Salmon and hydropower are the two great natural resources of the Columbia River Basin. At odds for most of this century, congressional action in 1980 injected a needed resolve into the conflict, inciting a sense of optimism. Since that time, optimism has largely given way to the dynamics of a classic struggle between environmental and energy interests. The climax of this particular struggle may well be our review of the parties’ efforts constituting what has been touted as the world’s largest program of biological restoration. 1

The Northwest Resources Information Center, Inc., and other environmental interests (collectively “NRIC”), the Confederated Tribes and Bands of the Yakima Indian Nation (the “Yakima Nation”), and the Aluminum Company of America and other companies purchasing power from the Bonneville Power Administration (the “Direct Service Industries” or “DSIs”) challenge the Pacific Northwest Electric Power and Conservation Planning Council’s final amendments to the Columbia River Basin Fish and Wildlife Program (the “Program” or “Strategy for Salmon”). These cases have been consolidated for decision on the petitions for review. 2

BACKGROUND

The flows of the Columbia River Basin 3 are the linchpin of the Pacific Northwest’s economy. More than one hundred and fifty dams are integrated to form the world’s largest hydropower system, generating over forty percent of the nation’s hydropower. The generation of kilowatts, however, is not the only reason for the Basin’s importance; the Basin is the habitat for what were once the world’s largest salmon runs. 4

Unfortunately, “[l]ike most of the American West, the Columbia River Basin has been developed, not managed as an ecosystem.” John M. Volkman & Willis E. MeConnaha, Through a Glass, Darkly: Columbia River Salmon, the Endangered Species Act, and Adaptive Management, 23 Envtl.L. 1249, 1250 (1993) [hereinafter Through a Glass ]. Adverse impacts have resulted from deforestation, over-fishing, irrigation practices, mining, grazing, urbanization, and hydropower operations and devel *1376 opment. Although precise estimates of the impacts of these activities are currently unknown, it is generally accepted that the Basin’s hydropower system is “a major factor in the decline of some salmon and steelhead runs to a point of near extinction.” 126 Cong.Ree. H10687 (1980) (letter from Comptroller General). The U.S. General Accounting Office (“GAO”) described the impact of the hydropower system on anadromous fish 5 in its September 4,1979, report to Representative John D. Dingell, Chairman of the House Subcommittee on Energy and Power, as follows:

H.R.Rep. No. 96-976, pt. I, 96th Cong., 2d Sess., at 46, 1980 U.S.Code Cong. & Ad. News at pp. 5989, 6044 (1980); see Public Util. Dist. 1 of Chelan County, Wash., 34 FERC ¶ 63,044 (1986) (finding of eleven percent mortality at one powerhouse); 126 Cong.Ree. H2062 (1980) (Rep. Bonker) (“Each dam on the upper [Columbia Rjiver can destroy 10 to 15 percent of the fish that pass over it enroute to the sea.”).

As a result of human activities, anadro-mous fish runs in the Columbia River Basin have dwindled from an estimated ten to sixteen million fish annually, before European settlement in the Northwest, to about two- and-one-half million today. Robert C. Loth-rop, The Misplaced, Role of Cost-Benefit Analysis in Columbia Basin Fishery Mitigation, 16 Envtl.L. 517, 522 (1986) [hereinafter Misplaced Role of Cost-Benefit Analysis ]. Of this annual loss, eighty percent is attributable to hydropower development and operation. Endangered and Threatened Species; Proposed Endangered Status for Snake River Sockeye Salmon, 56 Fed.Reg. 14,055, 14,058 (1991).

There was early recognition by some that highlight the importance of the Basin’s salmon runs. In the 1850’s, Northwest tribes ceded to the United States a total of sixty-four million acres of land in exchange for certain rights and for the exclusive use of land reservations. The tribes were careful, however, to reserve the right to continue, as they always had, to take fish at their usual fishing grounds. 6 Environmental factors af *1377 fecting run sizes were studied as early as 1875. Salmon is King, 16 Envtl.L. at 351. And Congress expressed concern about the Basin’s salmon runs in 1937 when it passed the Bonneville Project Act. 7

In the wake of devastating losses of salmon and steelhead in the mid-1970s, 8 Congress enacted the Fish and Wildlife Coordination Act in 1976. 16 U.S.C. §§ 661 -666c. The Coordination Act provided that fish and wildlife have “equal consideration” in the planning and development of hydropower projects. 9 In 1978, however, Federal fishery agencies considered invoking the protection of the Endangered Species Act (“ESA”), 16 U.S.C. §§ 153H4, for Snake River runs. Despite the mandate of the Coordination Act, it was not until the threat of action under the ESA that the depletion of anadromous fish in the Columbia River Basin was considered to be more than merely an issue that would resolve itself over time. Also in 1978, Congress was preparing to revise Northwest electric policies in light of forecasts predicting serious power shortages during critical water years expected in the 1980’s. H.R.Rep. No. 96-976, pt. II, 96th Cong., 2d Sess., at 31 (1980); see Forelaws on Bd. v. Johnson, 743 F.2d 677 , 679 (9th Cir.1984). Recognizing the tremendous, detrimental impact of dams on the fish runs, Congress acknowledged “that no longer [should] fish and wildlife be given a secondary status.” 126 Cong.Rec. H10681 (1980) (Rep. Dingell). In 1980, Congress enacted the Northwest Power Act (“NPA” or “Act”). 16 U.S.C. §§ 839 -839L

The NPA marked an important shift in federal policy. Continually declining fish runs had revealed the failures of previous legislative efforts requiring that “equal consideration” be given to fish and wildlife affected by resource exploitation. The NPA ensured the “equitable treatment” of fish and wildlife; it marked the shift of the burden of uncertainty — of proving specific harm to salmon from particular activities — from the salmon to the hydropower system, or so was its intent. 10 In doing so, it created a new obligation on the region and various Federal agencies to protect, mitigate, and enhance *1378 fish and wildlife. See 126 Cong.Rec. 10682 (Rep. Dingell).

The Act created a “pluralistic intergovernmental and public review process.” Michael C. Bluxnm, Fulfilling the Parity Promise: A Perspective on Scientific Proof, Economic Cost, and Indian Treaty Rights in the Approval of the Columbia Basin Fish and ■Wildlife Program, 13 Envtl.L. 103,112 (1982) [hereinafter Parity //]. At the hub of this process, Congress established the Pacific Northwest Electric Power and Conservation Planning Council (“Council”), directing it to create “a program to protect, mitigate, and enhance” the Columbia River Basin’s fish and wildlife “to the extent affected by the development and operation” of the Basin’s hydropower system. 11 16 U.S.C. § 839b(h)(l)(A), (h)(10)(A). In making fish and wildlife a “co-equal partner” with the hydropower industry, the NPA adopted several innovations, marking the Act for its legislative craftsmanship. See 126 Cong.Rec. H9852 (Rep. Swift) (“[The Act] represents] legislative craftsmanship at its very best”); see also Michael C. Blumm and Andy Simrin, The Unraveling of the Parity Promise: Hy-dropower, Salmon, and Endangered Species in the Columbia Basin, 21 Envtl.L. 657, 667 (1991) [hereinafter Parity V]. '

First, the Act expressly required “textual consistency”; that is, that its provisions, together with other applicable laws, specifically including environmental laws, be construed in a consistent manner. 16 U.S.C. § 839 ; see Parity II, 13 Envtl.L. at 118-21. Second, the Act called for a systemwide remedial program to cover the entire Basin, rather than the more traditional approach of focusing on individual projects. H.R.Rep. No. 96-976, pt. II, 96th Cong., 2d Sess., at 38 (“The goal of the program is not to increase the obligations of water project owners and operators, but rather to go beyond a project-by-project approach on a river system whose multiplicity of projects and interdependent biological species makes [sic] a project-by-project approach unsatisfactory for all involved parties.”); see Parity V, 21 Envtl.L. at 667. Third, the Act shifted focus in wildlife mitigation from merely creating substitute resources, such as salmon hatcheries, to emphasizing changes in hydro project operations. Fourth, “Congress lowered the burden of proof for undertaking action by requiring that the remedial program (1) be based only on ‘best available scientific knowledge,’ not scientific certainty; (2) favor biological outcomes over economic ones, and (3) defer to the recommendations of agencies and Indian tribes with fish and wildlife expertise.” Id. at 667-68 (footnotes omitted). 12 Fifth, the Act tapped revenues of the Basin’s hydropower system as a source for financing the biological restoration. 16 U.S.C. § 839b(h)(10); 126 Cong.Rec. H9845 (Rep. Lujan) (“[The NPA] provides for the rebuilding of the salmon and steelhead runs that have been damaged by hydro development on the Columbia River, with the cost to be paid by the rate-payers in the Northwest, not by U.S. taxpayers.”). Finally, the NPA called for energy conservation to be developed as a resource ahead of traditional resources. 16 U.S.C. § 839 ; 126 Cong.Rec. H9848 (Rep. Pritchard) (“[The Act] treats energy conservation as a resource, making it the top priority in meeting the region’s energy needs. I know of no other legislation which carries such a strong mandate for implementing conservation.”).

