APPENDIX A:

EXXON & SPILL "SCIENCE"

Exxon’s decision to conduct its own damage assessment program was a major financial committment which could not have occurred without some expectation–or plan–of comensurate returns.

Exxon is one of America’s most successful companies: in 1993, Exxon reported sales of $111 billion and profits of $5.3 billion (Economist 1994). Exxon consistently outperforms the other top oil firms on both after-tax return on capital and on its shareholders’ funds (Economist 1994). After a century of selling oil, the concept of cost efficiency is well entrenched at every level (Economist 1994).

As with the government research, the bulk of Exxon’s studies were confidential until April 1993, when Exxon presented the results of its damage assessment program two months after government scientists presented their studies. In sharp contrast to the government studies, Exxon’s studies found minimal residual oil, rapid recovery of natural resources, and virtually zero long-term effects. The comparative analyses in section 1 of this paper show that that Exxon consistently underestimated Exxon Valdez oil concentrations and injury in its key scientific studies.

By manipulating data and selectively presenting information (NOAA 1993), Exxon constructed its own "reality" of spill impacts in an attempt to mitigate its outstanding liability of tens of billions of dollars in civil lawsuits over the spill. Similarly, Exxon also manipulated data and people during its 1989 oil spill cleanup in order to make people believe the cleanup was successful and damages were minimized, even though Exxon had itself concluded almost immediately that they could not clean up the spill (Mauer 1992, Mayer 1993, Pyle 1993). Exxon promoted the person in charge of public relations during the spill, Don Cornett, to chief of public relations (Mauer 1993). Exxon also tried to manipulate the public impression of the government’s data set on the Exxon Valdez oil spill (Exxon 1993), but the attempt backfired when NOAA organized its own media response (NOAA 1993).

The theorem for this sort of reality construction, i.e. "appearances can be deceiving" (Edelman 1988), and further examples of Exxon falsifying data and/or manipulating information to influence public perception are presented elsewhere (CBS 60 Minutes October 1993, Cohen et al. 1992, Exxon 1992, Fry 1993, Ott 1989, U.S. Congress 1990, Wall Street Journal 1993).

Marine Mammals

About 4,000-6,000 marine mammals were killed in the Exxon Valdez oil spill, including whales, seals, sea otters, more than in any other spill on record.

Killer Whales - Pre-spill research in Prince William Sound documented about 182 killer whales in nine family units or "pods" of two distinct types (Dahlheim & Matkin 1993, EVOS Trustees 1992 p. 23-24, Matkin et al. 1993). Resident pods are characterized by strong maternal groups with low birth and death rates, resulting in stable numbers over time. Resident pod members remain together for life: members missing for more than one season are considered dead.

Post-spill research (Dahlheim & Matkin 1993, Matkin et al. 1993) found that the resident AB pod, a group of 36 whales in 1988, lost 7 whales in 1989 and 6 whales in 1990. The mortality rate for the AB pod was 3.1% in 1988, compared to 19.4% in 1989, 20.7% in 1990, 4.3% in 1991 (one whale), and 0 in 1992. The 1989 and 1990 mortality rates are unprecedented for North Pacific resident killer whales and strongly correlate with the Exxon Valdez oil spill. The potential for additional mortalities within the AB pod exists due to loss of breeding age females, disruption of social structure, and physiological abnormalities (collapsed dorsal fin in males): recovery of the population to pre-spill levels (36) could take at least 15 years (Dahlheim & Wright 1993).

Nine members of the transient AT pod have also been missing since the oil spill (Matkin et al. 1993). However, the cause of these disappearances is less well understood than is the case for the resident AB pod, because pod members may immigrate and emmigrate between groups..

Sea Otters - Between 3,500 to 5,000 sea otters were estimated to have died by acute exposure to oil in the spill-impacted area, including 1,600 from Prince William Sound alone. Up to 2,200 more sea otters may have died in Prince William Sound in the first three years since the spill, or about 38% of the total pre-spill population in the sound (EVOS Trustees 1992 p. 20-23).

