The Secretary of Commerce appeals the district court’s grant of summary judgment in favor of Earth Island.1 The district court held that the Secretary’s Initial Finding, triggering a change in the dolphin-safe label standard, was not in accordance with the law and constituted an abuse of discretion because the Secretary failed to (1) obtain and consider preliminary data from the eongressionally mandated stress studies and (2) apply the proper legal standard to the available scientific information. We affirm.
I. Factual and Procedural Background
This ease concerns congressional efforts to protect dolphins in the Eastern Tropical Pacific Ocean (“ETP”), which covers between five and seven million square miles and extends from the southern Californian to the South American coastlines. In the ETP, yellowfin tuna schools swim below dolphin groups, which are visible as they break the surface to breathe and leap into the air. Since 1959, fishermen in the ETP have pursued and chased the air breathing dolphin groups in order to catch the yel-lowfin tuna below. In this year-round process, referred to as “setting on dolphins,” the fishermen use explosives, chase boats, and helicopters to drive the dolphins and tuna into the center of purse seine nets. Floats and weights support the nets, which close like a purse around all trapped inside. From 1959 to 1972, millions of dolphins were killed in the nets. Public outrage over the ETP dolphin deaths led to a variety of legislation and ultimately a dolphin-safe labeling standard. A review of the legislation leading to the standard and the potential easing of that standard is necessary for an understanding of the present controversy.
In 1972, public outcry over the ETP dolphin deaths led Congress to enact the Marine Mammal Protection Act (“MMPA”), 16 U.S.C. §§ 1361 et seq, which had the reduction of ETP dolphin deaths as one of its goals. The MMPA directed the Secretary of the Treasury to “ban the importation of commercial fish or products from fish which have been caught with commercial fishing technology which *1061results in the incidental kill or incidental serious injury of ocean mammals in. excess of United States Standards.” Id. § 1371(a)(2). Having conducted research required by the MMPA, the National Marine Fisheries Service found that three ETP dolphin stocks were depleted-the coastal dolphin (42 Fed. Reg. 64, 548-60 (1977)), northeastern offshore spotted dolphin (58 Fed. Reg. 58, 285 (1993)), and the eastern spinner dolphin (58 Fed. Reg. 45,-006 (1993)).2 In 1984,1988, and 1992, Congress amended the MMPA to ban importation of tuna that failed to meet certain conditions regarding dolphin mortality. 16 U.S.C. §§ 1371(a)(2)(B) & 1411 et seq. In 1990, responding to consumer concern and American tuna processors labeling changes, Congress enacted the Dolphin Protection Consumer Information Act (“DPCIA”) under which tuna sold in the United States could not have a “dolphin safe” label if the tuna had been caught using purse seine nets intentionally deployed on or to encircle dolphins. 16 U.S.C. § 1385.
The American legislation and corresponding standards caused the loss of a large market for those countries that continued to set on ETP dolphins with purse seine nets. In 1992, the United States and other nations with purse seine fishing vessels in the ETP negotiated the International Dolphin Conservation Program (“La Jolla Agreement”), in which they “agreed to maintain dolphin kill levels at or below a ‘dolphin mortality limit’ assigned to each vessel, and to work toward reducing dolphin mortality to levels approaching zero.” Brower v. Daley, 93 F.Supp.2d 1071, 1074 (N.D.Cal.2000). Three years later, the La Jolla Agreement was formalized into a binding agreement, the Panama Declaration, under which the United States agreed to seek changes in United States laws pertaining to tuna embargoes, market access, and the dolphin safe label. S. 397, 105th Cong., 143 Congr. Rec. 379-01 (1997). The Panama Declaration sought legislation to change immediately the dolphin safe labeling standard and to allow tuna caught with purse seine nets to be labeled “dolphin safe” as long as no dolphins were observed to be killed or seriously injured during the set.
In part to implement the Panama Declaration and eliminate the ban on tuna imports from countries complying with the La Jolla Agreement, on August 15, 1997, Congress enacted the International Dolphin Conservation Program Act (“IDC-PA”), Pub. L. No. 105-42, 111 Stat. 1122. While there had been success in lowering dolphin mortality rates,3 Congress remained concerned that, even if dolphins were not killed or seriously injured in the purse seine nets, the physiological stress they suffered during the year-round chase and encirclement would impede the dolphin populations’ recovery. Accordingly, Congress rejected Panama Declaration language which sought an immediate change in the dolphin-safe label. H.R. Rep. No. 105-74 (pt. 1), reprinted in 1997 U.S.C.C.A.N. 1628. Congress included in the IDCPA a requirement of specified research projects directed toward assessing the prevalence and magnitude of fishery-induced stress in the ETP dolphins.
