ORDER
CHRISTENSEN, District Judge.The plaintiffs challenge two projects in the Gallatin National Forest — the Bozeman Municipal Watershed Project and the East Boulder Project. The Forest Service authorized the Bozeman Project on March 5, 2012. It involves logging and burning on several thousand acres over a 5-12 year time frame. The Service will need to construct 7 new miles of road and reopen 3 miles. The Service authorized the East Boulder Project on October 26, 2011. It will involve 650 acres of logging and 2 miles of temporary road construction. Some project activities are slated to take place in the Gallatin Fringe Inventoried Roadless Area, designated critical habitat for Canada lynx, and the Yellowstone Grizzly Bear Recovery Zone.
The plaintiffs raise several claims concerning the Projects’ potential impacts on Canada lynx, grizzly bears, old growth, *1199“snags,” sensitive species, and roadless areas. The Court partially grants summary judgment in favor of the plaintiffs and remands this case to the Forest Service in light of the Court’s recent lynx decision— Salix v. U.S. Forest Service, 944 F.Supp.2d 984, 2013 WL 2099811 (D.Mont. May 16, 2013). Because the plaintiffs in Salix did not meet the burden of identifying likely and irreparable harm tied to specific projects in Lynx Amendment forests, this Court did not grant any injunctive relief. This case is the first one the Court considers since its opinion in Salix that alleges specific harms caused by specific projects, thus providing the Court with the opportunity to articulate the approach for enjoining a specific project, and the burden that each party bears under this approach.
As for the remaining claims and issues, the Court grants summary judgment in favor of the defendants.
Standard
A party is entitled to summary judgment if it can demonstrate that “there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(a). Summary judgment is warranted where the documentary evidence produced by the parties permits only one conclusion. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 251, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986). Only disputes over facts that might affect the outcome of the lawsuit will preclude entry of summary judgment; factual disputes that are irrelevant or unnecessary to the outcome are not considered. Id. at 248, 106 S.Ct. 2505.
Analysis
The plaintiffs present four categories of issues in this case related to: (1) Canada lynx; (2) grizzly bears; (3) Goshawk and marten monitoring, snags, and sensitive species; and (4) roadless areas and wilderness study areas.
I. Canada lynx
The plaintiffs argue that the Projects violate the Endangered Species Act (“ESA”) and the National Environmental Policy Act (“NEPA”) because, among other things, the agencies’ analysis of the Projects’ potential effects on lynx and lynx critical habitat rely on the Northern Rockies Lynx Amendment, which did not address the designation of lynx critical habitat. The Forest Service insists that it did not violate either the ESA or NEPA, but the plaintiffs have the better argument in light of this Court’s decision in Salix.
In 2007, the Forest Service adopted the Northern Rockies Lynx Amendment, which was programmatically added to the forest plans for 18 National Forests, including the Gallatin National Forest. Before adopting the amendment, the Forest Service engaged in formal consultation with the Fish and Wildlife Service under Section 7(a)(2) of the ESA, 16 U.S.C. § 1536(a)(2), to determine whether the amendment would jeopardize lynx or destroy or adversely modify its critical habitat. The Fish and Wildlife Service concluded the amendment would not have these effects.
At the time of consultation, the Fish and Wildlife Service had not yet designated any critical habitat for lynx on Forest lands. That designation happened on February 25, 2009. So consultation for the Lynx Amendment did not include any consideration of whether and how the amendment would affect lynx critical habitat.
The plaintiffs argue that the Projects’ lynx analyses are flawed because the agencies should have reinitiated consultation for the Lynx Amendment when lynx critical habitat was designated. The Forest Service, on the other hand, argues that reinitiation of consultation was not necessary on a programmatic level. Instead, the agencies are required to assess im*1200pacts on lynx critical habitat only on a site-by-site basis when the Forest Service proposes a new project.
A. Salix v. U.S. Forest Service
The Court recently addressed this issue in Salix. There, the Court concluded that designation of critical habitat triggers the need for reinitiation of consultation under Section 7(a)(2) of the ESA. 944 F.Supp.2d at 998-1001, 2013 WL 2099811 at *14-*16. Federal regulations require reinitiation of formal consultation in the following circumstances:
(a) If the amount or extent of taking specified in the incidental take statement is exceeded;
(b) If new information reveals effects of the action that may affect listed species or critical habitat in a manner or to an extent not previously considered;
(c) If the identified action is subsequently modified in a manner that causes an effect to the listed species or critical habitat that was not considered in the biological opinion; or
(d) If a new species is listed or critical habitat designated that may be affected by the identified action.
Id. at 995, at *15 (quoting 50 C.F.R. § 402.16). The Court concluded that both subsections (b) and (d) are implicated by the designation of critical lynx habitat. Id. Moreover, contrary to the Forest Service’s argument in this case, “The agencies cannot shift this analysis to the project level.” Id. (citations omitted).
In short, the Forest Service must reinitiate consultation on the Lynx Amendment to determine what effects the Amendment will have on designated lynx critical habitat on a programmatic level. Id. at 1000-01, at *16.
In Salix, the Court remanded the case to the agency so that it could reinitiate consultation. But it did not enjoin any specific projects because the plaintiffs did not challenge any projects. Id. at 1001, at *17. The circumstances here are different. The plaintiffs ask the Court to enjoin both the East Boulder Project and the Bozeman Project on the basis that the agencies failed to reinitiate consultation under § 7(a)(2) of the ESA.
This case, then, raises the following question: What is the standard for enjoining a specific project in light of a procedural, programmatic violation of the ESA?
B. Preliminary injunction standard
As the Court explained in Salix, the traditional preliminary injunction analysis set out in Winter v. Natural Resources Defense Council, 555 U.S. 7, 129 S.Ct. 365, 172 L.Ed.2d 249 (2008), does not apply to alleged ESA violations. Salix, 944 F.Supp.2d at 1001, 2013 WL 2099811 at *16 (citing Natl. Wildlife Fedn. v. Natl. Marine Fisheries Serv., 422 F.3d 782, 793 (9th Cir.2005)). In ESA cases, courts do not have equitable discretion to balance the parties’ competing interests. Natl. Wildlife Fedn. v. Burlington N. R.R., 23 F.3d 1508, 1510 (9th Cir.1994). The equitable scales are always tipped in favor of the endangered or threatened species. Id.; see also TVA v. Hill, 437 U.S. 153, 194, 98 S.Ct. 2279, 57 L.Ed.2d 117 (1978); Natl. Marine Fisheries Serv., 422 F.3d at 794.
The preliminary injunction standard for ESA cases is more liberal than the traditional standard, but that does not mean that an injunction should issue for every ESA violation. See Burlington N, 23 F.3d at 1511. This Court believes that a plaintiff should bear some evidentiary burden when asking for a preliminary injunction in an ESA case, otherwise the Court cannot begin to fashion an appropriate remedy.
Two lines of cases have developed in the Ninth Circuit that describe the burden (or *1201lack of a burden) that plaintiffs bear in ESA cases.
