Return to the river: an ecological vision for the recovery of the Columbia River salmon.

I. INTRODUCTION

In recent amendments to the Columbia River Basin Fish and Wildlife Program, the Northwest Power Planning Council called on the Bonneville Power Administration to fund an Independent Scientific Group (ISG; now called the Independent Scientific Advisory Board (ISAB)), a panel of 11 senior scientists charged to conduct a biennial review of the science underlying salmon and steelhead recovery efforts in the basin. Salmon and steelhead populations are depressed and declining(1) due, in part, to heavy development of the basin for hydroelectric power and the general failure of mitigation programs for the hydropower system on the Columbia and Snake rivers(2) (Figure 1). The Council also asked us to develop a conceptual foundation for the fish and wildlife program. This conceptual foundation would provide an overall framework of scientific principles and assumptions on which the basin-wide program of mitigation for hydropower development could be based and against which present and proposed mitigation efforts would be evaluated.

[Figure 1 ILLUSTRATION OMITTED]

In September 1996, ISG delivered its review and conceptual framework in a report, Return to the River, to the Council in a prepublication form suitable for public comment.(3) While ISG did not make specific management recommendations in Return to the River, we did include example actions consistent with scientific findings and our proposed alternative conceptual foundation. Prior to release of the report to the Council, it was peer reviewed by eight anonymous scientists. During the fall of 1996, the Council held a series of public meetings across the basin at which ISG members presented the main conclusions and entertained questions from Council members, legislators, policy makers, and interested public. Additionally, public and technical comments were solicited by the Council through June 1997. The report is now being revised and prepared for final publication and should be available in late 1998.

The purpose of this paper is to provide a summary of Return to the River, particularly its proposed conceptual foundation, for those involved with the environmental, legal, and policy aspects of hydropower and salmon recovery. The report focuses especially on the region's efforts to restore salmon and steelhead (hereafter referred to collectively as salmon) in the Basin. It should provide a flavor of the in-depth scientific reviews conducted on topics such as the diversity, structure, and status of salmon populations; freshwater habitats in tributaries and the mainstem; mechanisms of salmon migration; history of restoration efforts; studies of fish passage at dams; losses of migrants to predation; gas bubble disease; the barge/truck transportation alternative to in-river migration; salmon in the estuary and ocean; artificial production (hatcheries); harvest management; flow augmentation and drawdown as river management alternatives; and the role of monitoring and evaluation.

II. THE CONCEPTUAL FOUNDATION

A conceptual foundation is a set of scientific principles and assumptions that can give direction to management activities, including biological restoration programs. It is the perspective or filter through which information is viewed and interpreted. Recovery measures and research findings will take on different meanings when viewed through different filters or perspectives. It is a blueprint for the assembly of various data sets and other information.

Conceptual foundations often are not explicitly made, and this has been true for Columbia River fisheries management objectives and hydropower mitigation. Past programs may not have been successful, in part, due to lack of an explicit foundation and related objectives. Without an explicit conceptual foundation, it is difficult to scientifically analyze the basis for the region's efforts or to evaluate the effectiveness of those efforts.

However, the current suite of programs does have an underlying theme, even if not explicitly stated. Past mitigation efforts have assumed that natural ecological processes that historically led to, and supported, healthy salmon populations could to a large degree be circumvented, simplified, and controlled by humans and still have the salmon populations remain healthy and productive. For example, placing more young fish in the river (from hatcheries) would automatically result in more adult spawners. Salmon production could be increased by actions taken in the river without accounting for conditions in the estuary or ocean, and management actions would not inadvertently damage the environmental attributes of the river system that support salmon.

This very mechanistic concept has driven management toward actions that are best characterized as technological substitutes for ecological processes. Actions often respond to individual problems in isolation, rather than in the context of full salmon life histories, behaviors, and habitats. Although done in good faith, and often with technological tools and research results appropriate to the narrow situation, they are not sufficient, based on the continued decline of the basin's salmon populations. From this, we concluded that a new conceptual foundation is needed.

III. AN ALTERNATIVE CONCEPTUAL FOUNDATION

The conceptual foundation we propose departs from that embodied in the current Columbia River Basin Fish and Wildlife Program. Our alternative conceptual foundation emphasizes the importance of the ecological processes which sustain salmon populations. These processes need to be taken in the context of both ecological and social processes that link organisms, including humans, with their environment. In essence, it focuses on restoration of essential features of the environment that the fish need to carry out their entire life history, rather than on the use of technological substitutes for these features. These environments occur throughout the freshwater, estuarine, and ocean habitats where salmon complete their life histories.