Through the IDCPA, Congress amended the DPCIA and required the Secretary to make Initial and Final Findings as to “whether the intentional deployment on or *1062encirclement of dolphins with purse seine nets is having a significant adverse impact on any depleted dolphin stock in the [ETP].” 16 U.S.C. § 1385(g)(1) & (g)(2). The Secretary was to make the Initial Finding on the basis of research conducted before March 1,1999, information obtained under the International Dolphin Conservation Program, and any other relevant information.4 The IDCPA also amended the MMPA to provide details of the required research:
(a) Required research
(1) In general. — The Secretary shall, in consultation with the Marine Mammal Commission and the Inter-American Tropical Tuna Commission, conduct a study of the effect of intentional encirclement (including chase) on dolphins and dolphin stocks incidentally taken in the course of purse seine fishing for yellowfin tuna in the [ETP]. The study, which shall commence on October 1, 1997, shall consist of abundance surveys as described in paragraph (2) and stress studies as described in paragraph (3), and shall address the question of whether such encirclement is having a significant adverse impact on any depleted dolphin stock in the [ETP],
(2) Population abundance surveys. — The abundance surveys under this subsection shall survey the abundance of such depleted stocks and shall be conducted during each of the calendar years 1998, 1999, and 2000.
(3) Stress studies. — The stress studies under this subsection shall include—
(A) a review of relevant stress-related research and a 3-year series of necropsy samples from dolphins obtained by commercial vessels;
(B) a 1-year review of relevant historical demographic and biological data related to dolphins and dolphin stocks referred to in paragraph (1); and
(C) an experiment involving the repeated chasing and capturing of dolphins by means of intentional encirclement.
§ 1414a(a). The Secretary delegated the research to the National Marine Fisheries Service. A change in the dolphin-safe labeling standard from the existing, more restrictive standard to the less protective standard depended on NMFS’s answer as to “whether the intentional deployment on or encirclement of dolphins with purse seine nets is having a significant adverse impact on any depleted dolphin stock in the [ETP].”
On March 25, 1999, NMFS submitted its report to Congress. NMFS found that the currently depleted populations of both *1063northeastern offshore spotted dolphins and eastern spinner dolphins5 were “not increasing at the rate expected based on the low rate of reported mortalities from the fishery since 1991 and the reproductive potential for these populations.” Report at vi. NMFS noted the difficulty in attributing the cause of the low or declining growth rates. Id. at viii. NMFS identified only one possible non-fishery related explanation for the slow or declining growth rates-a large scale environmental variability in the ocean habitat. However, NMFS discounted that explanation: “The review of environmental conditions did not disclose any large-scale oceanographic regime shifts during recent decades [and therefore] it is unlikely that [the dolphin populations’] failure to grow can be explained by large-scale environmental variability.” Id. at 23.
Turning to fishery-related explanations for the slow or declining population growth rates, NMFS identified stress, separation of cows and calves (with subsequent death of calves), as well as under-reporting of direct kills. NMFS noted that none of these potential explanations is necessarily exclusive of the others. Id. However, NMFS reported that it did not have data from any of the three mandated stress research projects. Id. at 4. Therefore, regarding stress concerns, NMFS included only the “physiological and behavioral stress in mammals” literature review. After reviewing the literature, NMFS concluded that:
Although this review of existing literature regarding stress in mammals cannot provide a quantitative or definitive answer to the question of whether the tuna fishery is causing stress to affected dolphin populations, the available information and evidence point to the likelihood that physiological stress is induced by fisheries activities. It is therefore plausible that stress resulting from chase and capture in the ETP tuna purse-seine fishery could have a population level effect on one or more dolphin stocks.
Id. at 5.
NMFS then reported that it did not have evidence to determine whether there was physiological evidence of stress in individual dolphins from the affected dolphin populations, and that the answer probably would be available at “the completion of the necropsy sampling program.” Id. at 14. Therefore, NMFS concluded that:
Given the information available from research vessel abundance estimates, tuna vessel abundance indices and observed fishery mortality, the quantitative answers to the question, “In the period since 1991, has there been for any depleted stock a failure to grow at the expected rate ...” are “yes” for both northeastern offshore spotted and eastern spinner dolphins. The probabilities associated with these answers are quite high, well above the suggested thresholds .... When considered with the qualitative answer from oceanographic studies (that it is unlikely that such a failure to grow can be explained by large-scale environmental variability) and the qualitative answer from the literature review (that it is plausible that stress resulting from chase and encirclement could have population level effects) the information suggests but by no means conclusively that the fishery has *1064been the source of significant adverse impact on these two populations.
