Tucson Herpetologica v. Kempthorne

FOR PUBLICATION UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT TUCSON HERPETOLOGICAL SOCIETY;  DEFENDERS OF WILDLIFE; CENTER FOR BIOLOGICAL DIVERSITY; HORNED LIZARD CONSERVATION SOCIETY; SIERRA CLUB; WENDY HODGES; FRANCIS ALLAN MUTH, No. 07-16641 Plaintiffs-Appellants, v.  D.C. No. CV 04-0075 NVW KEN SALAZAR,* in his official OPINION capacity as Secretary of the Interior; ROWAN W. GOULD,** in his official capacity as Acting Director of the U.S. Fish and Wildlife Service, Defendants-Appellees.  Appeal from the United States District Court for the District of Arizona Neil V. Wake, District Judge, Presiding Argued and Submitted January 16, 2009—San Francisco, California Filed May 18, 2009 Before: John T. Noonan, A. Wallace Tashima, and William A. Fletcher, Circuit Judges. *Ken Salazar is substituted for his predecessor Dirk Kempthorne, as Secretary of the Interior, pursuant to Fed. R. App. P. 43(c)(2). **Rowan W. Gould is substituted for his predecessor H. Dale Hall, as Acting Director of the U.S. Fish and Wildlife Service. [5941] 5942 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR Opinion by Judge Tashima; Partial Concurrence and Partial Dissent by Judge Noonan 5944 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR COUNSEL Neil Levine, Denver, Colorado, for the plaintiffs-appellants. TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR 5945 Sambhav N. Sankar, Environment & Natural Resources Division, U.S. Department of Justice, Washington, DC, for the defendants-appellees. OPINION TASHIMA, Circuit Judge: Conservation organizations and individual biologists (collectively “Plaintiffs”) contend that the Secretary of the Interior’s (the “Secretary”) decision to withdraw a rule proposing that the flat-tailed horned lizard (the “lizard”) be listed as a threatened species is contrary to the requirements of the Endangered Species Act (“ESA” or the “Act”), 16 U.S.C. § 1531 et seq., and the Administrative Procedure Act (“APA”), 5 U.S.C. § 706 . They appeal from the district court’s order granting summary judgment in favor of the Secretary. The district court had jurisdiction under 28 U.S.C. § 1331 and 16 U.S.C. § 1540 (g)(1)(C) (authorizing “citizen suits” to compel the Secretary to perform non-discretionary duties required by the ESA). We have jurisdiction to review the district court’s final judgment under 28 U.S.C. § 1291 , and we reverse in part and remand. STATUTORY BACKGROUND The ESA provides “a means whereby the ecosystems upon which endangered species and threatened species depend may be conserved.” 16 U.S.C. § 1531 (b). In service of this goal, the Act directs the Secretary to maintain a list of threatened and endangered species, and defines the phrase “endangered species” as “any species which is in danger of extinction throughout all or a significant portion of its range.” Id. § 1532(6).1 A threatened species is one “which is likely to [1] Upon listing a species, the Secretary must issue regulations to provide for the species’ conservation. See 16 U.S.C. § 1533 (d). 5946 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR become an endangered species within the foreseeable future throughout all or a significant portion of its range.” Id. § 1532(20). “The term ‘species’ includes 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.” Id. § 1532(16).2 The Secretary has delegated his authority to administer the ESA to the Fish and Wildlife Service (“FWS”). 50 C.F.R. § 402.01 (b). The ESA requires the Secretary to consider five factors when determining whether a species is threatened or endangered: (1) the present or threatened destruction, modification, or curtailment of the species’ range; (2) overutilization for commercial, recreational, scientific, or educational purposes; (3) disease or predation; (4) the inadequacy of existing regulatory mechanisms; and (5) other natural or man-made factors affecting the species’ continued existence. 16 U.S.C. § 1533 (a)(1); 50 C.F.R. § 424.11 (c). The Secretary must reach a listing determination “solely on the basis of the best scientific and commercial data available to him.” 16 U.S.C. § 1533 (b)(1)(A). FACTUAL AND PROCEDURAL BACKGROUND The lizard at issue in this case is “a small, cryptically colored iguanid . . . that is restricted to flats and valleys of the western Sonoran desert.”3 58 Fed. Reg. 62,624 , 62,625 (Nov. 29, 1993). Its natural habitat stretches across parts of southern California, southwestern Arizona, and northern Mexico. In California, the lizard can be found in the Coachella Valley, [2] Whether a species is threatened with extinction in a “significant portion of its range” and whether a “distinct population segment” of a species is threatened with extinction are independent legal questions. See Nw. Ecosystem Alliance v. U.S. Fish & Wildlife Serv., 475 F.3d 1136, 1143 (9th Cir. 2007). Plaintiffs do not contend that any portion of the lizard’s population qualifies for protection as a threatened or endangered “distinct population segment,” and thus we consider only the former question here. [3] The lizard’s formal name is Phrynosoma mcallii. TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR 5947 the west side of the Salton Sea and Imperial Valley, and the east side of the Imperial Valley. In Arizona, the lizard can be found in the Yuma Desert south of the Gila River and west of the Gila and Butler Mountains. 