A Ten-Year Study of Shoreline Conditions in the Exxon Valdez Spill Zone, Prince William Sound, Alaska

David S. Page
Bowdoin College, Department of Chemistry, Brunswick, ME 04011

Edward S. Gilfillan
Bowdoin College, Environmental Studies Program, Brunswick, ME 04011

INTRODUCTION
On March 24, 1989 just past midnight, the tanker Exxon Valdez deviated from the shipping lane in Prince William Sound (PWS), Alaska to avoid icebergs and grounded on Bligh Reef resulting in the release of 37,000 tons (10.9 million gallons) of Alaska North Slope (ANS) crude oil. This was about 20% of the 180,000 tons of crude oil the vessel was carrying when it struck the reef. The salvage effort that took place immediately after the grounding saved the vessel from sinking, thus preventing a far larger oil spill from happening. Figure 1 shows the remaining cargo being off-loaded from the stricken vessel during the salvage operation. While the largest oil spill from a vessel in US history, Figure 2 shows that the Exxon Valdez oil spill is an "average" large oil spill in world terms. Other spills have been much larger, often involving the complete loss of a vessel and cargo. Examples include the Ixtoc-1 blowout off the coast of Mexico in 1978 (about 400 million gallons), the tanker Amoco Cadiz off Brittany, France in 1978 (69 million gallons), the tanker Torrey Canyon off the English coast in 1967 (38 million gallons) and the tanker Metula in the Straits of Magellan in 1973 (16 million gallons). As a result of these oil spills and others, there has been a considerable effort by government, academic and industry scientists to understand the fate and effects of petroleum in the marine environment. A key review of this work is the 1985 National Research Council (NRC) report "Oil in the Sea: Inputs, Fate and Effects". The NRC review found no evidence that the oceans' ecosystems are seriously threatened by oil spills. Petroleum inputs from accidental oil spills were found to be less important contributors to the annual input of petroleum to the marine environment than chronic discharges from urban runoff, industrial waste, and transportation activities. It is also important to keep in mind that petroleum is a natural product and is released into the marine environment in significant amounts naturally at many oil seeps around the world. The literature indicates that, while initial impacts of oil spills can be severe, there are very effective natural mechanisms that produce rapid recovery in most spills.

salvage operation

Figure 1. The salvage operation prevented the vessel from sinking and saved 80% of the cargo, thus preventing a far larger oil spill.

oil spills ranking

Figure 2. Ranking of major oil spills by amount of oil spilled.

Figure 3 shows the time course of the oil spill from PWS to the Gulf of Alaska. The Exxon Valdez spill was really 2 oil spills in one. In Prince William Sound, shorelines in the western part of PWS were exposed to petroleum that had not undergone much environmental degradation beyond the loss of volatile hydrocarbons. In the Gulf of Alaska outside of PWS, oil that left PWS underwent extensive environmental degradation (weathering) while on the water. The Kenai/Kodiak/Alaska Peninsula impact was very spotty and of a much more weathered (less toxic) nature (see Maki 1991). Most of the oil that left PWS was dispersed at sea .

gulf of alaska region

Figure 3. Time course of the movement of the leading edge of the oil from Prince William Sound into the Gulf of Alaska Region.

Prince William Sound
Prince William Sound is a large, partially enclosed body of water with many islands and a mainland having a complex convoluted coastline with an overall length of about 5,000 km. There are many bays, fjord and active glaciers associated with this productive sub-arctic environment. PWS has a total area of about 39,000 km2. Figure 4 shows the location of the grounding in the northern part of the sound and illustrates the vast area of PWS. Coastal ocean currents flowing from east to west enter PWS from the east and flow out through the west, causing a change of water every 3-4 weeks. This water flow contributed to the movement of oil out of the sound and resulted in very low concentrations of petroleum in the water column. Much of the topography of the present shoreline where oiling occurred is a result of uplifting (1-3 m) during the 1964 earthquake (Plafker 1965). This means that much of the present shoreline of the sound is geologically young.

vessel aground on bligh reef

Figure 4. The northern end of PWS showing the vessel aground on Bligh Reef in the context of the vast area of PWS.

