Appeal from the United States District Court for the District of Oregon Michael R. Hogan, District Judge, Presiding D.C. No. 6:07-cv-06283-HO.
The opinion of the court was delivered by: M. Smith, Circuit Judge
Argued and Submitted March 3, 2010 -- Portland, Oregon
Before: Richard A. Paez, Richard C. Tallman, and Milan D. Smith, Jr., Circuit Judges.
Opinion by Judge Milan D. Smith, Jr.; Dissent by Judge Paez
The League of Wilderness Defenders-Blue Mountains Bio-diversity Project, Cascadia Wildlands Project, and the Sierra Club (collectively, the Conservation Groups, or Groups) brought suit against John Allen, Forest Supervisor of the Des-chutes National Forest, and the U.S. Forest Service (collectively, the Forest Service), alleging that the Five Buttes Project (Project) violates the National Forest Management Act (NFMA) and the National Environmental Policy Act (NEPA). The district court agreed, and granted summary judgment and an injunction in favor of the Conservation Groups. The Forest Service appeals.
We reverse, vacate the injunction, and remand with directions to the district court to grant summary judgment in favor of the Forest Service.
FACTS AND PRIOR PROCEEDINGS
1. The Northwest Forest Plan
The NFMA, 16 U.S.C. §§ 1600-1614, describes the statutory framework and requirements under which the Forest Service must manage National Forest System lands. The NFMA requires the Forest Service to develop a forest plan for each unit of the forest system, id. § 1604(a), and all subsequent agency actions must be consistent with the governing plan, id. § 1604(i). As required by the NFMA, the Forest Service developed the Northwest Forest Plan (NWFP) to protect and enhance old-growth forest ecosystems in the Pacific Northwest and Northern California that serve as habitats for numerous species. See Record of Decision for Amendments to Forest Service and Bureau of Land Management Planning Documents Within the Range of the Northern Spotted Owl, Summary, April 13, 1994, available at http://www.reo.gov/ library/reports/newroda.pdf (FS ROD). The endangered northern spotted owl (spotted owl) is an indicator species for the Deschutes National Forest, which lies within the NWFP area. As an indicator species, the spotted owl is a "bellwether... for the other species that have the same special habitat or population characteristics." Inland Empire Pub. Lands Council v. Schultz, 88 F.3d 754, 762 n.11 (9th Cir. 1996).
2. The Davis Late Successional Reserve
In order to balance environmental and economic needs, the NWFP designates certain forest areas for logging and reserves other areas, called late successional reserves (LSRs), for conservation. Specifically, the NWFP created the Davis LSR to "protect and enhance conditions of late-successional and old-growth forest ecosystems, which serve as habitat for late-successional and old-growth related species including the northern spotted owl." Except as otherwise permitted by law, commercial logging activities are prohibited in LSRs.
Wildfire and other disturbances occur frequently within the Davis LSR. Most notably, in 2003, a major fire in the Davis LSR (the Davis fire) burned approximately 21,000 acres of forest, including 3,736 acres of spotted owl nesting, roosting, and foraging (NRF) habitat, approximately 16,000 of which suffered near complete tree mortality. In all, the Davis fire burned twenty-four percent of the Davis LSR. In response to the Davis fire, the Forest Service revised its Davis LSR assessment to reflect the "immediate need" to "reduc[e] the risk of large-scale loss in a portion of the existing late and old-structure stands that are susceptible to insect attack and/or wildfire." The objective of the Forest Service's Project is to reduce that risk, in part, by thinning some of the trees in the Davis LSR. Objection to this logging component of the Project is the gravamen of the Conservation Groups' complaint.
3. The Five Buttes Project
The Forest Service is tasked with developing area-specific projects to fulfill the NWFP's goals. The projects generally describe planned management and treatment activities in the relevant areas of the National Forest System lands. Treatment activities, or silviculture, include commercial thinning, regeneration cuts, salvage harvesting, and other activities intended to improve forest health.