Our alternative conceptual foundation includes a few key concepts. Sustained salmon productivity requires a network of complex and interconnected habitats which are created, altered, and maintained by natural physical and biological processes. Many of these physical and biological processes have been altered or eliminated by human actions. Restoration of habitats will require restoration of these processes to reestablish suitable habitats.

Diverse and perhaps most importantly interconnected habitats are crucial for spawning, rearing, migration, maintenance of food webs, and predator avoidance by migratory species (including both anadromous and fully-freshwater forms). Evolutionarily, salmon have adapted to the high environmental diversity in the watersheds of western North America and to the volatility of many environmental conditions in both freshwater and marine environments by developing diverse life history strategies, genetic diversity, and a complex hierarchical population structure termed metapopulation organization. Salmon can accommodate many human cultural features if a critical mix of natural features remain.

We encapsulate these concepts in the term normative river ecosystem for salmon. Salmon-bearing river systems can be described in terms of norms or standards for specific biological and physical features that are essential to maintain diverse and productive salmon populations (even if the system is culturally modified, as with hydropower dams). We emphasize that normative conditions are not the same as historical conditions. Although our imperfect knowledge of the historical conditions that yielded high salmon productivity has aided in establishing the norms, historical conditions are clearly unattainable. Even if we were to restore all of the physical river habitat to predam conditions, we would still have human settlements, surface water discharges, agriculture, forestry, and a host of introduced species. Our concept of normative conditions embraces both the natural and the cultural elements that make up the Columbia River ecosystem. It addresses the features essential to the maintenance of productive fish and wildlife populations and suggests directions for regional restoration efforts. It draws on a growing body of science that can define the physical and biological conditions needed by salmon in their ecosystem and points the way to altering our heavily developed river systems to increase normative conditions, salmon diversity, and production (Figure 2).

[Figure 2 ILLUSTRATION OMITTED]

IV. THE NORMATIVE SALMON ECOSYSTEM

The availability of connected, high-quality habitats is key among the conditions salmon need in the Columbia River Basin throughout their life cycle, migrating through various habitats from spawning streams to the Pacific Ocean. These habitats vary from fast-moving, gravel-bottom streams in the headwaters through turbulent mainstem channels, to tidal estuaries and the coastal and pelagic gyres and currents of the northern Pacific Ocean. The habitats provide, for example, gravels and intragravel water flow for spawning, shallows for the growth of food organisms and places for juvenile fish to feed, turbulent channels for accelerated movement of smolts downstream at the proper time, and brackish estuaries for the transition to salt-water life. Science recently has made significant strides in defining these geomorphological needs in quantitative ways.(4) Return to the River also documents many of these needs.

Mere existence of the various habitats is not enough, however. They must be connected so that these migratory fish can move from one habitat to another in a timely fashion. One factor contributing to salmon declines in the basin today is the fragmentation of habitat into pockets separated by areas of unsuitable or even lethal conditions. For example, excellent spawning habitat in certain tributaries (often located in protected wilderness areas) is functionally isolated from suitable mainstem migration corridors by intervening reaches that are too warm for downstream-migrating juveniles or are actually dewatered in the adult spawning migration season by irrigation diversions (Figure 3). The normative condition is to have access between these habitats when fish need to move through them.

[Figure 3 ILLUSTRATION OMITTED]

Another aspect of the normative system is the set of physical processes that naturally structure aquatic habitats to create and maintain the needed topographic features. Flooding, for example, is a feature of rivers that serves many important functions.(5) Many development activities restrict or eliminate flooding as a natural process. Floods physically move and reposition river substrates (sand, gravel, logs, etc.), thereby cleansing spawning gravels and creating backeddies and side channels where juveniles feed. They maintain groundwater flows. They provide soils for riparian vegetation and seasonally submerge that vegetation so that it can serve as substrate for growth of aquatic insects on which downstream migrating juvenile salmon feed. They also physically transport many young salmon in their downstream migrations. Connections among habitats can be the result of episodic physical events (that is, functional at only the crucial times), such as seasonal high runoff. These physical processes have been greatly altered in a highly regulated system such as the Columbia River Basin. Water management with a normative river perspective, however, can create seasonal high water and allow occasional severe flooding in wet years, which will do the work of habitat restoration for us.(6)

