FOR PUBLICATION

UNITED STATES COURT OF APPEALS

FOR THE NINTH CIRCUIT

EARTH ISLAND INSTITUTE, a non- No. 22-16483 profit corporation, D.C. No. Plaintiff-Appellant, 1:22-cv-00710-

AWI-EPG

v.

CICELY MULDOON, in her official OPINION capacity as Superintendent of Yosemite National Park; U.S.

DEPARTMENT OF THE INTERIOR;

UNITED STATES NATIONAL

PARK SERVICE, an agency of the United States Department of the Interior,

Defendants-Appellees.

Appeal from the United States District Court for the Eastern District of California Anthony W. Ishii, District Judge, Presiding

Argued and Submitted May 9, 2023 San Francisco, California

Filed September 12, 2023 Before: Mary H. Murguia, Chief Judge, and Michelle T. Friedland and Mark J. Bennett, Circuit Judges.

Opinion by Judge Friedland

SUMMARY *

Environmental Law

The panel affirmed the district court’s denial of Earth Island Institute’s motion for a preliminary injunction against the Superintendent of Yosemite National Park, the National Park Service, and the Department of the Interior (collectively, “the Agency”) to halt parts of two projects to thin vegetation in Yosemite National Park in preparation for controlled burns. Earth Island Institute sued the Agency under the National Environmental Policy Act (“NEPA”), arguing that it was unlawful for the Agency to approve the projects without conducting a full review of their environmental impacts. Applying the arbitrary and capricious standard, the panel upheld the Agency’s determination that the projects fell under a categorical exclusion called the “minor-change exclusion” that exempted them from the requirement that the Agency prepare an environmental assessment or an environmental impact statement. The projects fell under that categorical exclusion because they were “changes or

* This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader.

EARTH ISLAND INSTITUTE V. MULDOON 3

amendments” to the 2004 Fire Management Plan that would cause “no or only minimal environmental impact.” The panel held that the projects were consistent with the Fire Management Plan, contributing to its goals and using its methods, with only minor modifications. The Agency adequately explained its conclusion that those modifications would have “no or only minimal” environmental impacts, including on threatened and endangered species. In holding that the Agency’s determination was not arbitrary and capricious, the panel emphasized that the relevant issue was the expected environmental impact of the aspects of the projects that deviated from the Fire Management Plan, not the environmental impact of the projects overall. The panel acknowledged that even if a proposed project fits within a categorical exclusion, an agency may not rely on that exclusion if there are “extraordinary circumstances in which a normally excluded action may have a significant effect” on the environment. 40 C.F.R. § 1501.4(b). The panel rejected Earth Island Institute’s argument that the projects are highly controversial and upheld the Agency’s determination that no extraordinary circumstances were present. Because Earth Island Institute failed to meet the threshold inquiry of showing a likelihood of success on the merits, the panel did not consider the other preliminary injunction factors. Thomas C. Buchele (argued) and Haley Nicholson, Earthrise Law Center, Lewis & Clark Law School, Portland, Oregon; Bridgett A. Chevallier, Willamette Law Group, Oregon City, Oregon; for Plaintiff-Appellant. David S. Frankel (argued), Sommer H. Engels, Rachel Heron, and Andrew C. Mergen, Attorneys; Jeffrey S. Thomas, Trial Attorney; Todd Kim, Assistant Attorney General; United States Department of Justice, Environment & Natural Resources Division, Washington, D.C.; David W. Gehlert, Trial Attorney, United States Department of Justice, Environment & Natural Resources Division, Denver, Colorado; for Defendants-Appellees.

OPINION

FRIEDLAND, Circuit Judge:

In 2004, the National Park Service adopted a comprehensive plan for fire management in Yosemite National Park. In 2021 and 2022, the National Park Service approved two projects to thin vegetation in Yosemite in preparation for controlled burns. Those projects comported with the fire management plan except for minor alterations. The Earth Island Institute sued under the National Environmental Policy Act (“NEPA”), arguing that it was unlawful for the National Park Service to approve the projects without conducting a full review of their expected environmental impacts. The Institute then moved for a preliminary injunction to halt parts of the projects. The district court denied the motion for a preliminary injunction,

