ORDER

DANA L. CHRISTENSEN, Chief Judge.

Plaintiffs filed suit in February 2012 seeking judicial review of the United States Forest Service's Decision Notice and Finding of No Significant Impact (FONSI) authorizing implementation of the Fleecer Mountains Project ("Project") on the Beaverhead-Deerlodge National Forest ("the Forest"). Plaintiffs also challenge the Record of Decision and corresponding documents authorizing the implementation of the Revised Beaverhead-Deerlodge National Forest Land and Resource Management Plan ("Forest Plan" or "Plan").

Plaintiffs claim the Project and the Forest Plan violate Section 7 of the Endangered Species Act ("ESA") because the United States Forest Service ("Forest Service") failed to complete consultation with the United States Fish and Wildlife Service ("Wildlife Service") about the potential affects of the Project and Plan on grizzly bears and Canada lynx. They also claim the Project and Forest Plan violate the National Forest Management Act ("NFMA") and National Environmental Protection Act ("NEPA") in various ways. They contend the Environmental Assessment ("EA") for the Project failed to disclose and apply the best available science regarding elk, grizzly bears, and lynx; that the Forest Plan does not ensure elk, grizzly bear, and lynx viability; that the Forest Service predetermined that it would reach a finding of no significant impact for the Project; and that the Forest Service failed to disclose whether a point source water pollution permit was necessary for the Project and failed to consult with the State of Montana on this issue.

Defendants counter that they complied with their obligations under the ESA, NFMA, and NEPA, and also insist that Plaintiffs' allegations of standing are inadequate.

For the reasons discussed below, the parties' motions for summary judgment are granted in part and denied in part. As a threshold matter, the Court finds that Plaintiffs have standing to challenge both the Forest Plan and the Project. Summary judgment is also granted in favor of Plaintiffs on their claims under the ESA. The Court concludes that the case must be remanded to the Wildlife Service to consider whether lynx "may be present" in the Forest because the Wildlife Service improperly applied a stricter standard to that inquiry. Until the Wildlife Service conducts its analysis under the proper standard and the parties complete any consultation that might become necessary, the Project must be enjoined. The Forest Service's biological assessment of whether the Project "may affect" grizzly bears was also arbitrary and capricious, and a new biological assessment must be prepared.

Summary judgment is granted in favor of Plaintiffs on their claim that the Forest Plan's and Project's discussions of elk violate NEPA. Although the Forest Service did not act arbitrarily or capriciously in setting road density levels for the Forest, analyzing road density at the landscape and hunting unit scales, or defining secure areas for elk, the Court nevertheless finds that the Forest Service must supplement its EIS for the Forest Plan to explain or support, if possible, its decision to exclude temporary roads from the road density objectives and to correct the record to show that permitted and administrative roads are included in the objectives. The Project EA must also be supplemented with a full and fair discussion of the impact that temporary roads will have on elk during the Project's lifetime, an important aspect of the problem given the already high road density levels in the Project area.

Summary judgment is granted in favor of Defendants on Plaintiffs' claims that the Forest Plan and Project violate NEPA and NFMA in their consideration of Canada lynx. By incorporating the Lynx Direction and considering the impact of linkage and connectivity on wildlife including lynx, the Forest Service adequately considered the best available science and provided for the viability of lynx when developing the Forest Plan. The Project is consistent with the Plan's standards, and the Forest Service considered the Lynx Direction in its Environmental Assessment and Wildlife Report.

Summary judgment is also granted in favor of Defendants on Plaintiffs' claims that the Forest Plan and Project violate NEPA and NFMA in their consideration of grizzly bears. The Forest Plan includes enforceable standards and guidelines in the form of road density levels and secure area designations that are consistent with the best available science. A supplemental Environmental Impact Statement ("EIS") is not required under Norton v. S. Utah Wilderness Alliance ("SUWA"), 542 U.S. 55, 73, 124 S.Ct. 2373, 159 L.Ed.2d 137 (2004), unless the Forest Service decides upon concluding consultation with the Wildlife Service that the Forest Plan needs to be amended. The Project's Environmental Assessment is properly tiered to the Forest Plan, and the Forest Service did not fail to disclose and apply any science that was not adequately considered and disclosed when the Forest Plan was developed.

Summary judgment is also granted in favor of Defendants on Plaintiffs' remaining two arguments. The FONSI was not predetermined, and the Forest Service was not required to obtain a NPDES permit for stormwater runoff.

FACTS

The Fleecer Mountains Project is a timber sale or salvage project planned for an area northwest of Wise River, Montana, in the Beaverhead-Deerlodge National Forest. FP:A01:1.1 The Project area is in the Fleecer watershed and encompasses 102,424 acres, including private lands and 98,769 acres of National Forest Service lands within the Beaverhead-Deerlodge National Forest. Id.

According to the Forest Service's Decision Notice, the goals of the Project are to improve forest conditions in the area, supply wood products to the forest products industry, and secure habitat for westslope cutthroat trout. Id. About 95 percent of the lodgepole pine stands in the Project area have been affected by mountain pine beetle. Id. Many of the trees are dead or dying, and the value of the wood as a commercial product decreases as the wood deteriorates. Id. Additionally, Douglas Fir trees are dying due to recent drought conditions and dense stand conditions, which make the trees more susceptible to the western spruce budworm. Id. at 2. The Forest Service also aims to address the encroachment of Douglas fir on to grassland and sagebrush parks that were free of conifers historically and to improve the growing conditions and stand resiliency of Douglas fir by reducing stand density. Id. Removing conifers is expected to promote riparian-associated aspen stands, which have decreased in patch size and distribution. Id. at 3. Finally, the Project involves establishing new fish-movement barriers and replacing culverts in two creeks in order to secure westslope cutthroat trout habitat, which is threatened by competition from normative fish populations. Id.

Plaintiffs' claims concern three species: lynx, grizzly bear, and elk. The Forest is classified as unoccupied by lynx, but potential linkage areas have been identified on the north end of the Project area to the Anaconda Mountains and Anaconda-Pintler Wilderness and to the southwest to the Pioneer Mountains. FP:A02:96. The grizzly bear is the only federally listed wildlife species known to occur in the Forest; grizzlies have recently been sighted on the north end of the Forest and have been documented in the John Long Mountains, the Anaconda Range, and the Flint Creek Range. Id. at 99. While there are no reports of grizzlies in the Project area itself, the area has "fairly good connectivity" with the Anaconda-Pintler Wilderness, where there have been several recent sightings of grizzlies. Id. at 101. Elk are found throughout the Project area, which includes hunting units 319 and 341. Id. at 121.

Additional facts in the record are discussed as they become relevant in the analysis below.

ANALYSIS

I. Standing

A plaintiff must have standing in order to present a justiciable case under Article III of the United States Constitution. Summers v. Earth Island Inst., 555 U.S. 488, 498, 129 S.Ct. 1142, 173 L.Ed.2d 1 (2009). An organizational plaintiff has standing to sue if at least one identified member would have standing to sue in his or her own right, the "interests at stake are germane to the organization's purposes," and the member's participation is not necessary to the claim or the relief requested. Friends of the Earth, Inc. v. Laidlaw Envtl. Srvcs. (TOC), Inc., 528 U.S. 167, 181, 120 S.Ct. 693, 145 L.Ed.2d 610 (2000); Summers, 555 U.S. at 498, 129 S.Ct. 1142. Three elements are essential to member standing: injury in fact, causation, and redressability. An "injury in fact" must be "(a) concrete and particularized and (b) actual or imminent, not conjectural or hypothetical." Friends of the Earth, Inc., 528 U.S. at 180-181, 120 S.Ct. 693 (citing Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-561, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992)). Additionally, the injury must be "fairly traceable to the challenged action" and likely to be redressed by a favorable decision. Id.

To establish an injury in fact, an organization must show, through specific facts, Fed.R.Civ.P. 56(e), that at least one member has concrete and personal interests, connected to specific, affected areas of the environment, that have been or will be directly harmed by the challenged government action. Summers, 555 U.S. at 493, 498, 129 S.Ct. 1142 (citations omitted). Alleging generalized harm to the environment is not sufficient, but an affidavit showing that an alleged environmental harm "affects the recreational or even the mere esthetic interests" of a member will suffice. Id. at 493, 129 S.Ct. 1142 (citation omitted). The affiant must demonstrate "a connection to the area of concern," White Tanks Concerned Citizens, Inc. v. Strock, 563 F.3d 1033, 1038-1039 (9th Cir. 2009) (citation omitted), and aver that he or she has used an area and has "specific and concrete" plans to return. Summers, 555 U.S. at 495, 129 S.Ct. 1142.

It is not sufficient for members to declare they use unspecified portions of a large area, which may or may not be affected by the challenged activity. Lujan v. Natl. Wildlife Fedn., 497 U.S. 871, 887-89, 110 S.Ct. 3177, 111 L.Ed.2d 695 (1990). Members must name specific, affected subareas that they use and enjoy in order to demonstrate a particularized threat of injury. W. Watersheds Project v. Kraayenbrink, 632 F.3d 472, 484 (9th Cir.2011), cert. denied, ___ U.S. ___, 132 S.Ct. 366, 181 L.Ed.2d 232 (2011) (citation omitted). This requirement, however, does not prevent plaintiffs from challenging actions with wider geographic effects such as a forest-wide plan. Summers, 555 U.S. at 495, 129 S.Ct. 1142.

Here, Plaintiffs have submitted the declaration of Michael Garrity, the Executive Director of Plaintiff Alliance for the Wild Rockies ("the Alliance").2 Though much of Garrity's affidavit is "conclusory and completely devoid of specific facts," Lujan, 497 U.S. at 898, 110 S.Ct. 3177, it nonetheless provides a sufficient factual basis to support standing. Garrity alleges that he has used the Fleecer Mountains Project area for "vocational and recreational purposes" including "wildlife observation and study, hiking, camping, and quiet contemplation in nature." (Doc. 16 at 2, 4.) He "intend[s] to continue to use and enjoy the lands within the Beaverhead-Deerlodge National Forest and Fleecer Mountains area, frequently and on an ongoing basis in the future" and plans "to continue to visit the area, including in the spring/summer of 2013, 2018, and 2023." (Id. at 3.) He alleges he will be unable to pursue his recreational interests in the area if the Project goes forward due to "logging, burning, road-building, road use, road reconstruction, etc[.]" (Id.) These "ecological and esthetic degradations will render the area unsuitable" for his prior uses of the area. (Id.)

Citing Lujan v. National Wildlife Federation, the Forest Service insists that Garrity should have identified the subareas of the Forest and Project area that he used more specifically. In Lujan, the plaintiff expressed an interest in a two-million acre area, but the site-specific project at issue only affected a 4,500-acre area within that; it was not clear the plaintiff was likely to suffer an injury in fact. 497 U.S. at 886-87, 110 S.Ct. 3177. Here, the "Fleecer Mountain Project area" is specific enough. By alleging he has recreated and worked in the Project area itself, Garrity has shown that he "use[s] the area affected by the challenged activity and not an area roughly in the vicinity of the project site." Summers, 555 U.S. at 499, 129 S.Ct. 1142. Though the Project area encompasses over 100,000 acres, it is a specific timber sale project in an area defined by the Forest Service that Garrity has visited and plans to use again. "Vegetation treatment" (clearcutting, commercial, precommercial, and noncommercial thinning, and conifer removal) will occur on some 3,043 acres of the Project area, FP:A01:4, which are distributed widely over the eastern half of the Project area, see id. at 49, and more sporadically over the western half of the Project area, see id. at 47. Additionally, the Project will involve opening temporary roads, reopening some closed roads, closing some roads and trails to the public, and burning, and the Forest Service recognized in its Environmental Assessment for the Project that the "sounds and sights attributed to harvest activity such as timber falling, log hauling, noise, dust, and occasional road delays" are likely to affect people enjoying various forms of dispersed recreation in the area. FP:A02:241. Even if the Court were to adopt the Sixth Circuit's decision in Heartwood, Inc. v. Agpaoa, 628 F.3d 261, 268 (6th Cir.2010), rehg. denied (May 11, 2011), that standing requires plaintiffs to "identify particular segments of a river, sections and sub-sections of a forest, or passes in a mountain range that they use and will continue to use, and that agency action will detrimentally affect," Garrity has expressed a legally protected interest in an area specific enough to establish standing.

