W. FLETCHER, Circuit Judge

Rebecca Kay Smith (argued), Public Interest Defense Center, Missoula, Montana; Timothy M. Bechtold, Bechtold Law Firm, Missoula, Montana; for Plaintiffs–Appellants.

Jeffrey S. Beelaert (argued), Travis J. Annatoyn, Allen M. Brabender, and Andrew C. Mergen, Attorneys, Environment and Natural Resources Division; Jeffrey H. Wood, Acting Assistant Attorney General; Kathryn Williams–Shuck, United States Department of the Interior; Alan Campbell, United States Department of Agriculture; Mark S. Smith, Assistant United States Attorney; United States Department of Justice, Washington, D.C.; for Defendants–Appellees.

Before: Ferdinand F. Fernandez, William A. Fletcher, and Michael J. Melloy, Circuit Judges.

W. FLETCHER, Circuit Judge:

This appeal challenges the United States Forest Service's proposed Lonesome Wood Vegetation Management 2 Project ("Lonesome Wood 2" or "project"), designed to reduce the threat of wildfire in a populated area of the Gallatin National Forest in Montana. If implemented, Lonesome Wood 2 would entail thinning just over 2,500 acres of forest land, including 495 acres of old growth forest. Appellants Native Ecosystems Council and the Alliance for the Wild Rockies (collectively, "Council") brought suit in federal district court to enjoin the project, contending that it violates the Endangered Species Act ("ESA"), the National Forest Management Act ("NFMA"), the National Environmental Policy Act ("NEPA"), and the Administrative Procedure Act ("APA"). The district court initially enjoined the project, but after twice remanding the case to the Forest Service to remedy defects in Biological Opinions concerning two listed species under the ESA, it granted the Forest Service's motion to dissolve the injunction. We affirm.

I. Background

In 2005, the Forest Service assessed the risk of wildfire near Hebgen Lake in the Gallatin National Forest. The Forest Service concluded that accumulation of fuel in the area posed a serious risk to the private homes, campgrounds, and recreational areas near the lake. To mitigate the risk, the Forest Service developed and approved a project to thin large trees on about 1,750 acres, including about 495 acres of old-growth forest; to thin small trees on about 825 acres; to slash and/or selectively burn on about 325 acres; and to build about six miles of temporary roads.

The Forest Service issued an Environmental Assessment for Lonesome Wood 2 in December 2007, followed by a Decision Notice ("DN") and Finding of No Significant Impact ("FONSI") in April 2008. The Council filed suit challenging the project in January 2009. Before the district court could rule, grizzly bears were relisted as a threatened species under the ESA, resulting in different consultation and management criteria. The Forest Service withdrew the DN and FONSI in order to prepare an Environmental Impact Statement. A Final Environmental Impact Statement ("FEIS") was issued in October 2012, and a Record of Decision ("ROD") approving the project was issued in December 2012.

The Council again challenged the project, filing suit in March 2013 and alleging violations of ESA, NFMA, NEPA, and the APA. The parties filed cross-motions for summary judgment. On December 5, 2014, the district court granted partial summary judgment to the Council on its ESA claim. The district court concluded that the United States Fish and Wildlife Service ("FWS"), in its Biological Opinions ("BiOps") evaluating the effect of Lonesome Wood 2 on two listed species—grizzly bears and Canada lynx—did not perform a site-specific analysis of the project's impact. Rather, FWS relied entirely on the project having satisfied the criteria for an exemption for fuel-treatment projects near human habitation, which exempted the project from otherwise-applicable standards. (We describe the exemption in detail in Section III.A., infra .) The court enjoined the project and remanded for preparation of site-specific BiOps. It granted partial summary judgment to the Forest Service on the Council's remaining claims.

On July 1, 2015, FWS submitted new BiOps. On August 31, 2015, the district court again held the BiOps inadequate because they continued to rely on the project having satisfied the criteria for the exemption. On April 12, 2016, FWS submitted a third pair of BiOps. These BiOps specifically addressed the project's expected environmental effects. The BiOp dealing with the lynx concluded that the "the Lonesome Wood 2 project, as proposed, is not likely to jeopardize the continued existence of Canada lynx." On July 7, 2016, the court concluded that the April 2016 BiOps were sufficiently site-specific to satisfy the Forest Service's consultation obligation under the ESA. The court dissolved the injunction and allowed the project to proceed.