Our conceptual foundation, which focuses on life history diversity among fish stocks, emphasizes the linkage between habitat and environmental diversity and life history diversity. Habitat diversity, created and maintained by natural riverine processes, is the template upon which salmon life history diversity is expressed. This diversity encompasses the broad array of habitats colonized by spawners (from mainstems to small streams, lakes, and oceans), the multiple habitats used for feeding and rearing (also in mainstems, tributaries, and lakes), as well as the different sizes of juvenile salmon at migration, and their varied migration times. Historically, there were apparently adults and juveniles moving upstream or downstream throughout the year. Currently, there are fairly distinct seasonal runs of adults headed upriver for specific spawning areas and a preponderance of juveniles migrating downstream during the normal spring freshet season (April - June). This restriction on migration timing is emphasized and maintained by operation of the mainstem dams that provide a narrow window of suitable migration conditions. Increasing normative conditions in the basin would help preserve the remaining stock diversity and allow it to increase locally, which serves an important function besides fully seeding or occupying available space.

Stock diversity provides a buffer for extreme environmental changes. The habitat continuum appears always to have been affected in a very dynamic fashion by natural daily, seasonal, interannual, and long-term (often cyclic) changes in factors such as temperature and water availability, and by natural disasters (volcanoes, landslides, cataclysmic floods, etc.). This has resulted in a complex habitat mosaic that is dynamic in time and space. Salmon have interacted with this dynamic habitat template over time, and diversified into myriad stocks and life history patterns. Such diversity protected the long-term stability and productivity of the Columbia River Basin salmon, because life history diversity spreads the risk of mortality (extinction at the population and species level)in fluctuating environments.

While life history diversity describes structure within a salmonid population, a relatively new concept, the metapopulation, describes structure among groups of populations. Metapopulations are spatially-structured groups of local populations linked by dispersal and interbreeding of individuals.(7) Persistence of a metapopulation is determined by the balance of local population decline or extinction and augmentation or recolonization following dispersal from neighboring populations. The application of the metapopulation concept to conservation is currently being debated by scientists and managers(8) and has only recently been applied to the conservation and management of salmonid fishes(9) Thus, the inclusion of the metapopulation in our conceptual foundation is on the basis of a hypothesis that requires further empirical evaluation.

The historical condition of interconnected populations within a larger metapopulation has been disconnected by river development, but can be restored, at least partially, by establishing salmon refuges for core populations and ensuring connectivity to satellite areas (Figure 4). For example, the Hanford, Washington reach of the mid-Columbia River, where the Hanford nuclear reservation precluded dam-building, contains a flowing riverine 50-mile section of the mainstem Columbia River and has the most thriving inland salmon population of the basin.(10) ISG believes that the Hanford fall chinook stock can serve as a core population for rebuilding chinook stocks in other parts of the basin, if the stock and the Hanford Reach are protected.(11)

[Figure 4 ILLUSTRATION OMITTED]

The historical condition of having an array of populations with diverse characteristics is perhaps most important in relation to fluctuating environmental conditions in the estuary and ocean. Whereas we can actively manage environmental conditions in the freshwater environment, at least within certain limits, we have little control over the estuary and virtually none over the ocean (except for regulating harvest). Salmon species and stocks are known to occupy discrete areas of the ocean.(12) When highly cyclic ocean conditions such as El Nino occur, that affect salmon in various parts of the ocean differently,(13) we must depend on having some salmon available from a wide range of populations so that some can be more successful than others in varying environmental conditions. That way, the entire group of populations is not extirpated by fluctuations in the environment.

The present hydrosystem, unfortunately, is operated in ways that reduce stock diversity. For example, juvenile fish bypasses are operated only during the main part of a stock's downstream migration, thus selecting against those that migrate especially early or late in the season; water flows are managed to facilitate downstream migration of only the major portion of a run; and hatcheries have purposefully or inadvertently selected brood stock from parts of a spawning run to match production schedules rather than to preserve run diversity. If we recognize that stock diversity is a normative feature ensuring the continued productivity of salmon in general, then we view it as a vital feature to protect and enhance rather than as a curiosity.

The biological and physical processes involved in downstream migration of juvenile salmon were researched as part of our report to identify necessary environmental conditions and whether proposed management approaches (e.g., reservoir drawdown and augmented flows) made sense biologically. The traditional conceptual model for juvenile migration has been one of flushing smolts by river flows to the ocean. In its place, we propose a model, which we call "spiraling," in which juveniles alternately (often in a daily cycle) move for periods and rest and feed for other periods (i.e., spiral their way downstream through periods of activity and rest). Fish need resting and feeding habitats in mainstem migration corridors, as well as open channels for movement. We evaluated available feeding habitat and food materials in the normative system. Finally, when features of a riverine channel and unsteady flow (e.g., vortices, turbulent bursts, and waves) are replaced by less turbulent and stratified flows of reservoirs, migration is impeded.

