We took this case en banc to clarify some of our environmental jurisprudence with respect to our review of the actions of the United States Forest Service.
The Lands Council and Wild West Institute (collectively, Lands Council) moved for a preliminary injunction to halt the Mission Brush Project (the Project), which called for the selective logging of 3,829 acres of forest in the Idaho Panhandle National Forest (IPNF). As the basis for the preliminary injunction, Lands Council claimed that Ranotta McNair and the United States Forest Service (collectively, the Forest Service), failed to comply with the National Forest Management Act (NFMA), 16 U.S.C. § 1600 et seq., the National Environmental Policy Act (NEPA), 42 U.S.C. § 4231 et seq., and the Administrative Procedure Act (APA), 5 U.S.C. § 701 et seq., in developing and implementing the Project.
Boundary County, City of Bonners Ferry, City of Moyie Springs, Everhart Logging, Inc., and Regehr Logging, Inc. (collectively, Intervenors) intervened on behalf of the Forest Service. The district court denied Lands Council’s motion for a preliminary injunction. A three-judge panel of this court reversed the district court’s decision and remanded for entry of a preliminary injunction in
Lands Council v. McNair,
I. FACTUAL AND PROCEDURAL BACKGROUND
A. Mission Brush Area
The Mission Brush Area (or Project Area) encompasses approximately 31,350 acres and is located in the northeastern portion of the Bonners Ferry Ranger District. Approximately 16,550 acres of the Project Area are National Forest System lands, which are home to a variety of species (or their habitats), including the northern gray wolf, Canada lynx, grizzly bear, black-backed woodpecker, flammu-lated owl, fisher, western toad, pileated woodpecker, and the white-tailed deer. The Project Area is also home to old-growth trees.
*985 The current structure and composition of the forest in the Project Area differs significantly from the forest’s historic composition. While the Project Area previously consisted of relatively open ponderosa pine and Douglas-fir stands, today it is crowded with stands of shade-tolerant, younger Douglas-firs and other mid-to-late-suceessional species. The suppression of naturally occurring fires, past logging practices, and disease are primarily responsible for this shift in forest composition.
The increased density of trees has proven deleterious to the old-growth trees and the Project Area’s ecology. First, old-growth trees need relatively open conditions to survive and maintain their growth rates. Second, the increased density is causing a decline in the health and vigor of all trees because they must compete for moisture, sunlight, and nutrients, and the densely clustered trees are less tolerant of insects and disease. Third, dense, dry forests are at risk for large, stand-replacing fires, due to the build-up of fuels. Lastly, wildlife species that prefer a relatively open forest composition with more old-growth trees have suffered a decline in habitat.
B. Mission Brush Project
The Forest Service proposed the Project, in part, to restore the forest to its historic composition, which, in the Forest Service’s assessment, is more likely to be sustainable over time. But this is not the Project’s only objective. According to the Supplemental Final Environmental Impact Statement (SFEIS) that the Forest Service issued in April 2006, the overall “objectives of the project are to begin restoring forest health and wildlife habitat, improv[e] water quality and overall aquatic habitat by reducing sediment and the risk of sediment reaching streams, and provid[e] recreation opportunities that meet the varied desires of the public and the agency while reducing negative effects to the ecosystem.” The Project proposes to accomplish these varied objectives through a number of actions, such as improving roads that presently contribute to sediment in the watersheds, decommissioning roads posing a great risk of contributing to sediment, ensuring that the Project Area has acceptable toilets and wheelchair accessible pathways to toilets, installing a boat ramp and fishing dock, and improving trails.
After considering multiple approaches on how best to accomplish the Project’s goals with respect to forest composition, including one no-action alternative, the Forest Service chose to implement a modified version of Alternative 2. In relevant part, Alternative 2 calls for silvicultural treatments on 3,829 acres of forest, fuels treatments on 3,698 acres, and ecosystem burns without harvest on 238 acres. The silvicultural treatments proposed include commercial thinning, 1 regeneration cuts, 2 and sanitation salvage harvesting. 3
*986 As a part of the Project, the Forest Service plans to treat 277 acres of dry-site old-growth stands in order to increase the overall quality of dry-site old-growth stands and scattered old-growth Douglas-fir, and to improve and maintain trees that could be old-growth in the future. Despite its plans to perform treatments within old-growth stands, the Project will not involve harvesting allocated old-growth trees. The Forest Service represented in the SFEIS that the allocated old-growth in the IPNF has not been harvested for several years, and that its “focus is on maintaining [existing] old growth stands ... and allocating additional stands for future old growth as they mature.” In those units containing old-growth trees, the Forest Service has identified those non-old-growth trees it plans to harvest.
The Project is expected to generate 23.5 million board feet of timber, which has been, or will be, sold pursuant to three timber sale contracts: the Brushy Mission Sale, the Haller Down Sale, and the Mission Fly By Sale. The Forest Service sold the Brushy Mission Sale to Everhart Logging, and the Haller Down Sale to Regehr Logging. The Forest Service received no bids for the Mission Fly By Sale, which contains all but fourteen of the old-growth acres that are part of the Haller Down Sale. Though logging under the Brushy Mission and Haller Down sales has already begun, the injunction imposed by the district court pursuant to the three-judge panel opinion in
Lands Council,
C. Procedural History
In late 2002, the Forest Service decided to undertake management activities in the Mission and Brush Creek areas. In 2003, the Forest Service issued a draft Environmental Impact Statement (EIS). After receiving public comments, the Forest Service released its final EIS and Record of Decision (ROD) in June 2004. Lands Council appealed the ROD. The Forest Service upheld the Project, but ordered the preparation of a supplemental EIS in light of this court’s decision in
Lands Council v. Powell (Lands Council I),
II. STANDARD OF REVIEW AND JURISDICTION
We have jurisdiction to review a district court’s denial of a preliminary injunction under 28 U.S.C. § 1292(a)(1). A district court’s decision regarding preliminary injunctive relief is subject to “limited and deferential” review.
