WILDEARTH GUARDIANS, Plaintiff - Appellant, v. TAMARA CONNER, in her official capacity as District Ranger, Leadville Ranger District, San Isabel National Forest, United States Forest Service; UNITED STATES FOREST SERVICE, a federal agency of the United States Department of Agriculture, Defendants - Appellees.
No. 17-1334
United States Court of Appeals for the Tenth Circuit
April 15, 2019
Before HARTZ, SEYMOUR, and EID, Circuit Judges.
PUBLISH; Appeal from the United States District Court for the District of Colorado (D.C. No. 1:15-CV-00858-CMA)
John R. Mellgren, (Peter M.K. Frost, with him on the briefs), Western Environmental Law Center, Eugene, Oregon, for Plaintiff-Appellant.
Sommer H. Engels (Jeffrey H. Wood, Acting Assistant Attorney General, Eric Grant, Deputy Assistant Attorney General, Andrew C. Mergen, Allen M. Brabender, and Barclay T. Samford, Attorneys, United States Department of Justice, Environment & Natural Resources Division, Washington, D.C., and Tyler Clarkson, Deputy General Counsel, and Kenneth Capps, Office of General Counsel, United States Department of Agriculture, with her on the brief), United States Department of Justice, Environment & Natural Resources Division, Washington, D.C for Defendants-Appellees.
The dispute before us concerns the Tennessee Creek Project (the Project), an effort of the United States Forest Service (the Service) in the San Isabel and White River National Forests to protect the forest from insects, disease, and fire; improve wildlife habitat; and maintain watershed conditions. In 2014 the Service published an environmental assessment (EA) of the Project, followed by a Decision Notice (DN) and Finding of No Significant Impact
WildEarth Guardians sought review in the United States District Court for the District of Colorado, arguing that the Service had violated the National Environmental Policy Act (NEPA),
I. BACKGROUND
A. NEPA framework
We have called NEPA the “centerpiece of environmental regulation in the United States.” New Mexico ex rel. Richardson v. Bureau of Land Mgmt., 565 F.3d 683, 703 (10th Cir. 2009). The statute’s “twin aims” are to ensure that agencies consider the environmental effects of their actions and inform the public of having done so. Baltimore Gas & Elec. Co. v. Nat. Res. Def. Council, 462 U.S. 87, 97 (1983). It does not compel “agencies to elevate environmental concerns over other appropriate considerations.” Id. Instead, NEPA’s mandate is that agencies “pause before committing resources to a project and consider the likely environmental impacts of the preferred course of action as well as reasonable alternatives.” Richardson, 565 F.3d at 703; see also Marsh v. Oregon Nat. Res.’s Council, 490 U.S. 360, 371 (1989). It “merely prohibits uninformed—rather than unwise—agency action.” Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 351 (1989).
If an agency is considering an action that might affect the environment, it must follow a process prescribed by NEPA and its implementing regulations. See
Because NEPA provides no private cause of action, see Utah Envtl. Congress v. Russell, 518 F.3d 817, 823 (10th Cir. 2008), challenges to an EA or FONSI must be brought under the Administrative Procedure Act (APA), which instructs us to review whether an agency’s action was “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law,”
B. Factual Background
1. The Canada lynx
The Canada lynx is native to the snowy, high-altitude coniferous forests of Colorado’s Southern Rockies. These mountains provide the conditions necessary for lynx habitat: elevated forests dominated by spruce-fir, lodgepole pine, and aspen-conifer mix, and populated by snowshoe hare for lynx to prey on. The U.S. Fish and Wildlife Service (FWS) designated the lynx as a threatened species in 2000. See 65 Fed. Reg. 16052 (March 24, 2000). A year earlier the Colorado Division of Wildlife began releasing lynx into the wild to augment the very small population. By 2007, it had released 218 lynx.