EARTH ISLAND INSTITUTE V. MULDOON 5

holding that the National Park Service had sufficiently evaluated the environmental impact of the projects. We affirm. I. A. Ever since Congress established the first national park in 1872, there has been a “tension” between the goals of preserving nature, which includes fire, and making “nature available for the enjoyment of all Americans” by suppressing fire. Hal K. Rothman, Blazing Heritage: A History of Wildland Fire in the National Parks 6 (2007). The history of fire management in Yosemite epitomizes that tension. Established by Congress in 1890, id. at 15, Yosemite National Park sits on the western slope of the Sierra Nevada, the highest and longest mountain range in California. Historically, that area was subject to periodic fires typically of low-to-moderate intensity. During those periodic burns, small plants would die, while larger, fire-resistant trees would survive. The smaller plants would then grow back, thickening the understory—the layer of vegetation beneath the main canopy of the forest—until another fire occurred and the cycle started anew. Government efforts to suppress fire disrupted that fire cycle, starting around the late nineteenth century. Id. at 6– 7. When the National Park Service was founded in 1916, it began leading these efforts, adopting a policy of extinguishing any fires in national parks, including Yosemite, as soon as possible. See id. at 7. Fire suppression altered the composition of the forest as small plants that would have burned in periodic fires instead accumulated. Over time, like other national park forests, Yosemite’s forest became much denser and its canopy became increasingly closed, allowing less sunlight to get through to the understory. Id. at 8. In the 1960s, the National Park Service realized that fire suppression had unintended consequences. See id. at 7, 117. Fire suppression had led to a greater density of burnable vegetation that increased the intensity of the fires that did occur, making it more difficult for firefighters to control them. See id. at 20. Accordingly, the National Park Service reassessed its fire policy, allowing some naturally ignited wildfires to burn and even intentionally starting additional fires—a technique known as “prescribed” or “controlled” burning. See id. at 117–19. In 2004, after nearly five years of preparation that included public notice and comment, the National Park Service published an 800-page document proposing an updated approach to fire management in Yosemite and analyzing its expected environmental impacts: the Final Yosemite Fire Management Plan/Environmental Impact Statement (“2004 Environmental Impact Statement” or “Impact Statement”). The Impact Statement acknowledged that what the National Park Service had been doing up to that point was not enough. Despite allowing fire to “play a more natural role in park ecosystems” for decades, the National Park Service’s fire-management program had “not been able to meet park land management objectives of ecosystems restoration.” In fact, the “long-term buildup of fuels”—such as small trees and dense understory—that had occurred under the fire-suppression policy persisted in many areas. The National Park Service concluded that “[i]ncreased application of prescribed fire” was needed to adequately “reduce forest fuels.” But the accumulation of fuel over the

EARTH ISLAND INSTITUTE V. MULDOON 7

decades of fire suppression made prescribed burning difficult. The National Park Service recognized that it was unsafe to initiate burns in many areas that desperately needed them because the density of fuel could cause fires to burn too intensely to control. As a result, the 2004 Environmental Impact Statement considered other fuel-reduction techniques that could be used to treat overgrown areas, either on their own or in preparation for prescribed burns. Many of the fuel-reduction techniques involved cutting down and removing trees and other vegetation (whether mechanically or by hand), a process called “thinning.” Some of the fuel-reduction techniques described in the Impact Statement were permitted under the National Park Service’s then-existing fire- management program, but many were new, such as using mechanical tools to remove vegetation. The 2004 Environmental Impact Statement articulated two high-level goals: (1) “to reduce the threat of wildland fire to public safety” and (2) “to return the influence of natural fire to park ecosystems” so that the park ecosystems would be “restored to, and maintained in, as natural a condition as possible.” It then considered four different paths that the National Park Service could take to attempt to achieve those goals—four “Alternatives”—and analyzed the expected results and environmental impacts of each. Alternative A represented the status quo. Under Alternative A, “the fire management program would continue to use the [fuel-reduction] strategies of the existing 1990 fire management plan.” The National Park Service concluded that maintaining the status quo would not meet the ecosystem restoration goals described in the Impact Statement. The other three alternatives (Alternatives B through D) leveraged the new fuel-reduction techniques described in the Impact Statement but differed in the aggressiveness of their application and thus in their timeline for achieving restoration goals. The Impact Statement’s recommended course of action—Alternative D—was the intermediate option and was expected to achieve restoration goals within fifteen to twenty years. Alternative D was deemed “environmentally preferred” because it would cause “the least damage to the environment” and “best protect[], preserve[], and enhance[] historic, cultural, and natural resources.” Unsurprisingly, in a subsequent Record of Decision, the National Park Service adopted Alternative D. The 2004 Environmental Impact Statement and Record of Decision (together, the “Fire Management Plan” or “Plan”), provided a blueprint for Alternative D’s implementation, describing what actions were to be taken and where. 1 For instance, the Plan contemplated that trees less than 20 inches in diameter would be thinned within 200 feet from the centerline of certain identified roads to serve as boundaries for prescribed burns. In sequoia groves, by

1 In the Record of Decision, the National Park Service reviewed the four alternatives presented in the 2004 Environmental Impact Statement and selected Alternative D, noting that the action it was selecting was “unchanged from what is presented” in the portion of the Impact Statement that described that alternative. The National Park Service instructed the reader to “[r]efer to the [2004] Final Yosemite Fire Management Plan/Environmental Impact Statement for complete details on the selected action.” In the remainder of this opinion, “Fire Management Plan” refers to the portions of the 2004 Environmental Impact Statement that were approved in the Record of Decision—i.e., the portions of the Impact Statement describing and analyzing Alternative D.

EARTH ISLAND INSTITUTE V. MULDOON 9

contrast, the Plan permitted thinning of non-sequoia trees only less than 12 inches in diameter. 2 B. Although the Fire Management Plan predicted that Alternative D would be complete within roughly twenty years, progress has been slow. In that time, fewer acres have been treated than would have been had the National Park Service implemented Alternative A—the status-quo alternative. The National Park Service attributes much of the delay to a drought that plagued California from 2012 to 2016, killing millions of trees in the Sierra Nevada. The high density of the resulting dead biomass, combined with the high density of live trees and understory, make it extraordinarily difficult for firefighters to control and contain fires, increasing the preparation that must be done before a prescribed burn can occur. Still, progress continues, one project at a time. Two such projects are at issue here: the Wawona Project and the Yosemite Valley Project (collectively, the “Projects”). 3 The National Park Service approved the Wawona Project in August 2021. The Project will reduce fuel along certain roads and trails in preparation for a prescribed burn