The Council appealed the district court's order granting partial summary judgment to the Forest Service and its later order dissolving the injunction. We have jurisdiction over the entirety of the appeal. See 28 U.S.C. § 1291 ; Alsea Valley All. v. Dep't of Commerce , 358 F.3d 1181, 1184 (9th Cir. 2004).

II. Standard of Review

We review an agency's compliance with the ESA, NFMA, and NEPA under the "arbitrary and capricious" standard of the APA. Defs. of Wildlife v. Zinke , 856 F.3d 1248, 1256–57 (9th Cir. 2017) (ESA); Native Ecosystems Council v. Tidwell , 599 F.3d 926, 932 (9th Cir. 2010) (NFMA); Lands Council v. Powell , 395 F.3d 1019, 1026 n.5 (9th Cir. 2005) (NEPA). The APA "requires an agency action to be upheld unless it is found to be ‘arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.’ " Zinke , 856 F.3d at 1256–57. "An agency's action is arbitrary and capricious if the agency fails to consider an important aspect of a problem, if the agency offers an explanation for the decision that is contrary to the evidence, if the agency's decision is so implausible that it could not be ascribed to a difference in view or be the product of agency expertise, or if the agency's decision is contrary to the governing law." Powell , 395 F.3d at 1026 (internal citations omitted).

We review summary judgment rulings de novo. Zinke , 856 F.3d at 1256. "We review the validity of a district court's order granting dissolution of an injunction for an abuse of discretion." N. Alaska Envtl. Ctr. v. Lujan , 961 F.2d 886, 889 (9th Cir. 1992). A district court abuses its discretion when it makes an error of law, which we review de novo. Id.

III. Discussion

A. Endangered Species Act

The ESA requires federal agencies to ensure that their actions are not "likely to jeopardize the continued existence of any endangered species or threatened species," using the "best scientific and commercial data available." 16 U.S.C. § 1536(a)(2). The Council challenges an exemption contained in the 2007 "Lynx Amendments" to the Gallatin Forest Plan, contending that it is not based on the "best scientific ... data available."

In 2000, after eight years of litigation, Canada lynx were listed by FWS as a threatened species under the ESA. In 2001, the Forest Service signed a Lynx Conservation Agreement with FWS under which the Forest Service agreed not to proceed with any projects that would be "likely to adversely affect" Canada lynx until its forest plans were amended. In 2004, the Forest Service issued a Draft Environmental Impact Statement containing proposed amendments to forest plans that would protect Canada lynx habitat. In March 2007, after consultation with FWS under Section 7 of the ESA, the Forest Service issued an FEIS and ROD adopting the Northern Rockies Lynx Management Direction (the "Lynx Amendments") to govern its management of Canada lynx habitat. The Lynx Amendments were then incorporated into the forest plans for the Gallatin and seventeen other national forests in Idaho, Montana, Wyoming and Utah.

In 2007, when the Lynx Amendments were adopted, FWS had designated 1,841 square miles as critical habitat for the Canada lynx, 1,389 square miles of which were located in the Northern Rocky Mountains "critical habitat unit." None of the designated critical habitat was located within a national forest. Cottonwood Envtl. Law Ctr. v. U.S. Forest Serv. , 789 F.3d 1075, 1077–78 (9th Cir. 2015). Shortly thereafter, FWS announced that its critical habitat determination had been "improperly influenced by then deputy assistant secretary Julie MacDonald and, as a result, may not be supported by the record, may not be adequately explained, or may not comport with the best available scientific and commercial information." Id. at 1078 (citation and internal quotation marks omitted).

In 2009, FWS revised its Canada lynx critical habitat designation upward from 1,841 to 39,000 square miles, more than 10,000 square miles of which were in the Northern Rocky Mountain critical habitat unit. Id. A significant amount of the newly designated Northern Rocky Mountain critical habitat is located in eleven national forests. Id. In 2015, we affirmed a district court determination that the designation of Canada lynx critical habitat in these national forests required reinitiation of Section 7 consultation with FWS and possible revision of the Lynx Amendments. Id. at 1077.