Mortalities of weanlings were abnormally high in western Prince William Sound 1990-1992: weanlings eat more intertidal mussels than adults, and the intertidal mussels are known to remain heavily contaminated with Exxon Valdez oil in the western sound (Gorbics 1993, Rotterman & Monet 1993). Mortalities of prime age adults were also abnormally high for the three years following the spill: chronic sublethal problems such as physiological impairment and variation in DNA content of blood lymphocytes may indicate long-term problems (EVOS Trustees 1992). High mortalities of prime age adults in eastern Prince William Sound also occurred, following the reintroduction of animals released from treatment centers (Monet & Rotterman 1993).

Since these studies indicate that sea otters in Prince William Sound continue to be stressed and are still experiencing high mortalities, the estimated time for recovery is unknown.

Harbor Seals - A government study (Frost & Lowry 1993) estimated 345 harbor seals died in Prince William Sound because of the oil spill, but this number probably greatly underestimated the total number of seals killed throughout the sound and the Gulf of Alaska. Debilitating brain lesions, caused by exposure to oil, were found in harbor seals in oiled areas in 1989: the lesions would have interfered with nerve transmissions and would account for the animals’ lethargic behavior. "This brain damage would have made it very difficult for seals to perform normal tasks such as swimming, diving, feeding, and escaping from predators... [It] probably made them more susceptible to drowning, being hit by boats, or being caught by predators such as killer whales" (p. 301). Presence of high levels of oil break-down products in seal tissues (bile, blubber, mammary tissue and milk) in 1990 show that seals were either still encountering oil in their environment, or they were burning stored fat reserves that had high levels of oil contamination, or both.

Pup production was lower in oiled areas in 1989 than in 1990 or 1991, relative to similar comparisons in unoiled areas. By 1992, seal numbers were still lower in oiled areas than they were before the spill. However, recovery is difficult to predict, because the seal population in this area was declining before the spill from unknown causes. (It is declining even faster now in oiled areas than unoiled areas.)

Humpback Whales - The single available study (von Ziegesar & Dahlheim 1993) did not find any measureable impact of the Exxon Valdez oil spill on reproductive success or habitat use of humpback whales. However, in 1989 humpback whales may have avoided (they were not seen) in Lower Knight Island Passage, a preferred but heavily-oiled habitat. They were observed in this area in 1988 and 1990.

Terrestrial Mammals

River Otters - It was impossible to determine how many river otters died from the oil spill, because of the lack of pre-spill data and the likelihood that sick otters were not found by cleanup workers. However, long-term effects from oil, such as generally poor body condition, a less diverse diet, avoidance of preferred habitat because of oil, larger home ranges, elevation of blood parameters indicative of damage and a depressed immune system, were found in otters in oiled areas for two years after the spill (Bowyer et al. 1993). These problems suggest that river otters may have major delayed effects from the oil spill as weakened animals die. Feeding on oil-contaminated food, such as intertidal mussels, will further delay recovery.

Brown Bear, Deer and possibly Mink - were affected to an unknown but possibly minimal extent.

Birds

More seabirds were killed in the Exxon Valdez oil spill than in any other spill on record. Direct mortalities after the spill were estimated between 300,000-645,000 total birds. Lost chick production for 1989-1991 was estimated at over 300,000.

Bald Eagles - An estimated 900 bald eagles were killed by oil throughout the spill area in 1989, or about 11% of the population (Bowman & Schempf 1993). In Prince William Sound, reproductive (nest) failures of up to 85% occurred near heavily-oiled areas in 1989, with confirmation of oil-contamination in prey and eggs (Bowman & Schempf 1993). The population appears to have recovered by 1990, indicating the effects of the oil spill were short term.

