Greenpeace Action v. Franklin

OPINION

CYNTHIA HOLCOMB HALL, Circuit Judge:.

On June 26, 1991, Greenpeace Action (“Greenpeace”) filed a complaint against the Secretary of Commerce (the “Secretary”) alleging violations of the National Environmental Policy Act (“NEPA”), 42 U.S.C. §§ 4321-4370d, and section 7(a)(2) of the Endangered Species Act (“ESA”), 16 U.S.C. § 1536(a)(2), by the Secretary and the National Marine Fisheries Service (the “Service”). Greenpeace sought declaratory relief and an injunction against continued pollock fishing in the Gulf of Alaska until the Service complied with the law. On September 30, 1991, Greenpeace moved for summary judgment and a permanent injunction. The Service filed a cross motion for summary judgment. The district court denied Greenpeace’s motions and granted the Service’s motion in an order entered on October 11, 1991. Greenpeace appeals from that order. The district court had jurisdiction under 28 U.S.C. § 1331 (federal question), 5 U.S.C. § 702 (Administrative Procedure Act) and 16 U.S.C. § 1540(g) (Endangered Species Act, citizen suit). We have jurisdiction over this appeal pursuant to 28 U.S.C. § 1291.

I

FACTS

This case arose out of concern over the fate of the Steller sea lion, which inhabits the waters of the northern Pacific Ocean. Between 1960 and 1989, the Steller sea lion’s Alaskan population suffered a precipitous decline, resulting in its classification in 1990 as a “threatened species” under the ESA. See Listing of Steller Sea Lions as Threatened Under the Endangered Species Act, 55 Fed. Reg. 49,204 (Dep’t of Comm.1990) (final rule). Though the Steller sea lion’s decline has abated over the last three years, it has not stopped. The harvesting of pollock, a groundfish that comprises about half of the Steller sea lion’s diet, has been cited as a likely factor in the Steller sea lion’s decline.

Greenpeace contends that studies — including the Service’s own data — demonstrate that pollock fishing is the “leading factor” in the sea lion’s decline. Greenpeace emphasizes that the primary danger is not the depletion of the over-all biomass of pollock in the Gulf, but rather localized depletion; the Steller sea lion’s proximity to its food source is crucial, and fisheries¹ and sea lions often compete for the same stock of pollock. The Secretary’s final rule listing the Steller sea lion as a threatened species takes a less definitive position on the effects of pollock fishing:

Some data show a high negative correlation between the amount of walleye pollock caught and sea lion abundance trends in the eastern Aleutians and central Gulf of Alaska. It is possible that a reduction in availability of pollock, the most important prey species in most areas, is a contributing factor in the decline in the number of Steller sea lions in western and central Alaska.

55 Fed.Reg. at 49,208 (emphasis added).

Pursuant to section 302(h)(1) of the Fisheries Conservation Management Act (“Magnu-son Act”), 16 U.S.C. § 1852(h)(1), the North Pacific Fishery Management Council (“the Council”) issued its Fishery Management Plan (“the Plan”) and an environmental impact statement (EIS) for the Gulf of Alaska Groundfish Fishery in 1978. An amendment to the Plan established a procedure for setting annual harvest levels for various species. Every September the Council makes public a preliminary Stock Assessment and Fishery Evaluation Report, as well as preliminary specifications for the acceptable biological catch (ABC) and the total allowable catch (TAC). The ABC is a measure of the size of the catch that the ecosystem can sustain. The TAC is the total tonnage of fish that fishermen may retain in a particular year, Preliminary TACs are replaced by final TACs when they are approved by the Secretary.

in September 1990, the Council made its ■preliminary recommendations for 1991. It recommended a pollock TAC of 73,400 metric tons (mt), the same as the 1990 level. In December 1990, the Council’s assessment of the 1991 fishing stock was released, and based on that report, the Council revised its proposed TAC to 130,000 mt, a 41 percent increase over the 1990 level. Greenpeace sent a letter to the Secretary objecting to the proposed TAC, alleging that the 41 percent increase over the 1990 level would violate ESA. Without the preparation of an EIS, Greenpeace alleged, implementing the increase would also violate NEPA. The letter charged that the Council had not adequately considered the effect of its plan on the Stel-ler sea lion and recommended that the 1991 TAC remain at the 1990 level. It attached a report by the Aquatic Resources Conservation Group voicing concern over the modell-ing technique used to arrive at the proposed ABC and TAC and recommending an ABC of 103,400 mt and a TAC of 71,010 mt.

The Council’s proposed TAC was never implemented. The 1991 fishery opened under an interim TAC, equal to one quarter of the 1990 quota. That limit was exceeded by mid-February and the fishery was closed. In March 1991, the Secretary deferred approval of the recommended TAC for further evaluation of its impact on the Steller sea lion’s food supply. At that time he entered into a consultation with the Service, pursuant to section 7(a)(2) of the ESA.²

In the course of the section 7 consultation, the Service collected and analyzed new data, and in June issued a biological opinion recommending a TAC of 103,400 mt, allocated temporally and geographically to prevent local depletions of pollock.³ It also recommended the implementation of a ten nautical mile (nm) no-trawl zone around the Steller sea lion rookeries. The opinion concluded that if implemented under the proposed conditions, the 1991 TAC was not likely to jeopardize the Steller sea lion. It stated that the effects of pollock harvesting on the Steller sea lion’s ability to obtain food were uncertain, but that the proposed TAC left available a stock of pollock sufficient both to reproduce and to meet the Steller sea lion’s annual food needs. It also determined that various measures it proposed were adequate to prevent temporary local depletions that may affect Steller sea lion feeding success. In June, the Service produced an environmental assessment, pursuant to 40 C.F.R. § 1501.4(a) — (c) (1991), analyzing the impact of the TAC on the environment and concluding that an EIS was unnecessary.⁴