Sw. Voter Registration Educ. Project v. Shelley,
A district court abuses its discretion in denying a request for a preliminary injunction if it “base[s] its decision on an erroneous legal standard or clearly erroneous findings of fact.”
Id.
(citation omitted). We review conclusions of law de novo and
*987
findings of fact for clear error.
Id.
Under this standard, “[a]s long as the district court got the law right, it will not be reversed simply because the appellate court would have arrived at a different result if it had applied the law to the facts of the case.”
Id.
(quoting
Earth Island Inst. v. U.S. Forest Serv. (Earth Island Inst. I),
“A preliminary injunction is appropriate when a plaintiff demonstrates ‘either: (1) a likelihood of success on the merits and the possibility of irreparable injury; or (2) that serious questions going to the merits were raised and the balance of hardships tips sharply in [the plaintiffs] favor.’ ”
Lands Council v. Martin (Lands Council II),
In deciding whether Lands Council is likely to succeed on the merits of its claims, we must remember that the APA provides the authority for our review of decisions under NEPA and NFMA.
Pit River Tribe v. U.S. Forest Serv.,
Review under the arbitrary and capricious standard “is narrow, and [we do] not substitute [our] judgment for that of the agency.”
Earth Island Inst. II,
III. DISCUSSION
Lands Council argues that, in developing the Project, the Forest Service violated the NFMA in two ways: (1) by failing to demonstrate the reliability of the scientific methodology underlying its analysis of the Project’s effect on wildlife, (specifically the flammulated owl and its habitat), 4 and (2) *988 by not complying with Standard 10(b) of the IPNF Forest Plan, which requires the Forest Service to maintain at least ten percent old-growth throughout the forest. Lands Council also argues that the Forest Service violated NEPA because, in Lands Council’s view, the Forest Service did not adequately address the uncertainty concerning its proposed treatment as a strategy to maintain species viability.
In essence, Lands Council asks this court to act as a panel of scientists that instructs the Forest Service how to validate its hypotheses regarding wildlife viability, chooses among scientific studies in determining whether the Forest Service has complied with the underlying Forest Plan, and orders the agency to explain every possible scientific uncertainty. As we will explain, this is not a proper role for a federal appellate court. But Lands Council’s arguments illustrate how, in recent years, our environmental jurisprudence has, at times, shifted away from the appropriate standard of review and could be read to suggest that this court should play such a role.
Below, we address each of Lands Council’s arguments. We first discuss the language and purpose of the NFMA and how, in
Ecology Center, Inc. v. Austin,
We are mindful, of course, that important environmental resources are at stake in cases such as this, and we strongly reaffirm that the Forest Service must fully comply with the requirements of the NFMA and NEPA. We conclude that the Forest Service has complied with those requirements in this case, and we affirm the district court’s denial of Lands Council’s request for a preliminary injunction.
A. The National Forest Management Act
1. Statutory Language And Purpose
The NFMA sets forth the statutory framework and specifies the procedural and substantive requirements under which the Forest Service is to manage National Forest System lands. Procedurally, the NFMA requires the Forest Service to develop a forest plan for each unit of the National Forest System. 16 U.S.C. § 1604(a). In developing and maintaining each plan, the Forest Service is required to use “a systematic interdisciplinary approach to achieve integrated consideration
*989
of physical, biological, economic, and other sciences.”
Id.
§ 1604(b). After a forest plan is developed, all subsequent agency action, including site-specific plans such as the Mission Brush Project, must comply with the NFMA and be consistent with the governing forest plan.
Id.
§ 1604(i);
see Idaho Sporting Cong., Inc. v. Rittenhouse,
Substantively, the NFMA requires the Secretary of Agriculture to develop guidelines “to achieve the goals of the Program,” including:
[P]rovid[ing] for diversity of plant and animal communities based on the suitability and capability of the specific land area in order to meet overall multiple-use objectives, and within the multiple-use objectives of a land management plan adopted pursuant to this section, provide, where appropriate, to the degree practicable, for steps to be taken to preserve the diversity of tree species similar to that existing in the region controlled by the plan ...
16 U.S.C. § 1604(g)(3)(B);
see Envtl. Prot. Info. Ctr. v. U.S. Forest Serv.,
The Project also must be consistent with the IPNF Forest Plan’s provisions regarding wildlife viability. See 16 U.S.C. § 1604(i). In the IPNF Forest Plan, the Forest Service designated the flammulat-ed owl, the only species at issue in this appeal, as a sensitive species. The IPNF Forest Plan requires the Forest Service to “[mjanage the habitat of species listed in the Regional Sensitive Species List to prevent further declines in populations which could lead to federal listing under the Endangered Species Act.” U.S. Dep’t of Agriculture, Forest Plan, Idaho Panhandle National Forests, at 11-28 (Aug. 1987), available at http://www.fs.fed.us/ *990 ipnf/eco/manage/forestplan/ [hereinafter IPNF Forest Plan].
Congress has consistently acknowledged that the Forest Service must balance competing demands in managing National Forest System lands. Indeed, since Congress’ early regulation of the national forests, it has never been the case that “the national forests were ... to be ‘set aside for non-use.’ ”
United States v. New Mexico,
Congress’ current vision of national forest uses, a broader view than Congress articulated in 1897, is expressed in the Multiple-Use Sustained Yield Act of 1960, 16 U.S.C. § § 528-31, which states that “[i]t is the policy of the Congress that the national forests are established and shall be administered for outdoor recreation, range, timber, watershed, and wildlife and fish purposes.” Id. § 528. The NFMA references 16 U.S.C. § § 528-531 and requires that plans developed for units of the National Forest System “provide for multiple use and sustained yield of the products and services obtained therefrom ... and [must] include coordination of outdoor recreation, range, timber, watershed, wildlife and fish, and wilderness[.]” Id. § 1604(e)(1). Thus, the NFMA is explicit that wildlife viability is not the Forest Service’s only consideration when developing site-specific plans for National Forest System lands.