The Lynx Amendments limit human activity in national forests in order to protect Canada lynx habitat. However, the Amendments contain an exemption for Forest Service fuel treatment projects in the wildland urban interface ("WUI") if they satisfy certain criteria. These criteria are that the project take place within one mile of a human community; that the totality of such projects affect no more than six percent of lynx habitat in any three adjoining lynx analysis units within a particular national forest; and that the totality of such projects affect no more than six percent of lynx habitat in the entirety of any particular national forest. It is uncontested that Lonesome Wood 2 satisfies the criteria for the WUI exemption and that the project area is not designated as Canada lynx critical habitat.

The Council contends that the Amendments' exemption from the otherwise-applicable standards for fuel treatment projects in the WUI is not based on the "best scientific ... data available," in violation of the ESA. The Council argues that a thesis prepared by Megan Kosterman in December 2014, in partial fulfillment of the requirements for the degree of Master of Science in Wildlife Biology at the University of Montana, compels revision or elimination of the WUI exemption. As summarized in its "abstract," Ms. Kosterman's thesis concludes:

The most important predictors for overall lynx reproductive success within occupied female home ranges were the connectivity of mature forest, intermediate (10–15%) amounts of young regenerative forest, young regenerating forest patches with low perimeter-area ratios, and the adjacency of mature forest to young regenerating forest types. Female lynx home ranges that contain greater than 50% mature forest and approximately 10–15% young regenerating forest appear to be the optimal composition of forest structure types. ... Incorporating these results into current and long-term land management plans will provide a valuable conservation tool to ensure the persistence of threatened Canada lynx populations in the western US.

Megan K. Kosterman, Correlates of Canada Lynx Reproductive Success in Northwestern Montana, at iii (2014).

The Forest Service responds that it has considered Ms. Kosterman's thesis, and has concluded that it does not require the Forest Service to withdraw its 2012 ROD approving Lonesome Wood 2. On April 20, 2016, Regional Forester Leanne Marten wrote a letter to one of the two appellants in this case:

Thank you for your March 18, 2016 email correspondence regarding the [Lynx Amendments] and the 2014 thesis produced by Megan Kosterman. My staff in the Regional Office have been fully aware of the thesis for some time. They have been working with Ms. Kosterman herself, the US Fish and Wildlife Service and scientists from the Forest Service Rocky Mountain Research Station (RMRS) to understand the relationship between the 2014 thesis and the [Lynx Amendments].

While the 2014 thesis provides valuable new information with potential to inform changes in Forest Service management of lynx and lynx habitat, the relationships between vegetation composition and lynx reproduction success in NW Montana described in the thesis are not well enough understood to determine if, or what, specific changes in management direction are warranted.

Marten praised Ms. Kosterman's thesis as a "success," but pointed out that her classifications of vegetation were different from the classifications used in the Lynx Amendments. Ms. Kosterman's classifications of vegetation were "deliberately imprecise in order to allow [her] to correlate lynx demography to habitat in a simple and rough sense," which made it difficult to use the thesis in assessing the efficacy of the Lynx Amendments. In Marten's words, "[T]he parameters and metrics used to demonstrate this relationship do not cross-walk well to the metric standards provided in the [Lynx Amendments]." Marten provided two examples of the differences in classification that, in her view, limited the utility of Ms. Kosterman's thesis. She then concluded:

Ms. Kosterman and RMRS scientists are working to publish her work in a peer reviewed scientific journal and it should also be noted that some of the findings in the original thesis may change through that process. Please also note that we fully recognize the importance of this study and its implication for improved management direction.

"The [ESA's] best available data requirement ... prohibits an agency from disregarding available scientific evidence that is in some way better than the evidence it relies on." Kern Cty. Farm Bureau v. Allen , 450 F.3d 1072, 1080 (9th Cir. 2006) (internal quotation marks and alterations omitted). "The standard does not, however, require an agency to conduct new tests or make decisions on data that does not yet exist." San Luis & Delta–Mendota Water Auth. v. Locke , 776 F.3d 971, 995 (9th Cir. 2014). "The determination of what constitutes the ‘best scientific data available’ belongs to the agency's ‘special expertise .... When examining this kind of scientific determination, as opposed to simple findings of fact, a reviewing court must generally be at its most deferential.’ " San Luis & Delta–Mendota Water Auth. v. Jewell , 747 F.3d 581, 602 (9th Cir. 2014) (quoting Baltimore Gas & Elec. Co. v. NRDC , 462 U.S. 87, 103, 103 S.Ct. 2246, 76 L.Ed.2d 437 (1983) ).

