The Oil Spill Recovery Institute: Past, Present, and Future Directions (2003)

One of the more profound outcomes of the 1989 Exxon Valdez oil spill was the recognition of our limited ability to realistically predict the effects of an oil spill on marine resources. The ongoing debate over long-term damages further highlights just how inadequate previous knowledge was in attempting to discern cause and effect in natural environments. This lack of knowledge was, on one level, an incomplete understanding of what resources were present. But even more fundamental was a lack of understanding of the structure and functioning of complex ecosystems. Our knowledge of the interconnectedness of systems and the basic controls on biological patterns in time and space was wholly inadequate to answer the questions being posed by regulators, litigators, responders, stakeholders, and the public in general. Without this fundamental knowledge, predictions, prevention, response, remediation, and resource damage assessment can be in jeopardy of being ineffective or inaccurate and thus potentially wasting financial resources and putting natural resources at risk.

One component of the OSRI mission is the portfolio of projects funded in the category called “predictive ecology” (Table 5-1). To facilitate the committee’s review of the OSRI portfolio, it was convenient to further subdivide predictive ecology into its modeling and nonmodeling efforts. This is in part due to the large effort being expended to develop the Nowcast/Forecast (NC/FC) model, which is a component of both the predictive ecology and applied technology programs. While many of the non-modeling activities may support the modeling efforts, in many instances they are more or less stand-alone projects, and the committee chose to

evaluate the modeling and nonmodeling activities separately (see Chapter 7).

Within the predictive ecology program (nonmodeling portion), OSRI has funded a diverse set of projects, including workshops, fellowships, a study of the toxic effects of oil on waterfowl, monitoring of rockfish, monitoring of river otters exposed to oil, herring and pollock monitoring, study of the distribution of intertidal invertebrates on the Copper River Delta, sensitivity mapping of coastal resources, and zooplankton/nekton monitoring (Table 5-1). Excluding the modeling components that OSRI categorizes within predictive ecology, funding of these nonmodeling projects totals a little more than $1 million in the FY98-01 period.

To gain a sense of the direction of and priorities within the Predictive Ecology program (nonmodeling components), the committee looked at the program’s largest projects: herring, pollock, zooplankton, and nekton monitoring in Prince William Sound; the Copper River ecological study; sensitivity mapping of coastal resources (in partnership with NOAA); and a continuation of the herring and pollock acoustic monitoring activity.

The Prince William Sound ecological resources monitoring projects focus on the three dominant pelagic animal biomass (most abundant) animals in the sound. These are herring, pollock, and copepods of the genus Neocalanus. Herring monitoring began in 1993 (pre-OSRI), pollock monitoring in 1995, and zooplankton monitoring in 2000. Acoustic and optical sensing techniques are used to monitor seasonal distributions and abundances.

Together the monitoring projects will provide information about the target animal populations and, to a lesser extent, their roles in the Prince William Sound ecosystem. Some aspects of the research have been published (Thomas and Thorne, 2001). However, the optimization strategy used in designing these programs has raised some concerns. Prince William Sound is a large embayment, and the winter and spring distributions of herring and pollock may not be fully identified by the current field design. In the past, fisheries biologists have suggested that juvenile herring from the sound may over-winter in the bays and coastal water of the Kenai Peninsula. Early studies of herring (1930s), when the main fishery was for adult herring for pickling, show a very different distribution than

most fisheries studies show today. Similarly, pollock stocks have shown dramatic shifts in biomass and distributions in the Gulf of Alaska, Prince William Sound, and the Bering Sea in recent years as more valuable fish stocks have been over-harvested. Climate change also is likely to have played a role, primarily through changing water temperature and related changes in primary production.

The dynamics of zooplankton populations are extremely complex. The acoustic survey methods in this project provide information about their abundance during the spring bloom period. Research has demonstrated significant interannual variability in zooplankton biomass and species composition in Prince William Sound (Cooney et al., 2001; Eslinger et al., 2001) and geographic variation in the bloom locations likely contributes to these differences. Given the current sampling design, it may be difficult to use the results of zooplankton studies in the spring to model spatial and temporal trends of large zooplankton species in the biological modules of the NC/FC model system.

The monitoring projects will provide valuable information about animal distributions during critical high aggregation times. This work should continue to evaluate the seasonal distributions and abundances of high biomass species. The stated objective is that the results of this monitoring will be integrated into the NC/FC model. However, at this point biological modeling components are not being developed.

Challenges remain in confirming the identity of targets in acoustic signals and calibrating the instruments accurately to biomass or other important indicators of stocks, such as size and composition (i.e., age distribution and sex). Denser geographic and temporal coverage is needed to fully understand the dynamics of these biomass dominants in Prince William Sound and adjacent areas. If the results are to be effectively integrated into biophysical models, supporting chemical and physical measurements, such as nutrients, particulates, chlorophyll, and oxygen, will have to be collected as well. Closer alignment of these studies with regional fisheries assessments and studies of other populations are important. The objectives of these projects are laudable, but full implementation may take years and exceed the financial wherewithal of OSRI. Leveraging, partnering, and collaboration with larger ecosystem programs in the area will be crucial to developing a truly predictive model of these important and complex ecosystems.

