary located about 30 km north of the eastern exit to the Gulf of Alaska (Hinchinbrook Entrance). The primary production cycle differs in these two regions, giving some support to the early SEA hypothesis of lake vs. river segregation of the food web. Zooplankton biomass increase is sequential to that of the phytoplankton and peaks in late May. The data support the hypothesis that interannual variation in the production cycle directly affects subsequent trophic levels. For example, the bloom period was shortest in 1995 (29 days) and longest in 1997 (36 days), and while the biomass as measured by chlorophyll was similar in all three years, the calculated carbon biomass (based on cell size and number) was much greater in 1996. Such results should indicate that 1996 was a better food year than 1994 or 1997, and this is reflected in the pink salmon success of the given year.

PAPER

Declining Primary Productivity in the North Pacific: Past Implications for Marine Mammal Populations and Changes Ahead

Populations of Steller sea lions (Eumetopias jubatus) and harbor seals (Phoca vitulina) have declined in the past decade to 10-50% of initial populations in the western Gulf of Alaska and Bering Sea. Several studies have sought to determine if these declines arise from "top down" or "bottom up" controls. We have used stable carbon isotope ratios in whale baleen as a means of testing the hypothesis that climate change has forced a decrease in ecosystem carrying capacity. Recent findings in laboratory and natural environments indicate that carbon isotope ratios of phytoplankton are closely linked to cell growth rates where other floristic and environmental conditions are similar. Once incorporated into phytoplankton, the isotope ratios are conservatively transferred into the food webs supporting consumer organisms.

Bowhead whale (Balaena mysticetus) baleen, grown while the whales fed in the Bering-Chukchi seas, provides a multiyear temporal record of isotope ratios in their zooplankton prey and, by proxy, the phytoplankton supporting the consumer food webs. By using baleen plates from 26 whales archived at the Los Angeles County Museum and recently taken by Native hunters, an isotopic record was constructed extending from 1947 to 1995. From this, we infer that seasonal primary productivity in the Bering Sea was at a higher rate over the period 1947-1966, then underwent a general decline continuing to the most recent samples (1995). Assuming a close similarity to the published relationships established between primary productivity and carbon isotope ratios, the decline in the Bering Sea carbon isotope ratios suggests a loss of 35-40% of the carrying capacity of 30 years ago. This drastic decline is evident in recent zooplankton biomass estimates and is very likely implicated in the continuing decline of marine mammal populations in the western Gulf of Alaska and Bering Sea. Seeking the environmental physical and chemical causes is the focus for future work.

POSTER

Spatial Analysis of Zooplankton Distributions Using Acoustic Data and Two-Dimensional Kriging

Due to patchy distributions and the expense of collecting and processing net samples, zooplankton population data are often characterized by broad confidence intervals, with little detailed information on vertical and horizontal distributions. Although acoustic techniques and optical plankton counters in combination with nets can supply much more detailed information on plankton distributions, application of classical statistical procedures to such data sets is complicated by the potential for pseudo replication due to autocorrelation of the closely spaced data points. In this paper, we present a model-based geostatistical technique, two-dimensional kriging, for handling spatial data sets with autocorrelation. The basic principles of the technique are outlined and examples are provided from acoustic data taken in the western Aleutian Islands. A new computational technique is proposed to handle the large data sets often encountered in nonrandomly sampled acoustic surveys. This technique has two fundamental advantages: (1) it accounts for autocorrelation of the data and permits rigorous statistical inferences to be generated, and (2) it provides an effective technique for visualizing the results. These advantages may make this a fundamental technique for identifying changes in the size and distribution of populations related to climatic and anthropogenic influences.

Food Chain Effects

PAPER

Trends of Forage Fish Availability in Relation to Seabird and Marine Mammal Foraging Activities in Prince William Sound and the Outer Kenai from 1995 to 1998