The Council argues that the conclusions of Ms. Kosterman's thesis are fatally inconsistent with the application of the WUI exemption, in the context of the specific project approved in Lonesome Wood 2. In light of the deference that we owe to the agency's expertise, and in light of FWS's site-specific BiOp analyzing the effect of Lonesome Wood 2 on the Canada lynx, we disagree. We decline to overrule the Forest Service's determination that Ms. Kosterman's thesis does not require it to reevaluate its approval of the project.

B. National Forest Management Act

The National Forest Management Act ("NFMA") requires that all national forests operate under "land and resource management plans," 16 U.S.C. § 1604(a), or "Forest Plans," and that all individual management actions within a forest unit "be consistent with each forest's overall management plan." Native Ecosystems Council v. U.S. Forest Serv. , 428 F.3d 1233, 1249 (9th Cir. 2005) ; see also 16 U.S.C. § 1604(i ). The Council contends that the Forest Service is in violation of the Forest Plan for the Gallatin National Forest, and therefore that action under Lonesome Wood 2 may not proceed.

The Forest Service promulgated the Forest Plan for the Gallatin National Forest in 1987. Then-governing Forest Service regulations required the Forest Service "to maintain viable populations of existing native and desired non-native vertebrate species in the planning area." 36 C.F.R. § 219.19 (1987). The Forest Service was also required to "identif[y] and select[ ] ... management indicator species," defined as species whose "population changes are believed to indicate the effects of management activities." Id. § 219.19(a)(1). The regulations required the Forest Service to monitor the management indicator species' population trends and "relationships to habitat changes." Id. § 219.19(a)(6). On November 9, 2000, these regulations were rescinded and replaced. See Native Ecosystems Council v. U.S. Forest Serv. , 428 F.3d 1233, 1237 (9th Cir. 2005). If a Forest Plan incorporates regulations that are later rescinded, the Forest Service must continue to comply with those regulations until the Plan is amended. See In re Big Thorne Project , 857 F.3d 968, 974 & n.3 (9th Cir. 2017).

Under the heading "Goals," the Gallatin Forest Plan provides, "The goals for the Gallatin National Forest are: ... 7. Provide habitat for viable populations of all indigenous wildlife species and for increasing populations of big game animals." Under the heading "Forest–Wide Standards," the Plan provides, " ‘Indicator species,’ which have been identified as species groups whose habitat is most likely to be affected by Forest management activities, will be monitored to determine population change." The Plan further provides that the Forest Service will "[d]etermine population trends of indicator species and relationships to habitat changes," specifically citing 36 C.F.R. § 219.19(a)(6). The Council contends that these provisions of the Forest Plan incorporate the now-rescinded regulations, and that the Forest Service is not in compliance with them. See 16 U.S.C. § 1604(i ) ; Native Ecosystems Council v. U.S. Forest Serv. , 418 F.3d 953, 961 (9th Cir. 2005). The Council further contends that this failure to comply renders the Forest Service's approval of Lonesome Wood 2 unlawful.

Many required actions under Forest Plans are not "final agency actions" under the APA and therefore are not subject to free-standing challenges. Neighbors of Cuddy Mountain v. Alexander , 303 F.3d 1059, 1067 (9th Cir. 2002). Such actions may be challenged, however, as part of a challenge to "a specific, final agency action, the lawfulness of which hinges on these practices." Id. The Forest Service has not argued on appeal that alleged violations of the above-described provisions of the Forest Plan are unconnected to the lawfulness of Lonesome Wood 2. We regard the failure of the Forest Service to advance this argument as a tacit factual concession that Lonesome Wood 2—in particular, its thinning of just over 2,500 acres of trees—is sufficiently connected to the alleged violations of the Forest Plan to satisfy the relatedness test articulated in Neighbors of Cuddy Mountain . We therefore proceed to consider whether the Forest Service has violated the above-described provisions of the Gallatin Forest Plan.