The environment of Prince William Sound is far from fragile. The continuing stress of physical factors from harsh winters and severe storms has produced plant and animal communities that are very resilient. The Exxon Valdez oil spill is one of many sources of hydrocarbons in PWS (e.g. see: Page et al., 1995a; Page, et al., 1996; Page, et al., 1999). PWS has a long history of human activity and there are numerous sites of both present and past human activities including active settlements, fish hatcheries, fish camps and recreational campsites in addition to abandoned settlements, canneries, sawmills, and copper mining camps (Page, et al., 1999). Hydrocarbons in the marine environment derived from fossil fuel use in these activities are superimposed on a regional background of natural petroleum hydrocarbons derived from oil seeps and eroding petroleum source rocks in the eastern Gulf of Alaska . Figure 5 summarizes the various sources of hydrocarbons in the PWS environment from past and present human activity and natural sources. Natural oil seeps (Figure 6) are common in the Gulf of Alaska region. PWS supports diverse and abundant plant and animal communities. Seabirds and marine mammals in PWS and the Gulf of Alaska were directly affected by the spill in 1989 and a 45 million dollar wildlife rescue and rehabilitation program was undertaken right after the spill .

hydrocarbon sources

Figure 5. Hydrocarbon sources in PWS. By 1999, any remaining traces of the Exxon Valdez oil spill are a very minor contributor to the total input of petroleum hydrocarbons from all sources.

oil creek

Figure 6. Petroleum is a natural product and is naturally present in the environment at many locations in the Gulf of Alaska region. Oil creek is a salmon stream on the Alaska Peninsula that has petroleum inputs from natural seeps and is representative of many seeps throughout the region.

There are major commercial herring and salmon fisheries in PWS. There is no evidence of adverse effects from the spill on these fisheries . It is important to keep in mind that the vast majority of the salmon streams in PWS were not oiled at all (see Maki et al., 1995). Fishery statistics in PWS are highly variable from year to year as shown in Figure 7. This figure also shows that the all-time record salmon runs were in 1990, 1991, 1994 and 1999, indicating no long-term adverse spill effects on the commercial fisheries.

pink salmon runs

Figure 7. Pink salmon runs (millions of fish) in Prince William Sound for the period 1970-1999 (Data source: Alaska Department of Fish and Game, Division of Commercial Fisheries annual fin fish management reports for the PWS area.).

Approximately 486 miles or 16% of the total shoreline of the sound was oiled to some degree in 1989 (Maki 1991; Neff et al. 1995). This means that 84% of the shoreline never was oiled at all. As with most oil spills, the shoreline was very discontinuously oiled. Most of the shoreline impact in PWS was light (>70%) and recovery from the initial effects of the spill were rapid in most places. Figure 8 shows the distribution of oil on the shorelines of western PWS in 1989 and 1992.

oil distribution

Figure 8. The distribution of oil on the shorelines of western PWS in: (A.) 1989 and (B.) 1992. Through a combination of an extensive cleanup program and natural recovery, there was a dramatic and ongoing removal of oil from the shorelines by biological and physical means. By 1992, remnants of the spill were present at a small number of "worst case" locations.

The Cleanup
A massive cleanup effort was undertaken right after the spill. Directed by the Federal On-Scene Coordinator, in consultation with state agencies, Exxon engaged in cleanup operations in PWS and the Gulf of Alaska during the summers of 1989, 1990 and 1991 and the spring of 1992. At the height of the cleanup effort in 1989, over 11,000 people were involved. A key element of the cleanup program was the shoreline survey program carried out by Shoreline Cleanup Assessment Teams (SCAT). These shoreline surveys were conducted by teams of experienced professionals that included a marine ecologist, an oil spill geomorphologist, an archaeologist and other representatives of government agencies, landowners, and Exxon. These surveys, begun in April of 1989 provided data on shoreline physical characteristics and oiling conditions on and were used to set priorities and methods for shoreline cleanup and to protect sensitive natural and cultural resources (see Neff et al. 1995). The SCAT program provided data that not only supported the cleanup efforts, but also produced a database of shoreline oiling information that supported scientific efforts.