The Secretary considered the recommendations of the Service, as well as those of the Steller Sea Lion Recovery Team, and ultimately set both the ABC and TAC for 1991 at 103,400 mt and adopted the emergency measures proposed in the June biological opinion. These measures went into effect and the fishery was reopened on June 13, 1991. See Groundfish of the Gulf of Alaska; Groundfish of the Bering Sea and Aleutian Islands Area, 56 Fed.Reg. 28,112 (Dep’t of Comm.1991) (notice of initial harvest specifications for pollock).⁵

Catches of groundfish unexpectedly jumped at the end of the third quarter of 1991, and although the Secretary closed the fishery early, the third quarter harvest was in excess of the quarterly allowance. In September, before reopening the fishery for the fourth quarter, the Service prepared a new environmental assessment. The assessment noted that the third-quarter overrun was small compared to the total pollock biomass and that the pollock biomass was at its highest levels since the early 1980s. The Service therefore concluded that the Steller sea lion’s decline could not be attributed to a decline in pollock abundance and that none of the management options being considered for the fourth quarter would significantly affect the quality of the human environment. Under these circumstances, the Service determined that NEPA did not require that an EIS be prepared before the fourth quarter fishery could be opened.

The Service also prepared a section 7 biological opinion for the fourth quarter, reiterating its position that the effect of commercial pollock fishing on the Steller sea lion was uncertain. It concluded that given this uncertainty, the small size of the proposed fourth quarter harvest (3 percent of the estimated exploitable Gulf biomass), and the emergency management safeguards, the fourth quarter harvest “was not likely to jeopardize the continued existence of the Steller sea lion.” When the fourth quarter fishery closed on October 25, 1991, the total harvest for 1991 was 6,000 metric tons below the allowed level.

On January 28, 1991, before the Secretary issued his final rule, Greenpeace sent a notice of intent to file a citizen’s suit under the ESA, objecting to the increase in the TAC proposed by the Council in December, 1990. The notice asserted that the proposed increase was likely to jeopardize the Steller sea lion and if it was not disallowed, Greenpeace would sue.

On June 26, 1991, a week after the adoption of the final rule, Greenpeace filed a complaint in federal district court seeking declaratory and injunctive relief against the Secretary. The Complaint charged that the Secretary’s implementation of the 1991 TAC violated the ESA because it was done without preparation of an adequate biological opinion and without consideration of the best available scientific and commercial data concerning the status of the fishery and its potential impact on the Steller sea lion. It further alleged that implementation of the TAC without preparation of an EIS or an adequate environmental assessment violated NEPA.

With respect to the alleged violations of the ESA, Greenpeace sought an order directing the Secretary to comply with the Act by ensuring that the continued harvest of pollock “is not likely to jeopardize the continued existence and recovery of the threatened Steller sea lion” and by preparing “a legally and scientifically adequate biological opinion ... concerning the risks posed by pollock fishing” to the Steller sea Hon. With respect to the alleged violations of NEPA, Greenpeace sought an order directing the Secretary to prepare “a legally adequate [environmental impact statement] or [environmental assessment] examining and disclosing the significant environmental effects of implementing the 1991 TAC.” Greenpeace also requested an injunction prohibiting further pollock harvesting until the Secretary com-pfied with these duties.

The parties filed cross motions for summary judgment, and on October 10, 1991 the district court granted summary judgment for the defendants. The court held that nothing in the record supported a claim that the Secretary had violated the ESA. The record showed that his decision not to undertake or consider additional studies before issuing his finding of “no jeopardy” was not arbitrary or capricious, and that the finding “was based on relevant factors and demonstrated no clear error of judgment.” It also concluded that nothing in the record supported a charge that the Secretary had violated NEPA. The record demonstrated that the Secretary had adequately assessed the impact of his action on the environment and had taken protective measures, and that his decision not to undertake an EIS was not arbitrary or capricious. The district court denied Greenpeace’s motion for a stay of the opening of the fourth quarter fishery pending appeal. On October 16, 1991, an emergency motions panel of this Court also denied Greenpeace’s motion for a stay.

II

MOOTNESS

The 1991 fishing season has ended and the 1991 TAC has expired. Consequently, Appellees argue that the issues presented by this appeal are moot. They maintain that because the 1992 amendments to the Fishery Management Plan' and the 1992 TAC have been determined based on an entirely new administrative record, this Court cannot afford Greenpeace the declaratory or injunctive rehef it seeks. They also argue that the issues presented are not “capable of repetition yet evading review.” Appellees’ burden of demonstrating the mootness of this case is a heavy one. See Headwaters, Inc. v. Bureau of Land Management, 893 F.2d 1012, 1015 (9th Cir.1989).

Although we cannot grant Greenpeace effective rehef on its claim that the 1991 TAC was unlawfully approved, we conclude that this action “is one of those extraordinary cases in which the complained of activity may be repeated and yet evade review.” Alaska Fish & Wildlife Fed’n v. Dunkle, 829 F.2d 933, 939 (9th Cir.1987), cert. denied, 485 U.S. 988, 108 S.Ct. 1290, 99 L.Ed.2d 501 (1988). Government actions fall within this category if (1) the duration of the challenged action is too short to allow full litigation before it ceases, and (2) there is a reasonable expectation that the plaintiffs will be subjected to it again. Id.