68 Fed. Reg. 331 , 332 (Jan. 3, 2003). Very little is known about the lizard’s range in Mexico, but it has been observed south of the California border in Baja California, south of Arizona from the international border to the Piñacate Region, and around Puerto Peñasco and Bahía de San Jorge, Sonora. FWS estimates that man-made factors are responsible for the destruction of 1,103,201 acres of the lizard’s estimated 4,875,624-acre historic range. 71 Fed. Reg. 36,745 , 36,749-51 (Jun. 28, 2006). Agricultural and urban development have resulted in fragmentation of the lizard’s remaining habitat. 68 Fed. Reg. at 332. Fragmentation creates isolated subpopulations that, because of their reduced size, have an increased probability of extinction. 58 Fed. Reg. 62,624 , 62,626-27 (Nov. 29, 1993). Remaining lizard habitat in the United States can be divided into four distinct geographical sections: the Coachella Valley, west side of the Imperial Valley/Salton Sea, and east side of the Imperial Valley in California, and the Yuma Valley in Arizona. 68 Fed. Reg. at 332. 1993 Listing The Secretary first proposed listing the lizard as threatened in 1993, citing documented and anticipated population declines.4 See 58 Fed. Reg. at 62,624. The proposed listing identified urban expansion, off-highway vehicle (“OHV”) use, energy development, and military activities as the primary threats to the lizard’s remaining habitat, and concluded that three of the five statutory listing criteria had been satisfied. Id. at 62,625-28. [4] In Defenders of Wildlife v. Norton, 258 F.3d 1136, 1138-40 (9th Cir. 2001) (“Defenders”), the court recounted the administrative history preceding the Secretary’s 1993 proposed listing in detail. We do not repeat it here. 5948 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR Upon proposing a species for listing, the ESA requires the Secretary to make a final determination (or formally extend the time for making a final decision) on whether to list within twelve months. See 16 U.S.C. § 1533 (b)(6)(A). That time period lapsed without the Secretary issuing a final decision, and Defenders of Wildlife, along with several of the plaintiffs in this action, sued the Secretary to compel a determination on whether or not to list the lizard. See Defenders, 258 F.3d at 1139 . In response to a court order requiring a final decision within sixty days, the Secretary issued a rule withdrawing the 1993 proposed listing (the “1997 withdrawal”). 62 Fed. Reg. 37,852 (July 15, 1997). The 1997 withdrawal relied, in part, on the adoption of the Flat-Tailed Horned Lizard Conservation Agreement (the “Conservation Agreement”). Id. at 37,853-55 . In 1997, seven state and federal agencies jointly entered into the Conservation Agreement, and agreed to implement a management strategy to address threats to the lizard’s remaining habitat on public lands.5 Id. The parties to the Conservation Agreement agreed to take steps aimed at “reducing threats to the species, stabilizing the species’ populations, and maintaining its ecosystem.” In addition to reliance on the Conservation Agreement, the 1997 withdrawal concluded that population trend data did not conclusively demonstrate significant declines, and that many of the threats cited in the 1993 proposed listing had either been reduced or eliminated. See id. at 37,859-60 . The Secretary also found that known threats to lizard habitat on private lands did not warrant listing, because adequate habitat existed on public lands to ensure the species’ continued viability. See id. at 37,860 . [5] The management strategy revolves around the designation of lizard management areas (“MAs”) on public lands under the jurisdiction of the Conservation Agreement signatories. 62 Fed. Reg. at 37,854. The MAs contain 437,000 acres of lizard habitat — approximately thirty-five percent of the lizard’s remaining domestic range. Id. TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR 5949 Defenders challenged the 1997 withdrawal in the U.S. District Court for the Southern District of California. Defenders, 258 F.3d at 1140 . In 1999, the district court entered summary judgment for the Secretary, and Defenders appealed. Id. We reversed and remanded, concluding that the Secretary had failed to consider whether the lost and threatened portions of the lizard’s range amounted to a “significant portion” of the species’ overall range, as the ESA requires. Id. at 1145 .6 2003 Withdrawal On remand in Defenders, the district court ordered the Secretary to reinstate the 1993 proposed withdrawal within sixty days, and to issue a final decision on listing within twelve months of reinstatement. 66 Fed. Reg. 66,384 (Dec. 26, 2001). Following a 120-day public comment period, the Secretary again issued a rule withdrawing the proposed listing (the “2003 withdrawal”). 