II. ESA

Plaintiffs claim that Defendants violated section 7 of the Endangered Species Act, the "heart of the ESA," W. Watersheds Project, 632 F.3d at 495, by failing to consult with the Wildlife Service to evaluate the consequences of the Forest Plan and the Fleecer Mountains Project on grizzly bears and lynx. 16 U.S.C. § 1536(a)(2). Section 7 requires an agency to ensure that no discretionary action will "jeopardize the continued existence of any endangered species or threatened species or result in the destruction or adverse modification of [critical] habitat of such species." 16 U.S.C. § 1536(a)(2); 50 C.F.R. § 402.12(a). "Only after the [agency] complies with § 7(a)(2) can any activity that may affect the protected [species] go forward." P. Rivers Council v. Thomas, 30 F.3d 1050, 1055-57 (9th Cir.1994).

The Forest Service's first step in complying with section 7 is to obtain from the Wildlife Service "a list of any listed or proposed species or designated or proposed critical habitat that may be present in the action area." 16 U.S.C. § 1536(c)(1); 50 C.F.R. § 402.12(c)-(d) (emphasis added). If the Wildlife Service advises that a listed species or critical habitat may be present, the Forest Service must complete a biological assessment to determine if the proposed action "may affect" or is "likely to adversely affect" the listed species. 16 U.S.C. § 1536(c)(1); 50 C.F.R. §§ 402.12(f), 402.14(a), (b)(1); Forest Guardians v. Johanns, 450 F.3d 455, 457 (9th Cir.2006). Once the biological assessment is completed, it must be shared with the Wildlife Service. (50 C.F.R. § 402.12). "If [the Wildlife Service] advises that no listed species or critical habitat may be present, the Federal agency need not prepare a biological assessment and further consultation is not required." 50 C.F.R. § 402.12(d).

A determination by the Forest Service in a biological assessment that an action "may affect" a listed species or critical habitat gives rise to a consultation requirement under section 7 of the ESA. Karuk Tribe of Cal. v. U.S. Forest Serv., 681 F.3d 1006, 1027 (9th Cir.2012). The Ninth Circuit holds that "the minimum threshold for an agency action to trigger consultation with the Wildlife Service is low." W. Watersheds Project, 632 F.3d at 496. "[A]ny possible effect, whether beneficial, benign, adverse, or of an undetermined character, triggers the formal consultation requirement." Id. (citing 51 Fed. Reg. 19,949; Cal. ex rel. Lockyer v. U.S. Dept. of Agric., 575 F.3d 999, 1018-19 (9th Cir.2009)).

There are two forms of consultation: formal and informal. Karuk Tribe of Cal., 681 F.3d at 1027. Formal consultation is obligatory where the Forest Service has determined that an action is "likely to adversely affect" a listed species. But it is not required if 1) the Forest Service finds, either in its biological assessment or through informal consultation, that while a project "may affect" a listed species, the species is "not likely to be adversely affected" and 2) the Wildlife Service concurs in writing. 50 C.F.R. §§ 402.12(j)-(k), 402.14(b)(1), 402.13(a).

The Administrative Procedure Act governs review of agencies' actions under section 7. W. Watersheds Project, 632 F.3d at 496 (citation omitted). The Court must determine whether the agencies' actions were "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." Id. (citation omitted). "Deference to an agency's technical expertise and experience is particularly warranted with respect to questions involving scientific matters." United States v. Alpine Land & Reservoir Co., 887 F.2d 207, 213 (9th Cir.1989). However, the "presumption of agency expertise may be rebutted if the decisions, even though based on scientific expertise, are not reasoned." Greenpeace v. NMFS, 80 F.Supp.2d 1137, 1147 (W.D.Wash.2000).

A. Lynx

In the instant case, in compliance with section 7(c) and 50 C.F.R. § 402.12(c)-(d), the Wildlife Service provided the Forest Service a list of threatened or endangered species that "may be present on the Beaverhead-Deerlodge National Forest." FP:I13-20; 50 C.F.R. § 402.12(c)-(d) (emphasis added). Though the Canada lynx is listed as a threatened species under the ESA, FP:X-04:2, the Wildlife Service did not include lynx on the "may be present" list on the grounds that the Forest is not "occupied" by lynx. Thus, the Forest Service reasoned it was not required to prepare a biological assessment or pursue further consultation regarding the possible effects of the Forest Plan or Project on lynx. 50 C.F.R. § 402.12(d). Because the Forest Service concluded it was not required to prepare a biological assessment, the Wildlife Service's obligation to issue a written concurrence was not triggered.

Plaintiffs insist that the Wildlife Service's decision to import the definition of "occupied" habitat into the ESA's "may be present" standard was arbitrary and inconsistent with the plain meaning and purposes of section 7 of the ESA. They point to numerous pieces of information that could support a finding that lynx "may be present" on the Forest even if they do not "occupy" it. Defendants counter that the agency's interpretation of ESA's "may be present" standard was rational and is entitled to deference. They claim that occupancy is relevant because lynx are highly mobile and can occur in isolated instances far from lynx populations and that the other evidence Plaintiffs argue they should have considered is stale and unreliable.

The Wildlife Service's determination was based at least in part on a series of interagency agreements between the Forest Service and the Wildlife Service and the agencies' interpretation of those agreements.3 After the lynx was listed as a threatened species in 2000, the Forest Service, the Wildlife Service, and the Bureau of Land Management entered an interim agreement, the Lynx Conservation Assessment and Strategy, to guide lynx conservation actions on federal land. BDNF:L1-280. In 2006, the Wildlife Service and the Forest Service agreed to an amendment that defined the requirements for occupancy, identified occupied lynx habitat, and determined that management guidelines were only required to be implemented in occupied habitat. BDNF:L1-333. They also agreed the Beaverhead-Deerlodge National Forest was unoccupied. Id. In 2007, the Forest Service published the Northern Rockies Lynx Management Direction ("Lynx Direction"), which superseded the prior agreements. BDNF:L1-370; FP:H-14.

The Lynx Direction incorporated the 2006 amendment's definition for occupancy. For a forest to be deemed occupied, there must be 1) two verified lynx observations since 1999 "unless they are verified to be transient individuals" or 2) evidence of reproduction in the area. Id. at 33. The Lynx Direction classified the Forest as unoccupied, secondary lynx habitat. FP:H-14:11, 33, 45.

The Lynx Direction notes that the role of unoccupied, secondary lynx habitat in sustaining lynx populations is "unclear." Id. at 36, 45.

Id. The Forest Service conceded in the Lynx Direction that "management actions could adversely affect unoccupied secondary lynx habitat" and that "if and when lynx attempt to establish home ranges in secondary areas, individual lynx could be affected." Id. at 37-39. However, it concluded that there is "no evidence" to suggest that unoccupied secondary habitat is necessary to the viability of lynx populations. Id. at 45. Thus, the Forest Service determined that the Lynx Direction "shall" be applied in occupied, core habitat, but need only be "considered" in unoccupied secondary areas. Id. at 35, 42.

The Lynx Direction anticipated that the Wildlife Service would not include lynx on species lists for forests deemed unoccupied, because the Wildlife Service did not include lynx on species lists for unoccupied forests under the 2006 amendment. Id. at 33. In a memorandum regarding the 2006 amendment, the Wildlife Service advised Ecological Services Project Leaders that "the lynx should not appear on species lists for proposed Federal actions on national forests determined to be unoccupied by lynx.... Compliance with section 7(a)(2) of the Endangered Species Act is not required for the lynx under this circumstance." FWS:004652-53.4 Defendants do not point to similar guidance provided after the 2007 Lynx Direction was published, but argue that the Wildlife Service followed the same policy. Doc. 22 at 32-33; doc. 25 at 20.

If "the agency has chosen a definition that comports with the text and purposes of the [ESA] and is not `decidedly irrational, it is not [the Court's] place to second-guess its judgment." Bassiri v. Xerox Corp., 463 F.3d 927, 933 (9th Cir.2006). The problem here is that the Wildlife Service has imported into section 7 of the ESA a standard — the definition of "occupied" habitat — that is inconsistent with the statutory and regulatory language. The "may be present" standard is, on its face, much broader than the Wildlife Service's requirement that a forest "be occupied" by the species. The Forest Service all but admits that lynx may be present on the Forest. See e.g. FP:A02:343 ("The analysis in the EA does not state that lynx do not occur in the mountain range. What it says is that the Forest (and project area) is currently considered `unoccupied' by the USFWS."); BDNF:L1-370:36-39 (recognizing that unoccupied secondary areas provide connectivity and linkages for lynx between core areas as well as foraging habitat). Similarly, the Wildlife Service, in its Biological Opinion on the Lynx Direction, emphasized that unoccupied habitat should be managed to "continue[] to facilitate and allow dispersal of lynx" and to "avoid or reduce effects on lynx." BDNF:L1-384; BDNF:L1-370:39. Thus, both agencies recognize that lynx may "occur," travel through, or forage in "unoccupied" areas, and that management actions in unoccupied areas may affect those transient lynx as well as any lynx attempting to establish new home areas.

If a species "may" be present, the ESA obligates the agency to perform a biological assessment or inter informal consultation with the Wildlife Service to ensure that the proposed action will not adversely affect the species. Defendants have not provided a reasoned basis for its construction that the "may be present" standard requires occupancy. The Wildlife Service itself, analyzing section 7, once rejected the argument of a commenter who urged the Service "to include only species actually known or believed to occur in the action area":

Interagency Cooperation — Endangered Species Act of 1973, as Amended; Final Rule, 51 FR 19926-01 (Jun. 3, 1986). Defendants now attempt to distinguish this response by distinguishing "migratory species" from transient species like the lynx, but the focus of the commentary was not limited in this manner. The Wildlife Service clearly rejected a standard which would require a species to be "actually known or believed to occur" in an area because it would conflict with the statutory language.

Although evidence of the presence of lynx in the Forest is not overwhelming, some evidence nevertheless exists that lynx may be present. Specifically, Squires et al. (2003), documented one set of lynx tracks in 2001 in the Anaconda range, part of which lies within the Big Hole landscape area, which is within the analysis area for wildlife security for the Project. Berg (2009) identified "possible" and "probable" lynx tracks in the Forest. Additionally, Berg found that though "most of the BDNF was ... likely not good lynx habitat," there were "significant exceptions" to this. FWS:004347. For example, "[r]esident lynx may have been present in the West Fork/Middle Fork Rock Creek vicinity" given the abundance of snowshoe hare and forest structure, and "habitat was also pretty good in the Pioneer Mountains." Id. Berg also concluded:

Id. A 2003 Wildlife Service map also suggests the Forest may be within the range of resident and dispersing lynx, FP:O-04: 264, and radio-collared lynx are actually known to have traveled through mountain ranges in the Forest, though they did not stay in the Forest for long, FP:J075a:3.

None of this evidence is reliable enough to fit the criteria for "occupancy," and other evidence cited is arguably stale. But the Wildlife Service's decision to reject the evidence entirely is arbitrary and capricious, particularly considering the Wildlife Service's earlier position that the "may be present" standard does not require actual occurrence. On its face, the question of whether lynx "may be present" in an area is less rigorous than the question of whether lynx "occupy" an area. Applying the occupancy definition to the first step in the process "create[s] a metric more stringent than, and contrary to, what the ESA dictates." Alliance for Wild Rockies v. Lyder, 728 F.Supp.2d 1126, 1137 (D.Mont. 2010).

Of course, the Wildlife Service's construction is expedient — the agencies undoubtedly anticipate that they would conclude that an action in an unoccupied area that allegedly complies with the Lynx Direction would not adversely affect the lynx — but that does not permit the agencies to take the procedural shortcut that has occurred here. The agencies must first determine whether a species "may be present," under a reasonable interpretation of the Act's plain language. Only then should they consider the likelihood that the species will be affected, and that inquiry should be based on the performance of a biological assessment or informal consultation. Because the Wildlife Service substituted its "occupancy" standard for the ESA's "may be present" standard, the agencies did not enter into informal consultation or have the opportunity to agree in writing that the action is "not likely to adversely affect" the lynx, despite the fact there is some evidence that lynx "may" be in the area. 50 C.F.R. § 402.13(a).