From 1998 to 2000, biologists from the Service, Bureau of Land Management, National Pаrk Service, and the FWS jointly compiled information on lynx in the contiguous United States, culminating in the 2000 publication of the Canada Lynx Conservation Assessment and Strategy (LCAS). The LCAS instructed agencies to map Lynx Analysis Units (LAUs), which are geographic areas approximating the size of the home range of a female lynx to be used “to begin the analysis of potential direct and indirect effects of projects or activities on individual lynx, and to monitor habitat changes.” Bill Ruediger et al., USDA Forest Service, USDI Fish and Wildlife Service, USDI Bureau of Land Management, and USDI National Park Service, Canada Lynx Conservation Assessment and Strategy, 6-2 (2d ed. 2000). It also recommended various conservation measures to protect lynx habitat on federal lands. For example, if an agency intended to take management action within an LAU, the LCAS instructed it to map lynx denning and foraging habitat and to ensure that at least 10% of the lynx habitat in the LAU would remain denning habitat. See LCAS at 7-4. And because the lynx population in the Southern Rockies is limited by thе availability of snowshoe hare (the primary prey for lynx), the LCAS also recommended various measures to maintain the horizontal cover (e.g., shrubs, understory trees, and low limbs) necessary for snowshoe-hare habitat. See LCAS at pp. 7-4–7-6. A month after the LCAS was published, regional managers of the Service and the FWS in the Southern Rockies signed the Lynx Conservation Agreement, committing themselves to consider the LCAS’s recommendations before undertaking new actions in lynx habitat. The agreement was revised and extended in 2005, and amended in 2006.
In 2008 the Service adopted the Southern Rockies Lynx Amendment (SRLA). This document superseded the Lynx Conservation Agreement, and it amended the Land and Resource Management Plans of eight National Forests in the Southern Rockies. Its purpose was to strike “a reasonable balance in providing for the conservation of lynx habitat while also allowing appropriate levels of humаn uses to occur.” Aplt. App. at 227. The SRLA imposes seven standards on agencies, such as a standard stating that “[t]imber management projects shall not regenerate1 more than 15 percent of lynx habitat on [National Forest System] lands within an LAU in a ten-year period.” Aplee. App. at GA7 (footnotes omitted). The SRLA also includes nonmandatory guidelines that recommend “actions that will normally be taken
The standards and guidelines of the SRLA were adopted only after completion of an EIS. The draft EIS for the SRLA (issued in 2004) and the supplemental draft EIS (issued in 2006) received nearly 300 comments. The final EIS explored five different alternatives, at least one of which—Alternative B—would have included greater protection for lynx denning habitat than Alternative F, which the Forest Service ultimately adopted. For one thing, Alternative B would have included a standard similar to the one in the LCAS that 10% of lynx habitat in each LAU must be maintained as dеnning habitat. The Service explained in its EIS that such a standard was probably unnecessary because most LAUs already have between 20% and 40% denning habitat, in which case the availability of denning habitat would not be a limiting factor for lynx.
Indeed, research after the study that led to the 10% standard in the LCAS concluded that lynx use “a greater variety of habitat for denning” than previously thought. Aplt. App. at 232. The research showed that lynx den sites “are found in both mature and younger forests that have a large amount of cover and downed, large woody debris. . . . [L]ynx have used all kinds of deadfall for den sites, so it is likely almost any forest does supply denning habitat. . . . The research does not indicate a certain minimum amount of denning habitat is required for lynx.” Aplee. Fed. R. App. P. 28(j) Letter of 11/17/2018, attachment at T01678.
Yet the SRLA still protected denning habitat. Guideline VEG G11 states that “[i]f denning habitat appears to be lacking in the LAU, then projects should be designed to retain some coarse woody debris, piles, or residual trees to provide denning habitat in the future.” Aplee. App. at GA9 (footnotes omitted). The SRLA also advises agencies to protect certain types of vegetation that are beneficial to lynx and their prey, and to ensure that lynx denning habitat is near snowshoe-hare winter habitat:
Guideline VEG G1: Vegetation management projects should be planned to recruit a high density of conifers, hardwoods, and shrubs where such habitat is scarce or not available. Priority for treatment should be given to stem-exclusion, closed-canopy structural stage stands to enhance habitat conditions for lynx or their prey (e.g. mesic, monotypic lodgepole stands). Winter snowshoe hare habitat should be near denning habitat.
Aplee. App. at GA9 (footnotes omitted).