2 The National Park Service amended the 2004 Fire Management Plan in 2017 to “allow more opportunities for fire to play its critical role in the ecosystem.” The amended Plan increased the areas of Yosemite in which prescribed burns were permitted. 3 The full names of the Projects are (1) “Biomass removal and thinning to protect sequoias, wildlife habitat and communities—Wawona Road to Merced Grove” (“Wawona Project”), and (2) “Biomass removal and thinning—Yosemite Valley, Wawona, and Yosemite West” (“Yosemite Valley Project”). scheduled for the fall of 2023. The goal of the thinning and subsequent prescribed burning is to protect the Merced and Tuolumne groves of giant sequoias and the towns near them. The Project contemplates thinning non-sequoia trees with diameters less than 20 inches in designated areas, including in the Merced Sequoia Grove. Dead and downed trees will then be removed from the area, either by being hauled offsite or by being piled and burned. The National Park Service approved the Yosemite Valley Project in April 2022. Like the Wawona Project, the Yosemite Valley Project aims to protect the identified areas by thinning trees less than 20 inches in diameter in preparation for a prescribed burn. 4 The actions contemplated by the Projects deviate somewhat from the measures described in the Fire Management Plan. The Wawona Project authorizes removal of certain non-sequoia trees less than 20 inches in diameter in the Merced Sequoia Grove, whereas the Fire Management Plan had contemplated removal of trees in sequoia groves only with diameters less than 12 inches. 5 The National Park Service explained that this deviation was necessary because tree density had increased in the years since the Fire Management Plan’s adoption, making the removal only of trees less than 12 inches in diameter “insufficient to protect the sequoias.” The Projects also expand the locations in which fuel-reduction work will take place. Both Projects add new road segments along which thinning will occur that were not included in the Fire Management Plan. And the

4 The Yosemite Valley Project does not involve any thinning in sequoia groves. 5 The Fire Management Plan generally permitted the removal of trees up to 20 inches in diameter, but not in sequoia groves.

EARTH ISLAND INSTITUTE V. MULDOON 11

Wawona Project contemplates thinning trees within 209 feet from the centerline of certain roads, whereas the Fire Management Plan contemplated the removal of trees up to only 200 feet from a road’s centerline. Although NEPA generally requires federal agencies to prepare a lengthy environmental impact analysis of new projects before their adoption, the National Park Service concluded that both Projects were categorically excluded from that requirement under regulations implementing NEPA because the Projects were “changes or amendments to an approved plan” (i.e., the Fire Management Plan) that “would cause no or only minimal environmental impact.” C. In May 2022, a member of the California-based environmental organization Earth Island Institute observed trucks carrying timber through Yosemite. After some investigation, the Institute learned of the Wawona Project and initiated this lawsuit in the United States District Court for the Eastern District of California. The suit challenged the Wawona Project’s approval as procedurally deficient under NEPA. The lawsuit named as Defendants Cicely Muldoon (the Superintendent of Yosemite National Park) in her official capacity, the National Park Service, and the Department of the Interior (collectively, “the Agency”). 6 The Institute filed a motion for a preliminary injunction two days later, requesting that the court enjoin the Agency from “allowing or implementing any of the logging, thinning, or biomass removal authorized by” the Wawona

6 The National Park Service is a division of the Department of the Interior and manages Yosemite as part of the National Park System. See 54 U.S.C. § 100101(a). Project, with limited exceptions. The Agency then informed the Institute about the Yosemite Valley Project and produced records documenting the Agency’s approval of both Projects. In response, the Institute amended its Complaint and preliminary injunction motion to include the Yosemite Valley Project. The amended motion requested that all “logging, thinning, or biomass removal” be enjoined except for within certain areas and except for the thinning of trees under 12 inches in diameter in the Merced Sequoia Grove. The parties agreed to an expedited briefing schedule that would resolve the preliminary injunction motion before the Agency was required to produce the complete administrative record. After receiving that briefing and a partial administrative record, the district court found that the Institute had shown that it was likely to suffer irreparable harm absent an injunction. The district court nonetheless denied relief because it concluded both that the Institute had not shown that it was likely to succeed on the merits (at least on the “limited record” presently before it) and that the “balance of equities/public interests tip[ped] firmly in favor of denying the injunction.” The Institute appealed a few days later. Because the Agency informed the Institute that it planned to work on the Projects during the appeal, the Institute moved for an injunction pending appeal in both the district court and our court. Both motions were denied. The Institute also moved to supplement the record on appeal to include additional documents from the administrative record that had been produced to the Institute after this appeal was filed. We denied that motion too.

EARTH ISLAND INSTITUTE V. MULDOON 13

II.

“To obtain a preliminary injunction, a plaintiff ‘must establish that he is likely to succeed on the merits, that he is likely to suffer irreparable harm in the absence of preliminary relief, that the balance of equities tips in his favor, and that an injunction is in the public interest.’” Kennedy v. Warren, 66 F.4th 1199, 1206 (9th Cir. 2023) (quoting Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008)). “A plaintiff alternatively can meet his burden under the first element by raising ‘serious questions going to the merits’ if ‘the balance of hardships tips sharply in [his] favor.’” Id. (alteration in original) (quoting All. for the Wild Rockies v. Cottrell, 632 F.3d 1127, 1131 (9th Cir. 2011)). “Likelihood of success on the merits is a threshold inquiry and is the most important factor.” Env’t Prot. Info. Ctr. v. Carlson, 968 F.3d 985, 989 (9th Cir. 2020). “We review a district court’s denial of a preliminary injunction for abuse of discretion.” United States v. California, 921 F.3d 865, 877 (9th Cir. 2019). “The district court’s interpretation of the underlying legal principles, however, is subject to de novo review and a district court abuses its discretion when it makes an error of law.” Id. at 877–78 (quoting Sw. Voter Registration Educ. Project v. Shelley, 344 F.3d 914, 918 (9th Cir. 2003) (en banc)).