In the first line of cases, the Ninth Circuit has held that the plaintiff does not bear an initial burden to show that irreparable harm is likely; irreparable harm is presumed: “ ‘It is not the responsibility of the plaintiffs to prove, nor the function of the courts to judge, the effect of a proposed action on an endangered species when proper procedures have not been followed.’” Wash. Toxics Coalition v. EPA, 413 F.3d 1024, 1035 (9th Cir.2005) (quoting Thomas v. Peterson, 753 F.2d 754, 765 (9th Cir.1985)); see also Thomas, 753 F.2d at 764 (“Irreparable damage is presumed to flow from a failure properly to evaluate the environmental impact of a major federal action.” (citation and internal quotation marks omitted)).
Under this line of cases, though, the presumption of irreparable harm is rebut-table. Even if an agency has violated a procedural requirement under the ESA, the agency can avoid an injunction if it can show that the challenged action will not jeopardize the species or destroy or adversely modify its critical habitat: “We have not expressly stated who bears the burden of showing that the action is non-jeopardizing, but the burden should be on the agency, the entity that has violated its statutory duty.” IcL; see also id. at 1035 (“[N]on-jeopardizing agency actions [can] continue during the consultation process.” (citing Sierra Club v. Marsh, 816 F.2d 1376, 1376 (9th Cir.1987))).
In the second line of cases, the Ninth Circuit has held that the plaintiff — not the agency defendant — bears the initial burden. The plaintiff must show at the outset that irreparable harm is “at least likely” in order for an injunction to be issued: “Federal courts are not obligated to grant an injunction for every violation of the law. The plaintiff must make a showing that a violation of the, ESA is at least likely in the future.” Burlington N., 23 F.3d at 15111; see also S. Yuba River Citizens League v. Natl. Marine Fisheries Serv., 804 F.Supp.2d 1045, 1054 (E.D.Cal.2011), reconsideration denied in part, 851 F.Supp.2d 1246 (E.D.Cal.2012); see also Natl. Marine Fisheries Serv., 422 F.3d at 796 (rejection by the district court of a biological opinion and a finding of irreparable harm are “precisely the circumstances in which our precedent indicates that the issuance of an injunction is appropriate”).
In Salix, the Court pointed to this second line of cases and observed that, unless plaintiffs can substantiate their claims by identifying likely irreparable harm, the Court cannot craft an injunction that is tailored to remedy a specific harm. 944 F.Supp.2d at 1001-02, 2013 WL 2099811 at *17 (citing Burlington N., 23 F.3d at 1511; Natural Resources Defense Council, Inc. v. Winter, 508 F.3d 885, 886 (9th Cir. 2007)).
In Salix, the plaintiffs failed to meaningfully allege, much less show any likelihood, of harm resulting from a specific project. Regardless of what burden the plaintiffs initially bore, there was no way to craft a tailored injunction. Here, though, the plaintiffs have alleged specific harms re-*1202suiting from specific projects, so the Court must address the burden that each party bears.
The two lines of cases briefly described above, which address those burdens, are difficult to reconcile in their strictest reading. Under the first line of cases, the plaintiff does not have to make an initial showing that irreparable harm is likely. The harm is presumed, but the agency may rebut that presumption. Under the second line of cases, though, the plaintiff must make an initial showing that irreparable harm is likely.
While hard to reconcile when read in their strictest sense, the two lines of cases can be pragmatically harmonized in a burden shifting approach.
C. A burden-shifting approach to preliminary injunctions in ESA cases
As a first step in the burden shifting approach, a plaintiff must substantiate its claim by alleging a specific irreparable harm resulting from the ESA violation. See Salix, 944 F.Supp.2d at 1001-02, 2013 WL 2099811 at *17. After all, if the plaintiff cannot identify a specific harm, the Court cannot tailor an injunction to remedy that harm.
In Burlington Northern, 23 F.3d at 1511, the Ninth Circuit implied that “harm” in the ESA context is a violation of the ESA — that is, causing “jeopard[y] [to] the continued existence of any endangered species or threatened species or result in the destruction or adverse modification of [critical habitat].... ” 16 U.S.C. § 1536(a)(2). So, at the outset, the plaintiff must allege that, as a result of the ESA violation, a project will jeopardize the continued existence of a specific endangered or threatened species or will destroy or adversely modify its critical habitat.
If the plaintiff alleges a specific harm, then the court should, at that point, presume the harm is irreparable for purposes of issuing an injunction. Thomas, 753 F.2d at 764 (“Irreparable damage is presumed to flow from a failure properly to evaluate the environmental impact of a major federal action.” (citation and internal quotation marks omitted)). At this stage, “ ‘It is not the responsibility of the plaintiff to prove, nor the function of the courts to judge, the effect of a proposed action on an endangered species when proper procedures have not been followed.’ ” Wash. Toxics, 413 F.3d at 1035 (quoting Thomas, 753 F.2d at 765).
If the plaintiff has alleged a specific harm, then, at the second step, the burden shifts to the agency which must show that the action will not jeopardize the species or destroy or adversely modify its critical habitat. Id. In terms of critical habitat, this means that the agency must show that the affected critical habitat will remain “functional.” That is, the physical and biological features of critical habitat — which are commonly described in terms of “primary constituent elements,” see 50 C.F.R. § 424.12(b) — will not be altered to an extent that appreciably reduces the conservation value of the critical habitat, and neither the recovery nor the survival of the species will be jeopardized. See Gifford Pinchot Task Force v. U.S. Fish & Wildlife Serv., 378 F.3d 1059, 1069-71 (9th Cir. 2004) (discussing 50 C.F.R. § 402.02); Natl. Wildlife Fedn. v. Natl. Marine Fisheries Serv., 524 F.3d 917, 931-33 (9th Cir. 2008) ; 74 Fed.Reg. 8616-01, 8644 (Feb. 25, 2009) .
That being said, the agencies cannot meet this burden by relying solely on their compliance with standards or guidelines that are derived from or the product of the underlying ESA violation. To permit such reliance would violate NEPA because the agencies would be “ ‘failing] to consider an important aspect of the prob*1203lem.’ ” Lands Council v. McNair, 537 F.3d 981, 993 (9th Cir.2008) (quoting Motor Vehicle MJrs. Assn., Inc. v. St. Farm Mut. Auto. Ins., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983)). That is the problem the defendants confront in this case. For example, the Forest Service cannot demonstrate that the Projects will not harm lynx critical habitat by relying solely on the Projects’ compliance with the Lynx Amendment. The Lynx Amendment did not address lynx critical habitat, so it “fail[s] to consider an important aspect of the problem.” Lands Council, 537 F.3d at 993. The agencies, however, might be able to meet their burden by showing that the Projects will have no adverse effects on the primary constituent elements of lynx critical habitat, without regard to compliance with any standards in the Lynx Amendment.2
If the agency comes forward with evidence that the project will not jeopardize a species or destroy or adversely modify its critical habitat, then, consistent with Burlington Northern and National Marine Fisheries Service, the plaintiff must come forward with its own evidence of irreparable harm — that is, contrary to the agency’s evidence, the action will harm a species or its critical habitat. If the plaintiff makes this showing, then an injunction should be issued.