Attempting to balance environmental and energy considerations, the Act states that fish and wildlife protection measures cannot jeopardize “an adequate, efficient, economical, and reliable power supply.” 16 U.S.C. § 839b(h)(5). 13 Also, the Council’s authority with respect to fish and wildlife measures is constrained; the Council “can guide, but not command, federal river management.” Through a Glass, 23 Envtl.L. at 1253; see 16 U.S.C. §§ 839b(h)(10), 839b(i), 839b(j); Northwest Resource Info. Ctr. v. National Marine Fisheries Serv., 25 F.3d 872 , 874 (9th *1379 Cir.1994) (“The Bonneville Power Administration [BPA] must act consistently with the Council’s program but in the end has final authority to determine its own decisions.”); Public Utility Dist. 1 v. BPA, 947 F.2d 386 , 392 (9th Cir.1991) (Section 839b(h)(ll)(A)(ii) “does not mean the Administrator and other agencies are limited to the program in exercising their responsibilities under the Act.”) (emphasis added), cert. denied, — U.S. —, 112 S.Ct. 1759 , 118 L.Ed.2d 422 (1992). 14 On the other hand, the Act requires federal water managers to act in a manner consistent with the Council’s fish and wildlife program. 15

Congressional mandate that the Council, once appointed and functioning, begin immediately to adopt a fish and -wildlife program, reflected Congress’s sense of urgency in the NPA; indeed, “the emphasis of the entire structure of the legislation is on prompt action.” Northwest Resource Info. Ctr. v. NMFS, 818 F.Supp. 1399, 1401 (W.D.Wash.1993), aff 'd, 25 F.3d 872 (9th Cir.1994); see 126 Cong.Rec. 10683 (Rep. Dingell) (“Adoption of the [fish and wildlife] program should not be delayed pending adoption of the plan. The program is to be adopted and implemented whether or not the plan is adopted. It exists independent of the plan and a part of it as well”); H.R.Rep. No. 96-976, pt. I, 96th Cong., 2d Sess., at 56 (“The Committee wants to stress that adoption of [fish and wildlife] recommendation[ ] is not intended to be delayed pending adoption of any plan or revision_”). Seven months after the Act was signed into law, the Council’s prompt action yielded an extensive set of recommendations from a coalition of federal and state fish and wildlife agencies and Indian tribes. See Parity V, 21 Envtl.L. at 670-74 (discussion of the recommendations). The coalition recommended a goal of restoring the Basin’s anadromous fish runs to their levels predating construction of the McNary Dam in 1953. Id. at 672. The key recommendation of the coalition concerned mainstem streamflows needed to help transport migrating juvenile anadromous fish downstream to the ocean during spring. 16 The coalition asserted that *1380 optimum fish survival required average weekly spring flows of 300-350 thousand cubic feet per second (kefs) at the Dalles Dam on the lower Columbia River. Id. at 671. In order to reduce the impact on hydropower generation, however, the fish and wildlife agencies recommended minimum flows of 220 kefs, subject to cutbacks of twenty-five percent during dry water years and increases in better-than-average water years, and warned that new biological information might necessitate higher flows. 17 The Indian tribes, on the other hand, refused to endorse the concession to hydropower interests, maintaining that Indian treaties entitled the tribes to optimum flows.

The Council promulgated the Columbia Basin Fish and Wildlife Program in November 1982. 18 The Council rejected the recommended goal of run sizes equal to those predating 1953. Rather, the program directed BPA to study fish losses from hydropower development and operations, as well as current and potential carrying capacity. To judge the efficacy of the program, the Council directed the development of area-by-area and stock-by-stock goals by 1984. By mid-1986 these goals remained undefined. 19

The centerpiece of the 1982 program, however, was the Council’s response to recommended mainstem flows. Rather than adopt a “sliding scale” of flows, “the Council avoided fixed flow levels and proposed a volumetric approach”:

Parity V, 21 Envtl.L. at 675 (footnotes omitted). The Council characterized this approach as the ‘Water Budget.” The Water Budget had the advantages over fixed flows of costing less in foregone hydropower revenues and involving fishery managers in day-to-day system operations, but it was limited to the period between April 15 and June 15. Id. Fishery managers claimed then, as is claimed in the present petitions, that a significant proportion of the spring, summer, and fall chinook runs extend past June 15.

In 1984, the Council approved a set of comprehensive amendments. The amendments included a five-year “Action Plan” setting priorities for program measures and implementation deadlines. They also established three interim goals: (1) the increase of quality and quantity of anadromous fish produced in the Basin (by improving mainstem passage, providing Water Budget flows, protecting against adverse effects of new hydro-power development, and increasing system-wide production capability); (2) the protection of ratepayer investment in the program (by improving harvest controls and monitoring the effectiveness of program measures); *1381 and (3) the progress of wildlife and resident fish measures where they did not conflict with anadromous fish measures. See Michael C. Blumm, Reexamining the Parity Promise: More Challenges Than Successes to the Implementation of the Columbia Basin Fish and Wildlife Program, 16 Envtl.L. 461, 476 (1986) [hereinafter Parity IV].

The Council adopted further amendments in 1985 and 1986, but amendments in 1987 established a new framework for implementing the program. The new framework sought greater accounting for the destructive effects of hydropower on anadromous fish at each stage of their life cycle. The new system focused on integrating into the system-wide program subbasin planning efforts designed to identify objectives and constraints on local fish production. The principal interim goal of the new system was to double the existing fish runs from approximately two- and-one-half million to five million annually. 20 The Council set no time-frame, however, for attaining this goal. Nevertheless, to facilitate implementation of the new system and achieve the double population goal, the Council gave priority to areas above the Bonneville Dam on the Columbia River, emphasizing improved mainstem passage. 21

Despite the Council’s progress in ensuring the installation of bypass systems and sufficient spills, salmon and steelhead populations generally declined through 1987. This trend continued the next three years. As a consequence, the National Marine Fisheries Service (“NMFS”) was petitioned in 1990 by special interests to list three Snake River stocks and one Columbia River stock under the ESA. Over the next two years, NMFS listed the three Snake River stocks and determined that the Columbia River stock was extinct. 22

The ESA proceedings spurred the region’s political leaders to convene a “Salmon Summit” in 1990 of various agencies and interested groups to develop a regional plan to end the salmon decline. While the Summit produced several interim measures, the participants returned to the Council for direction.

In 1991, the Council adopted a four-phase decision-making process to amend the fish and wildlife program. The first three phases of this process culminated in the Council’s adoption of the Strategy for Salmon in December 1992. See 56 Fed.Reg. 56935 (1992).

The Council received numerous proposals to its request for amendments to the fish and wildlife program. Fish and wildlife agencies and Indian tribes (collectively “agencies and tribes” or “fishery managers”) emphasized two points: (1) that substantial increases in Columbia River and Snake River spring and summer flows were necessary, 23 and (2) bio *1382 logical objectives, such as water particle travel time, 24 were necessary to measure the efficacy of restoration efforts. These two ideas formed the basis of a comprehensive regimen intended to replace the Water Budget.

While fishery managers recommended significant changes — substantial increases in mainstem flows and the adoption of specific biological objectives — power interests and DSIs defended the status quo — conservative steps calculated to produce positive, albeit minor, improvements. In particular, these latter interests vigorously challenged the biological benefits of flows above 85 kefs in the lower Snake River and 200 kefs in the lower Columbia River.

During the second phase of the process, which focused on salmon survival and harvest, the Council rejected the consensus of agencies and tribes that flows should be significantly increased, and adopted flows very close to those recommended by power interests and DSIs. As for biological objectives, the Council deferred consideration until the third phase, which would deal with salmon habitat and production.

Ultimately, the Council’s Strategy for Salmon called for both immediate and intermediate-term actions to enhance salmon survival in the rivers: increased river velocities, dam screens, water spills, reduced predation, and downstream barge transportation of juveniles. With respect to increased flows, the Council adopted the following:

No obvious biological objectives were included in the Program.

