Modesto Irrigation District v. Gutierrez

OPINION SCHROEDER, Circuit Judge: This Endangered Species Act (“ESA”) case is a challenge to the decision of the National Marine Fisheries Service (“NMFS”) to list the steelhead, a type of Pacific salmon, as a threatened species in California’s Central Valley. In listing the steelhead, NMFS defined it as a distinct species under the ESA, separate from rainbow trout, another type of Pacific salmon that breeds with and looks like the steelhead. The separate listing was a departure from the prior NMFS policy of classifying interbreeding Pacific salmon as a single species. Plaintiffs are irrigation districts in California’s Central Valley whose operations are impeded by the listing. They contend that the listing violated the ESA because steelhead and rainbow trout interbreed, and the statute therefore requires NMFS to treat them as a single species. Plaintiffs also contend that NMFS violated the Administrative Procedure Act (“APA”) by failing adequately to explain its decision to adopt a new policy for classifying the fish. We agree with the district court that under the ESA, interbreeding is not alone determinative of whether organisms must be classified alike where, as here, they develop and behave differently. We also find that NMFS’ explanation for its change of policy satisfies the standards set forth in the Supreme Court’s recent decision in F.C.C. v. Fox Television Stations, Inc., — U.S. — -, 129 S.Ct. 1800, 173 L.Ed.2d 738 (2009). We therefore affirm. THE STEELHEAD AND RAINBOW TROUT This case turns upon the distinctions between the steelhead and rainbow trout, types of Pacific salmon that comprise the Oncorhynchus mykiss (“O. mykiss ”) species as scientifically defined. The fish are born in fresh water, but the steelhead migrate to the ocean anywhere from hours to years after their birth. To transition from fresh water to salt water, steelhead undergo a “smolt” stage, and then after one to five years in the sea, return to the original stream to spawn. Because of their migration pattern, steelhead are known as the anadromous form of 0. myk-iss. The rainbow trout, on the other hand, remain in fresh water their entire lives and are commonly known as the resident form of the 0. mykiss species. While the two fish grow to differing sizes as adults and have different predators and prey, they do interbreed to some extent, and the offspring can take on the form of either. An excess of steelhead can regenerate the population of rainbow trout, but the reverse does not seem to be the case. All parties agreed in the district court and before us that the steelhead population is in decline in the Central Valley of California. This unique biology of 0. mykiss has complicated government regulation of the *1028 species. In 1974, the United States Fish and Wildlife Service (“FWS”) and NMFS entered into a memorandum of understanding defining their respective jurisdictions under the ESA. FWS has jurisdiction over predominantly freshwater fish including rainbow trout, and NMFS has jurisdiction over ocean dwellers, including steelhead. This concurrent jurisdiction has required the agencies to work together in classifying 0. mykiss under the ESA. STATUTORY AND REGULATORY HISTORY The ESA defines the term “species” to “include[ ] any sub-species of fish or wildlife or plants, and any distinct population segment of any species of vertebrate fish or wildlife which interbreeds when mature.” 16 U.S.C. § 1532(16). Congress added the term “distinct population segment” (“DPS”) to the definition of species in 1978 so that agencies could “provide different levels of protection to different populations of the same species.” Nat’l Ass’n of Home Builders v. Norton, 340 F.3d 835, 842 (9th Cir.2003). Congress did not define DPS, nor is the term used in the scientific community. As a result of this limited guidance from Congress, NMFS has struggled for two decades over how to apply the term DPS with respect to steel-head and rainbow trout. NMFS’ first attempt at defining a DPS in this context occurred in 1991 when the agency adopted a policy designed specifically for classifying Pacific salmon such as 0. mykiss. See Policy on Applying the Definition of Species Under the Endangered Species Act to Pacific Salmon, 56 Fed.Reg. 58,612 (Nov. 20, 1991) (“ESU Policy”). This policy did not use the statutory term DPS, but instead created another new term called the evolutionary significant unit (“ESU”). NMFS also determined, however, that an ESU was the functional equivalent of a DPS. See id. Under the ESU Policy, a salmon stock had to satisfy two main criteria before NMFS could place the stock in a distinct ESU: (1) It must be substantially reproductively isolated