Distributions of surface schooling forage fish, seabird numbers, seabird behavior, and selected marine mammal species were observed via aerial surveys from 1995 to 1998 in the oil spill-affected region. Forage fish species composition and species-specific abundance varied from year to year. Foraging patterns observed in gulls (mainly black-legged kittiwakes) also varied in response to changes in fish distribution. Marine mammal activity (mainly sea lions and humpback whales) was less variable from year to year, but still linked to abundance of surface schooling fishes. The annual and seasonal variability in fish distribution was compared with changes in ocean conditions. Seasonal changes in vertical fish distribution, affecting the availability of fish as forage, were observed. In general, juvenile herring (age 1) appeared in surface waters by May when thermoclines were fully established and water temperatures significantly increased. Mature pre-spawning capelin appeared in surface schools in June associated with huge foraging gull flocks. Sand lance appeared in huge numbers by July, the timing possibly related to when post-metamorphic age-0 juveniles have developed distinctive schooling behavior rather than to oceanographic conditions. Water temperatures generally peaked by that time. The variability observed in the distribution of forage fish and seabird foraging activity may be partially explained by the variations in oceanographic conditions. This environmentally induced variability has implications on seasonal and annual availability of forage fish as prey to apex predators. A long-term (10 yrs +), broad-scale data set of forage fish distribution is needed to better understand how large-scale changes in ocean conditions affect availability of fish as prey.

PAPER

A Perspective on Harbor Seal Trophic Interactions in Prince William Sound and the Gulf of Alaska

The number of harbor seals (Phoca vitulina) in Prince William Sound (PWS) has declined to approximately one-fourth of the 1975 population, and food web dynamics have been investigated as a possible factor. Stable carbon and nitrogen isotope ratios (delta13C and delta15N) are established in the primary producers of the food chain and provide information of prey consumed at different locations and trophic levels within PWS and the adjacent Gulf of Alaska. Offshore prey have more depleted isotope values than the same species within PWS. Prey from pelagic and benthic environments have similar delta15N values but the benthos is more enriched in 13C than pelagic species. Satellite tracking data of harbor seals tagged by Alaska Department of Fish and Game personnel reveal some seals leave the sound to feed in the gulf for extended periods of time. The nutritional values of these prey relative to pelagic prey, such as herring and capelin, are unknown but likely lower. Stable isotope values from archived and modern harbor seal tissues from 1950-1996 showed that no significant difference in the delta15N values occurred during the past 47 years and this indicated no trophic shift occurred during that time. However, during that same period, a decrease in the delta13C became evident in the bone collagen of three species of phocids, including harbor seals, and otariids throughout the Bering Sea and Gulf of Alaska. That decline tracked the decline in zooplankton biomass estimates and primary productivity rates as exhibited by the stable isotope records in bowhead whale baleen. All these data in combination may indicate that the carrying capacity of the northeastern Pacific Ocean and Bering Sea has declined since the 1960s and implies serious consequences for top trophic level organisms which require high primary productivity to support the prey density necessary for successful recruitment.

Sound Ecosystem Assessment

PAPER

Sound Ecosystem Assessment (SEA): Ecological Controls of Pink Salmon and Herring Production in Prince William Sound, Alaska

Oil spilled from the grounded tanker Exxon Valdez in late March 1989 dispersed to much of Prince William Sound, fouling beaches and the surface waters. Four years after the spill, pink salmon and herring populations had fallen to very low levels, prompting questions about the long-term health of the ecosystem. In response to strongly voiced concerns by the local fishing communities, the EVOS Trustee Council funded the Sound Ecosystem Assessment (SEA) program in 1994 to describe and model factors limiting the recovery of pink salmon and herring. SEA investigators and others speculated that the non-recovery was possibly linked to an oil-induced shift in the dominant fish community resulting in the realignment of food webs supporting and involving these two species. SEA evolved as a multi-project, interdisciplinary approach to understanding the factors limiting production of pink salmon and herring. Over the last five years, investigators have described and modeled specific aspects of the physical oceanography, plankton dynamics, and interactions between juvenile pink salmon and herring populations and their predators and competitors. These studies, coupled with historical production information for pink salmon and herring, provide a window into processes that regulate survivals during early life history. Juvenile herring and salmon are particularly vulnerable to bird and fish predation during early marine residence. Our work demonstrates that several factors including water temperature and plankton forage for all consumers modifies salmon fry mortality. In addition to predation on larval and postlarval herring, mortalities associated with starvation during the winter also establish recruitment levels to adult herring populations. We discuss these findings in relation to the recovery status of pink salmon and herring in Prince William Sound. We also present evidence that oceanographic conditions influencing planktonic forage in the region shifted dramatically in 1992, with implications for the survivals of both species.