Since 1980, at the latest, the debate concerning the decline of anadromous fish runs in the Columbia River Basin has revolved around the issue of river flows. Not surprisingly, how the Council’s Program addresses this primary issue is the basis for the instant petitions for review. NRIC, the Yakima Nation, and the DSIs each argue for restriking the balance between salmon and energy. Nevertheless, Petitioners find common ground in arguing that the Council failed to explain the basis for its flow decisions and failed to adopt and apply proper standards and objectives.

Standards of Review

In construing a statute, we look first to the statute’s language. California ex. rel. State Water Resources Bd. v. FERC, 877 F.2d 743, 746 (9th Cir.1989), aff'd, 495 U.S. 490 , 110 S.Ct. 2024 , 109 L.Ed.2d 474 (1990); Pacificorp v. Bonneville Power Admin., 856 F.2d 94, 96 (9th Cir.1988). “Absent a clearly expressed legislative intention to the contrary, that language must ordinarily be regarded as conclusive.” Consumer Prod. Safety Comm’n v. GTE Sylvania, Inc., 447 U.S. 102, 108 , 100 S.Ct. 2051, 2056 , 64 L.Ed.2d 766 (1980).

Congress directed that the Administrative Procedure Act (“APA”), 5 U.S.C. §§ 701-06 , govern our review of Council actions. Adoption of the Program is a final action subject to judicial review for the purposes of the APA. 16 U.S.C. § 839f(e)(l)(A); see 5 U.S.C. § 553 .

We have previously considered the Council’s decisions under the APA. Seattle Master Builders Ass’n v. Pacific Northwest Elec. Power and Conservation Planning Council, 786 F.2d 1359 (9th Cir.1986), cert. denied, 479 U.S. 1059 , 107 S.Ct. 939 , 93 L.Ed.2d 989 (1987). We ruled that:

Id. at 1366 . Review under this standard is to be “searching and careful,” but remains “narrow,” and a court is not to substitute its judgment for that of the agency. Marsh v. Oregon Natural Resources Council, 490 U.S. 360, 378 , 109 S.Ct. 1851, 1861 , 104 L.Ed.2d 377 (1989). This is especially appropriate where, as here, the challenged decision implicates substantial agency expertise. See United States v. Alpine Land and Reservoir Co., 887 F.2d 207, 213 (9th Cir.1989) (“Deference to an agency’s technical expertise and experience is particularly warranted with re *1384 spect to questions involving ... scientific matters.”), cert. denied, 498 U.S. 817 , 111 S.Ct. 60 , 112 L.Ed.2d 35 (1990).

DISCUSSION

Petitioners agree in arguing that the Council failed to explain a “rational connection between the facts found and the choice[s] made” in the Strategy for Salmon. See Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Automobile Ins. Co., 463 U.S. 29, 43 , 103 S.Ct. 2856, 2866 , 77 L.Ed.2d 443 (1983); National Wildlife Fed’n v. FERC, 801 F.2d 1505, 1515 (9th Cir.1986). More specifically, NRIC and the Yakima Nation mount a procedural challenge to the Program under § 839b(h)(7), alleging that the Council failed to explain, in the Program, a statutory basis for its rejection of river flow recommendations of fishery managers. 25 NRIC and the Yakima Nation also contend that the Council failed to properly consider the criteria in § 839b(h)(6) for adopting program measures. Underlying each of these claims is an assertion that the Council failed to give due deference to fishery managers.

The DSIs join NRIC and the Yakima Nation in arguing that the Council failed to explain its rationale for adopting the program measures of the Strategy for Salmon. The DSIs assert their challenge from a different perspective, however, arguing that the Council improperly tilted the balance between salmon and energy in favor of the salmon by failing to engage in a critical cost-benefit analysis of each program measure.

Responding to these challenges, the Council contends that the NPA and the record support each of the Strategy for Salmon’s measures.

I. NRIC’s and the Yakima Nation’s Petitions

NRIC and the Yakima Nation first charge that the Council’s Program failed to explain a statutory basis for its rejection of the recommendations of fishery managers as required under § 839b(h)(7).

Section 839b(h)(7) states:

This section mandates that the Council “shall explain in writing, as part of the program,” a statutory basis for rejecting the recommendations of fishery managers. NRIC and the Yakima Nation emphasize that the Council’s Program failed to explain any basis for its decisions rejecting the agencies’ and tribes’ recommendations. However, NRIC and the Yakima Nation argue that, to the extent the Council may have “explained in writing, as part of the program,” grounds for its rejection of recommendations of fishery managers, the Council failed to explain and rely *1385 upon at least one of the three permissible statutory bases in § 839b(h)(7) allowing such action. The Council responds with numerous citations to the administrative record, claiming that such references fulfill the written statutory explanation requirement of § 8S9b(h)(7). We find the Council’s position to be inconsistent with the clear and unambiguous language of the Act; it states that the Council “shall explain in writing, as part of the program,’’ a statutory basis 26 for rejecting recommendations. 27 (Emphasis added). This mandate is particularly forceful with respect to the recommendations of those to whom the statute gives deference — fish and wildlife agencies and Indian tribes.

Furthermore, for a court to fulfill its function under the appropriate standard of review, it:

Moon v. U.S. Dept. of Labor, 727 F.2d 1315, 1318 (D.C.Cir.1984).

Study of the Strategy for Salmon evokes an undeniable sense of the tremendous effort and commitment invested to create such a program. The Council has the unenviable task of sorting through the multitude of diverse recommendations and compiling measures that form the framework of an aggressive, effective, and balanced fish and wildlife program. To ensure a balance of interests in the program, the Act specifically requires the Council to explain, in the program, its reasons for rejecting recommendations. This requirement is critical to our review of Council decisions, but, more importantly, it forces the Council to hold out its final decisions and their rationale for public consideration and scrutiny. The Council neglected this mandate of the Act.

The Strategy for Salmon fails to explain the reasons for the Council’s decisions rejecting recommendations of the agencies and tribes. 28 The Council, however, directs our attention to other documents in the administrative record to convince us of its compliance with § 839b(h)(7). Most notable are documents related to amendments created during the first and second phases of the Council’s decision-making process.

The Council completed Phase Two amendments on December 11, 1991. 57 Fed.Reg. 56935, 56936 (1992). It noticed the completion of Phase Three and the adoption of the Strategy for Salmon on December 1, 1992, a year later. Id. During the interim, the Council “reopened for additional comment” its consideration of measures adopted in the first two phases. Id. Fishery managers renewed their positions during this period. The Council concluded the third phase by announcing that it “consider[ed] its September, 1992 decision [adopting the Program] to be a final decision with respect to all three phases of the salmon and steelhead amendments process, and the phase three rule su-percedes the phase one and two amendments in their entirety.” Id. (emphasis added).

This sequence of events establishes the Council’s intent that the Strategy for Salmon supercede all earlier amendments to the Program. If the Council intended otherwise, it should have specifically incorporated desired materials into the Program. The Council’s reliance here on documents expressly super-ceded by its final decision is somewhat analogous to a court’s reliance on a judicial deci *1386 sion it previously vacated; reliance on such documents (or a vacated judicial decision), if allowed, would undermine the validity and authoritativeness of final decisions. We refuse, therefore, the Council’s invitation to consider these documents in our determination of whether the Council complied with the written statutory explanation requirement of § 839b(h)(7). 29

The Council emphasizes one document in particular, however, which was created during Phase Three, to demonstrate its compliance with § 839b(h)(7): the Response to Comments for Phase Three (“P3 Response”). This document is separate from the Strategy for Salmon, but it does explain some of the Council’s rationale for certain decisions. Nevertheless, underscoring the fact that the Program itself fails to explain any basis, much less a statutory basis, for the Council’s decisions rejecting recommendations of fishery managers, we find that the P3 Response fails, as well, to explain a statutory basis justifying such decisions.

In sum, § 839b(h)(7) requires the Council to explain, in the Program, a statutory basis for its rejection of recommendations. The Council failed to do so here with respect to the recommendations of agencies and tribes and was, therefore, not in accordance with the NPA. See 5 U.S.C. §§ 706 (2)(A), (2)(D). 30 As a consequence, we remand this matter to the Council with instructions that it comply with the written statutory explanation requirement of § 839b(h)(7).