Because technological fixes at dams have been an important part of salmon management in the Columbia River Basin, we reviewed research on juvenile fish passage at mainstem dams. This was done, in part, to see if there were relationships between fish passage results and the behavior of juvenile salmon. Not surprisingly, passage schemes that make use of the normal behavior of both adult and juvenile salmon are the most successful. The normal surface orientation of smolts and their natural tendency to follow flows has not been used effectively in most attempts to engineer fish passage routes through dams (usually involving entrainment in deep turbine flows, screening from these flows, and guidance of fish through gatewells and piping to the downstream side of the dam). New concepts of surface fish bypasses, based on successful passage at Wells Dam, are more "normative" (more in concert with the normal behavior of smolts).

The normative river (ecosystem) concept has both scientific validity and potential usefulness in guiding water resource management in the Columbia River Basin toward strategies that will foster salmon recovery. It does not imply that we must return to a pristine, pre-development state. It requires that we learn the critical features of ecosystem and salmon performance and then strive to manage our cultural features (hydropower, irrigation withdrawals, navigation, flood control, etc.) in ways that more closely approximate those normative features. What we actually do will depend on where the actions are to be taken and what amount of alteration has already taken place (Figure 5). Headwater spawning areas can be left pristine. Storage reservoirs already in place can be managed by reregulating to achieve normative features by providing more normal seasonal cycles of flow and temperature.(14) Mainstem reservoirs can be managed to provide habitat, including, in some cases, drawdown or removal of some dams. Areas presently exhibiting normative conditions that are producing salmon (such as the Hartford Reach) can be made into refuges. Mainstem power projects can be designed and operated to more closely mimic key features of the normative river, such as reregulation of flows to stabilize daily fluctuations in flow that allow food web development in shallow water habitats. These management steps will clearly require more than just scientific input. They will also require an open and public regional discussion. Our conceptual foundation, based on the normative river concept, could guide and focus the region's efforts toward a common goal.

[Figure 5 ILLUSTRATION OMITTED]

V. LEGAL AND POLICY CHALLENGES

Restoration of normative conditions poses significant legal, social, and economic challenges. Many of its concepts run counter to the philosophy that has guided development of the Columbia River Basin for much of this century. For this reason, restoration of normative conditions will require an examination of the values that undergird Columbia River management. Coupled with the significant economic and social cost of restoring many normative features, the debate over salmon restoration will continue to be heated. However, the conceptual foundation outlined above at least provides a scientific basis for a debate that has often used scientific uncertainty to mask underlying questions of social values and cost.

Within this debate, at least two major legal and policy challenges can be identified that are relevant to the Return to the River. These challenges include regional recovery goals and the role science plays in shaping regional policies.

A. Goals for Regional Recovery

At the present time, there are two major laws that are the basis for regional recovery efforts. The Northwest Power Act of 1981(15) created the Northwest Power Planning Council and charged it with development of a program to "protect, mitigate, and enhance" fish and wildlife resources of the Columbia Basin as affected by development of the river's hydroelectric potential.(16) It is this program that was reviewed by the Independent Scientific Group as discussed above. The second relevant law is the federal Endangered Species Act (ESA or Act)(17) which, in the case of salmon, is administered by the National Marine Fisheries Service (NMFS). All salmon populations in the Snake River (the largest tributary to the Columbia River) are currently listed as endangered under the Act and other groups of salmon are under consideration for listing.

These two laws present the region with alternative, although not incompatible, approaches. The Northwest Power Act suggests a broad perspective, calling for the river to be treated as a system and addressing broad-scale problems resulting from hydroelectric development.(18) The ESA, in contrast, focuses more narrowly on restoration of specific populations listed under the Act, although it includes all factors affecting these populations, not just hydropower development.(19) Because they are based on separate bodies of law and are administered by different organizations with different jurisdictional perspectives (federal, in the case of NMFS, and regional, in the case of the Council), the restoration programs of the Council and NMFS are not well coordinated. The emergency nature of action under the ESA has resulted in abandonment of broader regional restoration.