1. Species Viability

The Council argues that the Forest Service failed to comply with a Forest Plan obligation to ensure species viability. The Forest Plan sets a "goal" of "[p]rovid[ing] habitat for viable populations of all indigenous wildlife species and for increasing populations of big game animals." The Council argues both that the Lonesome Wood 2 project is incompatible with this goal, and that the goal incorporates the 1987 regulation requiring USFS "to maintain viable populations of existing native and desired non-native vertebrate species in the planning area." 36 C.F.R. § 219.19 (1987). We disagree with both arguments.

The definitions section of the Forest Plan defines "goal" as

[a] concise statement that describes a desired condition to be achieved. It is normally expressed in broad, general terms and is timeless in that it has no specific date by which it is to be completed. Goal statements form the principal

basis from which objectives are developed.

We apply Auer deference to USFS's interpretation, adopting the agency's interpretation of parts of the Plan that are "susceptible to more than one meaning unless the interpretation is plainly erroneous or inconsistent with the [Plan]." Siskiyou Reg'l Educ. Project v. U.S. Forest Serv. , 565 F.3d 545, 555, 555 n.9 (9th Cir. 2009).

The Forest Service argues that the Forest Plan "goals" are merely aspirational and thus impose no obligations on USFS. We are unwilling to go so far. The only indication in the just-quoted definition that "goals" are only aspirational—that is, purely optional—is the word "desired." That word could conceivably be read to indicate a mere preference. This reading, however, fits poorly with other parts of the Plan discussing the relationships among goals, objectives, and standards. The Plan explains, for example, that "[a]n objective forms the basis for further planning, to define the precise steps to be taken and the resources to be used in achieving identified goals ." (Emphasis added.) Another part of the Plan explains that the "objectives" listed in the Forest Plan are "brief summaries" of how the forest is to be managed, and that "[a] complete understanding of the management direction can be attained by reading the Forest-wide goals and standards ...." (Emphasis added.) These and similar passages make clear that the Forest Plan's "goals" are not just a wish list that imposes no obligations.

Nevertheless, we hold that Lonesome Wood 2 does not violate the goal at issue here. First, the Lonesome Wood 2 project is not incompatible with the Forest Plan. The Plan's definition of "goals" allows flexibility in the manner and timing of their achievement. Under the Plan, the Forest Service fulfills its obligations as long as it does not take actions incompatible with the stated goal of "[p]rovid[ing] habitat for viable populations of all indigenous wildlife species and for increasing populations of big game animals" in the forest as a whole. We can imagine projects that would be incompatible with this goal. But the Council does not allege, and has offered no evidence to suggest, that Lonesome Wood 2 is such a project. Second, the goal at issue does not incorporate the 1987 viability regulation. The goal neither cites § 219.19 nor tracks its language.

2. Monitoring of Management Indicator Species

The Council further argues that the Forest Service failed to comply with its Gallatin Forest Plan obligation to monitor population trends for two management indicator species ("MIS"). The Plan incorporates the 1987 MIS monitoring regulation, both mirroring the regulation's language and citing the relevant provision. Under the Plan, the Service is required to "determine population trends of indicator species and relationships to habitat changes" on a forest-wide basis every five years with a "moderate" degree of "expected precision" and "expected reliability." 36 C.F.R. § 219.12(k)(4a–4c).

In 2011, the Forest Service issued a Management Indicator Species Assessment ("2011 Assessment" or "Assessment") for six MIS designated in the Gallatin Forest Plan—the grizzly bear, bald eagle, elk, wild trout, northern goshawk, and pine marten. The 2011 Assessment updates information provided in a previous Forest Service report covering the 2004–2006 period. The Council contends that the Forest Service's monitoring of the goshawk and pine marten, as described in the Assessment, does not comply with its MIS obligations under the Forest Plan.