The regulation challenged was in effect for less than one year, making it difficult to obtain effective judicial review. Id,.; Maryland People’s Counsel v. FERC, 761 F.2d 768, 773 (D.C.Cir.1985). The major issue— whether the Secretary has adequately examined the effects of pollock fishing on the Steller sea lions — is likely to recur in future years. In fact, the Secretary has relied on the same biological opinion in support of the 1992 TAC, and has declined to prepare an EIS. See Groundfish of the Gulf of Alaska, Groundfish Fishery of the Bering Sea and Aleutian Islands Area, 57 Fed.Reg. 2,683 (Dep’t of Comm.1992) (final rule). Finally, there is a continuing public interest in determining the standards governing the Secretary’s decision to authorize a certain level of pollock fishing in the Gulf of Alaska. See Joint Bd. of Control v. United States, 832 F.2d 1127, 1130 (9th Cir.1987), cert. denied, 486 U.S. 1007, 108 S.Ct. 1732, 100 L.Ed.2d 196 (1988); Alaska Fish & Wildlife Fed’n, 829 F.2d at 939.

Appellees argue that this action need not have evaded review because Greenpeace could have preserved the issue by obtaining a preliminary injunction against all pollock fishing in the Gulf of Alaska. See Headwaters, 893 F.2d at 1016. We disagree. Although the harvest of pollock could have been enjoined, the expiration of the 1991 TAC could not. No injunction could have preserved this challenge to a short-term rule. We conclude that this action is not moot.

Ill

NEPA VIOLATIONS

The district court upheld the Service’s decision not to prepare an EIS for the fourth quarter fishery, finding that the record did not support the need for such action. It found that although commercial fishing may have adversely affected the Steller sea lion in the past, the dangers posed by fishing had been diminished in the fourth quarter by the implementation of the. mitigation measures. The court also found there to be no significant controversy regarding the effects of the fourth quarter fishery that would warrant an impact statement. Finally, the court found that the Secretary adequately addressed the effects of the 1991 TAC on the Steller sea lion and other species, and adequately explained how the mitigation measures would protect the sea lion.

A

Standard of Review

In this Circuit, we have heretofore reviewed an agency’s decision not to prepare an initial EIS under a “reasonableness” standard. See, e.g., Sierra Club v. United States Forest Service, 843 F.2d 1190, 1192 (9th Cir.1988). The Supreme Court’s decision in Marsh v. Oregon Natural Resources Council, 490 U.S. 360, 109 S.Ct. 1851, 104 L.Ed.2d 377 (1989), requires us to reexamine this practice. Although that case reviewed an agency’s decision not to supplement an existing EIS, distinguishing the particular decision at issue as one that involved “a factual dispute” about the impact of new scientific information on prior scientific conclusions, the Court’s reasoning appears to be equally apposite to cases that present challenges to an agency’s decision not to prepare an initial EIS.

When an agency has attempted to comply with NEPA’s dictates and an aggrieved party challenges its determination that the proposed action does not warrant an EIS, the case often presents “a factual dispute the resolution of which implicates substantial agency expertise” and “ ‘requires a high level of technical expertise.’ ” Marsh, 490 U.S. at 376, 377, 109 S.Ct. at 1860, 1861 (quotations omitted). In such a case, it would contravene Marsh to review the agency’s determination under a standard other than “arbitrary and capricious.”⁶

Our cases reviewing challenges to an agency’s decision not to prepare an initial EIS after Marsh have thus far continued to apply the reasonableness standard of review. See Seattle Comm. Council Fed’n v. FAA, 961 F.2d 829, 832 (9th Cir.1992); Northwest Envtl. Defense Ctr. v. Bremen, 958 F.2d 930, 936 (9th Cir.1992); LaFlamme v. FERC, 945 F.2d 1124, 1128-29 (9th Cir.1991). Although Seattle Community Council Federation and LaFlamme did not mention Marsh or its implications,⁷ Northwest Environmental Defense Center remarked on the possibility that Marsh’s arbitrary and capricious standard might apply. That case found it unnecessary to decide the question, however, since the court’s review under a reasonableness standard produced the same result. 958 F.2d at 936 n. 5.

In this ease, Greenpeace complains of the lack of “a legally adequate” EIS or environmental assessment. But Greenpeace’s challenges to the Service’s failure to prepare an EIS turn on factual determinations, not legal principles. The challenges do not concern whether or not NEPA applies, because the Service has assumed that it does and has proceeded under NEPA’s requirements, preparing two environmental assessments that each resulted in a finding of no significant impact. The questions Greenpeace now raises are as follows: (1) whether the fishery management measures implemented to mitigate possible adverse effects were adequate to render such presumed effects not significant; (2) whether there was sufficient public controversy at the time the Service announced the TAC to compel the conclusion that the TAC’s effect was significant; (3) whether the management measures would be effective to prevent further declines among the Steller sea lion; and (4) whether the Service’s assumption that the TAC’s effect on harbor seals would be similar to its effect on the sea lion was permissible. Each requires resolution of factual disputes between the Service’s scientific conclusions and those of Greenpeace’s experts.