2. We Overrule Ecology Center
Lands Council argues that the Forest Service violated the NFMA because it has not demonstrated the reliability of the scientific methodology underlying its analysis of the effect of the Project’s proposed treatment on the flammulated owl and its habitat. Relying primarily on Ecology Center, Lands Council specifically contends that the Forest Service erred by not verifying its prediction regarding the effect of treatment on old-growth species’ habitat with observation or on-the-ground analysis. We disagree, and hereby overrule Ecology Center. We also hold that the district court did not abuse its discretion in concluding that Lands Council is unlikely to succeed on the merits of this claim.
In
Ecology Center,
we relied on
Lands Council I
when we grafted onto our jurisprudence a broad rule that, in effect, requires the Forest Service to always “demonstrate the reliability of its scientific methodology” or the hypotheses underlying the Service’s methodology with “on the ground analysis.”
See Ecology Ctr.,
In
Lands Council I,
we reviewed the Forest Service’s approval of a timber harvest as part of a watershed restoration project in the IPNF.
In analyzing the quality of the soil in the project area, the Forest Service had not taken soil samples from the activity area, *991 but instead had relied on samples from other areas in the forest and on aerial photographs. Id. Despite the Forest Service’s representation that it had “tested similar soils within the Forest, and similar soils act the same way,” we rejected the Forest Service’s choice of scientific methodology because it was based entirely on a spreadsheet model with no on-site inspection or verification. Id. at 1034-35. We explained that “/ujnder the circumstances of this case, the Forest Service’s basic scientific methodology, to be reliable, required that the hypothesis and prediction of the model be verified with observation. The predictions of the model ... were not verified with on the ground analysis.” Id. at 1035 (emphasis added). We then held that the “Forest Service’s reliance on the spreadsheet models, unaccompanied by on-site spot verification of the model’s predictions, violated NFMA.” Id.
In
Ecology Center,
we applied an on-the-ground analysis requirement to our review of the Lolo National Forest Post Burn Project, in which the Forest Service proposed logging in old-growth forest and post-fire habitats.
Ecology Center
even suggests that such an analysis must be on-site, meaning
in the location of the proposed action.
There, we rejected the Forest Service’s argument that its on-the-ground soil analysis was “sufficiently reliable because it utilized data from areas with ecological characteristics similar to the proposed harvest units.”
Id.
at 1070. We noted that, as in
Lands Council I,
the Forest Service had not tested “much of
the activity area.” Id.
(quoting
Lands Council I,
We made three key errors in Ecology Center. First, we read the holding of Lands Council I too broadly. Second, we created a requirement not found in any relevant statute or regulation. And, third, we defied well-established law concerning the deference we owe to agencies and their methodological choices. Today, we correct those errors.
In
Lands Council I,
we expressly limited our holding that “on-site spot verification” was required for soil analysis to “the circumstances of [that] case.”
The Forest Service is at liberty, of course, to use on-the-ground analysis if it deems it appropriate or necessary, but it is
*992
not required to do so. As Judge McKeown explained in her dissent in
Ecology Center,
“there is no legal basis to conclude that the NFMA requires an on-site analysis where there is a reasonable scientific basis to uphold the legitimacy of modeling. NFMA does not impose this substantive requirement, and it cannot be derived from the procedural parameters of NEPA.”
The NFMA unquestionably requires the Forest Service to “provide for diversity of plant and animal communities ... in order to meet overall multiple-use objectives.” 16 U.S.C. § 1604(g)(3)(B). Similarly, the IPNF Forest Plan requires the Forest Service to “[m]anage the habitat of species listed in the Regional Sensitive Species List to prevent further declines in populations which could lead to federal listing under the Endangered Species Act.” IPNF Forest Plan, supra, at 11-28. However, despite imposing these substantive requirements on the Forest Service, neither the NFMA and its regulations 6 nor the IPNF Forest Plan specify precisely how the Forest Service must demonstrate that its site-specific plans adequately provide for wildlife viability.
Granting the Forest Service the latitude to decide how best to demonstrate that its plans will provide for wildlife viability comports with our reluctance to require an agency to show us, by any particular means, that it has met the requirements of the NFMA every time it proposes action. We have approved of forest plans when they are “based on the current state of scientific knowledge.”
See Seattle Audubon Soc’y v. Moseley,
Here, the record demonstrates that the federal defendants considered the viability of plant and animal populations based on the current state of scientific knowledge. Because of the inherent flexibility of the NFMA, and because there is no showing that the federal defendants overlooked any relevant factors or made any clear errors of judgment, we conclude that their interpretation and application of the NFMA’s viability regulations was reasonable.
Id. (citations omitted). Thus, we defer to the Forest Service as to what evidence is, or is not, necessary to support wildlife viability analyses.
Were we to grant less deference to the agency, we would be ignoring the APA’s arbitrary and capricious standard of review.
Ecology Center
illustrates the consequences of failing to grant appropriate deference to an agency. In
Ecology Cen
*993
ter,
we rejected reports establishing that soil analysis was conducted in the project area as “too few and of poor quality.”
See
Instead, our proper role is simply to ensure that the Forest Service made no “clear error of judgment” that would render its action “arbitrary and capricious.”
See Marsh v. Or. Natural Res. Council,
This approach respects our law that requires us to defer to an agency’s determination in an area involving a “high level of technical expertise.”