Murres - Estimated acute mortality of adult murres was 300,000, with some colonies lossing 50-70% of the breeding birds (Nysewander et al. 1993). Numbers have continued to remain low, with low reproductive success possibly attributed to young, inexperienced breeders, later than normal breeding, reduced densities, and disruption of synchronized breeding behavior. Estimates of resulting lost production are at least 300,000 chicks. Recovery time is estimated to be 18 to 50 years (Heinemann 1993).

Harlequin Ducks - There was a direct mortality of at least 400 harlequin ducks in 1989 (Patten 1993). Near total reproductive failure has occurred in western Prince William Sound since the spill through 1993, delaying recovery and adding further impact by failure to achieve normal levels of replacement of old and dying ducks. Oil-contaminated tissues and poor body condition were found in many ducks in oiled areas: these ducks feed heavily on mussels and other invertebrates in intertidal areas. There is concern that unless steps are taken to remove oil from mussel beds, a local extinction of harlequin ducks may occur within the western sound (Patten 1993).

Black Oystercatchers - Estimates of black oystercatchers killed by oil in 1989 in the total spill area were 120 adults with a lost potential chick production of 1290 (Sharp & Cody 1993). However, black oystercatchers have continued to show a chronic or delayed effect from the oil spill: in 1992, there were 42% less birds than expected in oiled areas compared to unoiled areas (Klosiewski & Laing 1993). Females feeding on oil-contaminated mussels were reproductively impaired: they produced fewer eggs of smaller volume than females in unoiled areas (Sharp and Cody 1993). Young fed oil-contaminated mussels grew more slowly, despite being fed more food, but fledging success was similar to unoiled sites (Sharp and Cody 1993). Recovery time is unknown, as these birds continue to experience long-term impacts.

Other Birds - Significant and substantial spill-induced declines were found in populations of harlequin ducks, black oystercatchers, pigeon guillemots, northwestern crows, scoters, loons, mew gulls, cormorants and arctic terns: the most intense injury was exhibited by harlequin ducks, black oystercathers, pigeon guillemots and northwest crows, all summer breeding resident species that feed in intertidal areas (Klosiewski & Laing 1933). The number of species showing significant oil-induced declines increased from 1989 to 1990 to 1991, indicating a long-term or delayed effect of the oil (Peterson 1993).

Fish

Several species of coastal and offshore fish exhibited evidence of oil impacts over a large geographic area, although the greatest long-term damage was found to occur in eggs and larvae in extensive intertidal zones in the spill-impacted areas.

Pink Salmon - Egg mortalities of wild stocks in spill-impacted areas were high in 1989 in the intertidal zone, and progressively higher in 1990 in the upper intertidal zone, and in 1991 across all stream zones, indicating potential genetic damage (Bue et al. 1993). Larvae from oiled streams showed gross morphological deformities such as club fins and curved spines. Preliminary data from incubation studies, when compared to fall 1993 field data, confirm that "there is a difference in embryo survival between oil-contaminated and control streams and that these differences are carried by the parents" (Bue 1993).

Although salmon exposed to oil as fry (or eggs) in 1989, returned as adults in record numbers in 1990 (and 1991), the fish were unable to produce competitive offspring: 1992 (and 1993) pink salmon runs failed disastrously, 30% (and 20%) of anticipated in Prince William Sound, while all other runs throughout Gulf of Alaska were strong in 1993 (ADF&G 1994). Alaska Department of Fish & Game stated that: "it is probable that the extremely low returns of salmon in 1993 are symptomatic of widespread ecological damage within Prince William Sound."

Sockeye Salmon - Spill-related overescapement (from fishery closures due to oil) in Kenai, Kodiak and Chignik areas resulted in poor juvenile escapement, which may lead to declines in adult sockeye of 20%-50% in Kodiak and 90% in Kenai returns in 1994, 1995 and perhaps beyond (Schmidt, et al. 1993).

Herring - Herring spawned intertidally during the spill and resulting embryos were genetically damaged, morphologically abnormal (89%), and had a high incidence of eye tumors (Biggs et al. 1993). Larval growth was severely reduced. The 1989 year class largely failed to survive to adulthood: it was virtually missing in the 1993 spawning migration (Funk et al. submitted 1993).