III.

The Institute argues that the Agency failed to comply with NEPA in its approval of the Projects. We disagree. A. In NEPA, Congress took a procedural approach to environmental protection. Rather than mandating that agencies take specific actions or achieve certain outcomes, NEPA “requires federal agencies to take a ‘hard look’ at the environmental consequences of their actions” and to explain their decisions to the public. Env’t Def. Ctr. v. Bureau of Ocean Energy Mgmt., 36 F.4th 850, 872 (9th Cir. 2022) (quoting Kern v. U.S. Bureau of Land Mgmt., 284 F.3d 1062, 1066 (9th Cir. 2002)), cert. denied sub nom. Am. Petroleum Inst. v. Env’t Def. Ctr., 143 S. Ct. 2582 (2023). Under NEPA, before an agency may take an action that will “significantly affect[] the quality of the human environment,” it must prepare an environmental impact statement: a “detailed statement” describing (among other things) the “reasonably foreseeable environmental effects of the proposed agency action,” “any reasonably foreseeable adverse environmental effects which cannot be avoided should the proposal be implemented,” and “a reasonable range of alternatives to the proposed agency action.” 42 U.S.C. § 4332(C). The agency must notice its draft environmental impact statement for public comment and respond to comments before it can finalize the statement and move forward with the underlying action. See Mountain Cmtys. for Fire Safety v. Elliott, 25 F.4th 667, 674 (9th Cir. 2022); 40 C.F.R. § 1502.9(b), (c). To determine whether an action will significantly affect the quality of the human environment and thus require an environmental impact statement, an agency may first prepare a “less demanding” document—called an “environmental assessment.” Mountain Cmtys., 25 F.4th at 674–75. If the environmental assessment concludes that the action will not significantly impact the environment, the agency may forego preparing an environmental impact statement. Id. at 675. At issue here is an even more expedited track available for a limited set of agency actions, under which preparation

EARTH ISLAND INSTITUTE V. MULDOON 15

of an environmental assessment or environmental impact statement is not required. For reasons of “efficiency,” NEPA’s implementing regulations encourage agencies to identify “categories of actions that normally do not have a significant effect on the human environment.” 40 C.F.R. § 1501.4(a). Once such a “categorical exclusion” is defined, an agency need not prepare an environmental assessment or environmental impact statement for an action that it concludes fits within the exclusion, so long as no “extraordinary circumstances” indicate that the action will nonetheless have a significant effect. Id. § 1501.4(b). As we have previously explained, “[a]pplication of a categorical exclusion is not an exemption from NEPA; rather, it is a form of NEPA compliance, albeit one that requires less than where an environmental impact statement or an environmental assessment is necessary.” Ctr. for Biological Diversity v. Salazar, 706 F.3d 1085, 1096 (9th Cir. 2013). The Department of the Interior has announced a range of categorical exclusions. See 49 Fed. Reg. 39,233, 39,235–37 (Oct. 4, 1984). At issue here is Categorical Exclusion B-1, which we will refer to as the “minor-change exclusion.” That exclusion covers “[c]hanges or amendments to an approved plan, when such changes would cause no or only minimal environmental impact.” Id. at 39,235. 7 B. The Agency did not prepare an environmental impact statement or environmental assessment for either of the Projects. Instead, it concluded that both Projects were exempt from the environmental-impact-statement

7 The Institute does not challenge the validity of the minor-change exclusion, but rather only the exclusion’s application here. requirement under the minor-change exclusion because they are “changes or amendments” to the Fire Management Plan that will cause “no or only minimal environmental impact.” Applying the deferential arbitrary and capricious standard, we uphold the Agency’s conclusion. 1. “An agency satisfies NEPA if it applies its categorical exclusions and determines that neither an [environmental assessment] nor an [environmental impact statement] is required, so long as the application of the exclusions to the facts of the particular action is not arbitrary and capricious.” California v. Norton, 311 F.3d 1162, 1176 (9th Cir. 2002) (quoting Bicycle Trails Council of Marin v. Babbitt, 82 F.3d 1445, 1456 n.5 (9th Cir. 1996)). Agency action is arbitrary and capricious

only if the agency relied on factors Congress did not intend it to consider, entirely failed to consider an important aspect of the problem, or offered an explanation that runs counter to the evidence before the agency or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.

Env’t Def. Ctr., 36 F.4th at 871 (quoting Defs. of Wildlife v. Zinke, 856 F.3d 1248, 1257 (9th Cir. 2017)). The challenged Projects are easily characterized as “changes” or “amendments” to the Fire Management Plan. Both Projects fall within the ambit of the Fire Management Plan, contributing to its goals and using its methods, with minor modifications. Whereas the Plan expressly permitted