See Selkirk Conservation Alliance v. Forsgren,
Finally, this approach also acknowledges that “[w]e are not free to ‘impose on the agency [our] own notion of which procedures are ‘best’ or most likely to further some vague, undefined public good.’ ”
Churchill County v. Norton,
Thus, as non-scientists, we decline to impose bright-line rules on the Forest *994 Service regarding particular means that it must take in every case to show us that it has met the NFMA’s requirements. Rather, we hold that the Forest Service must support its conclusions that a project meets the requirements of the NFMA and relevant Forest Plan with studies that the agency, in its expertise, deems reliable. The Forest Service must explain the conclusions it has drawn from its chosen methodology, and the reasons it considers the underlying evidence to be reliable. We will conclude that the Forest Service acts arbitrarily and capriciously only when the record plainly demonstrates that the Forest Service made a clear error in judgment in concluding that a project meets the requirements of the NFMA and relevant Forest Plan.
For these reasons, we overrule Ecology Center and affirm that Lands Council I’s requirement of on-the ground analysis was limited to the circumstances of that particular case.
3. Reliability of the Forest Service’s Analysis Concerning The Effects Of Treating Old-Growth Habitat On The Flammulated Owl
Lands Council argues that the Forest Service violated the NFMA by failing to demonstrate the reliability of the scientific methodology underlying its analysis of the Project’s effect on wildlife, specifically the flammulated owl and its habitat. But the Forest Service supported its conclusions about the impact of the Project on the flammulated owl and its habitat with studies it deemed reliable. Moreover, the Forest Service did conduct on-the-ground analysis of the flammulated owl in an area straddled by the Mission Brush Project Area, even though, by overruling Ecology Center, this opinion confirms that such analysis is not required. See supra Part III.A.2.
These studies, together with the Forest Service’s reasonable assumption that enhancing the amount of flammulated owl habitat in the long-term will maintain the flammulated owl population, lead us to conclude that the Forest Service did not act arbitrarily and capriciously in determining that the Project met the substantive requirements of the NFMA and the IPNF Forest Plan regarding species diversity.
See Inland Empire Pub. Lands Council,
The Forest Service has provided studies evidencing that flammulated owls prefer old-growth habitat. See Montana Partners in Flight, Bird Conservation Plan — Flam- mulated Owl (2001) [hereinafter Montana Partners], at 2 (noting “a strong association between Flammulated Owls and old-growth ponderosa-pine/Douglas-fir habitat”); Idaho Partners in Flight, Idaho Bird Conservation Plan-Version 1.0; Implementation Schedule (2000), at 144 (“Old-growth trees are strongly correlated with nesting, singing, and foraging sites [of flammulated owls].”); R. Reynolds & B. Linkhart, Flammulated Owls in Pondero-sa Pine: Evidence of Preference for Old Growfth (1992), at 167 (“[Flammulated] owls settled into areas having greater proportions of old-growth ponderosa pine/Douglas-fir.”).
The Forest Service has also provided studies supporting its determination that the Project will maintain flammulated owl habitat because flammulated owls live in *995 old-growth habitat post-treatment. One study on the habitat selection of flammu-lated owls in British Columbia notes that “most [flammulated] owls were found in mature-old (100-200 yr) growth stands of Douglas fir that had been selectively harvested 20-30 years prior to [the] surveys.” R. Howie & R. Ritcey, Distribution, Habitat Selection, and Densities of Flammu-lated Owls in British Columbia (1987), at 251. While the study does not conclude that logging improves flammulated owl habitat, it documents flammulated owl presence within logged old-growth stands. 7 Other studies document this presence as well. See Montana Partners, supra, at 3 (acknowledging presence of flammulated owls in selectively logged sites in the Northern Rockies and stating that “[t]he Forest Service has an opportunity to manage restored acres to meet both the micr-ohabitat and landscape parameters of identified wildlife species, including the Flammulated Owl”).
Moreover, although it was not required to, the Forest Service conducted an on-the-ground analysis of flammulated owls in the Bonners Ferry Ranger district within the IPNF. Dawson Ridge Flammulated Owl Habitat Monitoring (June 30, 2006). The Dawson Ridge study monitored five 1/5 acre plots of flammulated owl habitat in an area that was treated with thinning and underburning in the mid-1970s, logged in 2000, and underburned in 2002. Id. at 1, 3. The researchers received one flammulated owl response in the 2006 survey, and recorded additional responses in 1999 and 2000. Id. at 1. It is within the Forest Service’s expertise, not ours, to determine the significance of these responses.
Although we acknowledge that this record is relatively sparse and approaches the limits of our deference, we nevertheless conclude that there is sufficient evidence to defer to the Forest Service’s conclusion that this survey response indicates that flammulated owls are using the monitored area. To determine whether deference is warranted, we look to the sufficiency of the evidence, not the size of the record. The Dawson Ridge study concluded that “[m]onitoring surveys confirm that owls are using the area after harvest,” and stated:
Although it is inappropriate at this time to assume that any of these silvicultural treatments improved (i.e., changed habitat from an unsuitable to suitable condition) flammulated owl habitat[,] it is encouraging given the management history of Dawson Ridge that owls are using the area. However, these positive responses do imply that our dry forest silvicul-tural practices are at least maintaining suitable habitat.
Id. at 3 (emphasis in original). Of course, neither the NFMA nor the IPNF Forest Plan require the Forest Service to improve a species’ habitat to prove that it is maintaining wildlife viability. 8
*996 Finally, the Forest Service used a habitat suitability model to analyze the potential effects of the proposed Project on the flammulated owl. Studies in the record reference the required size and continuity of habitat that the owls need to survive. The habitat suitability model predicted the change in suitable habitat 9 that would result from the treatment proposed in each of the Forest Service’s alternatives for the Project. The Forest Service explained its methodology for calculating the amount of habitat that would be suitable for the flam-mulated owl after the treatment. The Forest Service used vegetation characteristics to determine stands that were currently suitable habitat for flammulated owl, and wildlife biologists conducted site visits and interpreted aerial photographs to determine the suitability of stands deemed “capable.”