Id. at 22 (emphasis added).6
On May 7,1999, the Secretary issued his Initial Finding (64 Fed. Reg. 24590-01) concluding “that there is insufficient evidence that chase and encirclement by the tuna purse seine fishery ‘is having a significant adverse impact’ on depleted dolphin stocks in the ETP.” Pursuant to the Initial Finding, the dolphin safe label standard changed effective February 2, 2000, to permit the use of “dolphin safe” labeling when purse seine nets are used, as long as no dolphins were killed or seriously injured during the particular set in which the tuna were caught.
Earth Island challenged the validity of the Secretary’s Initial Finding under the Administrative Procedures Act, 5 U.S.C. § 706 et seq. as arbitrary, capricious, an abuse of discretion, and contrary to law.7 Earth Island claimed that the Secretary failed to obtain and consider preliminary data from the congressionally mandated stress research projects, and failed to determine whether, on the basis of the best available scientific evidence, the use of purse seine nets is having a significant adverse impact on the depleted dolphin populations. Earth Island also asserted that the Secretary failed to apply the proper legal standard to the scientific information available, abused his discretion, and acted arbitrarily and capriciously by failing to find significant adverse impact given the best available evidence.
The district court found that the Secretary’s Initial Finding was not in accordance with the law and an abuse of discretion because the Secretary failed to (1) obtain and consider preliminary data from the congressionally mandated stress studies and (2) apply the proper legal standard to the available scientific information. The district court concluded that:
[I]t would flout the statutory scheme to permit the Secretary to fail to conduct mandated research, and then invoke a lack of evidence as a justification for removing a form of protection for a depleted species, particularly given that the evidence presently available to the Secretary is all suggestive of a significant adverse impact.
Brower, 93 F.Supp.2d at 1089. The district court granted plaintiffs motion for summary judgment and set aside the Secretary’s Initial Finding “until such time as the Secretary has an opportunity to consider preliminary results from the Congressionally mandated stress research studies.” Id. The Secretary appealed.
II. Jurisdiction and Standard of Review
We have jurisdiction over the district court’s judgment pursuant to 28 U.S.C. § 1291.
*1065We review the district court’s grant of summary judgment de novo. Sierra Club v. Babbitt, 65 F.3d 1502, 1507 (9th Cir.1995). In determining whether summary judgment was properly granted, we must “view the case from the same position as the district court” and apply the same standards. Nevada Land Action Ass’n v. U.S. Forest Serv., 8 F.3d 713, 716 (9th Cir.1993). We may affirm the district court grant of summary judgment on any basis supported in the record. Hells Canyon Alliance v. U.S. Forest Serv., 227 F.3d 1170, 1176 (9th Cir.2000). We review the district court’s factual findings for clear error. Russian River Watershed Prot. Comm. v. City of Santa Rosa, 142 F.3d 1136, 1140 (9th Cir.1998).
An agency’s interpretation or application of a statute is a question of law reviewed de novo. Partridge v. Reich, 141 F.3d 920, 923 (9th Cir.1998). In reviewing an agency’s statutory construction, we must reject those constructions that are contrary to clear congressional intent or frustrate the policy that Congress sought to implement. Chevron, U.S.A., Inc. v. NRDC, 467 U.S. 837, 843 n. 9, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984); see also NRDC v. EPA, 966 F.2d 1292, 1297 (9th Cir.1992) (“On questions of statutory construction, courts must carry out the unambiguously expressed intent of Congress.”).
Under the Administrative Procedures Act, we review the Secretary’s Initial Finding, in light of the administrative record, to determine if the Finding is “ ‘arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law’ or if the [Finding] failed to meet statutory, procedural, or constitutional requirements.” Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 413-14, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971); see also 5 U.S.C. § 706(2)(A)-(D). This inquiry, while narrow, must be searching and careful. Marsh v. Oregon Natural Res. Council, 490 U.S. 360, 378, 109 S.Ct. 1851, 104 L.Ed.2d 377 (1989). An agency’s action may be arbitrary and capricious if “the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.” Southwest Ctr. for Biological Diversity v. U.S. Forest Serv., 100 F.3d 1443, 1448 (9th Cir.1996).