EARTH ISLAND INSTITUTE V. MULDOON 17

thinning of trees less than 12 inches in diameter in sequoia groves, the Wawona Project permits thinning trees less than 20 inches in diameter in one sequoia grove. The Wawona Project also contemplates thinning trees within 209 feet of the centerline of certain roads, whereas the Fire Management Plan contemplated the removal of trees up to only 200 feet from a road’s centerline. And both Projects extend the Fire Management Plan to some new road segments. Beyond those modifications, the Projects exactly follow the Fire Management Plan. The Agency also adequately explained its conclusion that those modifications would have “no or only minimal” environmental impact. For each Project, the Agency prepared a collection of documents called an “Exclusion Package” that described the Project’s parameters and explained the Agency’s conclusion that the Project fell within the scope of the minor-change exclusion. The Packages show that the Agency expressly considered the impact that the Projects will have on twenty-nine resources, ranging from air quality to wildlife habitats and vegetation. For example, regarding the impact on “wildlife and/or wildlife habitat,” the Agency noted that “[t]hinning vegetation, pile burning, and associated noise and disturbance may have impacts [on] wildlife communities and habitat” and that wildlife behavior may be impacted by food scraps left by humans working on the sites. But the Agency nonetheless concluded that, overall, the “[i]mpacts from this action are expected to be beneficial to forest habitat health and [are] intended to thwart the potential negative, extensive impacts from large, catastrophic fire, which could result from not taking action.” Similarly, the Agency recognized that the Projects may impact vegetation but determined that “[i]mpacts from this action are expected to be beneficial to forest health” by reducing the risk of catastrophic fire. 8 The Exclusion Packages also show that the Agency expressly considered the impact of the Projects on threatened and endangered species. Beginning with the Wawona Project, the Agency concluded that the Project “is not likely to adversely affect threatened, endangered, or rare species and/or their critical habitat,” and described mitigation measures that it would undertake to avoid adversely affecting two such species (the Pacific Fisher and the Great Grey Owl). Before arriving at that conclusion, the Agency consulted with the Fish and Wildlife Service regarding the impact of the Project on the Pacific Fisher. The Fish and Wildlife Service concluded that the Wawona Project “may affect, but is not likely to adversely affect,” that species, because the Project would remove only conifers less than 20 inches in diameter, thus “retain[ing] the most important habitat features for fisher[s].” When performing that consultation, the Fish and Wildlife Service emphasized that “fisher habitat will remain suitable and protected from future catastrophic wildfires that destroy habitat.” The Agency reached a similar conclusion with respect to the Yosemite Valley Project, once again formally consulting with the Fish and Wildlife Service before concluding that the Project was “not likely to adversely affect the Fisher.” We emphasize that the relevant issue here is the expected environmental impact of the aspects of the Projects that

8 The Institute does not suggest that any significant beneficial environmental impacts cause the Projects to fall outside the scope of the minor-change exclusion. We therefore consider only the Institute’s contention that the Agency insufficiently analyzed the potential for the Projects to have significant adverse environmental consequences.

EARTH ISLAND INSTITUTE V. MULDOON 19

deviate from the Plan, not the expected environmental impact of the Projects themselves. This is evident from the text of the minor-change exclusion, which encompasses “[c]hanges or amendments to an approved plan, when such changes would cause no or only minimal environmental impact.” 49 Fed. Reg. at 39,235 (emphasis added). The Institute appropriately criticizes the Packages for analyzing the Projects as a whole, rather than expressly describing how the parts of the Projects that deviate from the Fire Management Plan may impact the environment. But the Agency’s conclusion that the Projects as a whole will not significantly impact the environment entails the conclusion that the aspects of the Projects that are challenged here—the changes from the Plan—will not significantly impact the environment either. 9 2. The Institute argues that the Agency’s invocation of the minor-change exclusion was arbitrary and capricious for three reasons. Its arguments are unpersuasive.

9 In its opposition to the Institute’s motion for a preliminary injunction, the Agency relied on declarations from scientists and administrators familiar with the Projects. The Institute moved to strike all references to the declarations within the merits portion of the Agency’s briefing, arguing that the Agency’s reliance on those declarations was impermissible “post-hoc rationalization[] and analysis.” The district court denied the Institute’s motion to strike, although the court then relied on only limited parts of the challenged declarations in its merits analysis. Because we hold that the Agency’s explanations in the Exclusion Packages are sufficient to justify the Agency’s invocation of the minor-change exclusion, we need not decide whether it was appropriate for the district court to consider any additional information in the challenged declarations. i. First, the Institute argues that the minor-change exclusion should not be read to encompass the Projects because there exists another, potentially more applicable exclusion. We need not consider whether another exclusion might apply, because “in selecting a [categorical exclusion] for a project,” an agency “only needs to cite and rely on one” exclusion, “even if other[s] . . . may apply.” Mountain Cmtys., 25 F.4th at 679–80; see also id. at 680 (“[T]he fact that a project fits into one [categorical exclusion] does not mean that it could not also have fit into another one.” (quotation marks omitted)). ii. The Institute next argues that the Projects cannot be considered “changes or amendments” to the Fire Management Plan because they “implement, rather than amend” the Plan. In making that argument, the Institute draws from cases in which we have distinguished between agencies’ “programmatic” statements and “site-specific” analyses. In Pit River Tribe v. United States Forest Service, 469 F.3d 768 (9th Cir. 2006), for example, the Department of the Interior attempted to justify its approval of power plant leases by relying on a “programmatic environmental impact statement” that it had previously prepared. Id. at 773, 781. The programmatic statement had “not address[ed] the environmental implications” of approving power plant development “in particular locations,” leaving such site- specific analyses to be prepared “for each lease area prior to any leasing action.” Id. at 773. We held that the Department could not rely upon that programmatic statement as its environmental analysis for specific power plant leases that the plaintiffs there challenged, because the approval of each