Thus, we are squarely faced with a case that falls within the Supreme Court’s reasoning in Marsh. We must decide, as a threshold matter, whether this case also falls within its rule.⁸ We conclude that it does. We see no reason to impose a different standard of review on an agency’s decision not to prepare an initial EIS when the agency itself applied the same ‘rule of reason’ in making that determination, see 490 U.S. at 373-74, 109 S.Ct. at 1859, and the dispute is of the same factbound character as that presented in Marsh, see id. at 376-77, 109 S.Ct. at 1860-61. Although the difference between the two standards may not be of “great pragmatic consequence,” id. at 377 n. 23, 109 S.Ct. at 1861 n. 23, we conclude that when a litigant challenges an agency determination on grounds that, in essence, allege that the agency’s “expert review ... was incomplete, inconclusive, or inaccurate,” id. at 376-77, 109 S.Ct. at 1861, the greater degree of deference expressed by the arbitrary and capricious standard is appropriate.

We note that our sister circuits which had previously followed a standard of reasonableness have adopted the Supreme Court’s reasoning in Marsh in reviewing challenges to initial EIS determinations reached following preparation of an environmental assessment. See Los Ranchos, 956 F.2d at 972-73; Sabine River Authority v. United States Dep’t of Interior, 951 F.2d 669, 678-79 & n. 2 (5th Cir.1992); Goos, 911 F.2d at 1291-92; North Buckhead Civic Ass’n v. Skinner, 903 F.2d 1533, 1538 (11th Cir.1990). We join them in holding that review of a factbound challenge to an agency’s determination not to prepare an initial EIS, made after considerable agency review of a project’s environmental impact, is governed by the arbitrary and capricious standard.

This standard requires us to ensure that an agency has taken the requisite “hard look” at the environmental consequences of its proposed action, carefully reviewing the record to ascertain whether the agency decision is “founded on a reasoned evaluation ‘of the relevant factors.’” Marsh, 490 U.S. at 373-74, 378, 109 S.Ct. at 1859 (quoting Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 824, 28 L.Ed.2d 136 (1971)). This inquiry into the facts is to be searching and careful. Overton Park, 401 U.S. at 416, 91 S.Ct. at 823. But “[w]hen specialists express conflicting views, an agency must have discretion to rely on the reasonable opinions of its own qualified experts even if, as an original matter, a court might find contrary views more persuasive.” Marsh, 490 U.S. at 378, 109 S.Ct. at 1861. Once we are satisfied that an agency’s exercise of discretion is truly informed, “we must defer to ‘th[at] informed discretion.’ ” Id. at 377, 109 S.Ct. at 1861 (quoting Kleppe v. Sierra Club, 427 U.S. 390, 412, 96 S.Ct. 2718, 2731, 49 L.Ed.2d 576 (1976)). Greenpeace suggests four reasons why we should find the actions it challenges to have been arbitrary and capricious.

B

First, Greenpeace argues that the Service has admitted pollock fishing may have significant adverse effects on the Steller sea lion. An agency must prepare an EIS if “substantial questions are raised as to whether a project ... may cause significant degradation of some human environmental factor.” LaFlamme v. FERC, 852 F.2d 389, 397 (9th Cir.1988) (internal quotations omitted). “The plaintiff need not show that significant effects will in fact occur, but if the plaintiff raises substantial questions whether a project may have a significant effect, an EIS must be prepared.” Id.; see also Founder tion for N. Am. Wild Sheep v. USDA, 681 F.2d 1172, 1177-78 (9th Cir.1982) (standard is whether “the plaintiff has alleged facts which, if true, show that the proposed project may significantly degrade some human environmental factor”) (internal quotations omitted).

The Service’s environmental assessment of the fourth quarter fishery recognizes that commercial fisheries “have been identified in the literature as possible factors in the population decline,” but suggests that “Steller sea lion declines can not [sic] be attributed to declines in gulf-unde pollock abundance.” (emphasis added). The environmental assessment does, however, “speculate” that “[IJocalized depletions of pollock and other Steller sea lion prey may have occurred due to [the] spatial and temporal concentration of fishing effort that could have contributed to the decline of the Steller sea lion.”

Despite its uncertainty, the Service operated from the assumption that localized deple-tions might be having an adverse impact on the Steller sea lion, and concluded it was best to take precautions. It therefore recommended that the fourth quarter fishery be subject to the management measures described in the June biological opinion and environmental assessment. Although the effectiveness of these measures was unknown, the Service concluded that if the measures were implemented, opening the fourth quarter fishery under the 1991 TAC would not pose a great enough threat to the Steller sea lion to warrant an impact statement. The issue, then, is not whether the uncertainty surrounding the effect of pollock depletions on the Steller sea lion mandated the preparation of an EIS. It is whether, assuming that pollock depletion has had an adverse impact, the 1991 TAC in combination with the mitigation measures formed such an adequate buffer against that depletion that any possible depletion would be too minor to warrant an impact statement.

Greenpeace presented to the district court affidavits raising questions about the Service’s analyses and conclusions regarding the impact of the 1991 TAC and the effectiveness of the management measures. These affidavits do not themselves demonstrate that the fourth quarter fishery, as implemented under the Secretary’s management measures, posed a danger to the Steller sea lion. Rather, they criticize the Service for failing to assess the effects of the TAC on the availability (as opposed to abundance) of pollock or to refute the evidence of a causal connection between pollock fishing and the decline of the Steller sea lion. They also question the Service’s methodology, which they claim ignored hydroacoustic survey estimates measuring the biomass of pollock, and instead employed less rehable methods for estimating biomass.