The Project was designed in part to address the need identified in the updated Davis LSR assessment to reduce risks to the LSR from fire and disease. The Project covers approximately 160,000 acres (including the 48,900-acre Davis LSR) and authorizes management treatments, including commercial logging, across approximately 5,522 acres. It authorizes commercial logging in 618 acres of NRF habitat in the Davis LSR.*fn1
The Project's prescribed treatments differ depending on whether they are to occur within spotted owl habitat or home ranges. Within spotted owl home ranges, treatments would be "less intense or not done at all," while NRF habitat outside the home range would be treated depending on vegetation type and crown fire potential. The Project is structured so that no spotted owls will be directly harmed. Five Buttes Project Environmental Impact Statement 391 (EIS) ("There is no commercial thinning of NRF habitat proposed within an occupied spotted owl home range."). The stated goal of the Project is to accelerate the development of large trees and NRF habitat to promote the objectives of the Davis LSR.
b. Environmental Impact Statement and Record of Decision
As required by NEPA, 42 U.S.C. § 4321 et seq., the Forest Service prepared a draft Environmental Impact Statement (EIS) for the Project that describes its expected environmental impacts. Id. § 4332(2)(C). The Forest Service circulated a draft EIS, received comments from the public, and eventually issued a final EIS and a record of decision (ROD). In creating the EIS, the Forest Service conducted computerized simulations to determine the effects of wildfires on the Project area under three different treatment scenarios (A, B, and C) and to evaluate what level of treatment, if any, was needed to protect and preserve the Davis LSR.
Alternative A, or "No Action," described the expected effects of fire and disease on the Project area if no additional thinning or fuels treatments are implemented. In other words, it modeled the effects of "allow[ing] current processes to continue, along with associated risks and benefits, in the [Project] area."
Alternative B involved the most intensive treatments. It called for management activities across 5,522 acres and commercial harvesting of around 18.9 million board feet of lumber. This alternative proposed the greatest amount of commercial thinning within spotted owl NRF habitat (2,822 acres).
The Forest Service ultimately adopted Alternative C, which includes some treatments within the Davis LSR. Alternative C was developed to address landscape-scale fire prevention and retention of spotted owl habitat. It would "strategically place fuels treatments on the landscape to coordinate with past treatments to create and maintain fuel modifications around identified habitats." This alternative calls for management of 7,798 acres and would harvest around 14.4 million board feet of lumber. It proposes commercial thinning on 2,023 acres of NRF habitat and would not involve treatment within any occupied spotted owl territory, but "habit modification would occur within unoccupied home ranges and could affect the ability of new owls to locate and establish a territory." Under this alternative, "thinning would generally only remove trees less than 21 inches in diameter and less than 5 percent of all trees removed would exceed 21 inches in diameter." Harvesting of trees larger than twenty-one inches "would only occur to meet basal area objectives or to lessen disease spread."
The EIS states that Alternative C would reduce average burn probability by 40 percent over Alternative A, and that it "protects owl home ranges the best." The Forest Service estimates that, if no preventative action is taken, the risk of large-scale loss of late-structure forest is extremely high, and the risk of a Davis-like problem fire is moderate to high. EIS 84 ("Due to current fuel loadings... much of the landscape is classified as moderate to high risk of experiencing a Problem Fire similar to the Davis Fire."); see also EIS 359 (describing the risk of large-scale loss of large trees and late structure forest as "extremely high").*fn2 On appeal, the Conservation Groups contest this characterization, stating that "when the Forest Service considered all relevant factors, including ignition sources and location, it concluded that another large fire in the Five Buttes Project area was highly unlikely." They cite the Ager study, see infra pages 11572-74, for evidence that "conditions for a large wildfire event are rare within the study area."*fn3
In response to public concerns regarding the long-term effects on spotted owl habitat, the Forest Service stated that "[b]ased on modeling, the return to NRF conditions will take 2-5 decades depending on the thinning intensity prescribed.... There is no commercial thinning of NRF habitat proposed within an occupied spotted owl home range; all other thinning within these areas would be small-diameter (3 inches or less) and stands would remain in NRF condition following activities."
In response to comments opposing the harvest of larger trees, the Forest Service stated that the modeling had "deter-mined that small-diameter thinning alone... in most places, would not change the vegetation and fuels structure enough to reduce fire risk," and would not solve the problem of over-competition among larger trees. Also, limited harvest of larger trees would only take place outside NRF habitat, so that these areas would remain in NRF condition.
The EIS also contains a table and description of past, present, and reasonably foreseeable future actions in an effort to analyze the cumulative impact of the Project. Additionally, cumulative impacts are discussed in detail throughout the EIS in relation to each affected area, type of vegetation, or animal. See, e.g., EIS 66-70 (discussing the existing conditions of vegetation to create baseline); EIS 90 (discussing the cumulative effects of past fires and past and future fuels treatments); EIS 117 (discussing cumulative impacts on spotted owls).
On June 8, 2007, the Forest Supervisor for the Deschutes National Forest signed the ROD, which adopted Alternative C with minor modifications and incorporated the EIS analysis. The ROD incorporates and responds to public concerns, and outlines the Forest Service's final decision on the Project. It also reports the results of the computerized simulations and explains the Forest Service's decision to adopt Alternative C.