68 Fed Reg. 331 (Jan. 3, 2003).7 Plaintiffs then brought this action challenging the 2003 withdrawal, arguing that the withdrawal failed to comply with the remand order in Defenders. In 2005, the district court held that the 2003 withdrawal failed to comply with Defenders because it “assumed without explanation that large swaths of lost habitat were of no significance at all.” The district court rejected Plaintiffs’ other chal6 Defenders also noted that the 1997 withdrawal failed to analyze the putative benefits of the Conservation Agreement and management strategy in a site specific manner. 258 F.3d at 1146 . [7] The 2003 withdrawal noted that FWS solicited the opinion of four independent lizard experts. 68 Fed. Reg. at 340-41. Of the four, two recommended listing the species as threatened, one did not express a firm opinion, and one concluded that listing was not warranted. Id. Kevin Young, the biologist that did not favor listing, stated that a “significant portion of the [lizard’s] range” has indeed been lost, but concluded that listing would likely direct resources away from efforts to protect the species on public lands, and toward unproductive efforts to protect lizard habitat on private lands. 5950 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR lenges to the 2003 withdrawal, concluding that the Secretary’s assessment of threats to the lizard’s current range was reasonable and supported by the evidence in the administrative record. The Secretary did not appeal the remand order. Pursuant to that order, the Secretary withdrew the 2003 withdrawal, and restored the lizard to proposed listing status. 70 Fed. Reg. 72,776 (Dec. 7, 2005). 2006 Withdrawal In 2006, after another round of public comment, the Secretary again issued a rule withdrawing the proposed listing (the “2006 withdrawal”). 71 Fed. Reg. 36,745 (June 28, 2006). The 2006 withdrawal quantified the lizard’s lost range, explained why that range is not “significant” within the meaning of the ESA, and incorporated the findings in the 2003 withdrawal by reference. Id. at 36,749-51 . Plaintiffs then filed a supplemental complaint challenging the 2006 withdrawal. In its final 2007 order, the district court granted summary judgment in favor of the Secretary. Plaintiffs appealed, challenging both the district court’s 2005 order (insofar as it upheld the agency’s assessment of threats to the lizard’s current range) and its 2007 order (upholding the agency’s assessment of the significance of the lizard’s lost range). STANDARD OF REVIEW We review the district court’s summary judgment ruling de novo. Nw. Envtl. Advocates v. Nat’l Marine Fisheries Serv., 460 F.3d 1125 , 1132 (9th Cir. 2006). Under the APA, this court must set aside final agency decisions that are “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706 (2)(A). “Review under the arbitrary and capricious standard ‘is narrow, and [we do] not substitute [our] judgment for that of the agency.’ ” Lands Council v. McNair, 537 F.3d 981, 987 (9th TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR 5951 Cir. 2008) (en banc) (quoting Earth Island Inst. v. U.S. Forest Serv., 442 F.3d 1147, 1156 (9th Cir. 2006)) (alterations in original). The Secretary has an obligation, however, to “state a rational connection between the facts found and the decision made.” Gifford Pinchot Task Force v. U.S. Fish & Wildlife Serv., 378 F.3d 1059 , 1065 (9th Cir. 2004). DISCUSSION I. The Lizard’s Lost Historical Range A. The Secretary’s Compliance with Defenders The district court determined that Defenders requires the Secretary to engage in a two-step process. The Secretary was directed first to quantify the lizard’s historical range in order to establish a “temporal baseline,” and then to determine whether the lost habitat, measured against that baseline, amounts to a “significant portion” of the species’ overall range. On appeal, the Secretary challenges the district court’s entire approach.8 [1] Defenders stated that a species can be extinct “ ‘throughout . . . a significant portion of its range’ if there are major geographical areas in which it is no longer viable but once was.” 258 F.3d at 1145 (ellipsis in original). Before the Secretary can properly consider whether the areas where the lizard is “no longer viable” are significant, it is of course necessary to identify those areas. On appeal, the Secretary clings to his argument that lost historical habitat is largely irrelevant [8] Plaintiffs contend that if the Secretary wished to challenge the district court’s 2005 remand order, he ought timely to have appealed that order immediately after it was issued. See Alesa Valley Alliance v. Dep’t of Commerce, 358 F.3d 1181, 1184 (9th Cir. 2004) (explaining the general rule that only an agency may appeal a court’s remand order). Regardless, the issue on appeal is whether the 2006 withdrawal complied with the requirements laid out in Defenders, and thus we cannot avoid interpreting that case on this appeal. 