III. Analysis
The starting point in statutory interpretation is the language of the statute itself. Group Life & Health Ins. Co. v. Royal Drug Co., 440 U.S. 205, 210, 99 S.Ct. 1067, 59 L.Ed.2d 261 (1979). In construing federal statutes, we presume that the ordinary meaning of the words chosen by Congress accurately express its legislative intent. Mills Music, Inc. v. Snyder, 469 U.S. 153, 164, 105 S.Ct. 638, 83 L.Ed.2d 556 (1985). “The meaning of statutory language, plain or not, depends on context.” Holloway v. United States, 526 U.S. 1, 7, 119 S.Ct. 966, 143 L.Ed.2d 1 (1999) (internal citations omitted). Context in this regard relates to “the design of the statute as a whole and to its object and policy.” Gozlon-Peretz v. United States, 498 U.S. 395, 407, 111 S.Ct. 840, 112 L.Ed.2d 919 (1991) (quoting Crandon v. United States, 494 U.S. 152, 158, 110 S.Ct. 997, 108 L.Ed.2d 132 (1990)). In determining a statutory provision’s meaning, we “may consider the purpose of the statute in its entirety, and whether the proposed interpretation would frustrate or advance that purpose.” United States v. Mohrbacher, 182 F.3d 1041, 1049 (9th Cir.1999) (citations omitted).
In urging this court to reverse the district court, the Secretary and amici stress *1066that this case involves international concerns and competing policies for protecting dolphins. That it does, but it is not our role to make policy decisions about ETP dolphin conservation. Such decisions are within Congress’s bailiwick, and both the Secretary and this court must defer to congressional intent as reflected in the IDCPA.
The Required Secretarial Findings
In the section entitled “Secretarial Findings”, the IDCPA mandates the Secretary’s course of action: “[T]he Secretary shall ... make an initial finding regarding whether the intentional deployment on or encirclement of dolphins with purse seine nets is having a significant adverse impact on any depleted dolphin stock in the [ETP].” 16 U.S.C. § 1385(g)(1) (emphasis added); see also id. § 1385(g)(2) (final finding). The common meaning of “whether” is “whichever of the two,” Merriam Webster’s Collegiate Dictionary (10th ed. 1993), and should be read to mean “whether or not”. Therefore, the IDCPA’s mandatory language required the Secretary to make an Initial Finding whether or not the purse seine net fishery was having a significant adverse impact on any depleted ETP dolphin stock.
However, urging us to read § 1385(h)8 as establishing a default to the less protective label standard, the Secretary argues that he is not required to find affirmatively that there is not a significant adverse impact. The Secretary contends that if he fails to find evidence of significant adverse impact, then the less protective dolphin labeling standard will go into effect. We reject this interpretation for several reasons.
First, the Secretary’s interpretation is at odds with the statute’s structure. Under § 1385(g), “Secretarial Findings,” the IDCPA required the Secretary to make a finding whether or not the fishery-related activities were adversely impacting the dolphins. This finding requires a “yes” or “no” answer: “Yes,” there was a significant adverse impact or “no,” there was no significant adverse impact. Section 1385(h), “Certification by Captain and Observer,” does not determine the parameters or scope of the Initial Finding. Nor does it change the Secretary’s burden of proof as written in § 1385(g), “Secretarial Findings”.
Second, Congress rejected Panama Declaration language which sought an immedi*1067ate change in the dolphin safe label. It would be inconsistent with that history and congressional concern to interpret the statute as establishing the new less-protective labeling standard as the default.9 Just as we may not substitute our judgment for the agency’s, Overton Park, 401 U.S. at 416, 91 S.Ct. 814, the agency may not ignore Congress, Conner v. Burford, 848 F.2d 1441, 1454 (9th Cir.1988).