PAPER

Physical Oceanography of Prince William Sound, Alaska

The purpose of the physical oceanographic part of the Sound Ecosystem Assessment (SEA) is to identify the dominant physical processes that influence the distribution and abundance of pink salmon and Pacific herring in Prince William Sound (PWS), Alaska. Two aspects of the physical oceanography that most impact the biological components of PWS (especially phytoplankton and zooplankton) are surface stratification and circulation, including "flushing" rates and extent. Observations of temperature (T), salinity (S), and current velocity combined with simulations from a numerical circulation model show that some aspects of the circulation and water mass properties of PWS are fairly predictable, while others are not. The water column is well mixed in winter (January through March). The water mass properties in March are the coldest and saltiest of all the seasons observed, and the most spatially uniform. In April, the surface waters start to warm and freshen, so that a temperature minimum layer is formed. The extent and intensity of this layer varies from year to year, as well as seasonally. The cold underlying water may prolong surface stratification formation and lead to a longer spring bloom. In May and June, the surface stratification strengthens. In the central sound, isopycnal doming is observed in April through September, with varying intensity, which may be responsible for bringing nutrients to the upper layer. In April through June the central sound circulation may be cyclonic or anticyclonic. The sense of the circulation seems to be determined by a combination of Gulf of Alaska inflow and surface T/S properties, while wind forcing modulates the strength. Upper layer flow into the sound occurs at Hinchinbrook Entrance and Montague Strait in response to easterly wind bursts. Deeper outflow accompanies the inflow at Hinchinbrook Entrance. This mechanism might contribute to the flushing of zooplankton out of PWS.

PAPER

Observed and Modeled Plankton Dynamics in Prince William Sound, Alaska

The subarctic waters of Prince William Sound exhibit strong seasonality in the timing, magnitude, and duration of the pelagic marine production cycle at lower trophic levels. Following a period of intense winter mixing, phytoplankton respond to increasing light levels and upper-layer stability by initiating a bloom that usually begins in mid- to late April. This bloom can be characterized by different diatom species each year. Phytoplankton become nutrient limited in late April and early May, and stocks exhibit a sharp decline at this time. A slight recovery can occur in late May and early June when a large component of the grazing community leaves the surface water. The composition of the phytoplankton community shifts to small flagellates as inorganic nutrients become limiting. Plant biomass is channeled into growing copepod and other zooplankton populations, some reaching peaks in May, June, and/or July. Modeling results support the notion that upper-layer stability controls the amounts of forage retained in pelagic food webs each year. When the surface layers are relatively unstable in April, the bloom is less intense but extended in duration so that zooplankters can efficiently graze available food stocks. Conversely, when the upper layers are more stable, the bloom is very intense but short-lived. Under these different conditions, much of the organic matter sinks before it can be eaten by zooplankton. A well-defined statistical relationship between shelf convergence south of Prince William Sound and amounts of zooplankton sampled in the southwestern portion of the sound changed radically in the historical record after 1992. The implications of this change are explored by modeling the retention of zooplankton populations during the early spring, and examining the influence of modeled stocks entering the region from the shelf at this same time.

PAPER

Ecological Processes Influencing Mortality of Juvenile Pink Salmon in Prince William Sound, Alaska

Predation appeared to be the primary mechanism causing mortality of juvenile pink salmon (Oncorhynchus gorbuscha) during the spring bloom in Prince William Sound, Alaska. Two planktivores, herring (Clupea pallasi) and walleye pollock (Theragra chalcogramma) probably consumed the greatest numbers of juvenile salmon, several piscivores (Gadus macrocephalus, Microgadus proximus, and Salvelinus malma) probably consumed the second greatest numbers, and an assemblage of nearshore demersal fishes (Cottidae, Hexagrammidae, and Sebastes spp.) probably consumed the smallest numbers of salmon. Mortality of pink salmon released from Wally H. Noerenberg Hatchery, adjacent to our study area, was greatest in 1994 (-0.0091), decreasing in 1995 (-0.0074), 1996 (-0.0070), and 1997 (-0.0056). This decline in mortality corresponded to a decline in age 3+ pollock densities in our study area. Relative attack/capture probabilities for planktivores feeding on juvenile salmon decreased over salmon lengths from 3 to 5 cm, but increased over this same length range for piscivores and demersal fishes feeding on salmon. Over the length range from 6 to 7 cm, relative attack/capture probabilities appeared to decrease for all predator groups. Our results indicate that the copepod Neocalanus plays a unique role in modifying predation losses of juvenile pink salmon. Reduced densities of this large calanoid copepod probably cause: (1) reduced growth when juvenile salmon densities are high, (2) greater predation losses to planktivores as these predators switch to alternative prey, and (3) dispersion of foraging juvenile salmon offshore leading to greater predation losses to piscivores and demersal fishes. Analyses of historical coded-wire tag data from PWS hatcheries indicated that mortality of juvenile pink salmon was reduced when the copepod bloom was prolonged (p = 0.013), but increased juvenile growth rate and body weight upon ocean entry were also strongly correlated (p < 0.001) with reduced mortality.