Based on its analysis, the Forest Service concluded that, though the disturbance imposed by the Project may have short-term negative impacts in the immediate vicinity of harvesting, there would be no decrease in suitable habitat in the short-term, and the Project “would promote the long-term viability of suitable Flammulated Owl habitat.” The Forest Service also concluded that the Project’s effects “would not indicate local or regional change in habitat quality or population status, allowing Flammulated Owls to maintain their current distribution,” and that it would not contribute to a trend toward a “Federal listing” under the Endangered Species Act or cause a loss of viability.
Today, as we have in the past, we approve, based on the record before us, of the Forest Service’s use of the amount of suitable habitat for a particular species as a proxy for the viability of that species.
See, e.g., Inland, Empire Pub. Lands Council,
In Inland Empire Public Lands Council, the plaintiffs challenged the Forest Service’s analysis of a timber sales project’s impact on seven sensitive species in the Kootenai National Forest. Id. at 757. The plaintiffs, several environmental groups, claimed that the Forest Service did not satisfy the NFMA because its population viability analysis was insufficient to ensure viable populations of the relevant species. Id. at 759-60. We approved of the Forest Service’s “habitat viability analysis,” which measured the amount of suitable habitat for the species at issue and then used that figure as a proxy to estimate a species’ population. Id. at 763. Using this “habitat as a proxy approach,” the Forest Service concluded that a species would remain viable on the basis of whether “the threshold percentage of each type of habitat remaining in the chosen alternative [after harvesting] was greater than the percentage required for that species to survive.” Id. at 759. 10 We charac *997 terized the Forest Service’s assumption that maintaining acreage necessary for survival would ensure a species’ survival as “eminently reasonable” and deferred to the Forest Service’s methodology. See id. at 760-61. 11
To always require a particular type of proof that a project would maintain a species’ population in a specific area would inhibit the Forest Service from conducting projects in the National Forests. We decline to constrain the Forest Service in this fashion. Were we to do so, we may well be complicit in frustrating one or more of the other objectives the Forest Service must also try to achieve as it manages National Forest System lands. See 16 U.S.C. § 528 (noting Congress’ policy that the National Forests are to be “administered for outdoor recreation, range, timber, watershed, and wildlife and fish purposes”).
The case before us resembles
Inland Empire.
As explained, the record includes studies describing the quality and quantity of habitat necessary to sustain the viability of flammulated owls, and the Forest Service has determined what habitat is currently suitable for the flammulated owl and what habitat would be suitable after the proposed project. While the project involves a disturbance in the forest to some extent, it is for the Forest Service to determine how the Project will affect the habitat of flammulated owls. In this case, the Forest Service has concluded that the current amount of suitable habitat will be maintained and that flammulated owls will be able to maintain their current distribution. That a proposed project involves some disturbance to the forest does not prohibit the Forest Service from assuming that maintaining a sufficient amount of suitable habitat will maintain a species’ viability. Indeed, the project in
Inland Empire
involved a plan to harvest trees, and that disturbance did not render the habitat as a proxy approach inapplicable.
To the extent we suggested in
Idaho Sporting Congress v. Thomas,
Of course, a reviewing court still must ensure that the Forest Service’s use of “habitat as a proxy” is not arbitrary and capricious. We therefore hold that when the Forest Service decides, in its expertise, that habitat is a reliable proxy for species’
*998
viability in a particular case, the Forest Service nevertheless 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.
See Earth Island Institute II,
Though some of our cases limit the use of habitat as a proxy,
see, e.g., Oregon Natural Resources Council Fund v. Goodman,
As explained, in this case, the Forest Service detailed the methodology it used for determining the amount of suitable habitat and acknowledged the assumptions underlying its use of habitat as a proxy. Although it is true that no flammulated owls were located in suitable habitat in a number of presence surveys, the Forest Service acknowledges that the nesting boxes used may have been placed too low on trees in some of these surveys. Moreover, the Forest Service has represented that it is difficult to detect flammulated owls, and we recognize that “monitoring difficulties do not render a habitat-based analysis unreasonable, so long as the analysis uses all the scientific data currently available.”
See Envtl. Prot. Info. Ctr.,
In light of the discussion above, the rule we set forth in
Native Ecosystems Council
remains good law: the Forest Service may meet wildlife “viability requirements by preserving habitat, but only where both the Forest Service’s knowledge of what quality and quantity of habitat is necessary to support the species and the Forest Service’s method for measuring the existing amount of that habitat are reason
*999
ably reliable and accurate.”
On the basis of the studies provided by the Forest Service and the Forest Service’s reasonable assumption that maintaining suitable habitat for the flam-mulated owl will also maintain a viable population of flammulated owls, we conclude that the district court did not abuse its discretion in deciding that Lands Council is not likely to succeed on this aspect of its NFMA claim.
4. Forest Service’s Compliance With Standard 10(b) Of The IPNF Forest Plan
The NFMA requires the Forest Service to comply with its established forest plan in all subsequent actions. 16 U.S.C. § 1604(i);
Inland Empire Pub. Lands Council,
The Forest Service has shown that it has complied with Standard 10(b), and Lands Council’s contentions to the contrary are not supported by reliable evidence. The Forest Service presented two independent monitoring tools to determine the percentage of old-growth acres in the IPNF, each of which found that the forest contained approximately twelve percent old-growth.' The first tool, the National Forest Inventory and Analysis (FÍA) program, “provides a congressionally mandated, statistically-based, continuous inventory of the forest resources of the United States.” The program’s design and methods are “scientifically designed, publicly disclosed, and repeatable.... There are also stringent quality control standards and procedures.” Using the FIA data, the Forest Services concluded that 11.8 percent of the IPNF is old-growth.