Finally, this default construction should be avoided because it would lead to absurd results. Griffin v. Oceanic Contractors, Inc., 458 U.S. 564, 575, 102 S.Ct. 3245, 73 L.Ed.2d 973 (1982) (Statutory interpretations “which would produce absurd results are to be avoided if alternative interpretations consistent with the legislative purpose are available.”). Such a default construction would render the required stress studies irrelevant. The use of a default does not encourage active, aggressive fact-finding and research or conclusive answers. For example, the Secretary could deliberately drag his feet in commencing studies or while conducting studies and then conclude there was insufficient evidence to warrant finding a significant adverse impact on the ETP dolphin stocks. Similarly, the Secretary could limit the studies’ breadth and then discover that there was insufficient evidence to warrant finding a significant adverse impact on the ETP dolphin stocks. Following the Secretary’s interpretation, under the above scenarios, the dolphin labeling standard would default to the less protective standard for the Initial Finding, because the Secretary failed to find evidence of significant adverse impact, and the less protective standard would be in place during the interim period between the Initial and Final Findings. A practical consequence of this default would be increased pressure on the Secretary to keep the same dolphin labeling standard, that is, to have the same result for the Final Finding.
We reject the Secretary’s default construction and hold, as Congress required, that the Secretary must affirmatively find whether or not there is a significant adverse impact before the dolphin safe labeling standards can be relaxed.
Commencement of the Stress Studies
The deference accorded an agency’s scientific or technical expertise is not unlimited. Defenders of Wildlife v. Babbitt, 958 F.Supp. 670, 679 (D.D.C.1997). The presumption of agency expertise can be rebutted when its decisions, while relying on scientific expertise, are not reasoned. Id. We defer to agency expertise on methodology issues, “unless the agency has completely failed to address some factor consideration of which was essential to [making an] informed decision.” Inland Empire Pub. Lands Council v. Schultz, 992 F.2d 977, 981 (9th Cir.1993) (internal citations omitted).
The IDCPA statutory language clearly and unambiguously required the Secretary to commence a study, consisting of abundance surveys and stress studies, on October 1, 1997.10 16 U.S.C. § 1414a(a). The *1068IDCPA provided explicit, detailed instructions for the stress studies. Additionally, the IDCPA required that:
Between March 1, 1999, and March 31, 1999, the Secretary shall, on the basis of the research conducted before March 1, 1999, under section 1414a of this title, information obtained under the International Dolphin Conservation Program, and any other relevant information, make an initial finding regarding whether the intentional deployment on or encirclement of dolphins with purse seine nets is having a significant adverse impact on any depleted dolphin stock in the [ETP].
16 U.S.C. § 1385(g)(1) (emphasis added).
Trying to support his Initial Finding, which did not incorporate any stress study evidence, the Secretary asserts that the legislative history indicates that Initial Finding would be made on the basis of limited evidence. This contention is without merit. Issuing the Initial Finding using limited evidence is very different from issuing the Initial Finding without obtaining or considering any data from stress studies -that are specifically required. Given the mandatory, detailed, and explicit language regarding the stress studies, the Secretary was required to conduct stress tests and consider some preliminary results prior to issuing the Initial Finding.11 See, e.g., NRDC v. EPA, 966 F.2d 1292, 1300 (9th Cir.1992) (“This court must uphold adherence to the law, and cannot condone the failure of an executive agency to conform to express statutory requirements.”).
The Stress Studies
The Secretary contends that the district court erroneously found that NMFS failed to comport with the spirit and the letter of the law by unreasonably delaying the stress studies and failing to collect, analyze, and report on any stress study data. We reject this contention.
In determining whether NMFS unreasonably delayed the stress studies, 5 U.S.C. § 706(1), we balance the following “TRAC” factors:
(1) the time agencies take to make decisions must be governed by a “rule of reason”[;] (2) where Congress has provided a timetable or other indication of the speed with which it expects the agency to proceed in the enabling statute, that statutory scheme may supply content for this rule of reason [;] (3) delays that might be reasonable in the sphere of economic regulation are less tolerable when human health and welfare are at stake [;] (4) the court should consider the effect of expediting delayed action on agency activities of a higher or competing priority[;] (5) the court should also take into account the nature and extent of the interests prejudiced by the delay[;] and (6) the court need not “find any impropriety lurking behind agency lassitude in order to hold that agency action is unreasonably delayed.”
Independence Mining Co. v. Babbitt, 105 F.3d 502, 507 n. 7 (9th Cir.1997) (quoting *1069Telecommunications Research & Action v. FCC (TRAC), 750 F.2d 70, 80 (D.C.Cir.1984) (citations omitted)).
The IDCPA required a literature review of relevant stress-related research, a three-year series of dolphin necropsy samples obtained by commercial vessels, a one-year review of relevant historical demographic and biological data, and an experiment involving the repeated chasing and capturing of dolphins by means of intentional encirclement. 16 U.S.C. § 1414a(a). At the time of the Initial Finding, NMFS had commenced and completed only the literature review. NMFS, which had not conducted any research on the three required studies, provided brief overviews on and updates of the three studies (necropsy study, one-year historical study, and chase and recapture project) in the report.