from other nonspecific population units; and (2) [I]t must represent an important component in the evolutionary legacy of the species. Id. In 1996, in conjunction with FWS, NMFS promulgated a more general policy defining the ESA’s operative concept of DPS, a policy that would apply to all animal, bird and fish species. See Policy Regarding the Recognition of Distinct Vertebrate Population Segments Under the Endangered Species Act, 61 Fed.Reg. 4,722 (Feb. 7, 1996) (“DPS Policy”). The DPS Policy, like the ESU Policy, required NMFS to examine multiple factors in determining whether an organism qualified as a DPS. See 61 Fed.Reg. at 4,725. There was, however, a key difference between the ESU and DPS Policies. Under the ESU Policy, a type of Pacific salmon had to be “substantially reproductively isolated” from other salmon stock before it could be classified in its own ESU, whereas, under the DPS Policy, an organism could be placed its own DPS so long as it was “markedly separated from other populations of the same taxon as a consequence of physical, physiological, ecological, or behavioral factors.” See Endangered and Threatened Species: Request for Comment on Alternative Approach to Delineating 10 Evolutionary Significant Units of West Coast Oncorhynchus mykiss, 70 Fed. Reg. 67,130, 67,131 (Nov. 4, 2005) (describing the history of the policies). NMFS and FWS concluded in 1996 that the ESU Policy was, despite this difference, “consistent with” the DPS Policy and that both agencies would continue to apply the ESU Policy to Pacific salmon, and apply the *1029 DPS Policy to all other organisms. See 61 Fed.Reg. at 4,722. When NMFS applied the ESU Policy to Pacific salmon in 1997, it determined that interbreeding steelhead and rainbow trout populations should be classified in the same ESU because they were not “substantially reproductively isolated.” See Endangered and Threatened Species: Listing of Several Evolutionary Significant Units (ESUs) of West Coast Steelhead, 62 Fed.Reg. 43,937, 43,941 (Aug. 18, 1997). NMFS noted, however, that “conclusive evidence d[id] not yet exist regarding the relationship of resident and anadromous 0. mykiss,” which in most respects was “poorly understood.” See id. at 43,938, 43,941. Although NMFS and FWS agreed in 1997 that interbreeding populations of steelhead and rainbow trout should be placed in the same ESU, there was some dispute about how to protect threatened 0. mykiss populations. In many regions the steelhead population was declining and NMFS desired to list those populations as threatened or endangered. FWS, however, opposed any listing that would have protected the rainbow trout, having determined that its populations were in no way threatened. See id. at 43,941. FWS thus did not want to list any ESU containing both steelhead and rainbow trout, even where the steelhead population within that ESU was declining. NMFS addressed FWS’ concern by deciding that when an ESU contained both steelhead and rainbow trout, NMFS would list only the steelhead population within that ESU as threatened or endangered. See 70 Fed.Reg. at 67,130. By listing subdivisions of an ESU, NMFS was able to protect declining steelhead populations while acceding to FWS’ desire not to list rainbow trout. Such subdivision listings were challenged in court, however. In Alsea Valley Alliance v. Evans, 161 F.Supp.2d 1154, 1161-64 (D.Or.2001), the court held that the ESA requires that agencies list (or not list) entire DPSs/ESUs, thus precluding any listings below the ESU/DPS level. The court in Alsea found it was contrary to the ESA for NMFS to list only the steel-head population of an ESU containing both steelhead and rainbow trout. NMFS therefore had to revisit all of its prior steelhead listings. In 2004, NMFS promulgated a number of new listings to conform with the Alsea decision. Relevant to this litigation, NMFS defined the California Central Valley 0. mykiss ESU as comprising all steel-head and rainbow trout that “co-oc-curfred]” below impassible barriers. See Endangered and Threatened Species: Proposed Listing Determinations for 27 ESUs of West Coast Salmonids, 69 Fed.Reg. 33,-102, 33,115 (June 14, 2004). It proposed listing this ESU as threatened. Id. at 33,103. There were those, however, who opposed the new listings on the ground that it was inappropriate to classify steelhead and rainbow trout in the same ESU. They pointed out that even though steelhead and rainbow trout interbreed to some extent, they differ in behavior and in their ability to sustain themselves as a distinct segment of Pacific salmon. NMFS received several comments from environmentalists who “felt that rainbow trout and steelhead should be considered separate ESUs for biological reasons (differences in behavior, morphology, and ecology); or for policy or legal reasons (such as implementing the purposes of the ESA).” Endangered and Threatened Species: Final Listing Determinations for 10 Distinct Population Segments of West Coast Steel-head, 71 Fed.Reg. 834, 836 (Jan. 5, 2006). *1030 FWS also disagreed with the proposed listings. The Director of the FWS sent a letter to NMFS in June 2005 suggesting that NMFS had erred in not considering all the biological differences between the two fish when classifying them in the same ESU. FWS indicated that NMFS had placed too much emphasis on the fact that the two fish were a part of the same scientifically defined species, and suggested that it would be appropriate to stop applying the ESU Policy to 0. mykiss and instead use the more general DPS Policy so that steelhead and rainbow trout could be treated differently. See 70 Fed.Reg. at 67,130. At about this same time, NMFS obtained new scientific reports illuminating the differences between steelhead and rainbow trout. These reports indicated that, despite the reproductive exchange between the two fish, “it seem[ed] unlikely that a population of resident trout c[ould] consistently reestablish a steelhead population.” Independent Science Advisory Board, Viability ofESUs Containing Multiple Types of Populations 29 (2005); see also Salmon Recovery Science Review Panel, Report for the meeting held December 200k 12 (“[R]esident populations by themselves should not be relied upon to maintain long-term viability of an ESU.”). In response to both FWS’ concerns and the new scientific information highlighting the importance of protecting the steelhead, NMFS determined that it should delay the proposed listings for six months. On November 4, 2005, NMFS proposed to abandon the ESU Policy with respect to O. mykiss. NMFS proposed using the DPS Policy instead to classify O. mykiss so the steelhead and rainbow trout could be treated separately under the ESA. See 70 Fed.Reg. at 67,131. In support of the change in policy, NMFS stated: [I]t is appropriate that we consider departing from our past practice of applying the ESU Policy to O. mykiss stocks, and instead apply the DPS Policy in determining “species” of O. mykiss for listing consideration. Such an approach would also be consistent with use of the DPS Policy by the agencies in defining DPSs of Atlantic salmon ... The primary difference in the application of the two policies is that the ESU Policy relies on “substantial reproductive isolation” as the primary factor in delineating a group of organisms, while the DPS Policy relies on “marked separation” to delineate the group. Within a discrete group of O. mykiss populations, the resident and anadromous life forms of 0. mykiss remain “markedly separated” as a consequence of physical, physiological, ecological, and behavioral factors. Despite the apparent lack of reproductive isolation between the two forms within a given population or group of populations, under the DPS Policy anadromous and resident 0. mykiss may not warrant delineation as part of the same DPS. Id. After receiving fifteen total comments, both opposed to and in support of the policy change, NMFS formally announced in January 2006 it would apply the DPS Policy to 0. mykiss. In support of the policy change, NMFS provided two justifications. NMFS first pointed to administrative consistency, noting that it and FWS shared jurisdiction over Atlantic salmon and had applied the DPS Policy to classify this type of fish. NMFS also stated that it was no longer appropriate to apply the ESU Policy to 0. mykiss, because it was different from other types of Pacific salmon. The full justification stated: Although the ESU Policy did not by its terms apply to steelhead, the DPS Policy states that NMFS will continue to *1031 implement the ESU Policy with respect to “Pacific salmonids” (which include 0. mykiss). FWS, however, does not use our ESU policy in any of its ESA listing decisions. In a previous instance of shared jurisdiction over a species (Atlantic salmon), we and FWS used the DPS policy in our determination to list the Gulf of Maine DPS of Atlantic salmon as endangered (65 FR 69459; November 17, 2000). Given our shared jurisdiction over O. mykiss, and consistent with our approach for Atlantic salmon, we believe application of the joint DPS policy here is logical, reasonable, and appropriate for identifying DPSs of 0. mykiss. Moreover, use of the ESU policy — originally intended for Pacific salmon— should not continue to be extended to 0. mykiss, a type of salmonid with characteristics not typically exhibited by Pacific salmon. NMFS and FWS also intend to continue to evaluate application of the statutory term “distinct population segment” in a process outside the context of a species-specific listing. 