The 2011 Assessment describes the northern goshawk, a forest raptor, as an "old growth dependent species, dry Douglas fir sites." The Assessment recounts that during the 2005 breeding season, the Forest Service conducted a survey of goshawks in its Northern Region, comprising thirteen national forests and grasslands in Idaho, Montana, and the Dakotas. The survey sampled 114 Potential Sampling Units ("PSUs") in the Region, of which ten were located in the Gallatin National Forest. The survey detected goshawks in about forty percent of the PSUs overall. It detected goshawks in two of the ten PSUs in the Gallatin, and found an active nest in an additional PSU. After 2005, surveys of goshawks in the Gallatin were performed in connection with specific projects, and detection data from those surveys were compiled in 2010. Eighteen sites were resurveyed in 2010, and goshawks were detected at eight of those sites. The Forest Service did not initiate the 2010 survey until mid-July due to the wet and cold weather, so data from that count should be interpreted cautiously. The 2011 Assessment concluded that "goshawks are present and distributed across the Gallatin National Forest, but population trends cannot be determined from existing data."

The FEIS for Lonesome Wood 2 recounts that in 2000 a goshawk was detected in the Trapper Creek drainage, about 3/4 miles southwest of the closest fuel treatment area under the project. Surveys of the project area were conducted in 2003, 2004, 2007, 2010, 2011, and 2012, but no goshawks were detected. An "auditory response" was heard on July 15, 2010, but the response was not conclusive because gray jays can mimic a goshawk call. The FEIS recounts that "inherent weaknesses exist in surveying for goshawks," for "nest sites can be difficult to locate," and "outside of the breeding season, goshawks are largely silent."

The 2011 Assessment described pine marten, a fur-bearing species, as an "old growth dependent species, moist spruce sites." Pine marten are trapped commercially. All pine marten pelts are registered and tagged, which allows a year-to-year count of trapped animals. The statewide numbers for pine marten pelts during the fifteen years preceding the Assessment range between 653 and 1,323 per year. The 2008/09 state-wide number was 844. From 1999 to 2009, 760 pine martens were trapped in the Gallatin National Forest. Just over 40 were trapped in 1999–2001; just over 60 in 2002–2004; just under 100 in 2005–2007; and just over 60 in 2008–2009. The Assessment recounts that the numbers for the Gallatin "seem to parallel statewide trends." Further, the Assessment recounts that the Forest Service "has cooperated with Wild Things Unlimited ... to conduct winter carnivore track and/or camera surveys every year since 1997." These surveys "showed that pine marten are very common in the Gallatin and Madison Mountain Ranges[.]"

Citing a 2009 study, the FEIS recounts that the pine marten population has been "relatively stable or slightly declining ... on a statewide basis." The FEIS recounts further that pine marten were observed in the project area along Hebgen Lake Road in 2010. Under the preferred alternative, pine marten habitat would be reduced in the project area, "but remaining habitat would continue to support martens in the analysis area."

Based on the foregoing, we conclude that the Forest Service has sufficiently monitored these two MIS to comply with its obligation under the Forest Plan.C. National Environmental Policy Act

The National Environmental Policy Act ("NEPA") requires agencies to prepare an Environmental Impact Statement ("EIS") for any agency action that "significantly affect[s] the quality of the human environment." 42 U.S.C. § 4332(C). An EIS complies with NEPA if it shows that the agency took a "hard look" at the environmental consequences of its proposed action. Native Ecosystems Council , 418 F.3d at 960 (quoting Klamath–Siskiyou Wildlands Ctr. v. Bureau of Land Mgmt. , 387 F.3d 989, 993 (9th Cir. 2004) ). "In reviewing the adequacy of an EIS, we apply the ‘rule of reason’ standard, which requires ‘a pragmatic judgment whether the EIS's form, content and preparation foster both informed decision-making and informed public participation.’ " Id. ( quoting California v. Block , 690 F.2d 753, 761 (9th Cir. 1982) ). An agency fails to meet its "hard look" obligation when it "rel[ies] on incorrect assumptions or data" in drafting an EIS or presents information that is "so incomplete or misleading that the decisionmaker and the public could not make an informed comparison of alternatives." Id. at 964–65. "It surely follows that the data the Forest Service provides to the public to substantiate its analysis and conclusions must also be accurate." WildEarth Guardians v. Mont. Snowmobile Ass'n , 790 F.3d 920, 926 (9th Cir. 2015). The Council contends that the FEIS for Lonesome Wood 2 is misleading or inaccurate in several respects and is therefore deficient.