The second tool, the IPNF stand-level old-growth map, found a similar percentage using a method that “was designed and implemented independently from the FIA inventory.” This method utilizes stand information gathered by Forest Service personnel, which is inputted into the Timber Stand Management Record System (TSMRS) database. 12 Using this database, the Forest Service concluded that 12.1 percent of the IPNF is old-growth.
Lands Council’s argument that the Forest Service is not currently meeting Standard 10(b) is based on its own report. The report, Lost Forests, documented the results of a sampling, performed by Lands *1000 Council under the direction of a forest pathologist, of 3,000 acres that the Forest Service claimed to be old-growth. The report concluded that seventy percent of the surveyed area did not meet the Forest Service’s own standards for old-growth. The Forest Service’s expert, Dr. Arthur Zack, a forest ecologist, disagreed with the methodology and findings of the report. Dr. Zack found the report “contradictory and unclear about what criteria [it] used for making old growth determinations.” Dr. Zack called the report “not credible” because Lands Council used outdated versions of Forest Service databases and did not use “a representative, non-biased sample design.”
“When specialists express conflicting views, an agency must have discretion to rely on the reasonable opinions of its own qualified experts even if, as an original matter, a court might find contrary views more persuasive.”
Marsh,
The Forest Service has also established that it will not harvest any old-growth trees as a part of the Project. Despite its plans to perform treatments within old-growth stands, the treatment will not involve harvesting allocated old-growth. The Forest Service represented in the SFEIS that the IPNF has not harvested allocated old-growth for several years, and that its “focus is on maintaining [existing] old growth stands.”
In
Lands Council I,
we held that “[b]e-cause no old growth forest is to be harvested under the Project, ... it cannot be said that the Project itself violates the IPNF Plan’s requirement to maintain ten percent of the forest acreage as old growth forest.”
The district court did not abuse its discretion in concluding that Lands Council was not likely to succeed on the merits of this aspect of its NFMA claim.
B. National Environmental Policy Act
1. Statutory Language And Purpose
NEPA, unlike the NFMA, does not impose any substantive requirements on federal agencies — it “exists to ensure a process.”
Inland Empire Pub. Lands Council, 88
F.3d at 758. NEPA aims
to
make certain that “the agency ... will have available, and will carefully consider, detailed information concerning significant environmental impacts,” and “that the relevant information will be made available to the larger [public] audience.”
Robertson v. Methow Valley Citizens Council,
To that end, NEPA requires agencies to take a “hard look” at the environmental consequences of their actions by preparing an EIS for each “major Federal action[]
*1001
significantly affecting the quality of the human environment.” 42 U.S.C. § 4332(C);
Sierra Club v. Bosworth,
(i) the environmental impact of the proposed action, (ii) any adverse environmental effects which cannot be avoided should the proposal be implemented, (iii) alternatives to the proposed action, (iv) the relationship between local short-term uses of man’s environment and the maintenance and enhancement of long-term productivity, and (v) any irreversible and irretrievable commitments of resources which would be involved in the proposed action should it be implemented.
42 U.S.C. § 4332(C). We hold that when the Forest Service provides a full and fair discussion of environmental impacts and its EIS includes these necessary components, the Forest Service has taken the requisite “hard look.”
We have previously faulted the Forest Service for not addressing uncertainties relating to a project “in any meaningful way” in an EIS.
See Seattle Audubon Soc’y v. Espy,
We reaffirm, however, that the Forest Service must acknowledge and respond to comments by outside parties that raise significant scientific uncertainties and reasonably support that such uncertainties exist. This requirement comports with NEPA’s regulations, as well as with interpretations of NEPA offered by the Supreme Court and other circuits.
See, e.g.,
40 C.F.R. § 1500.1(b) (providing that the agency “must insure that environmental information is available to public officials and citizens” and this “information must be of high quality” as “[ajccurate scientific analysis, expert agency comments, and public scrutiny are essential to implementing NEPA”);
id.
§ 1502.9(a) (requiring that the agency “make every effort to disclose and discuss at appropriate points in the [EIS] all major points of view on the environmental impacts of the alternatives including the proposed action”);
id.
§ 1503.4(a) (stating that “[a]n agency preparing a final environmental impact statement shall assess and consider comments both individually and collectively, and shall respond ... in the final statement”);
id.
§ 1502.22 (providing that “[w]hen an agency is evaluating reasonably foreseeable significant adverse effects on the human environment in an environmental impact statement and there is incomplete or unavailable information, the agency shall always make clear that such information is
*1002
lacking”);
Izaak Walton League of Am. v. Marsh,
2. Lands Council’s Contentions That The Forest Service Violated NEPA By Failing To Address Scientific Uncertainty
Lands Council argues that the Project runs afoul of NEPA because the Forest Service “fail[ed] to include a full discussion of the scientific uncertainty surrounding its strategy for maintaining species viability.” Lands Council has not argued that the Forest Service’s SFEIS was deficient in any other respect. We conclude that the Forest Service has satisfied NEPA.
Lands Council first points to two papers it cited in its administrative appeal, one by Pfister et al. and one by Veblen, 13 to establish that the Project’s strategy of managing old-growth to maintain species viability is controversial. The Forest Service did not fail to conduct a “full and fair discussion” of environmental impacts as NEPA requires by not discussing these articles brought to its attention by Lands Council because they did not raise uncertainties about its methodology. Instead, on the whole, the papers Lands Council cites actually lend support to the Project’s proposed treatment of old-growth habitat.