This approach does not require the Court to balance the equitable interests of the parties (which is prohibited in the ESA context), but it does require the Court to balance the evidence of harm presented by the plaintiff and agency. If this balancing is a close question, then the Court should err on the side of protection and issue an injunction. See Natl. Wildlife Fedn., 23 F.3d at 1510-11.
This approach is consistent with the only Ninth Circuit case that attempts to harmonize the two lines of cases described above: Southwest Center for Biological Diversity v. U.S. Forest Service, 307 F.3d 964 (9th Cir.2002), opinion withdrawn, 355 F.3d 1203 (2004). The opinion in that case was withdrawn and therefore has no precedential effect, but its reasoning is nonetheless persuasive.
In Southwest Center for Biological Diversity, the Ninth Circuit held that when there is a procedural violation of the ESA, there is a presumption of irreparable harm that generally requires an injunction. Id. at 972-73. But there is a “narrow exception”: A project need not be enjoined if it will not harm the species or its critical habitat. Id. at 973. An excerpt is worth quoting at length:
This case does not present a ‘substantial procedural violation’ of the sort that required an injunction in Thomas and [Biodiversity Legal Foundation v.] Badgley [284 F.3d 1046 (9th Cir.2002) ]. While courts must generally impose an injunction given a procedural violation of ESA, this case fits a narrow exception. We believe that Marsh, 816 F.2d 1376, carves out an exception for cases such as the present one. In Marsh we held that “Congress intended that the consultation process would operate so as to prevent substantive violations of the act.” Id. at 1389 (citation omitted). Thereafter, we found that the protections afforded in Section 7(d) continue until the requirements of Section 7(a) are met. Thus, with respect to non-jeopardizing activities, we held that after consultation was reinitiated in Marsh, the parameters of Section 7(d) would govern the ongoing action. Id. This supports a conclusion *1204that non-jeopardizing agency action may take place during the consultation process in light of the protections of Section 7(d) where the action will not result in substantive violations of the act.
Furthermore, both the present case and Marsh are distinguishable from Thomas on one important point-in Thomas the procedural violation included nothing from which the agency, nor the court, could make any determination of how the action impacted the species or whether the action put the endangered species in jeopardy. In the case at hand, as in Marsh, there was evidence in the record for the court to review the impact on the species during the consultation period. The district court noted that the Forest Service was taking mitigating measures to ensure that the cattle grazing would have little, if any, impact on the loach minnow while formal consultation was taking place. Cattle were excluded from riparian areas, and these areas were being monitored. The consultation was ongoing and was nearing completion. The district court also found that the record supported a finding that the conditions were actually improving given the protective measures that had already been undertaken. These measures do fulfill the purpose of the act as required under Badgley, which is to provide protection to endangered and threatened species.
Id. at 973.
Although Southwest Center for Biological Diversity is not binding precedent, it does serve as guidance in this case, and together with the case law discussed above, supports the burden shifting approach for determining whether a project specific injunction should be issued to remedy a procedural ESA violation. The Court adopts this approach, which is summarized as follows:
1. A plaintiff must initially allege a specific irreparable harm resulting from the ESA violation so that the Court can tailor an injunction to remedy the specific harm. If the plaintiff does so, then the Court presumes that the challenged action will cause irreparable harm.
2. The agency can rebut this presumption by showing that the challenged action will not jeopardize the species or destroy or adversely modify its critical habitat.
3. If the agency comes forward with evidence that the challenged action will not jeopardize the species or destroy or adversely modify its critical habitat, then an injunction should be issued only if the plaintiff produces evidence that such harm is at least likely. If the evidence from both sides presents a close question, then the court should err on the side of issuing an injunction.
D. Application of the burden shifting framework to this case
Here, unlike in Salix, the plaintiffs allege specific harms caused by the Projects and the agencies’ ESA violation. In their complaint,3 they allege that the Projects will adversely impact thousands of acres of lynx habitat by, among other things, damaging denning habitat, foraging habitat, and snowshoe hare habitat.4 Giv*1205en the Forest Service’s failure to reinitiate consultation after the designation of lynx critical habitat, the Court therefore presumes, subject to rebuttal, that the Projects will irreparably harm lynx critical habitat. Thomas, 753 F.2d at 764; Wash. Toxics, 413 F.3d at 1035.
In order to rebut the presumption of irreparable harm, the agencies must show that the Projects will not destroy or adversely modify lynx critical habitat. See Wash. Toxics, 413 F.3d at 1035; 16 U.S.C. § 1536(a)(2). This means the agencies must show that the Projects will not reduce the functionality of lynx critical habitat by (1) altering the primary constituent elements of lynx critical habitat to an extent that appreciably reduces the conservation value of the critical habitat or (2) otherwise jeopardizing the recovery or survival of lynx. See Gifford Pinchot Task Force, 378 F.3d at 1069-71 (discussing 50 C.F.R. § 402.02); Natl. Marine Fisheries Serv., 524 F.3d at 931-33; 74 Fed.Reg. at 8644.
The final rule for lynx critical habitat states that the primary constituent element for lynx critical habitat is:
1. Boreal forest landscapes supporting a mosaic of differing successional forest stages and containing:
a. Presence of snowshoe hares and their preferred habitat conditions, which include dense understories of young trees, shrubs or overhanging boughs that protrude above the snow, and mature multistoried stands with conifer boughs touching the snow surface;
b. Winter snow conditions that are generally deep and fluffy for extended periods of time;
c. Sites for denning that have abundant coarse woody debris, such as downed trees and root wads; and
d. Matrix habitat (e.g., hardwood forest, dry forest, non-forest, or other habitat types that do not support snowshoe hares) that occurs between patches of boreal forest in close juxtaposition (at the scale of a lynx home range) such that lynx are likely to travel through such habitat while accessing patches of boreal forest within a home range.
74 Fed.Reg. at 8638.
The agencies admit that both Projects would adversely affect lynx critical habitat in the Project Areas. The East Boulder Project would degrade hundreds of acres of snowshoe hare habitat, denning habitat, and matrix habitat. The Forest Service, in its Final Wildlife Effects Report and Environmental Assessment wrote: “[S]ubstantial amounts of cover could be removed for lynx and their prey species” in addition to areas in the Lynx Analysis Unit that are already “considered a permanent habitat loss for lynx.” And the Fish and Wildlife Service repeatedly wrote in its biological opinion for lynx that hundreds of acres of lynx critical habitat would be adversely affected due to the reduction of snowshoe hare habitat.