In summary, the Wildlife Service's "occupancy" standard bypasses the procedural protections of section 7, allowing it to ignore any evidence that does not fit the more rigorous standard. The agencies' interagency agreement to reach this result in unoccupied areas cannot override the statutory and regulatory language of the ESA or the Ninth Circuit's direction that "the minimum threshold for an agency action to trigger consultation with the Wildlife Service is low," W. Watersheds Project, 632 F.3d at 496, and "any possible effect, whether beneficial, benign, adverse, or of an undetermined character, triggers the formal consultation requirement." Id. (citations omitted). Accordingly, the Project must be enjoined until the Wildlife Service reconsiders its listing determination in accordance with this opinion.

B. Grizzly Bears

At the time the Forest Plan was issued in January 2009, consultation on the grizzly bear was not required under section 7 because the grizzly bear was not listed as an endangered or threatened species. FP:H-05:1. But on September 21, 2009, the Wildlife Service was ordered to reclassify the Yellowstone Distinct Population Segment as threatened. Id. at 2. The Wildlife Service prepared a new species list for the Forest, listing the grizzly bear as a species that "may be present." Id.

As required, the Forest Service prepared a Biological Assessment for the Grizzly Bear (Ursos arctos horribilis) for the Beaverhead-Deerlodge Revised Forest Plan (2009), recognizing that a revised (or first) biological assessment is required if a newly listed species may be present in an area affected by a continuing agency action. FP:H05:01. The Biological Assessment only analyzed the four areas within the Forest that lie within the Yellowstone Distinct Population Segment (DPS) area as defined by the Wildlife Service. Id. It concluded that "[w]hile individual resource effects are expected to be minimal, implementation of the [Forest Plan] across the spectrum of resource areas is likely to adversely affect the threatened grizzly bear."5 Id. at 2 (emphasis in original).

The Forest Service initiated formal consultation when it provided its biological assessment to the Wildlife Service in August 2010. FP:H06:6. As required, the Wildlife Service then prepared a biological opinion, likewise limiting its analysis to areas where grizzlies were deemed likely to be present:

FP:H06:15. The Wildlife Service concluded that though the Forest Plan would result in incidental take of individual bears "due to the Revised Forest Plan direction for access management, sanitation/food storage and livestock grazing," it would not jeopardize the likelihood of grizzly bear survival or recovery. Id. at 46. The Forest Plan, as applied to the areas analyzed, thus passed section 7 muster.

Since then, there have been verified grizzly bear observations in areas outside the action area as it was defined in the 2010 biological opinion. Since the Complaint in this case was filed, the Forest Service has reinitiated consultation to consider the effects of the Forest Plan on grizzly bears in the remaining Forest areas. (Doc. 22 at 34.)

Plaintiffs argue that Defendants were required to engage in section 7 consultation for the Forest Plan as a whole and the Project specifically.

1. Consultation on the Forest Plan

By reinitiating consultation on the Forest Plan, the Forest Service has rendered moot at least portions of the Plaintiffs' third claim for relief. That claim asks the Court to enjoin the Project and declare that it violates the law (doc. 1 at 31) based on the Defendants' failure to 1) complete a biological assessment and biological opinion on the Forest Plan for the entire Forest, using current information concerning possible grizzly bear distribution, and 2) amend the Forest Plan to provide measures aimed at recovering and conserving the grizzly bear population in the Forest (id. at 24). Plaintiffs do not challenge the 2009 Biological Assessment or 2010 Biological Opinion insofar as they concern the action area analyzed; their concern is that other areas of the Forest should have been analyzed as well. Because this will occur in the reinitiated consultation, and because Plaintiffs point to no possibility of continuing violations, declaratory relief concerning the initial consultation and injunctive relief requiring further consultation is unnecessary. Defenders of Wildlife v. Martin, 454 F.Supp.2d 1085, 1102-04 (E.D.Wash.2006) (citing, among other cases, Forest Guardians, 450 F.3d at 462). Thus, these issues are moot.

However, the Project must be enjoined until Defendants complete the reinitiated consultation for grizzly bears. It is "well-settled that a court can enjoin agency action pending completion of section 7(a)(2) requirements." Wash. Toxics Coalition v. EPA, 413 F.3d 1024, 1034 (9th Cir.2005). Section 7 provides that "[a]fter initiation of consultation required under subsection (a)(2) of this section, the Federal agency ... shall not make any irreversible or irretrievable commitment of resources with respect to the agency action which has the effect of foreclosing the formulation or implementation of any reasonable and prudent alternative measures which would not violate subsection (a)(2) of this section." 16 U.S.C. § 1536(d).

Plaintiffs have met their threshold burden to obtain injunctive relief. Through this Order, they have shown actual success on the merits of their claim, and they have demonstrated a likelihood of irreparable harm absent injunctive relief. The record contains ample evidence that the Project will impact the Fleecer Project area in ways that are known to affect grizzly bears, including the development of new roads and "disturbance effects" during the Project's lifetime. Timber sales constitute per se irreversible and irretrievable commitments of resources under § 7(d). P. Rivers Council, 30 F.3d at 1057; Lane Co. Audubon Soc. v. Jamison, 958 F.2d 290, 295 (9th Cir.1992). It is possible that changes to the Forest Plan, if deemed necessary through consultation, may alter the plan for the Project. If the Project went forward before consultation was completed, it could result in irreparable harm.

To avoid an injunction, the acting agency must prove the action is non-jeopardizing. Wash. Toxics Coalition, 413 F.3d at 1035. Requiring this proof of Defendants "is consistent with the purpose of the ESA and what [the Ninth Circuit has] termed its institutionalized caution mandate." Id. Defendants have not met this burden. The Forest Plan provides "management direction" for the Project, FP:A02:16, and the reinitiated consultation may result in changes to that direction.

In summary, the Project cannot go forward until consultation on the effects of the Forest Plan on the grizzly bear is complete. It would constitute an irretrievable commitment of resources that might be contrary to the result of the consultation process and could lead to irreparable harm, and it would foreclose the implementation of alternative measures.

2. Consultation on the Project

Plaintiffs also argue that the agencies should have engaged in Section 7 consultation for the Project. (Doc. 15 at 15.).

a. "May be present" determination

It is not entirely clear whether Defendants recognized that grizzly bears "may be present" in the Project area. On one hand, the Forest Service recognized that grizzly bears "may be present" based on the Wildlife Service's 2010 and 2011 lists of threatened and endangered species that may be present in the Forest as a whole. FP:I13-5:14; FP:I13-20. Additionally, the Forest Service claims to have produced a biological assessment, which is only required if a species "may be present." On the other hand, the Forest Service argues that it did not determine that grizzlies "may occur" in the Project area (doc. 22 at 29), and it did not inquire of the Wildlife Service whether any listed species "may be present" in the Project area itself. 16 U.S.C. § 1536(c)(1); 50 C.F.R. § 402.12 (the list should include species that "may be present in the action area") (emphasis added); Thomas v. Peterson, 753 F.2d 754, 763 (9th Cir.1985) (citing 16 U.S.C. § 1536(c)(1)).

The Wildlife Report recognizes that grizzly bears have been documented in areas of the Forest other than the Yellowstone DPS distribution area. FP:I13-05:15. It notes that several grizzlies have recently "moved into" the area between the John Long Mountains, Flint Creek range, and Pintler range, which "has the capability of becoming a key grizzly linkage zone between the Boulder/Garnet mountain range complex and the Anaconda-Pintler Wilderness." Id. at 15-16. However, occupancy in this area has not been corroborated by the Wildlife Service. Id. at 16.

The Project area lies to the south of the Flint Creek range and to the south and east of the Anaconda-Pintler Wilderness. Id. It has "fairly good connectivity [with the Anaconda-Pintler Wilderness], because there is only a narrow, low-speed, low-traffic highway between them, with few developed areas." Id. Despite the good connectivity, no grizzlies have been sighted in the Project area itself. Id.

As with the lynx, it is not necessary that grizzly bears occupy an area to satisfy the "low threshold" for consultation. Grizzly bears "may be present" if "transitory bears [] might move through the project area." FP:A02:141. The Project area has "fairly good connectivity" with areas with known grizzly bear activity, FP:I13-05:15-16, and implicit in the Wildlife Report's and Environmental Assessment's discussion of the importance of preserving this connectivity is the possibility that grizzly bears, which are known to occur nearby and are a wide-ranging species, might travel through the Project area.

In any case, the first question in the section 7 analysis — whether grizzly bears may be present in the action area — must have been answered in the affirmative given that the Forest Service and Wildlife Service recognize that grizzly bears "may be present" in the Forest, the Forest Service did not seek an opinion on whether they "may be present" in the Project area specifically, and the Forest Service claims to have prepared a biological assessment and focuses its arguments on its "no effect" determination. If the questions were not answered in the affirmative, the Forest Service would have abused its discretion since there is no evidence the Wildlife Service excepted the Project area from its listing determination for the Forest.

b. "May affect" determination

When a listed species "may be present," the Forest Service is required to prepare a biological assessment in order to determine if the proposed action "may affect" the species. 50 C.F.R. § 402.12; see also Or. Nat. Resources Council v. Allen, 476 F.3d 1031 (9th Cir.2007); Native Ecosystems Council v. Dombeck, 304 F.3d 886, 901-02 (9th Cir.2002); Thomas, 753 F.2d at 763 (holding that the Forest Service's failure to prepare a biological assessment prior to its decision to build a road for a timber project was a substantial procedural violation of the ESA). Here, the Forest Service claims that its Fleecer Mountains Project Wildlife Report, FP:I13-05, which was incorporated into the Environmental Assessment ("EA") for the Project, FP:A02:91, is a biological assessment that satisfies the requirements of section 7 of the ESA. (Doc. 25 at 18.) The Wildlife Report and EA conclude that the Project would have "no effect" on grizzly bears, thus obviating the need for consultation. Karuk Tribe of Cal., 681 F.3d at 1027 ("An agency may avoid the consultation requirement only if it determines that its action will have `no effect' on a listed species or critical habitat.") (citation omitted).

Both the Wildlife Report and EA contain a brief discussion on the distribution of grizzly bears in the Forest and near the Project area and some of the potential impacts that the Project might have on grizzly bears that may travel through the Project area. The Wildlife Report states that the Project would result in changes to "stand structure and species composition," as well as "disturbance effects due to increased traffic, human activity, and equipment use during project activities." FP:I13-05:60. However, the report concludes that the Project would have "no effect" on grizzly bears because it would maintain connectivity with areas where bears are known to occur.

In reaching this conclusion, the Forest Service was not guided by any specific direction for management of grizzly bears in the Forest Plan "[b]ecause the project area lies outside of the mapped grizzly bear distribution." Id. at 61. The only Forest Plan direction that the Forest Service applied was the Plan's open motorized road and trail density goals, which are intended to improve wildlife security for various animals, including grizzlies. Id.; BDNF:A1-40:499. The Forest Service found that there would be no "net increase" in road density in the Project area and that "existing travel restrictions would be followed, maintaining fall secure areas." FP:I13-05:61. Elsewhere, the Wildlife Report and Environmental Assessment note that road density will increase during Project activities because the Project calls for the construction of temporary roads. FP:I13-05:77-78; FP:A02:158. These roads would not be open to public use and would be obliterated upon the Project's completion. Id. The Forest Service also determined there would be no net decrease in secure areas for grizzly bears-the percentages of the Project area that qualify as secure areas in the summer and in the fall would be the same after the Project as before the Project. FP:A02:133. With no net decrease in connectivity, the Forest Service determined the Project would have "no effect" on grizzly bears that might traverse the area. Id.; FP:I13-05:61.

1. Whether the analysis was arbitrary and capricious

Plaintiffs dispute that the Wildlife Report and EA count as a biological assessment. Neither party analyzes this issue in depth, and although it is not clear the reports were intended to constitute a biological assessment under the ESA, the Court finds that it is not necessary to resolve the question. Even if the Wildlife Report and EA are analyzed as a biological assessment, the Forest Service's conclusion that the Project will have no effect on grizzly bears was arbitrary and capricious.