5952 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR to the recovery of the species, and thus the ESA does not require him to consider it. [2] The Secretary attempts to re-litigate Defenders rather than accept the opinion’s basic conclusions. Tellingly, a memorandum from the Office of the Solicitor, Department of Interior, to the Director of FWS, dated March 16, 2007, acknowledges that Defenders “appears to” require consideration of historic range, but maintains that this court’s interpretation “is not supported by [the ESA].” District courts applying Defenders have also concluded that the holding requires consideration of historic range.9 See Defenders of Wildlife v. Sec’y of Interior, 354 F. Supp. 2d 1156, 1168-69 (D. Or. 2005) (interpreting Defenders to require FWS to analyze the gray wolf’s entire lost historical range before removing the species from the endangered species list); Defenders of Wildlife v. Norton, 239 F. Supp. 2d 9, 20 (D.D.C. 2002) (offering a similar interpretation of Defenders). We agree that Defenders requires the Secretary to analyze lost historical range. To comply with the district court’s 2005 limited remand order, the Secretary quantified the lizard’s range of approximately one hundred years ago — the period before agricultural and commercial development began to encroach on the lizard’s habitat — and identified the overall habitat loss from that baseline. 71 Fed. Reg. at 36,751. The Secretary then concluded that, for reasons we discuss below, the lost portions of [9] A district court in the Tenth Circuit has disagreed with this court’s analysis in Defenders. See Ctr. for Biological Diversity v. Norton, 411 F. Supp. 2d 1271, 1279 (D.N.M. 2005) (holding that the statutory phrase “significant portion of its range” refers to “biological significance,” and crediting the Secretary’s argument — specifically rejected in Defenders — that in order to be “significant,” the species as a whole must be threatened with extinction). In rejecting Defenders’ gloss on the ESA, however, the district court does not suggest or imply that the opinion does not require consideration of historical range. TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR 5953 the lizard’s range are not geographically, biologically, or otherwise significant. Id. Defenders explained that the Secretary maintains a “wide degree of discretion in delineating ‘a significant portion of its range,’ since the term is not defined in the statute.” 258 F.3d at 1145 . However, he “must at least explain [his] conclusion that the area in which the species can no longer live is not a ‘significant portion of its range.’ ” Id.10 Thus, Defenders left the appropriate criteria for testing “significance” undefined, but made clear that the Secretary must develop some rational explanation for why the lost and threatened portions of a species’ range are insignificant before deciding not to designate the species for protection. In the 2006 withdrawal, the Secretary offered a set of reasons for discounting the significance of the lizard’s lost historical range (approximately 23% of the species’ baseline range). He explained that: (1) lizard populations persist across most of the species’ current range despite habitat loss and fragmentation; (2) much of the lizard’s lost habitat was converted to agricultural, commercial, and residential development long ago; (3) the lost portions of the lizard’s range do not carry any special biological or genetic importance for the species as a whole; and (4) lost historical habitat represents a [10] Defenders accorded the Secretary’s interpretation of the statutory phrase “in danger of extinction throughout all or a significant portion of its range” no deference. 258 F.3d at 1145 n.11. The court explained that the Secretary had entirely ignored an aspect of the phrase (the use of the disjunctive “or”), and thus there was no interpretation to defer to. See id. Here, we owe the Secretary’s interpretation of the statute deference under Chevron U.S.A. Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837 (1984), because he offers an interpretation of an ambiguous statutory phrase, and offers that interpretation via formal notice and comment rulemaking. See United States v. Mead Corp., 533 U.S. 218, 230-31 (2001) (holding that, although notice and comment procedures are not required to trigger Chevron deference, such procedures are “significant . . . in pointing to Chevron authority”). 5954 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR relatively small portion of the lizard’s baseline range. See 71 Fed. Reg. at 36,751. Plaintiffs are correct that the Secretary’s stated reasons all rely, to varying degrees, on the premise that lizard populations persist throughout most of the species’ remaining range. For example, the 2006 withdrawal states that lost portions of the lizard’s range are not necessary for the preservation of gene flow because the isolated lizard populations outside of the Coachella Valley “are large enough to be self-sustaining.” 