As to the Bozeman Project, the Forest Service wrote that the Project “would affect about 2,673 acres of lynx habitat in some way,” including altering hundreds of acres “to an unsuitable condition,” reducing denning habitat, impacting foraging habitat, and negatively impacting snowshoe hare habitat. Prescribed burning would impact up to an additional 409 acres “to an unsuitable condition.” And planned fuel breaks along roads would result in “heavy treatment” of an additional 209 *1206acres of lynx habitat. All told, the Forest Service claims the Bozeman Project would “bring the total unsuitable lynx habitat within the project area to 1,164 acres (8% of lynx habitat in the project area).” And, as with the East Boulder Project, the Fish and Wildlife Service concluded in its biological opinion that the Bozeman Project would adversely affect lynx critical habitat by destroying snowshoe hare habitat.
These are only some of the highlights of the potential impacts. The NEPA documents and analyses from the agencies describe additional adverse effects in both Project Areas.
Despite these effects, the defendants insist that neither Project will adversely modify lynx critical habitat because the amount of critical habitat that would be adversely affected is relatively small in comparison to the total unaffected critical habitat. Their argument has some support in Ninth Circuit law. See e.g. Butte Envtl. Council v. U.S. Army Corps of Engrs., 620 F.3d 936, 948 (9th Cir.2010) (“An area of a species’ critical habitat can be destroyed without appreciably diminishing the value of a critical habitat for the species’ survival or recovery.”); see also Conservation Congress v. U.S. Forest Serv., 720 F.3d 1048, 1056-58 (9th Cir. 2013).
The problem here, however, is that the agencies reached their no-adverse-modification conclusion by relying heavily on the flawed Lynx Amendment and the biological opinion for the Lynx Amendment. The Forest Service blends its analysis of impacts on primary constituent elements for lynx critical habitat with its analysis of standards from the Lynx Amendment. And the Fish and Wildlife Service bases its conclusion of no adverse modification largely, if not entirely, on the Projects’ compliance with the standards in the Lynx Amendment. It finds, for example, that sufficient critical habitat will remain after the Projects are implemented because the biological opinion for the Lynx Amendment determined that up to 52,200 acres of fuel reduction treatment can take place in the Wildland Urban Interface on the Gallatin National Forest under exceptions to the Lynx Amendment standards.
The problem with these analyses is that they are fundamentally dependent on the Projects’ purported compliance with the flawed Lynx Amendment. As discussed above, the agencies did not reinitiate consultation on the Lynx Amendment after the designation of lynx critical habitat, so they have not determined how the standards and guidelines in the amendment will impact lynx critical habitat. For example, in light of the critical habitat designation, would the agencies still conclude that up to 52,200 acres of fuel reduction can occur in the Wildland Urban Interface on the Forest? Perhaps, but that is for the agency to determine, not this Court.
While the agencies discussion of the Lynx Amendment does not preclude a finding of no adverse modification for the Projects challenged here, the agencies have not provided an independent justification for such a finding apart from their reliance on the Lynx Amendment, which is an essential component of the second step of the burden shifting approach adopted by this Court.
The agencies might have been able to make this showing by demonstrating that the Projects will not alter the primary constituent elements of lynx critical habitat, without regard to compliance with the Lynx Amendment. But the agencies did not do this. Their analyses of the primary constituent elements for lynx critical habitat and their analyses of the standards and guidelines in the flawed Lynx Amendment are inextricably intertwined and incapable of separation. As a result, the agencies’ determination that the Projects will not *1207adversely modify lynx critical habitat is unreliable. The agencies have therefore failed to meet their burden of showing that the Projects will not adversely modify lynx critical habitat. See Wash. Toxics, 413 F.3d at 1035. For that reason, the Projects must be enjoined pending reinitiation of consultation and any subsequent, required NEPA analyses.
II. Grizzly bears
The plaintiffs raise several claims regarding grizzly bears. They argue that there is no incidental take statement that applies to the Projects, that the Forest Service has violated its annual reporting requirement for grizzly bears, and that the NEPA analysis for grizzly bears is inadequate. Each argument fails.
A. Incidental take statement
When an agency action is likely to cause a “take”5 of a listed species, the Fish and Wildlife Service may issue an “incidental take statement” that sets out the predicted impact on the species, as well as the terms and conditions of the action that will minimize takes. See 16 U.S.C. § 1536(b)(4); 50 C.F.R. § 402.14(i).
The plaintiffs argue that the agencies violated the ESA because there is no valid incidental take statement for the Gallatin Forest Plan that applies to grizzly bears outside the Yellowstone Grizzly Bear Recovery Zone, which they contend is a problem because the Forest Service acknowledges that the Bozeman Project, which is outside the Recovery Zone, will likely result in non-lethal takes of grizzly bears. The Service, however, insists that there is a valid incidental take statement that applies to the Bozeman Project — the incidental take statement accompanying the September 2006 Biological Opinion on the Effects to Grizzly Bears and Bald Eagles from the Gallatin National Forest Travel Plan.
The incidental take statement for the 2006 Travel Plan covers incidental takes associated with access management activities — e.g. roads and road use. The statement covers the entire Gallatin National Forest, both inside and outside the Recovery Zone.
There is also an incidental take statement that applies specifically to the anticipated effects on grizzly bear from helicopter logging for the Bozeman Project.
As for the East Boulder Project, the agencies determined that the Project will not adversely affect grizzly bears, so no specific incidental take statement was issued. Nevertheless, to the extent that an incidental take statement might be required for the East Boulder Project, the Forest Service maintains that the 2006 Travel Plan incidental take statement covers the Project.
The plaintiffs acknowledge that the 2006 Travel Plan incidental take statement “potentially addresses” Forest-wide management for grizzly bears outside the Recovery Zone, including the Bozeman Project Area. But, they argue, there is a problem with this incidental take statement — namely, the “secure habitat” standard.
For areas outside the Recovery Zone, the 2006 Travel Plan incidental take statement used “secure habitat” as a surrogate measure of take. In other words, with this approach, the agencies use the amount of secure habitat in an area to predict or measure takes. More secure habitat means fewer takes. The Service defines “secure habitat” as an area at least 10 acres in size and more than 500 meters from an open or gated motorized access route or reoccurring helicopter flight line.
*1208The plaintiffs argue that the “secure habitat” standard is no longer valid and that it does not represent the best available science because it is derived from the Grizzly Bear Conservation Strategy, which applies only once grizzlies are delisted under the ESA. Since Yellowstone grizzly bears were relisted in 2010 and remain listed, the Conservation Strategy — and, therefore, the secure habitat standard — no longer apply.
The Forest Service does not appear to disagree with the plaintiffs’ assessment that the Conservation Strategy is no longer binding. Instead, the Service argues that, regardless of whether it is binding, it still represents the best available science and that the agencies appropriately relied on it for the Travel Plan’s incidental take statement.
The plaintiffs, on the other hand, argue that analyzing only secure habitat does not represent the best available science. Instead, the agencies should have also analyzed road density outside the Recovery Zone. They point to the Interagency Grizzly Bear Committee Guidelines and the Schwartz et al. (2010) study in support of their argument. The Schwartz study, for example, states: “The most important predictors of survival in our best model were the amount of secure habitat within a bear’s home range and road densities outside of secure habitat.”