' We note that the Council’s failure to comply with § 839b(h)(7)’s mandate' limits our ability to review the Council’s actions in these appeals. It has also hampered Petitioners efforts to scrutinize and challenge the Council’s decisions. Nevertheless, most of the parties’ remaining contentions relate more to statutory interpretation of the fish and wildlife provisions of the Act, rather than the Council’s substantive decisions. We recognize that many of these statutory construction issues have plagued efforts to finalize the Strategy for Salmon. Therefore, we carefully consider the statutory interpretation issues with the expectation that our effort will aid the parties’ efforts on remand.

NRIC and the Yakima Nation contend that the Council failed to comply with the criteria set out in § 839b(h)(6) for adopting specific program measures. To decide this issue, we must determine the meaning of § 839b(h)(7)’s mandate that the Council give “due weight” to the recommendations and expertise of fishery managers when resolving inconsistencies in recommendations of program measures, and the nature of the criteria in § 839b(h)(6).

NRIC and the Yakima Nation claim that the fish and wildlife provisions of the Act and their legislative history reveal a congressional intent that the Council give a high degree of deference to fishery managers. The Council, on the other hand, argues that fishery managers are entitled to merely a nominal degree of deference, and urges that the discretion granted to it in the fish and wildlife provisions strictly limits judicial review of its decisions.

There is no question that § 839b(h)(7) requires that deference be given to the recommendations and expertise of agencies and tribes — the question is how much deference is due.

In Seattle Master Builders, 786 F.2d at 1367 , we stated:

(Emphasis added). Seattle Master Builders involved model conservation standards the Council adopted in its conservation and electricity usage plan for the region. According *1387 ly, we considered the power planning provisions of § 839b only — we did not consider the fish and wildlife provisions, nor did we even refer to § 839b(h). Id. at 1366 (“In this action petitioners challenge only that portion of the 1983 plan which establishes model energy conservation standards for new residential construction. Consequently, we so limit the following discussion.”).

The power plan provisions of the Act are cast in broad terms. For instance, section 839b(e)(2) provides that the “plan shall set forth a general scheme for implementing conservation measures and developing resources....” In establishing this general scheme, the Council is to consider several enumerated factors and “other criteria that may be set forth in the plan” by the Council. 16 U.S.C. § 839b(e)(2)(A)-(C). Furthermore, the NPA provides that the plan shall include seven elements that should be described “in such detail as the Council deems appropriate.” 16 U.S.C. § 839b(e)(3). Finally, the Council is to adopt “model conservation standards” in the plan that “reflect geographic and climatic differences within the region and other appropriate considerations....” 16 U.S.C. § 839b(f)(l). Relevant to this provision, we noted in Seattle Master Builders, 786 F.2d at 1368 , that “[t]he Act does not require the Council to follow any particular method or timetable for forecasting the amount of energy demand.”

These provisions grant the Council considerable flexibility in preparing a power plan; indeed, the Council’s function under these provisions is essentially legislative.

In stark contrast are the fish and wildlife provisions of § 839b(h). To initiate the development of a fish and wildlife program, Congress required the Council to solicit recommendations from the region’s fishery managers. The Act requires the Council to give notice of all the recommendations and supporting materials to, among others, the Administrator, federal and state fish and wildlife agencies, Indian tribes, federal agencies that manage, operate, or regulate the hydro-system. 16 U.S.C. § 839b(h)(4)(A). The Council is then required to provide for public participation and comment on the recommendations. 16 U.S.C. § 839b(h)(4)(B). From this process, “[t]he Council shall develop a program on the basis of such recommendations, supporting documents, and views and information obtained through public comment and participation, and consultation with the agencies, tribes, and customers referred to in subparagraph (A) of paragraph (4).” 16 U.S.C. § 839b(h)(5). In short, the NPA requires the Council to develop the program from sources outside the Council. In doing so, the Council must adopt program measures that are consistent with the purposes of the Act — the protection, mitigation, and enhancement of fish and wildlife, “while assuring the region an adequate, efficient, economical, and reliable power supply.” 16 U.S.C. § 839b(h)(5); see also 16 U.S.C. § 839b(h)(7). If reconciliation of recommendations is required, the Council must give “due weight” to fishery managers’ recommendations, and may disregard such recommendations only after explaining in the program a statutory basis for its decision. Id.

The difference between the power plan and the fish and wildlife provisions of § 839b, then, is a contrast of generous discretion and bound discretion; the difference highlights the limited discretion of the Council under 839b(h)(7).

Although an agency’s interpretation of the statute under which it operates is entitled to some deference, “this deference is constrained by our obligation to honor the clear meaning of a statute, as revealed by its language, purpose, and history.” International Broth. of Teamsters, Chauffeurs, Warehousemen and Helpers of America v. Daniel, 439 U.S. 551 , 566 n. 20, 99 S.Ct. 790 , 800 n. 20, 58 L.Ed.2d 808 (1979). In other words, “[i]n assessing action by an administrative agency, the court must temper its standard of review according to the degree of discretion Congress has given to the agency concerned.” Farley v. Sullivan, 983 F.2d 405, 407 (2d Cir.1993). “[T]he degree of respect to be accorded to an agency’s statutory interpretation of the limit of its authority varies, depending upon such factors as the interpretation’s inherent reasonableness, its consistency with prior precedent, its basis in statutory text and legislative history, and whether it is based on experience or expertise peculiarly within the agency’s ken.” Office of Consum *1388 ers’ Counsel v. FERC, 655 F.2d 1132 , 1141 (D.C.Cir.1980).

We conclude that § 839b(h) binds, more than unleashes, the Council’s discretion with respect to fish and wildlife issues. Indeed, we are convinced that the fish and wildlife provisions of the NPA and their legislative history require that a high degree of deference be given to fishery managers’ interpretations of such provisions and their recommendations for program measures.

Section 839b(h)(2) of the Act requires that the Council solicit recommendations from agencies and tribes for:

The Council “shall [then] determine whether each recommendation received is consistent with the purposes of this chapter.” 16 U.S.C. § 839b(h)(7). One of the primary purposes of the chapter is that “[t]he Council shall include in the [fish and wildlife] program measures which it determines” will meet the criteria set forth in § 839b(h)(6). 16 U.S.C. § 839b(h)(6). Inconsistencies between recommendations shall be resolved by the Council, “giving due weight to the recommendations, expertise, and legal rights and responsibilities” of agencies and tribes. 16 U.S.C. § 839b(h)(7) (emphasis added).

Congress’s intent is manifest in the NPA’s legislative history. Representative Dingell, one of the NPA’s principal sponsors, emphasized, “[c]learly, the [C]ouncil should rely heavily on the fish and wildlife agencies of the State and Federal Governments and not try to become a superfish and wildlife entity.” 126 Cong.Rec. E10683 (1980). The House Committee on Interstate and Foreign Commerce stated:

H.R.Rep. No. 96-976, pt. I, 96th Cong., 2d Sess., at 57 (emphasis added).

Hence, Congress realized that furtherance of the purpose of the Act, that fish and wildlife be on a par with energy, required that the Council defer to the recommendations of agencies and tribes. Of course, the reason for this deference to fishery managers is their unique experience and expertise in fish and wildlife. Congress intended that the Council not simply tap this resource of information and advice, but that it “heavily rely” upon it.

In promulgating § 839b, Congress anticipated the onus fish and wildlife issues would place on the Council in creating a hydropower plan that balanced fish and wildlife with energy. Congress recognized, in particular, that fish and wildlife issues were, and should be, outside the expertise of the Council and the hydropower regulating agencies. Nonetheless, the need for experience and expertise with respect to fish and wildlife was plain. Looking to those having responsibility for managing such resources, Congress found the experience and expertise on which the Council should rely to frame a fish and wildlife program. Accordingly, Congress required in § 839b that fishery managers be given a high degree of deference in the development of a fish and wildlife program for the Basin.

In light of the NPA’s legislative history and text, it follows that fishery managers, as well as the Council, be given deference in interpreting the fish and wildlife provisions of the Act. This conclusion is consistent with our holding in Public Util. Dist. 1, 947 F.2d at 390, that the BPA is due deference in *1389 interpreting the power plan provisions of the NPA because it was involved in the drafting of the Act. The role that fishery managers had in the promulgation of the NPA’s fish and wildlife provisions demands no less of us here. Furthermore, the unique experience and expertise of fishery managers makes their interpretations of § 839b, especially § 839b(h)(6), particularly helpful. We find it inherently reasonable to give agencies and tribes, those charged with the responsibility for managing our fish and wildlife, a high degree of deference in the creation of a program and the interpretation of the Act’s fish and wildlife provisions.