71 Fed. Reg. at 36,748. Elsewhere, the Secretary finds that the habitat lost early in the 20th Century (the majority of the lizard’s overall lost range) is not significant because the species’ “continued persistence over a span of nearly 100 years is a strong indication that the species will continue to persist into the foreseeable future despite the loss of historical habitat.” Id. at 36,751. It is insufficient, under Defenders, to point to one area or class of areas where lizard populations persist to support a finding that threats to the species elsewhere are not significant; the ESA requires a more thorough explanation. See 258 F.3d at 1146 (faulting the Secretary for concluding that the availability of habitat on public lands alone renders lost habitat on private lands insignificant). Plaintiffs argue that this is precisely what the Secretary has done here. [3] The Secretary’s explanations, however, exhibit more nuance than Plaintiffs acknowledge. Although he places considerable weight on the lizard’s persistence throughout most of its remaining range, reliance on persistence is not per se inconsistent with Defenders. The 2006 withdrawal analyzes the lizard’s lost habitat in a site-specific manner, and cites lizard persistence to corroborate its conclusion that the lost portions of the lizard’s range do not provide any unique or critical function for the well-being of the species. Moreover, the Secretary offers two supplemental reasons that do not depend on persistence alone. He cites to a study of migration TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR 5955 between isolated lizard populations and reasonably concludes that the lizard’s lost range does not represent a critical pathway for maintenance of genetic diversity. 71 Fed. Reg. at 36,750. He also notes that most of the lizard’s lost range was converted to agricultural or commercial uses decades ago, is generally not recoverable, and is thus of limited significance to the lizard’s long-term survival. Id. at 36,751. Taking these reasons together, and according the Secretary due interpretive deference, we conclude that his understanding of the phrase “significant portion of its range” does not run afoul of Defenders.11 B. Is the Lizard Persisting Throughout Most of its Range? The next question is whether the administrative record supports the Secretary’s finding that lizard populations outside of the Coachella Valley are persisting. An action will be deemed arbitrary and capricious where the agency offers an explanation for an action “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.” Motor Vehicle Mfrs. Ass’n of U.S. v. State Farm Mut. Auto Ins. Co., 463 U.S. 29, 43 (1983). “While our deference to the agency is significant, we may not defer to an agency decision that ‘is without substantial basis in fact.’ ” Sierra Club v. U.S. EPA, 346 F.3d 955, 961 (9th Cir. 2003) (quoting FPC v. Fla. Power & Light Co., 404 U.S. 453, 463 (1972)). [4] The Secretary argues that FWS relied on population studies to conclude that the lizard “is persisting in the vast [11] Plaintiffs also argue that the Secretary erred in failing to consider the cumulative significance of both the lost historical range, and the currentlythreatened Coachella Valley range. The argument is unavailing. Both the 2003 and 2006 withdrawals offer explanations for the insignificance of threatened and lost range, and those explanations clearly apply to the lizard’s condition as a whole. Moreover, the 2006 withdrawal explicitly incorporates the findings from the 2003 withdrawal, and thus its analysis can be presumed to take the conclusions in that document into account. 5956 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR majority of its range.” Appellee’s Br. at 22. Yet, in the 2003 withdrawal, the Secretary reported that “[i]nformation concerning population dynamics of flat-tailed horned lizard populations is limited and inconclusive.” 68 Fed. Reg. at 332; see also Teresa Woods & Steve Morey, Uncertainty and the Endangered Species Act, 83 Ind. L.J. 529 , 531-32 (2008) (explaining the difficulty in carrying out population studies from which inferences about extinction risk can be made, given budgetary constraints and short regulatory and courtordered time frames for decision-making). Apparently, the Secretary infers from the uncertainty in the population studies that lizard populations “remain[ ] viable throughout most of [the lizard’s] current extant range.” 71 Fed. Reg. at 36,751. This conclusion is conceptually distinct from and relied on different evidence than the Secretary’s determination that the species does not face significant threats in most of its current range. The latter conclusion is premised on an estimate of existing and potential threats to lizard habitat, existing management plans that mitigate those threats, and a variety of other risk assessment factors. The persistence finding, however, relies solely on the conclusion that lizard populations are in fact viable and stable throughout most of the species’ current range. It is this conclusion that ultimately requires reversal. [5] If the science on population size and trends is underdeveloped and unclear, the Secretary cannot reasonably infer that the absence of evidence of population decline equates to evidence of persistence. The absence of conclusive evidence of persistence, standing alone, without persuasive evidence of widespread decline, may not be enough to establish that the Secretary must list the lizard as threatened or endangered. See Cook Inlet Beluga Whale v. Daley, 156 F. Supp. 2d 16, 21-22 (D.D.C. 2001) (holding that the ESA does not require listing “simply because the agency is unable to rule out factors that could contribute to a population decline”); cf. Balt. Gas & Elec. Co. v. Natural Res. Def. Council, Inc., 462 U.S. 87, 103 (1983) (holding that when examining decisions made under TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR 5957 conditions of scientific uncertainty “a reviewing court must generally be at its most deferential”). But this is a different case. The Secretary affirmatively relies on ambiguous studies as evidence of persistence (i.e., stable and viable populations), and in turn argues that this “evidence” of persistence satisfies Defenders’ mandate and proves that the lizard’s lost range is insignificant for purposes of the ESA. This conclusion is unreasonable. The studies do not lead to the conclusion that the lizard persists in a substantial portion of its range, and therefore cannot support the Secretary’s conclusion. [6] Both parties acknowledge that the formerly common “scat count” method of estimating lizard population size has been discredited; thus, that nearly all of the existing data on lizard populations (including FWS’ 1998 Population Viability Assessment — which itself concluded that population trends could not be reliably determined) must be set aside as unhelpful. See 68 Fed. Reg. at 333 (“The relationship between scat counts and lizard abundance is unclear, or weak at best.” (citations to scientific studies omitted)). Results from studies utilizing an allegedly more accurate “capture-mark-recapture” methodology are just now emerging. 68 Fed. Reg. at 333. In the 2006 withdrawal, the Secretary cites one such study for the proposition that, between 2003 and 2005, in two discrete sections of the lizard’s current range (both within designated lizard MAs), there is no evidence of a “large decline in population” for the areas for which the researchers had more than one year of data. 71 Fed. Reg. at 36,751. This single attenuated finding represents the extent of the agency’s evidentiary support for its sweeping conclusion that viable lizard populations persist throughout most of the species’ current range.12 [12] The Secretary states that viable populations persist in most of the lizard’s current range, and we can therefore infer that he believes that populations persist in Mexico (where an estimated 59% of remaining lizard habitat exists). 68 Fed. Reg. at 332. Yet, the Secretary has not produced any evidence on lizard population outside of the United States; indeed, the agency acknowledges that “the distribution of the species in Mexico is poorly understood . . . .” Id. 5958 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR Contrary to the lesson the Secretary draws from the study (i.e., that lizard populations in the study areas show no sign of decline), the author’s primary conclusion is that the study’s population estimates can serve as a “baseline for future monitoring.” Further, the study’s author warns that the population estimates it reports “should be viewed with caution as they were based on sparse data.” We thus conclude that the administrative record does not support the Secretary’s determination that lizard populations persist throughout most of the species’ current range. The Secretary’s erroneous reliance on lizard persistence, however, does not end our inquiry. See 5 U.S.C. § 706 (stating that courts reviewing agency decisions should take “due account . . . of the role of prejudicial error”). “In circumstances where an agency errs, we may evaluate whether such an error was harmless.” Gifford Pinchot Task Force, 378 F.3d at 1071; see also Nat’l Ass’n of Home Builders v. Defenders of Wildlife, 127 S. Ct. 2518, 2530 (2007) (citing with approval PDK Labs, Inc. v. U.S. DEA, 362 F.3d 786, 799 (D.C. Cir. 2004) (“In administrative law, as in federal civil and criminal litigation, there is a harmless error rule . . . .” )). We have held that the harmless error doctrine “may be employed only ‘when a mistake of the administrative body is one that clearly had no bearing on the procedure used or the substance of decision reached.’ ” Gifford Pinchot Task Force, 378 F.3d at 1071 (quoting Buschman v. Schweiker, 676 F.2d 352, 358 (9th Cir. 1982)); see New York, New Haven & Hartford R.R. First Mortgage 4% Bondholders Comm. v. United States, 289 F. Supp. 418, 440 (S.D.N.Y. 1968) (Friendly, J.) (remanding where an agency’s erroneous factual finding “went to the heart of its analysis”) (“Bondholders Comm.”); cf. Carnegie Natural Gas Co. v. FERC, 968 F.2d 1291 , 1294 (D.C. Cir. 1992) (“We will . . . sustain an agency decision resting on several independent grounds if any of those grounds validly supports the result, unless there is reason to believe the combined force of these otherwise independent grounds influenced the outcome.”). TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR 5959 We thus must determine whether the Secretary’s stated reasons — after setting aside the erroneous persistence finding — would have persuaded him that the lizard’s lost historical range is not significant. The Secretary’s conclusion that the lizard’s lost