The first sentence of the Wildlife Report's section on possible effects to the grizzly bear states that the Project will result in "disturbance effects due to increased traffic, human activity, and equipment use during project activities." FP:I13-05:60. The Ninth Circuit has held that "[a]ny possible effect, whether beneficial, benign, adverse, or of an undetermined character, triggers the formal consultation requirement." Ctr. for Biological Diversity v. U.S. Bureau of Land Mgt., 698 F.3d 1101, 1122 (9th Cir.2012) (emphasis in original; internal quotation marks and citation omitted); Cal. ex rel. Lockyer, 575 F.3d at 1018-19 (citing 51 Fed.Reg. 19926, 19949); see also Final ESA Section 7 Handbook, 3-12 ("Is not likely to adversely affect [is] the appropriate conclusion when effects on listed species are expected to be discountable, or insignificant, or completely beneficial."). This "low" standard "allow[s] Federal agencies to satisfy their duty to insure under section 7(a)(2) [that species are not jeopardized]." Ctr. for Biological Diversity, 698 F.3d at 1122. To recognize that the Project will result in "disturbance effects" and then conclude that the Project will have "no effect" on grizzly bears is arbitrary and capricious. While the "disturbance effects" may be discountable or insignificant given the maintenance of connectivity for grizzly bears that might travel through the area, "any possible effect" requires the Forest Service to obtain the concurrence of the Wildlife Service in order to avoid consultation.

Nor can the Court accept the Forest Service's "post hoc rationalizations for agency action." N.W. Envtl. Defense Ctr. v. Bonneville Power Admin., 477 F.3d 668, 688 (9th Cir.2006) (citation omitted). The Forest Service explains in its briefing that it reached a "no effect" determination because there is no evidence of a grizzly bear ever being in the Project area and the nearest sightings of grizzly bears occurred 12 or more miles away from the Project. The Wildlife Report and EA did note that grizzly bears had not been documented in the Project area, though they had been documented in nearby areas. FP:I13-05:15, FP:A02:100-01. But until briefing on this motion, the Forest Service did not rely on these facts to conclude that the Project would have no effect on the species. Rather, the Forest Service concluded that the Project would have "no effect" on grizzly bears because it would "maintain the connectivity for transitory bears that might move through the project area." FP:A02:141. The Wildlife Report and EA concentrated on the preservation of road density and secure area levels to show that grizzlies that might pass through the area would not be impacted. It did not conclude that grizzly bears would not pass through the area. This contrasts sharply with the revised EA in Alliance for the Wild Rockies v. Bradford, 864 F.Supp.2d 1011 (D.Mont.2012), which clearly articulated the Forest Service's argument that the project at issue would not impact grizzly bears because there were none in the area. The Forest Service cannot now rely on a different rationale than that put forward in the document purportedly reviewing the proposed agency action. It is required to "present a rational connection between the facts found and the conclusions made" in the biological assessment itself. Or. Nat. Resources Council Fund v. Brong, 492 F.3d 1120, 1131 (9th Cir. 2007) (internal quotation marks and citation omitted).

The conclusion that the Forest Service's no-effect determination was arbitrary and capricious is bolstered by the Forest Service's failure to consider other "relevant factors." Selkirk Conservation Alliance v. Forsgren, 336 F.3d 944, 953-54 (9th Cir. 2003). For example, the main reason articulated in the Wildlife Report and EA for concluding that the Project would not affect grizzly bears was that road density levels before and after the Project would be the same. But the Forest Service failed to analyze whether the temporary increase in road density and the temporary decrease in summer secure areas during the Project's 5-10 year duration would affect transitory grizzly bears. See P. Coast Fedn. of Fishermen's Assns., Inc. v. NMFS, 265 F.3d 1028, 1037 (9th Cir.2001) (finding that the record lacked scientific evidence to support the decision to disregard the short-term effects of a timber project). Similarly, the Forest Service failed to consider whether the Project would impact whitebark pine in the area, and whether this would affect grizzly bears that might travel through the area. The Forest Service urges the Court to find that its path "may reasonably be discerned" in the record because it noted that overstory trees in the Fleecer Mountains area have been "moderately to severely impacted" from a "combination of blister rust and mountain pine beetle." FP:A02:53. But the Wildlife Report and EA make no connection between this finding and any potential effects on species that might rely on whitebark pine as a food source, and the Forest Service asks too much of the Court's ability to extrapolate from this finding that the whitebark pine in the Project area is now "insufficient to sate the appetite of a grizzly bear." It is arbitrary and capricious to fail to "entirely fail to consider an important aspect of the problem." Lands Council I, 537 F.3d at 987 (internal quotation marks and citation omitted).

In summary, the Forest Service arbitrarily concluded that the Project would have no effect on grizzly bears after admitting it would result in disturbance effects, and it failed to analyze important aspects of the problem, including whether the construction and use of temporary roads during Project activities may affect grizzly bears that may travel through the area. It is also possible that changes to the Forest Plan might be necessary after reinitiated consultation on the effects of the Plan is completed, and the Project analysis will need to take such changes into consideration. Thus, the case is remanded to the Forest Service to prepare a new biological assessment addressing whether the Project "may affect" grizzly bears that may be present in the area and, if the biological assessment concludes consultation is necessary, to take appropriate steps.

III. NEPA and NFMA

Under NFMA, the Forest Service must develop a Land Resource Management Plan (a forest plan) for each national forest, 16 U.S.C. § 1604(f)(1), which "provide[s] for diversity of plant and animal communities ... in order to meet overall multiple-use objectives," 16 U.S.C. § 1604(g)(3)(B). Subsequently, all projects planned within a forest must be consistent with the forest plan as well as any regulations in effect at the time of the decision. Native Ecosystems Council v. U.S. Forest Serv., 428 F.3d 1233, 1249 (9th Cir.2005) (citing 16 U.S.C. 1604(1)); Native Ecosystems Council v. Tidwell, 599 F.3d 926, 932 n. 9 (9th Cir.2010) (citation omitted).

NFMA regulations have been revised and replaced and subsequently withdrawn many times over the past several years. At this point, the Forest Service may elect to follow the 1982 regulations "[u]ntil the Department promulgates superseding planning regulations," 36 C.F.R. § 219.35(b) (2011). The Forest Plan at issue here "was begun and completed under the 1982 planning regulations." BDNF:I1-01:12. But the Forest Service elected to apply the transition provision of the 2000 regulations to the Project. AR A-01:33. Thus, the Forest Plan must ensure wildlife viability under the 1982 regulations, and the Project analysis must apply the best available science under the transition provision of the 2000 regulations.

The 1982 planning regulations includes a viability and management indicator species regulation that was located at 36 C.F.R. § 219.19 until it was withdrawn in November 2000. The 1982 viability regulation provides in part:

§ 219.19. Under the regulation, the Forest Service must also identify management indicator species by which the effects of management activities can be assessed; evaluate planning alternatives in terms of the quantity and quality of management indicator species habitat and population trends; consider the effects of recreation and pest and fire management on the Wildlife Service populations; monitor trends of the management indicator species; and identify "critical habitat" for threatened and endangered species. Id.

Beyond these statements, neither NFMA nor its regulations "specify precisely how the Forest Service must demonstrate that its site-specific plans adequately provide for wildlife viability." Lands Council v. McNair, 537 F.3d 981, 992 (9th Cir.2008) ("Lands Council I"), overruled in part on other grounds by Winter v. Nat. Resources Def. Council, Inc., 555 U.S. 7, 129 S.Ct. 365, 172 L.Ed.2d 249 (2008), as recognized by Am. Trucking Assn. v. City of Los Angeles, 559 F.3d 1046, 1052 (9th Cir.2009). Generally, the Court should "defer to the Forest Service as to what evidence is, or is not, necessary to support wildlife viability analyses." Id. "Viability analysis that uses all currently available scientific data is considered sound." Lands Council v. McNair, 629 F.3d 1070, 1081 (9th Cir.2010) ("Lands Council II"). But agencies are not obligated to adopt the specific findings of any particular study. Kern Co. Farm Bur. v. Allen, 450 F.3d 1072, 1081 (9th Cir.2006). Additionally, if the Forest Service elects to substitute direct monitoring of a species population with habitat monitoring, it "must both describe the quantity and quality of habitat that is necessary to sustain the viability of the species in question and explain its methodology for measuring this habitat." Lands Council II, 629 F.3d at 1081 (quoting Lands Council I, 537 F.3d at 987-88). These viability rules apply to the Forest Plan at issue.

The transition provision from the 2000 planning regulations states that the Forest Service must consider the best available science in its site-specific project analyses until new planning regulations, replacing the 1982 planning regulations and viability requirement, are finalized. 36 C.F.R. § 219.35(a) (2011). Analysis of whether the Forest Service considered the best available science is also pertinent to a viability analysis under the 1982 regulation. Case law indicates the two standards are inextricably entwined. See e.g. Ecology Ctr., 574 F.3d at 659 ("The 1982 Rule, as incorporated into the Forest Plan, requires only that the Forest Service maintain habitat capable of supporting "viable populations" of old-growth dependent species. Relying on the best available science, as required by the 2000 Rule, the Forest Service determined that maintaining old-growth levels often percent below 5500 feet was sufficient to support such species."); W. Watersheds Project v. Salazar, 766 F.Supp.2d 1095, 1114 (D.Mont.2011) aff'd in part, 494 Fed.Appx. 740 (9th Cir. 2012) ("Based on the best available science before it, the Forest Service reasonably determined that the Yellowstone bison herd is viable and genetically diverse.") The virtue of using the best available science standard to assess whether a Forest Plan ensures viability is emphasized by a series of decisions in which the Ninth Circuit has been reluctant to fully incorporate the 1982 Rule into plans based on general viability language in the plan or even references to the 1982 rule. E.g. Earth Is. Institute v. U.S. Forest Serv., 697 F.3d 1010, 1015-16 (9th Cir.2012). See also Earth Island Institute v. Carlton, 626 F.3d 462, 472 (9th Cir.2010); id. at 477 (Reinhart, J., dissenting).

Under NEPA, the Court must "simply [] ensure that the Forest Service made no `clear error of judgment' that would render its action `arbitrary and capricious.'" Lands Council I, 537 F.3d at 991. A decision is arbitrary and capricious "only if the agency relied on factors Congress did not intend it to consider, entirely failed to consider an important aspect of the problem, or offered an explanation 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." Id. at 987 (citations and internal quotation marks omitted).

A. Elk

1. Forest Plan

The Forest Plan identifies elk as a management indicator species and indicates that the Forest Service will "maintain[] habitat conditions for elk security." BDNF:A1-41:55. However, Plaintiffs claim that the Forest Plan fails to ensure elk viability because the Forest Service did not disclose and consider the best available science in its analysis of road density6 or of elk secure areas.

The Forest Service is not using habitat as a proxy for the population of elk in the Forest — it relies on the State's counts and specifically looks at population sizes in each hunting unit within the Forest. BDNF:A1-40:501. When the Plan was adopted, elk population trends were stable or increasing across most of the Forest, even though the previous forest plans (generally represented by Alternative 1) had provided less habitat security than the adopted alternative (modified Alternative 6). BDNF:A1-40:499, 517-19, table 175. However, the Forest Service is still required to analyze elk habitat under the terms of the Forest Plan itself, which lists as one of its primary objectives that "[c]onditions for self-sustaining or viable populations of native and desired non-native plant and animal species are supported within the natural capability of the ecosystem." BDNF:A1-41:19. The Plan also identifies what is considered a viable population:

BDNF:A1-41:314.

Plaintiffs argue that the Forest Plan fails to ensure the viability of elk, fails to comply with the best available science for elk, and fails to explain any departures from the best available science.

a. The Forest Service complied with the general requirements of the 1982 viability rule.