The Service contends that its decisions were based on a review of adequate scientific data and that the criticisms offered by Greenpeace’s affiants merely represent a difference of scientific opinion. We concur in the Service’s assessment of Greenpeace’s complaints. The Service’s conclusions are clearly based on substantial — though not dis-positive — scientific data, and not on mere speculation. Moreover, the June 5,1991 biological opinion reveals that the Service did indeed consider the hydroacoustic survey estimates.

In Friends of Endangered Species, Inc. v. Jantzen, 760 F.2d 976 (9th Cir.1985), we upheld a decision by the Fish and Wildlife Service not to prepare an impact statement in connection with its issuance of a permit allowing the taking of Mission Blue butterflies. In response to Friends’ argument that the agency had based its decision on faulty data, we held that “NEPA does not require that we decide whether an [environmental assessment] is based on the best scientific methodology available, nor does NEPA require us to resolve disagreements among various scientists as to methodology.” Id. at 986.

While Greenpeace cites several instances in which this Circuit has set aside an agency decision not to prepare an EIS, only in Wild Sheep did the court mandate reconsideration after the agency had prepared an environmental assessment.⁹ In Wild Sheep, the Forest Service’s failure to address “certain crucial factors, consideration of which was essential to a truly informed decision whether or not to prepare an EIS,” rendered its conclusion that no statement was necessary unreasonable. 681 F.2d at 1178. Unlike Wild Sheep, the record in this case reveals no complete failure to consider crucial factors. Contrary to Greenpeace’s assertions, its affidavits do not set forth facts demonstrating that the fishery harvest specifications and mitigation measures may have “disastrous effects” on the Steller sea lion. They demonstrate only that there is uncertainty as to how pollack fishing affects the sea lion, which is undisputed, and that the Service did not prove harm to the sea lion was impossible. To set aside the Service’s determination in this ease would require us to decide that the views of Greenpeace’s experts have more merit than those of the Service’s experts, a position we are unqualified to take.¹⁰

C

Next Greenpeace argues that substantial public controversy over the potential adverse affects of pollock fishing mandate preparation of an EIS. The existence of a public controversy over the effect of an agency action is one factor in determining whether the agency should prepare such a statement. See 40 C.F.R. § 1508.27(b)(4) (1991); LaFlamme, 852 F.2d at 400-01; Jones v. Gordon, 792 F.2d 821, 828-29 (9th Cir.1986); Wild Sheep, 681 F.2d at 1182. A federal action is controversial if “ ‘a substantial dispute exists as to [its] size, nature, or effect.’ ” Wild Sheep, 681 F.2d at 1182 (quoting Rucker v. Willis, 484 F.2d 158, 162 (4th Cir.1973)).

Greenpeace argues that the affidavits of scientists it presented to the district court demonstrate “widespread uncertainty and dispute throughout the scientific community concerning the environmental significance of pollock fishing” in the Gulf of Alaska.¹¹ We agree with the district court that Greenpeace may not establish a scientific controversy post hoc, through the affidavits of its own scientists and the experts it has hired, when at the time of the Service’s action, there existed no substantial dispute that should have alerted the Service to the concerns that Greenpeace now raises.¹²

In Wild Sheep, we found a substantial dispute where the agency “received numerous responses from conservationists, biologists, and other knowledgable individuals, all highly critical of the [environmental assessment] and all disputing [its] conclusion....” 681 F.2d at 1182. But in Jantzen, we recognized that where “virtual agreement exists among local, state, and federal government officials, private parties, and local environmentalists,” the criticisms of the plaintiff and its experts are not sufficient to demonstrate the existence of a public controversy. 760 F.2d at 986-87.

Public comment was solicited in September, November, and December 1990, when the proposed 1991 TAC was first noticed and recommended for approval by the Secretary, and in June 1991, when the final 1991 TAC, as modified, was approved and mitigating management measures were implemented. Neither occasioned an outpouring of public protest such as that involved in Wild Sheep. See 681 F.2d at 1175 & n. 10, 1180, 1182. Greenpeace did protest, supplying a report by the Aquatic Resources Conservation Group. The ABC of 103,400 mt recommended in this report was the same as the ABC actually adopted. Although the report recommended different mitigation measures than were finally adopted, there was no dispute that mitigation measures should be imposed.¹³ Regardless of what Greenpeace’s experts now say, at the time the 1991 TAC and emergency measures were implemented, there was no substantial controversy about the effects of those measures. Furthermore, the Service was able to develop a consensus among parties who had objected to the original proposed TAC of 130,000 mt that the revised TAC and emergency measures were adequate to preserve the Steller sea lion’s food supply. The history of the implementation of the fourth quarter fishing measures is not one of substantial public controversy. Nor is there any merit to the contention that an EIS must be prepared whenever qualified experts disagree, as Greenpeace contends. If this type of disagreement were all that was necessary to mandate an EIS, the environmental assessment process would be meaningless. An agency’s careful evaluation of the impact of its proposed action, its collection and review of evidence, and its reasoned conclusions as to what the date reveals would be for naught if by simply filing suit and supplying an affidavit by a hired expert, predicated upon the same facts relied upon by the agency but reaching a different conclusion, a litigant could create a controversy necessitating an EIS.