(1)Necropsy Study: The necropsy study “involves placing trained necropsy technicians onboard commercial tuna vessels to collect tissue samples from dolphins killed in the ETP fishery.” Report at 5. “Collection of these tissue samples requires that technicians be placed onboard non-U.S. tuna vessels fishing on dolphins because U.S. vessels currently do not fish on dolphins in the ETP.” Id. “Mexico had agreed to participate in the necropsy sampling program,” and “work with NMFS to conduct a pilot necropsy study.” Id. at vi, 5. In a December 10, 1998, necropsy meeting, Mexican representatives “repeatedly express[ed] strong commitment to facilitating the necropsy program to the full extent possible” and identified four appropriate fishing vessels, along with reporting that additional vessels would be available in the future. By the time of the Initial Finding, the only activity related to the necropsy study that had taken place was a three-day training session for technicians in January 19-21,1999.
The Secretary argues unconvincingly that NMFS could not obtain necropsy samples due to a lack of cooperation. NMFS did not request cooperation in a timely fashion, waiting until May 1998, seven months after the IDCPA directed NMFS to begin the study, to first request assistance. Thereafter, the record and Report to Congress indicate Mexico’s continued enthusiasm and willingness to cooperate. Moreover, the “lack of cooperation argument” does not explain why NMFS could begin necropsy sampling only after, rather than before, the Initial Finding. Finally, given Mexico’s cooperation, the Secretary fails to show that the study’s alleged complexity prevented NMFS from promptly commencing the study and obtaining preliminary data.
(2) One-Year Historical Study: “The intent of the historical and demographic data project is to assess the potential of existing archived samples from the fishery for determining fishery-related stress in ETP dolphins and to proceed with the most promising apparent avenues of research utilizing the samples.” Report at 6. “The [sample] archive includes several thousand samples for each year and consists of reproductive organs, teeth, blood stains, and skin. In addition, over 500 biopsy samples (skin samples taken by retrievable dart) were taken during the 1998 abundance survey in the ETP.” Id. NMFS noted that “[t]hese samples, while not necessarily ideal for all purposes, represent a readily available source of material for examination.” The record does not demonstrate, and the Secretary cannot explain, why readily available preliminary results from the one-year historical study were not used for the Initial Finding.
(3) Chase and Recapture Research Project: “The chase-recapture samples are intended to provide information about dynamic changes in physiological systems af*1070fected by chase, capture and release.” Report at 7. NMFS held a planning workshop for the chase and recapture research project in July 1997, before the IDCPA was enacted. In the Report to Congress, NMFS asserted that because of the “especially complex” nature of this project, it would “initiate additional formal research planning in mid-1999” and conduct the experiment between February and April 2001. NMFS did not explain why it chose to wait two years before initiating additional research planning. And NMFS’s assertions about complexity do not explain why it could not have obtained preliminary results.
Congress mandated that the agency make an Initial Finding on the basis of the specific research prescribed in 16 U.S.C. § 1414a, including stress studies. The record fails to show any compliance or valid excuse for the failure to comply. The Secretary’s emphasis on other work completed before the Initial Finding, including the mandatory abundance study, the stress literature review, and environmental variability analysis is irrelevant. Completion of other studies does not relieve the Secretary from progressing with clearly mandated studies. Applying the TRAC factors, it is clear that the Secretary unreasonably delayed the stress studies.
The agency invoked the lack of stress-related information to trigger a change in the dolphin-safe label standard. This puts the cart before the horse. The agency was required by law to conduct stress research as a prerequisite to its decision making. By failing to obtain and consider preliminary data from any of the mandated stress research projects before the Initial Finding, the Secretary unreasonably delayed action. 5 U.S.C. § 706(1). In addition, by failing to obtain and consider preliminary data from any of the mandated stress research projects in its Initial Finding, the Secretary abused his discretion, and acted arbitrarily and capriciously and not in accordance with the law, 5 U.S.C. § 706(2)(A) & (D). See Southwest Ctr. for Biological Diversity, 100 F.3d at 1448 (agency action may be arbitrary and capricious where the agency has “entirely failed to consider an important aspect of the problem”).-
The Best Available Scientific Evidence Standard
Scientific findings in marine mammal conservation area are often necessarily made from incomplete or imperfect information. The Secretary and Earth Island agree that the Initial Finding was to be determined using the “best available evidence” standard. Cf. Conner, 848 F.2d at 1454 (with best available data standard Congress required agency to consider the scientific information presently available and intended to give “the benefit of the doubt to the species”); Am. Tunaboat Ass’n v. Baldrige, 738 F.2d 1013 (9th Cir.1984) (discussing best available evidence standard under MMPA); Earth Island Inst. v. Brown, 865 F.Supp. 1364, 1373 n. 10 (N.D.Cal.1994) (discussing MMPA and concluding that there is no “indication that Congress intended to render all depletion determinations irrelevant simply because the available data may not always be complete or precise”). However, the Secretary determined that there was insufficient information to determine whether the fishery was having a significant adverse impact on the ETP dolphin stock. That determination was contrary to law and an abuse of his discretion.