The Forest Service generally complied with the requirements set forth by the 1982 Rule. The agency identified management indicator species, including elk. BDNF:A1-40:504. It then evaluated each planning alternative in terms of its effect on elk habitat and population trends, concentrating on vehicle access management as the primary management tool, as recommended by Christensen. BDNF:A1-40:498-99, 544; BDNF:L1-055:4; see also analysis at BDNF:A1-40:502, 517-24, 544, 546. Though the Forest Service does not independently monitor the elk population, it relies on the counts provided by Montana Fish, Wildlife, and Parks (MTFWP). E.g. BDNF:A1-40:501-02; BDNF:A1-41:16; A1-41:284 (identifying MTFWP reports on population numbers as a highly reliable performance measure for elk habitat effectiveness). This generally satisfies the requirements listed under the viability regulation, § 219.19 (1982).

b. Road density objectives

Plaintiffs argue that the best available science cited in the Forest Plan does not support the road density thresholds that were adopted to ensure elk viability.

1) Generally

The Forest Service considered and largely adopted the recommendations presented by Christensen et al. (1993), a study the parties agree is among the best available science on the subject of protecting elk habitat. First, the Forest Service recognized road density as the most significant consideration for protecting elk populations and habitat. Compare BDNF:A1-40:498, 518 with BDNF:L1-055:4. As recommended, the Forest Plan measures elk habitat effectiveness by the number of miles of open road per section, BDNF:A1-41:1465; BDNF:L1-055:4-5, and it significantly increased the percentage of Forest closed to wheeled travel in both the summer and winter, BDNF:A1-40:519.

Second, the Forest Service implemented a seasonal approach to regulating road density. It implemented more stringent route density objectives for the fall hunting season, recognizing that hunting is "the primary source of elk mortality," compare BDNF:A1-40:498, 519-20; BDNF:A1-40:508 with BDNF:L1-055:5, and it closed more big game winter range to motorized travel in 11 out of 12 landscape areas than was closed under the previous plans, BDNF:A1-40:520. This "season of use" approach was strongly advocated by Christensen. BDNF:L1-055.

Third, the Forest Service worked with the State extensively, coordinating with the Montana State Elk Management Plan and relying on Montana Fish Wildlife and Parks' elk population objectives and counts. BDNF:A1-40:498-99, 502, 544; BDNF:A1-41:16. Christensen stressed the importance of collaborating with State biologists and planners in developing forest standards and guidelines for elk. BDNF:L1-055:6-8. Although the agencies did not agree on all points, the Forest Service and Montana Fish Wildlife and Parks agreed on many others, and particularly agreed to work together to manage elk populations in the Forest. See generally BDNF:A1-40:911-35.

2) Quantitative requirements

Plaintiffs insist the Forest Plan fails to disclose and comport with Christensen's quantitative "thresholds" for preserving "habitat effectiveness" for elk.

The explicit recognition that the Forest Plan's "vehicle access management" strategy is "[b]ased on Christensen et al. (1993)" sufficiently incorporates Christensen's quantitative analysis into the Forest Plan's discussion. BDNF:Al-40:498. By citing the specific study upon which the Plan relied, the Forest Service complied with 40 C.F.R. § 1502.24, providing the public notice of its reliance on Christensen's analysis. Earth Island Inst. v. U.S. Forest Serv., 351 F.3d 1291, 1300-01 (9th Cir. 2003), overruled on other grounds by Lands Council I, 537 F.3d 981 ("An agency must `identify any methodologies used' and `make explicit reference by footnote to the scientific and other sources relied upon for conclusions in the [EIS].'") (quoting 40 C.F.R. § 1502.24).

The Forest Service implemented quantitative objectives for habitat effectiveness that are consistent with Christensen's analysis. Christensen explains that "habitat effectiveness" is "used as a measure of summer range ability to support elk" and is assessed by measuring road density. BDNF:L1-055:3-4. Contrary to Plaintiffs' arguments, Christensen does not require habitat effectiveness of 70% or greater for all summer range. Christensen lays out different density levels that fit different management goals. Id. at 4-5, graph. Where elk are a "primary resource consideration," habitat effectiveness should be 50% or greater (less than 1.7 mi/mi2), and where "benefit[ting]" elk summer range and "retain[ing] high use" by elk is an even higher priority, habitat effectiveness should be 70% or greater (less than 1 mi/mi2). Id. at 5. A ratio of road mileage to land area that is lower than 50% will "mak[e] only minor contributions to elk management goals" and suggests that "elk are not a consideration" in those areas. Id.

In eight of the eleven landscape areas in the Forest for which road density objectives are set out,7 the Forest Service implemented year-round road-density objectives that achieve habitat effectiveness of 50% or greater. BDNF:A1-41:53-54, table 13. In these areas, open motorized route density levels range from 0 mi/mi2 (Madison landscape) to 1.6 mi/mi2 (Jefferson River landscape). Christensen condones such density levels when elk are at least a "primary consideration" in an area. The Forest Plan does not suggests that elk merit a higher priority level, particularly given that "elk populations Forest-wide have fundamentally met or exceeded State objectives with the [previous plans'] motorized roads/trails footprint across the BDNF." BDNF:A1-40:518. Thus, the Forest Service's determination that 50% habitat effectiveness in these landscapes is sufficient to support a viable population of elk is reasonable.

Based on the Forest Service's road density goals, elk are not a primary consideration in three other landscapes, however. The Boulder River and Clark Fork-Flints landscapes have objectives of 1.9 mi/mi2 and the Upper Clark Fork landscape has an objective of 2.0 mi/mi2. Though these road densities indicate elk are not a priority in these landscapes, the densities are lower than under previous plans.

These densities are also consistent with the Forest Plan's stated management goals for these areas. In the Boulder River landscape, the Forest Service places an emphasis on dispersed recreation "with many roads and trails on gentle terrain that accommodates motorized vehicles," and the area receives heavy recreational use. BDNF:AR-41:97. "Roads are dense in several parts of the area and were built to accommodate timber harvest, mining and access to private property." Id. Most of the management areas in the landscape are managed for timber production, livestock grazing, road-based recreation, and access to forest resources. Id. at 99, 101, 107, 109, 113. For these management areas, elk are not listed as a primary consideration, except where winter range is closed to winter motorized travel. Id. at 99, 101, 107, 109, 113. Three other management areas are managed in part for elk security, id. at 107, or general wildlife habitat, id. at 103, 105. The remaining management area does discuss elk or wildlife habitat but big game winter range is closed to motorized activities in winter. Id. at 111. Despite the road density, the Plan indicates that elk are common in the landscape. Id. at 97.

The Clark Fork Flint Landscape includes a variety of management areas, some of which have very high road density (such as the areas around Georgetown Lake and Discovery Ski Basin), BDNF:AR-41:117, 119, 125, others of which have low road density, id. at 123, 127, 128, and others of which have a combination, id. at 121, 129, 131. As may be expected, the management goals for each area also range considerably. Forest products contribute significantly to the economies of the communities in this area, and there is significant recreational use as well as timber production, livestock grazing, and mining. Id. at 117, 119, 121.

The Upper Clark Fork Landscape contains the Butte urban area and is heavily used by Butte residents. BDNF:AR-41:249. Private and public land is intermixed along the edges of the forest, with "private homes and patented mining claims mixed in with forested vegetation and various access needs." Id. "The landscape is the smallest on the Forest, yet the level of development and recreation use is high." Id. Two of the management areas are managed for concentrated recreation, motorized and non-motorized recreation, timber harvest, and livestock grazing, id. at 251, 255, a third is managed to protect water quality, id. at 253, and the fourth is specifically managed for fall and winter secure habitat, among other uses, id. at 257.

Because the higher road densities are consistent with the management goals for these three landscapes, it is reasonable to conclude that the elk habitat effectiveness objectives are actually consistent with Christensen.8 Additionally, given that elk viability in the Forest has not been impacted by the higher road densities in these areas, it is reasonable for the Forest Service to conclude that low habitat effectiveness in these particular landscapes will not threaten elk viability in the Forest overall.

3) Unit of analysis for road-density objectives

The Forest Plan sets out year-round road density objectives at the landscape scale and fall-season road density objectives at the hunting unit scale. Plaintiffs argue that Christensen recommends a smaller scale of analysis.

The decisions of the Forest Service to use the landscape and hunting unit scales were not arbitrary and capricious. Idaho Sporting Cong., Inc. v. Rittenhouse, 305 F.3d 957, 973-74 (9th Cir.2002) ("[T]he choice of analysis scale must represent a reasoned decision and cannot be arbitrary.") In the first place, the Forest Service did not disregard a clear recommendation from Christensen. That was the problem in the case cited by Plaintiffs, Native Ecosystems v. Weldon, 848 F.Supp.2d 1207, 1217 (D.Mont.2012), vacated 2012 WL 5986475 (Nov. 20, 2012). In that case, the Forest Service repeatedly recognized Hillis as the best available science, but ignored Hillis's conclusion that "analysis unit boundaries should be defined by the elk herd home-range." Id. at 1217. When the Forest Service calculated road density at two other scales without explanation, its determination was arbitrary and capricious. Id.

Here, Christensen's recommendation is not as clear as Plaintiffs suggest:

BDNF:L1-055:5. The Forest Plan does not implement a specific project or analyze the effects of a project. Rather, it sets objectives that apply "forestwide" and are intended to be achieved over the next ten to fifteen years. BDNF:A1-41:19. It provides a "meaningful, larger scale perspective[]" from which to assess project-specific decisions. Accordingly, the Forest Service's use of the landscape and hunting unit scales in the Forest Plan does not disregard Christensen's recommendation that project-level effects be analyzed at a smaller scale.

The decision to use the landscape and hunting unit scales is reasonable. Though the landscapes are generally larger than the 30,000 to 150,000 acre range described in Christensen, it is not clear from Christensen that a larger area necessarily provides a less meaningful estimate of landscape conditions. The Northern Region of the Forest Service decided in 1995 to base revision of its forest plans on landscape areas, explaining:

BDNF:L2-01:5. The decision to set road density objectives at the landscape level comports with that philosophy. See Wildland CPR, Inc. v. United States Forest Service, 872 F.Supp.2d 1064, 1076-77 (D.Mont.2012). And because "[landscapes in Montana tend to be isolated mountain ranges and the valley bottoms around them," BDNF:L2-01:4, the Forest Service did not have to make such sweeping generalizations as to render landscape-based analysis arbitrary. Id. at 1077. Assessing road density at the smaller, hunting unit scale during the fall season recognizes that hunting is the primary source of elk mortality and that road density increases elk vulnerability. BDNF:L1-055:6. The scale is compatible with the Montana State Elk Management Plan, BDNF:A1-40:498, 502, 910, 913, and "provides a better scale at which to manage habitat for elk and use monitoring information gathered by MTFWP," id. at 544.

There is no evidence that smaller scales of analysis would be more appropriate for setting road density objectives at the forestwide level, and the Forest Service did not abuse its discretion in deciding to use the landscape and hunting unit scales.

4) Roads only open to permitted and administrative use and temporary roads

Plaintiffs argue that the Forest Plan arbitrarily excludes permitted and administrative roads from its road density calculations. Each table in the Forest Plan that lays out year-round or fall road-density objectives includes a footnote indicating that roads that are only open to permitted and administrative use are not included. BDNF:A1-41:53-55. No explanation is provided for why administrative and permitted roads are not included, and the footnote directly contradicts a recommendation in Christensen:

BDNF:L1-055:4. While the inclusion of the footnote without any explanation appears to be arbitrary and capricious, the footnotes were evidently typos, and the road density objectives actually do include administrative and permitted roads. This correction appears in a "protocol developed to help [Forest Service employees] apply the goals, objectives, and standards found in the Wildlife Habitat section of the Revised Forest Plan" that was sent to all Beaverhead-Deerlodge National Forest Employees on August 12, 2011. FP:I13-7:1-4. If this protocol is accurate, the Forest Service actually complied with Christensen's recommendation to include administrative and permitted roads. However, that is not evident in the Forest Plan or its administrative record, and this must be corrected on remand in a supplemental EIS.

Though Defendants concede that administrative and permitted roads must be included in the Plan's road density objectives, they insist that "temporary" roads are not included and do not need to be.

Christensen et al. (1993) does not support the exclusion of temporary roads. See Native Ecosystems Council, 848 F.Supp.2d at 1219. While the study does not speak specifically to "temporary" roads except to advise that the Forest Service "[i]dentify temporary roads where they are an option," temporary roads are not excepted from Christensen's conclusion that "[a]ny motorized vehicle use on roads will reduce habitat effectiveness." BDNF:L-055:4 (emphasis added).