EARTH ISLAND INSTITUTE V. MULDOON 21

individual lease required consideration of site-specific factors that had not been analyzed in the programmatic statement. Id. at 783–84. Pit River does not help the Institute because the relationship between the Projects and the Plan at issue here is significantly different from the relationship between the power plant leases and programmatic statement there. In Pit River, the challenged leases allowed power plant development in specific geographic locations, the environmental impact of which had not been analyzed in the programmatic statement. But here, the Plan considered the environmental impacts of performing the relevant fuel- reduction techniques in specifically identified parts of Yosemite. Take, for example, the Plan’s discussion of one fuel- reduction technique: “machine crushing/shredding.” Under the Plan, this technique is permitted only in specified parts of Yosemite. After outlining where machine crushing/shredding may be performed, the Plan analyzed the expected environmental impacts of performing the technique in those areas. That analysis was not a theoretical discussion of how machine crushing/shredding could impact the environment in the abstract but rather a site-specific discussion of how the technique was likely to impact the environment if implemented in the areas contemplated by the Plan, including discussion of the predicted impacts on certain endangered species and on local communities. The Plan took a similar approach for the other fuel-reduction techniques it permitted. Unlike the plan discussed in Pit River, then, the Fire Management Plan is not purely programmatic; instead, it contains both programmatic and site-specific elements. And the Projects will utilize the Plan’s fuel-reduction techniques (though on somewhat larger trees) in the same parts of Yosemite (with minor additions) that were analyzed in the Plan. 10 The new aspects of the Projects are thus appropriately characterized as “changes” or “amendments” to the Plan. iii. Finally, the Institute attacks the Agency’s explanation for its invocation of the minor-change exclusion, arguing that the Agency “failed to demonstrate” that the Projects’ deviations from the Fire Management Plan would result in no or only minimal environmental impacts. According to the Institute, the Agency did not “sufficiently analyze site- specific impacts of the Projects” because the Exclusion Packages “offer only the equivalent of checked boxes,” which “do[] not satisfy NEPA” under Klamath-Siskiyou Wildlands Center v. Bureau of Land Management, 387 F.3d 989 (9th Cir. 2004). To begin, Klamath-Siskiyou was reviewing an environmental assessment, not the invocation of a categorical exclusion. Id. at 993–94. “Where agency action falls under a categorical exclusion, it need not comply with the requirements of an environmental impact statement” or environmental assessment. See Salazar, 706 F.3d at 1097

10 The Institute argues that the Projects as a whole require further environmental review because they should be considered “implementations” under our caselaw, not that the aspects of the Projects that deviate from the Plan trigger treatment as implementations. As we explain in the next Section, the Agency has reasonably concluded that those deviations are sufficiently minor to fall within the scope of the minor-change exclusion. And, again, the Institute does not challenge the validity of the minor-change exclusion. See supra note 7. We therefore need not consider whether some larger expansion of a previously adopted plan would trigger a need for further environmental review under Pit River.

EARTH ISLAND INSTITUTE V. MULDOON 23

(cleaned up). Klamath-Siskiyou is therefore of limited relevance to this case. In any event, the Institute’s argument fails because the Exclusion Packages are considerably more detailed than the “checked boxes” we criticized in Klamath- Siskiyou. There, the relevant portion of the agency’s analysis contained only “a list of environmental concerns such as air quality, water quality, and endangered species, with a ‘Yes’ and ‘No’ checkbox to indicate whether the respective condition . . . will be ‘affected.’” Klamath- Siskiyou, 387 F.3d at 995. For some of the resources for which the “No” box was checked in Klamath-Siskiyou, there was an accompanying asterisk noting that the “affected critical elements would be impacted by implementing the proposed action,” with no further explanation. Id. The record here is different. Although the Agency’s consideration of the Projects’ resource impacts was formatted in a table (like the analysis in Klamath-Siskiyou), the table entries here are more substantive, giving a sentence or two of explanation for the majority of the entries. It is true that the Exclusion Packages do not discuss the Projects’ expected environmental impacts with the same level of detail as the Fire Management Plan. But that is not surprising; the entire point of categorical exclusions is to reduce the administrative burden of approving certain projects. See Norton, 311 F.3d at 1176 (“In many instances, a brief statement that a categorical exclusion is being invoked will suffice.”); Salazar, 706 F.3d at 1097 (holding that imposing procedures required for environmental impact statements on the invocation of a categorical exclusion would be “inconsistent with the efficiencies that the abbreviated categorical exclusion process provides”). Although the analysis in the Exclusion Packages is brief, the Packages demonstrate that the Agency considered the relevant factors when reaching its conclusion that the Projects are likely to have no or only minimal environmental impact, and nothing in the record suggests that this conclusion was unreasonable. See Alaska Ctr. for Env’t v. U.S. Forest Serv., 189 F.3d 851, 859 (9th Cir. 1999) (“Once the agency considers the proper factors and makes a factual determination on whether the impacts are significant or not, that decision implicates substantial agency expertise and is entitled to deference.”). *** It is always possible to quibble with an agency’s explanation; a motivated litigant will be able to identify parts of any agency explanation that could have been more precise or thorough. But the arbitrary and capricious standard does not demand perfection. As the Supreme Court has instructed, we “should ‘uphold a decision of less than ideal clarity if the agency’s path may reasonably be discerned.’” FCC v. Fox Television Stations, Inc., 556 U.S. 502, 513–14 (2009) (quoting Bowman Transp., Inc. v. Arkansas-Best Freight Sys. Inc., 419 U.S. 281, 286 (1974)). In light of the deference we owe the Agency and the purpose behind “the abbreviated categorical exclusion process,” Salazar, 706 F.3d at 1097, we hold that the Institute is not likely to prevail on its argument that the Agency improperly concluded that the Projects fall within the scope of the minor-change exclusion. 11