The SRLA’s standards and guidelines also aim to protect the winter habitat of snowshoe hare, which are the primary source of food for lynx. The SRLA describes snowshoe-hare winter habitat as “places where young trees or shrubs grow densely—thousands of woody stems per acre—and tall enough to protrude above the snow during winter, so snowshoe hare can browse on the bark and small twigs.” Aplt. App. at 239. Several studies have identified spruce-fir stands as providing the highest snowshoe-hare densities of forest types in the region, but snowshoe hare can also populate aspen and lodgepole-pine stands, so long as the stands provide enough horizontal cover, which the SRLA defines as “visual obscurity provided by vegetation that extends to the ground or snow surface.” Aplt. App. at 655.
Further, the SRLA Implementation Guide provides clarification and guidance on how agencies should remap LAUs. An LAU should approximate the size of the home range of a female lynx in the Southern Rockies (one study reported that this averaged about 18,500 acres). It also must contain at least 6,400 acres of “primary vegetation,” Aplee. App. at GA45, which
2. The Tennessee Creek Project
a. Scope
The Project is a response to a mountain-pine-beetle epidemic that impacted forest stands on the White River and San Isabel National Forests, and an associated threat to headwaters that serve communities along Colorado’s Front Range. It will be implemented over 10 to 15 years “to create forest conditions that are more resilient to outbreaks of insects, disease and wildfire; to improve habitat for threatened, endangered and sensitive species and other important wildlife species; and to provide for sustainable watershed conditions.” Aplt. App. at 506. Planned action involves a mix of clearcutting, thinning, and prescribed burns.
The dominant forest type in the 16,450-acre Project area is lodgepole pine (11,096 acres), although there are also significant spruce-fir stands (2,177 acres) and aspen stands (564 acres). The lodgepole pines are currently vulnerable to beetle infestations and the spread of dwarf mistletoe. In 2014, 40% of the lodgepole-pine stands were already infected by dwarf mistletoe. The spruce-fir and aspen stands currently have a low incidence of insects and disease, but they will become more susceptible as they age.
b. Lynx habitat in the project area
Each of the three main forest types in the Project area—lodgepole-pine, spruce-fir, and aspen—can provide lynx habitat (although lodgepole-pine stands are not considered habitat once they develop into mature monocultures). About 9,480 acres of the Project area are mapped lynx habitat falling within one of two LAUs—the Tennessee Pass LAU and the Massive LAU. Most of the Project area also falls within the 67,500-acre Tennessee Pass Linkage Area, an area connecting blocks of lynx habitat. During a 2013 two-month research project, three lynx were trapped and released within the spruce-fir and lodgepole forests of the LAUs that overlap with the Project area.
The Project will include conservation measures to protect this lynx habitat and to follow the standards and guidelines of the SRLA:
- The Service will not treat spruce-fir stands (which often provide lynx denning habitat and high snowshoe-hare densities) in mapped lynx habitat, except that dead trees will be removed from areas affected by beetles or disease (always leaving 10% for denning habitat).
- The Service will exclude from treatment any tree stands with greater than 35% dense horizontal cover. Such stands provide lynx denning habitat and the highest-quality winter habitat for snowshoe hare.
- Most clearcutting will take place in climax lodgepole stands and in mature lodgepole monocultures, which are either low-quality habitat or uninhabitable for lynx and snowshoe hare.
- Any adjacent clear-cuts will be kept at least 200 feet apart to allow travel corridors for wildlife.
- Thinning will be “in a mosaic fashion that would mimic natural disturbances,” Aplt. App. at 587.
- The Service will track the acreage of lynx habitat treated by the Project and report that data to the FWS as required by the SRLA.
- In the event that mechanical trampling or other harvest or salvage activity damages lynx or snowshoe-hare
winter habitat, the Service will track that damage and count it toward any limits imposed by the SRLA.
c. The Project’s EA
In 2013 the Service issued a draft EA, as well as a draft biological assessment (BA) that primarily analyzed the Project’s effects on lynx. See
Although the EA quantified the amount of each type of treatment, it did not specify the treatment locations. Rather, the Service intends to identify 300 to 500 acres for thinning and clearcutting each year over the next 10 to 15 years. It has asserted that this flexible approach is necessary for reacting to on-the-ground conditions, such as a beetle infestation or fire risk.