A major OSRI predictive ecology project is a study of the lower Copper River Delta ecosystem. The research focuses on the distribution of bivalves, amphipods, and insect larvae. The study area and target species

appear to have been chosen based on their potentially high sensitivity to petroleum hydrocarbons and their importance to the diet of shorebirds frequenting the area. The Copper River Delta is an important seasonal habitat for migratory birds using the Alaskan coastal flyway. This study provides basic ecological information about a unique habitat. There have been few other studies of the ecology of river estuary systems in subarctic environments, such as the Copper River estuary. Such ecosystems, including the Upper Cook Inlet, are oligotrophic (high biomass and low diversity), due mainly to climatic conditions and high sediment loads derived from glacial till, and usually are not particularly sensitive to the organic enrichment stress produced by an oil spill. However, they are quite productive and, particularly in the case of the Copper River Delta, are known to be important foraging areas for migratory birds.

This project is well designed and will provide valuable information about this type of delta habitat. However, as currently designed, the project does not address some questions whose answers would be needed to respond to or assess the impacts of a future oil spill in the larger area. For example, there is no chemistry component for evaluating existing polynuclear aromatic hydrocarbon (PAH) sediment concentrations or PAH source characterization in the existing program.

Although the Copper River ecosystem is clearly of ecological and economic importance, especially to the community of Cordova, the ecosystem is not representative of other Arctic and subarctic biomes and thus, it may prove difficult to extrapolate the findings to other settings. In the future, representativeness of Arctic and subarctic environments should be an important consideration when choosing sites for long-term monitoring and study.

OSRI has provided significant support ($215,000) to a project conducting environmental sensitivity index mapping for Alaskan coastlines. The project supports standard mapping of resources in Prince William Sound and the Aleutian Islands, focusing on coastal regions so responders would be better able to protect sensitive areas in the event of a spill. The Hazardous Materials Response Division of NOAA’s Office of Response and Restoration (HAZMAT), which coordinates advice on science and natural resource issues for the US Coast Guard, and which is charged with directing spill response under OPA 90. The mapping is used by HAZMAT in identifying resources at risk from a spill. The focus of the mapping is on shoreline type and degree of exposure, which determines its sensitivity to oiling. Shoreline protection and cleanup priorities are set using this, as well as other information. It also provides basic background information on

shore types useful in natural resource damage assessments. HAZMAT has been developing such data for the entire U.S. shoreline, and thus the OSRI funds are essentially a supplement to that effort. This project is within the OSRI mission.

The LIDAR Proof of Concept project (also called Remote Sensing Cooperative Agreement) is a one-year project that builds on the herring and pollock acoustic monitoring program initiated in FY00. LIDAR in conjunction with established acoustic methods represents a potential for precise estimates of these key fisheries. Funding within this project will enable a proof-of-concept effort to establish the viability of this technique. This project was allocated $100,000 in FY00.

In general, the OSRI predictive ecology projects (nonmodeling) are responsive to the OSRI mission. However, some projects are less clearly connected to the OSRI mission. One possible concern is the OSRI emphasis on Prince William Sound, when the legislative mandate is geographically much broader. The committee understands and supports the need to focus on the resources most at risk from oil spills, given the limited financial resources. This would be mainly Prince William Sound and possibly Cook Inlet (particularly if oil exploration and development increases again).

The legislative mandate implies that studies of long-term ecological effects of oil spills are appropriate projects for OSRI, although to date this has not been an area of emphasis. The legislation tasks OSRI with assisting the Exxon Valdez oil spill trustees in assessing environmental damages from the Exxon Valdez oil spill, focused geographically on the northern Gulf of Alaska region, primarily Prince William Sound. There remains great potential for research on long-term ecological effects.

OSRI has given little attention to the second half of the legislative mandate “... to understand the long range effects of Arctic and subarctic oil spills on…the economy, and the lifestyle and well-being of the people who are dependent on them….” A limited amount of OSRI research is directed at the human occupants of the area and how they interact with resources and how oil spills influence these interactions. The legislative mandate is extremely broad and OSRI has limited resources, so this decision to focus is understandable.

There is still great need for improved understanding and targeted monitoring of sensitive areas at risk from oil spills or chronic releases of petroleum (NRC, 2002). In particular, there is a continuing need for a better understanding of the physical and biological effects of oil spills in Arctic and subarctic environments. Such information is essential for developing optimal spill prevention, response, and remediation strategies that limit collateral harm to the environment. In general, knowledge of biological resources and how to protect them is still limited. There is a great need for research to improve methods and strategies for bioremediation of oil-contaminated Arctic and subarctic marine and wetland ecosystems. There also is a need to better understand the physical, chemical, and biological fates (weathering) of petroleum in cold environments. Knowledge of biological resources, and how to protect them, is frequently lacking; marine mammals and birds offshore in the Beaufort and Chukchi seas are at particular risk. The following suggestions are illustrations of possible directions and are not intended as a comprehensive list or to replace strategic planning by OSRI.