The importance of these size-related parameters to mortality is consistent with planktivores as the dominant predator. Our results indicate that bottom-up processes affecting the duration of the spring Neocalanus bloom and juvenile salmon foraging behavior modify top-down processes involving the timing of feeding mode shifts toward piscivory and size-dependent predation losses of juvenile pink salmon.

PAPER

Life History of Herring in Prince William Sound, Alaska

Following the Exxon Valdez oil spill it became readily apparent that little was known about the early life history of herring in Prince William Sound. When the fishery collapsed in 1993, the need for that knowledge became readily apparent. The Sound Ecosystem Assessment (SEA) program addressed this problem. The herring component of this project focused on physical and biological variables across space and time during 1995 to 1997. Effects on the success of spawning were examined. Dispersal of larvae from known spawning sites was simulated using a physical circulation model which treated larvae as passive drifters. Surveys of the sound discovered that bays were used as principal nursery areas within the sound. Mechanisms to retain the juveniles within these bays were studied. The spatial distribution, seasonal variability, density, and growth of juvenile herring were determined. Spatial and temporal trends in feeding, food availability, isotopic signatures, and energetics of juveniles were examined during the first two years. It appeared that the amount of energy a juvenile herring acquired during its first summer was critical to survival through its first winter. Little to no feeding took place during the winter, and the smallest fish starved in the lab and disappeared in the field. Differences in feeding and energetics were apparent among nursery areas during both the summer and winter seasons, indicating that the quality of the habitat was not equal in all areas of the sound. Oceanography differences accounted for these inequalities.

PAPER

Inter-Seasonal and Annual Changes in the Distribution and Abundance of Walleye Pollock (Theragra chalcogramma) in Prince William Sound

A series of hydroacoustic surveys were conducted from 1993 to 1998 that assessed the abundance and distribution of walleye pollock (Theragra chalcogramma) and other dominant pelagic fishes in Prince William Sound. These surveys found pre-spawning walleye pollock to be pelagic in distribution, highly contagious in their wintertime distribution, and least contagious at the start of their spring feeding migration. Vertical distribution of the pre-spawners was between 150 and 300 m with highest densities found adjacent to steep shoreline drop-offs. Vertical distribution and contagiousness of the feeding adult walleye pollock was highly variable in the spring, observed as individuals or in schools ranging in depth from the very near surface (40 m) to over 150 m. These distributions appear to be closely associated with the location of prey in the water column. In June 1994, a strong year class of age-0 walleye pollock first appeared in the midwater layer at about 20 m in depth, and moved inshore to the bays and shorelines as the summer and fall progressed.

We estimated the biomass of pre-spawning pollock at 38,000 t in 1995, 37,000 t in 1997, and 114,000 t in 1998. In 1998, the biomass jumped due to the recruitment of a large 1994 cohort. Because of the numerical abundance of walleye pollock in Prince William Sound, predation by adult and juvenile walleye pollock could easily impact the survival of juvenile fish such as pink salmon fry as they migrate to the ocean in the spring. However, walleye pollock appear to be targeting macrozooplankton (copepods, euphausiids, and pteropods) during the spring outmigration of salmon fry. The implications of adult walleye pollock feeding on macrozooplankton in the spring are discussed with respect to climate change and the survival of other marine species.

Subsistence, Communities, and Human Dimensions

PAPER

Distribution of Herring and Other Forage Fish as Observed by Resource Users

This study was initiated in 1997 to document the historical distribution of forage fish through qualitative interviews with key respondents in several communities. The information will be provided to researchers with the Alaska Predator Ecosystem experiment (APEX) and the Sound Ecosystem Assessment (SEA). It is designed to provide researchers with population and habitat use information over a longer period and broader area than can be known through existing records and current data collection efforts.

A combination of key respondent interviews and mapping was used to document the locations of juvenile herring, herring of unknown age, mixed adult and juvenile herring, herring spawn, unidentified small fishes, capelin, sand lance, eulachon, cod, and pollock. Researchers concentrated on recording information about juvenile herring and small fishes other than juvenile herring, such as sand lance and capelin.