71 Fed.Reg. at 834. Pursuant to the DPS Policy, NMFS listed the Central Valley steelhead in its own DPS, apart from rainbow trout, and then listed steelhead as threatened. THIS LITIGATION Plaintiffs Modesto Irrigation District and other irrigation and water districts (collectively, “MID”) formed a coalition in 2006 to file this suit under the APA in the Eastern District of California challenging the listing. MID’s principal arguments were that NMFS had not adequately justified the change of policy, and that NMFS’ classification of steelhead in its own DPS violated the plain language of the ESA because the steelhead and rainbow trout were not sufficiently biologically different to justify separate treatment. The Center for Biological Diversity (“CBD”) intervened to support the government. The district court granted the government’s motion for summary judgment in a published opinion. See California State Grange v. Nat’l Marine Fisheries Serv., 620 F.Supp.2d 1111 (E.D.Cal.2008) (“Grange ”). The opinion dealt with this and related litigation not now before us. As relevant to this appeal, the district court first rejected MID’s contention that the ESA required interbreeding organisms to be included in the same DPS. The ESA defines a species as “any subspecies of fish or wildlife or plants, and any distinct population segment of any species of vertebrate fish or wildlife which interbreeds when mature.” 16 U.S.C. § 1532(16). The district court found this definition ambiguous in that it did not clearly state that all interbreeding organisms must be placed in the same DPS. Grange, 620 F.Supp.2d at 1174-76. The district court also noted that Congress added the term DPS to the species definition in order to provide agencies with greater flexibility in their classifications. It concluded that MID’s proposed limiting construction would be contrary to this intent. Id. at 1176. The district court also found that NMFS had provided sufficient justification for the change from the ESU to DPS Policy, applying then controlling law. MID appeals and contends that under a proper interpretation of the ESA, the steelhead and rainbow should be classified in the same DPS because, to some extent, they interbreed. MID also contends that the policy change for 0. mykiss from the ESU Policy to the DPS Policy was not adequately explained or justified and hence was arbitrary and capricious. WHETHER THE ESA REQUIRES SIMILAR TREATMENT FOR STEEL-HEAD AND RAINBOW TROUT The ESA provides that “[t]he term ‘species’ includes any subspecies of fish or *1032 wildlife or plants, and any distinct population segment of any species of vertebrate fish or wildlife which interbreeds when mature.” 16 U.S.C. § 1532(16). MID contends that in defining a species as a “[DPS] ... which interbreeds when mature,” the ESA requires agencies to place all interbreeding organisms in the same DPS. MID thus argues that NMFS violated the plain language of the ESA by placing steelhead in their own DPS, because steelhead sometimes interbreed with rainbow trout. The parties agree that the analysis set forth in Chevron U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), guides the present inquiry. Under Chevron, we first determine whether Congress has “directly spoken to the precise question at issue.” 467 U.S. at 842, 104 S.Ct. 2778. If Congress’ intent is clear, that ends the matter. Id. at 842-43, 104 S.Ct. 2778. If, however, the statute is ambiguous, we must next determine whether the agency’s interpretation of the statute is permissible. Id. at 843, 104 S.Ct. 2778. For MID, the statute is not ambiguous and we need not proceed beyond Chevron’s first step. MID contends that the ESA made interbreeding the single, definitive characteristic of a DPS; so long as interbreeding can take place, the fish should be in the same DPS. NMFS’ position is that interbreeding is a necessary, but not a sufficient condition for classification as a DPS. Some organisms may be distinguished by other factors that justify different classifications. The agency explained this in its final rule: The ESA requirement that a group of organisms must interbreed when mature to qualify as a DPS is a necessary but not exclusive condition. Under the definition, although all organisms that belong to a DPS must interbreed when mature (at least on some time scale), not all organisms that share some reproductive exchange with members of the DPS must be included in the DPS. The DPS policy outlines other relevant considerations for determining whether a particular group should be delineated as a DPS (i.e., “marked separation” as a consequence of physical, physiological, ecological or behavioral factors). 