One of the stated objectives of the ecological studies supported by OSRI is to develop “baseline” data about the structure and function of marine ecosystems in the northern Gulf of Alaska that are at risk from future oil spills. The baseline data are intended to provide the basis for documenting and quantifying ecological injury, should a spill occur. Baseline inventories of resources or benchmarks have real limitations in preparing for and assessing damage after an oil spill. A National Research Council review of a Bureau of Land Management (BLM) program determined that the baseline approach is not a useful way to characterize the effects of human perturbations on natural systems (NRC, 1978). A fundamental restructuring of BLM’s environmental studies program was recommended. The NRC conclusions are applicable to OSRI and how it selects its portfolio of projects. Studies of natural resources need to provide a dynamic vision of the systems being studied, because they change on a wide variety of spatial and temporal scales. Inventories or descriptive studies should be performed only within the context of understanding the structure and functioning of ecosystems and not as a static picture of resources at one point in time. Efforts to develop techniques for monitoring populations and implementing those techniques over multiple years will provide a more fruitful approach to understanding resources.

There is considerable debate, and little resolution, concerning the extent to which the methods used in oil spill response themselves cause injury to the shoreline and coastal marine resources, beyond the effects of the oil itself. Research is needed on the effects of different oil spill response options, such as use of chemical dispersants, beach cleaners, high-pressure or hot water washing, bioremediation, etc., on water column and intertidal biological communities. There is a need to develop methods for removing oil from the environment efficiently while minimizing adverse effects on marine organisms and ecosystems. Because the window of opportunity (length of time after a spill when dispersants are effective) for dispersant use is so narrow, it is essential that they be pre-approved for use in areas where a spill might occur. However, that would require careful studies to define the parameters that control the window of opportunity for dispersant use for a particular crude oil under a set of “typical” environmental conditions. A systematic and robust net environmental benefit analysis for dispersant use versus non-use would provide decision makers with the information needed to make effective use of dispersants.

Biodegradation played an important role in the removal of oil from the water column and shore of the sound and Gulf of Alaska following the Exxon Valdez spill. The estimate (Wolfe et al., 1994) that 50 percent of the spilled oil biodegraded is very approximate, but probably in the ballpark. The bioremediation (enhanced biodegradation) studies had mixed results, but did prove the concept that enhanced biodegradation can contribute to removal of oil from the shore. At the time of the spill, most spill response experts felt that bioremediation would not work in a subarctic environment like the Gulf of Alaska, because ambient temperatures are too low. However Bragg et al. (1994), Atlas (1995), Lindstrom and Braddock (2002), and others have shown that bioremediation works in the Alaskan marine environment. More research is needed to better understand the chemical and environmental factors controlling natural and enhanced biodegradation of petroleum.

When assessing active treatment strategies, such as dispersant application, it is critical to have a clear understanding of the effects these treatments will likely have on biodegradation processes, because biodegradation is an important process leading to removal of contaminants from the environment. The results of existing studies examining the effects of commercial dispersants on biodegradation are specific to the product used,

with some dispersants increasing and some diminishing biodegradation. Dispersants can potentially affect the relative degradation of different classes of hydrocarbons. Foght and Westlake (1982), for example, found that Corexit 9527 affected alkane degraders differently than aromatic degraders, depending on the nutrient regime. Overall these types of studies make clear that understanding naturally occurring biodegradation of petroleum hydrocarbons in the environment is of interest and the potential effects of treatment strategies are very important to appropriately decide on a response to an oil spill. Predictive models also need much more accurate information on the toxicity of hydrocarbons to a wide range of species inhabiting coastal areas.

There is also a need to develop methods for restoring oil-impacted marine ecosystems and for facilitating natural recovery of impacted ecosystems, in accordance with the current NRDA regulations. Most marine resource populations and their supporting ecosystems begin recovery as soon as concentrations of oil decline to levels that are not toxic or otherwise harmful to populations. If large, long-lived species, such as marine birds and mammals, are affected, recovery of their populations may be slow. Human efforts to facilitate recovery probably should focus on habitat cleanup and protection. Similarly, intertidal communities of plants and animals often are severely harmed by oil washing ashore. However, the injured populations quickly recolonize affected habitats when the oil concentration and toxicity decline. The most effective remediation/restoration strategy would be to remove oil from the spill zone as quickly as possible.

OSRI should consider developing an oil spill restoration program, perhaps in collaboration with industry, the Environmental Protection Agency, or NOAA, to develop and evaluate methods for removing oil from the sea surface and the shoreline that cause minimal additional harm to biological resources.