The EA includes nine pages analyzing the proposed action’s possible effects on lynx, as well as an appendix assessing its adherence to each SRLA objective, standard, and guideline. Because the Service does not yet know precisely which of the 9,480 acres of mapped lynx habitat will be treated, it took the conservative approach of assuming that all lynx habitat in the Project area will be treated.
In that worst-case scenario, the Service found, clearcutting would temporarily convert 6% of lynx habitat in each LAU to nonhabitat—well below the SRLA’s 15% limit on how much lynx habitat within an LAU may be regenerated in a 10-year period. Further, the clear-cut area would become habitat in 15 to 30 years, when trees would have grown enough to be available to snowshoe hare above the snow level, although denning habitat that is clear-cut would take more than 150 years to regenerate fully. The Service also found that thinning treatments could cause lynx habitat to be temporarily degraded but not rendered unsuitable, and within 30 years those thinning treatments will have created more snowshoe-hare winter habitat by
After reviewing objections to the EA and the draft DN/FONSI, Leadville District Ranger Tamara Conner issued a final (slightly revised) DN/FONSI in November 2014. It announced that an EIS was not necessary and that the Service would proceed with the EA’s proposed alternative.
3. Procedural History
WildEarth brought this action in the United States District Court for Colorado against the Service and Tаmara Conner (in her official capacity as the Leadville District Ranger) for allegedly violating NEPA. WildEarth claimed that the Service’s EA and FONSI were inadequate because they did not “disclose, analyze, and otherwise take a hard look” at the Project’s environmental effects, particularly its effects on lynx. Aplt. App. at 051. The district court granted judgment in favor of the agency.
II. DISCUSSION
On appeal WildEarth raises two claims: (1) the Service violated NEPA because its EA did not adequately assess the Project’s effects on lynx; and (2) the Service violated NEPA by not producing an EIS. We address each in turn.
A. Standard of Review
WildEarth brings its challenges to the Service’s actions under the APA. Our task is therefore to determine whether those actions were “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.”
short of a clear error of judgment
In accordance with this standard, we assess the adequacy of the Service’s action by asking whether its method of analyzing environmental effects “had a rational basis and took into consideration the relevant factors.” Utah Shared Access Alliance, 288 F.3d at 1212–13. The analysis need not have used the best possible methodology, nor do we ask whether it could have discussed environmental impacts in more detail. See id. at 1212–13. Instead, we review whether the agency’s decision was reasoned, and we defer to the agency’s expertise and discretion. See id. at 1213; see also Utah Envtl. Cong. v. Russell, 518 F.3d 817, 824 (10th Cir. 2008) (“Deference to the agency is especially strong where the challenged decisions involve technical or scientific matters within the agency’s area of expertise.” (internal quotation marks omitted)).
B. Environmental Assessment
WildEarth argues that the EA was inadequate because it did not sufficiently evaluаte the Project’s effects on lynx. We disagree.
Before even beginning work on the EA, the Service had devoted a great deal of attention to the lynx, assessing what type of habitat was needed and what actions would imperil them. The LCAS, published in 2000, presented a number of recommendations. It was incorporated in the Lynx Conservation Agreement of regional managers of the Service and the FWS in the Southern Rockies, and the agreement was updated in 2005 and 2006. Then, after considerable input from the public, the Service in 2008 promulgated the SRLA, which superseded the Agreement and amended Land and Resource Management Plans for eight national forests in the southern Rockies. Based on an EIS, it contained a number of standards and guidelines. With this scientific knowledge in hand, the Service could reasonably assess the maximum impact that the Project could have on the lynx and conclude it was unlikely to adversеly affect them. In particular, it could conclude that the Project will not violate any SRLA standards even in a worst-case scenario in which every acre of mapped lynx habitat in the Project area is treated.
WildEarth nonetheless argues that the Service needed to collect and disclose several additional types of data. We address each of its arguments.