range holds no critical genetic value for the species finds some support in the record, as does his determination that much of the lizard’s lost historical range was converted to other uses decades ago and is thus not recoverable.13 Neither reason is entirely dependent on lizard persistence. [7] Nonetheless, the 2006 withdrawal repeatedly refers to lizard persistence as persuasive evidence that the species’ lost historical range is not significant. The Secretary offers persistence as both an independent, and indeed primary, basis for discounting the importance of lost range, and as support for several other key conclusions. Because a reliance on the lizard’s persistence throughout most of its current range cuts to “the heart of [the agency’s] analysis,” Bondholders Comm., 289 F. Supp. at 440 , we cannot readily say that the erroneous finding clearly had no bearing on the Secretary’s ultimate decision to withdraw the proposed listing. On remand, the Secretary may be persuaded that, absent reliable evidence of population persistence, the lizard’s lost historical range is indeed significant. II. Threats to the Lizard’s Current Range [8] The 2003 withdrawal reports that the Coachella Valley in California is the only segment of the lizard’s current range where the species is in immediate jeopardy. See 68 Fed. Reg. at 345. The Secretary concluded that the likely extinction of [13] The conclusion that the overall lost range represents a relatively small portion of the lizard’s overall range (approximately 23%) — although not entirely dependent on lizard persistence — amounts to a conclusory and unhelpful factor for determining the significance of the species’ lost range. That is, the Secretary has not explained why the loss of nearly one-quarter of the species overall range is “relatively small.” 5960 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR the Coachella Valley lizard population is not significant because of the population’s relatively small size (approximately sixteen thousand acres, or, one percent if the lizard’s remaining range, 68 Fed. Reg. at 334), isolation from other remaining lizard populations, and lack of importance for maintenance of genetic diversity. See id. at 348. In essence, the Secretary determined that the Coachella Valley is home to a very small (in comparison with the three other segments of the lizard’s remaining domestic range) non-unique lizard population. As such, the demise of the Coachella Valley population would not be significant within the meaning of the ESA. This is a reasonable approach to assessing the significance of threatened range, and one that complies with Defenders. See 258 F.3d at 1145 (requiring only that the Secretary “explain [his] conclusion that the area in which [the lizard] can no longer live is not a ‘significant portion of its range’ ”). Plaintiffs also challenge the Secretary’s assessment of threats to the lizard’s remaining habitat outside of the Coachella Valley. [9] First, they object to the Secretary’s assertion that the Conservation Agreement and management strategy have diminished threats to the lizard’s remaining habitat on public lands, and emphasize the management strategy’s slow and still incomplete implementation. The Secretary acknowledges that the Conservation Agreement has not yet been fully implemented, but points to specific conservation benefits that the agreement has achieved since it came into being in 1997. See 68 Fed. Reg. at 347 (noting that pesticide use and OHV racing have been limited in the lizard MAs). Moreover, the 2003 withdrawal states that its assessment of threats to the species’ current range is not “dependent on full implementation” of the Conservation Agreement’s management strategy. Id. Kevin Young, one of the four independent biologists that reviewed FWS’ assessment of the lizard’s viability, offers support for the Secretary’s position. He opined that the Conservation Agreement’s protective measures alone were “adequate” to TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR 5961 protect the species’ remaining domestic range (outside of the Coachella Valley). We conclude that the limited benefits that the 2003 withdrawal points to are supported by the record, and the Secretary did not err in taking the Conservation Agreement into account. Next, Plaintiffs argue that OHV use in the lizard’s current range presents a much greater threat to the species than the Secretary acknowledges. Plaintiffs’ argument, however, relies on inferences from indeterminate scientific evidence. Plaintiffs have shown that OHV use is on the rise throughout the lizard’s remaining range.14 The available studies analyzing OHV-related impacts on the lizard, however, do not conclusively show that OHV use amounts to a significant threat to the species’ viability.15 Both Plaintiffs and the Secretary point to scientific studies supporting their respective views on the effects of OHVs, but the merits of the conflicting studies is not a proper subject for this court to resolve. See Nw. Ecosystem Alliance v. U.S. Fish & Wildlife Serv., 475 F.3d 1136, 1150 (9th Cir. 2007) (“[W]e must defer to the agency’s interpretation of complex