D

The district court concluded that whatever harmful impact pollock fishing might have had on the Steller’s food supply in the past, the risk of further harm had been adequately diminished by the emergency management measures. Greenpeace argues that the Service’s implementation of those measures does not relieve it of the obligation to prepare an EIS. It maintains that because the Service did not know whether these measures would be effective, implementing them could not absolve the Service of its obligation to prepare a statement. Greenpeace’s affiants question the effectiveness of these measures, and they contend that the Service has not presented any data or analyses to suggest that these measures will do any good.

The Service does concede that the effectiveness of the mitigation measures is uncertain, but contends that this uncertainty is largely because the cause of the Steller sea lion’s decline is itself uncertain. In any ease, “so long as significant measures are undertaken to ‘mitigate the project’s effects,’ they need not completely compensate for adverse environmental impacts.” Jantzen, 760 F.2d at 987 (internal citation omitted). The Stel-ler Sea Lion Recovery Team, which had objected to the originally proposed TAC, stated that “[g]iven the limitations of our current understanding of sea lion feeding and nutritional requirements [the mitigation measures] represent a reasonable approach to the regulation of the Gulf pollock fishery.”

Moreover, the suggestion that the Service based these measures on pure speculation is false. The spatial and temporal allocation of the 1991 fishery was premised on an evaluation of the past distribution of pollock fishing and rates of Steller sea lion decline. And the 10 nm no-trawl zone was based on studies of Steller sea lion foraging patterns. In Wild Sheep, we found the agency’s decision not to prepare an impact statement unreasonable despite its adoption of mitigation measures, because the efficacy of these measures had been attacked in numerous responses to an original draft of the environmental assessment. 681 F.2d at 1180. As already noted, the efficacy of the mitigation measures, while under attack now, was not attacked when they were disclosed to the public in the July emergency rule. Although the Service was uncertain about the potential efficacy of the measures, the measures were carefully considered, based on evidence from scientific studies, and appear reasonably designed to protect the Steller sea lion.

E

Finally, Greenpeace argues that even if the Service adequately studied the effects of its actions on the Steller sea lion, it failed to take a “hard look” at the impact of pollock fishing on harbor seals and other species in the Gulf of Alaska. This charge has some merit. The September environmental assessment states: “[The Service] is proceeding under the assumption that due to the similarity in timing and locations of harbor seal and Steller sea lion declines the declines of the two species may have the same causative agent.' Thus the discussion that follows may apply equally well to both species.” One of Greenpeace’s affiants, Anne Hoover-Miller, states that there is evidence suggesting that commercial fishing has caused a decline in the population of the harbor seal and that the Service’s report is entirely inadequate to evaluate the effects of its measures on that animal. Indeed, she suggests, the 10 nm no-trawl zone around the sea lion rookeries may divert pollock fishing towards the feeding areas of the harbor seals.

There is no doubt the Service’s analysis of the effects of pollock fishing on the harbor seal is far less rigorous than the study of its effects on the Steller sea lion. But it is equally clear that there is little data available to analyze these effects. Hoover-Miller’s affidavit discussing the effects is itself highly speculative, and in no way undermines the Service’s assumption that its conclusions about the Steller sea lion could be applied to the harbor seal. At the same time, the environmental assessment indicated that the Service had begun a comprehensive assessment of the harbor seals to determine the status of the population. Given the fact that the harbor seal was not a threatened or endangered species, the lack of available data on the harbor seal, and the apparent reasonableness of the Service’s assumption that its analyses of the Steller sea lion could be applied to the harbor seal, we cannot regard its failure to take a “harder look” at the effects of the action on the harbor seal to be arbitrary and capricious.

F

In conclusion, we disagree with Greenpeace that the Service’s “Finding of No Significant Impact” was based on speculation and not supported by adequate evidence. The Service took a careful look at the effects of the 1991 fishery on the Steller sea lion. Although Greenpeace has demonstrated that some scientists dispute the Service’s analyses and conclusions, such a showing is not a sufficient basis for us to conclude that the Service’s action was arbitrary or capricious. If it were, agencies could only act upon achieving a degree of certainty that is ultimately illusory.

IV

ESA VIOLATIONS

Section 7 of the Endangered Species Act imposes on an agency a duty to “insure” that any action it takes is “not likely to jeopardize the continued existence” of a threatened species. Stop H-3 Ass’n v. Dole, 740 F.2d 1442, 1459 (9th Cir.1984) (quoting 16 U.S.C. § 1536(a)(2)), cert. denied, 471 U.S. 1108, 105 S.Ct. 2344, 85 L.Ed.2d 859 (1985). An action would “jeopardize” a species if it “reasonably would be expected ... to reduce appreciably the likelihood of both the survival and recovery” of the species by reducing its reproduction, numbers or distribution. 50 C.F.R. § 402.02 (1991). In fulfilling its duty, the agency “shall use the best scientific and commercial data available.” 16 U.S.C. § 1536(a)(2). When an agency relies on the analysis and opinion of experts and employs the best evidence available, the fact that the evidence is “weak,” and thus not dispositive, does not render the agency’s determination “arbitrary and capricious.” Stop H-3, 740 F.2d at 1460; see also Pyramid Lake Paiute Tribe of Indians v. United States Dep’t of Navy, 898 F.2d 1410, 1415 (9th Cir.1990).