The ESA requires the agencies to use the best available science when conducting their analysis. 16 U.S.C. § 1536(a)(2). “Courts grant considerable deference to agencies on issues requiring great technical expertise, including the important question of what is the best available science.” W. Watersheds Project v. Salazar, 766 F.Supp.2d 1095, 1114 (D.Mont.2011) (citing Ecology Ctr. v. Castaneda, 574 F.3d 652, 658-59 (9th Cir. 2009)). Courts do not weigh competing scientific analyses. Id. (citing Lands Council v. McNair, 537 F.3d 981, 988 (9th Cir.2008), overruled on other grounds, Am. Trucking Assns. Inc. v. City of L.A., 559 F.3d 1046, 1052 (9th Cir.2009)). A plaintiff alleging that an agency did not use the best available science should be able to “cite[ ] ... scientific studies that indicate the [agency’s] analysis is outdated or flawed” or “scientific information directly undermining” the agency’s conclusion. Ecology Ctr., 574 F.3d at 659-60.
Here, the agencies are entitled to deference on their determination of the best available science and the use of the secure habitat metric. While the Schwartz study states that a road density metric is also a useful measure of grizzly bear survival, that conclusion does not “directly underminfe]” the agencies’ use of only the secure habitat standard or its opinion that the Conservation Strategy represents the best available science. See Ecology Ctr., 574 F.3d at 659-60. The study does not say that measurement of road density is necessary to measure impacts on grizzly bears. Moreover, the Schwartz study recognizes the importance of the secure habitat metric, which depends, in part, on a measurement of roads in the given area.
This is not an instance where the agencies’ own scientists identified the best available science or applicable standards but then the agencies failed to follow it. See e.g. Native Ecosystems Council v. Weldon, 848 F.Supp.2d 1207, 1217 (D.Mont. 2012). Instead, the agencies identified what they believe is the best available science, explained why they relied on it, and adhered to it in their analysis. Moreover, the agencies did not ignore road density, as the plaintiffs seem to suggest. The record shows that the agencies thoroughly considered it in their analysis.
The Court will not attempt to weigh competing scientific analyses, assuming that the Schwartz study somehow conflicts *1209with the agencies’ use of the secure habitat standard. This is an appropriate issue warranting deference.
The incidental take statement for the 2006 Travel Plan therefore applies here.
B. Annual reporting requirement
The plaintiffs argue that, even if there is a valid incidental take statement, the agencies are violating the terms and conditions of that statement. The plaintiffs’ argument is based on the annual reporting requirement in the incidental take statement for the 2006 Travel Plan. The statement requires the Forest Service to annually report changes in secure habitat and roads, among other things. The plaintiffs claim the Service failed to do this for a number of years.
The Forest Service correctly responds that it did not need to submit reports for the 2007-2009 years because the grizzly bear was delisted during that time. Although the reports may not have been prepared as regularly as the plaintiffs would like, the Forest Service nevertheless submitted an annual report to Fish and Wildlife Service in 2012, addressing secure habitat changes for 2011 and prior years. The Forest Service has therefore not violated its annual reporting requirement for the incidental take statement.
C. NEPA
Finally, the plaintiffs argue that the NEPA analysis for the Bozeman and East Boulder Projects, as it pertains to grizzly bears, is inadequate for a few reasons. First, they argue that the analyses for the Projects fail because the Forest Service failed to analyze the secure habitat standard that it relied on in the incidental take statement. Second, the plaintiffs argue that the incidental take statements for the 2006 Travel Plan and the Bozeman Project allow unlimited takes for the duration of helicopter logging activity. Finally, they argue that the ongoing effects of displacement from roads and helicopter activity after completion of the Bozeman Project are not covered by any incidental take statement. The arguments have no merit.
1. Secure habitat standard
The plaintiffs insist that, in its NEPA analysis for the Projects, the Forest Service never applied the secure habitat standard from the 2006 Travel Plan incidental statement, upon which it relied. The Court disagrees.
As an initial matter, the Service was not required to apply this standard to the East Boulder Project because the agencies concluded that the Project would not likely have an adverse affect on grizzly bears. So no incidental take statement — and no application of the secure habitat standard — was required for the Project.
In its incidental take statement that applied specifically to the Bozeman Project, the Fish and Wildlife Service wrote:
The road use associated with this project would not impart any effects to grizzly bears in addition to those analyzed in the 2006 biological opinion and the proposed project would be in compliance with the incidental take statement of that opinion.
Throughout the Bozeman Project Biological Opinion, the Fish and Wildlife Service repeatedly wrote that the Project would not run afoul of the 2006 Travel Plan Biological Opinion and incidental take statement. The Fish and Wildlife Service wrote in the Bozeman Project Biological Opinion:
The 2006 biological opinion on the effects of the Travel Management Plan on grizzly bears (U.S. Fish and Wildlife Service 2006) provides an incidental take statement concerning the effects of roads on grizzly bears. In doing so, the *1210effects of the existing forest roads and temporary project roads were analyzed and the effects of access management on the Forest, including the action area, were fully considered in the analysis in the 2006 biological opinion. The road use associated with this project would not impart any effects to grizzly bears in addition to those analyzed in the 2006 biological opinion and the proposed project would be in compliance with the incidental take statement of that opinion. Therefore, consultation on the effects of roads to grizzly bears is complete and the use of existing roads and construction and use of temporary roads will not be considered further.
While the Fish and Wildlife Service did not repeat the precise secure habitat standards from the 2006 Travel Plan Biological Opinion, the agencies did not act in an “arbitrary and capricious” manner.
As discussed above, review under the arbitrary and capricious standard is narrow. An agency’s action is arbitrary and capricious only if the agency: (1) relied on factors Congress did not intend it to consider, (2) entirely failed to consider an important aspect of the problem, or (3) 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. Lands Council, 537 F.3d at 987.
Here, the record shows that the agencies adequately considered the Bozeman Project’s effects on grizzly bears within the specific Project Area, and they did not offer an explanation that is counter to the evidence or that is implausible.
2. Helicopter takes
The plaintiffs next argue that the Bozeman Project’s incidental take statement violates NEPA because it allows unlimited takes during the duration of the helicopter logging activity. This argument fails.
The Fish and Wildlife Service wrote in the Bozeman Project Biological Opinion that helicopter logging might result in some non-lethal takes: “Helicopter logging in occupied grizzly bear habitat during the non-denning period may elicit a response in grizzly bears. Effects may range from a simple awareness of the helicopter, short-term disturbance or flight response or displacement from an area.”
The Service thoroughly explained that, since these non-lethal takes are not easily quantified, it had to employ a surrogate measure for incidental takes. That surrogate measure was 144 days of helicopter activity over a five year span, with multiple helicopter flights per day. If the helicopter activity must exceed 144 days, then the Service must reinitiate consultation.