Initially, the goal of the cleanup was to remove as much oil from affected shorelines as possible, with some locations being treated several times in 1989 (Figure 9). Cleanup methods used in 1989 included the manual removal of oil with sorbent pads, low- and moderate-pressure cold and warm water washing coupled with nearshore oil skimmers, mechanical removal of oiled sediments and tilling of shoreline material and bioremediation. Most of the cleanup effort was directed to the upper and middle intertidal shoreline zones that received most of the oil. Lower intertidal areas, which are biologically much more productive were generally not oiled. The physical removal of oil in 1989 and the natural cleaning of oiled shores during storms in the winter of 1989/90 brought about a dramatic reduction in oil remaining in PWS in the spring of 1990 and allowed less intrusive cleanup techniques to be used in subsequent years. These included tilling, physical removal of tar mats and the spreading of oil-soluble fertilizer to promote microbial degradation of petroleum residues (bioremediation). These measures coupled with natural oil degradation processes were very successful in reducing the amount of remaining residues of the spill and in June of 1992, representatives of the federal and state governments determined that no additional cleanup of shoreline was warranted, and the cleanup program ended. Figure 10 illustrates the dramatic recovery of the oiled shorelines for a boulder beach that was heavily oiled in 1989.

clean-up operations

 Figure 9. 1989 cleanup operations showing the deployment of workers on the shorelines and the use of omnibarges to wash the shorelines with warm water to mobilize the oil and allow it to be collected by skimmers. The goal was to remove as much oil from the shorelines as possible to give natural recovery process a head start.

 shoreline recovery

Figure 10. The dramatic recovery of a heavily oiled shoreline from 1989-1992. The 1989 picture shows pools of oil on an exposed boulder beach. The 1992 picture of the same beach shows no oil. Many exposed north facing shores on the islands in PWS were in the direct path of the spill an were heavily oiled in 1989. Through a combination of natural and human processes, most of the oil from the spill was gone from PWS by 1992.

By the end of 1992, the total miles of oiled shoreline had decreased from 486 in 1989 to approximately 6.2, with 92% of the oiling being very light. Heavily oiled shoreline areas decreased from 87.6 miles in 1989 to approximately 0.06 miles in 1992 as a result of cleanup operations and natural processes of recovery (Neff et al. 1995). Table 1 summarizes the dramatic decrease in oiling levels in PWS going from 1989 to 1992. By 1999, remnants of the spill can be found as isolated deposits of weathered oil at a small number of specific locations that had originally been very heavily oiled. The vast majority of these deposits are the top of the tide zone in a form and location on the shoreline that poses minimal, if any, threat to wildlife.

Table 1. Linear miles of surface oiling of shores in Prince William Sound 1989-92 (Adapted from Neff, et al., 1995).

Surveyed
Oiling Category
Year
Miles
Very
Light
Narrow/
Light
Medium/
Moderate
Wide/
Heavy
Total
Oiled

1989

901

139

203

58.4

87.4

486

1990

689

201

49.7

28.6

13.0

292

1991

240

42.3

9.3

7.5

0.06

59.1

1992

19.9

5.4

0.5

.4

0.1

6.4

Scientific Studies Document Rapid Recovery
A large number of scientific studies were conducted by Exxon-supported scientists in 1989, 1990 and 1991 to assess the fate and effects of the spill on the shorelines of Prince William Sound and the Gulf of Alaska (for a collection of papers on these studies, see Wells et al. 1995). In addition, studies of the effects of the spill were also conducted by scientists supported by State of Alaska and Federal agencies.