scientific data.”); see also Marsh v. Or. Natural Res. Council, 490 U.S. 360, 378 (1989) (“When 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.”). [14] A study that the Secretary relies on reports that over one-quarter of the lizard’s remaining habitat in California is open to legal OHV recreation, and illegal OHV recreation in closed areas is “not uncommon.” [15] Although the notion that lizards and OHVs can co-exist in the same immediate area is counter-intuitive, the Secretary directs us to a recent study reporting that lizards can be found in areas heavily impacted by OHVs, and another reporting frequent lizard observations in an area of high Border Patrol traffic. On the other hand, Plaintiffs point to a study suggesting that fewer lizards can be found in areas where OHVs regularly operate. 5962 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR [10] Finally, Plaintiffs challenge the Secretary’s treatment of scattered threats posed by energy and mineral development projects, increased Border Patrol activity, and the possible construction of large-scale infrastructure projects in the lizard’s current range. Plaintiffs’ arguments follow the same course as their attack on the agency’s analysis of OHV use. In short, they have not presented conclusive evidence to rebut the Secretary’s determination that such threats, either alone or in concert, are not likely to cause the “destruction, modification, or curtailment of [the species’] habitat or range.” 16 U.S.C. § 1533 (a)(1)(A). CONCLUSION For the reasons set forth above, we reverse and remand the judgment of the district court with instructions that the matter be further remanded to the Secretary so that the Secretary can again consider whether to withdraw the proposed listing of the lizard. REVERSED and REMANDED. NOONAN, Circuit Judge, concurring and dissenting: This case began in 1993 when the Secretary of the Interior first proposed listing the species. After careful and conscientious consideration by this court, it is now in 2009 remanded to continue to be litigated for an indefinite time. The pattern of the litigation is scarcely unfamiliar in environmental cases. Congress has enacted law designed to conserve species of wildlife threatened with extinction. 16 U.S.C. § 1531 et seq. A federal agency has been entrusted with enforcement. Using its expertise, the agency has determined what protection should be afforded a particular species. Its determination has been challenged by a private nonprofit organizations concerned with the existence of the species. The district court has TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR 5963 heard the resulting litigation more than once, and this court has heard it more than once. The various decisionmakers and participants — the agency, the nonprofits, the district court, and the court of appeals — are not motivated by private passion or grudge, but seek to see the fair application of broad legislation to highly particularized and often elusive data. The legal system does not confide the definitive judgment to the agency entrusted with enforcement of the law but subjects that judgment first to the challenges of the nongovernmental organizations and then to the supervision of judges who are not expert in the scientific matters at stake and not familiar with the species whose survival is at stake. As if this interplay of governmental and private groups did not create room for tension, misunderstandings, and passionate disagreement, the problems in this case have been exacerbated by the simple absence of information. How many flattailed horned lizards are there? No one knows the answer to that question. Nor does anyone know how many lizards disappeared when portions of their range disappeared. It is supposed that a diminution in range correlates with a diminution in lizards. This hypothesis is plausible. It has not been shown to be probable. Yet the case turns on what measures are necessary to keep this unknown population in existence. The court concludes that the Secretary erred in finding that the lizard has not lost a significant portions of its range. The old method of counting lizards is out. A new method has not been tried very much. It’s anybody’s guess whether the lizards are multiplying or declining. In a guessing contest one might defer to the government umpire. The court, however, finds the Secretary’s conclusion impacted by over-reliance on fragmenting evidence of the lizard’s persistence; so the court decides to give the Secretary another crack at the problem. If the Secretary does not know what the lizard population was to begin with, or what it was in 1993, or what it is now 5964 TUCSON HERPETOLOGICAL SOCIETY v. SALAZAR in May 2009, how will he know if it is increasing, staying the same, or declining? A style of judging, familiar to readers of the old English reports, characterizes the judge as dubitante. That is probably the most accurate term for me, which leads me to concur in the majority opinion insofar as it rejects the contentions of the Tucson Herpetological Society and to dissent from the remand whose command to the Secretary of the Interior is, Guess again.