1. Need to specify treatment locations
First, WildEarth argues that the Service was obligated to specify the sizes, locations, and treatment planned for each of the treatment units and the locations of the 21 miles of temporary road expected to be built. According to WildEarth, our decision in Richardson holds that an environmental assessment must include such “site-specific” detail about a project area so that a proper analysis can be performed. Aplt. Reply Br. at 9. Sometimes. But that depends on the circumstances.
In Richardson the federal Bureau оf Land Management (BLM) produced a draft EIS proposing a land-use plan that would allow drilling on New Mexico’s Otero Mesa. See 565 F.3d at 688, 690. The plan limited drilling to within 492 feet of existing roadways to protect desert grassland from habitat fragmentation. See id. at 690. Three years later, the BLM issued a final EIS adopting a modified plan that did not limit drilling to areas close to existing roadways. See id. at 692. Instead, this alternative opened most of the Mesa to drilling, so long as only 5% of the surface area of the Mesa was subject to drilling at any one time. See id. Despite this major change to the plan, the BLM barely updated
But Richardson did not hold that an agency’s EA or EIS always must specify the precise locations within a project area that will be affected. The problem in Richardson was simply that there had been no environmental assessment of the ultimate plan. The earlier assessment contemplated a significantly different project from what was later selected. That is hardly the case here. The EA analyzed what cоuld happen whatever sites were eventually chosen for treatment by the Project, so long as the Project restrictions were satisfied. The Service’s analysis accounted for the uncertainty about treatment locations by evaluating the Project’s effects on lynx in a worst-case scenario in which all the mapped lynx habitat in the Project area is treated, and by including conservation measures to protect high-quality lynx habitat, such as not treating healthy spruce-fir stands or any stands with greater than 35% dense horizontal cover. Moreover, the Service had a valid reason for not identifying specific treatment sites in its EA: it intends to select treatment units based on changing on-the-ground conditions over the 10 to 15 years of the Project. NEPA leaves “substantial discretion to an agency to determine how best to gather and assess information” about a project’s environmental impacts. Biodiversity Conservation Alliance v. U.S. Forest Serv., 765 F.3d 1264, 1270 (10th Cir. 2014). The Sеrvice used that discretion reasonably, assessing the Project’s maximum possible effect on lynx habitat while also conserving agency resources and retaining flexibility to respond to changing conditions. See Utah Shared Access Alliance, 288 F.3d at 1213 (“By conducting an EA, an agency considers environmental concerns yet reserves its resources for instances where a full EIS is appropriate.” (internal quotation marks omitted)). We note that the Service was not postponing the requisite environmental analysis until it picks the specific sites for treatment under the Project; rather, it was saying that such future analysis would be unnecessary because, in its expert opinion, whatever sites it ultimately chooses (within the constraints imposed by the Project), there would not be a negative impact on the lynx.3
2. Disclosure of precommercial thinning
Repeating the same error just discussed, WildEarth specifically argues that the Service violated NEPA by not disclosing the locations of its preidentifiеd precommercial thinning units. But such disclosure
Relatedly, WildEarth contends in its reply brief that the Service failed to adhere to the SRLA’s VEG S5 standard governing precommercial thinning, because it did not specify the amount of precommercial thinning that would affect snowshoe-hare winter habitat. But this argument is waived because it was not raised in the opening brief. See Silverton Snowmobile Club v. U.S. Forest Serv., 433 F.3d 772, 783 (10th Cir. 2006) (“The failure to raise an issue in an opening brief waives that issue.” (brackets and internal quotation marks omitted)).
3. Need to identify denning habitat
WildEarth also argues that the Service could not truly understand the Project’s impact on lynx without knowing how much affected habitat would be denning habitat. It ignores, however, the expеrt opinion relied on by the Service that denning habitat is not a constraint on the lynx in the Project area. In particular, a study conducted shortly before issuance of the EIS prepared for the SRLA noted how adept lynx are in creating dens: “[L]ynx have used all kinds of deadfall for den sites, so it is likely almost any forest does supply denning habitat. . . . The research does not indicate a certain minimum amount of denning habitat is required for lynx.” Aplee. Fed. R. App. P. 28(j) Letter of 11/17/2018, attachment at T01678. Lynx denning habitat is therefore not expected to be a “limiting factor” for lynx in the Southern Rockies. Aplt. App. at 232 (SRLA). Moreover, to further ensure that lynx in the Project area will not suffer from any shortage of denning habitat, the Project will avoid treating healthy spruce-fir stands or any tree stands with greater than 35% dense horizontal cover. In light of those conservation measures and the low likelihood of LAUs having insufficient denning habitat to begin with, the Service did not need to quantify denning habitat to conclude that the Project will not adversely affect lynx. See Utah Shared Access Alliance, 288 F.3d at 1212–13 (the Service must use a methodology with a “rational basis” but does not need to use the best possible methodology or create the most detailed EA possible).