The Endangered Species Act requires agencies to make determinations on the basis of the best scientific data available. Thus, a review of ESA case law provides insightful and analogous provisions and analysis. In Conner v. Burford, 848 F.2d 1441, 1454 (9th Cir.1988), this court held *1071that an agency’s claim of insufficient information to prepare comprehensive biological opinions violated the ESA requirement that opinions use best data available, and ordered the agency to comply with the ESA requirement. See also Greenpeace v. Nat’l Marine Fisheries Serv., 55 F.Supp.2d 1248, 1261-62 (W.D.Wash.1999) (best scientific data available standard requires less than conclusive proof; Secretary must issue biological opinion); Defenders of Wildlife v. Babbitt, 958 F.Supp. 670, 679-81 (D.D.C.1997) (Secretary must determine whether any species is threatened or endangered using the best available evidence).
As shown in the Report, the available information from the mandated abundance study and the stress literature review indicated that the fishery was having a significant adverse impact on the dolphin stocks. The abundance survey revealed that the dolphins were not recovering at expected levels, while the stress literature indicated that “stress resulting from chase and capture in the ETP tuna purse-seine fishery could have a population level effect on one or more dolphin stocks.” The record and the Report do not provide any contradictory conclusions, and NMFS was unable to attribute the failure to recover to any source other than the fishery. In fact, NMFS specifically ruled out the only potential non-fishery explanation for the slow population recovery-a large scale change in the ocean environment. Here, all of the evidence indicated that dolphins were adversely impacted by the fishery.
Given the best available evidence standard and IDCPA’s statutory mandate to determine whether or not the chase and netting of dolphins are having a significant adverse impact on the depleted ETP dolphin stocks, the Secretary cannot use insufficient evidence as an excuse for failing to comply with the statutory requirement. See, e.g., Conner, 848 F.2d at 1454; Comm. for Humane Legislation Inc. v. Richardson, 540 F.2d 1141, 1150 (D.C.Cir.1976) (NMFS’s assertion that “there is no evidence that the porpoise populations would substantially increase or decrease as a result of the ... reissuance of general permit” to set nets on dolphins was “not responsive” to statutory mandate). By claiming insufficiency of evidence, the Secretary acted contrary to law and abused his discretion. 5 U.S.C. § 706(2).
IV. Conclusion
For the foregoing reasons, the district court’s grant of summary judgment for Earth Island is AFFIRMED.
. Appellants and appellees will be referred to as "Secretary” and "Earth Island,” respectively.
. "Depleted,” a term of art under the MMPA, signifies that a species or population stock has fallen below its optimum sustainable population. 16 U.S.C. § 1362(1)(A).
. In 1972, the estimated annual ETP dolphin mortality rate caused by the purse seine net fishery was 423,678. By 1992, the estimated rate had dropped to 15,550. These numbers do not reflect the proportional impact on the total population.
. (g) Secretarial findings
(1) Between March 1, 1999, and March 31, 1999, the Secretary shall, on the basis of the research conducted before March 1, 1999, under section 1414a of this title, information obtained under the International Dolphin Conservation Program, and any other relevant information, make an initial finding regarding whether the intentional deployment on or encirclement of dolphins with purse seine nets is having a significant adverse impact on any depleted dolphin stock in the [ETP],
16 U.S.C. § 1385(g)(1).
Congress also required the Secretary to make a Final Finding by December 31, 2002 on the basis of the completed stress studies, information obtained under the International Dolphin Conservation Program, and any other relevant information:
(2) Between July 1, 2001, and December 31, 2002, the Secretary shall, on the basis of the completed study conducted under section 1414(a) of this title, information obtained under the International Dolphin Conservation Program, and any other relevant information, make a [final] finding regarding whether the intentional deployment on or encirclement of dolphins with purse seine nets is having a significant adverse impact on any depleted dolphin stock in the [ETP],
16 U.S.C. § 1385(g)(2).
. Sparse and unreliable data on the coastal spotted dolphins, the third depleted stock, failed to provide population abundance estimates. Report at 20-21. Therefore, NMFS concluded that “it is not possible at this time to determine if chase and encirclement by the purse seine fishery is having a significant adverse impact on the coastal stock of spotted dolphinsId. at 21.