Though Pfister states that “producing ‘old-growth habitats’ through active management is an untested hypothesis,” he approves of active management in certain circumstances that apply to the Project Area. See R.D. Pfister et al., Contract Review of Old-Growth Management on School Trust Lands: Supplemental Biodiversity Guidance 8/02/00, at 11. Pfister states that “initial restoration cutting treatments appear necessary to restore old-growth stands historically sustained by relatively frequent low-to mixed-intensity fire.” Id. at 15. As explained, the Project Area was historically sustained by such fires. Also relevant to the Project Area, Pfister recommends restoration cutting to “reduce unsustainable post-settlement density increases in [old-growth] pine/fir stands.” Id.
Veblen questioned whether suppression of naturally occurring forest fires resulted in unnatural fuel buildup, and also stated that this “premise and its implications need to be evaluated by ... area-specific research” in forests targeted for “fuels or ecological restoration projects.” T.T. Veblen, Key Issues in Fire Regime Research for Fuels Management and Ecological Fire Restoration (2003). The Forest Service has done precisely what Veblen advocates—“area-specific research.” The Forest Service conducted “field reconnaissance” in the Project area to obtain area-specific information on the history and intervals of fire. The Forest Service also modeled the fire risk for the Project Area.
Lands Council also contends, more generally, that the Forest Service did not cite adequate evidence that the Project will improve the habitat of old-growth species and did not adequately examine adverse impacts from logging within old-growth stands. We disagree.
*1003 In the SFEIS, the Forest Service discussed how the treatment proposed as a part of the Project would maintain dry-forest, old-growth stands and cited literature explaining that such treatment improves tree vigor and resistance to insects and disease. The Forest Service also modeled the treatment proposed in the different alternatives it considered, which demonstrated that the Project provided the greatest reduction in the risk of stand-replacing fires, thereby benefiting old-growth habitat. Also, Lands Council claims to have submitted a comment stating that the Forest Service did not cite “any evidence that its managing for old growth habitat strategy will improve old growth species habitat over the short-term or long-term.” The Forest Service responded, again with citations to literature, by stating that the Project’s proposed treatment would return the Project Area to old-growth and, in turn, improve tree vigor.
The Forest Service did not ignore that there may be some adverse impact from logging in old-growth stands. The Forest Service acknowledged possible short-term, negative impacts in the immediate vicinity of harvest units for the flam-mulated owl. But the Forest Service explained, based on its habitat suitability model, that its actions would not decrease suitable habitat in the short-term and would enhance it in the long-term. Finally, as we have already explained, we allow the Forest Service to use habitat as a proxy when the Forest Service concludes, in its expertise, that it is reasonable to assume that a project will maintain a species’ viability if the project will maintain suitable habitat for the species. Though the Forest Service must explain the methodology it used for its habitat suitability analysis, which the Forest Service did here, NEPA does not require us to “decide whether an [EIS] is based on the best scientific methodology available.”
See Friends of Endangered Species, Inc. v. Jantzen,
We conclude that the Forest Service took the requisite “hard look” at the environmental impacts of the Project to satisfy NEPA. Thus, the district court did not abuse its discretion in concluding that Lands Council is unlikely to succeed on the merits of its NEPA claim.
C. Injunctive Relief
“A preliminary injunction is appropriate when a plaintiff demonstrates ‘either: (1) a likelihood of success on the merits and the possibility of irreparable injury; or (2) that serious questions going to the merits were raised and the balance of hardships tips sharply in [the plaintiffs] favor.’ ”
Lands Council II,
As a threshold matter, a district court abuses its discretion if it “base[s] its decision on an erroneous legal standard or clearly erroneous findings of fact.”
Earth Island Inst. II,
Here, the district court applied the correct preliminary injunction standard and concluded that Lands Council established neither a fair chance of success on the merits nor the existence of serious questions going to the merits. The district court noted that Lands Council did not point to irreparable harm “beyond the general allegation that environmental harm is irreparable,” and refused to “presume that in all environmental cases that irreparable harm will outweigh all other considerations.” Importantly, Lands Council has not argued that the district court applied an incorrect legal standard or made clearly erroneous findings of fact in denying its request for a preliminary injunction.
We also agree with the district court that Lands Council has not shown that “the balance of hardships tips sharply in its favor.”
See Lands Council II,
Turning first to the significance of environmental injury, the Supreme Court has instructed us that “[e]nvironmental injury, by its nature, can seldom be adequately remedied by money damages and is often permanent or at least of long duration,
i.e.,
irreparable. If such injury is sufficiently likely, therefore, the balance of harms will usually favor the issuance of an injunction to protect the environment.”
Amoco Prod. Co. v. Vill. of Gambell,
Consistent with
Amoco Production Company,
we have held that the public interest in preserving nature and avoiding irreparable environmental injury outweighs economic concerns in cases where plaintiffs were likely to succeed on the merits of their underlying claim,
see Earth Island Inst. II,
Our law does not, however, allow us to abandon a balance of harms analysis just because a potential environmental injury is at issue.
See Forest Conservation Council v. U.S. Forest Serv.,
Intervenors raise hardships that must be balanced against Lands Council’s claims of potential environmental injury. Inter-venors contend that enjoining the project will force the timber companies that purchased the Sales to lay off some or all of their twenty-seven workers, in addition to other indirect harm to the struggling local economy. We must also consider the public’s interest.
See Kootenai Tribe of Idaho,
We conclude that the district court did not abuse its discretion in denying Lands Council’s request for a preliminary injunction.
CONCLUSION
For the reasons explained above, we conclude that Lands Council was not likely to succeed on any of its claims under the NFMA or NEPA. We also conclude that Lands Council has not shown that, if we allow the Forest Service to proceed with *1006 the Mission Brush Project, the balance of hardships tips sharply in its favor. The district court’s denial of Lands Council’s request for a preliminary injunction is AFFIRMED.
Notes
. The glossary to the SFEIS explains that the primary objective of commercial thinning is to "stimulate growth of the residual stand, increase total yield, and utilize material that is suppressed,” by removing "approximately 1/3 of the stand, leaving larger trees evenly spaced with crowns free to grow before canopy closure occurs again.”