The Fish and Wildlife Service’s analysis is not arbitrary and capricious. It does not permit unlimited takes. Instead, recognizing the difficulty in quantifying the number of non-lethal takes, it establishes a surrogate measure of 144 days of helicopter flights over five years. The Service did not fail to consider an important part of the problem, and it did not offer an explanation that runs counter to the evidence or is implausible. Lands Council, 537 F.3d at 987.
3. Ongoing effects
The plaintiffs’ final argument with respect to grizzly bears is that the agencies have failed to consider the ongoing effects on grizzly bears after completion of the Projects. The plaintiffs insist that takes of grizzly bears beyond the five-year project duration is not permitted by any incidental take statement. Again, the Court disagrees.
*1211The incidental take statement expressly recognizes that “under use [by grizzly bears] could last for the five years of activity, and for a few years ... until wary females regain full use of their home range.” This under use is part of the take contemplated by the agencies, anticipated by the agencies, and permitted by the incidental take statement.
III. Goshawk and marten monitoring, snags, and sensitive or management indicator species
The plaintiffs argue that the Forest Service has not complied with Forest Plan and NEPA requirements for Northern goshawk and marten monitoring, snag density, and monitoring for several sensitive species. These claims fail.
A. Goshawk and marten monitoring
The plaintiffs first claim that the Forest Service has not properly monitored two management indicator species: the Northern goshawk and the American pine marten. In particular, they argue, the Service has not determined population trends for the two species, and it has improperly used the presence of old growth habitat as a proxy for goshawk and marten populations. The Service responds that the plaintiffs have misconstrued the requirements of the Forest Plan. The Service has the more persuasive position.
The Forest Plan for the Gallatin National Forest identifies goshawk and marten as management indicator species for old growth sites. The Plan states that “indicator species ... will be monitored to determine population changes.” It directs the monitoring to take place every five years with “moderate” precision and reliability.
The plaintiffs claim that, instead of directly monitoring goshawk and marten, the Service used a proxy-on-proxy or “habitat proxy” approach. In other words, the Service measured the amount of old growth as a proxy for measuring the actual number of goshawk and marten. The plaintiffs claim this approach was inadequate and that there is inadequate old growth in the Project Areas.
The problem with the plaintiffs’ argument is that there is no evidence in the record that the Forest Service ever took this approach. Instead, the record shows that the Service did, in fact, directly monitor goshawk and marten populations.
In its 2011 monitoring, the Service reported that goshawk are “fairly common and widely distributed in the roaded (or more managed) portions” of Forest lands in the Northern Region. Numerous goshawk sightings were reported, and the Service concluded that goshawk appear “to be stable and cycling at low numbers” with “more than enough habitat to maintain a minimum viable population.”
Moreover, the East Boulder Project EA offers extensive analysis and discussion of the impacts on goshawk and pine marten, as well as their habitat. The EA concludes that the Service is satisfying the Forest Plan standards for these two species.
While the plaintiffs repeatedly claim that the Projects will degrade goshawk and marten habitat, they do not point to a single standard that the Service has purportedly violated with respect to goshawk, marten, or their habitat. The Forest Service’s only requirement under the Plan with respect to goshawk and marten is to monitor them to “determine population changes.” The Service did this. Even though the Service was not able to derive statistically significant population trends as of 2011, the Service was able to look back to data from 2005 and determine how the population is changing — it “appear[s] to be stable and cycling at low numbers.” *1212The Service has therefore met its obligations under the Forest Plan.
Finally, the plaintiffs also argue that there are little or no goshawk or marten in the Project area, so these species are not appropriate species to monitor as management indicator species. But, again, the plaintiffs do not point to any standard that requires these species to be found in the Project area. Contrary to plaintiffs’ arguments, the Service did not violate any aspect of the Forest Plan or NEPA with respect to its monitoring of goshawk or pine marten.
B. Snag standards
The Forest Plan requires the Forest Service to manage sensitive species habitat in order to maintain the species. The plaintiffs argue that several of these species are dependent on “snags” — standing dead trees at least 18 feet tall and at least 10 inches in diameter at breast height— and that the snag standards are not being met in the East Boulder Project Area.
The snag standards are in the Forest Plan’s Snag Amendment. The standards are intended to guide the Forest Service when they plan timber harvests. There are two standards relevant here. First, when harvesting timber, the Service must leave an average of 30 snags per 10 acres. If there are not 30 snags per acre, then “the largest available dead trees will be left as snags.” Second, the Service must leave an average of 30 “live snag replacement trees” per 10 acres.6
The Forest Service wrote a report for snag habitat and snag dependent species in preparation for the East Boulder Project. The Service found that the “East Boulder project would reduce the number of snags, down logs, and live replacement trees in the project area.” And, “Removal of these materials could reduce habitat suitability for snag-dependent species.”
The report also notes that the snag standards are not being met in many areas throughout the Project Area: “Many of the proposed treatment units currently have few snags available and do not currently meet the Forest Plan snag standard of 30 snags per 10 acres.” The Service explained that the reason for the low snag density in these areas is that “[s]nag poor units have typically had past management (thinned, cut, or partially cut with the past 20 years or so) resulting in younger, healthy trees, with few snags.”
The plaintiffs take these findings to mean that the Project cannot go forward. Not so. In areas with a snag density greater than 30 snags per 10 acres, nothing prevents the Forest Service from removing snags, as long as the Service leaves at least 30 snags per 10 acres. And, relevant here, for areas where the snag density is less than 30 snags per 10 acres, the snag standards require the Forest Service to leave “the largest available dead trees ... as snags.” In both the snag report and the EA, the Forest Service stated that it would do just that. The Project will therefore not violate the snag standards.
C. Monitoring sensitive species
The plaintiffs also make claims related to specific sensitive species, arguing that the Forest Service has failed to monitor those species and failed to consider project impacts on those species. Those species include the flammulated owl, long-eared and long-legged myotis, and black backed woodpecker.
*1213The Forest Plan states that “habitat that is essential for species identified in the Sensitive Species list developed for the Northern Region will be managed to maintain these species.” Contrary to the plaintiffs contention, the Forest Service analyzed impacts on the above species in its EA. The Forest Service explained that the Project will not have an impact on the black-backed woodpecker and “will not likely result in a trend toward federal listing or reduced viability” of the flammulated owl and long-eared and long-legged myotis. The Forest Service did not fail to take a hard look at the Project impacts on these species.
As to the plaintiffs’ argument regarding monitoring, they do not point to any monitoring requirement in the Forest Plan. The Forest Plan requires only that “habitat that is essential for species identified in the Sensitive Species list developed for the Northern Region will be managed to maintain these species.” The plaintiffs have not shown how the Forest Service has violated this directive.
IV. Roadless and wilderness study areas
Some of the proposed timber harvesting for the Bozeman Project takes place within an inventoried roadless area and near a wilderness study area. The plaintiffs assert that the harvesting violates the Roadless Rule and NEPA. This claim fails.