Respondents from Cordova, Homer, Seward, Chenega Bay, and Tatitlek were identified through chain referral and interviewed in person. Respondents included retired professional biologists, professional spotter pilots-salmon fishermen, charter captains, fishermen, pilots, and other experienced resource users. Responses were mapped on navigational charts. Associated information was recorded on Excel spreadsheets and in indexed notes. Most interviews were tape recorded. Results include an analysis of the observations, maps of the observations in Prince William Sound and the outer Kenai Peninsula, and summary of the ecological knowledge shared by respondents. There is also a discussion of methods for identifying informative classes of respondents and possible future monitoring efforts.

Nearshore Predators and Their Prey

PAPER

Interactions between Sea Otters, Sea Urchins, and Kelp following the Exxon Valdez Oil Spill: Making Practical Uses of Paradigms in Ecology

One of the most widely held paradigms in ecology involves interactions between sea otters, sea urchins, and kelp. Sea urchins that graze kelp are a preferred food of sea otters. Where otters are abundant, there are few large sea urchins and dense stands of kelp. Areas without otters have high densities of large urchins and less abundant kelp. Based on this paradigm, we predicted that a lack of recovery of sea otters in heavily oiled portions of Prince William Sound would be accompanied by an increase in the number of large urchins and possibly by a decrease in kelp. We measured sea otter, sea urchin, and kelp populations after the Exxon Valdez oil spill to help assess the recovery status of otters, to document changes in the community, and to examine how these changes fit with the existing sea otter-sea urchin-kelp paradigm.

The density of sea otters was reduced by about 87% in a heavily oiled area around northern Knight Island after the spill, and from 1995 through 1998, averaged 62% of the pre-spill density. Densities of otters in an unoiled area off Montague Island were about 8-20 times higher than those at northern Knight Island. Where otter densities were reduced, the size distribution of sea urchins was strongly skewed toward larger individuals and urchin densities within preferred urchin habitats were higher. These changes are consistent with predictions of the sea otter-sea urchin-kelp paradigm and provide further evidence of a lack of recovery by otters. However, few urchins at either Knight or Montague islands were larger than 40 mm (a size common in areas without otters) and most were found in cryptic habitats (a common response to sea otter predation). Furthermore, there were few indications of increased grazing of kelp by urchins and kelp densities remained high in oiled areas. Thus, sea otters continue to exert strong top-down control of the system even at the greatly reduced density observed at northern Knight Island.

PAPER

The Pigeon Guillemot in Prince William Sound: Evidence of Injury, Status of Recovery, and Factors Limiting Populations following the Exxon Valdez Oil Spill

The objectives of our study were to determine the extent to which pigeon guillemot (Cepphus columba) populations in Prince William Sound (PWS) were injured and have recovered following the Exxon Valdez oil spill (EVOS), and to determine what factors are limiting their recovery. Guillemot populations in PWS were impacted by EVOS. Population censuses conducted before and after the spill demonstrate that densities of guillemots declined significantly more along oiled shorelines than along unoiled shorelines, and that impacted populations have not recovered. Populations at Jackpot Island (unoiled site) have increased since the spill, while those at Naked Island (oiled site) have not. In an attempt to understand what factors are limiting recovery, we compared the reproductive ecology of guillemots at these two sites. We also examined the physiological health of PWS guillemots to test the hypothesis that residual oil contamination is limiting recovery. We found no differences in reproductive success between the two sites, although chicks appeared to grow more slowly and fledge at lower weights at the oiled site. Our investigations of chick provisioning suggest that meal sizes were lower at the oiled site compared to the unoiled site, resulting in chicks being fed fewer calories per day. In addition, high-lipid prey fishes were more common in the chick diets at Jackpot, suggesting that prey quality may differ among sites. Our analyses of blood and tissues sampled from guillemot chicks at the two sites provide no evidence of exposure to residual hydrocarbons. Furthermore, none of the biomarkers that we assayed provide any indication that the physiological health of the chicks at Naked Island was compromised. Analyses of adult blood parameters are not yet completed. Current evidence suggests that recovery has been constrained by patterns of food availability; however, high rates of predation also appear to be limiting population growth.

PAPER

Chronic Effects of the Exxon Valdez Oil Spill on River Otters: Is Recovery Complete?