The definition section of the FEIS does not support the exclusion of temporary roads either. "Road density" is defined as the "[n]umber of miles of open road per square mile." BDNF:A1-40:1463. While "open road" may suggest that restricted-use roads are not included in the definition, Defendants have admitted that administrative and permitted roads are, in fact, included in the definition. A "temporary road" is listed as one type of "road." Id. It is defined as a "road[] authorized by contract, permit, lease, other written authorization, or emergency operation not intended to be part of the forest transportation system and not necessary for long-term resource management," id., and as "[a] road or trail necessary for emergency operations, or authorized by contract, permit, lease, or other written authorization that is not a forest road or trail that is not included in the Forest Transportation Atlas (36 CFR 212.1 (2005) Transportation System)," id. at 1464. In other words, a temporary road may be an administrative or permitted road, which Defendants say are included in the tables.

Neither the Forest Plan nor the FEIS discuss what effect temporary roads will have on elk viability. In their briefing, Defendants argue that including temporary roads would be nonsensical in areas where the road density objectives are lower than the actual road density at the time the Plan was adopted. In these areas, they assert, no management activities requiring temporary roads would ever be allowed. While this may be true, the Forest Service failed to develop its analysis in the record for the Forest Plan itself, and provided no explanation for its departure from the best available science or from the definitions contained in the FEIS. It "entirely failed to consider an important aspect of the problem," Lands Council I, 537 F.3d at 993, and must address this issue on remand in a supplemental EIS.

c. Elk secure areas

Plaintiffs also challenge the Forest Plan's definition of "secure areas" for elk.

Elk "secure areas" or "security areas" "allow elk to remain in a specific area while under stress from hunting." Christensen et al. (1993), BDNF:L-055:7. Various factors that contribute to security including "size, extent, distance from roads, and vegetative characteristics" work together to make a particular area "secure." Id. But the main concern is that secure areas are "large enough and far enough away from open roads to provide security." Id.

The Forest Plan uses both road density objectives and "secure areas" to promote elk and other wildlife viability. However, road density is the primary management tool used. The goals for wildlife habitat include the following:

BDNF:A1-41:53-54.10

Criteria for secure areas are not discussed in the body of the Plan. However, the Plan's glossary includes the following definitions:

BDNF:A1-41:296,302.

According to the protocol sent to Forest Service employees in 2011, the inclusion of the "Elk Security Area" definition was a mistake:

FP:I13-1:5-6. The protocol also advised that in applying the "Wildlife Secure Areas and Connectivity" goal for "ungulates and large carnivores[,]" the glossary's definition of "secure areas" should be used. Id. at 3. This definition — which states that a secure area is 10 acres or larger and is at least 1/3 of a mile from an open route — is consistent with the definition used throughout the FEIS when discussing the adopted alternative and Alternatives 3, 4, and 5. BDNF:A1-40:260, 498, 499, 503, 510, 912.

Plaintiffs insist that Defendants must use the definition of "elk security area" that was mistakenly included in the glossary for the Forest Plan even though the FEIS repeatedly defines "secure areas" for both elk and grizzly to be areas that are 10 acres or larger and are 1/3 mile from the closest open road. Though the inclusion of the inapplicable definition is confusing, the public was not misled. Id. The FEIS clearly lays out the standard the Forest Service chose to apply.

Plaintiffs also object to the "undisclosed and unsupported assumption" that a definition for secure areas that was originally developed for grizzly bears is sufficient for elk. Doc. 24 at 31. The FEIS states that secure areas are intended to contribute to the viability of both grizzly bears and elk, as well as other ungulates and large carnivores. BDNF:A1-40:498, 503, 510, 512, 513, 519, 523, 912. It also explains that the definition is based on the definition in the 2006 Final Conservation Strategy for the Grizzly Bear in the Yellowstone Ecosystem and Grizzly Bear Amendment ("Grizzly Bear Amendment") and "the 500 meter threat response buffer for elk noted in Wisdom et al. (2004)." Id. at 498, 503, 519, 523. Elsewhere, the FEIS notes that the definition "employs new science in assessing the threat response of elk to motorized disturbance." Id. at 510. No science besides Wisdom et al. (2004) is cited for this proposition.

Plaintiffs insist that Christensen et al. (1993) indicates a grizzly bear secure area is not sufficient for elk. Again, Christensen et al. (1993) is not as clear as Plaintiffs claim. Christensen states that appropriate definitions for security areas for elk vary across the region, depending on cover and other factors, and that the Forest Service should work with state biologists to develop a definition that fits each specific Forest. BDNF:L-055:7. Christensen notes that one definition, the "Hillis paradigm," developed in Hillis et al. (1991), "may be appropriate for other forests to use as a general guide." Id. (emphasis added). Hillis recommends security areas cover 30% of an area and be 250 or more acres, nonlinear in shape, and over .5 miles from open roads, id. — the same definition mistakenly included in the Forest Plan glossary for "elk security areas." However, Christensen does not suggest that this definition is always appropriate.

The applicability of Wisdom et al. (2004) to the question of how a secure area for elk should be defined is questionable. The paper discusses an experiment designed to measure effects of off-road recreation on mule deer and elk. BDNF:L1-404:2-11. The researchers established 20 miles of "off-road transects"11 in an experimental forest closed to all other human activity. Id. Pairs of people then traversed the transects twice daily over a five-day period; some traveled by ATV, others by mountain bike, others by horseback, and others by foot. Id. The researchers monitored the effect of this traffic on radio-collared elk and mule deer. Id. They found that elk were more likely to exhibit a flight response when near ATV or mountain bike traffic than horseback riders or hikers. Id. Elk exhibited a flight response 62% of the time they were within 100 meters of ATVs, 43% of the time they were within 500 meters of ATVs, and 25% of the time they were within 1000 meters of ATVs. Id. at 13; BDNF:A1-40:523-24. Wisdom et al. (2004) concludes that additional research is necessary and recommends three basic strategies to mitigate the negative effects of off-road traffic on elk: feature opportunities for different off-road activities in different watersheds; restrict each recreational activity to specified trails or roads; and reduce road density. Id. at 11.

Nowhere does the study "note" "the 500 meter threat response for elk," BDNF:A1-40:498, or suggest that a 1/3 of a mile (536 meter) buffer zone is recommended to protect elk habitat or create a secure area for elk. In fact, the study notes that "higher probabilities of elk flight continued beyond 820 yards (750 m) from horseback riders, and 1,640 yards (1,500 m) from mountain bike and ATV riders." BDNF:L1-404:8. The FEIS does not explain how the Forest Service extrapolated from Wisdom et al. (2004) that a 1/3-mile buffer zone from roads is sufficient to create secure areas for elk. Nor does the FEIS disclose why a 10-acre area is sufficient to constitute a secure area. Wisdom et al. (2004) does not address this at all, and no other "recent science" in the Forest Plan record is cited in support. Clearly, a 10-acre area is significantly smaller than a 250-acre area — which is the only specific recommendation offered in the best available science cited in the FEIS — and the FEIS provides no explanation for the discrepancy.

On the other hand, Christensen — a more recent study than Wisdom and the science primarily relied on by the Forest Service — recognized that what makes a secure area adequate to protect elk will vary from region to region. Additionally, the Forest Service definition of secure areas is supported by the relevant state agency. Among several other comments regarding elk habitat, Montana Fish Wildlife and Parks stated that the FEIS "definition provides for greater management options to protect important wildlife habitats...." Id. at 913. This suggests that the Forest Service and state biologists "cooperat[ed] and coordinat[ed] with State biologists ... to formulate criteria" for a secure areas in the Beaverhead-Deerlodge National Forest, as recommended by Christensen. Id. Also in keeping with Christensen, the FEIS repeatedly states that overall road density is the primary management tool for ensuring elk viability, and the Forest Plan reiterates that elk security will be achieved by managing road and trail density.

By ensuring that some secure areas, however defined, are available, the agency at least "consider[ed] an important aspect of the problem," Lands Council I, 537 F.3d at 987, while concentrating on the most effective management tool available, road density. The Court must "defer to the Forest Service as to what evidence is, or is not, necessary to support wildlife viability analyses." Lands Council I, 537 F.3d at 992. Although Wisdom et al. (2004) does not clearly recommend that elk secure areas be defined as the Forest Service has chosen, the Forest Service's definition does not "run counter to [Wisdom et al. (2004) or other] evidence before the agency" and it is not so "implausible that it could not be ascribed to a difference in view or the product of agency expertise." Lands Council I, 537 F.3d at 987. Accordingly, the Forest Service's decision to ensure elk viability by setting road density objectives and providing secure areas as defined in the FEIS is entitled to deference.

In summary, the Forest Service complied with the general requirements of the 1982 viability regulation for elk and adequately disclosed the science upon which it relied to determine appropriate road density levels for areas with different management goals. It did not act arbitrarily or capriciously in setting road density objectives and analyzing road density at the landscape and hunting unit scales. Nor did it act arbitrarily and capriciously in applying the EIS definition of secure areas to elk. However, the Forest Service did not explain or support its decision to exclude temporary roads from the road density objectives. In a supplemental EIS, the agency must analyze this determination and correct the record to show that permitted and administrative roads are included in the road density objectives for each landscape and hunting unit.

2. The Project

a. NFMA

The Project EA is consistent with the Forest Plan, applying the Forest Plan's definition of secure habitat and maintaining the requisite levels of secure areas and road density after the Project's completion. Accordingly, the Project satisfies NFMA. Plaintiffs again demand that the Project analysis should use the Hillis definition for elk secure areas (areas larger than 250 acres at least .5 miles from open roads, covering at least 30% of an analysis area) rather than the definition adopted in the Forest Plan (areas larger than 10 acres with at least a 1/3-mile buffer from open roads). The Forest Service properly applied the definition discussed throughout the Forest Plan FEIS, a definition which the Court has already concluded was reasonable. After the Project's completion, the amount of secure habitat will be the same as was laid out in the Forest Plan. FP:A02:95-96, 140, 147, 158.

Also as concluded above, Christensen does not mandate a 50% or 70% effectiveness standard in all landscapes or smaller units, so the Plaintiffs' arguments that the Upper Clark Fork Landscape and the Big Hole Landscape "fail" these standards are not persuasive. It is true that the road density in the Project area, both before and after the Project, is high. But there is no requirement in NFMA, the Forest Plan, or Christensen that elk be a primary consideration in all areas, and the road density levels are consistent with the Forest Plan.

Christensen recommends that road density be assessed at a smaller scale when analyzing project-level impacts, and that this analysis then needs to be considered against a larger scale. BDNF:L1-055:5. In line with this recommendation, the Project EA measures road density at a smaller, project-specific scale, the watershed scale, FP:A02:200, and at the hunting unit and landscape scales, id. at 95-96.12 At the watershed scale, the road density levels in the Project area exceed the levels recommended by Christensen for areas where elk are at least "a primary resource consideration." BDNF:L1-055:3-4 (less than 1.7 mile of road per square mile). The levels range from 1.8 to 3.5 miles/square mile, FP:A02:200, indicating that the Project area will, at best, "mak[e] only minor contributions to elk management goals" and "elk are not a consideration" there, BDNF:L1-055:3-4. At the landscape scale, elk are not a consideration in the Upper Clark Fork Landscape either and are only a primary consideration in the Big Hole Landscape. BDNF:A02:95-96.

The high road density levels are not caused or increased by the Project; there will be no net increase in road density after the Project's completion. FP:A02:140. During the fall hunting season, even temporary roads will be closed in Project secure areas, so there will no net increase in road density in secure areas during the hunting season over the course of the Project. FP:A02:140; see also FP:A02:29-30. This is consistent with the Forest Plan's Wildlife Standard 1. FP:A02:140. Because the Project is consistent with the Forest Plan, which appropriately relied on the best available science is setting road density and secure area levels, it does not violate NFMA.

b. NEPA

Though the Project satisfies NFMA, the EA did not provide a full and fair discussion of the impact that temporary roads will have on elk during the Project's lifetime. Particularly given the high permanent road density in the area, this constitutes a failure to discuss an important aspect of the problem, a violation of NEPA. As the Project EA recognizes, elk populations are below the state objectives in both hunting units in the Project area. FP:A02:120-21. Though this may be attributable to state hunting regulations,13 the Forest Service provides no analysis to rule out the possibility that its definition of secure area and road density levels may have contributed to the decrease. Additionally, the Project EA fails to discuss how the already-high road density, combined with the increase in temporary roads and temporary decrease in summer secure habitat, will impact elk during the Project's five to ten year lifetime. The Forest Service "entirely failed to consider an important aspect of the problem," Lands Council I, 537 F.3d at 987, and must provide this discussion on remand.