The shoreline ecology program (SEP) was begun in 1990 and 1991 and continued in 1998 and 1999 (Boehm et al. 1995; Gilfillan et al. 1995; Page et al. 1995; Page, et al. 1999a). The goal of the SEP was to understand and assess the recovery of the intertidal and nearshore subtidal biological communities in the spill zone and to determine the persistence of petroleum residues from the spill. The study design was influenced by the extensive experience of the scientists involved with prior spill studies. A key part of the study design was taking into account natural ecological process that can produce biological effects that could be confused with spill effects. An important part of post-spill studies is the use of the correct definition of "recovery." It is not appropriate to define recovery as a return to "pre-spill conditions" because natural ecosystems are in a state of constant change. For example, temperature and salinity are two physical variables that can have profound ecological effects. Figure 11 shows that there were large changes in temperature and salinity in the spill zone after 1989. Other non-spill sources of environmental stress in PWS include, storm events, predation by sea otters, sea stars, birds and killer whales and storm induced shoreline damage by ice and logs (Gilfillan, et al., 1995). Plant and animal communities are in a constant state of flux, termed "inter-annual variation". There is no reason to expect that the community of animals and plants present at some location this year will be the same in the future. The ecology of an area will change over time in the absence of any human perturbation, such as an oil spill. One can detect recovery by comparing biological and chemical measurements made at oiled sites with those from comparable unoiled reference sites. An area is recovered when it is no longer different from an appropriate reference area (Figure 12). Comparing groups of oiled and unoiled sites increases the statistical power to detect any oil spill effect.

Recovery has been documented for all types of environments affected by oil spills (e.g. see: NRC, 1985). The presence of weathered oil does not connote injury. If recovery is defined as when the last molecule of spilled oil is degraded, then it will be a long time coming. In every spill situation there will be locations where, for one reason or another, oil will be protected from rapid degradation. These sites are invariably few in number and small in area. If one defines recovery as that time when the vast majority (99.9 +%) of the impact zone is indistinguishable from unoiled reference areas, then recovery from oil spills is rapid.

annual surface water means

Figure 11. A. Sea surface temperature at the mouth of Resurrection Bay just west of PWS from 1985 through 1999. B. Sea surface salinity at the mouth of Resurrection Bay from 1985 to 1999. This natural variation will produce ecological changes that can be confused with spill effects.

baseline conditions, not pre-spill conditions

Figure 12. Recovery is defined as a return to baseline conditions &emdash; what the spill zone would have been like had the spill not occurred. Since natural systems are in a constant state of change, defining recovery as a return to pre-spill conditions is not valid.

The study design for the SEP (Page et al., 1995) involved concurrent sediment sampling for chemical, biological and toxicological analysis to provide data to assess recovery. The study design had two components, each with different objectives. A 1990 stratified random sampling (SRS) program was used to extrapolate findings on recovery to all of PWS. A 1990-1999 fixed (non-randomly selected) site program was initiated to understand the process of recovery of "worst-case" oiled shorelines over time. The locations of the SEP sites are shown in Figure 13.

SEP site locations

Figure 13. The locations of the 1990-1999 Shoreline Ecology Program study sites in PWS.

The SRS program was comprised of 64 sampling sites that were stratified across four levels of oiling (none or reference, light, moderate, and heavy) and the four principal habitat types in PWS (exposed bedrock [16.8 % of the shoreline], sheltered bedrock [56.6 % of the shoreline], boulder/cobble [22.9 % of the shoreline], and pebble/gravel [3.7 % of the shoreline]).

shoreline habitats

Figure 14. The 4 main shoreline habitat types in the Exxon valdez spill zone. Salt marsh/soft sediment environments where oil is known to persist are rare in the spill zone.

There were four randomly selected sampling sites within each oiling-level, habitat-type strata. These sites were randomly selected from large groups of candidate sites using the computer-geographical information system (GIS) database produced by the SCAT program. For each site, 3 transects were sampled over 5 intertidal and subtidal elevations. These transects were 30 m apart and were found to behave independently in most statistical analyses. This gave the study design a high statistical power to detect any effect of the spill if present. The SRS sites (Figure 13) were sampled only in 1990. The overall study design is summarized in Figure 15.