Our previous research on river otters (Lutra canadensis) indicated that these mustelids suffered chronic effects from the Exxon Valdez oil spill. We continued our studies on populations of otters inhabiting previously oiled and non-oiled areas of Prince William Sound, Alaska, during spring and summer of 1996 and 1997. Unlike data from 1996, otters did not exhibit significant differences in values for haptoglobin or endothelial P4501A1 between oiled and non-oiled areas in 1997. Likewise, no significant differences occurred in body mass (corrected for age class, sex, and total length of body), or age structure of populations inhabiting our two study areas. No significant differences existed in the abundance of marine fishes between oiled and non-oiled areas, or between latrine sites and random sites in either study area. We believe it is unlikely that food in our study sites could be limiting populations of river otters. We discovered that some of our telemetered animals moved between the two study sites. This reduced our power to detect a difference in blood values or P450 between oiled and non-oiled areas. Moreover, some animals were using freshwater habitats at our non-oiled site; we confirmed data from telemetry with a stable-isotope analysis of otter hair to assess diet (i.e., marine or freshwater). These different foraging strategies have the potential to bias our data on body condition because the prey base of freshwater systems is lower than for marine environments. Despite these confounding factors, we believe otters are recovering from the effects of the oil spill.

PAPER

Quantification of Cytochrome P450 1A as a Bioindicator of Exposure of Nearshore Vertebrate Predators to Residual Oil from the Exxon Valdez Oil Spill

A major component of the Nearshore Vertebrate Predator (NVP) ecosystem study addressed the question of continued exposure to residual oil from the 1989 Exxon Valdez oil spill as a possible factor limiting recovery of the predator species. To evaluate exposure, we measured expression of cytochrome P450 1A (CYP1A), an enzyme induced by polycyclic aromatic hydrocarbons, found in oil, and halogenated aromatic hydrocarbons. We compared CYP1A levels in animals from oiled and non-oiled areas of western Prince William Sound. Three methods were used to measure CYP1A: (1) quantify the amount of CYP1A in cells of skin samples, using an immuno-histochemical assay and CYP1A specific antibodies, (2) quantify ethoxyresorufin O-deethylase rates, catalyzed by CYP1A, in liver samples, and (3) quantify the mRNA for CYP1A, using cDNA probes, in peripheral blood lymphocytes. Different methods were necessitated by differences in tissues required for the technique and availability among species. Samples were collected from four predator species, harlequin ducks, river otters, sea otters, and Barrow's goldeneyes, and one prey species, masked greenlings, during 1996 to 1998. For all five species, average CYP1A levels were significantly higher in oiled areas than in non-oiled areas. The consistent difference between oiled and non-oiled areas is clear evidence of greater contaminant exposure for organisms in oiled areas relative to non-oiled reference sites, continuing at least into 1998. However, we cannot state conclusively that the source of this contamination was the 1989 spill, versus other petroleum or organochlorine contaminants in the environment. Continuing research on elevated CYP1A levels will include sampling animals across a broader geographical area, identification of other potential contaminants in the study areas, and dosing studies to interpret how oil exposure and variation in CYP1A expression might relate to individual behavior and health.

PAPER

Responses of River Otters to Oil Contamination: A Controlled Study of Biological Stress Markers and Foraging Success

Studies following the Exxon Valdez oil spill strongly indicate a correlation between oil contamination and physiological stress in coastal river otters (Lutra canadensis). This circumstantial evidence requires verification through controlled experiments. This project is designed to explore experimentally the effects of oil contamination on physiological and behavioral responses in river otters. Fifteen young adult male river otters captured in PWS in spring 1998 and transferred to the Alaska SeaLife Center via air are exposed to two levels of oil contamination under controlled conditions. Samples of blood, tissues, and feces are collected for analysis of biomarkers and immunological examinations. In addition, behavioral observations on foraging behavior are being conducted to explore the effects of oil contamination on foraging success.

POSTER

Hydrocarbons on Prince William Sound River Otter Fur

Approximately eight years after the Exxon Valdez oil spill, river otters were trapped from the shoreline in both oiled (Knight Island) and non-oiled (Jack Pot Bay) areas of Prince William Sound. River otters were captured at latrine sites from May to August in both 1996 and 1997 using either Hancock or #11 Sleep Creek double jaw leghold traps. Once in captivity, the animals were swiped with gauze presoaked in isopropanol. After wiping both sides and the underside of the river otter body, the gauze was wrapped in foil and frozen until analyzed in the lab in Fairbanks, Alaska. The gauze was extracted with isopropanol and petroleum hydrocarbons in the wipe extracts were analyzed by GC-MS. Hexadecane and pentadecane were detected, while l-methyl naphthalene was not. Besides these hydrocarbons, phenanthrene, chrysene, pentacosane, and hexacosane were also studied. This data is useful when evaluating the role of fur and grooming in exposing river otters to hydrocarbons.