4. Independent Review and Approval by Outside Agencies
The NWFP requires the Regional Ecosystem Office (REO)*fn4 to review Forest Service projects to ensure consistency with the NWFP. Here, the REO concluded that Alternative C was consistent with the NWFP, stating that the "proposed treatment in the LSR meets the objectives for managing LSRs," especially because "[t]reatments are placed on the landscape to alter fuel profiles in strategic locations such that it minimizes the likelihood of loss of valuable late- and old-structure habitat." The REO also found that the proposed treatments "will result in a balanced mix of multi-story stands that are more... conducive to late-successional species, with single story stands that are more resilient and less susceptible to loss," and that "[t]reatments will focus on retaining the late and old structure components that are both desirable for late-successional species and adapted to the local fire regimes."
Although not required to do so, out of an abundance of caution, the Forest Service also requested formal review of the Project by the United States Fish and Wildlife Service (FWS). The FWS conducted an extensive empirical analysis and published an eighteen-page opinion stating that "the [Project] is not likely to jeopardize the continued existence of the spotted owl" because the NWFP "provides a well distributed set of reserves which protect suitable habitat across the range of the spotted owl," and because "no suitable habitat will be removed within spotted owl home ranges or core areas." It concluded that spotted owl "home ranges will be maintained through fire risk reduction and stand density treatments" and that the short-term loss of some spotted owl habitat was justified by the "long-term benefits to owl habitat across the landscape by reducing fuel loads, strategic placement of treatments that reduce the risk of [fires], and reducing tree stocking densities to promote development of large trees in the future."
The Conservation Groups brought suit against the Forest Service, alleging that the Project calls for logging within the Davis LSR that does not comply with the NWFP. The Conservation Groups also claimed that the Forest Service's EIS violates NEPA because the EIS does not adequately consider cumulative effects and does not respond to opposing views regarding logging and the prevention of catastrophic fires.
Regarding the NFMA challenge, the district court recognized that "[r]isk reduction activities are permitted in LSRs under limited circumstances." However, in light of statements in the ROD recognizing possible negative effects on the LSR and spotted owls, the court found that "[t]he findings in the ROD are not strong enough to meet" the NWFP requirement that commercial thinning projects focused on older stands must "clearly result in greater assurance of long-term maintenance of habitat."
With respect to the NEPA claim, the district court found that the cumulative impact discussion in the EIS was deficient for lack of detailed, quantitative information about past projects. League of Wilderness Defenders-Blue Mountains Bio-diversity Project v. Weldon, No. 07-6238-HO, 2008 WL 4279807, at *4 (D. Or. Sept. 11, 2008) (District Court Order) ("[T]hese [past] projects have not been quantified with time, place, and scale data."). The court also found that the Forest Service's use of an "aggregate effects" approach was "in contradiction to current Ninth Circuit law." Id. at *5. The court declined to reach the Conservation Groups' second NEPA claim regarding disclosure of opposing scientific opinions. Id.
The district court granted summary judgment in favor of the Conservation Groups, enjoined the Forest Service from any additional logging, and remanded to the Forest Service for preparation of a new ROD that complies with the NFMA and NEPA.*fn5 The Forest Service appeals the district court's final judgment.
STANDARDS OF REVIEW AND JURISDICTION
Under the Administrative Procedure Act (APA), a reviewing court can only reverse an agency decision if that decision was "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2)(A). To address what had been a gradual divergence from this highly deferential standard by our court, we recently went en banc and unanimously held that:
Review under the arbitrary and capricious standard is narrow, and [we do] not substitute [our] judgment for that of the agency. Rather, we will reverse a decision as arbitrary and capricious only if the agency relied on factors Congress did not intend it to consider, entirely failed to consider an important aspect of the problem, or offered an explanation that runs counter to the evidence before the agency or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.
Lands Council v. McNair, 537 F.3d 981, 987 (9th Cir. 2008) (en banc) (internal quotation marks and citations omitted) (alterations in original). This deference is highest when reviewing an agency's technical analyses and judgments involving the evaluation of complex scientific data within the agency's technical expertise. Id. at 993.
For NEPA claims, "[w]e employ a rule of reason [standard] to determine whether the [EIS] contains a reasonably thorough discussion of the significant aspects of the probable environmental consequences." Ctr. for Biological Diversity v. U.S. Forest Serv., 349 F.3d 1157, 1166 (9th Cir. 2003) (internal quotation marks omitted) (second alteration in original). We must ensure "that the agency has taken a 'hard look' at the environmental consequences of the proposed action," id. (quoting Churchill County v. Norton, 276 F.3d 1060, 1072 (9th Cir. 2001)), and we must uphold the agency decision as long as the agency has "considered the relevant factors and articulated a rational connection between the facts found and the choice made," Selkirk Conservation Alliance v. Forsgren, 336 F.3d 944, 953-54 (9th Cir. 2003) (quoting Wash. Crab Producers, Inc. v. Mosbacher, 924 F.2d 1438, 1441 (9th Cir. 1990)).