4. Need to quantify winter habitat
WildEarth contends that the Service should have quantified the amount of winter lynx habitat that will be affected. But such habitat analysis in the EA was not necessary, because the Service reasonably found that the Project will preserve existing high-quality winter habitat, target stands that provide poor or no winter habitat, and even generate new winter habitat in those treated areas. As discussed above, lynx winter habitat is closely correlated with snowshoe-hare winter habitat, which exists in areas with horizontal cover above the snowline. Such conditions tend to exist in healthy spruce-fir stands as well as in other forest stands with greater than 35% dense horizontal covеr—all of which the Service will exclude from treatment. The Service will instead target stands considered nonhabitat or poor habitat for lynx and snowshoe hare. Clearcutting will generally occur in climax lodgepole stands and in mature lodgepole monocultures, which provide only “marginal to poor horizontal cover for snowshoe hare.” Aplt. App. at 588. The Service therefore anticipates that this clearcutting will “not have an effect on snowshoe hare” other than improving winter foraging conditions within 15 to 30 years by allowing new horizontal cover to develop. Aplt. App. at 588. Likewise, thinning
WildEarth argues that the Service’s approach to winter habitat is nonetheless inadequate because it treats snowshoe-hare winter habitat as a proxy for lynx winter habitat instead of accounting for the distinct needs of lynx in winter. To show that snowshoe-hare winter habitat is not a proxy for lynx winter habitat, WildEarth points to a single sentence in the third edition of the LCAS that states, “Winter habitat may be more limiting for lynx.” Aplt. App. at 374. But that sentence is not conveying that lynx winter habitat may be more limiting than snowshoe-hare winter habitat; it is conveying that lynx winter habitat may be more limiting than lynx habitat in other seasons. Indeed, the article cited by that sentence specifically ties lynx winter habitat to snowshoe-hare winter habitat, explaining that lynx in winter are particularly dependent on the availability of snowshoe hare. WildEarth has failed to point to any source contradicting the statement in the EA that “[l]ynx winter habitat is not exactly the same as snowshоe hare winter habitat though the two are very closely associated.” Aplt. App. 494. We are not persuaded that it was unreasonable for the EA to treat snowshoe-hare availability as the key factor for lynx winter habitat.
5. Need for baseline data
Next, WildEarth makes two arguments that the EA was inadequate for failing to include “baseline data” regarding lynx denning and winter habitat in the Project area. Aplt. Br. at 25. First, relying on two Ninth Circuit cases, it contends that agencies have a general duty to ascertain baseline data during the NEPA process to determine the likely impacts of their actions. See N. Plains Res. Council, Inc. v. Surface Transp. Bd., 668 F.3d 1067 (9th Cir. 2011) and Nat’l Parks & Conservation Ass’n v. Babbitt, 241 F.3d 722 (9th Cir. 2001), abrogated on other grounds by Monsanto Co. v. Geertson Seed Farms, 561 U.S. 139, 157 (2010). Those cases, however, are readily distinguishable. In Northern Plains Resource Council there was no dispute that the EIS prepared for the Surface Transportation Board failed to include studies necessary to determine the impact of the planned railroad on, among other things, the pallid sturgeon (an endangered fish) and the sage grouse; rather, what was planned was to study the effects as part of mitigation mеasures during execution of the project. See N. Plains Res. Council, Inc., 668 F.3d at 1083–85. The circuit court held that these delayed studies were not “sufficient to meet the Board’s NEPA obligations to determine the projected extent of the environmental harm to enumerated resources before a project is approved.” Id. at 1084. It pointed out that “an agency must support its conclusions with studies that the agency deems reliable,” and “[s]uch analyses must occur before the proposed action is approved, not afterward.” Id. at 1083. In the case before us, in contrast, the Service determined, based on data and studies it deemed reliable, that the Project would not have an adverse impact on the lynx. Similarly, in National Parks the circuit court held that the EA was inadequate because it described various environmental effects as “unknown” but proposed only future research and monitoring to determine those effects. 241 F.3d at 732–33. The court said, “That is precisely the information and understanding thаt is required before a decision that may have a significant adverse impact on the environment is made, and precisely