71 Fed.Reg. at 838. NMFS thus agrees the ESA requires organisms to interbreed in order to qualify as a “distinct population segment,” or species as defined by the statute. It contends that because steel-head interbreed, the ESA’s species definition has been satisfied. We agree with the district court that the statutory definition of a species as a “[DPS] ... which interbreeds when mature” is “grammatically ambiguous.” See Grange, 620 F.Supp.2d at 1175. We also agree that the government’s reading of Section 1532(16) finds support in the statutory language. There is no indication within the text that Congress intended to create a rigid limitation on an agency’s discretion to define the statutorily undefined concept of a “distinct population segment.” Cf. Strunk & White, The Elements of Style 59 (3d ed. 1979) (stating that phrases using the relative pronoun ‘which’ can be unclear due to the pronoun’s frequent misuse). The “[DPS] ... which interbreeds when mature” language in Section 1532(16) is phrased in terms of interbreeding as a condition for a DPS to qualify as a species under the ESA. Yet it does not necessarily indicate that interbreeding must be the sole defining characteristic of a DPS, because, by definition, a DPS must also be “distinct.” The statute certainly cannot be read to support, unambiguously, MID’s contention that all interbreeding organisms must be treated the same. *1033 Indeed, this court has already determined that “distinct population segment” is an ambiguous term and that the DPS Policy defining it is entitled to Chevron deference. See Nw. Ecosystem Alliance v. U.S. Fish and Wildlife Serv., 475 F.3d 1136, 1141-44 (9th Cir.2007). By inventing the term DPS and then declining to define it, Congress intended to provide agencies with discretion in deciding the composition of a DPS. See Nat’l Ass’n of Home Builders, 340 F.3d at 842 (noting that Congress added DPS to the species definition to allow agencies to “provide different levels of protection to different populations of the same species”). MID’s suggestion that the “interbreeds when mature” language somehow restricts that discretion runs counter to Congressional intent and our decision in Northwest Ecosystem Alliance. Under our circuit law, the DPS Policy rests on a permissible interpretation of the ESA. Moreover, contrary to MID’s suggestion, the decision in Alsea does not support MID’s proposed statutory interpretation. See 161 F.Supp.2d at 1161-64. There, NMFS defined an ESU to consist of both naturally spawned and hatchery spawned Oregon coast coho salmon, but listed only the naturally spawned population as threatened. Id. The court found this listing arbitrary and capricious because the ESA permits listing only an entire species as threatened or endangered. Id. at 1162. Alsea held that once NMFS defines an ESU/DPS, it cannot list only a portion of that DPS as threatened or endangered. See id. at 1161-63. Here there has been no subdivision below the DPS level. NMFS determined that the Central Valley steelhead warranted its own DPS. NMFS then listed that entire DPS as threatened, thereby curing any Alsea problem. MID’s reliance on Alsea is therefore misplaced. For these reasons, we conclude that Section 1532(16) of the ESA does not require that interbreeding organisms be placed in the same DPS. THE GOVERNMENT PROVIDED AN ADEQUATE RATIONALE FOR THE CHANGE IN POLICY NMFS changed its policy when it applied the DPS Policy to 0. mykiss after it had previously applied the ESU Policy, and it is undisputed that such a policy change requires an explanation. See F.C.C. v. Fox Television Stations, Inc., — U.S. -, 129 S.Ct. 1800, 1810-11, 173 L.Ed.2d 738 (2009). MID questions whether NMFS provided sufficient explanation for the change. We hold that the record reflects that NMFS engaged in a careful decision-making process and provided a sufficient explanation for its decision to apply the DPS Policy. The district court held, on the basis of aging authority, that NMFS’ decision to adopt the DPS Policy for O. mykiss was supported by the record and sufficiently explained. See Grange, 620 F.Supp.2d at 1161-62 (relying on Motor Vehicle Mfrs. Ass’n of the United States v. State Farm, Mut. Auto. Ins. Co., 463 U.S. 29, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983)). Subsequent to the district court’s decision, however, the Supreme Court provided new guidance to courts reviewing an agency’s change in policy, that in effect confirmed the validity of the district court’s analysis. See Fox, 129 S.Ct. at 1810-11. In Fox, the Court examined the FCC’s change in its definition of what constituted an indecent broadcast. Id. at 1806-07. Under the FCC’s old policy, a broadcast with