Most ecological systems have the ability to recover naturally from perturbations such as oil spills (e.g., Neff 1987, Baker et al. 1990). However, in certain circumstances active restoration of habitats following a spill may enhance the rate of natural recovery. The appropriateness of habitat restoration following an oil spill depends on the predicted rate of ecological recovery without restoration, the effectiveness of restoration techniques for enhancing the natural recovery rate, and the cost of implementing restoration activities. Habitat restoration following an oil spill may not be appropriate when 1) natural recovery will occur within a reasonable period of time, 2) restoration techniques are not technically feasible, 3) restoration is not physically possible because of environmental, political, or safety considerations, 4) restoration activities may further injure biological communities and the habitat, or 5) restoration costs would be grossly disproportionate relative to the value of the affected resources. Approaches to evaluating habitat restoration options for application to oil spill sites are in an early stage of development (e.g., Cairns and Buikema 1984, Cicchetti 1989, pers. comm.). In addition, information on rates of natural recovery after oil spills and the feasibility of using restoration to enhance natural recovery is inadequate for many habitats. This report represents an initial effort to assess technical feasibility and practicability of habitat restoration following oil spills and to develop an approach for evaluating tradeoffs between natural recovery and restoration. Objectives The objectives of this study are to determine circumstances under which 1) active restoration of natural resources following oil spills should be pursued or 2) partial active restoration or no active restoration would be preferred alternatives. These objectives were met by developing a conceptual model for evaluating restoration alternatives and applying the model to five hypothetical spill scenarios. The following tasks were completed as part of this study: · Develop a conceptual model for evaluation of habitat restoration vs. natural recovery · Review selected case studies of oil spills in marine and river and stream habitats · Evaluate measures of natural recovery, factors affecting natural recovery, and recovery rates in selected aquatic ecosystems after oil spills · Evaluate technical feasibility, practicability, and ecological impacts of restoration techniques · Evaluate natural resource damages relative to the costs of restoration under low-intensity restoration, high-in-tensity restoration, and natural recovery scenarios. Regulatory Background and Definitions The legal basis for recovery of natural resource damages caused by a release of a hazardous substance or an oil spill is primarily the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), 42 U.S.C. 9601 et. seq., as amended by the Superfund Amendments and Reauthorization Act, and the Clean Water Act (CWA), 33 U.S.C. 1251 et. seq. Under regulations promulgated pursuant to Section 9651(c) of CERCLA, parties responsible for an oil spill may be held liable for damages to natural resources in addition to cleanup costs, in claims brought under Section 311 of CWA, 33 U.S.C. Section 1321. Guidelines for performing a natural resource damage assessment (NRDA) under CERCLA were developed by the U.S. Department of the Interior and were originally published as a Final Rule in the Federal Register on August 1, 1986 (51 FR 27674). In response to state, industry, and environmental group petitions, the U. S. Court of Appeals for the District of Columbia Circuit reviewed the 1986 'RDA regulations in Ohio v. U.S. Department of Interior, 880 F. 2d 432 (D.C. Cir. 1989) and Colorado v. U.S. Department of Interior, 880 F. 2d 481 (D.C. Cir. 1989). Based on its review, the court remanded the regulations to the U. S. Department of the Interior for revision. Revisions to the Final Rule were published as a Notice of Proposed Rulemaking on April 29, 1991 (56 FR 19752). The proposed changes to the rule have undergone public review, and it is anticipated that the revised Finial Rule will be issued in early 1992. The Type A procedures (a computer model requiring minimal field investigation) for coastal and marine environments are currently being modified and are scheduled for proposed rulemaking in December 1991 (56 FR 19753). A proposed Type A rule for the Great Lakes environment is scheduled for release in the winter of 1991. While the NRDA regulations were being revised under CERCLA, the U.S. Congress passed the Oil Pollution Act of 1990 (Public Law 101-380). This new law authorizes the National Oceanic and Atmospheric Administration (NOAA) to write regulations that address liability for natural resource damages incurred as a result of oil spills. Once the NOAA regulations are completed, the CERCLA NRDA regulations will apply only to releases of hazardous substances. Definitions of key terms used in the NRDA regulations are provided below. In the NRDA regulations, response is distinguished from restoration as follows: · Response means removal, remedy, and remedial actions [§ 11.14(jj)]. Removal means the cleanup of released hazardous substances from the environment, actions necessary to monitor, assess, and evaluate the release, the disposal of removed material, or actions necessary to prevent, minimize, or mitigate damage to the public health or welfare or to the environment that may otherwise result from the release. Remedy or remedial action means those actions consistent with a permanent remedy taken instead of or in addition to removal actions in the event of a release, to prevent or minimize the release so it does not migrate to cause substantial danger to present or future public health or welfare or the environment. Such actions may include storage, confinement, perimeter protection, neutralization, cleanup, recycling, destruction, incineration, or monitoring (CERCLA Part 101). · Restoration or rehabilitation means actions undertaken to return an injured resource to its baseline condition (the condition that would have existed had the discharge not occurred, or the 'without-spill condition') as measured in terms of the injured resource's physical, chemical, or biological properties or the services it previously provided, when such actions are in addition to response actions, and when such actions exceed the level of response action determined to be appropriate for the site pursuant to the National Contingency Plan (NCP) [§ 11.14(ll)]. The division or break point between response actions and restoration activities is far from distinct. Certain oil spill cleanup techniques may be implemented as either response actions or restoration actions, depending on the timing of the activity relative to initial response and the use of the technique in combination with species enhancement efforts as part of a specific restoration plan. For example, seawater flushing of an intertidal seagrass habitat might be interpreted as a restoration activity if it is conducted 1) after the on-scene coordinator [the Coast Guard or U.S. Environmental Protection Agency (EPA)] has decided, pursuant to the NCP, that the initial response effort is complete and 2) before replanting of seagrass, to enhance the survival of introduced plants. On the other hand, some activities are clearly restoration efforts according to the definitions of response and restoration above. For example, direct enhancement of species populations (e.g., replanting of seagrass or marsh vegetation or restocking of a clam species) would definitely fall under the category of restoration. For the purposes of this report, the term restoration includes oil spill cleanup techniques used before or in support of activities that are intended to directly enhance animal or plant populations. The terms low-intensity restoration and high-intesity restoration are used to denote the relative level of effort expended on active restoration. Natural recovery is the return of habitats and biological communities to a baseline state as a result of unassisted natural physical,