The fixed site component of the SEP program assessed recovery at 12 non-randomly chosen sites of special interest or concern (Figure 13). These sites were mainly exposed boulder/cobble and pebble/gravel habitats and represented some of the worst oiling conditions in the Sound. Included were two soft-sediment sites, habitats not common enough in PWS to be included in the SRS program. Four of the 12 fixed sites had received no cleanup treatment. Fixed sites were sampled in 1990, 91, 98, and 99. For purposes of statistical comparison 9 reference sites from the SRS program (five boulder/cobble and four pebble/gravel) were also sampled in these same years along with the 12 fixed sites.

shoreline ecology program

Figure 15. Summary of the study design for the Shoreline Ecology Program (SEP).

Overall Recovery Results
The SRS study demonstrated that, based on biological community structure parameters, approximately 90% of the plant and animal communities throughout the spill zone had recovered by the summer of 1990. There was a rapid and ongoing decrease in petroleum residues from the spill, even at the heavily oiled study sites. This result was consistent with the fact that most of the spill zone was lightly oiled in 1989 (Table 1). By 1991, any remaining effects were found to be limited to a small number of "worst case," locations, where the initial oiling was very heavy or cleanup activities had been limited. Sediment toxicity measurements done as part of the 1990 and 1991 SEP showed that, by 1991, the oil had weathered to the point where its toxic effect on animals living in the shoreline sediments was not distinguishable from that at unoiled sites (Boehm et al. 1995). It was estimated (Neff, et al. 1995) that there was an approximately 75% overall decrease in the amount of surface and subsurface oil in PWS each year between 1989 and 1993 and presumably thereafter.

Shoreline Chemistry
Polycyclic aromatic hydrocarbons (PAH) are taken as an indicator of petroleum persistence since they are known to weather more slowly than aliphatic hydrocarbons and are associated with the toxicity of petroleum in the environment. A conservative threshold measure of the potential for PAH toxicity in biota is the "effects range low" (ER-L) value of 4022 ppb for the total PAH (TPAH ) concentration in sediment (Long et al. 1995). By 1999, the TPAH concentrations at all of the boulder/cobble worst-case sampling stations were at least 3 times lower than this ER-L value, thus indicating low potential for adverse effect. By 1999, all oiled boulder/cobble samples had PAH concentrations close to the range of those for the unoiled reference sites. Table 2 summarizes the 1990-1999 sediment chemistry results for the worst-case boulder/cobble sites sampled.

Table 2. 1990-1999 decrease in total PAH (TPAH) concentration in sediments at worst case and unoiled reference sites in PWS. Data are given as the averages (± standard deviation) for 3 stations at each tide level (see text). The 1999 data are all lower than the 4022 pbb ER-L 10% probability of adverse effect sediment quality guideline for TPAH.

Site

SCAT Segment

TPAH (ppb) 1990

TPAH (ppb) 1999

Upper Intertidal Zone

     

Knight Island Lower Passage

KN-103

8955 ± 15407

39 ± 28

SW Green Island

GR-09

14919 ± 12190

135 ± 68

SW Green Island

GR-1A

11728 ± 12504

812 ± 68

SE Knight Island

KN-405

2508 ± 2384

64 ± 12

NE Latouche Island (Setaside)

LA15

447 ± 93

98 ± 34

NW Latouche Island

LA-21

176 ± 171

169 ± 129

Latouche Island - Sleepy Bay

LA19

8025 ± 7136

1177 ± 1469

Knight Island - Point Helen

KN405

263 ± 267

89 ± 31

Unoiled reference site average

 

34 ± 54

69 ± 150

       

Middle Intertidal Zone

     

Knight Island Lower Passage

KN-103

60639 ± 104849

48 ± 31

SW Green Island

GR-09

361 ± 97

60 ± 10

SW Green Island

GR-1A

1724 ± 451

336 ± 25

SE Knight Island

KN-405

29 ± 36

88 ± 45

NE Latouche Island (Setaside)

LA15

477 ± 328

325 ± 444

NW Latouche Island

LA-21

8041 ± 11679

154 ± 43

Latouche Island - Sleepy Bay

LA19

566 ± 433

187 ± 40

Knight Island - Point Helen

KN405

583 ± 314

43 ± 32

Unoiled reference site average

 