a “fleeting” or non-literal expletive was rarely deemed indecent. Id. The new policy made such “fleeting” and non-literal expletives actionably indecent. Id. The Second Circuit struck down the change, ruling that the FCC was required to pro *1034 vide a thorough justification of why the new policy was necessary, and that the agency had failed to do so. Id. at 1810. The Second Circuit interpreted the Supreme Court’s decision in State Farm, as requiring courts to conduct a “more searching review” of an agency’s decision to abandon an old policy in favor of a new one. Fox, 129 S.Ct. at 1810. The Supreme Court reversed and upheld the FCC’s new indecency definition. It held that neither the APA nor State Farm require courts to subject an agency’s policy change to heightened judicial review. Id. Rather, the Court noted that an agency need only satisfy certain criteria in order to justify a policy change. First an agency must “display awareness that it is changing position.” Id. at 1811. An agency must then explain its reasoning, but it need not demonstrate that the new policy is an improvement. It must: show that there are good reasons for the new policy. But it need not demonstrate to a court’s satisfaction that the reasons for the new policy are better than the reasons for the old one; it suffices that the new policy is permissible under the statute, that there are good reasons for it, and that the agency believes it to be better, which the conscious change of course adequately indicates. Id. Some circumstances require that an agency provide a greater justification for changing a policy than for adopting a new one. In such cases, the justification must be greater than what “would suffice for a new policy created on a blank slate.” Id. The Court suggested such heightened explanation may be appropriate, for example, if the new policy rests upon factual findings that contradict those which underlay the previous policy or if the old policy has “engendered serious reliance interests.” Id. Neither side has argued that such a heightened standard applies here. Agencies thus have flexibility to alter policies under Fox. The other Circuits that have already applied its standard have recognized that the Supreme Court intended to provide agencies considerable latitude to change course. So long as the agency “fully explain[s] how its new construction is permissible under the statute, that there are good reasons for the new construction, and that [the agency] believes the new interpretation to be better,” a previous position is “no obstacle” to adoption of a new course. Ad Hoc Shrimp Trade Action Comm. v. United States, 596 F.3d 1365, 1372 (Fed.Cir.2010) (upholding policy change where agency determined that a “better reading” of the controlling statute required a departure from previous practice); see also Westar Energy, Inc. v. Fed. Energy Regulatory Comm’n, 568 F.3d 985, 988-89 (D.C.Cir.2009) (finding that where agency “carefully explained” policy change it must be upheld under Fox). At the same time, because judicial deference is not unlimited, departures from prior policy must be recognized and explained. Courts will not “assume [an agency] has engaged in reasoned decision making” when it “implicitly” departs from its prior precedent and provides no explanation for doing so. See Dillmon v. Nat’l Transp. Safety Bd., 588 F.3d 1085, 1091 (D.C.Cir.2009). Here NMFS recognized the policy change and explained it. The explanation for the change appeared in the final rule as follows: Although the ESU Policy did not by its terms apply to steelhead, the DPS Policy states that NMFS will continue to implement the ESU Policy with respect to “Pacific salmonids” (which include O. mykiss). FWS, however, does not use our ESU policy in any of its ESA listing *1035 decisions. In a previous instance of shared jurisdiction over a species (Atlantic salmon), we and FWS used the DPS policy in our determination to list the Gulf of Maine DPS of Atlantic salmon as endangered ... Given our shared jurisdiction over 0. mykiss, and consistent with our approach for Atlantic salmon, we believe application of the joint DPS policy here is logical, reasonable, and appropriate for identifying DPSs of 0. mykiss. Moreover, use of the ESU policy — originally intended for Pacific salmon — should not continue to be extended to 0. mykiss, a type of salmonid with characteristics not typically exhibited by Pacific salmon. NMFS and FWS also intend to continue to evaluate application of the statutory term “distinct population segment” in a process outside the context of a species-specific listing. 