The Council’s contention that § 839b grants it virtually unfettered discretion in creating a fish and wildlife program is unfounded; the Act’s fish and wildlife provisions significantly temper the Council’s discretion. At any rate, we do not decide here whether the Council abused its discretion in giving or failing to give proper deference to fishery managers because the record as it stands precludes such an inquiry. We note again that the Council failed to explain, in the Program, its reasons for rejecting the recommendations of fishery managers. This failure of the Council is disturbing given that it adopted, for the most part, the flows and measures recommended by power interests and DSIs, despite the overwhelming consensus among agencies and tribes in favor of significantly higher flows and more scientifically-based biological objectives.

We next turn to the nature of the § 839b(h)(6) criteria. The Council claims that the criterion in § 39b(h)(6) merely “set the agenda but do not dictate the substance of measures.” As a result, the Council urges, its “determinations necessarily involve a great deal of judgment, not well suited for judicial review.” The Council misconstrues § 839b(h)(6).

The criteria in § 839b(h)(6) are mandatory. The provision states that the Council shall include program measures that adhere to each of the criterion. As recognized above, the fish and wildlife provisions of the Act significantly circumscribe the Council’s discretion with respect to fish and wildlife. Construing the criteria in § 839b(h)(6) as substantive ■ criteria is consistent with that construction.

The legislative history is also helpful in characterizing the nature of the criteria in § 839b(h)(6). In discussing § 839b(h), the House Interior and Insular Committee stated that “[t]he section sets out specific criteria such recommendations must meet in order to be included in the program.” H.R.Rep. No. 96-976, pt. II, 96th Cong., 2d Sess., at 37, 1980 U.S.Code Cong. & Ad.News at pp. 5989, 6035. The Committee stated further, with respect to § 839b(h)(6), that it:

H.R.No.Rep. 96-976, pt. II, 96th Cong., 2d Sess'., at 44 (emphasis added). The House Committee on Interstate and Foreign Commerce added that “[a]ny determination [under § 839b(h) ] must set forth the reasoning in support of that determination and will be subject to judicial review.” H.R.Rep. No. 96-976, pt. I, 96th Cong., 2d Sess., at 57.

Finally, we previously decided that other provisions of § 839b(h) impose substantive requirements. In Confederated Tribes and Bands of the Yakima Indian Nation v. FERC, 746 F.2d 466, 473 (9th Cir.1984), we held that the FERC failed to meet its obligation to give fish “equitable treatment” under § 839b(h)(ll)(A)(i) of the NPA. We construed this requirement as a “substantive” obligation. Id.; see also National Wildlife Fed’n, 801 F.2d at 1514-15 . The criteria in § 839b(h)(6) are no less substantive, and measures adopted thereunder no less reviewable by this Court. 31

Section 839b(h)(6) provides:

We now focus on each of the criterion in this provision.

The first criterion requires that measures “complement the existing and future activities of the Federal and the region’s State fish and wildlife agencies and appropriate Indian tribes.” 16 U.S.C. § 839b(h)(6)(A). Since adoption of the Program, there have been various proceedings under the ESA concerning listed anadromous fish in the Basin. Several ESA actions have been filed, the lead case by the State of Idaho against NMFS and other agencies in 1992. Idaho Dept. of Fish and Wildlife v. NMFS, 850 F.Supp. 886 (D.Or.1994). That case challenges government actions in operating the Federal Columbia River Power System (“FCRPS”) in 1993. More specifically, the Idaho Department of Fish and Game claims that NMFS and other Federal agencies have violated the ESA by “(1) failing to insure that FCRPS operations are not likely to jeopardize listed species; (2) omitting consideration of all relevant scientific factors; (3) failing to include all reasonable and prudent mitigation measures to reduce incidental take of listed species; (4) limiting consideration of short and long term impacts and measures to the immediate nine-month operational period under consideration; and (5) operating the FCRPS between April 15 and May 26,1993[,] prior to completion of the biological opinion.” Id. at 887 .

A major contention in the case concerns the adequacy of NMFS’s 1993 Biological Opinion. The district court, experienced in these particular matters, granted Idaho’s summary judgment, finding that the Biological Opinion was “arbitrary and capricious and otherwise not in accordance with the meaning and underlying purposes of the Endangered Species Act, § 7(a)(2), with respect to the chosen jeopardy standard and their consideration of reasonable and prudent alternatives to avoid jeopardy.” Id. at 900. The court commented:

Id.

Idaho Dept, of Fish and Game is relevant to the instant case not just because it involves, ultimately, the same issue — what to do about preserving and restoring the salmon — but because it urges policy and operations in a direction away from the status quo towards affirmative action. NRIC and the Yakima Nation press for the same direction from the Council, the instant case being their vehicle for doing so. While we do not now decide whether the Council’s measures complement existing and future activities of fish and wildlife agencies and Indian tribes in the ESA actions, the Council’s rejection of the agencies’ and tribes’ consensus as to increased flows and biological objectives does not appear to square well with these efforts.

The second criterion requires that measures be based on and supported by the “best available scientific knowledge.” 16 U.S.C. § 839b(h)(6)(B). That the standard requires only the best available scientific knowledge ensures action in the promulgation and implementation of a fish and wildlife program. Textual consistency with § 839b(h)(7) requires that the Council defer to the agencies’ and tribes’ recommendations as to what is the best available scientific knowledge.- Moreover, the standard requires only the best available scientific knowledge, not data. Section 839b(h)(6)(B) suggests, therefore, that reasonable inferences and predictions may be drawn from the best available scientific knowledge. In this case, determining whether the program measures are based on and supported by the best available scientific evidence is prevented by the Council’s failure to comply with § 839b(h)(7)’s written státutory explanation requirement.

The third criterion requires that alternative measures be evaluated for effectiveness in achieving sound biological objectives. 16 U.S.C. § 839b(h)(6)(C). Sound biological objectives relate the biological needs of fish and wildlife to the operations of the hydropower system. If two measures are equally effective in achieving the same biological objectives, the Council must adopt the one with less economic cost. Id.

NRIC and the Yakima Nation claim that the Council failed to adopt sound biological objectives. 32 They claim that the Council’s failure to adopt a particle travel time objective resulted in the adoption of flow measures in the absence of a meaningful evaluation of their effectiveness. The Council relies on the Program’s doubling goal, rebuilding targets, and performance measures to demonstrate that it did, in fact, adopt sound biological objectives.

NRIC’s and the Yakima Nation’s instant claim was markedly described earlier by the Fish and Wildlife Service (“FWS”) when it commented on the final amendments to the Program:

(July 1992 letter from the Regional Director of FWS to Ted Hallock, Chairman of the Council). We are persuaded that the Program fails to incorporate the sound biological objectives contemplated in the Act. For instance, the goal of doubling population is a policy statement; to that end, it is an untell-ing policy statement because it sets no firm deadline for attaining the goal. The policy’s statement of direction and philosophy is a far cry from the specificity of a discrete biological objective. This is also true with respect to the rebuilding targets and performance standards in the Program. Moreover, none of these framework elements appear to have guided the Council in evaluating alternative recommendations and measures; rather, the Council utilized these elements to convey and estimate the effectiveness of measures already adopted. Indeed, the record evidences the Council’s intent ultimately to refrain from specifically adopting biological objectives. The Council’s post hoc adoption of the doubling goal, rebuilding targets, and performance standards as sound biological objectives, despite the recommendations of agencies and tribes, ignores the purposes and requirements of § 839b(h)(6)(C). On remand, the Council should consider measures in light of specific biological objectives, giving the recommendations of agencies and tribes “due weight” in its determination of which biological objectives should be utilized.

The fourth criterion requires that measures “be consistent with the legal rights of appropriate Indian tribes in the region.” 16 U.S.C. § 839b(h)(6)(D). Congress recognized the preexisting rights that tribes reserved for themselves, in contrast to rights granted to them by the Government. In this light, it is reasonable to conclude that measures that would allow the extinction of Snake River fall chinook, for instance, upon which the Yakima people largely depend for their livelihood, may very well be inconsistent with the Yakima Nation’s treaty reserved fishing rights. See Washington State Commercial Passenger Fishing Vessel Ass’n, 443 U.S. at 674-82, 99 S.Ct. at 3068-72 . We note this possibility only to sensitize the parties to this Court’s duty to honor the purposes of the Act while, at the same time, strictly construing its language.