A. Impacts on roadless characteristics and Wilderness potential
The plaintiffs claim that the Bozeman Project violates NEPA because the Forest Service did not disclose or take a hard look at the Project’s impacts on roadless characteristics and Wilderness potential. The Service, though, adequately addressed the potential impacts.
The Bozeman Project authorizes commercial logging, prescribed burning, and mechanical slash piling in the Gallatin Fringe Inventoried Roadless Area. It also authorizes temporary road construction, commercial logging, prescribed burning, and mechanical slash piling in an unroaded area outside of, but adjacent to, both the Gallatin Fringe Inventoried Roadless Area and the Hyalite Porcupine Buffalo Horn Wilderness Study Area.
The plaintiffs argue that the Forest Service violated NEPA by failing to consider the irreversible adverse effects that logging and temporary road construction would have on roadless characteristics and Wilderness potential. As discussed above, an agency violates NEPA in only one of three circumstances: (1) it relied on factors Congress did not intend it to consider, (2) it entirely failed to consider an important aspect of the problem, or (3) it 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. Lands Council, 537 F.3d at 987.
Here, the plaintiffs do not argue that the Service relied on factors that Congress did not intend it to consider. And the record shows that the Service considered important aspects of the problem — impacts on roadless characteristics and Wilderness potential. The question, then, is whether the .Forest Service offered an adequate explanation for why the Bozeman Project would not have long-term, significant impacts on roadless characteristics and Wilderness potential.
The Forest Service concluded in the Final Environmental Impact Statement that there would be short term impacts but no long-term, irreversible impacts:
There would be no irreversible (total loss) loss of resources of either the Gal*1214latin Fringe IRA or the unroaded lands, which would eliminate possibility of the roadless area to be designated as wilderness at some future date because the vegetation would recover over time and temporary roads do not eliminate the area from being considered for its wilderness potential. There would be some short term (15-25 years) irretrievable changes to roadless character because of the vegetation disturbance and temporary road construction.
The plaintiffs counter: “The agency cannot discount impacts by pointing to a possible time decades into the future where the effects may not longer be evident....” The plaintiffs cite two cases in support of their argument: Lands Council v. Martin, 529 F.3d 1219 (9th Cir.2008), and Smith v. U.S. Forest Service, 33 F.3d 1072 (9th Cir.1994). Relying on these cases, the plaintiffs argue that the Forest Service cannot log or build temporary roads in roadless areas, unroaded areas, or areas that might someday be designated as Wilderness under the Wilderness Act.
The two cases, though, stand only for the proposition that, when an agency is considering logging in roadless areas, the logging is “environmentally significant” and “its environmental consequences must be considered.” Lands Council, 529 F.3d at 1230 (citing Smith, 33 F.3d at 1078-79). The cases do not prohibit logging or temporary road construction in roadless areas or areas that might someday be designated Wilderness. And the plaintiffs do not point to any other sources that set out such a prohibition.
The Forest Service acknowledges that the Bozeman Project will result in impacts that might last up to 25 years. But, as the Service explained, the impacts are temporary and would no longer be evident over time. The plaintiffs have not shown that the Service “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.” Lands Council, 537 F.3d at 987. The Service adequately considered the Bozeman Project’s impacts on roadless characteristics and Wilderness potential.
B. The Roadless Rule
The 2001 Roadless Rule generally “prohibits ... timber harvest in inventoried roadless areas” on national forests. 66 Fed.Reg. 3244. There are exceptions, though. Section 24.13(b)(1) provides:
(1) The cutting, sale, or removal of generally small diameter timber is needed for one of the following purposes and will maintain or improve one or more of the roadless area characteristics as defined in § 294.11.
(i) To improve threatened, endangered, proposed, or sensitive species habitat; or
(ii) To maintain or restore the characteristics of ecosystem composition and structure, such as to reduce the risk of uncharacteristic wildfire effects, within the range of variability that would be expected to occur under natural disturbance regimes of the current climatic period____
In other words, whether the Service may harvest timber in an inventoried roadless area is a three-step inquiry. First, the timber to be harvested must be “generally small diameter.” Second, the harvest must be needed for one of the two listed purposes. Third, the harvest must maintain or improve one or more of the roadless area characteristics as defined in § 294.11.
1. “Generally small diameter timber”
The plaintiffs claim that the timber that the Forest Service plans to harvest is not “generally small diameter timber.”
*1215The Forest Service commented in the preamble to the Roadless Rule that the term “generally small diameter timber” is a relative term “[bjecause of the great variation in stand characteristics between vegetation types in different areas.” 66 Fed.Reg. 3244, 3257. Consequently, determinations of what constitutes “generally small diameter timber” are “best made through project specific analysis or land and resource management plan NEPA analyses, as guided by ecological considerations.” 66 Fed.Reg. 3244, 3257.
Here, the Record of Decision states, “Generally small diameter timber will be removed with the emphasis on retaining healthy trees that are more resilient to wildland fires.” The Service further explained that “average diameter of trees to be removed is 10-12 inches and the overall stand diameter will increase when the fuel reduction project is complete.”
Some larger trees will be harvested, but the trees to be harvested are generally smaller compared to the rest of the trees in the treatment areas because the average tree diameter in the stands will increase once the trees are harvested. The Roadless Rule provides the Service flexibility when determining what constitutes “generally small diameter timber.” And the Ninth Circuit has expressly held that agencies are entitled to a great deal of deference when it comes to how they interpret “vague, undefined” directives. Lands Council, 537 F.3d at 993. The Service did not make a “clear error of judgment,” id., when it concluded that the timber to be harvested is “generally small diameter timber.”
2. “Maintain or restore the characteristics of ecosystem composition and structure”
Next, the plaintiffs argue that the timber harvest will not “maintain or restore the characteristics of ecosystem composition and structure.” Again, the plaintiffs’ argument fails.
The Roadless Rule specifically states that wildfire risk is an element of ecosystem composition and structure that the Service may control through timber harvesting. Timber harvesting may be permitted in roadless areas if, for example, its purpose is “to reduce the risk of uncharacteristic wildfire effects, within the range of variability that would be expected to occur under natural disturbance regimes of the current climatic period.” 66 Fed.Reg. 3244, 3273 (36 C.F.R. § 294.13(b)(1)(ii)).
Here, the Forest Service claims that the timber harvest will restore the natural composition of the timber stands because the areas are currently an unnatural, high fire hazard. The Service explained in the Record of Decision that the Project will “restore ecosystem composition and structure” by “retaining trees that are spaced so they are more resilient to fires, and reducing fuel loads to reduce the risk of uncharacteristic wildfire effects.”
The Forest Service also explained why stand conditions are currently unnatural in the roadless areas:
Although fire is the primary disturbance process in this forest type, wildfire has not been allowed to burn in this area due to past fire policy, the presence of municipal watershed and adjacent private land development. As a result the stand conditions are likely much denser than they would have been historically and are more likely to support stand-replacing fires.
This rationale is explained in more detail in the Service’s fire and fuels specialist report.