B. Lynx

1. Forest Plan

Plaintiffs claim that the Forest Service violated NFMA's viability requirements by failing to include mandatory standards to ensure the viability of lynx in the Forest Plan.

The Forest Plan incorporates the Northern Rockies Lynx Management Direction, which sets out specific standards for lynx management. BDNF:A1-41:57. However, the Lynx Direction's Record of Decision specifically directs that unoccupied forests are not required to follow the Lynx Direction until they are deemed occupied by lynx. BDNF:A1-41:448,476. Instead, the Lynx Direction, and specifically the direction regarding linkage habitat, need only be "considered." BDNF:A1-41:448. Because the Beaverhead-Deerlodge National Forest is currently unoccupied, the Direction is not binding at this point, even though it is incorporated in the Plan. BDNF:A1-40:522, 1069; BDNF:A1-41:57, 485. If lynx do eventually occupy the Forest, the Lynx Direction's specific standards will become mandatory. BDNF:A1-41:57, 485.

The Court must defer to the Forest Service's conclusion, based on the best available science in the Lynx Direction itself, that incorporating the Lynx Direction for possible future application is sufficient to provide for lynx viability. Lands Council I, 537 F.3d at 992 (the Court must "defer to the Forest Service as to what evidence is, or is not, necessary to support wildlife viability analyses"); Lands Council II, 629 F.3d at 1081 ("Viability analysis that uses all currently available scientific data is considered sound."). Like the Forest Plan, the Lynx Direction was adopted under the 1982 NFMA regulations, and it concluded that maintaining occupied lynx habitat would sufficiently provide for lynx viability. BDNF:A1-41:487. It noted that "[t]here is currently no evidence that suggest[s] that unoccupied secondary habitat is considered necessary for a viable population of lynx," a determination that was supported by the Wildlife Service's 2007 Biological Opinion concluding that the Lynx Direction was not likely to jeopardize the continued existence of lynx within the contiguous United States Distinct Population Segment and would contribute to the survival and recovery of lynx in the northern Rockies ecosystem. Id. at 487-88.

The conclusion that agencies need not apply the Lynx Direction in unoccupied forests does not conflict with the viability regulation, which only requires the Forest Service to maintain viable populations of "existing" species in a planning area. The regulation does not define the likelihood of presence that is necessary for a species to be considered to "exist" in a place, but the Lynx Direction adequately explains the significance of occupied versus unoccupied territory to species viability.

Plaintiffs insist that not requiring the Forest Service to aggressively conserve now-unoccupied lynx habitat rewards its alleged past failures to protect lynx. But the cases Plaintiffs rely on for this proposition are inapplicable. In Seattle Audubon Society v. Evans, the Forest Service claimed a species was no longer viable once it was listed under the ESA and thus the viability regulation no longer applied. 952 F.2d 297, 301-02 (9th Cir.1991). That is not the case here — the Forest Service recognizes the importance of preserving lynx viability in the Lynx Direction, which was incorporated into the Forest Plan. Plaintiffs merely dispute the relative importance of unoccupied habitat, a subject on which the Forest Service is entitled to deference.

In Native Ecosystems Council v. Tidwell, the Ninth Circuit rejected the Forest Service's use of a proxy-by-proxy approach for monitoring the viability of the sage grouse, a management indicator species, where the best available science did not support the Forest Service's methodology and showed that the sagebrush habitat proxy failed to track the sage grouse population. 599 F.3d at 929, 935. There is no evidence here that the Forest Service's distinction between occupied and unoccupied habitat is contradicted by the best available science or that it fails to preserve lynx viability. Rather, the evidence shows the opposite. By incorporating the Lynx Direction and considering the impact of linkage and connectivity on wildlife including lynx, BDNF:A1-40:499, the Forest Plan's analysis relies on and considers the best available science and sufficiently provides for the viability of lynx, should they occupy the area. No more is required.

2. The Project

The Forest Service satisfied its obligation to "consider" the Lynx Management Direction in planning the Project; as already discussed, it was not required to apply the Direction because the area is unoccupied by lynx.

Both the Project EA and the Wildlife Report consider the Direction. The Project EA references "Attachment A" in the wildlife section of the record, noting that it shows how the Lynx Direction was taken into consideration during Project planning. In Attachment A, the Forest Service Lynx Coordinator considered whether the Project would maintain habitat connectivity with the lynx analysis units in the Project area (Lynx Direction Standard ALL S1), concluding:

FP:I13-09:4. The Coordinator also determined the Project activities would not occur within any linkage area and considered the remaining standards and guidelines in the Lynx Direction, including those related to vegetation management, livestock management, and human use. Id. at 4-12.

The Wildlife Report also discusses several of these standards. Among other things, it identifies the lynx analysis units in the Project area and discusses effects from past management actions as well as how the Project would affect potential lynx and snowshoe hare habitat. FP:I13-05:44, 84-86. The Wildlife Report concludes the Project would not affect the most suitable habitat for lynx because it will focus on dead and dying lodgepole pine instead of subalpine fir and spruce, and snags and downed wood would be retained in the areas that are salvaged to provide habitat components for the future. Id. at 84; FP:A02:164. Both the Wildlife Report and Project EA address connectivity and linkage issues, recognizing that the Forest Plan provides for the maintenance of connectivity for all species, including lynx, by setting aside secure habitat and limiting road density levels. FP:I13-05:10-13; FP:A02:95-96. Road density and the distribution of secure habitat, as defined in the Forest Plan, would be unchanged after the Project's completion. Id.

Plaintiffs assert that the Forest Service "fails to explain how the hundreds of acres of logging in lynx habitat could maintain habitat connectivity for lynx in those areas." (Doc. 15 at 27.) Although the Forest Service does not address this question directly, its path may be reasonably discerned. The Project meets the Lynx Direction's vegetation standards and will not change net road density or net secure habitat, thus maintaining habitat connectivity. The Forest Service's discussion was sufficient and its conclusions were not arbitrary or capricious. Moreover, the Wildlife Report emphasizes that the Project area is considered unoccupied and that there is little evidence lynx have done more than move through the area in recent years. Id. at 46. New field surveys will be conducted in 2013 to determine if lynx have moved into the area. Id.

Having adequately considered the Lynx Direction in developing the Project, the Forest Service complied with the Forest Plan and discussed the best available science for maintaining lynx viability.

C. Grizzly bear

1. Forest Plan

Plaintiffs claim that the Forest Plan fails to ensure grizzly bear viability and that the EIS for the Forest Plan must be supplemented due to the relisting of the grizzly bear or the Forest Plan must be amended.

a. Whether the Forest Plan contains enforceable standards and guidelines for preserving grizzly bear viability

Plaintiffs claim the Forest Plan does not provide enforceable standards and guidelines to ensure the viability of grizzly bears across the Forest. Defendants counter that the Forest Plan ensures grizzly bear viability by setting road density goals, providing secure habitat, and incorporating the 2006 Yellowstone Grizzly Bear Amendment. Defendants have the stronger argument.

As discussed above, the Forest Plan sets road density levels for each landscape and hunting unit in the Forest, recognizing that road density levels contribute to habitat security for grizzly bears and other large mammals. The Forest Plan also measures and designates "secure habitat" across the Forest, based on the definition in the Forest Plan Amendment for Grizzly Bear Habitat Conservation for the Greater Yellowstone Area National Forests (2006) ("Grizzly Bear Amendment"), BDNF:L1-365:16, and the Interagency Conservation Strategy Team's Final Conservation Strategy for the Grizzly Bear in the Yellowstone Ecosystem (2003), BDNF:L1-155:42. BDNF:A1-40:503. This definition is consistent with the best available science cited by Plaintiffs, the Interagency Grizzly Bear Report (1998) and Schwartz et al. (2010). (Pls.' SOF ¶¶ 71, 74.) These sources define secure habitat as "any habitat (exclusive of lakes) that is 10 acres or larger and 500 meters or more from a road." BDNF:A1-40:503. The Forest Plan expands the road buffer slightly to 1/3 mile from all motorized roads and trails. Id.

Secure habitat is intended to benefit grizzly bears as well as other species by providing areas that are less impacted by humans and that maintain connectivity and linkage across the Forest. Id. at 498, 503. The Forest Plan increased the quantity of secure habitat in the Forest over the status quo. Among the alternatives considered, the selected alternative provided the second-highest wilderness acreage, the third-highest summer secure habitat proportion, and the second-highest fall secure habitat proportion. BDNF:A-40:498, 503, 510, 512-13, 517, 519, 523. Forest-wide, the Plan provides 59% fall secure habitat and 52% summer secure habitat. Id. at 510.

The Grizzly Bear Amendment specifically recognized that "the draft revised forest plan for the Beaverhead-Deerlodge National Forest proposes road density standards for all Forest landscapes," that "[t]hese standards will allow only small changes in existing motorized access route density and associated secure habitat." It also recognized that the "maintenance of this level of secure habitat outside the Primary Conservation Area will provide additional assurances the population will be maintained above 400 grizzly bears as required by the Conservation Strategy." BDNF:A1-41:3923. Considering the provision of secure habitat across multiple forests, including the Forest here, the Amendment stated: "We believe the maintenance of over three million acres of long-term secure habitat, supplemented by over one million acres of short-term secure habitat outside the Primary Conservation Area [such as the secure habitat in the Forest], will provide the security necessary for bears to occupy many new areas within the Greater Yellowstone Area, improving chances for movement between important habitats." Id. at 19-20. The Grizzly Bear Amendment thus supports the definition and quantity of secure habitat provided in the Forest Plan.

In addition to providing for secure habitat and setting road density goals, the Forest Plan incorporates the Grizzly Bear Amendment in Wildlife Standard 6, applying it to the Beaverhead portion of the Forest only. BDNF:A1-41:57. The Grizzly Bear Amendment sets out standards and guidelines for "[m]anag[ing] grizzly bear habitat within the Primary Conservation area to sustain the recovered Yellowstone grizzly bear population."14 BDNF:A1-41:384. Because the Lee Metcalf Wilderness is the only portion of the Forest within the Primary Conservation area, these standards are only mandatory there. BDNF:A1-40:498. But the Grizzly Bear Amendment also provides guidance for areas outside the Primary Conservation Area that are deemed "biologically suitable and socially acceptable for grizzly bear occupancy." BDNF:A1-41:384. The goal for these areas is to "accommodate grizzly bear populations to the extent that accommodation is compatible with the goals and objectives of other uses." Id. Such areas are to be managed in accordance with the applicable forest plan and monitored for any changes in secure habitat. BDNF:A1-41:391-92. While the Forest Plan does not indicate what areas of the Beaverhead portion of the Forest are "biologically suitable and socially acceptable for grizzly bear occupancy," it does provide for monitoring of secure habitat across the Forest.

Plaintiffs' allegation that the Forest Plan does not include enforceable standards and guidelines to ensure the viability of grizzly bears is inaccurate. The road density goals, identification of secure habitat, and incorporation of the Grizzly Bear Amendment for part of the Forest constitute enforceable standards and guidelines, and secure habitat and road density restrictions are the key grizzly bear management tools recommended in the Interagency Grizzly Bear report (1998) and the Grizzly Bear Amendment (2006), as well as Schwartz et al. (2010). Plaintiffs insist the Forest Plan should have discussed the specific "parameters" recommended by the Interagency Grizzly Bear Committee in its 1998 report, "explain[ed] the importance of each parameter, and explain[ed] why the Forest Service chose not to implement those parameters in the Forest Plan." (Doc. 33 at 27.) However, it is sufficient that the Forest Plan incorporated the Grizzly Bear Amendment, which relied in part on the Interagency Grizzly Bear Report as well as the Interagency Conservation Strategy Team strategy (2003).15 FP:H11:95-96. An agency need not cite every publication relied on by publications it has in turned relied on. The Forest Service relied on the best available science and outlined standards to maintain the viability of grizzly bears.16

b. Whether the Forest Plan must be amended or a supplemental EIS is required

Plaintiffs argue that the relisting of the grizzly bear is a significant new circumstance that requires a supplemental Environmental Impact Statement (EIS) and the amendment of the Forest Plan.