The last criterion addresses the particular needs of anadromous fish. It requires that measures provide “improved survival of such fish” at hydropower facilities, 16 U.S.C. § 839b(h)(6)(E)(i), and “flows of sufficient quality and quantity between such facilities to improve production, migration, and survival of such fish as necessary to meet sound biological objectives,” 16 U.S.C. § 839b(h)(6)(E)(ii). This criterion accents the importance of the Council strictly complying with the Act’s mandates with respect fish and wildlife, especially that requiring that a high degree of deference be given to fishery managers. The Council’s failure to comply with § 839b(h)(7) prevents us from addressing the issue of whether the Council complied with this criterion in adopting the Strategy for Salmon’s measures. The record evokes in us, however, a strong sense of skepticism; without explanation, the Council rejected the consensus of most fishery man *1393 agers on the issues of flows and biological objectives in favor of the recommendations of power interests and DSIs.

In justifying its decisions with respect to flows, the Council relies heavily on the NMFS’s 1992 Biological Opinion to argue that the Program’s flows are of sufficient quality and quantity to comport with the NPA. The opinion, however, is not nearly so reassuring. The NMFS considered the Council’s overall plan for 1992 operations only. The opinion states:

The NMFS found that 1992 operations under the plan’s Strategy for Salmon were not likely to retard survival further or to preclude any progress “toward reversing the decline of listed and proposed species.” The NMFS recognized two things, however: (1) that the plan, as it related to endangered fish, did little more than maintain the status quo and (2) that the status quo was insufficient to maintain listed runs in future years. We do not take this opinion as either dispositive authority of the Council’s Program, or even as NMFS’s stamp of approval. 33

Significantly, the Council staffs analysis in November 1991 concluded that stocks with low productivity would “probably decline to extinction” with the Program’s measures. The staff further concluded that with the Program’s measures, “medium productivity stocks appear to stabilize but not rebuild. Rebuilding is seen only for those stocks in the best habitat that are presently limited only by passage survival rate.”

Both the NMFS’s 1992 Biological Opinion and the Council staffs analysis undermine, rather than bolster, the Council’s assertion that the Program is unquestionably consis1 tent with the purposes of the fish and wildlife provisions of the Act.

We recognize the standards in § 839b(h)(6) as substantive criteria that each program measure must meet. Unfortunately, we are precluded, for the most part, from determining whether the Council applied these standards correctly. We conclude with confidence, however, that the Council failed to utilize sound biological objectives in evaluating program measures. We are also concerned that the Council may not have given the required degree of deference to the recommendations of fishery managers. The Council should address these points on remand.

Finally, NRIC complains that the Council may not be “committed” to implementing the Program’s intermediate-term actions. The Council admits in the Strategy for Salmon that immediate measures with respect to river flows “do not appear to be enough, in themselves, to rebuild salmon runs.” Such disclaimers have been a common element of the Program since the Council first adopted a general plan in 1982. 34 The success of the Strategy for Salmon is, therefore, dependent on the Council committing itself to aggressive action to save and rebuild the Basin’s anadromous fish runs. We construe the Program as the Council’s binding commitment to the timely implementation of all measures if they satisfy the substantive criteria of the NPA.

II. The DSIs’ Petition

The DSIs complain that the Council violated the NPA by failing to conduct a critical *1394 cost-benefit analysis of each measure in the Program. They also claim that the Program’s economic impact and drawdown measures violate the Act.

The NPA requires, in part, that the overall Program assure the region “an adequate, efficient, economical, and reliable power supply.” 16 U.S.C. § 839b(h)(5). It also requires, under § 839b(h)(6)(C), that the Council “utilize, where effective alternative means of achieving the same sound biological objective exist, the alternative with the minimum cost.”

The DSIs argue that these two provisions combine to require the Council to engage in a critical cost-benefit analysis of each of the Program’s measures; that is, an evaluation of whether each of the Program’s measures imposes costs on the hydropower system that exceed the fish mitigation benefits each measure should achieve. The Council counters that it must only evaluate the costs of the fish and wildlife Program to ensure that the Program does not undermine the region’s power supply.

Program costs, generally, were addressed by Congress prior to enacting the NPA. The House Committee on Interstate and Foreign Commerce stated:

H.R.No.Rep. 96-976, pt. I, 96th Cong., 2d Sess., at 57 (emphasis added). The Committee further noted that:

Id. at 49 (emphasis added).

We conclude from our study of § 839b(h)(5) and the legislative history that: (1) Congress did not say the Council should perform a critical cost-benefit analysis of each measure; 35 (2) a fish and wildlife measure cannot be rejected solely because it will result in power losses and economic costs; and (3) the Council must assess overall power and economic impacts so that the Program does not cause an inadequate, inefficient, uneconomic, and unreliable power supply. These conclusions are underscored by the substantive criterion in § 839b(h)(6)(C).

Section 839b(h)(6)(C) requires individual program measures to be based on sound biological objectives. In those instances where two proposed measures equally achieve the biological objectives, the provision imposes a “cost-effectiveness” standard to determine which has the least economic cost. See 16 U.S.C. § 839a(4). In this regard, § 839b(h)(6)(C) emphasizes the achievement of predetermined biological objectives in a least-cost manner. See Misplaced Role of Cosh-Benefit Analysis, 16 Envtl.L. at 549-51. Unlike cost-benefit analysis, which “seeks to reduce and compare all values by a common denomination — the dollar,” see id. at 534 n. 60, cost-effectiveness, in this context, prevents cost considerations from precluding the biologically sound restoration of anadromous fish in the Columbia River Basin to the extent affected by hydro-power development and operations so long as an adequate, efficient, economical, and reliable power supply is assured.

Finally, engaging in a critical cost-benefit analysis of each program measure intended to protect, mitigate, and enhance fish and *1395 wildlife would work against such efforts. Cost-benefit analysis “measures only the magnitude of costs and benefits; it does not assess the distribution, or equity, of the resulting gains and losses.” Id. at 535. In other words, non-economie values the Program seeks to further, such as equity, ecology, conservation, and culture, would be ignored. In addition, the application of cost-benefit analysis to the NPA’s fish and wildlife provisions would eviscerate another design of the Act: to allow knowledge that is less than scientific uncertainty be a basis for sound decision-making, thereby shifting the burden of proof to those challenging fishery managers’ recommendations. We reject the DSIs’ contention that a critical cost-benefit analysis of each program measure is required.

The DSIs also complain that the Strategy for Salmon’s economic impact imposes unreasonable burdens on the hydro-power system, violating the NPA. To the extent the DSIs challenge specific measures under the criterion in § 839b(h)(6)(C), the record is insufficient for us to make a determination. It is worth noting, however, that the DSIs fail to demonstrate that any two proposed measures equally achieved the same biological objectives, requiring an analysis of the measures’ cost-effectiveness. Similarly, to the extent the DSIs challenge the Program under § 839b(h)(5), because of its overall economic impact, they fail to demonstrate that the Council’s actions were “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706 (A). The DSIs complain of increased costs and lost profits, but provide no specific evidence that the overall economic impact of the Program on the hydropower system is unreasonable; Congress expected increased costs and lost profits to the hydro-power system to the extent the system was responsible for damaging fish and wildlife in the region.

The DSIs’ last contention is that the Program’s reservoir drawdown measures are contrary to the Act. Specifically, they challenge the Council’s expressed intent to implement the drawdown strategy unless it was shown to be “structurally or economically infeasible, biologically imprudent.” The DSIs believe the Act imposes a stricter standard for adopting program measures than simply determining whether they are “infeasible” or “imprudent.” The DSIs ignore, however, the language of the drawdown measure that reservoir drawdowns will be rejected if they are “inconsistent with Sections 4(b)(5)— (7) of the Northwest Power Act.” This contention lacks merit.

CONCLUSION

The NPA adopted fish and wildlife restoration as a primary goal, leaving to the Council and interested parties the onerous burden of deciding how to restore such resources. In doing so, the Act placed a premium on prompt action, allowing decisions to be made on the best available scientific knowledge. It also limited the role of economic considerations in decision-making. Most importantly, however, the Act acknowledged fish and wildlife as an irreplaceable finite resource.