The plaintiffs respond that the Project activities will actually create unnatural conditions because it will reduce the canopy cover from 70-75% to 40-45%. There *1216are two problems with this critique. First, the plaintiffs do not provide any support for the argument that 70-75% canopy cover (and not 40-45%) is natural. Their explanation is that since no cutting has occurred in the proposed treatment area for the last several decades, the timber has been allowed to naturally grow and mature. But this rationale ignores the Service’s explanation that the growth has been unnatural because of aggressive fire suppression. Second, reducing the canopy cover is the very purpose of the harvest because the high canopy cover purportedly creates an unnatural fire hazard.
The plaintiffs have not shown that the timber harvest will create unnatural conditions or, more specifically, will not “maintain or restore the characteristics of ecosystem composition and structure.” 66 Fed.Reg. 3244, 3273 (36 C.F.R. § 294.13(b)(1)(ii)). The Forest Service has adequately explained how the timber harvest will “restore the characteristics of ecosystem composition and structure.” 66 Fed.Reg. 3244, 3273 (36 C.F.R. § 294.13(b)(1) (ii)) -
3. “Maintain or improve one or more of the roadless area characteristics”
The final criteria that must be met in order for a timber harvest to be permitted in a roadless area is: “The cutting, sale, or removal ... will maintain or improve one or more of the roadless area characteristics.” 66 Fed.Reg. 3244, 3273 (36 C.F.R. § 294.13(b)(1)). Here, the Forest Service claims that the timber harvest will maintain or improve the roadless area as a source for public drinking water. According to the Service, an uncharacteristic wildfire would cause excessive ash and runoff that would contaminate the supply of public water. The plaintiffs counter that the timber harvest is not necessary to improve the water supply because Bozeman plans to upgrade its water filtration system to handle that runoff.
The Forest Service explained in its Supplemental Final Environmental Impact Statement for the Bozeman Project how the timber harvest would improve the public water supply — a large wildfire in the watersheds could result in loss of water from a few days to several weeks, particularly if heavy rainfall occurs within 2 years of a major wildfire. Heavy rainfall would cause ash and sediment to contaminate the water supply on account of increased erosion resulting from a wildfire. The Forest Service acknowledged that the City of Bozeman intends to upgrade its water filtration system, but, even with the upgrade, “[Tjreatment would still be difficult for the system in the event of large influxes of sediment and ash.”
The plaintiffs only response is that the Forest Service has not sufficiently explained why the harvest is necessary in light of the filtration upgrade. The Forest Service, though, acknowledged the upgrade and explained that, even with the upgrade, ash and sediment runoff will still cause water treatment problems. The plaintiffs’ only evidence to the contrary is a newspaper article that says the filtration system “will be able to handle the sediment form a fire,” but that article alone is not sufficient evidence to undermine the Forest Service’s conclusion. Indeed, in that same article, the water treatment supervisor said, “I can’t guarantee it could handle everything....” Even if the filtration system helps remove contaminants, the plaintiffs have not refuted the Service’s conclusion that “treatment would still be difficult for the system in the event of large influxes of sediment and ash.”
Furthermore, in its amicus brief, the City of Bozeman confirms that the Sourdough Water Treatment Plant will not be able to process an unlimited quantity of *1217degraded source water and that the plant would very likely shut down during periods when it receives degraded water.
The Forest Service has adequately shown that the project will “maintain or improve one or more of the roadless area characteristics.” 66 Fed.Reg. 3244, 3273 (36 C.F.R. § 294.13(b)(1)). Since the Project meets all the requirements for the timber harvesting exception to the Road-less Rule, the Project does not violate the Rule.
Conclusion
The plaintiffs levy several claims, but the plaintiffs are entitled to summary judgment only on their claims regarding lynx critical habitat and reinitiation of consultation for the Northern Rockies Lynx Amendment under the Endangered Species Act. The defendants are entitled to summary judgment in all other respects.
IT IS ORDERED that the motions for summary judgment (docs. 21, 32) are GRANTED IN PART and DENIED IN PART. The Court GRANTS the plaintiffs’ motion for summary judgment (doc. 21) on their first and second claims for relief— that is, them claims related to lynx critical habitat and reinitiation of consultation for the Northern Rockies Lynx Amendment. To this extent, the defendants’ motion for summary judgment (doc. 32) is DENIED.
This matter is REMANDED to the Forest Service so that it may reinitiate consultation for the Northern Rockies Lynx Amendment, consistent with this Court’s decision in Salix v. U.S. Forest Service, 2013 WL 2099811, 944 F.Supp.2d 984 (D.Mont.2013).
The defendants are ENJOINED from implementing the East Boulder Fuels Reduction Project and Bozeman Municipal Watershed Fuels Reduction Project on the Gallatin National Forest, pending completion of reinitiated consultation and any further procedures that might be required under the National Environmental Policy Act in light of the findings from that consultation.
IT IS FURTHER ORDERED that the Court GRANTS the defendants’ motion for summary judgment (doc. 32) on all remaining claims and issues. To this extent, the plaintiffs’ motion for summary judgment (doc. 21) is DENIED. IT IS FURTHER ORDERED that the Clerk of Court is directed to enter judgment and close this case.
. When the Ninth Circuit held in Burlington Northern that "[t]he plaintiff must make a showing that violation of the ESA is at least likely in the future,” it relied solely on Amoco Production Co. v. Village of Gambell, 480 U.S. 531, 545, 107 S.Ct. 1396, 94 L.Ed.2d 542 (1987). In Amoco Production Co., the U.S. Supreme Court held that in NEPA cases, if the injury to the environment "is sufficiently likely, ... the balance of harms will usually favor the issuance of an injunction.” 480 U.S. at 545, 107 S.Ct. 1396. That case’s applicability to ESA cases might be dubious, since the traditional Winter standard, and the requisite showing of likelihood of irreparable harm, applies to NEPA cases but not ESA cases.
. The mere fact that an agency discusses or mentions the Lynx Amendment in the administrative record would not preclude an agency from meeting its burden. Instead, the Court must disregard that portion of the analysis and look to some other independent basis for evidence that the project will not cause any harm.
. The Court anticipates that the first step of the burden shifting approach, which requires the plaintiff to allege a specific irreparable harm resulting from an ESA violation, will be satisfied in most instances by specific allegations in the complaint.
. The plaintiffs do not specifically allege harm to lynx themselves. They allege only harm to critical habitat. This distinction is important because, as the Court observed in Salix, " 'The analysis of the effects to critical habitat is a separate and different analysis from that of the effects to the species, and may provide *1205greater regulatory benefits to the recovery of a species than listing alone.’ ” at 1000, at *15 (quoting the administrative record).
. "The term 'take' means to harass, harm, pursue, hunt, shoot, wound, kill, trap, capture, or collect, or to attempt to engage in any such conduct.” 16 U.S.C. § 1532(19).
. For Douglas fir and subalpine fir on rocky or shallow soils, the Service must designate 60 trees per 10 acres as replacement trees.