11 In addition to challenging the substance of the Exclusion Packages, the Institute challenges their form, arguing that the Agency was not permitted to “tier” its analysis to the Fire Management Plan and that, even if tiering was permitted, the Packages did not comport with regulatory requirements for tiering. “Tiering” is a term of art in the

EARTH ISLAND INSTITUTE V. MULDOON 25

C. Even if a proposed project fits within a categorical exclusion, the agency may not rely on that exclusion if there exist “extraordinary circumstances in which a normally excluded action may have a significant effect” on the environment. 40 C.F.R. § 1501.4(b). “In such extraordinary circumstances, a categorically excluded action would nevertheless trigger preparation of an [environmental impact statement] or an [environmental assessment].” Norton, 311 F.3d at 1168. NEPA’s implementing regulations list a range of situations in which extraordinary circumstances would exist. See 43 C.F.R. § 46.215. The Institute invokes one of these situations, arguing that extraordinary circumstances are present here because the Projects may have “highly controversial environmental effects.” Id. § 46.215(c). We conclude that the Agency’s determination that the potential effects of the Projects are not highly controversial was not arbitrary or capricious. 12

NEPA context, referring to the practice of incorporating by reference previously conducted environmental analyses into a subsequent environmental analysis. See 40 C.F.R. § 1501.11; Kern v. U.S. Bureau of Land Mgmt., 284 F.3d 1062, 1073 (9th Cir. 2002). We have held that agencies are permitted to refer to prior environmental analyses when invoking a categorical exclusion. See Salazar, 706 F.3d at 1098. Moreover, the regulation governing tiering that the Institute accuses the Agency of violating is applicable only to environmental impact statements and environmental assessments—not categorical exclusions. 40 C.F.R. § 1501.11; see also Salazar, 706 F.3d at 1098. 12 Aside from controversial effects, the Institute also briefly suggests that extraordinary circumstances exist due to the Projects’ potential significant impact on endangered or threatened species. See 43 C.F.R. § 46.215(h). We reject that argument because, as we explained above, “A project is highly controversial if there is a substantial dispute about [its] size, nature, or effect.” Safari Club Int’l v. Haaland, 31 F.4th 1157, 1179 (9th Cir. 2022) (quoting Bark v. U.S. Forest Serv., 958 F.3d 865, 870 (9th Cir. 2020)). “Mere opposition to an action does not, by itself, create a controversy within the meaning of NEPA regulations.” Id. (quoting Am. Wild Horse Campaign v. Bernhardt, 963 F.3d 1001, 1011 (9th Cir. 2020)). But “[w]here there is substantial evidence in the record that” a project may have highly controversial environmental effects, “the agency must at the very least explain why” it will not. Norton, 311 F.3d at 1177. The Institute argues that the Projects are “highly controversial” because some scientists dispute the Agency’s position that tree-thinning aids fire prevention and management—some going so far as to say that thinning can increase the risk of severe fires. In support of its position, the Institute offers declarations from Dr. Chad Hanson, a research ecologist at the Institute. Dr. Hanson asserts that “[p]ost-fire logging and clearcutting, which Defendants are in fact doing in the Park through the challenged logging projects . . . is perhaps the most highly controversial of all forest management activities.” He explains:

Unlike prescribed fire, managed wildfire, and thinning of genuinely small trees and underbrush, commercial logging operations like commercial thinning and post-fire logging fundamentally change the

the Agency reasonably determined that the Projects would have only minor impacts on the environment, including on endangered or threatened species. See supra Section III.B.1. The Institute offers no argument challenging this determination as to any species.

EARTH ISLAND INSTITUTE V. MULDOON 27

microclimate of a forest and often tend to increase overall fire severity and tree mortality. In fact, research by U.S. government scientists that promotes commercial thinning . . . acknowledges this, such as Prichard et al. (2021).

As an initial matter, Dr. Hanson overstates the relevant controversy by mischaracterizing the Projects. According to Dr. Hanson, there is significant scientific controversy surrounding “commercial thinning” and “post-fire logging.” But contrary to Dr. Hanson’s assertion, the Agency is not “in fact doing” these things. First, the Agency represents that “no entity profits from the sale of any timber cut during project work,” explaining that the “minimal monies received from collecting small trees and biomass are dedicated entirely to offsetting project costs and cover a small fraction of project expenses.” Other than Dr. Hanson’s conclusory accusations, there is no evidence to the contrary. Second, the thinning contemplated by the Projects is in preparation for a prescribed burn and accordingly is not appropriately characterized as post-fire logging. Similarly, the literature on which Dr. Hanson relies as evidence of scientific disagreement does not criticize thinning of the sort contemplated by the Projects. For example, Dr. Hanson cites a paper by Susan J. Prichard et al. as evidence of the purportedly relevant controversy. But that paper discusses the deficiencies of thinning conducted without subsequent prescribed burns—not thinning conducted in preparation for prescribed burns. See Susan J. Prichard et al., Adapting Western North American Forests to Climate Change and Wildfires: 10 Common Questions, 31 Ecological Applications, no. 8, Dec. 2021, at 10 (“On most sites, thinning alone achieves a reduction of canopy fuels but contributes to higher surface fuel loads [that] . . . can contribute to high-intensity surface fires.” (emphasis added)). The paper in fact supports the Agency’s position, concluding that “although the efficacy of thinning alone as a fuel reduction treatment is questionable and site dependent, there exists widespread agreement that combined effects of thinning plus prescribed burning consistently reduces the potential for severe wildfire across a broad range of forest types and conditions.” Id. In any event, to the extent a scientific controversy exists, that controversy concerns the approach taken by the Fire Management Plan, and our inquiry must be focused on the ways in which the Projects deviate from the Plan. That is, to show that the Projects are “highly controversial” such that extraordinary circumstances bar the Agency from relying on the minor-change exclusion, the Institute must show that the new aspects of the Projects will have controversial effects. Holding otherwise would mean that an agency’s minor change or amendment to a NEPA-compliant, yet controversial, plan would always require its own environmental impact statement or environmental assessment, even if everyone agreed that the change or amendment itself would have no environmental impact whatsoever. Here, the Institute has mounted a challenge only to the Projects, not to the Agency’s continued reliance on the 2004 Fire Management Plan. Although the Institute may disagree with the approach taken in the Plan, the Agency was aware of, and thoughtfully rejected, the objection the Institute raises. During the notice and comment phase of the development of the Fire Management Plan, some commenters flagged precisely the concern that the Institute