Unfortunately, the record reveals few profound successes resulting from these innovations in thinking. The Council’s approach seems largely to have been from the premise that only small steps are possible, in light of entrenched river user claims of economic hardship. Rather than asserting its role as a regional leader, the Council has assumed the role of a consensus builder, sometimes sacrificing the Act’s fish and wildlife goals for what is, in essence, the lowest common denominator acceptable to power interests and DSIs. The Council has failed at least two requirements of the NPA in its process, as well as § 706 of the APA: it has failed to explain a statutory basis for its rejection of recommendations of fishery managers and it has failed to evaluate proposed program measures against sound biological objectives. There is also concern that the Council may have failed to give proper deference to fishery managers and to fully comply with other substantive criteria for program measures. We, therefore, remand these matters to the Council for reconsideration consistent with this opinion.

. See Kai N. Lee & Jody Lawrence, Adaptive Management: Learning From the Columbia River Basin Fish and Wildlife Program, 16 Envtl.L. 431, 441 (1986) [hereinafter Adaptive Manage ment].

. Two intervenors have filed briefs: the United States Government and the State of Idaho. The United States intervened to clarify the potential impact these actions have on various Government agencies and other proceedings under the Endangered Species Act. The State of Idaho intervened to protect its interest in Snake River salmon and steelhead runs in Idaho, most of which are in serious decline.

. The Columbia River Basin is comprised largely of the Columbia River and its largest tributary, the Snake River. Geographically, management of the Basin most directly affects Idaho, Montana, Oregon, Washington, and our northern neighbor, Canada.

. The Columbia River Basin produces steelhead trout and four principal species of salmon: chi-nook, sockeye, coho, and chum. When discussing salmon species, biologists distinguish them by "seasonal races” or “runs” and "tributary stocks” and river “substocks.” Thus, the Snake River fall chinook is a fall race and a Snake River stock.

. Collectively, the steelhead and salmon are "anadromous” fish; that is, "any fish which spawns or is artificially produced in freshwater, reaches mature size while rearing in saltwater and returns to freshwater to reproduce." United States v. Washington, 384 F.Supp. 312, 405 (W.D.Wash.1974). Young salmon will not start the smoltification process (swimming downstream and adapting from freshwater to saltwater) until they achieve certain physical and physiological characteristics. John V. Byrne, Salmon is King — or is it?, 16 Envtl.L. 343, 352-53 (1986) [hereinafter Salmon is King].

. In Washington v. Washington State Commercial Passenger Fishing Vessel Ass'n, 443 U.S. 658 , 667 n. 11, 99 S.Ct. 3055 , 3071 n. 11, 61 L.Ed.2d 823 (1979), the Supreme Court told the tribes: "The paper [the treaty] secures your fish.” The Court also noted that the treaty guarantee of "the right of taking fish" was meaningful only if fish were available for the taking. Id. at 678 , 99 S.Ct. at 3071.

. Pub.L. No. 75-329, 50 Stat. 731 (1937) (codified as amended at 16 U.S.C. §§ 832-8321 ). The Senate passed a resolution directing the Commissioner of Fisheries to assess the effect of the Bonneville Dam on the Columbia fish runs and to recommend steps "to attain the full conservation of such fish_" See S.Doc. No. 87, 75th Cong., 1st Sess. (1937).

. Between 1973 and 1977, low flow years for the upper Columbia River Basin, virtually the entire flow was put through the turbines of the main-stem dams. This resulted in estimated losses of ninety-five percent of the salmon and steelhead in 1973 alone. H.R.Rep. No. 96-976, pt. I, 96th Cong., 2d Sess., at 47.

. The Coordination Act continued the emphasis, from the late 1930s, on hatcheries as a means to compensate for fish and wildlife losses. As a consequence, the focus was on procurement of substitute resources rather than minimizing adverse impacts on the environment.

. See, e.g., 126 Cong.Rec. E5105 (Rep. Dingell) ("[I]n developing the program, the Council and BPA should not slight wildlife needs. The fishery needs are important. Much emphasis in consideration of this legislation was on the anadromous and other fishery resources. However, from the beginning of our consideration of the bill in the House, we have stressed the need to protect, mitigate the adverse effects on, and enhance both fish and wildlife.”); 126 Cong.Rec. H10680 (Rep. Dingell) ("The conservation and enhancement of the great migratory fish and wildlife populations of the. Pacific Northwest, something of great concern to the sportsmen and conservationists of this Nation, are, for the first time, a matter of urgent priority under this legislation. They are place [sic] on a par with other purposes for Federal facilities in this area. If the fish populations of the Pacific Northwest are to be restored to the sportsmen, the Indians and the commercial fishermen, this is the mechanism which will do it.”); 126 Cong.Rec. H2062 (Rep. Bonker) (“Federal management of the Columbia River should be conducted so as to optimize production of salmon and steelhead as well as hydroelectricity.”); 126 Cong.Rec. H9846 (Rep. Lujan) ("The concept that protection, mitigation and enhancement of fish and wildlife while assuring an adequate, efficient, economical and reliable power supply is the most basic principle in developing the program. It is particularly important that this concept be recognized whenever a program is developed that is likely to be implemented before a plan is developed or amended.”); 126 Cong.Rec. H9848 (Rep. Dingell) (“Thd bill contains provisions relative to fish and wildlife on the Columbia River and its tributaries. These provisions will assure that the power-planning decisions [sic] fish and wildlife concerns are adequately met. Fish and wildlife, for the first time in this region, will be treated on a par with power and other purposes.").

. 5 The Act allows the States of Idaho, Montana, Oregon, and Washington to each appoint two persons to serve on the Council. 16 U.S.C. § 839b(a)(2).

. Specific measures were to satisfy, in addition to other criterion, river flows of sufficient quality and quantity to improve production, migration. and survival of anadromous fish. 16 U.S.C. § 839b(h)(6)(E).

.It is worth noting the statute assures a "power supply” not a "hydropower supply.” This highlights, again, conservation and the development of other resources as purposes of the NPA.

. The four federal water managers who affect Columbia River flows are BPA (as marketer of federal hydropower), the U.S. Army Corps of Engineers (as operator of most of the mainstem federal dams), the Bureau of Reclamation (as operator of several large storage dams, including the Grand Coulee Dam), and the Federal Energy Regulatory Commission ("FERC”).

. See 16 U.S.C. § 839b(h)(10)(A); see National Wildlife Fed’n v. FERC, 801 F.2d 1505, 1514-15 (9th Cir.1986). Moreover, it requires all four federal water management agencies to give “equitable treatment” to fish and wildlife. 16 U.S.C. § 839b(h)(ll)(A)(i). It also directs the federal agencies to take the program "into account ... to the fullest extent practicable” at each relevant stage of their decision making. 16 U.S.C. § 839b(h)(l)(A)(ii); see 126 Cong.Rec. 10683 (Rep. Dingell) ("Again it is important to stress that the fish and wildlife program, as well as the recommendations, may well proceed [sic] the plan. If so, the BPA and others should not delay implementation pending adoption of the plan which will incorporate this program.... [Subsection (h) also provides a directive to BPA and other Federal agencies responsible for the management and operation or regulation of hydro facilities on the Columbia or its tributaries to adequately protect, mitigate, and enhance for fish and wildlife affected by such facilities in a manner that insures equitable treatment for fish and wildlife with other purposes for which system and facilities are managed and operated or regulated. They must achieve the directive consistent with the purposes of this act and other applicable laws."); 126 Cong.Rec. H9845 (Rep. Lujan) ("While creating a mechanism for the acquisition and allocation of power by [BPA] within the region, the bill carefully limits the authority of the BPA administrator. It does not create anything resembling an 'energy czar.' ”); H.R.Rep. No. 96-976, pt. I, 96th Cong., 2d Sess., at 50 ("Lastly, these purposes provide that BPA, the Council, and other Federal agencies must carry out their duties and responsibilities and exercise their authorities to protect, mitigate, and enhance fish and wildlife...."); H.R.Rep. No. 96-976, pt. II, 96th Cong., 2d Sess., at 45, 1980 U.S.Code Cong. & Ad.News at pp. 5989, 6043 ("BPA shall act consistently with the regional plan, the [fish and wildlife] program ..., the purposes of this Act and other provisions of law.”).

.Other important recommendations included installation of juvenile bypass systems in each dam; interim spills at mainstem dams (until bypass systems were installed); research on the causes and solutions to predation and residual-ism; improved upstream adult passage; and increased effectiveness of both natural and hatchery production.

Spills are the most biologically benign way of safely passing fish at dams, but the water cannot then be utilized to generate power because it is not passed through the dams' turbines. Spills, therefore, have generally been considered an interim measure. Bypass systems, on the other *1380 hand, provide a long-term solution because they provide relatively effective safe fish passage without significantly reducing power generation.