. The Forest Service proposed both even-aged and uneven-aged regeneration cuts; the former is designed to regenerate a stand with a single age class, and the latter is designed to maintain an uneven-aged structure "by removing some trees in all size classes either singly, in small groups, or in strips.”
.The glossary to the SFEIS explains that sanitation salvage involves the "removal of trees to improve stand health by stopping or reducing the spread of insects and disease,” and the removal of dead, damaged, or dying trees.
. The scope of the arguments we address in this opinion comports with the representations that Lands Council made before this panel. Specifically, at oral argument, Lands Council represented that, for the purpose of obtaining a preliminary injunction, it is concerned only with the treatment of the 277 acres that qualify as old-growth habitat and the treatment’s effect on the flammulated owl and its habitat. Lands Council did not argue that the Forest Service’s analysis was deficient with respect to any other species.
We note, however, that Lands Council’s position as to why a preliminary injunction is necessary has been a constantly moving target. To illustrate, in its opening brief before the district court, Lands Council argued that the Forest Service had violated the NFMA by *988 failing to ensure habitat for old-growth species and viable populations of management indicator species. The only species Lands Council mentioned by name in its motion before the district court were the pileated woodpecker, black-backed woodpecker, grizzly bear, Canada lynx, and gray wolf; in its reply brief before the district court, Lands Council also mentioned the northern goshawk, the fisher, and the western toad.
In its opening brief on appeal, however, Lands Council changed the group of species about which it expressed concern, and argued that the Project will adversely affect the flam-mulated owl as well as the northern goshawk, the fisher, and the pileated woodpecker. Despite Lands Council’s changing list of species, the Forest Service has not argued that Lands Council has waived any arguments with respect to any particular species.
. The parties did not discuss what regulations implement 16 U.S.C. § 1604(g)(3)(B), but Lands Council suggested in a letter submitted under Rule 28(j) of the Federal Rules of Appellate Procedure that 36 C.F.R. § 219.19 is the applicable regulation. Section 219.19 required the Forest Service to manage wildlife habitat "to maintain viable populations of existing ... species” and required the Forest Service to designate management indicator species (MIS) to monitor and evaluate wildlife viability. 36 C.F.R. § 219.19 (2000). This regulation is no longer in effect,
see Envtl. Prot. Info. Ctr.,
We need not resolve Lands Council's eleventh hour suggestion that § 219.19 is applicable because the SFEIS and ROD incorporated that regulation. Lands Council has not claimed that the Forest Service violated any specific regulation regarding wildlife viability. In fact, Lands Council’s brief does not cite any regulations to support its argument that the Forest Service violated the NFMA or NEPA. Moreover, the species that Lands Council focuses on, the flammulated owl, is not an MIS under § 219.19. The parties do, however, agree that the Project must comply with the IPNF Forest Plan’s requirements regarding species viability. The Forest Plan’s requirement, which is discussed below, does not materially differ from § 219.19’s requirement that the Forest Service must manage wildlife habitat to maintain viable populations of existing species.
. Even if 36 C.F.R. § 219.19 is applicable to the Project because it was explicitly incorporated into the SFEIS and ROD, see supra n. 5, the transitional rule in effect until April 21, 2008 stated that for plans developed "using the provisions of the planning rule in effect prior to November 9, 2000,” which would include 36 C.F.R. § 219.19, "the Responsible Official may comply with any obligations relating to [MIS] by considering data and analysis relating to habitat unless the plan specifically requires population monitoring or population surveys for the species. Site-specific monitoring or surveying or a proposed project or activity area is not required, but may be conducted at the discretion of the Responsible official.” 36 C.F.R. § 219.14 (2007) (emphasis added). The Forest Service did not designate the flammulated owl as an MIS and, in any case, Lands Council does not argue that any language in the IPNF Forest Plan or the Project itself imposed a requirement for site-specific monitoring.
. Lands Council's briefs repeatedly suggest that the Forest Service must improve wildlife habitat but, as our discussion of the NFMA and the IPNF Forest Plan makes clear, neither the NFMA nor the IPNF Forest Plan require the Forest Service to establish that its plans will improve the habitat of a particular species.
. Lands Council observes that the Forest Service has conducted other surveys in the Project Area and has failed to locate any flammu-lated owls. While it is true that these studies did not record the presence of flammulated owls, nothing in the SFEIS suggests that these particular surveys were performed to determine if flammulated owls occupy treated old-growth habitat. Rather, the SFEIS is clear that these studies were performed when the Forest Service was examining suitable habitat for the flammulated owls. Contrary to Lands Council's suggestion, we cannot infer from these studies anything about the impact of the proposed Project on flammulated owls.
. The SFEIS defines "suitable habitat” as habitat "that currently has both fixed and variable stand attributes for a given species' habitat requirements.” This differs from "capable habitat,” which refers to a site's "inherent potential ... to produce essential habitat requirements of a species” though the site does not currently have all that a species requires.
. We have also allowed the Forest Service to use habitat as a proxy to measure a species’ population, and then to use that species' population as a proxy for
the
population of other species (proxy-on-proxy approach).
See Envtl. Prot. Info. Ctr.,
. Some of our sister circuits have been skeptical when the Forest Service has relied only on habitat analyses to satisfy its requirements under 36 C.F.R. § 219.19.
Compare Inland Empire Pub. Lands Council,
. In
Lands Council I,
. The Veblen paper is not a part of the administrative record, and only an excerpt from the Veblen paper is included in Lands Council’s brief. The full text is available at http://www.fs.fed.us/rm/pubs/rmrs_p029.html.
. Even though Lands Council expressed concern only about the 277 acres of old-growth habitat in the Project Area at oral argument, Lands Council seeks to enjoin the entire project without regard for the multiple-use objectives the Project seeks to promote.