. The record includes a "Scientists Statement Regarding the Setting of Nets on Dolphins in the [ETP]” (9/>s) and "Letter from Sixteen Marine Mammal Scientists" (3/24/99) ("[A]s scientists who have conducted research on cetaceans and have observed the impacts of human activity on them ... we believe that research on the effects of human activities on dolphins demonstrates that it is highly likely that the activities of the [ETP] tuna fleets are causing significant negative impacts on dolphins in addition to the direct mortalities counted by observers.”).
. Earth Island also challenged the Secretary's Initial Finding under the National Environmental Policy Act, 42 U.S.C. § 4321 et seq. The district court granted the Secretary's summary judgment motion on the NEPA claim holding that NEPA "only applies to 'discretionary federal action' " and did not apply in this case because the "Secretary's initial finding was a nondiscretionary act for purposes of NEPA.” Brower, 93 F.Supp.2d at 1090. Earth Island does not appeal that determination.
. (h) Certification by captain and observer.—
(1) Unless otherwise required by paragraph
(2), the certification by the captain under subsection (d)(2)(B)(i) of this section and the certification provided by the observer as specified in subsection (d)(2)(B)(ii) of this section shall be that no dolphins were killed or seriously injured during the sets in which the tuna were caught.
(2) The certification by the captain under subsection (d)(2)(B)(i) of this section and the certification provided by the observer as specified under subsection (d)(2)(B)(ii) of this section shall be that no tuna were caught on the trip in which such tuna were harvested using a purse seine net intentionally deployed on or to encircle dolphins, and that no dolphins were killed or seriously injured during the sets in which the tuna were caught, if the tuna were caught on a trip commencing—
(A) before the effective date of the initial finding by the Secretary under subsection (g)(1) of this section;
(B) after the effective date of such initial finding and before the effective date of the finding of the Secretary under subsection (g)(2) of this section, where the initial finding is that the intentional deployment on or encirclement of dolphins is having a significant adverse impact on any depleted dolphin stock; or
(C) after the effective date of the finding under subsection (g)(2) of this section, where such finding is that the intentional deployment on or encirclement of dolphins is having a significant adverse impact on any such depleted stock.
16 U.S.C. § 1385(h).
. See Report, A.R. 2141 ("Considerable concern about the potential effects of stress caused by [purse seine nets] led to inclusion in the IDCPA of research projects directed toward assessing the prevalence and magnitude of fishery-induced stress in the dolphins targeted by this fishery.”); S8299, 143 Congr. Rec. (IDCPA requires "the expeditious commencement of research to further study the effect of dolphin setting on dolphins. Tuna caught by dolphin sets may not be labeled dolphin safe until at least March 1999, at which time the Secretary of Commerce must review the preliminary results of the study, and make a determination as to whether or not dolphin setting is causing significant adverse impacts to depleted dolphin stocks in the ETP.”).
. " 'Shall' means shall.” Center for Biological Diversity v. Norton, 254 F.3d 833, 837-38 (9th Cir.2001) (quoting Forest Guardians v. *1068Babbitt, 174 F.3d 1178, 1187-88 (10th Cir.1999)); see also United States v. Monsanto, 491 U.S. 600, 607, 109 S.Ct. 2657, 105 L.Ed.2d 512 (1989) (by using "shall” "Congress could not have chosen stronger words to express its intent that forfeiture be mandatory”).
. Accordingly, the Secretary's assertion, that he had discretion to schedule the stress studies after the Initial Finding, is also without merit. This contention conflicts with the plain statutory terms (§ 1414a (requiring stress studies); § 1385(g)(1) (requiring Initial Finding to use § 1414a research)) and is not accorded deference. See, e.g., Board of Governors of the Fed. Reserve Sys. v. Dimension Fin. Corp., 474 U.S. 361, 368, 106 S.Ct. 681, 88 L.Ed.2d 691 (1986).