First, the Council contends that the FEIS provides an incomplete and misleading description of a 2005 peer-reviewed article by Susan Patla, "Monitoring Results of Northern Goshawk Nesting Areas in the Greater Yellowstone Ecosystem: Is Decline in Occupancy Related to Habitat Change?," published in The Journal of Raptor Research , 39 J. Raptor Res. 324 (2005). The Council complains that while the FEIS discusses the article, naming Patla and the date of publication, it neither provides the title nor indicates that the article was peer reviewed. The Council further complains that the FEIS criticizes the Patla article by relying on the analysis in a memorandum without revealing that the memorandum was prepared by a Forest Service employee, was unpublished, and was not peer reviewed. Finally, the Council complains that the FEIS does not disclose the fact that the Patla article "found declining goshawk occupancy and indications that the decline was significantly higher in logged areas."

The Forest Service responds that it had no obligation to provide a more detailed description of the Patla article and the memorandum. It responds, further, that Patla analyzed goshawks in a national forest that had significantly different characteristics from the Gallatin. Finally, it notes that Patla assessed goshawk occupancy rather than population trends, noting that she wrote in her article, "I assumed that occupancy results apply to the target population of known nesting areas monitored and may not reflect forest-wide population trends."

Second, the Council contends that the FEIS misrepresents the contents of an unpublished report, "Moose Population Viability Analysis," prepared by Forest Service employee Tyler (first name not provided). Tyler recounts in the report that Kurt Alt, a Montana wildlife program manager, "stated that moose are in decline [in] the Gallatin Forest, as well as across Montana. That decline has also been noticed in the general project area[.]" The Council contends that this statement is inconsistent with two statements by the Forest Service citing the Tyler report—a statement in the ROD that "the moose population is stable at the local and larger scales," and a response, during the comment period, that "there is no evidence that moose are declining, when viewed at larger spacial scales than the project level."

The Forest Service responds that the Council has misread the Tyler report. The report recites Alt's view of the moose population as declining, but then goes on to state Tyler's own view, which is generally consistent with the Forest Service's statements in the ROD and in its response. Tyler wrote, "[T]here is no evidence that moose should be considered a small or declining population at all the appropriate spacial scales for analysis, that is, at the global or continental level or across the intermountain west, or [Greater Yellowstone Area]. This statement is also valid for the Gallatin forest[.]"

Third, the Council points out that the ROD for Lonesome Wood 2 states that the 2011 MIS Assessment indicated that the populations of the goshawk and the pine marten in the Gallatin were "stable to increasing." A response to a comment to the Draft EIS also states that the goshawk population is "stable to increasing." The Council correctly notes that these statements are inaccurate. Contrary to the statements in the ROD and the response, the 2011 Assessment concluded, "[G]oshawks are present and distributed across the Gallatin National Forest, but population trends cannot be determined from the existing data," and that population "parameters" for the pine marten "indicate a relatively stable or slightly declining population on a statewide basis."

We agree with the Forest Service's responses to the Council's first and second points. The Forest Service did not act arbitrarily or capriciously in the manner in which it described and analyzed the Patla article, the internal memorandum, and the Tyler report.

However, the Council's third point is well taken. The Forest Service was flatly wrong in stating that the Assessment concluded that the populations of goshawk and pine marten were "stable to increasing." This mistake is troubling, given the importance of MIS monitoring in alerting the Forest Service and the public to effects of human activity in our national forests. It does not appear, however, that the mistake was a significant factor in the Forest Service's approval of Lonesome Wood 2. With respect to the goshawk, the Forest Service determined that the project would not affect forest-wide population trends, given that goshawk have not been detected in the project area and that mitigation measures are specified in the project's FEIS in order to protect any goshawk nests that might be discovered in the course of implementing the project. With respect to the pine marten, the Forest Service recounted accurately in the FEIS that its population was "relatively stable or declining on a statewide basis," and also determined that the project would not significantly affect forest-wide population trends. Further, it does not appear that this mistake prevented the public from making an informed decision about the likely effects of Lonesome Wood 2. We do not underestimate the importance of accurate descriptions of the results of MIS surveys. In the context of this particular project, however, we conclude that the Forest Service's mistake was not inconsistent with its having taken a "hard look" at the project.Conclusion

For the foregoing reasons, we affirm the district court's summary judgment order and order dissolving the injunction.

AFFIRMED.