EARTH ISLAND INSTITUTE V. MULDOON 29

now asserts: that thinning may increase the risk of severe fires or otherwise harm the environment. One commenter, for example, argued that thinning “could increase the fire risk” because “reducing the canopy can result in drying out of the land.” The Agency defended its decision over that objection, explaining that “[a]s a result [of thinning], fire will play a more natural role in these ecosystems while mitigating the risk of high intensity wildland fires and preventing an accumulation of unnaturally heavy fuels.” In response to another comment criticizing the Agency’s conclusion that thinning trees would aid its fire-management goals, the Agency stated that the thinning strategy was “based on the best scientific information” because it would return the park to conditions closer to those that “existed when wildland fires played a more benign and natural role.” The Institute does not point to any ways in which the relevant science has materially changed since 2004. The Institute cannot use this challenge to the Projects as a backdoor means to relitigate a decision that the Agency previously made, after notice and comment and a detailed environmental impact statement. Finally, the Institute argues that our precedent compels us to conclude that thinning of the sort at issue here is highly controversial. The Institute points to Bark v. United States Forest Service, in which we held that the Forest Service’s decision not to prepare an environmental impact statement for a logging project was arbitrary and capricious because the project was highly controversial. See 958 F.3d at 870– 71. The project at issue there would have used a technique called “variable density thinning,” which gives the agency flexibility in choosing which trees to cut, setting only a target range for the percentage of canopy cover rather than limiting thinning to trees of a certain size. Id. at 868. After preparing an environmental assessment, the agency concluded that the project would have no significant environmental effects. Id. at 869. We held that the agency’s conclusion was arbitrary and capricious because the agency failed to address arguments raised during the administrative process that challenged whether variable density thinning effectively reduces the risk of wildfires. Id. at 870–71. Bark is distinguishable from this case for two reasons. First, the Projects here cannot be characterized as involving variable density thinning. Whereas variable density thinning does not impose limits on which trees may be cut, the Projects do, limiting the Agency to trees below 20 inches in diameter. Second, whereas the project in Bark contemplated thinning without subsequent prescribed burns, the thinning here (as we have emphasized) will be conducted in preparation for prescribed burns. Indeed, the fact that the Bark thinning would not be followed by a prescribed burn was one of the reasons why we concluded that the scientific evidence demonstrated that a significant controversy existed. See id. at 871 (describing a study that concluded that “fuel treatments are unlikely to reduce fire severity and consequent impacts, because often the treated area is not affected by fire before the fuels return to normal levels”). Regardless, to the extent that Bark is relevant at all, it speaks only to the controversy surrounding the Fire Management Plan itself—not the aspects of the Projects that deviate from the Plan. The Institute’s strongest objection is that the Agency did not explain why it concluded that the Projects were not “highly controversial.” We acknowledge that it would have been preferable for the Agency to provide an explanation. Still, considering the record presently before us, the Institute has not offered substantial evidence that the Projects are

EARTH ISLAND INSTITUTE V. MULDOON 31

highly controversial in the relevant sense. See Safari Club, 31 F.4th at 1179 (holding that a project was not “highly controversial” because the plaintiffs had not shown “that the disputed parts of the [agency action] have highly controversial, uncertain, or unique environmental effects,” without suggesting that the agency offered any contemporaneous documentation explaining why it did not consider the projects highly controversial). “Employing the deferential [arbitrary and capricious] standard of review,” we conclude that the Institute has not shown that the Agency failed to “consider[] the relevant factors” when it “determined that no extraordinary circumstances were present.” Alaska Ctr., 189 F.3d at 859. D. The Institute’s remaining merits argument is that the Agency failed to take a “hard look” at the environmental impacts of the Projects, as NEPA requires it to do. But the Agency’s reasonable invocation of the minor-change exclusion fully satisfies its obligations under NEPA. See Salazar, 706 F.3d at 1096 (“Application of a categorical exclusion . . . is a form of NEPA compliance.”). Nothing more is required.

IV.

The Institute has not shown that it is likely to succeed on the merits—or even raised serious questions going to the merits. Because the Institute has failed to meet that “threshold inquiry,” we need not consider the other preliminary injunction factors. California v. Azar, 911 F.3d 558, 575 (9th Cir. 2018) (quoting Disney Enters., Inc. v. VidAngel, Inc., 869 F.3d 848, 856 (9th Cir. 2017)). For the foregoing reasons, we AFFIRM.