71 Fed.Reg. at 834. NMFS explicitly recognized it was changing policy when it determined the DPS Policy to be a better fit for 0. mykiss. See Fox, 129 S.Ct. at 1811 (requiring agencies to recognize that they are changing policy). While a new policy must rest upon a permissible construction of the controlling statute, see Fox, 129 S.Ct. at 1811, we have here and previously determined that the DPS Policy represents a permissible construction of the ESA. See Nw. Ecosystem Alliance, 475 F.3d at 1150 (“The DPS Policy is entitled to Chevron deference.”). The more serious question here is whether NMFS provided “good reasons” for the policy change. See Fox, 129 S.Ct. at 1811. MID does not argue that NMFS proffered bad reasons in choosing to use the DPS Policy. Rather it contends that the “good reasons” offered by NMFS have no support in the record. As the ESU Policy was designed specifically for Pacific salmon, MID points to NMFS’ lack of detailed explanation for its conclusion that 0. mykiss are subject to the DPS Policy because they differ materially from other Pacific salmon. Even if we were to agree with MID that NMFS’ explanation leaves several unanswered questions, the law does not require the explanation to be exhaustive. We must “uphold a decision of less than ideal clarity if the agency’s path may reasonably be discerned.” See id. (citation omitted). Here the notice and comment process reveals NMFS’ evolving understanding of 0. mykiss and reflects NMFS’ determination that the differences between steelhead and rainbow trout set 0. mykiss apart from other Pacific salmon. When NMFS first attempted to list the steelhead in 1997, it recognized there were “uncertainties about the relationship of resident and anadromous 0. mykiss.” See 71 Fed.Reg. at 834. The proposed rule, issued almost a decade later, describes at length the differences between the two fish, evidencing that the agency had since determined that the distinctions between steelhead and rainbow trout were expansive: Despite the apparent reproductive exchange between resident and anadro-mous 0. mykiss, the two life forms remain markedly separated physically, physiologically, ecologically, and behaviorally. Steelhead differ from resident rainbow trout physically in adult size and fecundity, physiologically by undergoing smoltification, ecologically in their preferred prey and principal predators, and behaviorally in their migratory strategy. Where the two life forms co-occur, adult steelhead typically range in size from 40-72 cm in length and 2-5 kg body mass, while adult rainbow trout typically range in size from 25-46 cm in length and 0.5-2 kg body mass. Steel-head females produce approximately 2,500 to 10,000 eggs, and rainbow trout *1036 fecundity ranges from 700 to 4,000 eggs per female, with steelhead eggs being approximately twice the diameter of rainbow trout eggs or larger. Steelhead undergo a complex physiological change that enables them to make the transition from freshwater to saltwater (smoltification), while rainbow trout reside in freshwater throughout their entire life cycle. While juvenile and adult steel-head prey on euphausiid crustaceans, squid, herring, and other small fishes in the marine environment, the diet of adult rainbow trout is primarily aquatic and terrestrial insects and their larvae, mollusks, amphipod crustaceans, fish eggs, and minnows. Finally, steelhead migrate several to hundreds of miles from their natal streams to the ocean, and spend up to 3 years in the ocean migrating thousands of miles before returning to freshwater to spawn. Rainbow trout, in contrast, may exhibit seasonal migrations of tens of kilometers but generally remain associated with their natal drainages. 70 Fed.Reg. at 67,132 (internal citations omitted). In later extending the comment period, NMFS noted that it had received and was considering a number of new scientific studies bearing on the relationship between steelhead and rainbow trout. See Endangered and Threatened Species: 6-month Extension of the Final Listing Determinations for Ten Evolutionary Significant Unites of West Coast Oncorhynehus mykiss, 70 Fed.Reg. 37,219, 37,220 (June 28, 2005). These reports demonstrated that, despite the reproductive exchange between the fish, rainbow trout probably cannot regenerate a steelhead population. See Independent Science Advisory Board, Viability of ESUs Containing Multiple Types of Populations 29 (2005) (stating that it is “unlikely that a population of resident trout can consistently reestablish a steelhead population”); Salmon Recovery Science Review Panel, Report for the meeting held December 200k 12 (2004) (“[Rjesident populations by themselves should not be relied upon to maintain long-term viability of an ESU.”). This record thus contains ample support for the reasons NMFS stated when it decided to use the DPS Policy. The final rule stated that the ESU Policy was no longer appropriate for classifying O. myk-iss because it is “a type of salmonid with characteristics not typically exhibited by Pacific salmon.” 