A supplemental EIS is required under NEPA if "[t]here are significant new circumstances or information relevant to environmental concerns and bearing on the proposed action or its impacts," 40 C.F.R. § 1502.9(c)(1)(ii), and there "remains major Federal actio[n] to occur," Norton v. S. Utah Wilderness Alliance ("SUWA"), 542 U.S. 55, 73, 124 S.Ct. 2373, 159 L.Ed.2d 137 (2004). The United States Supreme Court has found when a forest plan is approved, agency action is completed and that "[t]here is no ongoing `major Federal action' that could require supplementation." Id.; Cabinet Resource Group v. U.S. Fish & Wildlife Serv., 465 F.Supp.2d 1067, 1099 (D.Mont.2006). Plaintiffs do not point to any particular future projects that the Forest Plan here requires; the Plan merely sets standards and milestones for compliance that are implemented through specific projects. Cabinet Resource Group, 465 F.Supp.2d at 1099. Thus NEPA does not require a supplemental analysis unless the Forest Plan is amended or revised. Norton, 542 U.S. at 73, 124 S.Ct. 2373 (citing 43 C.F.R. § 1601.0-6).

The decision whether or not to amend a forest plan is reviewed for reasonableness. Kern v. U.S. Bureau of Land Mgt., 284 F.3d 1062, 1069 (9th Cir. 2002) (citing Northcoast Envtl. Ctr. v. Glickman, 136 F.3d 660, 667 (9th Cir. 1998)). Pacific Rivers Council v. U.S. Forest Service established that the Forest Service should "consider" amending a forest plan when a new species is listed after the forest plan is published. 689 F.3d 1012, 1031-32 (9th Cir.2012), cert. granted, U.S. Forest v. P. Rivers Council, ___ U.S. ___, 133 S.Ct. 1582, 185 L.Ed.2d 575 (2013). The court noted favorably the Forest Service's determination that amendment was appropriate in that case where the existing plans did not address the chinook salmon. Id. at 1056. But the court did not hold that amendment is always necessary; rather it held that given that the forest plans "may affect" the salmon, the Forest Service was required to initiate consultation on them under the ESA. Id.

At this point, the Forest Service has reasonably concluded that it need not amend the Forest Plan despite the relisting of the grizzly bear. When the Forest Service initiated consultation with the Wildlife Service after the grizzly bear was relisted as a threatened species, the Wildlife Service focused on the "action area" within the Forest where bears were known to occur or where the Wildlife Service predicted they might occur within the life of the Plan — the Madison, Gravelly, Tobacco Roots, and Highland Landscapes. FP:H06:13. The Wildlife Service determined that the Plan would not "appreciably reduce the likelihood of both the survival and recovery of the grizzly bears." FP:H06:43. Even though the Plan was developed before grizzly bears were listed, it adequately provided for the viability of grizzly bears in the areas analyzed by the Wildlife Service. Thus, the Forest Service's conclusion that the Plan need not be amended for these areas was reasonable.

Since grizzly bears have been seen outside the area originally analyzed, consultation has been reinitiated. The Wildlife Service's upcoming biological opinion will be relevant to the question of whether the Forest Plan needs to be amended for these areas. If the Plan does need to be amended, that action will constitute an agency action requiring NEPA analysis.

2. The Project

Plaintiffs argue that the Project EA violates NFMA by failing to disclose or apply the requirements for grizzly bears that are contained in the Interagency Grizzly Bear report (1998) or Schwartz et al. (2010), and that it violates NEPA by failing to address why the road density and secure habitat levels in the Project area are sufficient to protect grizzly bears.

The Project EA states:

FP:A02:141. However, the Forest Plan's secure habitat and road density goals apply forest-wide and are meant to protect grizzly bears as well as other species. Only the Grizzly Bear Amendment, incorporated as Wildlife Standard 6, would not apply to the Project area, because the area is not in the mapped Distinct Population Segment area. The EA explains that the Project is consistent with the Forest Plan's road density and secure area goals, noting that after the Project, there would be no net increase in road density and no decrease in secure habitat. FP:A02:140-41. The Project area would continue to contain 32% summer secure areas and 63% fall secure areas. Id. at 95-97, 140-41.

As Defendants assert, the Project EA is properly tiered to the Forest Plan EIS and incorporates by reference that publication's discussions concerning road density and secure habitat. 40 C.F.R. § 1508.28 ("Tiering is appropriate when the sequence of ... analyses is [] from a ... plan ... environmental impact statement to a ... site specific statement or analysis."). The net result of the Project is consistent with the Forest Plan's road density and secure habitat goals for the area, and Plaintiffs point to no site-specific reason additional analysis is necessary. Finally, although the Project EA does not cite to Schwartz et al. (2010), which was not available when the Forest Plan was drafted, Plaintiffs failed to submit the study for the Forest Service's consideration until the administrative appeal following the publication of the EA. In any case, Schwartz emphasizes the importance of regulating both road density and secure habitat, which the Forest Plan and Project both do. FP:O-06:359, 363. The Project thus complies with both NFMA and NEPA.

D. Whether the FONSI was a predetermined outcome

"The purpose of an EA is to provide the agency with sufficient evidence and analysis for determining whether to prepare an EIS or to issue a [Finding of No Significant Impact]" ("FONSI"). Metcalf v. Daley, 214 F.3d 1135, 1143 (9th Cir.2000) (citing 40 C.F.R. § 1508.9). Thus, "[t]he decision whether to prepare a FONSI should be based on the EA, of course, not the other way around." Davis v. Mineta, 302 F.3d 1104, 1112 (10th Cir. 2002) (emphasis added).

Plaintiffs contend that the outcome of the Fleecer Mountains Project EA was predetermined. They note that the Initiation Letter for the Project, dated November 19, 2009, included a timeline culminating in a deadline for publishing a Decision Notice or FONSI. FP:B06:1, 9. They also point to notes from an interdisciplinary team meeting held September 9, 2010, which state in part: "Please review the context and intensity factors for (non)significance when you estimate your effects.... I will need to refer to your reports to substantia[te] a FONSI." ARB-03:4. Plaintiffs' Statement of Undisputed Facts also suggests that the quick turnaround for the EA is suspicious.

While these factors could suggest the outcome of the EA was anticipated, they are not sufficient to find that it was predetermined. "Predetermination occurs only when an agency irreversibly and irretrievably commits itself to a plan of action that is dependent upon the NEPA environmental analysis producing a certain outcome, before the agency has completed that environmental analysis." Forest Guardians v. U.S. Fish & Wildlife Serv., 611 F.3d 692, 714 (10th Cir.2010). "Predetermination is not present simply because the agency's planning, or internal or external negotiations, seriously contemplated, or took into account, the possibility that a particular environmental outcome would be the result of its NEPA review of environmental effects." Id. at 715 (internal quotation marks and citation omitted); Defenders of Wildlife v. Hall, 807 F.Supp.2d 972, 984 (D.Mont.2011); see also Metcalf v. Daley, 214 F.3d 1135, 1143 (9th Cir.2000). While an internal deadline for a certain outcome may suggest predetermination, Am. Wildlands v. U.S. Forest Serv., CV 97-160-M-DWM, 1999 U.S. Dist. LEXIS 22243 (D.Mont. Apr. 16, 1999), it is not determinative. In Friends of Southeast's Future v. Morrison, for example, the Ninth Circuit affirmed a holding that a tentative operating schedule for planned timber supply did not commit the agency to a particular action. 153 F.3d 1059, 1063 (9th Cir.1998); Hall, 807 F.Supp.2d at 984 ("Setting internal deadlines is not the irretrievable commitment of resources needed for a court to conclude the Service predetermined the result of the environmental assessment.").

The Forest Service was not bound in any way by the timeline in the Initiation Letter to reach a finding of no significant impact, and the timeline did not foreclose the possibility an EIS would have to be prepared. Indeed, the same Letter recognizes the possibility that the Forest Service's NEPA analysis could determine that the Project would result in significant impacts. FP:B06:4. Similarly, the informal notes from the 2010 meeting are susceptible to more than one interpretation, and do not amount to an instruction that specialists were required to find no significant impacts. The arguably aggressive schedule for completing the EA does not demonstrate a predetermined outcome, either. Oceana, Inc. v. Locke, 725 F.Supp.2d 46, 67 (D.D.C.2010). Finally, the Forest Service did not make any agreements or contracts that relied on a finding of no significant impact before the EA's completion. Davis, 302 F.3d at 1112 (holding the EA outcome was prejudged where the consultant drafting the EA was contractually obligated to prepare a FONSI); Metcalf v. Daley, 214 F.3d at 1143-44 (holding the agency violated NEPA when it considered the environment only after committing in writing to supporting a proposal). There is simply not enough evidence to conclude the Forest Service committed itself to the Project irreversibly before completing the EA.

E. Discharge permits

The Forest Service did not disclose in the Project's draft or final EA whether it would need to obtain a permit for stormwater runoff from the logging roads under the Clean Water Act and that the Forest Service did not consult with the Montana Department of Environmental Quality to determine if such a permit would be necessary. Plaintiffs claim these constitute violations of NEPA. The Forest Service responds that a permit is not required, that it sent a scoping letter to the Montana DEQ and received no response, and that the NEPA regulation requiring consultation regarding environmental permits applies only to environmental impact statements, not environmental assessments.

The Supreme Court has recently clarified that a National Pollutant Discharge Elimination System (NPDES) permit is not required under the Clean Water Act or its implementing regulations for stormwater runoff from logging roads. Decker v. N.W. Envtl. Def. Ctr., ___ U.S. ___, 133 S.Ct. 1326, 1336-37, 185 L.Ed.2d 447 (2013). Because a Federal permit is unnecessary, the Forest Service was not required to disclose the potential need for such a permit under 40 C.F.R. § 1502.25(b).

NEPA regulations also require some level of consultation with state agencies. When preparing an EIS, an agency "shall... [r]equest the comments of [a]ppropriate state and local agencies which are authorized to develop and enforce environmental standards." 40 C.F.R. § 1503.1(2)(a)(i). In preparing an environmental assessment, on the other hand, an "agency shall involve environmental agencies, applicants, and the public, to the extent practicable." 40 C.F.R. § 1501.4. Section 1501.4 grants significant discretion to the Forest Service to determine whether consultation is necessary. See Taxpayers of Mich. Against Casinos v. Norton, 433 F.3d 852, 861 (D.C.Cir.2006); Greater Yellowstone Coal. v. Flowers, 359 F.3d 1257, 1279 (10th Cir.2004); see also Bering Strait Citizens for Responsible Resource Dev. v. U.S. Army Corps of Engineers, 524 F.3d 938, 952 (9th Cir.2008). Requiring consultation would go beyond the regulations' more general approach to EAs. In any case, because the permit was unnecessary, there was no harm from the Forest Service's failure to consult with the DEQ concerning stormwater runoff.

ORDER

For the reasons discussed above,

IT IS ORDERED that both motions are GRANTED in part and DENIED in part. Plaintiffs' motion (doc. 13) is GRANTED and Defendants' motion (doc. 20) is DENIED as to the claims under section 7 of the ESA and the NEPA claims concerning elk. Plaintiffs' motion is DENIED and Defendants' motion is GRANTED as to the NEPA and NFMA claims concerning grizzly bear and lynx; the NFMA claim that the Project failed to comply with the Forest Plan's Wildlife Standard 1; and the NEPA claims that the FONSI was predetermined and the Forest Service should have disclosed or consulted regarding the need to obtain a NPDES permit.

IT IS FURTHER ORDERED that the Defendants are ENJOINED from implementing the Fleecer Mountain Project, and this matter is REMANDED to the Wildlife Service and the Forest Service to address the deficiencies identified in this opinion.

The Clerk of Court is directed to (1) enter judgment for Plaintiffs and against Defendants in accordance with this Order and (2) close this case.