However, the perspective of the two laws and the goals of the two administering organizations are not incompatible and, indeed, should be complementary. The normative ecosystem conceptual foundation outlined above is as equally compatible with the broad goal of restoring the Columbia River fish and wildlife as it is with the narrower goal of restoring specific populations. While it is likely that restoration of normative conditions to achieve broader, system-wide goals will also address the needs of populations listed under the ESA, it is not clear that the reverse is true. Return to the River addresses the broad need to restore the Columbia River ecosystem which includes retaining and rebuilding populations currently listed under the ESA. The challenge for the region is to forge these two perspectives into an effective, science-based approach to achievement of regional goals.

B. The Role of Science in Shaping Regional Policies

Whether trying to preserve or restore natural river conditions, or implementing normative conditions without fully removing dams or draining reservoirs, federal agencies and Congress must have the political will to question the status quo. The 1994-1998 Biological Opinion for the Federal Columbia River Power System Operations,(20) recently upheld in American Rivers v. NMFS,(21) does not require the Corps of Engineers or the Bureau of Reclamation to operate the system to attain more normative river conditions. Instead, it calls on the river operators to make relatively minor, albeit expensive, modifications that leave the basic status quo in place. While the normative river concept is compatible with other uses of the river and its watershed, this is not to say that normative conditions can be obtained without significant alteration of these activities. Assuming that the region is serious in its desire for significant rebuilding of the native fish and wildlife system, significant movement toward normative conditions is not a status quo option. The status quo, in our opinion, is unlikely to result in significant improvement in the long-term status of fish and wildlife in the Columbia River. It is more likely to result in continued salmon declines and extinctions.

However, making significant changes will, in many cases, require painful changes including congressional alteration of authorized project operations. If adoption and implementation of a normative river approach is not compatible with the authorized purpose of a hydroelectric project, for example, Congress must alter the project authorizations and provide adequate funding to implement the changes. This would require the region to balance the protection and enhancement of salmon relative to other uses and make some difficult choices. For example, drawdown of reservoir elevations would limit, although not eliminate, the region's ability to use the Columbia River as a navigation corridor and to supply some irrigation needs. The Port of Lewiston and farmers who would be forced to ship their products by alternate routes, not surprisingly, oppose such changes.

Traditionally, balancing competing interests has resulted in maintenance of the status quo, posing a difficult challenge when normative conditions are the goal. Return to the River, and other recent reviews of the salmon problem,(22) provide a scientific foundation for salmon recovery. Consequently, the biggest challenge facing the region is not the biological uncertainties associated with salmon recovery efforts, but is whether the region is willing to face the fact that the we cannot have our cake and eat it too. Restoration of fish and wildlife in the Columbia River Basin will require difficult decisions, and will test whether the region's policy makers and elected officials can find the political will and strength necessary to endorse and implement a scientifically sound salmon recovery program.

VI. CONCLUSIONS

We strongly recommend the normative salmon ecosystem concept as the foundation for salmon recovery and hydropower mitigation in the Columbia River Basin. It is soundly based on state-of-the-art science in river basin geomorphology, ecology, and salmon biology. We have not recommended specific management actions, except as examples, but we firmly believe that important actions can be planned with this concept as a guide for allocating research funds and revising operational policies that will allow both salmon productivity and important cultural uses of the river, such as hydropower. The Pacific Northwest community has an exciting challenge to create more river-like conditions for fish in a highly developed watershed such as the Columbia River Basin.

(1) Willa Nehlsen et al., Pacific Salmon at the Crossroads: Stocks at Risk from California, Oregon, Idaho, and Washington, 16 FISHERIES 4, 4 (1991).

(2) NATIONAL RESEARCH COUNCIL, UPSTREAM: SALMON AND SOCIETY IN THE PACIFIC NORTHWEST 60-66, 231-125 (1996).

(3) INDEPENDENT SCIENTIFIC GROUP, RETURN TO THE RIVER: RESTORATION OF SALMONID FISHES IN THE COLUMBIA RIVER ECOSYSTEM (1996) [hereinafter RETURN TO THE RIVER].

(4) J. A. Stanford et al., A General Protocol for Restoration of Regulated Rivers, 12 REGULATED RIVERS 391, 398 (1996).

(5) Wolfgang J. Junk et al., The Flood Pulse Concept in River-Floodplain Systems, in PROCEEDINGS OF THE INTERNATIONAL LARGE RIVER SYMPOSIUM 110, 122 (D. P. Dodge ed., 1989).

(6) Stanford, supra note 4, at 391-413.