We review the district court's grant of summary judgment de novo, Feldman v. Allstate Ins. Co., 322 F.3d 660, 665 (9th Cir. 2003), and its grant of injunctive relief for abuse of discretion, Rolex Watch, U.S.A., Inc. v. Michel Co., 179 F.3d 704, 708 (9th Cir. 1999). We have jurisdiction under 28 U.S.C. § 1291.
I. Compliance with the NFMA and the NWFP
 The NWFP requires LSRs to be managed to "protect and enhance conditions of late-successional and old-growth forest ecosystems." NWFP 8. The NWFP has "[g]uidelines to reduce risks of large-scale disturbance," which provide that "[s]ilvicultural activities aimed at reducing risk shall focus on younger stands in [LSRs]," and "the scale of... treatments should not generally result in degeneration of currently suitable owl habitat or other late-successional conditions." To avoid degeneration, "logging and other ground-disturbing activities are generally prohibited" in LSRs. Or. Natural Res. Council Fund v. Brong, 492 F.3d 1120, 1126 (9th Cir. 2007) (quoting Seattle Audubon Soc'y, 871 F. Supp. at 1304-05).
 However, the NWFP permits logging activities in LSRs if: "(1) the proposed management activities will clearly result in greater assurance of long-term maintenance of habitat, (2) the activities are clearly needed to reduce risks, and (3) the activities will not prevent the [LSRs] from playing an effective role in the objectives for which they were established." Standards and Guidelines for Management of Late-Successional and Old-Growth Forest Related Species within the Range of the Northern Spotted Owl C-13 (April 13, 1994) [hereinafter NWFP S. & G.]. The NWFP acknowledges that some logging "may reduce the quality of habitat for late-successional organisms" and that "managers need to seek a balanced approach that reduces the risk of fire while protecting large areas of fire-prone late-successional forest." NWFP S. & G. B-7.
Our highest deference is owed to the Forest Service's technical analyses and judgments within its area of expertise, Lands Council, 537 F.3d at 993; nonetheless, our dissenting colleague would have us halt the Forest Service's Project because he does not like the Forest Service's approach to solving the problems addressed. We went en banc to foreclose precisely this type of second-guessing of the Forest Service. See id. at 988 (noting that "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" "act as a panel of scientists that instructs the Forest Service" how to perform its expert duties). The Forest Service thoroughly considered various reasonable approaches to "protect and enhance conditions" of the LRSs, NWFP S. & G. at C-11, and offered a plan that does not "run[ ] counter to the evidence before the agency or is so implausible that it could be not ascribed to a difference in view or the product of agency expertise," Earth Island Inst. v. U.S. Forest Serv., 442 F.3d 1147, 1156 (9th Cir. 2006), abrogated on other grounds by Winter v. Natural Re. Def. Council, Inc, 129 S.Ct. 365 (2008). Far from conflicting with the protection of LSRs, carefully controlled logging is a tool expressly authorized by the NWFP for long-term LSR maintenance.
 Our dissenting colleague simply articulates a different point of view, and an extreme one at that. He begins by implying that the law provides that under no circumstances can any old growth timber be cut from an LSR. Dissent at 11586-87. Eventually, however, our colleague reluctantly admits that is simply not the case. Dissent at 11587-88 ("[T]he NWFP... recognize[s] that in some LRSs [including the Davis LSR], stand management that includes older trees may be considered."). Thus, it is clear that the limited logging of LSR trees of all types and ages is permissible if such logging complies with the three NWFP requirements previously described. There is no dispute that the Forest Service correctly identified that standard. The question, then, is not whether such logging within the LSR is generally permitted, but whether the Forest Service's determination that the NWFP requirements for cutting some trees in an LSR were met was arbitrary or capricious. We now turn to that analysis.
After the Davis fire, the Forest Service determined that logging was necessary to protect the Davis LSR from future fires and that the Project met the three NWFP criteria for logging within LSRs. It noted that in the previous five years, "approximately 16,654 acres of NRF habitat ha[d] been lost mostly due to wildfires." The Project, in comparison, calls for thinning only 618 acres of NRF habitat. This would leave 93 percent of existing NRF habitat and 96 percent of spotted owl critical habitat in the Davis LSR. The Project would reduce fire risks by 40 percent over ...