35 ± 67

202 ± 590

1998 Shoreline Surveys
The authors conducted shoreline surveys of most of these worst-case sites in 1993, 1994, 1996, 1998 and 1999. In all cases, there was a dramatic decrease in the extent of oiling; where the remaining weathered oil residues were in isolated deposits at the top of the tide zone. Remaining deposits of weathered oil were generally found in protected pockets in boulder beach areas, where the immovable boulders slow down the natural removal and breakdown of oil by mechanical wave action. The biota appeared normal for the type of shoreline area surveyed. At all locations, an abundance of newly set young of the year of all species was observed, indicating a high level of recruitment. As in many locations characterized by unstable substrate and exposed to high wave energy, much of the surface biota at these sites are essentially seasonal, with abundant recruitment in the spring and summer when conditions are favorable, followed by disruption during winter storms. There is no visual evidence that the biota were stressed by oil.

Shoreline Biology
Biota affected by the Exxon Valdez recovered much more quickly than was popularly expected. The rapid recovery observed is a product of the physical environment in PWS as well as the extensive cleanup that was carried out after the spill. Over 95% of the area oiled by the spill consists of bedrock and boulder shorelines. Many of these areas are exposed to very strong wave action during winter storms. Far from being a benign environment sheltering a fragile flora and fauna, PWS is a very harsh environment. Only those species that are good recruiters, i.e., those that can replace killed individuals rapidly, can persist in Prince William Sound. These same animals and plants were able to rapidly recolonize bare space created by the oil spill and subsequent cleanup. The rapid recovery of oiled bedrock and boulder beaches following the Exxon Valdez oil spill is similar to that observed for the same types of environments following other spills such as the Amoco Cadiz oil spill in France (Page, et al, 1989; NRC, 1985).

The biological communities of the boulder/cobble beaches that were sampled in 1998 and 1999 illustrate the types of processes occurring in PWS. Where there is protection from wave exposure, either as wave shadow behind an offshore outcrop or in the form of a rock too large to move by wave action, sediments are stable. In these areas there will be many species of shoreline plants and animals and individual plants and animals may live for several years. The diversity of these biological communities is high. In areas of high wave exposure boulders and cobbles move during winter storms and beach sediments are unstable. In these unstable areas, the resident plants and animals may be exterminated during winter storms. It is common to see polished areas at the base of the larger rocks in exposed areas where the moving sediment particles (some as big as melons) have scrubbed off all the biota. In these areas it is not uncommon for juveniles of such species as barnacles, snails, algae, and mussels to heavily colonize rock surfaces during periods of favorable weather in the spring and summer. These animals and plants persist through the summer and fall and most are killed by sediment movement during the following winter. Consequently, in unstable sediment areas, many species are essentially annuals. This is perfectly normal and has no connection with the oil spill.

The relationship of plants and animals living in PWS to the remaining residues of the spill is such that the remaining oil is in advanced stages of decomposition as a result of naturally occurring processes. The vast majority of the remnants of the spill is in a form and location that is not available to the biota. It is located very high in the tide zone where water reaches only on the highest spring tides or during storms. It is remote from the lower part of the tide zone that is inhabited by animals and plants.

CONCLUSIONS
This study examined the worst case shoreline sites in PWS where oil spill effects would be expected to be the most severe and the most persistent. Natural inter-annual variability is the largest and most consistent signal observed in this study, not the residual effects of the oil spill. Any oil spill impact study must explicitly address the importance of inter-annual variability in both study design and data interpretation. Pre-spill data are not a valid benchmark. Oil spill studies must take factors other than the oil spill into account in the analysis and interpretation of results. These factors include sediment grain size, sediment organic carbon and habitat type and wave exposure. A few oiling effects were detected in 1990-1991; none were observed in 1998-1999. In the present study, large inter-annual differences in the biological community structure ecology of boulder/cobble beaches in Prince William Sound, seem to have been caused in large part by long-term changes in sea surface temperature and salinity. Effects of the oil spill were restricted to the two years after the spill and were subsequently superimposed by larger natural changes caused by climate change and other non-spill factors.