(7) Ilkka Hanski, Single-Species Metapopulation Dynamics: Concepts, Models, and Observations, 42 BIOLOGICAL J. LINNEAN SOC'Y 17, 17-19 (1991); I. Hanski & M. Gilpin, Metapopulation Dynamics: Brief History and Conceptual Domain, 42 BIOLOGICAL J. LINNEAN SOC'Y 3, 7-8 (1991).

(8) Susan Harrison, Metapopulations and Conservation, in LARGE-SCALE ECOLOGY AND CONSERVATION BIOLOGY 111 (P. J. Edwards et al. eds., 1994); Charles C. Mann & Mark L. Plummer, Are Wildlife Corridors the Right Path?, 270 SCIENCE 1428 (1995).

(9) BRUCE E. RIEMAN & JOHN D. MCINTYRE, DEMOGRAPHIC AND HABITAT REQUIREMENTS FOR CONSERVATION OF BULL TROUT 15 (1993); Phillip R. Mundy et. al., Selection of Conservation Units for Pacific Salmon: Lessons from the Columbia River, in EVOLUTION AND THE AQUATIC ECOSYSTEM: DEFINING UNIQUE UNITS IN POPULATION CONSERVATION 28, 29 (Jennifer L. Nielsen ed., 1995); Bruce E. Rieman & John D. McIntyre, Occurrence of Bull Trout in Naturally Fragmented Habitat Patches of Varied Size, 124 TRANSACTIONS AM. FISHERIES SOC'Y 285, 285-286 (1995); B. E. Rieman & J. D. McIntyre, Spatial and Temporal Variability in Bull Trout Redd Counts, 16 N. AM. J. FISHERIES MGMT. 132, 132-133 (1996).

(10) C. Huntington et al., A Survey of Healthy Native Stocks of Anadromous Salmonids in the Pacific Northwest and California, 21 FISHERIES 6, 6-14 (1996).

(11) David R. Geist, The Hanford Reach: What Do We Stand To Lose?, 11 ILLAHEE 130, 140 (1995); RETURN TO THE RIVER, supra note 3; Shauna Marie Whidden, The Hanford Reach: Protecting the Columbia's Last Safe Haven for Salmon, 26 ENVTL L. 265, 270 (1996). See generally Guido R. Rahr, III et al., Sanctuaries for Native Salmon: A Conservation Strategy for the 21st Century, 23 FISHERIES 6 (1998).

(12) PACIFIC SALMON LIFE HISTORIES 5, 123, 234, 315, 398, 450 (C. Groot & L. Margolis eds., 1991).

(13) John D. Fulton & R. J. LeBrasseur, Interannual Shifting of the Subarctic Boundary and Some of the Biotic Effects on Juvenile Salmonids, in EL NINO NORTH: NINO EFFECTS IN THE EASTERN SUBARCTIC PACIFIC OCEAN 237, 239 (Warren S. Wooster & David L. Fluharty eds., 1985).

(14) Stanford, supra note 4, at 404.

(15) Pacific Northwest Electric Power Planning and Conservation Act, 16 U.S.C. [subsections] 839-839h (1994 & Supp. II 1996).

(16) 16 U.S.C. [sections] 839b (1994 & Supp. II 1996).

(17) Endangered Species Act of 1973, 16 U.S.C. [subsections] 1531-1544 (1994 & Supp. II 1996).

(18) 16 U.S.C. [sections] 839 (1994).

(19) 16 U.S.C. [sections] 1533(f) (1994).

(20) NATIONAL MARINE FISHERIES SERV., U.S. DEP'T OF COMMERCE, THE 1994-1998 BIOLOGICAL OPINION FOR THE FEDERAL COLUMBIA RIVER POWER SYSTEM OPERATIONS (1994).

(21) 126 F.3d 1118 (9th Cir. 1997).

(22) NATIONAL RESEARCH COUNCIL, UPSTREAM' SALMON AND SOCIETY IN THE PACIFIC NORTHWEST (1996); PACIFIC SALMON AND THEIR ECOSYSTEMS: STATUS AND FUTURE OPTIONS (D. J. Stouder et al. eds., 1997).

THE INDEPENDENT SCIENTIFIC GROUP, Scientific Advisory Group to the Northwest Power Planning Council, Portland, Oregon. Members include: Richard N. Williams (Chair), Daniel L. Bottom, Lyle D. Calvin, Charles C. Coutant, Michael W. Erho, Jr., Christopher A. Frissell, James A. Lichatowich, William J. Liss, Willis E. McConnaha, Phillip R. Mundy, Jack A. Stanford, and Richard R. Whitney.