WildEarth’s second argument is that more baseline data about the Project area is necessary to monitor (what WildEarth calls “ground-truth“) the Service’s commitment not to treat areas of mapped lynx habitat with greater than 35% dense horizontal cover. Aplt. Br. at 22. In essence, WildEarth is saying that it does not trust the Service to do what it promises and needs additional information at this time so that it can later investigate whether the Service has lived up to its commitments. WildEarth does not provide any authority supporting its argument, and we are aware of none. We generally presume that government agencies comply with the law and NEPA creates no exception to this presumption. See Pit River Tribe v. U.S. Forest Service, 615 F.3d 1069, 1082 (9th Cir. 2010) (“[W]e presume that agencies will follow the law.“); cf. Poe v. Gerstein, 417 U.S. 281, 281–82 (1974) (it was appropriate for district court to issue a declaratory judgment but not an injunction against the State because court would not assume that State would decline to acquiesce in the decision). We decline to require additional steps to facilitate monitoring of whether the Service complies with what it proposes to do.
In sum, the record shows that the Service made a reasoned evaluation of how the Project will affect lynx. WildEarth contends that the Service needed to state in the EA precisely where the Project would do what and then evaluate the specific effects of those actions on the lynx. But the nature of the Project, which requires responding to conditions on the ground as they develop over the course of 10 to 15 years, makes such precision impraсticable. And the Service’s long study of the lynx and the requirements for its habitat enabled it to reasonably conclude that even in the worst-case scenario, the Project would not adversely affect that animal.
C. The FONSI
WildEarth’s second claim is that the Service erred by issuing a FONSI instead of conducting an EIS. An agency may issue a FONSI only if, after reviewing the direct and indirect effects of a proposed action, it concludes that the action “will not have a significant effect on the human environment.”
First, WildEarth contends that the sheer size of the Project—over 2,000 acres of clearcutting and 7,000 acres of thinning—bears on two significance factors. See
TOMAC offers no support for the proposition that an EIS is required when a project reaches a certain size. The relevant benchmark is whether the federal action “significantly affect[s] the quality of the human environment.”
42 U.S.C. § 4332(2)(C) . Large federal projects may, on the average, be more likely to meet this threshold. But there is no categorical rule that sizable federal undertakings always have a significant effect on the quality of the human environment.
Context is an important consideration. See
WildEarth next argues that the Project’s direct and cumulative impacts on lynx and lynx habitat will be significant because the Project will destroy some denning habitat for 150 years, degrade other winter and denning habitat, degrade linkage area, and render some lynx habitat unsuitable for up to 25 years. See
WildEarth also argues that the Project’s effects on lynx are “highly controversial” and “highly uncertain“—two other significance factors under § 1508.27(b)(4)–(5). Even in the absence of substantial public opposition, an action may be “highly controversial” if there is “a substantial dispute as to the size, nature, or effect of the action.” Middle Rio Grande Conservancy Dist. v. Norton, 294 F.3d 1220, 1229 (10th Cir. 2002). According to WildEarth, the Project’s effects are controversial and uncertain because the Service has not specified the size or location of the Project’s treatment units or their effects on lynx habitat, and the Service could resolve this controversy and reduce uncertainty simply by agreeing to conduct an EIS addressing these issues. But given that the Service reasonably concluded that the Project was unlikely to harm lynx regardless of treatment locations, it could properly conclude that there was no legitimate controversy.
The final significance factor that WildEarth directs us to is the “degree to which the action may adversely affect an endangered or threatened species or its habitat,”
III. CONCLUSION
We AFFIRM the district court’s order rejecting WildEarth’s objections to the Project.