The overall conclusions of this 10 year study are summarized as follows:

  1. Weathered remnants of the spill are present at a small number of sites in PWS. These sites were originally very heavily oiled and represent a minute fraction of the total shoreline area of the sound. Any isolated deposits of remaining oil residues from the spill are highly weathered and therefore not in a form that is available and toxic to biota.
  2. Between 1990 and 1999, there was a dramatic decrease in the concentrations of petroleum hydrocarbons at the "worst case" boulder/cobble" sites studied. The concentrations of PAH measured at these sites in 1999 were less than one-third of the ER-L sediment quality guideline of 4022 ppb for total PAH.
  3. Most remaining oil residues from the spill at the boulder/cobble sites studied are in the uppermost part of the intertidal zone that is only covered by the highest spring or storm tides. These events occur infrequently, on the order of 20 days a year. Thus any remaining spill residues are rarely made available, in any form, to the abundant biota in the middle and lower intertidal zones.
  4. The authors surveyed these sites in 1993, 1994, 1996, 1998 and 1999. The changes observed at these sites are consistent with natural processes of oil weathering and removal that we have observed at many other sites around Prince William Sound and with other oil spills world wide.
  5. Not only was there a large decrease in the total area of spill residues on the shorelines surveyed, but the nature of the remaining oil residues changed as well. Natural processes of physical and biological weathering during the period since1990/1991 have reduced the toxicity of any oil residues present in 1998 and 1999. Most of the remaining spill residues are asphalt-like in character, lacking the toxic low-molecular weight PAH fractions, and are not in a form that is available for absorption by marine plants and animals on the shore. Other sources of petroleum from on-going human activities in PWS are now a much more important source of petroleum hydrocarbons to the marine environment than the residues from the spill.
  6. This study found that natural inter-annual variability is now the largest and most consistent factor affecting biological communities in PWS, not the residual effects of the oil spill. By 1998 and 1999, no statistically significant effect of the spill on the ecological structure of biological communities could be detected. Any impact study must explicitly address the importance of inter-annual variability in both study design and data interpretation.
  7. In designing this study, the authors have tried to avoid the 3 cardinal sins of oil spill science. They are:
    • Only heavily oiled locations are studied that are chosen in a biased manner.
    • Unoiled reference sites that don’t have the same key ecological features as the oiled sites are chosen in a biased manner
    • All chemical and biological effects observed are assumed to be due to the oil spill

Acknowledgement
This study was supported by the ExxonMobil Company, U. S. A.

Websites for further study
www.valdezscience.com
http://www.bowdoin.edu/faculty/d/dpage/index

REFERENCES
Boehm, P. D., Page, D. S., Gilfillan, E. S., Bence, A. E., Burns, W. A. & Mankiewicz, P. J. (1998). Study of the fates and effects of the Exxon Valdez oil spill on benthic sediments in two bays in Prince William Sound, Alaska. I: Study design, chemistry and source fingerprinting. Environmental Science and Technology 32, 567-576.

Boehm, P. D., Page, D. S., Gilfillan, E. S., Stubblefield, W. A. & Harner, E. J. (1995). Shoreline Ecology Program For Prince William Sound, Alaska, Following The Exxon Valdez Oil Spill: Part 2&endash;Chemistry. In Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters, ASTM Special Technical Publication # 1219 (P. G. Wells, J. N. Butler and J. S. Hughes, ed.) American Society for Testing and Materials, Philadelphia, PA. pp. 347-397.

Gilfillan, E. S., Page, D. S., Boehm, P. D. & Harner, E. J. (1995). Shoreline Ecology Program for Prince William Sound, Alaska, Following the Exxon Valdez Oil Spill. Part 3: Biology. In Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters ASTM Special Technical Publication # 1219 (P. G. Wells, J. N. Butler and J. S. Hughes, ed.) American Society for Testing and Materials, Philadelphia, PA. pp. 398-443.

Johnson, C. B. and D. L. Garshelis (1995). Sea otter abundance, distribution, and pup production in Prince William Sound following the Exxon Valdez oil spill . "Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters ASTM Special Technical Publication # 1219". (P. G. Wells, J. N. Butler and J. S. Hughes, ed.) American Society for Testing and Materials Philadelphia, PA. pp. 894-929.