Pursuant to 16 U.S.C. § 1536(a)(2), the Service held consultations with the Secretary to assess the effect of the 1991 fishery on the Steller sea lion. These consultations led to the preparation of biological opinions concluding that the 1991 TAC, implemented under the emergency management measures, was not likely to jeopardize the continued existence of the Steller sea lion. We will not upset an agency’s assessment of its obligations under section 7 unless we determine it to be arbitrary and capricious. Paiute Tribe, 898 F.2d at 1414. We agree with the district court that the Secretary acted within his discretion in choosing the data on which to rely and adequately supported his conclusion that the management measures would mitigate any potential harm to the Steller sea lion caused by pollock fishing.

Greenpeace claims the district court erred in finding the Service’s determination reasonable because the Service admitted that pollock fishing may adversely affect the Steller sea lion, because the efficacy of the Service’s proposed protection measures was too speculative, and because the Service placed the burden of any uncertainty on the Steller sea lion. These three arguments boil down to one: The Service has recognized the possibility that the fishery could pose a threat to the Steller sea lion and has failed to offer evidence proving that the 1991 TAC, imple-merited under the proposed management measures, will not place the animal in jeopardy. Because it acted despite uncertainty about the effects of its actions, the Service violated its section 7 obligations. The Service argues that its “admissions” reflect nothing more than a recognition of the uncertainty in the scientific community about the causal relationship between pollock fishing and the decline of the Steller sea lion, and that this uncertainty does not make its “no jeopardy” determination arbitrary and capricious.

The Service has clearly satisfied its ‘procedural obligations under section 7. The Service’s “no jeopardy” determinations and proposals for the 1991 TAC and fishery management measures followed consultations with the Secretary, as required by section 7, in May, June, and September of 1991. At the close of the consultations, the Service prepared biological opinions, see 50 C.F.R. 402.-14(g) (1991), describing the consultations and setting forth its recommendations.

We hold that the Service has fulfilled its substantive duties as well. Despite Greenpeace’s assertions to the contrary, the Service supported its conclusions with ample data and analysis. The June biological opinion indicates that the Service, the Alaska Fisheries Science Center, and the National Marine Mammal Laboratory “analyzed all the available data on the pollock fishery and Steller sea lions” in the Gulf of Alaska. The Service also sought the recommendations of the Steller Sea Lion Recovery Team. The opinion demonstrates that the Service evaluated the spatial and temporal distribution of commercial fishing across the Gulf of Alaska. It then addressed not only the total biomass of pollock in the Gulf and the effects of fishery removals on that biomass, but also the spatial and temporal distribution of pollock across the Gulf. And despite Greenpeace’s claims to the contrary, the Service did not ignore hydroacoustic surveys of pollock biomass, but considered and compared them to bottom trawl surveys. Finally, while the Service has repeatedly conceded that it was uncertain about the effectiveness of its management measures, it premised these measures on a reasonable evaluation of available data, not on pure speculation.

The biological opinions indicate that the Service, an expert agency, consulted with other teams of experts to consider all relevant factors pertaining to the effects of the Gulf fishery on the Steller sea lion. And they indicate that the Service did not ignore data, as Greenpeace suggests. The Service’s decision to go ahead with the 1991 fishery under the proposed restrictions, despite some uncertainty about the effects of commercial pollock fishing on the Steller sea lion, was not a clear error of judgment. In Stop H-3 we reviewed a “no jeopardy” determination by the Federal Highway Administration pertaining to the impact of construction of Interstate H-3 in Hawaii on the Oahu Creeper. The Highway Administration consulted with the Fish and Wildlife Service to determine the effect of the highway on the Creeper. The Wildlife Service’s biological opinion stated:

“In essence, we have very little data for providing an opinion, but feel it would be unreasonable to request [an additional] study which would be unlikely to provide definitive results. We must, therefore, assume the Oahu Creeper would be like most species in that a highway would not split a population.

Based on the available information, which we grant is weak, it is our opinion the proposed project is not likely to jeopardize the continued existence of the Oahu Creeper.”

740 F.2d at 1458 (quoting biological opinion). Despite the uncertainty of the data, we held that the Highway Administration had complied with section 7 because it had based its decision on the best available scientific data and had grounded its decision in a consideration of the relevant factors. Id. at 1460. Applying this standard, the Service has undeniably fulfilled its duties under the ESA.

For the foregoing reasons, the decision of the district court is AFFIRMED.

1 . The term "fishery" refers to: "(A) one or more stocks of fish which can be treated as a unit for purposes of conservation and management and which are identified on the basis of geographical, scientific, technical, recreational and economic characteristics; and (B) any fishing for such stocks." 16 U.S.C. § 1802(8) (1988).

2 . Section 7(a)(2) requires that agencies, in consultation with the Secretary, "insure that any action authorized, funded, or carried out by such agency ... is not likely to jeopardize the continued existence of any endangered species or threatened species or result in the destruction or adverse modification of habitat of such species.” 16 U.S.C. § 1536(a)(2) (1988).

3 . The opinion proposed allocating 100,000 mt of the TAC to the Westem/Central area of the Gulf, and 3,400 mt to the Eastern Area. It also proposed apportioning fishing in the Westem/Cen-tral area across four calendar quarters. 25,000 mt could be fished each quarter and carryover would be limited to 50 percent of the initial quarterly allowance. So, if less than 25,000 mt were fished in any quarter, no more than 12,500 mt of that could be made up in the next quarter.