Another notable aspect of the recommendation was the coalition's rejection of transportation by truck or barge of juvenile fish around dams as an alternative to enhanced streamflows.

. The coalition estimated mortalities of more than ninety percent if the flows fell below 220 kefs.

. Because of the lack of scientific certainty upon which to base decisions and the Act's bias toward action, the Council adopted "adaptive management” as a conceptual approach and strategy for implementation:

Adaptive Management at 442 (emphasis added). “The central point [of this approach] is learning.” Kai N. Lee, Rebuilding Confidence: Salmon, Science, and Law in the Columbia Basin, 21 Envtl.L. 745, 781 (1991) [hereinafter Rebuilding Confidence ].

.See Michael C. Blumm, Implementing the Parity Promise: An Evaluation of the Columbia Basin Fish and Wildlife Program, 14 Envtl.L. 277, 291-92 n. 58 (1984) [hereinafter Parity III] (BPA's allegation that funding the goals study proposed by the fishery agencies and tribes was "unacceptable to BPA management" and "not consistent with ratepayer interests” because it *1381 would amount to "funding advocacy” and be inconsistent with “sound business principles").

. This goal was based on the Council’s 1986 losses study, which estimated that the hydropower system was responsible for five to ten million fish lost each year. Although the Council fixed such losses at eight million fish annually, it adopted the lesser goal of five million on the grounds that socioeconomic and biological conditions reduced the possibility of restoring runs to greater levels.

. The Council recognized the importance of preserving the genetic diversity of individual fish runs. For a discussion of the importance of genetic diversity in the Basin's salmon and steel-head, see Michael L. Goodman, Comment, Preserving the Genetic Diversity of Salmonid Stocks: A Call for Federal Regulation of Hatchery Programs, 20 Envtl.L. Ill (1990).

. For the ESA decisions by NMFS, see 56 Fed. Reg. 58619 (1991) (Endangered Status for Snake River Sockeye); 57 Fed.Reg. 14653 (1992) (Threatened Status for Snake River Spring/Summer and Fall Chinook); 56 Fed.Reg. 29553 (1991) (lower Columbia River Coho).

.The Columbia Basin Fish and Wildlife Authority (“CBFWA"), a fisheries coalition, proposed the "optimal" flows first recommended in 1981 by agencies and tribes (140 kefs in the Lower Snake River and 300 kefs in the Lower Columbia River), the major difference being the "sliding scale” proposed in 1981 versus the fixed flows proposed in 1991. See CBFWA, the Biological and Technical Justification for the Flow Proposal of the Columbia Basin Fish and Wildlife Authority (Feb. 1991). [Petitioners’ Joint ER at 140-203J The Council and others took the position that such flows were virtually impossible because of the lack of water in the Water Budget to perform the augmentation. In response, the State of Idaho and other entities developed a strategy called mainstem reservoir drawdown as a means to meet the CBFWA optimal flows, often referred to as the “Idaho Drawdown Plan."

Drawdowns are releases of water that reduce the cross-section of rivers by lowering water levels in mainstem reservoirs. The result is increased velocity of the streamflow with less water than flow augmentation would require. Most salmon advocates supported the combination of the CBFWA flows and the Idaho drawdown plan, even though the resulting flows still fell short of *1382 those with which the salmon evolved. Though some advocates supported the "minimum” flows and sliding scale recommended in 1981 (220 kefs in the lower Columbia River), nearly all agreed that substantially greater flows than 85 kefs in the lower Snake River and 200 kefs in the lower Columbia River were needed.

. Particle travel time was recommended as a method to measure smolt travel time and survival because it lends itself well to models and monitoring. This method, however, is based upon a causal relationship between smolt travel time and survival, which is a hotly contested issue in this case; the Council and DSIs rely on studies significantly downplying the importance of flows, finding faster flows as merely an unremarkable factor in travel time compared to fish size, the level of smolt development, food availability, competition, and predation pressure. Interestingly, the Council has strongly resisted calling for significantly higher flows for a designated period to study their effects. At any rate, the point of particle travel time is to measure the velocity of the river, as opposed to the volume of water. Velocity can be achieved either by increasing flow rates (augmentation), or reducing the cross-section of rivers by lowering water levels in mainstem reservoirs (mainstem reservoir drawdown). It is also worth pointing out that Idaho ( Idaho Code § 42-1501 ), Oregon (Or.Rev. Stat. §§ 537.332, 537.336), and Washington (Wash.Rev.Code Ann. § 90.03.247) recognize a relationship between smolt travel time and survival in state statutes protecting minimum in-stream flows.

. NRIC and the Yakima Nation support the CBFWA optimal flows in conjunction with the Idaho Drawdown Plan. The Yakima Nation stresses that the Council failed to provide adequate flows to protect Snake River fall chinook, as compared to spring and summer runs; in other words, the Program overemphasizes flows on the lower Snake River from April 16 through June 15.

NRIC and the Yakima Nation also unite in asserting that the flow measure period must not end until much later in the summer if it is to effectively protect spring, summer, and fall runs. The timing of the measures is a critical issue because high power needs in the winter pressure the hydropower system to store water in its reservoirs during spring and summer. Thus, the greatest need for flows by salmon and energy occur at virtually the same time each year.

.As § 839b(h)(7) provides, the Council may reject a recommendation only if it is (1) inconsistent with the purposes of the Act, (2) inconsistent with standards established for the Program, or (3) less effective than an adopted recommendation in achieving protection, mitigation, and enhancement. 16 U.S.C. § 839b(h)(7); H.R.Rep. No. 96-976, pt. II, 96th Cong., 2d Sess., at 44. "The quantity or quality of the data should not serve as a basis for turning down any recommendation." H.R.Rep. No. 96-976, pt. I, 96th Cong., 2d Sess., at 56.

. That Congress meant precisely what it stated in § 839(b)(h)(7) is highlighted by the statements of Representative A1 Swift: "the simplest point to make for the record is that where the bill uses the word 'shall' it means 'shall', not 'may' ", 126 Cong.Rec. E5091-92 (1980).

. The parties focus on Volume II of the Strategy for Salmon, which contains the Program's measures. Volume I is an overview of the Program. Volume II discusses in greater detail the conclusions and recommendations of Volume I.

. We note that none of the documents referred to by the Council explain a statutory basis for the Council's rejection of recommendations of fishery managers.

. The Yakima Nation and Council engage in a protracted argument over whether § 706(2)(A) or (2)(D) of the APA applies here as the standard of review. We need not address their argument, however, because the Council's failure to explain the basis for its decision, as required by § 839b(h)(7), was not in accordance with the law under either standard.

. In support of its contention that the criteria are not substantive, the Council cites a remark by Representative Dingell:

126 Cong.Rec. H10683 (1980). Representative Dingell's remark that the Council has “a certain amount of judgment” in making determinations under § 839b(h)(6) does not detract from the high degree of reliance required on the fishery managers' recommendations, nor from authority construing these criteria as substantive. The Council’s construction of § 839b(h)(6) would instill tremendous discretion in the Counsel, effectively circumventing congressional intent that fish and wildlife be assured balanced consideration with other interests, and that the Council not become a ''superfish” entity.

Moreover, with respect to Representative Din-gell’s suggestion that determinations under § 839b(h)(6) were not intended to provide a legal basis for a challenge, the Act specifically provides otherwise. Section 839f(e)(l)(A) states in pertinent part:

(Emphasis added).

. That the Council must adopt sound biological objectives is certainly implied if not expressly required in § 839b(h)(6)(C).

. Particularly noteworthy is a report to NMFS by Hydrosphere Resource Consultants in which there is support for the fishery managers' contention that the CBFWA flows in combination with the Idaho Drawdown Plan are attainable without material harm to the hydropower system. Hydrosphere Resource Consultants, Water Supplies to Promote Juvenile Anadromous Fish Migration in the Snake River, January 1991 (NMFS Contract No. 50ABNF900105).

. For example, the Water Budget has often gone unimplemented because of the priority power managers have given power sales and the Council’s reticence to assert itself. See Parity V, 21 Envtl.L. at 688-90; Parity TV, 16 Envtl.L. at 494-501. Also, the record indicates that several deadlines since late 1992 have passed with little or no action.

. See American Textile Mfrs. Inst. v. Donovan, 452 U.S. 490, 508-12 , 101 S.Ct. 2478, 2490-92 , 69 L.Ed.2d 185 (1980) (a cost-benefit analysis requirement will not be read into a statute; Congress makes specific language when it wants an agency to engage in cost-benefit decision-making).