Long, E. R., Macdonald, D. D., Smith, S. L. & Calder, F. D. (1995). Incidence of adverse biological effects within ranges of chemical concentrations in marine estuarine sediments. Environmental. Management 19, 81-97.

Maki, A. (1991). The Exxon Valdez Oil Spill: Initial Environmental Impact Assessment. Environmental Science and Technology 25: 24-29.

Maki, A. W., E. J. Brannon, L. G. Gilbertson, L. L. Moulton and J. R. Skalski (1995). An assessment of oil-spill effects on pink salmon populations following the Exxon Valdez oil spill - Part 2: Adults and escapement. "Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters ASTM Special Technical Publication # 1219". (P. G. Wells, J. N. Butler and J. S. Hughes, ed.) American Society for Testing and Materials. Phildaelphia, PA. pp. 585-625.

Neff, J. M., E. H. Owens, S. W. Stoker and D. M. McCormick (1995). Shoreline oiling conditions in Prince William Sound following the Exxon Valdez oil spill. "Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters ASTM Special Technical Publication # 1219". (P. G. Wells, J. N. Butler and J. S. Hughes ed.) American Society for Testing and Materials. Philadelphia, PA. pp. 312-346.

Page, D. S., Foster, J. C., Fickett, P. M. & Gilfillan, E. S. (1989). Long-term weathering of Amoco Cadiz oil in soft intertidal sediments. Proceedings of the 1989 Oil Spill Conference, Washington, D.C., American Petroleum Institute, pp. 401-406

Page, D. S., Boehm, P. D., Douglas, G. S. & Bence, A. E. (1995a). Identification of Hydrocarbon Sources in the Benthic Sediments of Prince William Sound and the Gulf of Alaska Following the Exxon Valdez Oil Spill. In Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters ASTM Special Technical Publication # 1219 (P. G. Wells, J. N. Butler and J. S. Hughes, ed.), American Society for Testing and Materials, Phildelphia, PA. pp. 41-83.

Page, D. S., Gilfillan, E. S., Boehm, P. D. & Harner, E. J. (1995). Shoreline Ecology Program for Prince William Sound, Alaska, Following the Exxon Valdez Oil Spill. Part I: Methodology. In Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters ASTM Special Technical Publication # 1219 (P. G. Wells, J. N. Butler and J. S. Hughes, ed.), American Society for Testing and Materials, Phildelphia, PA. pp. 263-296.

Page, D. S., P. D. Boehm, G. S. Douglas, A. E. Bence, W. A. Burns and P. J. Mankiewicz (1996). The natural petroleum hydrocarbon background in subtidal sediments of Prince William Sound, Alaska. Environmental Toxicology and Chemistry 15: 1266-1281.

Page, D. S., Gilfillan, E. S., Neff, J.M., Stoker, S. W. and Boehm, P. D., (1999a). 1998 Shoreline Conditions in the Exxon Valdez Oil Spill Zone in Prince William Sound. In: Proceedings of the 1999 Oil Spill Conference March 8-11, 1999, American Petroleum Institute, Washington, D.C., pp. 119-126.

Pearson, W. H., E. Moksness and J. R. Skalski (1995). A field and laboratory assessment of oil spill effects on survival and reproduction of pacific herring following the Exxon Valdez spill. "Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters ASTM Special Technical Publication # 1219". (P. G. Wells, J. N. Butler and J. S. Hughes ed.). American Society for Testing and Materials. Philadelphia, PA. pp. 626-661.

Wells, P. G., Butler, J. N. & Hughes, J. S., Ed. (1995). Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters, ASTM Special Technical Publication # 1219. Phildelphia, PA., American Society for Testing and Materials. 929 pp.

Wiens, J. A. (1995). Recovery of seabirds following the Exxon Valdez oil spill: An overview. "Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters ASTM Special Technical Publication # 1219". (P. G. Wells, J. N. Butler and J. S. Hughes ed.) American Society for Testing and Materials, Philadelphia, PA. pp. 854-893.