4 . The NEPA requires agencies to "include in every recommendation or report on proposals for ... major Federal actions significantly affecting the quality of the human environment, a detailed statement by the responsible official on — (i) the environmental impact of the proposed action." 42 U.S.C. § 4332(C). If a proposed action would normally require an environmental impact statement, an agency must prepare an environmental assessment to determine whether or not an environmental impact statement is necessary. 40 C.F.R. § 1501.4(a)-(c) (1991). If, based on the environmental assessment, the agency determines that the proposed action will not significantly affect the environment, it need not prepare an EIS and instead may issue a Finding of No Significant Impact (FONSI). 40 C.F.R. § 1508.13 (1991).

5 .In addition to allocating the harvest across the Western/Central and Eastern regions of the Gulf, the Secretary divided the Westem/Central region into two areas and allocated 50,000 mt to each area. 56 Fed.Reg. at 28,112, 28,114.

6 . Marsh appears to distinguish cases that turn on "the legal meaning of the term ‘significant’ or ... the predominantly legal question whether established and uncontested historical facts presented by the administrative record satisfy this standard.” 490 U.S. at 376, 109 S.Ct. at 1860. Accordingly, we express no opinion as to the standard by which to review such challenges.

7 . Of course, the agency actions challenged in these cases which were affirmed under the reasonableness standard would likewise have withstood scrutiny under the arbitrary and capricious standard mandated by Marsh.

8 . For purposes of our decision, we need not and do not decide whether a reasonableness standard still applies to threshold questions of NEPA applicability, such as the determination whether "major federal action” exists. See Goos v. ICC, 911 F.2d 1283, 1292 (8th Cir.1990); cf. Village of Los Ranchos de Albuquerque v. Marsh, 956 F.2d 970, 973 n. 4 (10th Cir.) (en banc) (expressly reserving the question), cert. denied, - U.S. -, 113 S.Ct. 59, 121 L.Ed.2d 27 (1992); see also Citizens Against Burlington, Inc. v. Busey, 938 F.2d 190, 196 (D.C.Cir.) (Thomas, J.) (reviewing the alternatives selected for discussion in EIS under a reasonableness standard), cert. denied, -U.S.-, 112 S.Ct. 616, 116 L.Ed.2d 638 (1991).

9 . Compare LaFlamme v. FERC, 852 F.2d 389, 399 (9th Cir.1988) (agency failed to prepare an environmental assessment, in violation of the NEPA, and based its decision on staff reports filed after license for challenged action was granted); Save the Yaak Comm. v. Block, 840 F.2d 714, 717-718 (9th Cir.1988) (environmental assessment "not prepared to examine the environmental impacts, but to 'evaluate various techniques of maintaining [] road’ " did not satisfy threshold requirements of 40 C.F.R. § 1508.9); Jones v. Gordon, 792 F.2d 821, 827-28 (9th Cir.1986) (agency maintaining its action fell within categorical exclusion for issuance of permits for scientific research prepared no environmental assessment, only a four sentence statement relying on fact that prior permits had been issued without an EIS); Steamboaters v. FERC, 759 F.2d 1382, 1392 (9th Cir.1985) (agency failed to prepare an environmental assessment).

10 . Wild Sheep was also decided under a reasonableness standard of review, and the court scrutinized the Forest Service's action with less deference than we must accord the Service's decision in the wake of Marsh.

11 . Greenpeacé argues that the very existence of uncertainty mandates the preparation of an impact statement and it cites to several cases holding that an impact statement must indicate whether there is uncertainty over the environmental effects of a proposed action. See Southern Oregon Citizens Against Toxic Sprays, Inc. v. Clark, 720 F.2d 1475, 1479 (9th Cir.1983), cert. denied, 469 U.S. 1028, 105 S.Ct. 446, 83 L.Ed.2d 372 (1984); Friends of the Earth v. Hall, 693 F.Supp. 904, 926 (W.D.Wash.1988). Because these cases deal not with whether an impact statement should be prepared, but with what information must be included in an impact statement after it has been judged necessary, they do not stand for the proposition that the existence of uncertainty mandates the preparation of an impact statement.

12 . Sierra Club v. U.S. Forest Service, 843 F.2d 1190 (9th Cir.1988), is not to the contrary. Except in certain circumstances, judicial review should be limited to the administrative record already in existence, rather than the new record made initially in the district court. Animal Defense Council v. Hodel, 840 F.2d 1432, 1436 (9th Cir.1988). We consider the affidavits to determine whether the Service considered all relevant factors in reaching its conclusion of no significant impact, and find that it did — although it did not reach the conclusion Greenpeace would have liked. See ante at 1334. In Sierra Club, the affidavits and trial testimony demonstrated that the agency had failed to consider all relevant factors and that there was a public controversy.

13 .Greenpeace sued for injunctive relief immediately following adoption of the TAC and the mitigation measures, rather than participating in further notice and comment procedures. We note that the Service limited opportunity for public input before recommending the emergency measures it adopted because of "the economic hardship to the fishing industry, and the need to implement emergency measures to protect Stel-ler sea lions simultaneously with the pollack fishery opening." 56 Fed.Reg. 28114. We express no opinion as to the propriety of Greenpeace’s choice to sue instead of submitting comments on the TAC and mitigation measures implemented in June, which could continue in effect for only three months and thus were amenable to modification. Nevertheless, we cannot characterize the agency’s action as "arbitrary” for failing to consider views that were never presented to it. Nor can we consider its failure to provide for earlier public comment arbitrary under these circumstances.