The 1988 year class, which was exposed to oil as 1-year old fry during spill, were reproductively impaired as adults in 1992.

In 1993, the herring fishery was closed due to a weak return (20%) with high incidence of visible lesions (ADF&F 1993b , Kocan accepted 1994). In 1989, uptake of oil in adults was found to cause internal tissue damage in 1989 and 1990 which may have weakened the fish’s immune system and ability to resist diseases (Arkoosh et al. 1993).

Other Fish - Dolly Varden and cutthroat trout survival and growth were lower in oiled areas in 1989 and 1990, with some recovery in 1991 (Hepler et al. 1993). Increased mortality and sublethal injuries (lesions) from oil exposure occurred in rockfish in 1989. Measureable amounts of Exxon Valdez oil were found in bile of halibut; pollock (Whitney 1990); rock, yellowfin and flathead sole; pacific cod; and in 1990 sablefish and Dover sole (EVOS Trustees 1992 p. 34).

The Exxon Valdez oil spill presented a rare opportunity for scientists to study the effects of oil in a pristine, productive, cold-water marine ecosystem. This opportunity, combined with the tremendous publicity surrounding the oil spill in Alaska, attracted hordes of researchers to Prince William Sound. Unfortunately, the only available source of funding for immediate response studies was through the NRDA process. This forced scientists to comply to the tort process, which severely restricted information flow and the scope of research (Cummings 1992 p. 385-387). NRDA studies tend to focus on species that will show the largest degree of damage in the shortest time for the least money. However, despite these problems, over 70 studies were initiated under the NRDA process.

While useful as research projects, many of the studies did not fit within the NRDA guidelines which stipulate that studies must be cost-effective, not duplicative, and use existing baseline data, among other things (Cummings 1992). These stipulations exist to make the NRDA process selective, i.e. so that it is used only to fill in the gaps with previous and on-going research. Since, supposedly, only those NRDA studies designed for litigation are granted confidentiality, the regulations protect the government’s case, while giving the public access to other spill response studies.

However, because of the nonselective screening process and lack of baseline research, the bulk of studies initiated in response to the Exxon Valdez spill were classified as NRDA studies and were granted the confidentiality accorded to all NRDA research. The net effect of this was that the public was kept virtually ignorant of spill impacts for nearly four years(4), which compromised the overall success of the cleanup, restoration, civil settlement, and preventing future spills.

There are other advantages to timely and full disclosure of data as well. Notably, it takes away a powerful bargining chip (e.g. access to the NRDA data) for governments to use against third party litigants. Governments guilty of contributing to industrial spills through complacency should not be allowed to buttress their case behind the shield of confidentiality. Only full disclosure of the NRDA data will enable the public to criticize its government and force necessary improvements in spill prevention and response.

Timely and full disclosure of data would also help prevent "sweetheart" settlements between the government(s) and the spiller for recovery of damages to natural resources, by allowing the public to comment meaningfully on settlement terms (Cummings 1992 p. 397).

The Exxon Valdez settlement history is a classic example of why damage assessment data need to be made public. The first sweetheart settlement fell apart in March 1991 when its immenient closure was leaked to the public and Native communities, the latter who demanded to be included in settlement negotiations. The second sweetheart settlement was finalized in October 1991, because the state and federal governments cut deals with the Alaska Natives and fishermen to allow these third party plaintiffs access to the NRDA data in trade for dismissal of their lawsuits against the two governments (for contributing to the spill by complacency in oversight roles). However, it became evident over a year later, when a confidential survey was released, that the final settlement significantly undersold the greater public interest (Phillips 1993b).

Although there are various mechcanisms for relief to the confidentiality provisions of CERCLA (Cummings 1993 p. 400-409), none would provide adequate satisfaction when applied to the collective experiences with Exxon Valdez oil spill and its aftermath.