Nutrition, Physiology, and Disease

POSTER

Alaskan Harbor Seals: Indices of Health, Nutrition, and Population Alteration

Because harbor seal populations in Prince William Sound have not recovered after the Exxon Valdez oil spill, this work asked, are the seals healthy and do they show signs of food limitation? We approached these issues by sampling a large number of adult seals (>300) to determine how blood chemistry varied by known factors, such as region, season, seal age, etc. About 40-50% of the variability in blood chemistries was explained by these factors.

We then examined the 50% of variability that could not be explained by these factors. First, we utilized criteria that would signify medical problems but did not find any patterns suggesting these animals were medically compromised. However, seals inside PWS demonstrated elevated levels of inflammatory stress. Second, we examined patterns that would suggest food limitation, but none of the predicted indicators in blood chemistry or anatomy were found. Finally, a statistical method known as "outlier frequency" indicated that seal populations could be distinguished by subtle changes in blood chemistry that may be indicative of chronic, low-level medical stress or differing fish diets.

We expanded our work in 1997 to include harbor seal pups. In 1997, we could not see significant medical, anatomical, or nutritional differences in these pups, but in the 1998 summer season we will expand our sampling pattern and size.

Work began in 1998 at the Alaska SeaLife Center to feed harbor seals controlled diets of differing fish (pollock, herring, salmon, etc.) to quantify how diet can change blood chemistry and anatomy in seals. This work will determine if the differences seen in harbor seal blood chemistry using the outlier frequency methods can be reproduced in the laboratory by differing fish diets.

These data suggest that declining seal populations are not driven by significant medical problems or food limitation.

Alaska Predator Ecosystem Experiment

PAPER

The Food Web Supporting Forage Fish Populations in Prince William Sound, Alaska

The forage fish, such as sand lance (Ammodytes hexapterus), herring (Clupea harengus), capelin (Mallotus villosus), and juvenile walleye pollock (Theragra chalcogramma), on which many seabirds depend for food, feed on zooplankton. A variety of jellyfish species also consume zooplankton as well as the egg and larval stages of fish. In this overview, we will explore how the seasonal abundances and distributions of zooplankton relate to the population patterns of forage fish and jellyfish. We will evaluate the extent of dietary overlap between forage fish and jellyfish and their potential for food competition. We will draw on data, including aerial, acoustic, and net surveys, from the APEX and SEA programs from 1994 to 1997, with an emphasis on comparisons among years that showed marked differences in environmental variables. We will speculate on how the different species respond to interannual and regional environmental variability, and on the relative importance of competition and environmental condition in determining relative abundances of forage species.

POSTER

Jellyfish in Prince William Sound: Abundance, Aggregation, and Feeding Rates on Zooplankton

Jellyfish are abundant predators of zooplankton, and therefore are potential competitors of forage fish species. We summarize data from the SEA and APEX projects on the distributions and abundance of the several jellyfish species. Jellyfish populations vary seasonally, with the greatest numbers found in June and July. Jellyfish populations also differed annually. In 1996, jellyfish were markedly more diverse and abundant as compared with other years (1994 to 1997). We examined the distributions and abundances of moon jelly (Aurelia aurita) aggregations from aerial, acoustic, and underwater video data. The diets of the abundant jellyfish species consisted primarily of copepods and larvaceans, which also were the main prey of the forage fish species. In situ feeding rates of zooplankton were calculated from gut contents and digestion rates, and combined with abundances of jellyfish and their zooplankton prey in order to determine their potential for competition for food with forage fish.

Relevant Alaska Sea Grant publications

ASG Legal Team Final Report

Lessons of the Exxon Valdez

A Challenge Met: Fishermen Fight the Spill of the Exxon Valdez

After the Exxon Valdez Oil Spill

Prevention, Response, and Oversight Five Years After the Exxon Valdez Oil Spill: Proceedings of an International Conference

The Selendang Ayu Oil Spill: Lessons Learned

North Aleutian Basin Energy-Fisheries: Workshop Proceedings