71 Fed.Reg. at 834. The final rule may not explicitly describe why and how O. mykiss differ from other species of Pacific salmon, but the rule does provide an extensive discussion of the similarities and differences between steelhead and rainbow trout that are not shared by other Pacific salmon. It “may reasonably be discerned,” then, that NMFS determined that O. mykiss is distinct from other types of Pacific salmon. See Fox, 129 S.Ct. at 1811. MID also points to scientific evidence in the record it suggests shows that steelhead and rainbow trout are not as different as NMFS states. Yet our task is not to second guess the agency’s action. We must defer to a reasonable agency action “even if the administrative record contains evidence for and against its decision.” See Trout Unlimited v. Lohn, 559 F.3d 946, 958 (9th Cir.2009). Moreover, NMFS provided reasons for the policy change that MID does not meaningfully challenge or dispute. The final rule cited precedent for applying the DPS Policy, stating that NMFS and FWS, in another instance of overlapping jurisdiction, had jointly applied the DPS Policy to Atlantic salmon. 71 Fed.Reg. at 834. The final rule explained it was in accord with FWS’ expressed preference for the DPS Policy. Id. at 837. These administrative justifications demonstrate that the agencies were seeking to adopt a consistent *1037 approach to a similar issue, and we view this to be a “good reason” for a policy change. MID’s principal reliance is on three cases, all decided before the Supreme Court’s decision in Fox. They are, hence, of questionable value, because it is doubtful any applied the deference Fox now requires. Even if we were to assume they did apply the proper deferential standard, however, they are all distinguishable in any event. In Defenders of Wildlife v. Hall, 565 F.Supp.2d 1160, 1171-78 (D.Mont.2008), the court enjoined a delisting because the government did not explain a sudden change in policy. The record in this case reflects how NMFS’ improved scientific understanding of the characteristics of O. mykiss led eventually to the policy change. In Northwest Environmental Defense Center v. Bonneville Power Administration, 477 F.3d 668, 686-90 (9th Cir.2007), the government relied excessively on language in a Congressional committee report. Here, following a lengthy notice and comment period, the government offered scientific evidence in support of the change. Finally, in Western Watersheds Project v. Kraayenbrink, 538 F.Supp.2d 1302, 1312-14 (D.Idaho 2008), the government justified a policy change on the basis of generalized concerns for cost and efficiency. The change in this case has been shown to be attributable to information about biological differences between the steelhead and rainbow trout. The agency’s pathway and rationale for the change was thus apparent, and its decision was not arbitrary or capricious. See Fox, 129 S.Ct. at 1811. In light of the evolving understanding of the differences between the fish, the desire for a flexible policy, and the depth of consideration that NMFS has given the issue over close to two decades, we conclude that the agency was not arbitrary or capricious in changing its policy in order to protect the steelhead. We therefore affirm the district court’s holding that the explanation for the policy change was sufficient. Its ruling was fully consistent with the Supreme Court’s subsequent decision in Fox. We are aware of the practical difficulties the classification decision creates for management of the Irrigation and Water Districts. The two types of fish look and behave the same during their early years in the rivers and streams in the area, and the practical effect of our decision is that plaintiffs may have to protect both the steelhead and rainbow trout, even though only the steelhead is threatened. The flexibility to make policy changes in response to such concerns, however, remains in the agencies administering the provisions of the ESA and not with the courts. CONCLUSION The two narrow issues before this panel involve first a question of statutory interpretation and second, whether an agency provided an adequate justification for a policy change. We agree with the district court that the definition of “species” in the ESA did not require NMFS to place interbreeding steelhead and rainbow trout in the same DPS. We further hold that, under the Supreme Court’s decision in Fox, NMFS sufficiently justified its decision to apply the DPS Policy to O. mykiss. We therefore affirm the district court. AFFIRMED.